From 49920bc66984a512f4bcc7735a61642cd0e4d6f2 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Thu, 13 Oct 2011 15:15:27 +0530 Subject: [PATCH 001/820] dmaengine: add new enum dma_transfer_direction This new enum removes usage of dma_data_direction for dma direction. The new enum cleans tells the DMA direction and mode This further paves way for merging the dmaengine _prep operations and also for interleaved dma Suggested-by: Jassi Brar Reviewed-by: Barry Song Signed-off-by: Vinod Koul --- include/linux/dmaengine.h | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-) diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index ace51af4369f..d946ef7f5e67 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -23,7 +23,6 @@ #include #include -#include #include /** @@ -75,6 +74,19 @@ enum dma_transaction_type { /* last transaction type for creation of the capabilities mask */ #define DMA_TX_TYPE_END (DMA_CYCLIC + 1) +/** + * enum dma_transfer_direction - dma transfer mode and direction indicator + * @DMA_MEM_TO_MEM: Async/Memcpy mode + * @DMA_MEM_TO_DEV: Slave mode & From Memory to Device + * @DMA_DEV_TO_MEM: Slave mode & From Device to Memory + * @DMA_DEV_TO_DEV: Slave mode & From Device to Device + */ +enum dma_transfer_direction { + DMA_MEM_TO_MEM, + DMA_MEM_TO_DEV, + DMA_DEV_TO_MEM, + DMA_DEV_TO_DEV, +}; /** * enum dma_ctrl_flags - DMA flags to augment operation preparation, @@ -267,7 +279,7 @@ enum dma_slave_buswidth { * struct, if applicable. */ struct dma_slave_config { - enum dma_data_direction direction; + enum dma_transfer_direction direction; dma_addr_t src_addr; dma_addr_t dst_addr; enum dma_slave_buswidth src_addr_width; @@ -490,11 +502,11 @@ struct dma_device { struct dma_async_tx_descriptor *(*device_prep_slave_sg)( struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags); struct dma_async_tx_descriptor *(*device_prep_dma_cyclic)( struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, - size_t period_len, enum dma_data_direction direction); + size_t period_len, enum dma_transfer_direction direction); int (*device_control)(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned long arg); @@ -520,7 +532,7 @@ static inline int dmaengine_slave_config(struct dma_chan *chan, static inline struct dma_async_tx_descriptor *dmaengine_prep_slave_single( struct dma_chan *chan, void *buf, size_t len, - enum dma_data_direction dir, unsigned long flags) + enum dma_transfer_direction dir, unsigned long flags) { struct scatterlist sg; sg_init_one(&sg, buf, len); From db8196df4bb6f117caa163aa73b0f16fd62290bd Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Thu, 13 Oct 2011 22:34:23 +0530 Subject: [PATCH 002/820] dmaengine: move drivers to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves dma/drivers/* to use new enum Cc: Jassi Brar Cc: Russell King Cc: Viresh Kumar Cc: Linus Walleij Cc: Nicolas Ferre Cc: Mika Westerberg Cc: H Hartley Sweeten Cc: Li Yang Cc: Zhang Wei Cc: Sascha Hauer Cc: Guennadi Liakhovetski Cc: Shawn Guo Cc: Yong Wang Cc: Tomoya MORINAGA Cc: Boojin Kim Cc: Barry Song Acked-by: Mika Westerberg Acked-by: Linus Walleij Acked-by: Viresh Kumar Acked-by: Nicolas Ferre Signed-off-by: Vinod Koul --- arch/arm/mach-ep93xx/include/mach/dma.h | 6 ++--- .../arm/plat-nomadik/include/plat/ste_dma40.h | 4 +-- drivers/dma/amba-pl08x.c | 24 ++++++++--------- drivers/dma/at_hdmac.c | 22 ++++++++-------- drivers/dma/coh901318.c | 12 ++++----- drivers/dma/coh901318_lli.c | 23 ++++++++-------- drivers/dma/coh901318_lli.h | 4 +-- drivers/dma/dw_dmac.c | 14 +++++----- drivers/dma/ep93xx_dma.c | 22 ++++++++-------- drivers/dma/fsldma.c | 4 +-- drivers/dma/imx-dma.c | 10 +++---- drivers/dma/imx-sdma.c | 10 +++---- drivers/dma/intel_mid_dma.c | 14 +++++----- drivers/dma/intel_mid_dma_regs.h | 2 +- drivers/dma/ipu/ipu_idmac.c | 4 +-- drivers/dma/mxs-dma.c | 8 +++--- drivers/dma/pch_dma.c | 12 ++++----- drivers/dma/pl330.c | 18 ++++++------- drivers/dma/shdma.c | 25 +++++++++--------- drivers/dma/ste_dma40.c | 26 +++++++++---------- drivers/dma/timb_dma.c | 18 ++++++------- drivers/dma/txx9dmac.c | 12 ++++----- include/linux/amba/pl08x.h | 4 +-- include/linux/dw_dmac.h | 2 +- include/linux/sh_dma.h | 2 +- 25 files changed, 150 insertions(+), 152 deletions(-) diff --git a/arch/arm/mach-ep93xx/include/mach/dma.h b/arch/arm/mach-ep93xx/include/mach/dma.h index 46d4d876e6fb..e82c642fa53c 100644 --- a/arch/arm/mach-ep93xx/include/mach/dma.h +++ b/arch/arm/mach-ep93xx/include/mach/dma.h @@ -37,7 +37,7 @@ */ struct ep93xx_dma_data { int port; - enum dma_data_direction direction; + enum dma_transfer_direction direction; const char *name; }; @@ -80,14 +80,14 @@ static inline bool ep93xx_dma_chan_is_m2p(struct dma_chan *chan) * channel supports given DMA direction. Only M2P channels have such * limitation, for M2M channels the direction is configurable. */ -static inline enum dma_data_direction +static inline enum dma_transfer_direction ep93xx_dma_chan_direction(struct dma_chan *chan) { if (!ep93xx_dma_chan_is_m2p(chan)) return DMA_NONE; /* even channels are for TX, odd for RX */ - return (chan->chan_id % 2 == 0) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + return (chan->chan_id % 2 == 0) ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM; } #endif /* __ASM_ARCH_DMA_H */ diff --git a/arch/arm/plat-nomadik/include/plat/ste_dma40.h b/arch/arm/plat-nomadik/include/plat/ste_dma40.h index 685c78716d95..38b041a40db4 100644 --- a/arch/arm/plat-nomadik/include/plat/ste_dma40.h +++ b/arch/arm/plat-nomadik/include/plat/ste_dma40.h @@ -187,7 +187,7 @@ static inline struct dma_async_tx_descriptor *stedma40_slave_mem(struct dma_chan *chan, dma_addr_t addr, unsigned int size, - enum dma_data_direction direction, + enum dma_transfer_direction direction, unsigned long flags) { struct scatterlist sg; @@ -209,7 +209,7 @@ static inline struct dma_async_tx_descriptor *stedma40_slave_mem(struct dma_chan *chan, dma_addr_t addr, unsigned int size, - enum dma_data_direction direction, + enum dma_transfer_direction direction, unsigned long flags) { return NULL; diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c index b7cbd1ab1db1..41c62fd0680d 100644 --- a/drivers/dma/amba-pl08x.c +++ b/drivers/dma/amba-pl08x.c @@ -882,9 +882,9 @@ static int prep_phy_channel(struct pl08x_dma_chan *plchan, ch->signal = ret; /* Assign the flow control signal to this channel */ - if (txd->direction == DMA_TO_DEVICE) + if (txd->direction == DMA_MEM_TO_DEV) txd->ccfg |= ch->signal << PL080_CONFIG_DST_SEL_SHIFT; - else if (txd->direction == DMA_FROM_DEVICE) + else if (txd->direction == DMA_DEV_TO_MEM) txd->ccfg |= ch->signal << PL080_CONFIG_SRC_SEL_SHIFT; } @@ -1102,10 +1102,10 @@ static int dma_set_runtime_config(struct dma_chan *chan, /* Transfer direction */ plchan->runtime_direction = config->direction; - if (config->direction == DMA_TO_DEVICE) { + if (config->direction == DMA_MEM_TO_DEV) { addr_width = config->dst_addr_width; maxburst = config->dst_maxburst; - } else if (config->direction == DMA_FROM_DEVICE) { + } else if (config->direction == DMA_DEV_TO_MEM) { addr_width = config->src_addr_width; maxburst = config->src_maxburst; } else { @@ -1136,7 +1136,7 @@ static int dma_set_runtime_config(struct dma_chan *chan, cctl |= burst << PL080_CONTROL_SB_SIZE_SHIFT; cctl |= burst << PL080_CONTROL_DB_SIZE_SHIFT; - if (plchan->runtime_direction == DMA_FROM_DEVICE) { + if (plchan->runtime_direction == DMA_DEV_TO_MEM) { plchan->src_addr = config->src_addr; plchan->src_cctl = pl08x_cctl(cctl) | PL080_CONTROL_DST_INCR | pl08x_select_bus(plchan->cd->periph_buses, @@ -1152,7 +1152,7 @@ static int dma_set_runtime_config(struct dma_chan *chan, "configured channel %s (%s) for %s, data width %d, " "maxburst %d words, LE, CCTL=0x%08x\n", dma_chan_name(chan), plchan->name, - (config->direction == DMA_FROM_DEVICE) ? "RX" : "TX", + (config->direction == DMA_DEV_TO_MEM) ? "RX" : "TX", addr_width, maxburst, cctl); @@ -1322,7 +1322,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_dma_memcpy( static struct dma_async_tx_descriptor *pl08x_prep_slave_sg( struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct pl08x_dma_chan *plchan = to_pl08x_chan(chan); @@ -1354,10 +1354,10 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg( */ txd->direction = direction; - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { txd->cctl = plchan->dst_cctl; slave_addr = plchan->dst_addr; - } else if (direction == DMA_FROM_DEVICE) { + } else if (direction == DMA_DEV_TO_MEM) { txd->cctl = plchan->src_cctl; slave_addr = plchan->src_addr; } else { @@ -1368,10 +1368,10 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg( } if (plchan->cd->device_fc) - tmp = (direction == DMA_TO_DEVICE) ? PL080_FLOW_MEM2PER_PER : + tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER_PER : PL080_FLOW_PER2MEM_PER; else - tmp = (direction == DMA_TO_DEVICE) ? PL080_FLOW_MEM2PER : + tmp = (direction == DMA_MEM_TO_DEV) ? PL080_FLOW_MEM2PER : PL080_FLOW_PER2MEM; txd->ccfg |= tmp << PL080_CONFIG_FLOW_CONTROL_SHIFT; @@ -1387,7 +1387,7 @@ static struct dma_async_tx_descriptor *pl08x_prep_slave_sg( list_add_tail(&dsg->node, &txd->dsg_list); dsg->len = sg_dma_len(sg); - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { dsg->src_addr = sg_phys(sg); dsg->dst_addr = slave_addr; } else { diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index fcfa0a8b5c59..7e76574e83ec 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -660,7 +660,7 @@ err_desc_get: */ static struct dma_async_tx_descriptor * atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct at_dma_chan *atchan = to_at_dma_chan(chan); @@ -678,7 +678,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, dev_vdbg(chan2dev(chan), "prep_slave_sg (%d): %s f0x%lx\n", sg_len, - direction == DMA_TO_DEVICE ? "TO DEVICE" : "FROM DEVICE", + direction == DMA_MEM_TO_DEV ? "TO DEVICE" : "FROM DEVICE", flags); if (unlikely(!atslave || !sg_len)) { @@ -692,7 +692,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, ctrlb = ATC_IEN; switch (direction) { - case DMA_TO_DEVICE: + case DMA_MEM_TO_DEV: ctrla |= ATC_DST_WIDTH(reg_width); ctrlb |= ATC_DST_ADDR_MODE_FIXED | ATC_SRC_ADDR_MODE_INCR @@ -725,7 +725,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, total_len += len; } break; - case DMA_FROM_DEVICE: + case DMA_DEV_TO_MEM: ctrla |= ATC_SRC_WIDTH(reg_width); ctrlb |= ATC_DST_ADDR_MODE_INCR | ATC_SRC_ADDR_MODE_FIXED @@ -787,7 +787,7 @@ err_desc_get: */ static int atc_dma_cyclic_check_values(unsigned int reg_width, dma_addr_t buf_addr, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { if (period_len > (ATC_BTSIZE_MAX << reg_width)) goto err_out; @@ -795,7 +795,7 @@ atc_dma_cyclic_check_values(unsigned int reg_width, dma_addr_t buf_addr, goto err_out; if (unlikely(buf_addr & ((1 << reg_width) - 1))) goto err_out; - if (unlikely(!(direction & (DMA_TO_DEVICE | DMA_FROM_DEVICE)))) + if (unlikely(!(direction & (DMA_DEV_TO_MEM | DMA_MEM_TO_DEV)))) goto err_out; return 0; @@ -810,7 +810,7 @@ err_out: static int atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc, unsigned int period_index, dma_addr_t buf_addr, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { u32 ctrla; unsigned int reg_width = atslave->reg_width; @@ -822,7 +822,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc, | period_len >> reg_width; switch (direction) { - case DMA_TO_DEVICE: + case DMA_MEM_TO_DEV: desc->lli.saddr = buf_addr + (period_len * period_index); desc->lli.daddr = atslave->tx_reg; desc->lli.ctrla = ctrla; @@ -833,7 +833,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc, | ATC_DIF(AT_DMA_PER_IF); break; - case DMA_FROM_DEVICE: + case DMA_DEV_TO_MEM: desc->lli.saddr = atslave->rx_reg; desc->lli.daddr = buf_addr + (period_len * period_index); desc->lli.ctrla = ctrla; @@ -861,7 +861,7 @@ atc_dma_cyclic_fill_desc(struct at_dma_slave *atslave, struct at_desc *desc, */ static struct dma_async_tx_descriptor * atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { struct at_dma_chan *atchan = to_at_dma_chan(chan); struct at_dma_slave *atslave = chan->private; @@ -872,7 +872,7 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, unsigned int i; dev_vdbg(chan2dev(chan), "prep_dma_cyclic: %s buf@0x%08x - %d (%d/%d)\n", - direction == DMA_TO_DEVICE ? "TO DEVICE" : "FROM DEVICE", + direction == DMA_MEM_TO_DEV ? "TO DEVICE" : "FROM DEVICE", buf_addr, periods, buf_len, period_len); diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c index 4234f416ef11..d65a718c0f9b 100644 --- a/drivers/dma/coh901318.c +++ b/drivers/dma/coh901318.c @@ -39,7 +39,7 @@ struct coh901318_desc { struct scatterlist *sg; unsigned int sg_len; struct coh901318_lli *lli; - enum dma_data_direction dir; + enum dma_transfer_direction dir; unsigned long flags; u32 head_config; u32 head_ctrl; @@ -1034,7 +1034,7 @@ coh901318_prep_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src, static struct dma_async_tx_descriptor * coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct coh901318_chan *cohc = to_coh901318_chan(chan); @@ -1077,7 +1077,7 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, ctrl_last |= cohc->runtime_ctrl; ctrl |= cohc->runtime_ctrl; - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { u32 tx_flags = COH901318_CX_CTRL_PRDD_SOURCE | COH901318_CX_CTRL_SRC_ADDR_INC_ENABLE; @@ -1085,7 +1085,7 @@ coh901318_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, ctrl_chained |= tx_flags; ctrl_last |= tx_flags; ctrl |= tx_flags; - } else if (direction == DMA_FROM_DEVICE) { + } else if (direction == DMA_DEV_TO_MEM) { u32 rx_flags = COH901318_CX_CTRL_PRDD_DEST | COH901318_CX_CTRL_DST_ADDR_INC_ENABLE; @@ -1274,11 +1274,11 @@ static void coh901318_dma_set_runtimeconfig(struct dma_chan *chan, int i = 0; /* We only support mem to per or per to mem transfers */ - if (config->direction == DMA_FROM_DEVICE) { + if (config->direction == DMA_DEV_TO_MEM) { addr = config->src_addr; addr_width = config->src_addr_width; maxburst = config->src_maxburst; - } else if (config->direction == DMA_TO_DEVICE) { + } else if (config->direction == DMA_MEM_TO_DEV) { addr = config->dst_addr; addr_width = config->dst_addr_width; maxburst = config->dst_maxburst; diff --git a/drivers/dma/coh901318_lli.c b/drivers/dma/coh901318_lli.c index 9f7e0e6a7eea..6c0e2d4c6682 100644 --- a/drivers/dma/coh901318_lli.c +++ b/drivers/dma/coh901318_lli.c @@ -7,11 +7,10 @@ * Author: Per Friden */ -#include #include -#include #include #include +#include #include #include "coh901318_lli.h" @@ -177,18 +176,18 @@ coh901318_lli_fill_single(struct coh901318_pool *pool, struct coh901318_lli *lli, dma_addr_t buf, unsigned int size, dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl_eom, - enum dma_data_direction dir) + enum dma_transfer_direction dir) { int s = size; dma_addr_t src; dma_addr_t dst; - if (dir == DMA_TO_DEVICE) { + if (dir == DMA_MEM_TO_DEV) { src = buf; dst = dev_addr; - } else if (dir == DMA_FROM_DEVICE) { + } else if (dir == DMA_DEV_TO_MEM) { src = dev_addr; dst = buf; @@ -215,9 +214,9 @@ coh901318_lli_fill_single(struct coh901318_pool *pool, lli = coh901318_lli_next(lli); - if (dir == DMA_TO_DEVICE) + if (dir == DMA_MEM_TO_DEV) src += block_size; - else if (dir == DMA_FROM_DEVICE) + else if (dir == DMA_DEV_TO_MEM) dst += block_size; } @@ -234,7 +233,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool, struct scatterlist *sgl, unsigned int nents, dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl, u32 ctrl_last, - enum dma_data_direction dir, u32 ctrl_irq_mask) + enum dma_transfer_direction dir, u32 ctrl_irq_mask) { int i; struct scatterlist *sg; @@ -249,9 +248,9 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool, spin_lock(&pool->lock); - if (dir == DMA_TO_DEVICE) + if (dir == DMA_MEM_TO_DEV) dst = dev_addr; - else if (dir == DMA_FROM_DEVICE) + else if (dir == DMA_DEV_TO_MEM) src = dev_addr; else goto err; @@ -269,7 +268,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool, ctrl_sg = ctrl ? ctrl : ctrl_last; - if (dir == DMA_TO_DEVICE) + if (dir == DMA_MEM_TO_DEV) /* increment source address */ src = sg_phys(sg); else @@ -293,7 +292,7 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool, lli->src_addr = src; lli->dst_addr = dst; - if (dir == DMA_FROM_DEVICE) + if (dir == DMA_DEV_TO_MEM) dst += elem_size; else src += elem_size; diff --git a/drivers/dma/coh901318_lli.h b/drivers/dma/coh901318_lli.h index 7a5c80990e9e..abff3714fdda 100644 --- a/drivers/dma/coh901318_lli.h +++ b/drivers/dma/coh901318_lli.h @@ -97,7 +97,7 @@ coh901318_lli_fill_single(struct coh901318_pool *pool, struct coh901318_lli *lli, dma_addr_t buf, unsigned int size, dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl_last, - enum dma_data_direction dir); + enum dma_transfer_direction dir); /** * coh901318_lli_fill_single() - Prepares the lli:s for dma scatter list transfer @@ -119,6 +119,6 @@ coh901318_lli_fill_sg(struct coh901318_pool *pool, struct scatterlist *sg, unsigned int nents, dma_addr_t dev_addr, u32 ctrl_chained, u32 ctrl, u32 ctrl_last, - enum dma_data_direction dir, u32 ctrl_irq_mask); + enum dma_transfer_direction dir, u32 ctrl_irq_mask); #endif /* COH901318_LLI_H */ diff --git a/drivers/dma/dw_dmac.c b/drivers/dma/dw_dmac.c index 9bfd6d360718..decca1c3c83d 100644 --- a/drivers/dma/dw_dmac.c +++ b/drivers/dma/dw_dmac.c @@ -696,7 +696,7 @@ err_desc_get: static struct dma_async_tx_descriptor * dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct dw_dma_chan *dwc = to_dw_dma_chan(chan); @@ -720,7 +720,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, prev = first = NULL; switch (direction) { - case DMA_TO_DEVICE: + case DMA_MEM_TO_DEV: ctllo = (DWC_DEFAULT_CTLLO(chan->private) | DWC_CTLL_DST_WIDTH(reg_width) | DWC_CTLL_DST_FIX @@ -777,7 +777,7 @@ slave_sg_todev_fill_desc: goto slave_sg_todev_fill_desc; } break; - case DMA_FROM_DEVICE: + case DMA_DEV_TO_MEM: ctllo = (DWC_DEFAULT_CTLLO(chan->private) | DWC_CTLL_SRC_WIDTH(reg_width) | DWC_CTLL_DST_INC @@ -1165,7 +1165,7 @@ EXPORT_SYMBOL(dw_dma_cyclic_stop); */ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, size_t period_len, - enum dma_data_direction direction) + enum dma_transfer_direction direction) { struct dw_dma_chan *dwc = to_dw_dma_chan(chan); struct dw_cyclic_desc *cdesc; @@ -1206,7 +1206,7 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan, goto out_err; if (unlikely(buf_addr & ((1 << reg_width) - 1))) goto out_err; - if (unlikely(!(direction & (DMA_TO_DEVICE | DMA_FROM_DEVICE)))) + if (unlikely(!(direction & (DMA_MEM_TO_DEV | DMA_DEV_TO_MEM)))) goto out_err; retval = ERR_PTR(-ENOMEM); @@ -1228,7 +1228,7 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan, goto out_err_desc_get; switch (direction) { - case DMA_TO_DEVICE: + case DMA_MEM_TO_DEV: desc->lli.dar = dws->tx_reg; desc->lli.sar = buf_addr + (period_len * i); desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan->private) @@ -1239,7 +1239,7 @@ struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan, | DWC_CTLL_FC(dws->fc) | DWC_CTLL_INT_EN); break; - case DMA_FROM_DEVICE: + case DMA_DEV_TO_MEM: desc->lli.dar = buf_addr + (period_len * i); desc->lli.sar = dws->rx_reg; desc->lli.ctllo = (DWC_DEFAULT_CTLLO(chan->private) diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c index b47e2b803faf..009851b2aeea 100644 --- a/drivers/dma/ep93xx_dma.c +++ b/drivers/dma/ep93xx_dma.c @@ -330,7 +330,7 @@ static void m2p_fill_desc(struct ep93xx_dma_chan *edmac) struct ep93xx_dma_desc *desc = ep93xx_dma_get_active(edmac); u32 bus_addr; - if (ep93xx_dma_chan_direction(&edmac->chan) == DMA_TO_DEVICE) + if (ep93xx_dma_chan_direction(&edmac->chan) == DMA_MEM_TO_DEV) bus_addr = desc->src_addr; else bus_addr = desc->dst_addr; @@ -443,7 +443,7 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac) control = (5 << M2M_CONTROL_PWSC_SHIFT); control |= M2M_CONTROL_NO_HDSK; - if (data->direction == DMA_TO_DEVICE) { + if (data->direction == DMA_MEM_TO_DEV) { control |= M2M_CONTROL_DAH; control |= M2M_CONTROL_TM_TX; control |= M2M_CONTROL_RSS_SSPTX; @@ -463,7 +463,7 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac) control |= M2M_CONTROL_RSS_IDE; control |= M2M_CONTROL_PW_16; - if (data->direction == DMA_TO_DEVICE) { + if (data->direction == DMA_MEM_TO_DEV) { /* Worst case from the UG */ control = (3 << M2M_CONTROL_PWSC_SHIFT); control |= M2M_CONTROL_DAH; @@ -803,8 +803,8 @@ static int ep93xx_dma_alloc_chan_resources(struct dma_chan *chan) switch (data->port) { case EP93XX_DMA_SSP: case EP93XX_DMA_IDE: - if (data->direction != DMA_TO_DEVICE && - data->direction != DMA_FROM_DEVICE) + if (data->direction != DMA_MEM_TO_DEV && + data->direction != DMA_DEV_TO_MEM) return -EINVAL; break; default: @@ -952,7 +952,7 @@ fail: */ static struct dma_async_tx_descriptor * ep93xx_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction dir, + unsigned int sg_len, enum dma_transfer_direction dir, unsigned long flags) { struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan); @@ -988,7 +988,7 @@ ep93xx_dma_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, goto fail; } - if (dir == DMA_TO_DEVICE) { + if (dir == DMA_MEM_TO_DEV) { desc->src_addr = sg_dma_address(sg); desc->dst_addr = edmac->runtime_addr; } else { @@ -1032,7 +1032,7 @@ fail: static struct dma_async_tx_descriptor * ep93xx_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len, size_t period_len, - enum dma_data_direction dir) + enum dma_transfer_direction dir) { struct ep93xx_dma_chan *edmac = to_ep93xx_dma_chan(chan); struct ep93xx_dma_desc *desc, *first; @@ -1065,7 +1065,7 @@ ep93xx_dma_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr, goto fail; } - if (dir == DMA_TO_DEVICE) { + if (dir == DMA_MEM_TO_DEV) { desc->src_addr = dma_addr + offset; desc->dst_addr = edmac->runtime_addr; } else { @@ -1133,12 +1133,12 @@ static int ep93xx_dma_slave_config(struct ep93xx_dma_chan *edmac, return -EINVAL; switch (config->direction) { - case DMA_FROM_DEVICE: + case DMA_DEV_TO_MEM: width = config->src_addr_width; addr = config->src_addr; break; - case DMA_TO_DEVICE: + case DMA_MEM_TO_DEV: width = config->dst_addr_width; addr = config->dst_addr; break; diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c index 8a781540590c..b98070c33ca9 100644 --- a/drivers/dma/fsldma.c +++ b/drivers/dma/fsldma.c @@ -772,7 +772,7 @@ fail: */ static struct dma_async_tx_descriptor *fsl_dma_prep_slave_sg( struct dma_chan *dchan, struct scatterlist *sgl, unsigned int sg_len, - enum dma_data_direction direction, unsigned long flags) + enum dma_transfer_direction direction, unsigned long flags) { /* * This operation is not supported on the Freescale DMA controller @@ -819,7 +819,7 @@ static int fsl_dma_device_control(struct dma_chan *dchan, return -ENXIO; /* we set the controller burst size depending on direction */ - if (config->direction == DMA_TO_DEVICE) + if (config->direction == DMA_MEM_TO_DEV) size = config->dst_addr_width * config->dst_maxburst; else size = config->src_addr_width * config->src_maxburst; diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c index d746899f36e1..678cd01dc42c 100644 --- a/drivers/dma/imx-dma.c +++ b/drivers/dma/imx-dma.c @@ -106,7 +106,7 @@ static int imxdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, imx_dma_disable(imxdmac->imxdma_channel); return 0; case DMA_SLAVE_CONFIG: - if (dmaengine_cfg->direction == DMA_FROM_DEVICE) { + if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) { imxdmac->per_address = dmaengine_cfg->src_addr; imxdmac->watermark_level = dmaengine_cfg->src_maxburst; imxdmac->word_size = dmaengine_cfg->src_addr_width; @@ -223,7 +223,7 @@ static void imxdma_free_chan_resources(struct dma_chan *chan) static struct dma_async_tx_descriptor *imxdma_prep_slave_sg( struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct imxdma_channel *imxdmac = to_imxdma_chan(chan); @@ -240,7 +240,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_slave_sg( dma_length += sg->length; } - if (direction == DMA_FROM_DEVICE) + if (direction == DMA_DEV_TO_MEM) dmamode = DMA_MODE_READ; else dmamode = DMA_MODE_WRITE; @@ -270,7 +270,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_slave_sg( static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic( struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { struct imxdma_channel *imxdmac = to_imxdma_chan(chan); struct imxdma_engine *imxdma = imxdmac->imxdma; @@ -316,7 +316,7 @@ static struct dma_async_tx_descriptor *imxdma_prep_dma_cyclic( imxdmac->sg_list[periods].page_link = ((unsigned long)imxdmac->sg_list | 0x01) & ~0x02; - if (direction == DMA_FROM_DEVICE) + if (direction == DMA_DEV_TO_MEM) dmamode = DMA_MODE_READ; else dmamode = DMA_MODE_WRITE; diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c index eab1fe71259e..065de5442c93 100644 --- a/drivers/dma/imx-sdma.c +++ b/drivers/dma/imx-sdma.c @@ -246,7 +246,7 @@ struct sdma_engine; struct sdma_channel { struct sdma_engine *sdma; unsigned int channel; - enum dma_data_direction direction; + enum dma_transfer_direction direction; enum sdma_peripheral_type peripheral_type; unsigned int event_id0; unsigned int event_id1; @@ -649,7 +649,7 @@ static int sdma_load_context(struct sdma_channel *sdmac) struct sdma_buffer_descriptor *bd0 = sdma->channel[0].bd; int ret; - if (sdmac->direction == DMA_FROM_DEVICE) { + if (sdmac->direction == DMA_DEV_TO_MEM) { load_address = sdmac->pc_from_device; } else { load_address = sdmac->pc_to_device; @@ -910,7 +910,7 @@ static void sdma_free_chan_resources(struct dma_chan *chan) static struct dma_async_tx_descriptor *sdma_prep_slave_sg( struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct sdma_channel *sdmac = to_sdma_chan(chan); @@ -1007,7 +1007,7 @@ err_out: static struct dma_async_tx_descriptor *sdma_prep_dma_cyclic( struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { struct sdma_channel *sdmac = to_sdma_chan(chan); struct sdma_engine *sdma = sdmac->sdma; @@ -1092,7 +1092,7 @@ static int sdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, sdma_disable_channel(sdmac); return 0; case DMA_SLAVE_CONFIG: - if (dmaengine_cfg->direction == DMA_FROM_DEVICE) { + if (dmaengine_cfg->direction == DMA_DEV_TO_MEM) { sdmac->per_address = dmaengine_cfg->src_addr; sdmac->watermark_level = dmaengine_cfg->src_maxburst; sdmac->word_size = dmaengine_cfg->src_addr_width; diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c index 9e96c43a846a..6deda25fd0a8 100644 --- a/drivers/dma/intel_mid_dma.c +++ b/drivers/dma/intel_mid_dma.c @@ -394,10 +394,10 @@ static int midc_lli_fill_sg(struct intel_mid_dma_chan *midc, midc->dma->block_size); /*Populate SAR and DAR values*/ sg_phy_addr = sg_phys(sg); - if (desc->dirn == DMA_TO_DEVICE) { + if (desc->dirn == DMA_MEM_TO_DEV) { lli_bloc_desc->sar = sg_phy_addr; lli_bloc_desc->dar = mids->dma_slave.dst_addr; - } else if (desc->dirn == DMA_FROM_DEVICE) { + } else if (desc->dirn == DMA_DEV_TO_MEM) { lli_bloc_desc->sar = mids->dma_slave.src_addr; lli_bloc_desc->dar = sg_phy_addr; } @@ -631,13 +631,13 @@ static struct dma_async_tx_descriptor *intel_mid_dma_prep_memcpy( if (midc->dma->pimr_mask) { cfg_hi.cfgx.protctl = 0x0; /*default value*/ cfg_hi.cfgx.fifo_mode = 1; - if (mids->dma_slave.direction == DMA_TO_DEVICE) { + if (mids->dma_slave.direction == DMA_MEM_TO_DEV) { cfg_hi.cfgx.src_per = 0; if (mids->device_instance == 0) cfg_hi.cfgx.dst_per = 3; if (mids->device_instance == 1) cfg_hi.cfgx.dst_per = 1; - } else if (mids->dma_slave.direction == DMA_FROM_DEVICE) { + } else if (mids->dma_slave.direction == DMA_DEV_TO_MEM) { if (mids->device_instance == 0) cfg_hi.cfgx.src_per = 2; if (mids->device_instance == 1) @@ -681,11 +681,11 @@ static struct dma_async_tx_descriptor *intel_mid_dma_prep_memcpy( ctl_lo.ctlx.sinc = 0; ctl_lo.ctlx.dinc = 0; } else { - if (mids->dma_slave.direction == DMA_TO_DEVICE) { + if (mids->dma_slave.direction == DMA_MEM_TO_DEV) { ctl_lo.ctlx.sinc = 0; ctl_lo.ctlx.dinc = 2; ctl_lo.ctlx.tt_fc = 1; - } else if (mids->dma_slave.direction == DMA_FROM_DEVICE) { + } else if (mids->dma_slave.direction == DMA_DEV_TO_MEM) { ctl_lo.ctlx.sinc = 2; ctl_lo.ctlx.dinc = 0; ctl_lo.ctlx.tt_fc = 2; @@ -731,7 +731,7 @@ err_desc_get: */ static struct dma_async_tx_descriptor *intel_mid_dma_prep_slave_sg( struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct intel_mid_dma_chan *midc = NULL; diff --git a/drivers/dma/intel_mid_dma_regs.h b/drivers/dma/intel_mid_dma_regs.h index aea5ee88ce03..c6de919a6401 100644 --- a/drivers/dma/intel_mid_dma_regs.h +++ b/drivers/dma/intel_mid_dma_regs.h @@ -262,7 +262,7 @@ struct intel_mid_dma_desc { unsigned int lli_length; unsigned int current_lli; dma_addr_t next; - enum dma_data_direction dirn; + enum dma_transfer_direction dirn; enum dma_status status; enum dma_slave_buswidth width; /*width of DMA txn*/ enum intel_mid_dma_mode cfg_mode; /*mode configuration*/ diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c index 6815905a772f..0cee3b30cd77 100644 --- a/drivers/dma/ipu/ipu_idmac.c +++ b/drivers/dma/ipu/ipu_idmac.c @@ -1362,7 +1362,7 @@ static void ipu_gc_tasklet(unsigned long arg) /* Allocate and initialise a transfer descriptor. */ static struct dma_async_tx_descriptor *idmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len, - enum dma_data_direction direction, unsigned long tx_flags) + enum dma_transfer_direction direction, unsigned long tx_flags) { struct idmac_channel *ichan = to_idmac_chan(chan); struct idmac_tx_desc *desc = NULL; @@ -1374,7 +1374,7 @@ static struct dma_async_tx_descriptor *idmac_prep_slave_sg(struct dma_chan *chan chan->chan_id != IDMAC_IC_7) return NULL; - if (direction != DMA_FROM_DEVICE && direction != DMA_TO_DEVICE) { + if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV) { dev_err(chan->device->dev, "Invalid DMA direction %d!\n", direction); return NULL; } diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c index b4588bdd98bb..bdf4672b2553 100644 --- a/drivers/dma/mxs-dma.c +++ b/drivers/dma/mxs-dma.c @@ -377,7 +377,7 @@ static void mxs_dma_free_chan_resources(struct dma_chan *chan) static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg( struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long append) { struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan); @@ -450,7 +450,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg( ccw->bits |= CCW_CHAIN; ccw->bits |= CCW_HALT_ON_TERM; ccw->bits |= CCW_TERM_FLUSH; - ccw->bits |= BF_CCW(direction == DMA_FROM_DEVICE ? + ccw->bits |= BF_CCW(direction == DMA_DEV_TO_MEM ? MXS_DMA_CMD_WRITE : MXS_DMA_CMD_READ, COMMAND); @@ -472,7 +472,7 @@ err_out: static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic( struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { struct mxs_dma_chan *mxs_chan = to_mxs_dma_chan(chan); struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma; @@ -515,7 +515,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic( ccw->bits |= CCW_IRQ; ccw->bits |= CCW_HALT_ON_TERM; ccw->bits |= CCW_TERM_FLUSH; - ccw->bits |= BF_CCW(direction == DMA_FROM_DEVICE ? + ccw->bits |= BF_CCW(direction == DMA_DEV_TO_MEM ? MXS_DMA_CMD_WRITE : MXS_DMA_CMD_READ, COMMAND); dma_addr += period_len; diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c index a6d0e3dbed07..9944e8295498 100644 --- a/drivers/dma/pch_dma.c +++ b/drivers/dma/pch_dma.c @@ -99,7 +99,7 @@ struct pch_dma_desc { struct pch_dma_chan { struct dma_chan chan; void __iomem *membase; - enum dma_data_direction dir; + enum dma_transfer_direction dir; struct tasklet_struct tasklet; unsigned long err_status; @@ -224,7 +224,7 @@ static void pdc_set_dir(struct dma_chan *chan) mask_ctl = DMA_MASK_CTL0_MODE & ~(DMA_CTL0_MODE_MASK_BITS << (DMA_CTL0_BITS_PER_CH * chan->chan_id)); val &= mask_mode; - if (pd_chan->dir == DMA_TO_DEVICE) + if (pd_chan->dir == DMA_MEM_TO_DEV) val |= 0x1 << (DMA_CTL0_BITS_PER_CH * chan->chan_id + DMA_CTL0_DIR_SHIFT_BITS); else @@ -242,7 +242,7 @@ static void pdc_set_dir(struct dma_chan *chan) mask_ctl = DMA_MASK_CTL2_MODE & ~(DMA_CTL0_MODE_MASK_BITS << (DMA_CTL0_BITS_PER_CH * ch)); val &= mask_mode; - if (pd_chan->dir == DMA_TO_DEVICE) + if (pd_chan->dir == DMA_MEM_TO_DEV) val |= 0x1 << (DMA_CTL0_BITS_PER_CH * ch + DMA_CTL0_DIR_SHIFT_BITS); else @@ -607,7 +607,7 @@ static void pd_issue_pending(struct dma_chan *chan) static struct dma_async_tx_descriptor *pd_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len, - enum dma_data_direction direction, unsigned long flags) + enum dma_transfer_direction direction, unsigned long flags) { struct pch_dma_chan *pd_chan = to_pd_chan(chan); struct pch_dma_slave *pd_slave = chan->private; @@ -623,9 +623,9 @@ static struct dma_async_tx_descriptor *pd_prep_slave_sg(struct dma_chan *chan, return NULL; } - if (direction == DMA_FROM_DEVICE) + if (direction == DMA_DEV_TO_MEM) reg = pd_slave->rx_reg; - else if (direction == DMA_TO_DEVICE) + else if (direction == DMA_MEM_TO_DEV) reg = pd_slave->tx_reg; else return NULL; diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c index 571041477ab2..1e58eeb030d8 100644 --- a/drivers/dma/pl330.c +++ b/drivers/dma/pl330.c @@ -320,14 +320,14 @@ static int pl330_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned case DMA_SLAVE_CONFIG: slave_config = (struct dma_slave_config *)arg; - if (slave_config->direction == DMA_TO_DEVICE) { + if (slave_config->direction == DMA_MEM_TO_DEV) { if (slave_config->dst_addr) pch->fifo_addr = slave_config->dst_addr; if (slave_config->dst_addr_width) pch->burst_sz = __ffs(slave_config->dst_addr_width); if (slave_config->dst_maxburst) pch->burst_len = slave_config->dst_maxburst; - } else if (slave_config->direction == DMA_FROM_DEVICE) { + } else if (slave_config->direction == DMA_DEV_TO_MEM) { if (slave_config->src_addr) pch->fifo_addr = slave_config->src_addr; if (slave_config->src_addr_width) @@ -597,7 +597,7 @@ static inline int get_burst_len(struct dma_pl330_desc *desc, size_t len) static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic( struct dma_chan *chan, dma_addr_t dma_addr, size_t len, - size_t period_len, enum dma_data_direction direction) + size_t period_len, enum dma_transfer_direction direction) { struct dma_pl330_desc *desc; struct dma_pl330_chan *pch = to_pchan(chan); @@ -612,13 +612,13 @@ static struct dma_async_tx_descriptor *pl330_prep_dma_cyclic( } switch (direction) { - case DMA_TO_DEVICE: + case DMA_MEM_TO_DEV: desc->rqcfg.src_inc = 1; desc->rqcfg.dst_inc = 0; src = dma_addr; dst = pch->fifo_addr; break; - case DMA_FROM_DEVICE: + case DMA_DEV_TO_MEM: desc->rqcfg.src_inc = 0; desc->rqcfg.dst_inc = 1; src = pch->fifo_addr; @@ -687,7 +687,7 @@ pl330_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dst, static struct dma_async_tx_descriptor * pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flg) { struct dma_pl330_desc *first, *desc = NULL; @@ -702,9 +702,9 @@ pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, return NULL; /* Make sure the direction is consistent */ - if ((direction == DMA_TO_DEVICE && + if ((direction == DMA_MEM_TO_DEV && peri->rqtype != MEMTODEV) || - (direction == DMA_FROM_DEVICE && + (direction == DMA_DEV_TO_MEM && peri->rqtype != DEVTOMEM)) { dev_err(pch->dmac->pif.dev, "%s:%d Invalid Direction\n", __func__, __LINE__); @@ -747,7 +747,7 @@ pl330_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, else list_add_tail(&desc->node, &first->node); - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { desc->rqcfg.src_inc = 1; desc->rqcfg.dst_inc = 0; fill_px(&desc->px, diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c index 81809c2b46ab..592304fb41a6 100644 --- a/drivers/dma/shdma.c +++ b/drivers/dma/shdma.c @@ -23,7 +23,6 @@ #include #include #include -#include #include #include #include @@ -479,19 +478,19 @@ static void sh_dmae_free_chan_resources(struct dma_chan *chan) * @sh_chan: DMA channel * @flags: DMA transfer flags * @dest: destination DMA address, incremented when direction equals - * DMA_FROM_DEVICE or DMA_BIDIRECTIONAL + * DMA_DEV_TO_MEM * @src: source DMA address, incremented when direction equals - * DMA_TO_DEVICE or DMA_BIDIRECTIONAL + * DMA_MEM_TO_DEV * @len: DMA transfer length * @first: if NULL, set to the current descriptor and cookie set to -EBUSY * @direction: needed for slave DMA to decide which address to keep constant, - * equals DMA_BIDIRECTIONAL for MEMCPY + * equals DMA_MEM_TO_MEM for MEMCPY * Returns 0 or an error * Locks: called with desc_lock held */ static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan, unsigned long flags, dma_addr_t *dest, dma_addr_t *src, size_t *len, - struct sh_desc **first, enum dma_data_direction direction) + struct sh_desc **first, enum dma_transfer_direction direction) { struct sh_desc *new; size_t copy_size; @@ -531,9 +530,9 @@ static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan, new->direction = direction; *len -= copy_size; - if (direction == DMA_BIDIRECTIONAL || direction == DMA_TO_DEVICE) + if (direction == DMA_MEM_TO_MEM || direction == DMA_MEM_TO_DEV) *src += copy_size; - if (direction == DMA_BIDIRECTIONAL || direction == DMA_FROM_DEVICE) + if (direction == DMA_MEM_TO_MEM || direction == DMA_DEV_TO_MEM) *dest += copy_size; return new; @@ -546,12 +545,12 @@ static struct sh_desc *sh_dmae_add_desc(struct sh_dmae_chan *sh_chan, * converted to scatter-gather to guarantee consistent locking and a correct * list manipulation. For slave DMA direction carries the usual meaning, and, * logically, the SG list is RAM and the addr variable contains slave address, - * e.g., the FIFO I/O register. For MEMCPY direction equals DMA_BIDIRECTIONAL + * e.g., the FIFO I/O register. For MEMCPY direction equals DMA_MEM_TO_MEM * and the SG list contains only one element and points at the source buffer. */ static struct dma_async_tx_descriptor *sh_dmae_prep_sg(struct sh_dmae_chan *sh_chan, struct scatterlist *sgl, unsigned int sg_len, dma_addr_t *addr, - enum dma_data_direction direction, unsigned long flags) + enum dma_transfer_direction direction, unsigned long flags) { struct scatterlist *sg; struct sh_desc *first = NULL, *new = NULL /* compiler... */; @@ -592,7 +591,7 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_sg(struct sh_dmae_chan *sh_c dev_dbg(sh_chan->dev, "Add SG #%d@%p[%d], dma %llx\n", i, sg, len, (unsigned long long)sg_addr); - if (direction == DMA_FROM_DEVICE) + if (direction == DMA_DEV_TO_MEM) new = sh_dmae_add_desc(sh_chan, flags, &sg_addr, addr, &len, &first, direction); @@ -646,13 +645,13 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy( sg_dma_address(&sg) = dma_src; sg_dma_len(&sg) = len; - return sh_dmae_prep_sg(sh_chan, &sg, 1, &dma_dest, DMA_BIDIRECTIONAL, + return sh_dmae_prep_sg(sh_chan, &sg, 1, &dma_dest, DMA_MEM_TO_MEM, flags); } static struct dma_async_tx_descriptor *sh_dmae_prep_slave_sg( struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len, - enum dma_data_direction direction, unsigned long flags) + enum dma_transfer_direction direction, unsigned long flags) { struct sh_dmae_slave *param; struct sh_dmae_chan *sh_chan; @@ -996,7 +995,7 @@ static void dmae_do_tasklet(unsigned long data) spin_lock_irq(&sh_chan->desc_lock); list_for_each_entry(desc, &sh_chan->ld_queue, node) { if (desc->mark == DESC_SUBMITTED && - ((desc->direction == DMA_FROM_DEVICE && + ((desc->direction == DMA_DEV_TO_MEM && (desc->hw.dar + desc->hw.tcr) == dar_buf) || (desc->hw.sar + desc->hw.tcr) == sar_buf)) { dev_dbg(sh_chan->dev, "done #%d@%p dst %u\n", diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c index 467e4dcb20a0..0c6cbacb8321 100644 --- a/drivers/dma/ste_dma40.c +++ b/drivers/dma/ste_dma40.c @@ -216,7 +216,7 @@ struct d40_chan { struct d40_log_lli_full *lcpa; /* Runtime reconfiguration */ dma_addr_t runtime_addr; - enum dma_data_direction runtime_direction; + enum dma_transfer_direction runtime_direction; }; /** @@ -1854,7 +1854,7 @@ err: } static dma_addr_t -d40_get_dev_addr(struct d40_chan *chan, enum dma_data_direction direction) +d40_get_dev_addr(struct d40_chan *chan, enum dma_transfer_direction direction) { struct stedma40_platform_data *plat = chan->base->plat_data; struct stedma40_chan_cfg *cfg = &chan->dma_cfg; @@ -1863,9 +1863,9 @@ d40_get_dev_addr(struct d40_chan *chan, enum dma_data_direction direction) if (chan->runtime_addr) return chan->runtime_addr; - if (direction == DMA_FROM_DEVICE) + if (direction == DMA_DEV_TO_MEM) addr = plat->dev_rx[cfg->src_dev_type]; - else if (direction == DMA_TO_DEVICE) + else if (direction == DMA_MEM_TO_DEV) addr = plat->dev_tx[cfg->dst_dev_type]; return addr; @@ -1874,7 +1874,7 @@ d40_get_dev_addr(struct d40_chan *chan, enum dma_data_direction direction) static struct dma_async_tx_descriptor * d40_prep_sg(struct dma_chan *dchan, struct scatterlist *sg_src, struct scatterlist *sg_dst, unsigned int sg_len, - enum dma_data_direction direction, unsigned long dma_flags) + enum dma_transfer_direction direction, unsigned long dma_flags) { struct d40_chan *chan = container_of(dchan, struct d40_chan, chan); dma_addr_t src_dev_addr = 0; @@ -1901,9 +1901,9 @@ d40_prep_sg(struct dma_chan *dchan, struct scatterlist *sg_src, if (direction != DMA_NONE) { dma_addr_t dev_addr = d40_get_dev_addr(chan, direction); - if (direction == DMA_FROM_DEVICE) + if (direction == DMA_DEV_TO_MEM) src_dev_addr = dev_addr; - else if (direction == DMA_TO_DEVICE) + else if (direction == DMA_MEM_TO_DEV) dst_dev_addr = dev_addr; } @@ -2107,10 +2107,10 @@ d40_prep_memcpy_sg(struct dma_chan *chan, static struct dma_async_tx_descriptor *d40_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len, - enum dma_data_direction direction, + enum dma_transfer_direction direction, unsigned long dma_flags) { - if (direction != DMA_FROM_DEVICE && direction != DMA_TO_DEVICE) + if (direction != DMA_DEV_TO_MEM && direction != DMA_MEM_TO_DEV) return NULL; return d40_prep_sg(chan, sgl, sgl, sg_len, direction, dma_flags); @@ -2119,7 +2119,7 @@ static struct dma_async_tx_descriptor *d40_prep_slave_sg(struct dma_chan *chan, static struct dma_async_tx_descriptor * dma40_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t dma_addr, size_t buf_len, size_t period_len, - enum dma_data_direction direction) + enum dma_transfer_direction direction) { unsigned int periods = buf_len / period_len; struct dma_async_tx_descriptor *txd; @@ -2268,7 +2268,7 @@ static int d40_set_runtime_config(struct dma_chan *chan, dst_addr_width = config->dst_addr_width; dst_maxburst = config->dst_maxburst; - if (config->direction == DMA_FROM_DEVICE) { + if (config->direction == DMA_DEV_TO_MEM) { dma_addr_t dev_addr_rx = d40c->base->plat_data->dev_rx[cfg->src_dev_type]; @@ -2291,7 +2291,7 @@ static int d40_set_runtime_config(struct dma_chan *chan, if (dst_maxburst == 0) dst_maxburst = src_maxburst; - } else if (config->direction == DMA_TO_DEVICE) { + } else if (config->direction == DMA_MEM_TO_DEV) { dma_addr_t dev_addr_tx = d40c->base->plat_data->dev_tx[cfg->dst_dev_type]; @@ -2356,7 +2356,7 @@ static int d40_set_runtime_config(struct dma_chan *chan, "configured channel %s for %s, data width %d/%d, " "maxburst %d/%d elements, LE, no flow control\n", dma_chan_name(chan), - (config->direction == DMA_FROM_DEVICE) ? "RX" : "TX", + (config->direction == DMA_DEV_TO_MEM) ? "RX" : "TX", src_addr_width, dst_addr_width, src_maxburst, dst_maxburst); diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c index a4a398f2ef61..8c880729b094 100644 --- a/drivers/dma/timb_dma.c +++ b/drivers/dma/timb_dma.c @@ -90,7 +90,7 @@ struct timb_dma_chan { struct list_head queue; struct list_head free_list; unsigned int bytes_per_line; - enum dma_data_direction direction; + enum dma_transfer_direction direction; unsigned int descs; /* Descriptors to allocate */ unsigned int desc_elems; /* number of elems per descriptor */ }; @@ -235,7 +235,7 @@ static void __td_start_dma(struct timb_dma_chan *td_chan) "td_chan: %p, chan: %d, membase: %p\n", td_chan, td_chan->chan.chan_id, td_chan->membase); - if (td_chan->direction == DMA_FROM_DEVICE) { + if (td_chan->direction == DMA_DEV_TO_MEM) { /* descriptor address */ iowrite32(0, td_chan->membase + TIMBDMA_OFFS_RX_DHAR); @@ -278,7 +278,7 @@ static void __td_finish(struct timb_dma_chan *td_chan) txd->cookie); /* make sure to stop the transfer */ - if (td_chan->direction == DMA_FROM_DEVICE) + if (td_chan->direction == DMA_DEV_TO_MEM) iowrite32(0, td_chan->membase + TIMBDMA_OFFS_RX_ER); /* Currently no support for stopping DMA transfers else @@ -398,7 +398,7 @@ static struct timb_dma_desc *td_alloc_init_desc(struct timb_dma_chan *td_chan) td_desc->txd.flags = DMA_CTRL_ACK; td_desc->txd.phys = dma_map_single(chan2dmadev(chan), - td_desc->desc_list, td_desc->desc_list_len, DMA_TO_DEVICE); + td_desc->desc_list, td_desc->desc_list_len, DMA_MEM_TO_DEV); err = dma_mapping_error(chan2dmadev(chan), td_desc->txd.phys); if (err) { @@ -419,7 +419,7 @@ static void td_free_desc(struct timb_dma_desc *td_desc) { dev_dbg(chan2dev(td_desc->txd.chan), "Freeing desc: %p\n", td_desc); dma_unmap_single(chan2dmadev(td_desc->txd.chan), td_desc->txd.phys, - td_desc->desc_list_len, DMA_TO_DEVICE); + td_desc->desc_list_len, DMA_MEM_TO_DEV); kfree(td_desc->desc_list); kfree(td_desc); @@ -558,7 +558,7 @@ static void td_issue_pending(struct dma_chan *chan) static struct dma_async_tx_descriptor *td_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, unsigned int sg_len, - enum dma_data_direction direction, unsigned long flags) + enum dma_transfer_direction direction, unsigned long flags) { struct timb_dma_chan *td_chan = container_of(chan, struct timb_dma_chan, chan); @@ -606,7 +606,7 @@ static struct dma_async_tx_descriptor *td_prep_slave_sg(struct dma_chan *chan, } dma_sync_single_for_device(chan2dmadev(chan), td_desc->txd.phys, - td_desc->desc_list_len, DMA_TO_DEVICE); + td_desc->desc_list_len, DMA_MEM_TO_DEV); return &td_desc->txd; } @@ -775,8 +775,8 @@ static int __devinit td_probe(struct platform_device *pdev) td_chan->descs = pchan->descriptors; td_chan->desc_elems = pchan->descriptor_elements; td_chan->bytes_per_line = pchan->bytes_per_line; - td_chan->direction = pchan->rx ? DMA_FROM_DEVICE : - DMA_TO_DEVICE; + td_chan->direction = pchan->rx ? DMA_DEV_TO_MEM : + DMA_MEM_TO_DEV; td_chan->membase = td->membase + (i / 2) * TIMBDMA_INSTANCE_OFFSET + diff --git a/drivers/dma/txx9dmac.c b/drivers/dma/txx9dmac.c index cbd83e362b5e..6122c364cf11 100644 --- a/drivers/dma/txx9dmac.c +++ b/drivers/dma/txx9dmac.c @@ -845,7 +845,7 @@ txx9dmac_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src, static struct dma_async_tx_descriptor * txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, - unsigned int sg_len, enum dma_data_direction direction, + unsigned int sg_len, enum dma_transfer_direction direction, unsigned long flags) { struct txx9dmac_chan *dc = to_txx9dmac_chan(chan); @@ -860,9 +860,9 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, BUG_ON(!ds || !ds->reg_width); if (ds->tx_reg) - BUG_ON(direction != DMA_TO_DEVICE); + BUG_ON(direction != DMA_MEM_TO_DEV); else - BUG_ON(direction != DMA_FROM_DEVICE); + BUG_ON(direction != DMA_DEV_TO_MEM); if (unlikely(!sg_len)) return NULL; @@ -882,7 +882,7 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, mem = sg_dma_address(sg); if (__is_dmac64(ddev)) { - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { desc->hwdesc.SAR = mem; desc->hwdesc.DAR = ds->tx_reg; } else { @@ -891,7 +891,7 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, } desc->hwdesc.CNTR = sg_dma_len(sg); } else { - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { desc->hwdesc32.SAR = mem; desc->hwdesc32.DAR = ds->tx_reg; } else { @@ -900,7 +900,7 @@ txx9dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, } desc->hwdesc32.CNTR = sg_dma_len(sg); } - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { sai = ds->reg_width; dai = 0; } else { diff --git a/include/linux/amba/pl08x.h b/include/linux/amba/pl08x.h index 9eabffbc4e50..033f6aa670de 100644 --- a/include/linux/amba/pl08x.h +++ b/include/linux/amba/pl08x.h @@ -134,7 +134,7 @@ struct pl08x_txd { struct dma_async_tx_descriptor tx; struct list_head node; struct list_head dsg_list; - enum dma_data_direction direction; + enum dma_transfer_direction direction; dma_addr_t llis_bus; struct pl08x_lli *llis_va; /* Default cctl value for LLIs */ @@ -197,7 +197,7 @@ struct pl08x_dma_chan { dma_addr_t dst_addr; u32 src_cctl; u32 dst_cctl; - enum dma_data_direction runtime_direction; + enum dma_transfer_direction runtime_direction; dma_cookie_t lc; struct list_head pend_list; struct pl08x_txd *at; diff --git a/include/linux/dw_dmac.h b/include/linux/dw_dmac.h index 4bfe0a2f7d50..f2c64f92c4a0 100644 --- a/include/linux/dw_dmac.h +++ b/include/linux/dw_dmac.h @@ -127,7 +127,7 @@ struct dw_cyclic_desc { struct dw_cyclic_desc *dw_dma_cyclic_prep(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, size_t period_len, - enum dma_data_direction direction); + enum dma_transfer_direction direction); void dw_dma_cyclic_free(struct dma_chan *chan); int dw_dma_cyclic_start(struct dma_chan *chan); void dw_dma_cyclic_stop(struct dma_chan *chan); diff --git a/include/linux/sh_dma.h b/include/linux/sh_dma.h index cb2dd118cc0f..62ef6938da10 100644 --- a/include/linux/sh_dma.h +++ b/include/linux/sh_dma.h @@ -30,7 +30,7 @@ struct sh_desc { struct sh_dmae_regs hw; struct list_head node; struct dma_async_tx_descriptor async_tx; - enum dma_data_direction direction; + enum dma_transfer_direction direction; dma_cookie_t cookie; size_t partial; int chunks; From 20dd5a356efcb5bf574baa9adccc775158f13ae1 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:35:16 +0530 Subject: [PATCH 003/820] plat-samsung: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves samsung drivers to use new enum Cc: Ben Dooks Cc: Kukjin Kim Cc: Boojin Kim Acked-by: Kukjin Kim Signed-off-by: Vinod Koul --- arch/arm/plat-samsung/dma-ops.c | 4 ++-- arch/arm/plat-samsung/include/plat/dma-ops.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm/plat-samsung/dma-ops.c b/arch/arm/plat-samsung/dma-ops.c index 6e3d9abc9e2e..3df0bad036c2 100644 --- a/arch/arm/plat-samsung/dma-ops.c +++ b/arch/arm/plat-samsung/dma-ops.c @@ -35,14 +35,14 @@ static unsigned samsung_dmadev_request(enum dma_ch dma_ch, chan = dma_request_channel(mask, pl330_filter, (void *)dma_ch); - if (info->direction == DMA_FROM_DEVICE) { + if (info->direction == DMA_DEV_TO_MEM) { memset(&slave_config, 0, sizeof(struct dma_slave_config)); slave_config.direction = info->direction; slave_config.src_addr = info->fifo; slave_config.src_addr_width = info->width; slave_config.src_maxburst = 1; dmaengine_slave_config(chan, &slave_config); - } else if (info->direction == DMA_TO_DEVICE) { + } else if (info->direction == DMA_MEM_TO_DEV) { memset(&slave_config, 0, sizeof(struct dma_slave_config)); slave_config.direction = info->direction; slave_config.dst_addr = info->fifo; diff --git a/arch/arm/plat-samsung/include/plat/dma-ops.h b/arch/arm/plat-samsung/include/plat/dma-ops.h index 4c1a363526cf..12561152fb97 100644 --- a/arch/arm/plat-samsung/include/plat/dma-ops.h +++ b/arch/arm/plat-samsung/include/plat/dma-ops.h @@ -17,7 +17,7 @@ struct samsung_dma_prep_info { enum dma_transaction_type cap; - enum dma_data_direction direction; + enum dma_transfer_direction direction; dma_addr_t buf; unsigned long period; unsigned long len; @@ -27,7 +27,7 @@ struct samsung_dma_prep_info { struct samsung_dma_info { enum dma_transaction_type cap; - enum dma_data_direction direction; + enum dma_transfer_direction direction; enum dma_slave_buswidth width; dma_addr_t fifo; struct s3c2410_dma_client *client; From 1003cab8276cd34d9deab8ca9d148ee59f7728a7 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:39:23 +0530 Subject: [PATCH 004/820] media, video: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves media and video drivers to use new enum Cc: Mauro Carvalho Chehab Cc: Paul Mundt Cc: Guennadi Liakhovetski Acked-by: Guennadi Liakhovetski Acked-by: Guennadi Liakhovetski Signed-off-by: Vinod Koul --- drivers/media/video/mx3_camera.c | 2 +- drivers/media/video/timblogiw.c | 2 +- drivers/video/mx3fb.c | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/media/video/mx3_camera.c b/drivers/media/video/mx3_camera.c index c045b47803ad..ec98eae20cd8 100644 --- a/drivers/media/video/mx3_camera.c +++ b/drivers/media/video/mx3_camera.c @@ -251,7 +251,7 @@ static int mx3_videobuf_prepare(struct vb2_buffer *vb) sg_dma_len(sg) = new_size; buf->txd = ichan->dma_chan.device->device_prep_slave_sg( - &ichan->dma_chan, sg, 1, DMA_FROM_DEVICE, + &ichan->dma_chan, sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT); if (!buf->txd) return -EIO; diff --git a/drivers/media/video/timblogiw.c b/drivers/media/video/timblogiw.c index 84cd1b65b765..d542c8b9e2f4 100644 --- a/drivers/media/video/timblogiw.c +++ b/drivers/media/video/timblogiw.c @@ -564,7 +564,7 @@ static void buffer_queue(struct videobuf_queue *vq, struct videobuf_buffer *vb) spin_unlock_irq(&fh->queue_lock); desc = fh->chan->device->device_prep_slave_sg(fh->chan, - buf->sg, sg_elems, DMA_FROM_DEVICE, + buf->sg, sg_elems, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP); if (!desc) { spin_lock_irq(&fh->queue_lock); diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c index 7e3a490e8d76..e2caf884fe31 100644 --- a/drivers/video/mx3fb.c +++ b/drivers/video/mx3fb.c @@ -334,7 +334,7 @@ static void sdc_enable_channel(struct mx3fb_info *mx3_fbi) /* This enables the channel */ if (mx3_fbi->cookie < 0) { mx3_fbi->txd = dma_chan->device->device_prep_slave_sg(dma_chan, - &mx3_fbi->sg[0], 1, DMA_TO_DEVICE, DMA_PREP_INTERRUPT); + &mx3_fbi->sg[0], 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT); if (!mx3_fbi->txd) { dev_err(mx3fb->dev, "Cannot allocate descriptor on %d\n", dma_chan->chan_id); @@ -1104,7 +1104,7 @@ static int mx3fb_pan_display(struct fb_var_screeninfo *var, async_tx_ack(mx3_fbi->txd); txd = dma_chan->device->device_prep_slave_sg(dma_chan, sg + - mx3_fbi->cur_ipu_buf, 1, DMA_TO_DEVICE, DMA_PREP_INTERRUPT); + mx3_fbi->cur_ipu_buf, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT); if (!txd) { dev_err(fbi->device, "Error preparing a DMA transaction descriptor.\n"); From 1d0c81e876fe04295ef5df387ba057bf9cfae3d8 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:42:56 +0530 Subject: [PATCH 005/820] carma: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves carma drivers to use new enum Signed-off-by: Vinod Koul CC Ira W. Snyder --- drivers/misc/carma/carma-fpga-program.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/misc/carma/carma-fpga-program.c b/drivers/misc/carma/carma-fpga-program.c index 7ce6065dc20e..20ef1eac5599 100644 --- a/drivers/misc/carma/carma-fpga-program.c +++ b/drivers/misc/carma/carma-fpga-program.c @@ -513,7 +513,7 @@ static noinline int fpga_program_dma(struct fpga_dev *priv) * transaction, and then put it under external control */ memset(&config, 0, sizeof(config)); - config.direction = DMA_TO_DEVICE; + config.direction = DMA_MEM_TO_DEV; config.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; config.dst_maxburst = fpga_fifo_size(priv->regs) / 2 / 4; ret = chan->device->device_control(chan, DMA_SLAVE_CONFIG, From 05f5799cbe5c9e2c03f604b3de5783cf4d726227 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:45:11 +0530 Subject: [PATCH 006/820] mmc-host: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves mmc drivers to use new enum Cc: Nicolas Ferre Cc: Chris Ball Cc: Guennadi Liakhovetski Acked-by: Nicolas Ferre Acked-by: Guennadi Liakhovetski Signed-off-by: Vinod Koul --- drivers/mmc/host/atmel-mci.c | 10 +++++++--- drivers/mmc/host/mmci.c | 11 +++++++---- drivers/mmc/host/mxcmmc.c | 10 +++++++--- drivers/mmc/host/mxs-mmc.c | 7 ++++++- drivers/mmc/host/sh_mmcif.c | 4 ++-- drivers/mmc/host/tmio_mmc_dma.c | 4 ++-- 6 files changed, 31 insertions(+), 15 deletions(-) diff --git a/drivers/mmc/host/atmel-mci.c b/drivers/mmc/host/atmel-mci.c index fa8cae1d7005..ac87803ab70e 100644 --- a/drivers/mmc/host/atmel-mci.c +++ b/drivers/mmc/host/atmel-mci.c @@ -653,6 +653,7 @@ atmci_prepare_data_dma(struct atmel_mci *host, struct mmc_data *data) struct scatterlist *sg; unsigned int i; enum dma_data_direction direction; + enum dma_transfer_direction slave_dirn; unsigned int sglen; /* @@ -681,16 +682,19 @@ atmci_prepare_data_dma(struct atmel_mci *host, struct mmc_data *data) if (atmci_is_mci2()) mci_writel(host, DMA, MCI_DMA_CHKSIZE(3) | MCI_DMAEN); - if (data->flags & MMC_DATA_READ) + if (data->flags & MMC_DATA_READ) { direction = DMA_FROM_DEVICE; - else + slave_dirn = DMA_DEV_TO_MEM; + } else { direction = DMA_TO_DEVICE; + slave_dirn = DMA_MEM_TO_DEV; + } sglen = dma_map_sg(chan->device->dev, data->sg, data->sg_len, direction); desc = chan->device->device_prep_slave_sg(chan, - data->sg, sglen, direction, + data->sg, sglen, slave_dirn, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) goto unmap_exit; diff --git a/drivers/mmc/host/mmci.c b/drivers/mmc/host/mmci.c index 56e9a4168264..3abd51a9fec1 100644 --- a/drivers/mmc/host/mmci.c +++ b/drivers/mmc/host/mmci.c @@ -372,6 +372,7 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data, struct dma_chan *chan; struct dma_device *device; struct dma_async_tx_descriptor *desc; + enum dma_data_direction buffer_dirn; int nr_sg; /* Check if next job is already prepared */ @@ -385,10 +386,12 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data, } if (data->flags & MMC_DATA_READ) { - conf.direction = DMA_FROM_DEVICE; + conf.direction = DMA_DEV_TO_MEM; + buffer_dirn = DMA_FROM_DEVICE; chan = host->dma_rx_channel; } else { - conf.direction = DMA_TO_DEVICE; + conf.direction = DMA_MEM_TO_DEV; + buffer_dirn = DMA_TO_DEVICE; chan = host->dma_tx_channel; } @@ -401,7 +404,7 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data, return -EINVAL; device = chan->device; - nr_sg = dma_map_sg(device->dev, data->sg, data->sg_len, conf.direction); + nr_sg = dma_map_sg(device->dev, data->sg, data->sg_len, buffer_dirn); if (nr_sg == 0) return -EINVAL; @@ -424,7 +427,7 @@ static int mmci_dma_prep_data(struct mmci_host *host, struct mmc_data *data, unmap_exit: if (!next) dmaengine_terminate_all(chan); - dma_unmap_sg(device->dev, data->sg, data->sg_len, conf.direction); + dma_unmap_sg(device->dev, data->sg, data->sg_len, buffer_dirn); return -ENOMEM; } diff --git a/drivers/mmc/host/mxcmmc.c b/drivers/mmc/host/mxcmmc.c index 14aa213b00da..8a816639b46c 100644 --- a/drivers/mmc/host/mxcmmc.c +++ b/drivers/mmc/host/mxcmmc.c @@ -217,6 +217,7 @@ static int mxcmci_setup_data(struct mxcmci_host *host, struct mmc_data *data) unsigned int blksz = data->blksz; unsigned int datasize = nob * blksz; struct scatterlist *sg; + enum dma_transfer_direction slave_dirn; int i, nents; if (data->flags & MMC_DATA_STREAM) @@ -239,10 +240,13 @@ static int mxcmci_setup_data(struct mxcmci_host *host, struct mmc_data *data) } } - if (data->flags & MMC_DATA_READ) + if (data->flags & MMC_DATA_READ) { host->dma_dir = DMA_FROM_DEVICE; - else + slave_dirn = DMA_DEV_TO_MEM; + } else { host->dma_dir = DMA_TO_DEVICE; + slave_dirn = DMA_MEM_TO_DEV; + } nents = dma_map_sg(host->dma->device->dev, data->sg, data->sg_len, host->dma_dir); @@ -250,7 +254,7 @@ static int mxcmci_setup_data(struct mxcmci_host *host, struct mmc_data *data) return -EINVAL; host->desc = host->dma->device->device_prep_slave_sg(host->dma, - data->sg, data->sg_len, host->dma_dir, + data->sg, data->sg_len, slave_dirn, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!host->desc) { diff --git a/drivers/mmc/host/mxs-mmc.c b/drivers/mmc/host/mxs-mmc.c index d513d47364d0..996b9a91d6f4 100644 --- a/drivers/mmc/host/mxs-mmc.c +++ b/drivers/mmc/host/mxs-mmc.c @@ -153,6 +153,7 @@ struct mxs_mmc_host { struct dma_chan *dmach; struct mxs_dma_data dma_data; unsigned int dma_dir; + enum dma_transfer_direction slave_dirn; u32 ssp_pio_words[SSP_PIO_NUM]; unsigned int version; @@ -323,7 +324,7 @@ static struct dma_async_tx_descriptor *mxs_mmc_prep_dma( } desc = host->dmach->device->device_prep_slave_sg(host->dmach, - sgl, sg_len, host->dma_dir, append); + sgl, sg_len, host->slave_dirn, append); if (desc) { desc->callback = mxs_mmc_dma_irq_callback; desc->callback_param = host; @@ -432,6 +433,7 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host) int i; unsigned short dma_data_dir, timeout; + enum dma_transfer_direction slave_dirn; unsigned int data_size = 0, log2_blksz; unsigned int blocks = data->blocks; @@ -447,9 +449,11 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host) if (data->flags & MMC_DATA_WRITE) { dma_data_dir = DMA_TO_DEVICE; + slave_dirn = DMA_MEM_TO_DEV; read = 0; } else { dma_data_dir = DMA_FROM_DEVICE; + slave_dirn = DMA_DEV_TO_MEM; read = BM_SSP_CTRL0_READ; } @@ -517,6 +521,7 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host) WARN_ON(host->data != NULL); host->data = data; host->dma_dir = dma_data_dir; + host->slave_dirn = slave_dirn; desc = mxs_mmc_prep_dma(host, 1); if (!desc) goto out; diff --git a/drivers/mmc/host/sh_mmcif.c b/drivers/mmc/host/sh_mmcif.c index 557886bee9ce..0ae8d62dc976 100644 --- a/drivers/mmc/host/sh_mmcif.c +++ b/drivers/mmc/host/sh_mmcif.c @@ -230,7 +230,7 @@ static void sh_mmcif_start_dma_rx(struct sh_mmcif_host *host) if (ret > 0) { host->dma_active = true; desc = chan->device->device_prep_slave_sg(chan, sg, ret, - DMA_FROM_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); + DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); } if (desc) { @@ -278,7 +278,7 @@ static void sh_mmcif_start_dma_tx(struct sh_mmcif_host *host) if (ret > 0) { host->dma_active = true; desc = chan->device->device_prep_slave_sg(chan, sg, ret, - DMA_TO_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); + DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); } if (desc) { diff --git a/drivers/mmc/host/tmio_mmc_dma.c b/drivers/mmc/host/tmio_mmc_dma.c index 86f259cdfcbc..7a6e6cc8f8b8 100644 --- a/drivers/mmc/host/tmio_mmc_dma.c +++ b/drivers/mmc/host/tmio_mmc_dma.c @@ -77,7 +77,7 @@ static void tmio_mmc_start_dma_rx(struct tmio_mmc_host *host) ret = dma_map_sg(chan->device->dev, sg, host->sg_len, DMA_FROM_DEVICE); if (ret > 0) desc = chan->device->device_prep_slave_sg(chan, sg, ret, - DMA_FROM_DEVICE, DMA_CTRL_ACK); + DMA_DEV_TO_MEM, DMA_CTRL_ACK); if (desc) { cookie = dmaengine_submit(desc); @@ -158,7 +158,7 @@ static void tmio_mmc_start_dma_tx(struct tmio_mmc_host *host) ret = dma_map_sg(chan->device->dev, sg, host->sg_len, DMA_TO_DEVICE); if (ret > 0) desc = chan->device->device_prep_slave_sg(chan, sg, ret, - DMA_TO_DEVICE, DMA_CTRL_ACK); + DMA_MEM_TO_DEV, DMA_CTRL_ACK); if (desc) { cookie = dmaengine_submit(desc); From a485df4b4404379786c4bdd258bc528b2617449d Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:47:38 +0530 Subject: [PATCH 007/820] spi, serial: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves spi, serial drivers to use new enum Signed-off-by: Vinod Koul Cc: Grant Likely Cc: Alan Cox Acked-by: Mika Westerberg --- drivers/spi/spi-dw-mid.c | 8 ++++---- drivers/spi/spi-ep93xx.c | 9 ++++++--- drivers/spi/spi-pl022.c | 8 ++++---- drivers/spi/spi-topcliff-pch.c | 4 ++-- drivers/tty/serial/amba-pl011.c | 8 ++++---- drivers/tty/serial/pch_uart.c | 4 ++-- drivers/tty/serial/sh-sci.c | 4 ++-- 7 files changed, 24 insertions(+), 21 deletions(-) diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c index 130e55537db6..0a1fc8a27e3d 100644 --- a/drivers/spi/spi-dw-mid.c +++ b/drivers/spi/spi-dw-mid.c @@ -131,7 +131,7 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change) rxchan = dws->rxchan; /* 2. Prepare the TX dma transfer */ - txconf.direction = DMA_TO_DEVICE; + txconf.direction = DMA_MEM_TO_DEV; txconf.dst_addr = dws->dma_addr; txconf.dst_maxburst = LNW_DMA_MSIZE_16; txconf.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; @@ -147,13 +147,13 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change) txdesc = txchan->device->device_prep_slave_sg(txchan, &dws->tx_sgl, 1, - DMA_TO_DEVICE, + DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_DEST_UNMAP); txdesc->callback = dw_spi_dma_done; txdesc->callback_param = dws; /* 3. Prepare the RX dma transfer */ - rxconf.direction = DMA_FROM_DEVICE; + rxconf.direction = DMA_DEV_TO_MEM; rxconf.src_addr = dws->dma_addr; rxconf.src_maxburst = LNW_DMA_MSIZE_16; rxconf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; @@ -169,7 +169,7 @@ static int mid_spi_dma_transfer(struct dw_spi *dws, int cs_change) rxdesc = rxchan->device->device_prep_slave_sg(rxchan, &dws->rx_sgl, 1, - DMA_FROM_DEVICE, + DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_DEST_UNMAP); rxdesc->callback = dw_spi_dma_done; rxdesc->callback_param = dws; diff --git a/drivers/spi/spi-ep93xx.c b/drivers/spi/spi-ep93xx.c index 1cf645479bfe..0afd87dd9459 100644 --- a/drivers/spi/spi-ep93xx.c +++ b/drivers/spi/spi-ep93xx.c @@ -550,6 +550,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir) struct dma_async_tx_descriptor *txd; enum dma_slave_buswidth buswidth; struct dma_slave_config conf; + enum dma_transfer_direction slave_dirn; struct scatterlist *sg; struct sg_table *sgt; struct dma_chan *chan; @@ -572,6 +573,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir) conf.src_addr = espi->sspdr_phys; conf.src_addr_width = buswidth; + slave_dirn = DMA_DEV_TO_MEM; } else { chan = espi->dma_tx; buf = t->tx_buf; @@ -579,6 +581,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir) conf.dst_addr = espi->sspdr_phys; conf.dst_addr_width = buswidth; + slave_dirn = DMA_MEM_TO_DEV; } ret = dmaengine_slave_config(chan, &conf); @@ -630,7 +633,7 @@ ep93xx_spi_dma_prepare(struct ep93xx_spi *espi, enum dma_data_direction dir) return ERR_PTR(-ENOMEM); txd = chan->device->device_prep_slave_sg(chan, sgt->sgl, nents, - dir, DMA_CTRL_ACK); + slave_dirn, DMA_CTRL_ACK); if (!txd) { dma_unmap_sg(chan->device->dev, sgt->sgl, sgt->nents, dir); return ERR_PTR(-ENOMEM); @@ -978,7 +981,7 @@ static int ep93xx_spi_setup_dma(struct ep93xx_spi *espi) dma_cap_set(DMA_SLAVE, mask); espi->dma_rx_data.port = EP93XX_DMA_SSP; - espi->dma_rx_data.direction = DMA_FROM_DEVICE; + espi->dma_rx_data.direction = DMA_DEV_TO_MEM; espi->dma_rx_data.name = "ep93xx-spi-rx"; espi->dma_rx = dma_request_channel(mask, ep93xx_spi_dma_filter, @@ -989,7 +992,7 @@ static int ep93xx_spi_setup_dma(struct ep93xx_spi *espi) } espi->dma_tx_data.port = EP93XX_DMA_SSP; - espi->dma_tx_data.direction = DMA_TO_DEVICE; + espi->dma_tx_data.direction = DMA_MEM_TO_DEV; espi->dma_tx_data.name = "ep93xx-spi-tx"; espi->dma_tx = dma_request_channel(mask, ep93xx_spi_dma_filter, diff --git a/drivers/spi/spi-pl022.c b/drivers/spi/spi-pl022.c index 730b4a37b823..ddcddf4b1248 100644 --- a/drivers/spi/spi-pl022.c +++ b/drivers/spi/spi-pl022.c @@ -910,11 +910,11 @@ static int configure_dma(struct pl022 *pl022) { struct dma_slave_config rx_conf = { .src_addr = SSP_DR(pl022->phybase), - .direction = DMA_FROM_DEVICE, + .direction = DMA_DEV_TO_MEM, }; struct dma_slave_config tx_conf = { .dst_addr = SSP_DR(pl022->phybase), - .direction = DMA_TO_DEVICE, + .direction = DMA_MEM_TO_DEV, }; unsigned int pages; int ret; @@ -1051,7 +1051,7 @@ static int configure_dma(struct pl022 *pl022) rxdesc = rxchan->device->device_prep_slave_sg(rxchan, pl022->sgt_rx.sgl, rx_sglen, - DMA_FROM_DEVICE, + DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!rxdesc) goto err_rxdesc; @@ -1059,7 +1059,7 @@ static int configure_dma(struct pl022 *pl022) txdesc = txchan->device->device_prep_slave_sg(txchan, pl022->sgt_tx.sgl, tx_sglen, - DMA_TO_DEVICE, + DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!txdesc) goto err_txdesc; diff --git a/drivers/spi/spi-topcliff-pch.c b/drivers/spi/spi-topcliff-pch.c index 1d23f3831866..7eba26e20aeb 100644 --- a/drivers/spi/spi-topcliff-pch.c +++ b/drivers/spi/spi-topcliff-pch.c @@ -1051,7 +1051,7 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw) } sg = dma->sg_rx_p; desc_rx = dma->chan_rx->device->device_prep_slave_sg(dma->chan_rx, sg, - num, DMA_FROM_DEVICE, + num, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc_rx) { dev_err(&data->master->dev, "%s:device_prep_slave_sg Failed\n", @@ -1086,7 +1086,7 @@ static void pch_spi_handle_dma(struct pch_spi_data *data, int *bpw) } sg = dma->sg_tx_p; desc_tx = dma->chan_tx->device->device_prep_slave_sg(dma->chan_tx, - sg, num, DMA_TO_DEVICE, + sg, num, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc_tx) { dev_err(&data->master->dev, "%s:device_prep_slave_sg Failed\n", diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c index f5f6831b0a64..460ff7dbfd63 100644 --- a/drivers/tty/serial/amba-pl011.c +++ b/drivers/tty/serial/amba-pl011.c @@ -268,7 +268,7 @@ static void pl011_dma_probe_initcall(struct uart_amba_port *uap) struct dma_slave_config tx_conf = { .dst_addr = uap->port.mapbase + UART01x_DR, .dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE, - .direction = DMA_TO_DEVICE, + .direction = DMA_MEM_TO_DEV, .dst_maxburst = uap->fifosize >> 1, }; struct dma_chan *chan; @@ -301,7 +301,7 @@ static void pl011_dma_probe_initcall(struct uart_amba_port *uap) struct dma_slave_config rx_conf = { .src_addr = uap->port.mapbase + UART01x_DR, .src_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE, - .direction = DMA_FROM_DEVICE, + .direction = DMA_DEV_TO_MEM, .src_maxburst = uap->fifosize >> 1, }; @@ -480,7 +480,7 @@ static int pl011_dma_tx_refill(struct uart_amba_port *uap) return -EBUSY; } - desc = dma_dev->device_prep_slave_sg(chan, &dmatx->sg, 1, DMA_TO_DEVICE, + desc = dma_dev->device_prep_slave_sg(chan, &dmatx->sg, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) { dma_unmap_sg(dma_dev->dev, &dmatx->sg, 1, DMA_TO_DEVICE); @@ -676,7 +676,7 @@ static int pl011_dma_rx_trigger_dma(struct uart_amba_port *uap) &uap->dmarx.sgbuf_b : &uap->dmarx.sgbuf_a; dma_dev = rxchan->device; desc = rxchan->device->device_prep_slave_sg(rxchan, &sgbuf->sg, 1, - DMA_FROM_DEVICE, + DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); /* * If the DMA engine is busy and cannot prepare a diff --git a/drivers/tty/serial/pch_uart.c b/drivers/tty/serial/pch_uart.c index b46218d679e2..73201a11697a 100644 --- a/drivers/tty/serial/pch_uart.c +++ b/drivers/tty/serial/pch_uart.c @@ -747,7 +747,7 @@ static int dma_handle_rx(struct eg20t_port *priv) sg_dma_address(sg) = priv->rx_buf_dma; desc = priv->chan_rx->device->device_prep_slave_sg(priv->chan_rx, - sg, 1, DMA_FROM_DEVICE, + sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) @@ -906,7 +906,7 @@ static unsigned int dma_handle_tx(struct eg20t_port *priv) } desc = priv->chan_tx->device->device_prep_slave_sg(priv->chan_tx, - priv->sg_tx_p, nent, DMA_TO_DEVICE, + priv->sg_tx_p, nent, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) { dev_err(priv->port.dev, "%s:device_prep_slave_sg Failed\n", diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c index dbd32a1286d3..3683a69b45ee 100644 --- a/drivers/tty/serial/sh-sci.c +++ b/drivers/tty/serial/sh-sci.c @@ -1207,7 +1207,7 @@ static void sci_submit_rx(struct sci_port *s) struct dma_async_tx_descriptor *desc; desc = chan->device->device_prep_slave_sg(chan, - sg, 1, DMA_FROM_DEVICE, DMA_PREP_INTERRUPT); + sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT); if (desc) { s->desc_rx[i] = desc; @@ -1322,7 +1322,7 @@ static void work_fn_tx(struct work_struct *work) BUG_ON(!sg_dma_len(sg)); desc = chan->device->device_prep_slave_sg(chan, - sg, s->sg_len_tx, DMA_TO_DEVICE, + sg, s->sg_len_tx, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) { /* switch to PIO */ From 35e16581ed6bff55009a0bac34c755140407b03f Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:49:30 +0530 Subject: [PATCH 008/820] sound-soc: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves asoc drivers to use new enum Signed-off-by: Vinod Koul Cc: Liam Girdwood Cc: Mark Brown Acked-by: Mark Brown --- sound/atmel/abdac.c | 2 +- sound/atmel/ac97c.c | 10 +++++----- sound/soc/ep93xx/ep93xx-pcm.c | 4 ++-- sound/soc/imx/imx-pcm-dma-mx2.c | 6 +++--- sound/soc/samsung/dma.c | 4 ++-- sound/soc/sh/siu_pcm.c | 4 ++-- sound/soc/txx9/txx9aclc.c | 2 +- 7 files changed, 16 insertions(+), 16 deletions(-) diff --git a/sound/atmel/abdac.c b/sound/atmel/abdac.c index 6fd9391b3a6c..4fa1dbd8ee83 100644 --- a/sound/atmel/abdac.c +++ b/sound/atmel/abdac.c @@ -133,7 +133,7 @@ static int atmel_abdac_prepare_dma(struct atmel_abdac *dac, period_len = frames_to_bytes(runtime, runtime->period_size); cdesc = dw_dma_cyclic_prep(chan, runtime->dma_addr, buffer_len, - period_len, DMA_TO_DEVICE); + period_len, DMA_MEM_TO_DEV); if (IS_ERR(cdesc)) { dev_dbg(&dac->pdev->dev, "could not prepare cyclic DMA\n"); return PTR_ERR(cdesc); diff --git a/sound/atmel/ac97c.c b/sound/atmel/ac97c.c index 6e5addeb236b..cd9428b24a36 100644 --- a/sound/atmel/ac97c.c +++ b/sound/atmel/ac97c.c @@ -102,7 +102,7 @@ static void atmel_ac97c_dma_capture_period_done(void *arg) static int atmel_ac97c_prepare_dma(struct atmel_ac97c *chip, struct snd_pcm_substream *substream, - enum dma_data_direction direction) + enum dma_transfer_direction direction) { struct dma_chan *chan; struct dw_cyclic_desc *cdesc; @@ -118,7 +118,7 @@ static int atmel_ac97c_prepare_dma(struct atmel_ac97c *chip, return -EINVAL; } - if (direction == DMA_TO_DEVICE) + if (direction == DMA_MEM_TO_DEV) chan = chip->dma.tx_chan; else chan = chip->dma.rx_chan; @@ -133,7 +133,7 @@ static int atmel_ac97c_prepare_dma(struct atmel_ac97c *chip, return PTR_ERR(cdesc); } - if (direction == DMA_TO_DEVICE) { + if (direction == DMA_MEM_TO_DEV) { cdesc->period_callback = atmel_ac97c_dma_playback_period_done; set_bit(DMA_TX_READY, &chip->flags); } else { @@ -393,7 +393,7 @@ static int atmel_ac97c_playback_prepare(struct snd_pcm_substream *substream) if (cpu_is_at32ap7000()) { if (!test_bit(DMA_TX_READY, &chip->flags)) retval = atmel_ac97c_prepare_dma(chip, substream, - DMA_TO_DEVICE); + DMA_MEM_TO_DEV); } else { /* Initialize and start the PDC */ writel(runtime->dma_addr, chip->regs + ATMEL_PDC_TPR); @@ -484,7 +484,7 @@ static int atmel_ac97c_capture_prepare(struct snd_pcm_substream *substream) if (cpu_is_at32ap7000()) { if (!test_bit(DMA_RX_READY, &chip->flags)) retval = atmel_ac97c_prepare_dma(chip, substream, - DMA_FROM_DEVICE); + DMA_DEV_TO_MEM); } else { /* Initialize and start the PDC */ writel(runtime->dma_addr, chip->regs + ATMEL_PDC_RPR); diff --git a/sound/soc/ep93xx/ep93xx-pcm.c b/sound/soc/ep93xx/ep93xx-pcm.c index 8dfd3ad84b19..faf7d4f7412b 100644 --- a/sound/soc/ep93xx/ep93xx-pcm.c +++ b/sound/soc/ep93xx/ep93xx-pcm.c @@ -113,9 +113,9 @@ static int ep93xx_pcm_open(struct snd_pcm_substream *substream) rtd->dma_data.name = dma_params->name; if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - rtd->dma_data.direction = DMA_TO_DEVICE; + rtd->dma_data.direction = DMA_MEM_TO_DEV; else - rtd->dma_data.direction = DMA_FROM_DEVICE; + rtd->dma_data.direction = DMA_DEV_TO_MEM; rtd->dma_chan = dma_request_channel(mask, ep93xx_pcm_dma_filter, &rtd->dma_data); diff --git a/sound/soc/imx/imx-pcm-dma-mx2.c b/sound/soc/imx/imx-pcm-dma-mx2.c index 43fdc24f7e8d..7d28de9758f3 100644 --- a/sound/soc/imx/imx-pcm-dma-mx2.c +++ b/sound/soc/imx/imx-pcm-dma-mx2.c @@ -107,12 +107,12 @@ static int imx_ssi_dma_alloc(struct snd_pcm_substream *substream, } if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { - slave_config.direction = DMA_TO_DEVICE; + slave_config.direction = DMA_MEM_TO_DEV; slave_config.dst_addr = dma_params->dma_addr; slave_config.dst_addr_width = buswidth; slave_config.dst_maxburst = dma_params->burstsize; } else { - slave_config.direction = DMA_FROM_DEVICE; + slave_config.direction = DMA_DEV_TO_MEM; slave_config.src_addr = dma_params->dma_addr; slave_config.src_addr_width = buswidth; slave_config.src_maxburst = dma_params->burstsize; @@ -159,7 +159,7 @@ static int snd_imx_pcm_hw_params(struct snd_pcm_substream *substream, iprtd->period_bytes * iprtd->periods, iprtd->period_bytes, substream->stream == SNDRV_PCM_STREAM_PLAYBACK ? - DMA_TO_DEVICE : DMA_FROM_DEVICE); + DMA_MEM_TO_DEV : DMA_DEV_TO_MEM); if (!iprtd->desc) { dev_err(&chan->dev->device, "cannot prepare slave dma\n"); return -EINVAL; diff --git a/sound/soc/samsung/dma.c b/sound/soc/samsung/dma.c index 851346f7d68d..87a39575fb61 100644 --- a/sound/soc/samsung/dma.c +++ b/sound/soc/samsung/dma.c @@ -85,7 +85,7 @@ static void dma_enqueue(struct snd_pcm_substream *substream) dma_info.cap = (samsung_dma_has_circular() ? DMA_CYCLIC : DMA_SLAVE); dma_info.direction = (substream->stream == SNDRV_PCM_STREAM_PLAYBACK - ? DMA_TO_DEVICE : DMA_FROM_DEVICE); + ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM); dma_info.fp = audio_buffdone; dma_info.fp_param = substream; dma_info.period = prtd->dma_period; @@ -170,7 +170,7 @@ static int dma_hw_params(struct snd_pcm_substream *substream, dma_info.client = prtd->params->client; dma_info.direction = (substream->stream == SNDRV_PCM_STREAM_PLAYBACK - ? DMA_TO_DEVICE : DMA_FROM_DEVICE); + ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM); dma_info.width = prtd->params->dma_size; dma_info.fifo = prtd->params->dma_addr; prtd->params->ch = prtd->params->ops->request( diff --git a/sound/soc/sh/siu_pcm.c b/sound/soc/sh/siu_pcm.c index f8f681690a71..0193e595d415 100644 --- a/sound/soc/sh/siu_pcm.c +++ b/sound/soc/sh/siu_pcm.c @@ -131,7 +131,7 @@ static int siu_pcm_wr_set(struct siu_port *port_info, sg_dma_address(&sg) = buff; desc = siu_stream->chan->device->device_prep_slave_sg(siu_stream->chan, - &sg, 1, DMA_TO_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); + &sg, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) { dev_err(dev, "Failed to allocate a dma descriptor\n"); return -ENOMEM; @@ -181,7 +181,7 @@ static int siu_pcm_rd_set(struct siu_port *port_info, sg_dma_address(&sg) = buff; desc = siu_stream->chan->device->device_prep_slave_sg(siu_stream->chan, - &sg, 1, DMA_FROM_DEVICE, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); + &sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) { dev_err(dev, "Failed to allocate dma descriptor\n"); return -ENOMEM; diff --git a/sound/soc/txx9/txx9aclc.c b/sound/soc/txx9/txx9aclc.c index 3de99af8cb82..602bb68b8784 100644 --- a/sound/soc/txx9/txx9aclc.c +++ b/sound/soc/txx9/txx9aclc.c @@ -134,7 +134,7 @@ txx9aclc_dma_submit(struct txx9aclc_dmadata *dmadata, dma_addr_t buf_dma_addr) sg_dma_address(&sg) = buf_dma_addr; desc = chan->device->device_prep_slave_sg(chan, &sg, 1, dmadata->substream->stream == SNDRV_PCM_STREAM_PLAYBACK ? - DMA_TO_DEVICE : DMA_FROM_DEVICE, + DMA_MEM_TO_DEV : DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_CTRL_ACK); if (!desc) { dev_err(&chan->dev->device, "cannot prepare slave dma\n"); From 8af11fd3f013892c918a39acb691bc7469c5ad77 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 10:52:48 +0530 Subject: [PATCH 009/820] net-ks8842: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves net/ks8842 drivers to use new enum Signed-off-by: Vinod Koul Cc: David S. Miller Cc: Samuel Ortiz --- drivers/net/ks8842.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ks8842.c b/drivers/net/ks8842.c index 4a6ae057e3b1..de9f2e205962 100644 --- a/drivers/net/ks8842.c +++ b/drivers/net/ks8842.c @@ -459,7 +459,7 @@ static int ks8842_tx_frame_dma(struct sk_buff *skb, struct net_device *netdev) sg_dma_len(&ctl->sg) += 4 - sg_dma_len(&ctl->sg) % 4; ctl->adesc = ctl->chan->device->device_prep_slave_sg(ctl->chan, - &ctl->sg, 1, DMA_TO_DEVICE, + &ctl->sg, 1, DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP); if (!ctl->adesc) return NETDEV_TX_BUSY; @@ -571,7 +571,7 @@ static int __ks8842_start_new_rx_dma(struct net_device *netdev) sg_dma_len(sg) = DMA_BUFFER_SIZE; ctl->adesc = ctl->chan->device->device_prep_slave_sg(ctl->chan, - sg, 1, DMA_FROM_DEVICE, + sg, 1, DMA_DEV_TO_MEM, DMA_PREP_INTERRUPT | DMA_COMPL_SKIP_SRC_UNMAP); if (!ctl->adesc) From 0b934ccd707ff33a87f15a35a9916d1d8e85d30e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 4 Nov 2011 15:41:27 -0400 Subject: [PATCH 010/820] Xen: Export xen_biovec_phys_mergeable When Xen is enabled, using BIOVEC_PHYS_MERGEABLE in a module causes xen_biovec_phys_mergeable to be referenced, so it needs to be exported. Acked-by: Jens Axboe Signed-off-by: Matthew Wilcox --- drivers/xen/biomerge.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/xen/biomerge.c b/drivers/xen/biomerge.c index ba6eda4b5143..18c1bb6ffce3 100644 --- a/drivers/xen/biomerge.c +++ b/drivers/xen/biomerge.c @@ -11,3 +11,4 @@ bool xen_biovec_phys_mergeable(const struct bio_vec *vec1, return __BIOVEC_PHYS_MERGEABLE(vec1, vec2) && ((mfn1 == mfn2) || ((mfn1+1) == mfn2)); } +EXPORT_SYMBOL(xen_biovec_phys_mergeable); From b60503ba432b16fc84442a84e29a7aad2c0c363d Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Jan 2011 12:50:14 -0500 Subject: [PATCH 011/820] NVMe: New driver This driver is for devices that follow the NVM Express standard Signed-off-by: Matthew Wilcox --- Documentation/ioctl/ioctl-number.txt | 1 + drivers/block/Kconfig | 11 + drivers/block/Makefile | 1 + drivers/block/nvme.c | 1043 ++++++++++++++++++++++++++ include/linux/nvme.h | 343 +++++++++ 5 files changed, 1399 insertions(+) create mode 100644 drivers/block/nvme.c create mode 100644 include/linux/nvme.h diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt index 54078ed96b37..4840334ea97b 100644 --- a/Documentation/ioctl/ioctl-number.txt +++ b/Documentation/ioctl/ioctl-number.txt @@ -149,6 +149,7 @@ Code Seq#(hex) Include File Comments 'M' 01-03 drivers/scsi/megaraid/megaraid_sas.h 'M' 00-0F drivers/video/fsl-diu-fb.h conflict! 'N' 00-1F drivers/usb/scanner.h +'N' 40-7F drivers/block/nvme.c 'O' 00-06 mtd/ubi-user.h UBI 'P' all linux/soundcard.h conflict! 'P' 60-6F sound/sscape_ioctl.h conflict! diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig index 6f07ec1c2f58..35e56e1c948f 100644 --- a/drivers/block/Kconfig +++ b/drivers/block/Kconfig @@ -315,6 +315,17 @@ config BLK_DEV_NBD If unsure, say N. +config BLK_DEV_NVME + tristate "NVM Express block device" + depends on PCI + ---help--- + The NVM Express driver is for solid state drives directly + connected to the PCI or PCI Express bus. If you know you + don't have one of these, it is safe to answer N. + + To compile this driver as a module, choose M here: the + module will be called nvme. + config BLK_DEV_OSD tristate "OSD object-as-blkdev support" depends on SCSI_OSD_ULD diff --git a/drivers/block/Makefile b/drivers/block/Makefile index 76646e9a1c91..349539ad3ad9 100644 --- a/drivers/block/Makefile +++ b/drivers/block/Makefile @@ -23,6 +23,7 @@ obj-$(CONFIG_XILINX_SYSACE) += xsysace.o obj-$(CONFIG_CDROM_PKTCDVD) += pktcdvd.o obj-$(CONFIG_MG_DISK) += mg_disk.o obj-$(CONFIG_SUNVDC) += sunvdc.o +obj-$(CONFIG_BLK_DEV_NVME) += nvme.o obj-$(CONFIG_BLK_DEV_OSD) += osdblk.o obj-$(CONFIG_BLK_DEV_UMEM) += umem.o diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c new file mode 100644 index 000000000000..ef66eccc2aa2 --- /dev/null +++ b/drivers/block/nvme.c @@ -0,0 +1,1043 @@ +/* + * NVM Express device driver + * Copyright (c) 2011, Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define NVME_Q_DEPTH 1024 +#define SQ_SIZE(depth) (depth * sizeof(struct nvme_command)) +#define CQ_SIZE(depth) (depth * sizeof(struct nvme_completion)) +#define NVME_MINORS 64 + +static int nvme_major; +module_param(nvme_major, int, 0); + +/* + * Represents an NVM Express device. Each nvme_dev is a PCI function. + */ +struct nvme_dev { + struct list_head node; + struct nvme_queue **queues; + u32 __iomem *dbs; + struct pci_dev *pci_dev; + int instance; + int queue_count; + u32 ctrl_config; + struct msix_entry *entry; + struct nvme_bar __iomem *bar; + struct list_head namespaces; +}; + +/* + * An NVM Express namespace is equivalent to a SCSI LUN + */ +struct nvme_ns { + struct list_head list; + + struct nvme_dev *dev; + struct request_queue *queue; + struct gendisk *disk; + + int ns_id; + int lba_shift; +}; + +/* + * An NVM Express queue. Each device has at least two (one for admin + * commands and one for I/O commands). + */ +struct nvme_queue { + struct device *q_dmadev; + spinlock_t q_lock; + struct nvme_command *sq_cmds; + volatile struct nvme_completion *cqes; + dma_addr_t sq_dma_addr; + dma_addr_t cq_dma_addr; + wait_queue_head_t sq_full; + struct bio_list sq_cong; + u32 __iomem *q_db; + u16 q_depth; + u16 cq_vector; + u16 sq_head; + u16 sq_tail; + u16 cq_head; + u16 cq_cycle; + unsigned long cmdid_data[]; +}; + +/* + * Check we didin't inadvertently grow the command struct + */ +static inline void _nvme_check_size(void) +{ + BUILD_BUG_ON(sizeof(struct nvme_rw_command) != 64); + BUILD_BUG_ON(sizeof(struct nvme_create_cq) != 64); + BUILD_BUG_ON(sizeof(struct nvme_create_sq) != 64); + BUILD_BUG_ON(sizeof(struct nvme_delete_queue) != 64); + BUILD_BUG_ON(sizeof(struct nvme_features) != 64); + BUILD_BUG_ON(sizeof(struct nvme_command) != 64); + BUILD_BUG_ON(sizeof(struct nvme_id_ctrl) != 4096); + BUILD_BUG_ON(sizeof(struct nvme_id_ns) != 4096); + BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64); +} + +/** + * alloc_cmdid - Allocate a Command ID + * @param nvmeq The queue that will be used for this command + * @param ctx A pointer that will be passed to the handler + * @param handler The ID of the handler to call + * + * Allocate a Command ID for a queue. The data passed in will + * be passed to the completion handler. This is implemented by using + * the bottom two bits of the ctx pointer to store the handler ID. + * Passing in a pointer that's not 4-byte aligned will cause a BUG. + * We can change this if it becomes a problem. + */ +static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler) +{ + int depth = nvmeq->q_depth; + unsigned long data = (unsigned long)ctx | handler; + int cmdid; + + BUG_ON((unsigned long)ctx & 3); + + do { + cmdid = find_first_zero_bit(nvmeq->cmdid_data, depth); + if (cmdid >= depth) + return -EBUSY; + } while (test_and_set_bit(cmdid, nvmeq->cmdid_data)); + + nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(depth)] = data; + return cmdid; +} + +static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, + int handler) +{ + int cmdid; + wait_event_killable(nvmeq->sq_full, + (cmdid = alloc_cmdid(nvmeq, ctx, handler)) >= 0); + return (cmdid < 0) ? -EINTR : cmdid; +} + +/* If you need more than four handlers, you'll need to change how + * alloc_cmdid and nvme_process_cq work + */ +enum { + sync_completion_id = 0, + bio_completion_id, +}; + +static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) +{ + unsigned long data; + + data = nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(nvmeq->q_depth)]; + clear_bit(cmdid, nvmeq->cmdid_data); + wake_up(&nvmeq->sq_full); + return data; +} + +static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) +{ + return ns->dev->queues[1]; +} + +static void put_nvmeq(struct nvme_queue *nvmeq) +{ +} + +/** + * nvme_submit_cmd: Copy a command into a queue and ring the doorbell + * @nvmeq: The queue to use + * @cmd: The command to send + * + * Safe to use from interrupt context + */ +static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) +{ + unsigned long flags; + u16 tail; + /* XXX: Need to check tail isn't going to overrun head */ + spin_lock_irqsave(&nvmeq->q_lock, flags); + tail = nvmeq->sq_tail; + memcpy(&nvmeq->sq_cmds[tail], cmd, sizeof(*cmd)); + writel(tail, nvmeq->q_db); + if (++tail == nvmeq->q_depth) + tail = 0; + nvmeq->sq_tail = tail; + spin_unlock_irqrestore(&nvmeq->q_lock, flags); + + return 0; +} + +struct nvme_req_info { + struct bio *bio; + int nents; + struct scatterlist sg[0]; +}; + +/* XXX: use a mempool */ +static struct nvme_req_info *alloc_info(unsigned nseg, gfp_t gfp) +{ + return kmalloc(sizeof(struct nvme_req_info) + + sizeof(struct scatterlist) * nseg, gfp); +} + +static void free_info(struct nvme_req_info *info) +{ + kfree(info); +} + +static void bio_completion(struct nvme_queue *nvmeq, void *ctx, + struct nvme_completion *cqe) +{ + struct nvme_req_info *info = ctx; + struct bio *bio = info->bio; + u16 status = le16_to_cpup(&cqe->status) >> 1; + + dma_unmap_sg(nvmeq->q_dmadev, info->sg, info->nents, + bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); + free_info(info); + bio_endio(bio, status ? -EIO : 0); +} + +static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, + struct bio *bio, enum dma_data_direction dma_dir, int psegs) +{ + struct bio_vec *bvec; + struct scatterlist *sg = info->sg; + int i, nsegs; + + sg_init_table(sg, psegs); + bio_for_each_segment(bvec, bio, i) { + sg_set_page(sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); + /* XXX: handle non-mergable here */ + nsegs++; + } + info->nents = nsegs; + + return dma_map_sg(dev, info->sg, info->nents, dma_dir); +} + +static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, + struct bio *bio) +{ + struct nvme_rw_command *cmnd; + struct nvme_req_info *info; + enum dma_data_direction dma_dir; + int cmdid; + u16 control; + u32 dsmgmt; + unsigned long flags; + int psegs = bio_phys_segments(ns->queue, bio); + + info = alloc_info(psegs, GFP_NOIO); + if (!info) + goto congestion; + info->bio = bio; + + cmdid = alloc_cmdid(nvmeq, info, bio_completion_id); + if (unlikely(cmdid < 0)) + goto free_info; + + control = 0; + if (bio->bi_rw & REQ_FUA) + control |= NVME_RW_FUA; + if (bio->bi_rw & (REQ_FAILFAST_DEV | REQ_RAHEAD)) + control |= NVME_RW_LR; + + dsmgmt = 0; + if (bio->bi_rw & REQ_RAHEAD) + dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH; + + spin_lock_irqsave(&nvmeq->q_lock, flags); + cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail].rw; + + if (bio_data_dir(bio)) { + cmnd->opcode = nvme_cmd_write; + dma_dir = DMA_TO_DEVICE; + } else { + cmnd->opcode = nvme_cmd_read; + dma_dir = DMA_FROM_DEVICE; + } + + nvme_map_bio(nvmeq->q_dmadev, info, bio, dma_dir, psegs); + + cmnd->flags = 1; + cmnd->command_id = cmdid; + cmnd->nsid = cpu_to_le32(ns->ns_id); + cmnd->prp1 = cpu_to_le64(sg_phys(info->sg)); + /* XXX: Support more than one PRP */ + cmnd->slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); + cmnd->length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); + cmnd->control = cpu_to_le16(control); + cmnd->dsmgmt = cpu_to_le32(dsmgmt); + + writel(nvmeq->sq_tail, nvmeq->q_db); + if (++nvmeq->sq_tail == nvmeq->q_depth) + nvmeq->sq_tail = 0; + + spin_unlock_irqrestore(&nvmeq->q_lock, flags); + + return 0; + + free_info: + free_info(info); + congestion: + return -EBUSY; +} + +/* + * NB: return value of non-zero would mean that we were a stacking driver. + * make_request must always succeed. + */ +static int nvme_make_request(struct request_queue *q, struct bio *bio) +{ + struct nvme_ns *ns = q->queuedata; + struct nvme_queue *nvmeq = get_nvmeq(ns); + + if (nvme_submit_bio_queue(nvmeq, ns, bio)) { + blk_set_queue_congested(q, rw_is_sync(bio->bi_rw)); + bio_list_add(&nvmeq->sq_cong, bio); + } + put_nvmeq(nvmeq); + + return 0; +} + +struct sync_cmd_info { + struct task_struct *task; + u32 result; + int status; +}; + +static void sync_completion(struct nvme_queue *nvmeq, void *ctx, + struct nvme_completion *cqe) +{ + struct sync_cmd_info *cmdinfo = ctx; + cmdinfo->result = le32_to_cpup(&cqe->result); + cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; + wake_up_process(cmdinfo->task); +} + +typedef void (*completion_fn)(struct nvme_queue *, void *, + struct nvme_completion *); + +static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) +{ + u16 head, cycle; + + static const completion_fn completions[4] = { + [sync_completion_id] = sync_completion, + [bio_completion_id] = bio_completion, + }; + + head = nvmeq->cq_head; + cycle = nvmeq->cq_cycle; + + for (;;) { + unsigned long data; + void *ptr; + unsigned char handler; + struct nvme_completion cqe = nvmeq->cqes[head]; + if ((le16_to_cpu(cqe.status) & 1) != cycle) + break; + nvmeq->sq_head = le16_to_cpu(cqe.sq_head); + if (++head == nvmeq->q_depth) { + head = 0; + cycle = !cycle; + } + + data = free_cmdid(nvmeq, cqe.command_id); + handler = data & 3; + ptr = (void *)(data & ~3UL); + completions[handler](nvmeq, ptr, &cqe); + } + + /* If the controller ignores the cq head doorbell and continuously + * writes to the queue, it is theoretically possible to wrap around + * the queue twice and mistakenly return IRQ_NONE. Linux only + * requires that 0.1% of your interrupts are handled, so this isn't + * a big problem. + */ + if (head == nvmeq->cq_head && cycle == nvmeq->cq_cycle) + return IRQ_NONE; + + writel(head, nvmeq->q_db + 1); + nvmeq->cq_head = head; + nvmeq->cq_cycle = cycle; + + return IRQ_HANDLED; +} + +static irqreturn_t nvme_irq(int irq, void *data) +{ + return nvme_process_cq(data); +} + +/* + * Returns 0 on success. If the result is negative, it's a Linux error code; + * if the result is positive, it's an NVM Express status code + */ +static int nvme_submit_sync_cmd(struct nvme_queue *q, struct nvme_command *cmd, + u32 *result) +{ + int cmdid; + struct sync_cmd_info cmdinfo; + + cmdinfo.task = current; + cmdinfo.status = -EINTR; + + cmdid = alloc_cmdid_killable(q, &cmdinfo, sync_completion_id); + if (cmdid < 0) + return cmdid; + cmd->common.command_id = cmdid; + + set_current_state(TASK_UNINTERRUPTIBLE); + nvme_submit_cmd(q, cmd); + schedule(); + + if (result) + *result = cmdinfo.result; + + return cmdinfo.status; +} + +static int nvme_submit_admin_cmd(struct nvme_dev *dev, struct nvme_command *cmd, + u32 *result) +{ + return nvme_submit_sync_cmd(dev->queues[0], cmd, result); +} + +static int adapter_delete_queue(struct nvme_dev *dev, u8 opcode, u16 id) +{ + int status; + struct nvme_command c; + + memset(&c, 0, sizeof(c)); + c.delete_queue.opcode = opcode; + c.delete_queue.qid = cpu_to_le16(id); + + status = nvme_submit_admin_cmd(dev, &c, NULL); + if (status) + return -EIO; + return 0; +} + +static int adapter_alloc_cq(struct nvme_dev *dev, u16 qid, + struct nvme_queue *nvmeq) +{ + int status; + struct nvme_command c; + int flags = NVME_QUEUE_PHYS_CONTIG | NVME_CQ_IRQ_ENABLED; + + memset(&c, 0, sizeof(c)); + c.create_cq.opcode = nvme_admin_create_cq; + c.create_cq.prp1 = cpu_to_le64(nvmeq->cq_dma_addr); + c.create_cq.cqid = cpu_to_le16(qid); + c.create_cq.qsize = cpu_to_le16(nvmeq->q_depth - 1); + c.create_cq.cq_flags = cpu_to_le16(flags); + c.create_cq.irq_vector = cpu_to_le16(nvmeq->cq_vector); + + status = nvme_submit_admin_cmd(dev, &c, NULL); + if (status) + return -EIO; + return 0; +} + +static int adapter_alloc_sq(struct nvme_dev *dev, u16 qid, + struct nvme_queue *nvmeq) +{ + int status; + struct nvme_command c; + int flags = NVME_QUEUE_PHYS_CONTIG | NVME_SQ_PRIO_MEDIUM; + + memset(&c, 0, sizeof(c)); + c.create_sq.opcode = nvme_admin_create_sq; + c.create_sq.prp1 = cpu_to_le64(nvmeq->sq_dma_addr); + c.create_sq.sqid = cpu_to_le16(qid); + c.create_sq.qsize = cpu_to_le16(nvmeq->q_depth - 1); + c.create_sq.sq_flags = cpu_to_le16(flags); + c.create_sq.cqid = cpu_to_le16(qid); + + status = nvme_submit_admin_cmd(dev, &c, NULL); + if (status) + return -EIO; + return 0; +} + +static int adapter_delete_cq(struct nvme_dev *dev, u16 cqid) +{ + return adapter_delete_queue(dev, nvme_admin_delete_cq, cqid); +} + +static int adapter_delete_sq(struct nvme_dev *dev, u16 sqid) +{ + return adapter_delete_queue(dev, nvme_admin_delete_sq, sqid); +} + +static void nvme_free_queue(struct nvme_dev *dev, int qid) +{ + struct nvme_queue *nvmeq = dev->queues[qid]; + + free_irq(dev->entry[nvmeq->cq_vector].vector, nvmeq); + + /* Don't tell the adapter to delete the admin queue */ + if (qid) { + adapter_delete_sq(dev, qid); + adapter_delete_cq(dev, qid); + } + + dma_free_coherent(nvmeq->q_dmadev, CQ_SIZE(nvmeq->q_depth), + (void *)nvmeq->cqes, nvmeq->cq_dma_addr); + dma_free_coherent(nvmeq->q_dmadev, SQ_SIZE(nvmeq->q_depth), + nvmeq->sq_cmds, nvmeq->sq_dma_addr); + kfree(nvmeq); +} + +static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, + int depth, int vector) +{ + struct device *dmadev = &dev->pci_dev->dev; + unsigned extra = (depth + BITS_TO_LONGS(depth)) * sizeof(long); + struct nvme_queue *nvmeq = kzalloc(sizeof(*nvmeq) + extra, GFP_KERNEL); + if (!nvmeq) + return NULL; + + nvmeq->cqes = dma_alloc_coherent(dmadev, CQ_SIZE(depth), + &nvmeq->cq_dma_addr, GFP_KERNEL); + if (!nvmeq->cqes) + goto free_nvmeq; + memset((void *)nvmeq->cqes, 0, CQ_SIZE(depth)); + + nvmeq->sq_cmds = dma_alloc_coherent(dmadev, SQ_SIZE(depth), + &nvmeq->sq_dma_addr, GFP_KERNEL); + if (!nvmeq->sq_cmds) + goto free_cqdma; + + nvmeq->q_dmadev = dmadev; + spin_lock_init(&nvmeq->q_lock); + nvmeq->cq_head = 0; + nvmeq->cq_cycle = 1; + init_waitqueue_head(&nvmeq->sq_full); + bio_list_init(&nvmeq->sq_cong); + nvmeq->q_db = &dev->dbs[qid * 2]; + nvmeq->q_depth = depth; + nvmeq->cq_vector = vector; + + return nvmeq; + + free_cqdma: + dma_free_coherent(dmadev, CQ_SIZE(nvmeq->q_depth), (void *)nvmeq->cqes, + nvmeq->cq_dma_addr); + free_nvmeq: + kfree(nvmeq); + return NULL; +} + +static __devinit struct nvme_queue *nvme_create_queue(struct nvme_dev *dev, + int qid, int cq_size, int vector) +{ + int result; + struct nvme_queue *nvmeq = nvme_alloc_queue(dev, qid, cq_size, vector); + + result = adapter_alloc_cq(dev, qid, nvmeq); + if (result < 0) + goto free_nvmeq; + + result = adapter_alloc_sq(dev, qid, nvmeq); + if (result < 0) + goto release_cq; + + result = request_irq(dev->entry[vector].vector, nvme_irq, + IRQF_DISABLED | IRQF_SHARED, "nvme", nvmeq); + if (result < 0) + goto release_sq; + + return nvmeq; + + release_sq: + adapter_delete_sq(dev, qid); + release_cq: + adapter_delete_cq(dev, qid); + free_nvmeq: + dma_free_coherent(nvmeq->q_dmadev, CQ_SIZE(nvmeq->q_depth), + (void *)nvmeq->cqes, nvmeq->cq_dma_addr); + dma_free_coherent(nvmeq->q_dmadev, SQ_SIZE(nvmeq->q_depth), + nvmeq->sq_cmds, nvmeq->sq_dma_addr); + kfree(nvmeq); + return NULL; +} + +static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) +{ + int result; + u32 aqa; + struct nvme_queue *nvmeq; + + dev->dbs = ((void __iomem *)dev->bar) + 4096; + + nvmeq = nvme_alloc_queue(dev, 0, 64, 0); + + aqa = nvmeq->q_depth - 1; + aqa |= aqa << 16; + + dev->ctrl_config = NVME_CC_ENABLE | NVME_CC_CSS_NVM; + dev->ctrl_config |= (PAGE_SHIFT - 12) << NVME_CC_MPS_SHIFT; + dev->ctrl_config |= NVME_CC_ARB_RR | NVME_CC_SHN_NONE; + + writel(aqa, &dev->bar->aqa); + writeq(nvmeq->sq_dma_addr, &dev->bar->asq); + writeq(nvmeq->cq_dma_addr, &dev->bar->acq); + writel(dev->ctrl_config, &dev->bar->cc); + + while (!(readl(&dev->bar->csts) & NVME_CSTS_RDY)) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + } + + result = request_irq(dev->entry[0].vector, nvme_irq, + IRQF_DISABLED | IRQF_SHARED, "nvme admin", nvmeq); + dev->queues[0] = nvmeq; + return result; +} + +static int nvme_identify(struct nvme_ns *ns, void __user *addr, int cns) +{ + struct nvme_dev *dev = ns->dev; + int status; + struct nvme_command c; + void *page; + dma_addr_t dma_addr; + + page = dma_alloc_coherent(&dev->pci_dev->dev, 4096, &dma_addr, + GFP_KERNEL); + + memset(&c, 0, sizeof(c)); + c.identify.opcode = nvme_admin_identify; + c.identify.nsid = cns ? 0 : cpu_to_le32(ns->ns_id); + c.identify.prp1 = cpu_to_le64(dma_addr); + c.identify.cns = cpu_to_le32(cns); + + status = nvme_submit_admin_cmd(dev, &c, NULL); + + if (status) + status = -EIO; + else if (copy_to_user(addr, page, 4096)) + status = -EFAULT; + + dma_free_coherent(&dev->pci_dev->dev, 4096, page, dma_addr); + + return status; +} + +static int nvme_get_range_type(struct nvme_ns *ns, void __user *addr) +{ + struct nvme_dev *dev = ns->dev; + int status; + struct nvme_command c; + void *page; + dma_addr_t dma_addr; + + page = dma_alloc_coherent(&dev->pci_dev->dev, 4096, &dma_addr, + GFP_KERNEL); + + memset(&c, 0, sizeof(c)); + c.features.opcode = nvme_admin_get_features; + c.features.nsid = cpu_to_le32(ns->ns_id); + c.features.prp1 = cpu_to_le64(dma_addr); + c.features.fid = cpu_to_le32(NVME_FEAT_LBA_RANGE); + + status = nvme_submit_admin_cmd(dev, &c, NULL); + + /* XXX: Assuming first range for now */ + if (status) + status = -EIO; + else if (copy_to_user(addr, page, 64)) + status = -EFAULT; + + dma_free_coherent(&dev->pci_dev->dev, 4096, page, dma_addr); + + return status; +} + +static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, + unsigned long arg) +{ + struct nvme_ns *ns = bdev->bd_disk->private_data; + + switch (cmd) { + case NVME_IOCTL_IDENTIFY_NS: + return nvme_identify(ns, (void __user *)arg, 0); + case NVME_IOCTL_IDENTIFY_CTRL: + return nvme_identify(ns, (void __user *)arg, 1); + case NVME_IOCTL_GET_RANGE_TYPE: + return nvme_get_range_type(ns, (void __user *)arg); + default: + return -ENOTTY; + } +} + +static const struct block_device_operations nvme_fops = { + .owner = THIS_MODULE, + .ioctl = nvme_ioctl, +}; + +static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int index, + struct nvme_id_ns *id, struct nvme_lba_range_type *rt) +{ + struct nvme_ns *ns; + struct gendisk *disk; + int lbaf; + + if (rt->attributes & NVME_LBART_ATTRIB_HIDE) + return NULL; + + ns = kzalloc(sizeof(*ns), GFP_KERNEL); + if (!ns) + return NULL; + ns->queue = blk_alloc_queue(GFP_KERNEL); + if (!ns->queue) + goto out_free_ns; + ns->queue->queue_flags = QUEUE_FLAG_DEFAULT | QUEUE_FLAG_NOMERGES | + QUEUE_FLAG_NONROT | QUEUE_FLAG_DISCARD; + blk_queue_make_request(ns->queue, nvme_make_request); + ns->dev = dev; + ns->queue->queuedata = ns; + + disk = alloc_disk(NVME_MINORS); + if (!disk) + goto out_free_queue; + ns->ns_id = index; + ns->disk = disk; + lbaf = id->flbas & 0xf; + ns->lba_shift = id->lbaf[lbaf].ds; + + disk->major = nvme_major; + disk->minors = NVME_MINORS; + disk->first_minor = NVME_MINORS * index; + disk->fops = &nvme_fops; + disk->private_data = ns; + disk->queue = ns->queue; + sprintf(disk->disk_name, "nvme%dn%d", dev->instance, index); + set_capacity(disk, le64_to_cpup(&id->nsze) << (ns->lba_shift - 9)); + + return ns; + + out_free_queue: + blk_cleanup_queue(ns->queue); + out_free_ns: + kfree(ns); + return NULL; +} + +static void nvme_ns_free(struct nvme_ns *ns) +{ + put_disk(ns->disk); + blk_cleanup_queue(ns->queue); + kfree(ns); +} + +static int set_queue_count(struct nvme_dev *dev, int sq_count, int cq_count) +{ + int status; + u32 result; + struct nvme_command c; + u32 q_count = (sq_count - 1) | ((cq_count - 1) << 16); + + memset(&c, 0, sizeof(c)); + c.features.opcode = nvme_admin_get_features; + c.features.fid = cpu_to_le32(NVME_FEAT_NUM_QUEUES); + c.features.dword11 = cpu_to_le32(q_count); + + status = nvme_submit_admin_cmd(dev, &c, &result); + if (status) + return -EIO; + return min(result & 0xffff, result >> 16) + 1; +} + +/* XXX: Create per-CPU queues */ +static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) +{ + int this_cpu; + + set_queue_count(dev, 1, 1); + + this_cpu = get_cpu(); + dev->queues[1] = nvme_create_queue(dev, 1, NVME_Q_DEPTH, this_cpu); + put_cpu(); + if (!dev->queues[1]) + return -ENOMEM; + dev->queue_count++; + + return 0; +} + +static void nvme_free_queues(struct nvme_dev *dev) +{ + int i; + + for (i = dev->queue_count - 1; i >= 0; i--) + nvme_free_queue(dev, i); +} + +static int __devinit nvme_dev_add(struct nvme_dev *dev) +{ + int res, nn, i; + struct nvme_ns *ns, *next; + void *id; + dma_addr_t dma_addr; + struct nvme_command cid, crt; + + res = nvme_setup_io_queues(dev); + if (res) + return res; + + /* XXX: Switch to a SG list once prp2 works */ + id = dma_alloc_coherent(&dev->pci_dev->dev, 8192, &dma_addr, + GFP_KERNEL); + + memset(&cid, 0, sizeof(cid)); + cid.identify.opcode = nvme_admin_identify; + cid.identify.nsid = 0; + cid.identify.prp1 = cpu_to_le64(dma_addr); + cid.identify.cns = cpu_to_le32(1); + + res = nvme_submit_admin_cmd(dev, &cid, NULL); + if (res) { + res = -EIO; + goto out_free; + } + + nn = le32_to_cpup(&((struct nvme_id_ctrl *)id)->nn); + + cid.identify.cns = 0; + memset(&crt, 0, sizeof(crt)); + crt.features.opcode = nvme_admin_get_features; + crt.features.prp1 = cpu_to_le64(dma_addr + 4096); + crt.features.fid = cpu_to_le32(NVME_FEAT_LBA_RANGE); + + for (i = 0; i < nn; i++) { + cid.identify.nsid = cpu_to_le32(i); + res = nvme_submit_admin_cmd(dev, &cid, NULL); + if (res) + continue; + + if (((struct nvme_id_ns *)id)->ncap == 0) + continue; + + crt.features.nsid = cpu_to_le32(i); + res = nvme_submit_admin_cmd(dev, &crt, NULL); + if (res) + continue; + + ns = nvme_alloc_ns(dev, i, id, id + 4096); + if (ns) + list_add_tail(&ns->list, &dev->namespaces); + } + list_for_each_entry(ns, &dev->namespaces, list) + add_disk(ns->disk); + + dma_free_coherent(&dev->pci_dev->dev, 4096, id, dma_addr); + return 0; + + out_free: + list_for_each_entry_safe(ns, next, &dev->namespaces, list) { + list_del(&ns->list); + nvme_ns_free(ns); + } + + dma_free_coherent(&dev->pci_dev->dev, 4096, id, dma_addr); + return res; +} + +static int nvme_dev_remove(struct nvme_dev *dev) +{ + struct nvme_ns *ns, *next; + + /* TODO: wait all I/O finished or cancel them */ + + list_for_each_entry_safe(ns, next, &dev->namespaces, list) { + list_del(&ns->list); + del_gendisk(ns->disk); + nvme_ns_free(ns); + } + + nvme_free_queues(dev); + + return 0; +} + +/* XXX: Use an ida or something to let remove / add work correctly */ +static void nvme_set_instance(struct nvme_dev *dev) +{ + static int instance; + dev->instance = instance++; +} + +static void nvme_release_instance(struct nvme_dev *dev) +{ +} + +static int __devinit nvme_probe(struct pci_dev *pdev, + const struct pci_device_id *id) +{ + int result = -ENOMEM; + struct nvme_dev *dev; + + dev = kzalloc(sizeof(*dev), GFP_KERNEL); + if (!dev) + return -ENOMEM; + dev->entry = kcalloc(num_possible_cpus(), sizeof(*dev->entry), + GFP_KERNEL); + if (!dev->entry) + goto free; + dev->queues = kcalloc(2, sizeof(void *), GFP_KERNEL); + if (!dev->queues) + goto free; + + INIT_LIST_HEAD(&dev->namespaces); + dev->pci_dev = pdev; + pci_set_drvdata(pdev, dev); + dma_set_mask(&dev->pci_dev->dev, DMA_BIT_MASK(64)); + nvme_set_instance(dev); + + dev->bar = ioremap(pci_resource_start(pdev, 0), 8192); + if (!dev->bar) { + result = -ENOMEM; + goto disable; + } + + result = nvme_configure_admin_queue(dev); + if (result) + goto unmap; + dev->queue_count++; + + result = nvme_dev_add(dev); + if (result) + goto delete; + return 0; + + delete: + nvme_free_queues(dev); + unmap: + iounmap(dev->bar); + disable: + pci_disable_msix(pdev); + nvme_release_instance(dev); + free: + kfree(dev->queues); + kfree(dev->entry); + kfree(dev); + return result; +} + +static void __devexit nvme_remove(struct pci_dev *pdev) +{ + struct nvme_dev *dev = pci_get_drvdata(pdev); + nvme_dev_remove(dev); + pci_disable_msix(pdev); + iounmap(dev->bar); + nvme_release_instance(dev); + kfree(dev->queues); + kfree(dev->entry); + kfree(dev); +} + +/* These functions are yet to be implemented */ +#define nvme_error_detected NULL +#define nvme_dump_registers NULL +#define nvme_link_reset NULL +#define nvme_slot_reset NULL +#define nvme_error_resume NULL +#define nvme_suspend NULL +#define nvme_resume NULL + +static struct pci_error_handlers nvme_err_handler = { + .error_detected = nvme_error_detected, + .mmio_enabled = nvme_dump_registers, + .link_reset = nvme_link_reset, + .slot_reset = nvme_slot_reset, + .resume = nvme_error_resume, +}; + +/* Move to pci_ids.h later */ +#define PCI_CLASS_STORAGE_EXPRESS 0x010802 + +static DEFINE_PCI_DEVICE_TABLE(nvme_id_table) = { + { PCI_DEVICE_CLASS(PCI_CLASS_STORAGE_EXPRESS, 0xffffff) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, nvme_id_table); + +static struct pci_driver nvme_driver = { + .name = "nvme", + .id_table = nvme_id_table, + .probe = nvme_probe, + .remove = __devexit_p(nvme_remove), + .suspend = nvme_suspend, + .resume = nvme_resume, + .err_handler = &nvme_err_handler, +}; + +static int __init nvme_init(void) +{ + int result; + + nvme_major = register_blkdev(nvme_major, "nvme"); + if (nvme_major <= 0) + return -EBUSY; + + result = pci_register_driver(&nvme_driver); + if (!result) + return 0; + + unregister_blkdev(nvme_major, "nvme"); + return result; +} + +static void __exit nvme_exit(void) +{ + pci_unregister_driver(&nvme_driver); + unregister_blkdev(nvme_major, "nvme"); +} + +MODULE_AUTHOR("Matthew Wilcox "); +MODULE_LICENSE("GPL"); +MODULE_VERSION("0.1"); +module_init(nvme_init); +module_exit(nvme_exit); diff --git a/include/linux/nvme.h b/include/linux/nvme.h new file mode 100644 index 000000000000..9ba53584f722 --- /dev/null +++ b/include/linux/nvme.h @@ -0,0 +1,343 @@ +/* + * Definitions for the NVM Express interface + * Copyright (c) 2011, Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + */ + +#ifndef _LINUX_NVME_H +#define _LINUX_NVME_H + +#include + +struct nvme_bar { + __u64 cap; /* Controller Capabilities */ + __u32 vs; /* Version */ + __u32 ims; /* Interrupt Mask Set */ + __u32 imc; /* Interrupt Mask Clear */ + __u32 cc; /* Controller Configuration */ + __u32 csts; /* Controller Status */ + __u32 aqa; /* Admin Queue Attributes */ + __u64 asq; /* Admin SQ Base Address */ + __u64 acq; /* Admin CQ Base Address */ +}; + +enum { + NVME_CC_ENABLE = 1 << 0, + NVME_CC_CSS_NVM = 0 << 4, + NVME_CC_MPS_SHIFT = 7, + NVME_CC_ARB_RR = 0 << 11, + NVME_CC_ARB_WRRU = 1 << 11, + NVME_CC_ARB_VS = 3 << 11, + NVME_CC_SHN_NONE = 0 << 13, + NVME_CC_SHN_NORMAL = 1 << 13, + NVME_CC_SHN_ABRUPT = 2 << 13, + NVME_CSTS_RDY = 1 << 0, + NVME_CSTS_CFS = 1 << 1, + NVME_CSTS_SHST_NORMAL = 0 << 2, + NVME_CSTS_SHST_OCCUR = 1 << 2, + NVME_CSTS_SHST_CMPLT = 2 << 2, +}; + +#define NVME_VS(major, minor) (major << 16 | minor) + +struct nvme_id_ctrl { + __le16 vid; + __le16 ssvid; + char sn[20]; + char mn[40]; + char fr[8]; + __le32 nn; + __u8 rab; + __u8 rsvd77[178]; + __le16 oacs; + __u8 acl; + __u8 aerl; + __u8 frmw; + __u8 lpa; + __u8 elpe; + __u8 npss; + __u8 rsvd264[248]; + __le64 psd[32]; + __le16 oncs; + __le16 fuses; + __u8 fna; + __u8 vwc; + __le16 awun; + __le16 awupf; + __u8 rsvd778[246]; + __u8 cmdset[2048]; + __u8 vs[1024]; +}; + +struct nvme_lbaf { + __le16 ms; + __u8 ds; + __u8 rp; +}; + +struct nvme_id_ns { + __le64 nsze; + __le64 ncap; + __le64 nuse; + __u8 nsfeat; + __u8 nlbaf; + __u8 flbas; + __u8 mc; + __u8 dpc; + __u8 dps; + __u8 rsvd30[98]; + struct nvme_lbaf lbaf[16]; + __u8 rsvd192[192]; + __u8 vs[3712]; +}; + +enum { + NVME_NS_FEAT_THIN = 1 << 0, + NVME_LBAF_RP_BEST = 0, + NVME_LBAF_RP_BETTER = 1, + NVME_LBAF_RP_GOOD = 2, + NVME_LBAF_RP_DEGRADED = 3, +}; + +struct nvme_lba_range_type { + __u8 type; + __u8 attributes; + __u8 rsvd2[14]; + __u64 slba; + __u64 nlb; + __u8 guid[16]; + __u8 rsvd48[16]; +}; + +enum { + NVME_LBART_TYPE_FS = 0x01, + NVME_LBART_TYPE_RAID = 0x02, + NVME_LBART_TYPE_CACHE = 0x03, + NVME_LBART_TYPE_SWAP = 0x04, + + NVME_LBART_ATTRIB_TEMP = 1 << 0, + NVME_LBART_ATTRIB_HIDE = 1 << 1, +}; + +/* I/O commands */ + +enum nvme_opcode { + nvme_cmd_flush = 0x00, + nvme_cmd_write = 0x01, + nvme_cmd_read = 0x02, + nvme_cmd_write_uncor = 0x04, + nvme_cmd_compare = 0x05, + nvme_cmd_dsm = 0x09, +}; + +struct nvme_rw_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 nsid; + __u64 rsvd2; + __le64 metadata; + __le64 prp1; + __le64 prp2; + __le64 slba; + __le16 length; + __le16 control; + __le32 dsmgmt; + __le32 reftag; + __le16 apptag; + __le16 appmask; +}; + +enum { + NVME_RW_LR = 1 << 15, + NVME_RW_FUA = 1 << 14, + NVME_RW_DSM_FREQ_UNSPEC = 0, + NVME_RW_DSM_FREQ_TYPICAL = 1, + NVME_RW_DSM_FREQ_RARE = 2, + NVME_RW_DSM_FREQ_READS = 3, + NVME_RW_DSM_FREQ_WRITES = 4, + NVME_RW_DSM_FREQ_RW = 5, + NVME_RW_DSM_FREQ_ONCE = 6, + NVME_RW_DSM_FREQ_PREFETCH = 7, + NVME_RW_DSM_FREQ_TEMP = 8, + NVME_RW_DSM_LATENCY_NONE = 0 << 4, + NVME_RW_DSM_LATENCY_IDLE = 1 << 4, + NVME_RW_DSM_LATENCY_NORM = 2 << 4, + NVME_RW_DSM_LATENCY_LOW = 3 << 4, + NVME_RW_DSM_SEQ_REQ = 1 << 6, + NVME_RW_DSM_COMPRESSED = 1 << 7, +}; + +/* Admin commands */ + +enum nvme_admin_opcode { + nvme_admin_delete_sq = 0x00, + nvme_admin_create_sq = 0x01, + nvme_admin_get_features = 0x02, + nvme_admin_delete_cq = 0x04, + nvme_admin_create_cq = 0x05, + nvme_admin_identify = 0x06, + nvme_admin_abort_cmd = 0x08, + nvme_admin_set_features = 0x09, + nvme_admin_get_log_page = 0x0a, + nvme_admin_async_event = 0x0c, + nvme_admin_download_fw = 0x0d, + nvme_admin_security_recv = 0x0e, + nvme_admin_format_nvm = 0x10, + nvme_admin_security_send = 0x11, + nvme_admin_activate_fw = 0x14, +}; + +enum { + NVME_QUEUE_PHYS_CONTIG = (1 << 0), + NVME_CQ_IRQ_ENABLED = (1 << 1), + NVME_SQ_PRIO_URGENT = (0 << 1), + NVME_SQ_PRIO_HIGH = (1 << 1), + NVME_SQ_PRIO_MEDIUM = (2 << 1), + NVME_SQ_PRIO_LOW = (3 << 1), + NVME_FEAT_ARBITRATION = 0x01, + NVME_FEAT_POWER_MGMT = 0x02, + NVME_FEAT_LBA_RANGE = 0x03, + NVME_FEAT_TEMP_THRESH = 0x04, + NVME_FEAT_ERR_RECOVERY = 0x05, + NVME_FEAT_VOLATILE_WC = 0x06, + NVME_FEAT_NUM_QUEUES = 0x07, + NVME_FEAT_IRQ_COALESCE = 0x08, + NVME_FEAT_IRQ_CONFIG = 0x09, + NVME_FEAT_WRITE_ATOMIC = 0x0a, + NVME_FEAT_ASYNC_EVENT = 0x0b, + NVME_FEAT_SW_PROGRESS = 0x0c, +}; + +struct nvme_identify { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 nsid; + __u64 rsvd2[2]; + __le64 prp1; + __le64 prp2; + __le32 cns; + __u32 rsvd11[5]; +}; + +struct nvme_features { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 nsid; + __u64 rsvd2[2]; + __le64 prp1; + __le64 prp2; + __le32 fid; + __le32 dword11; + __u32 rsvd12[4]; +}; + +struct nvme_create_cq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct nvme_create_sq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __le32 rsvd12[4]; +}; + +struct nvme_delete_queue { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[9]; + __le16 qid; + __le16 rsvd10; + __le32 rsvd11[5]; +}; + +struct nvme_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 nsid; + __u32 rsvd2[14]; +}; + +struct nvme_command { + union { + struct nvme_common_command common; + struct nvme_rw_command rw; + struct nvme_identify identify; + struct nvme_features features; + struct nvme_create_cq create_cq; + struct nvme_create_sq create_sq; + struct nvme_delete_queue delete_queue; + }; +}; + +/* XXX: Sync with spec */ +enum { + NVME_SC_SUCCESS = 0x0, + NVME_SC_INVALID_OPCODE = 0x1, + NVME_SC_INVALID_FIELD = 0x2, + NVME_SC_CMDID_CONFLICT = 0x3, + NVME_SC_DATA_XFER_ERROR = 0x4, + NVME_SC_POWER_LOSS = 0x5, + NVME_SC_INTERNAL = 0x6, + NVME_SC_ABORT_REQ = 0x7, + NVME_SC_ABORT_QUEUE = 0x8, + NVME_SC_FUSED_FAIL = 0x9, + NVME_SC_FUSED_MISSING = 0xa, + NVME_SC_LBA_RANGE = 0x80, + NVME_SC_CAP_EXCEEDED = 0x81, + NVME_SC_NS_NOT_READY = 0x82, + NVME_SC_CQ_INVALID = 0x100, + NVME_SC_QID_INVALID = 0x101, + NVME_SC_QUEUE_SIZE = 0x102, + NVME_SC_WRITE_FAULT = 0x280, + NVME_SC_READ_ERROR = 0x281, +}; + +struct nvme_completion { + __le32 result; /* Used by admin commands to return data */ + __le32 rsvd; + __le16 sq_head; /* how much of this queue may be reclaimed */ + __le16 sq_id; /* submission queue that generated this entry */ + __u16 command_id; /* of the command which completed */ + __le16 status; /* did the command fail, and if so, why? */ +}; + +#define NVME_IOCTL_IDENTIFY_NS _IOW('N', 0x40, struct nvme_id_ns) +#define NVME_IOCTL_IDENTIFY_CTRL _IOW('N', 0x41, struct nvme_id_ctrl) +#define NVME_IOCTL_GET_RANGE_TYPE _IOW('N', 0x42, struct nvme_lba_range_type) + +#endif /* _LINUX_NVME_H */ From 3001082cac4bf6ffd09f72b39e6292ad6394ef17 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Jan 2011 09:10:15 -0500 Subject: [PATCH 012/820] NVMe: Factor out queue_request_irq() Two callers with an almost identical long string of arguments, and introducing a third soon. Time to factor out the commonalities. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index ef66eccc2aa2..b10e064795ed 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -568,6 +568,13 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, return NULL; } +static int queue_request_irq(struct nvme_dev *dev, struct nvme_queue *nvmeq, + const char *name) +{ + return request_irq(dev->entry[nvmeq->cq_vector].vector, nvme_irq, + IRQF_DISABLED | IRQF_SHARED, name, nvmeq); +} + static __devinit struct nvme_queue *nvme_create_queue(struct nvme_dev *dev, int qid, int cq_size, int vector) { @@ -582,8 +589,7 @@ static __devinit struct nvme_queue *nvme_create_queue(struct nvme_dev *dev, if (result < 0) goto release_cq; - result = request_irq(dev->entry[vector].vector, nvme_irq, - IRQF_DISABLED | IRQF_SHARED, "nvme", nvmeq); + result = queue_request_irq(dev, nvmeq, "nvme"); if (result < 0) goto release_sq; @@ -630,8 +636,7 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) return -EINTR; } - result = request_irq(dev->entry[0].vector, nvme_irq, - IRQF_DISABLED | IRQF_SHARED, "nvme admin", nvmeq); + result = queue_request_irq(dev, nvmeq, "nvme admin"); dev->queues[0] = nvmeq; return result; } From b3b06812e199f248561ce7824a4a8a9cd573c05a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Jan 2011 09:14:34 -0500 Subject: [PATCH 013/820] NVMe: Reduce set_queue_count arguments by one sq_count and cq_count are always the same, so just call it 'count'. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b10e064795ed..7efd7e92b637 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -777,12 +777,12 @@ static void nvme_ns_free(struct nvme_ns *ns) kfree(ns); } -static int set_queue_count(struct nvme_dev *dev, int sq_count, int cq_count) +static int set_queue_count(struct nvme_dev *dev, int count) { int status; u32 result; struct nvme_command c; - u32 q_count = (sq_count - 1) | ((cq_count - 1) << 16); + u32 q_count = (count - 1) | ((count - 1) << 16); memset(&c, 0, sizeof(c)); c.features.opcode = nvme_admin_get_features; @@ -800,7 +800,7 @@ static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) { int this_cpu; - set_queue_count(dev, 1, 1); + set_queue_count(dev, 1); this_cpu = get_cpu(); dev->queues[1] = nvme_create_queue(dev, 1, NVME_Q_DEPTH, this_cpu); From 1b23484bd012c078de2ea939249e2fb2e85a0a6e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Jan 2011 13:01:49 -0500 Subject: [PATCH 014/820] NVMe: Implement per-CPU queues Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 61 ++++++++++++++++++++++++++++++++++++-------- 1 file changed, 50 insertions(+), 11 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 7efd7e92b637..b6a213c98584 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -172,11 +172,17 @@ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) { - return ns->dev->queues[1]; + int qid, cpu = get_cpu(); + if (cpu < ns->dev->queue_count) + qid = cpu + 1; + else + qid = (cpu % rounddown_pow_of_two(ns->dev->queue_count)) + 1; + return ns->dev->queues[qid]; } static void put_nvmeq(struct nvme_queue *nvmeq) { + put_cpu(); } /** @@ -795,19 +801,51 @@ static int set_queue_count(struct nvme_dev *dev, int count) return min(result & 0xffff, result >> 16) + 1; } -/* XXX: Create per-CPU queues */ static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) { - int this_cpu; + int result, cpu, i, nr_queues; - set_queue_count(dev, 1); + nr_queues = num_online_cpus(); + result = set_queue_count(dev, nr_queues); + if (result < 0) + return result; + if (result < nr_queues) + nr_queues = result; - this_cpu = get_cpu(); - dev->queues[1] = nvme_create_queue(dev, 1, NVME_Q_DEPTH, this_cpu); - put_cpu(); - if (!dev->queues[1]) - return -ENOMEM; - dev->queue_count++; + /* Deregister the admin queue's interrupt */ + free_irq(dev->entry[0].vector, dev->queues[0]); + + for (i = 0; i < nr_queues; i++) + dev->entry[i].entry = i; + for (;;) { + result = pci_enable_msix(dev->pci_dev, dev->entry, nr_queues); + if (result == 0) { + break; + } else if (result > 0) { + nr_queues = result; + continue; + } else { + nr_queues = 1; + break; + } + } + + result = queue_request_irq(dev, dev->queues[0], "nvme admin"); + /* XXX: handle failure here */ + + cpu = cpumask_first(cpu_online_mask); + for (i = 0; i < nr_queues; i++) { + irq_set_affinity_hint(dev->entry[i].vector, get_cpu_mask(cpu)); + cpu = cpumask_next(cpu, cpu_online_mask); + } + + for (i = 0; i < nr_queues; i++) { + dev->queues[i + 1] = nvme_create_queue(dev, i + 1, + NVME_Q_DEPTH, i); + if (!dev->queues[i + 1]) + return -ENOMEM; + dev->queue_count++; + } return 0; } @@ -931,7 +969,8 @@ static int __devinit nvme_probe(struct pci_dev *pdev, GFP_KERNEL); if (!dev->entry) goto free; - dev->queues = kcalloc(2, sizeof(void *), GFP_KERNEL); + dev->queues = kcalloc(num_possible_cpus() + 1, sizeof(void *), + GFP_KERNEL); if (!dev->queues) goto free; From 821234603b265f59d7eebce16d9e8beca2a5752d Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Jan 2011 13:24:06 -0500 Subject: [PATCH 015/820] NVMe: Rename 'cycle' to 'phase' It's called the phase bit in the current draft Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b6a213c98584..3d917a87ea93 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -93,7 +93,7 @@ struct nvme_queue { u16 sq_head; u16 sq_tail; u16 cq_head; - u16 cq_cycle; + u16 cq_phase; unsigned long cmdid_data[]; }; @@ -364,7 +364,7 @@ typedef void (*completion_fn)(struct nvme_queue *, void *, static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) { - u16 head, cycle; + u16 head, phase; static const completion_fn completions[4] = { [sync_completion_id] = sync_completion, @@ -372,19 +372,19 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) }; head = nvmeq->cq_head; - cycle = nvmeq->cq_cycle; + phase = nvmeq->cq_phase; for (;;) { unsigned long data; void *ptr; unsigned char handler; struct nvme_completion cqe = nvmeq->cqes[head]; - if ((le16_to_cpu(cqe.status) & 1) != cycle) + if ((le16_to_cpu(cqe.status) & 1) != phase) break; nvmeq->sq_head = le16_to_cpu(cqe.sq_head); if (++head == nvmeq->q_depth) { head = 0; - cycle = !cycle; + phase = !phase; } data = free_cmdid(nvmeq, cqe.command_id); @@ -399,12 +399,12 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) * requires that 0.1% of your interrupts are handled, so this isn't * a big problem. */ - if (head == nvmeq->cq_head && cycle == nvmeq->cq_cycle) + if (head == nvmeq->cq_head && phase == nvmeq->cq_phase) return IRQ_NONE; writel(head, nvmeq->q_db + 1); nvmeq->cq_head = head; - nvmeq->cq_cycle = cycle; + nvmeq->cq_phase = phase; return IRQ_HANDLED; } @@ -557,7 +557,7 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, nvmeq->q_dmadev = dmadev; spin_lock_init(&nvmeq->q_lock); nvmeq->cq_head = 0; - nvmeq->cq_cycle = 1; + nvmeq->cq_phase = 1; init_waitqueue_head(&nvmeq->sq_full); bio_list_init(&nvmeq->sq_cong); nvmeq->q_db = &dev->dbs[qid * 2]; From 53c9577e9ca68a633c6e9df2b54eaecacfa77f62 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Jan 2011 13:42:34 -0500 Subject: [PATCH 016/820] NVMe: Fix admin IRQ claim on real hardware The admin IRQ is supposed to use the pin-based (or single message MSI) interrupt. Accomplish this by filling in entry[0]'s vector with the INTx irq number. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 3d917a87ea93..44a9d5edd8db 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -979,6 +979,7 @@ static int __devinit nvme_probe(struct pci_dev *pdev, pci_set_drvdata(pdev, dev); dma_set_mask(&dev->pci_dev->dev, DMA_BIT_MASK(64)); nvme_set_instance(dev); + dev->entry[0].vector = pdev->irq; dev->bar = ioremap(pci_resource_start(pdev, 0), 8192); if (!dev->bar) { From 36c14ed9caa957c686d4a48fd598a5ec2aa0331b Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 24 Jan 2011 07:52:07 -0500 Subject: [PATCH 017/820] NVMe: Use PRP2 for the nvme_identify ioctl DMA the result straight to userspace instead of bounce-buffering in the kernel. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 56 +++++++++++++++++++++++++++++++------------- 1 file changed, 40 insertions(+), 16 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 44a9d5edd8db..c0ef1dd1cc90 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -647,33 +647,57 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) return result; } -static int nvme_identify(struct nvme_ns *ns, void __user *addr, int cns) +static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) { struct nvme_dev *dev = ns->dev; - int status; + int i, err, count, nents, offset; struct nvme_command c; - void *page; - dma_addr_t dma_addr; + struct scatterlist sg[2]; + struct page *pages[2]; - page = dma_alloc_coherent(&dev->pci_dev->dev, 4096, &dma_addr, - GFP_KERNEL); + if (addr & 3) + return -EINVAL; + offset = offset_in_page(addr); + count = offset ? 2 : 1; + + err = get_user_pages_fast(addr, count, 1, pages); + if (err < count) { + count = err; + err = -EFAULT; + goto put_pages; + } + sg_init_table(sg, count); + for (i = 0; i < count; i++) + sg_set_page(&sg[i], pages[i], PAGE_SIZE, 0); + nents = dma_map_sg(&dev->pci_dev->dev, sg, count, DMA_FROM_DEVICE); + if (!nents) + goto put_pages; memset(&c, 0, sizeof(c)); c.identify.opcode = nvme_admin_identify; c.identify.nsid = cns ? 0 : cpu_to_le32(ns->ns_id); - c.identify.prp1 = cpu_to_le64(dma_addr); + c.identify.prp1 = cpu_to_le64(sg_dma_address(&sg[0]) + offset); + if (count > 1) { + u64 dma_addr; + if (nents > 1) + dma_addr = sg_dma_address(&sg[1]); + else + dma_addr = sg_dma_address(&sg[0]) + PAGE_SIZE; + c.identify.prp2 = cpu_to_le64(dma_addr); + } c.identify.cns = cpu_to_le32(cns); - status = nvme_submit_admin_cmd(dev, &c, NULL); + err = nvme_submit_admin_cmd(dev, &c, NULL); - if (status) - status = -EIO; - else if (copy_to_user(addr, page, 4096)) - status = -EFAULT; + if (err) + err = -EIO; - dma_free_coherent(&dev->pci_dev->dev, 4096, page, dma_addr); + dma_unmap_sg(&dev->pci_dev->dev, sg, nents, DMA_FROM_DEVICE); + put_pages: + for (i = 0; i < count; i++) + put_page(pages[i]); - return status; + return err; } static int nvme_get_range_type(struct nvme_ns *ns, void __user *addr) @@ -713,9 +737,9 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, switch (cmd) { case NVME_IOCTL_IDENTIFY_NS: - return nvme_identify(ns, (void __user *)arg, 0); + return nvme_identify(ns, arg, 0); case NVME_IOCTL_IDENTIFY_CTRL: - return nvme_identify(ns, (void __user *)arg, 1); + return nvme_identify(ns, arg, 1); case NVME_IOCTL_GET_RANGE_TYPE: return nvme_get_range_type(ns, (void __user *)arg); default: From 7b4fe9b1cb4b9a6f4ae23a12ef96d08d96e2a5da Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 26 Jan 2011 10:01:21 -0500 Subject: [PATCH 018/820] NVMe: Make nvme_common_command more featureful Add prp1, prp2 and the metadata prp to the common command, since the fields are generally used this way. Signed-off-by: Matthew Wilcox --- include/linux/nvme.h | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 9ba53584f722..1c0b5ef08959 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -142,6 +142,18 @@ enum nvme_opcode { nvme_cmd_dsm = 0x09, }; +struct nvme_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 nsid; + __u64 rsvd2; + __le64 metadata; + __le64 prp1; + __le64 prp2; + __u32 rsvd10[6]; +}; + struct nvme_rw_command { __u8 opcode; __u8 flags; @@ -284,14 +296,6 @@ struct nvme_delete_queue { __le32 rsvd11[5]; }; -struct nvme_common_command { - __u8 opcode; - __u8 flags; - __u16 command_id; - __le32 nsid; - __u32 rsvd2[14]; -}; - struct nvme_command { union { struct nvme_common_command common; From ff22b54fda2078fc3cd1bcdcb7a5ce5d08fd6591 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 26 Jan 2011 10:02:29 -0500 Subject: [PATCH 019/820] NVMe: Add nvme_setup_prps() Generalise the code from nvme_identify() that sets PRP1 & PRP2 so that it's usable for commands sent by nvme_submit_bio_queue(). Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 70 +++++++++++++++++++++++++++++--------------- 1 file changed, 46 insertions(+), 24 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index c0ef1dd1cc90..1e57737b1760 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -240,6 +240,36 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, bio_endio(bio, status ? -EIO : 0); } +/* length is in bytes */ +static void nvme_setup_prps(struct nvme_common_command *cmd, + struct scatterlist *sg, int length) +{ + int dma_len = sg_dma_len(sg); + u64 dma_addr = sg_dma_address(sg); + int offset = offset_in_page(dma_addr); + + cmd->prp1 = cpu_to_le64(dma_addr); + length -= (PAGE_SIZE - offset); + if (length <= 0) + return; + + dma_len -= (PAGE_SIZE - offset); + if (dma_len) { + dma_addr += (PAGE_SIZE - offset); + } else { + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + if (length <= PAGE_SIZE) { + cmd->prp2 = cpu_to_le64(dma_addr); + return; + } + + /* XXX: support PRP lists */ +} + static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, struct bio *bio, enum dma_data_direction dma_dir, int psegs) { @@ -261,7 +291,7 @@ static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct bio *bio) { - struct nvme_rw_command *cmnd; + struct nvme_command *cmnd; struct nvme_req_info *info; enum dma_data_direction dma_dir; int cmdid; @@ -290,27 +320,26 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH; spin_lock_irqsave(&nvmeq->q_lock, flags); - cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail].rw; + cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail]; if (bio_data_dir(bio)) { - cmnd->opcode = nvme_cmd_write; + cmnd->rw.opcode = nvme_cmd_write; dma_dir = DMA_TO_DEVICE; } else { - cmnd->opcode = nvme_cmd_read; + cmnd->rw.opcode = nvme_cmd_read; dma_dir = DMA_FROM_DEVICE; } nvme_map_bio(nvmeq->q_dmadev, info, bio, dma_dir, psegs); - cmnd->flags = 1; - cmnd->command_id = cmdid; - cmnd->nsid = cpu_to_le32(ns->ns_id); - cmnd->prp1 = cpu_to_le64(sg_phys(info->sg)); - /* XXX: Support more than one PRP */ - cmnd->slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); - cmnd->length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); - cmnd->control = cpu_to_le16(control); - cmnd->dsmgmt = cpu_to_le32(dsmgmt); + cmnd->rw.flags = 1; + cmnd->rw.command_id = cmdid; + cmnd->rw.nsid = cpu_to_le32(ns->ns_id); + nvme_setup_prps(&cmnd->common, info->sg, bio->bi_size); + cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); + cmnd->rw.length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); + cmnd->rw.control = cpu_to_le16(control); + cmnd->rw.dsmgmt = cpu_to_le32(dsmgmt); writel(nvmeq->sq_tail, nvmeq->q_db); if (++nvmeq->sq_tail == nvmeq->q_depth) @@ -667,8 +696,9 @@ static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) goto put_pages; } sg_init_table(sg, count); - for (i = 0; i < count; i++) - sg_set_page(&sg[i], pages[i], PAGE_SIZE, 0); + sg_set_page(&sg[0], pages[0], PAGE_SIZE - offset, offset); + if (count > 1) + sg_set_page(&sg[1], pages[1], offset, 0); nents = dma_map_sg(&dev->pci_dev->dev, sg, count, DMA_FROM_DEVICE); if (!nents) goto put_pages; @@ -676,15 +706,7 @@ static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) memset(&c, 0, sizeof(c)); c.identify.opcode = nvme_admin_identify; c.identify.nsid = cns ? 0 : cpu_to_le32(ns->ns_id); - c.identify.prp1 = cpu_to_le64(sg_dma_address(&sg[0]) + offset); - if (count > 1) { - u64 dma_addr; - if (nents > 1) - dma_addr = sg_dma_address(&sg[1]); - else - dma_addr = sg_dma_address(&sg[0]) + PAGE_SIZE; - c.identify.prp2 = cpu_to_le64(dma_addr); - } + nvme_setup_prps(&c.common, sg, 4096); c.identify.cns = cpu_to_le32(cns); err = nvme_submit_admin_cmd(dev, &c, NULL); From b8deb62cf271fa9381edc8cf52bcae2f0225c55a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 26 Jan 2011 10:08:25 -0500 Subject: [PATCH 020/820] NVMe: Zero the command before we send it Make sure there's no left-over bits set from previous commands that used this slot. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 1e57737b1760..25ca7af96469 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -322,6 +322,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, spin_lock_irqsave(&nvmeq->q_lock, flags); cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail]; + memset(cmnd, 0, sizeof(*cmnd)); if (bio_data_dir(bio)) { cmnd->rw.opcode = nvme_cmd_write; dma_dir = DMA_TO_DEVICE; From bd38c5557cf482fc195e2264b32ea62eed60730a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 26 Jan 2011 14:34:32 -0500 Subject: [PATCH 021/820] NVMe: Change NVME_IOCTL_GET_RANGE_TYPE to return all the ranges Factor out most of nvme_identify() into a new nvme_submit_user_admin_command() function. Change nvme_get_range_type() to call it and change nvme_ioctl to realise that it's getting back all 64 ranges. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 49 +++++++++++++++++--------------------------- 1 file changed, 19 insertions(+), 30 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 25ca7af96469..b28d188d10f8 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -677,18 +677,17 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) return result; } -static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) +static int nvme_submit_user_admin_command(struct nvme_dev *dev, unsigned long addr, + unsigned length, struct nvme_command *cmd) { - struct nvme_dev *dev = ns->dev; int i, err, count, nents, offset; - struct nvme_command c; struct scatterlist sg[2]; struct page *pages[2]; if (addr & 3) return -EINVAL; offset = offset_in_page(addr); - count = offset ? 2 : 1; + count = ((offset + length) > PAGE_SIZE) ? 2 : 1; err = get_user_pages_fast(addr, count, 1, pages); if (err < count) { @@ -704,13 +703,9 @@ static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) if (!nents) goto put_pages; - memset(&c, 0, sizeof(c)); - c.identify.opcode = nvme_admin_identify; - c.identify.nsid = cns ? 0 : cpu_to_le32(ns->ns_id); - nvme_setup_prps(&c.common, sg, 4096); - c.identify.cns = cpu_to_le32(cns); + nvme_setup_prps(&cmd->common, sg, length); - err = nvme_submit_admin_cmd(dev, &c, NULL); + err = nvme_submit_admin_cmd(dev, cmd, NULL); if (err) err = -EIO; @@ -723,34 +718,28 @@ static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) return err; } -static int nvme_get_range_type(struct nvme_ns *ns, void __user *addr) +static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) { - struct nvme_dev *dev = ns->dev; - int status; struct nvme_command c; - void *page; - dma_addr_t dma_addr; - page = dma_alloc_coherent(&dev->pci_dev->dev, 4096, &dma_addr, - GFP_KERNEL); + memset(&c, 0, sizeof(c)); + c.identify.opcode = nvme_admin_identify; + c.identify.nsid = cns ? 0 : cpu_to_le32(ns->ns_id); + c.identify.cns = cpu_to_le32(cns); + + return nvme_submit_user_admin_command(ns->dev, addr, 4096, &c); +} + +static int nvme_get_range_type(struct nvme_ns *ns, unsigned long addr) +{ + struct nvme_command c; memset(&c, 0, sizeof(c)); c.features.opcode = nvme_admin_get_features; c.features.nsid = cpu_to_le32(ns->ns_id); - c.features.prp1 = cpu_to_le64(dma_addr); c.features.fid = cpu_to_le32(NVME_FEAT_LBA_RANGE); - status = nvme_submit_admin_cmd(dev, &c, NULL); - - /* XXX: Assuming first range for now */ - if (status) - status = -EIO; - else if (copy_to_user(addr, page, 64)) - status = -EFAULT; - - dma_free_coherent(&dev->pci_dev->dev, 4096, page, dma_addr); - - return status; + return nvme_submit_user_admin_command(ns->dev, addr, 4096, &c); } static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, @@ -764,7 +753,7 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, case NVME_IOCTL_IDENTIFY_CTRL: return nvme_identify(ns, arg, 1); case NVME_IOCTL_GET_RANGE_TYPE: - return nvme_get_range_type(ns, (void __user *)arg); + return nvme_get_range_type(ns, arg); default: return -ENOTTY; } From 7fc3cdabba75c2516b8b645eb0ca7907aea70415 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 26 Jan 2011 17:05:50 -0500 Subject: [PATCH 022/820] NVMe: Create nvme_map_user_pages() and nvme_unmap_user_pages() These are generalisations of the code that was in nvme_submit_user_admin_command(). Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 70 ++++++++++++++++++++++++++++++++++---------- 1 file changed, 54 insertions(+), 16 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b28d188d10f8..f44d6cd87ea2 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -677,17 +677,22 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) return result; } -static int nvme_submit_user_admin_command(struct nvme_dev *dev, unsigned long addr, - unsigned length, struct nvme_command *cmd) +static int nvme_map_user_pages(struct nvme_dev *dev, int write, + unsigned long addr, unsigned length, + struct scatterlist **sgp) { int i, err, count, nents, offset; - struct scatterlist sg[2]; - struct page *pages[2]; + struct scatterlist *sg; + struct page **pages; if (addr & 3) return -EINVAL; + if (!length) + return -EINVAL; + offset = offset_in_page(addr); - count = ((offset + length) > PAGE_SIZE) ? 2 : 1; + count = DIV_ROUND_UP(offset + length, PAGE_SIZE); + pages = kcalloc(count, sizeof(*pages), GFP_KERNEL); err = get_user_pages_fast(addr, count, 1, pages); if (err < count) { @@ -695,29 +700,62 @@ static int nvme_submit_user_admin_command(struct nvme_dev *dev, unsigned long ad err = -EFAULT; goto put_pages; } + + sg = kcalloc(count, sizeof(*sg), GFP_KERNEL); sg_init_table(sg, count); sg_set_page(&sg[0], pages[0], PAGE_SIZE - offset, offset); - if (count > 1) - sg_set_page(&sg[1], pages[1], offset, 0); - nents = dma_map_sg(&dev->pci_dev->dev, sg, count, DMA_FROM_DEVICE); + length -= (PAGE_SIZE - offset); + for (i = 1; i < count; i++) { + sg_set_page(&sg[i], pages[i], min_t(int, length, PAGE_SIZE), 0); + length -= PAGE_SIZE; + } + + err = -ENOMEM; + nents = dma_map_sg(&dev->pci_dev->dev, sg, count, + write ? DMA_TO_DEVICE : DMA_FROM_DEVICE); if (!nents) goto put_pages; - nvme_setup_prps(&cmd->common, sg, length); + kfree(pages); + *sgp = sg; + return nents; - err = nvme_submit_admin_cmd(dev, cmd, NULL); - - if (err) - err = -EIO; - - dma_unmap_sg(&dev->pci_dev->dev, sg, nents, DMA_FROM_DEVICE); put_pages: for (i = 0; i < count; i++) put_page(pages[i]); - + kfree(pages); return err; } +static void nvme_unmap_user_pages(struct nvme_dev *dev, int write, + unsigned long addr, int length, + struct scatterlist *sg, int nents) +{ + int i, count; + + count = DIV_ROUND_UP(offset_in_page(addr) + length, PAGE_SIZE); + dma_unmap_sg(&dev->pci_dev->dev, sg, nents, DMA_FROM_DEVICE); + + for (i = 0; i < count; i++) + put_page(sg_page(&sg[i])); +} + +static int nvme_submit_user_admin_command(struct nvme_dev *dev, + unsigned long addr, unsigned length, + struct nvme_command *cmd) +{ + int err, nents; + struct scatterlist *sg; + + nents = nvme_map_user_pages(dev, 0, addr, length, &sg); + if (nents < 0) + return nents; + nvme_setup_prps(&cmd->common, sg, length); + err = nvme_submit_admin_cmd(dev, cmd, NULL); + nvme_unmap_user_pages(dev, 0, addr, length, sg, nents); + return err ? -EIO : 0; +} + static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) { struct nvme_command c; From a53295b6998f62d961c29e54051c1cf1d738c2b3 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 16:13:29 -0500 Subject: [PATCH 023/820] NVMe: Add NVME_IOCTL_SUBMIT_IO Allow userspace to submit synchronous I/O like the SCSI sg interface does. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 43 +++++++++++++++++++++++++++++++++++++++++++ include/linux/nvme.h | 18 ++++++++++++++++++ 2 files changed, 61 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index f44d6cd87ea2..40fb2e1bdfe4 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -780,6 +780,47 @@ static int nvme_get_range_type(struct nvme_ns *ns, unsigned long addr) return nvme_submit_user_admin_command(ns->dev, addr, 4096, &c); } +static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) +{ + struct nvme_dev *dev = ns->dev; + struct nvme_queue *nvmeq; + struct nvme_user_io io; + struct nvme_command c; + unsigned length; + u32 result; + int nents, status; + struct scatterlist *sg; + + if (copy_from_user(&io, uio, sizeof(io))) + return -EFAULT; + length = io.nblocks << io.block_shift; + nents = nvme_map_user_pages(dev, io.opcode & 1, io.addr, length, &sg); + if (nents < 0) + return nents; + + memset(&c, 0, sizeof(c)); + c.rw.opcode = io.opcode; + c.rw.flags = io.flags; + c.rw.nsid = cpu_to_le32(io.nsid); + c.rw.slba = cpu_to_le64(io.slba); + c.rw.length = cpu_to_le16(io.nblocks - 1); + c.rw.control = cpu_to_le16(io.control); + c.rw.dsmgmt = cpu_to_le16(io.dsmgmt); + c.rw.reftag = cpu_to_le32(io.reftag); /* XXX: endian? */ + c.rw.apptag = cpu_to_le16(io.apptag); + c.rw.appmask = cpu_to_le16(io.appmask); + /* XXX: metadata */ + nvme_setup_prps(&c.common, sg, length); + + nvmeq = get_nvmeq(ns); + status = nvme_submit_sync_cmd(nvmeq, &c, &result); + put_nvmeq(nvmeq); + + nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); + put_user(result, &uio->result); + return status; +} + static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { @@ -792,6 +833,8 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, return nvme_identify(ns, arg, 1); case NVME_IOCTL_GET_RANGE_TYPE: return nvme_get_range_type(ns, arg); + case NVME_IOCTL_SUBMIT_IO: + return nvme_submit_io(ns, (void __user *)arg); default: return -ENOTTY; } diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 1c0b5ef08959..0aaecb059d14 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -340,8 +340,26 @@ struct nvme_completion { __le16 status; /* did the command fail, and if so, why? */ }; +struct nvme_user_io { + __u8 opcode; + __u8 flags; + __u16 control; + __u32 nsid; + __u64 metadata; + __u64 addr; + __u64 slba; + __u16 nblocks; + __u16 block_shift; + __u32 dsmgmt; + __u32 reftag; + __u16 apptag; + __u16 appmask; + __u32 result; +}; + #define NVME_IOCTL_IDENTIFY_NS _IOW('N', 0x40, struct nvme_id_ns) #define NVME_IOCTL_IDENTIFY_CTRL _IOW('N', 0x41, struct nvme_id_ctrl) #define NVME_IOCTL_GET_RANGE_TYPE _IOW('N', 0x42, struct nvme_lba_range_type) +#define NVME_IOCTL_SUBMIT_IO _IOWR('N', 0x43, struct nvme_rw_command) #endif /* _LINUX_NVME_H */ From 51814232ecae90f888c902e252306df8d017f0dd Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 16:18:08 -0500 Subject: [PATCH 024/820] NVMe: Read the model, serial & firmware rev from the controller Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 40fb2e1bdfe4..12e37c1cf057 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -58,6 +58,9 @@ struct nvme_dev { struct msix_entry *entry; struct nvme_bar __iomem *bar; struct list_head namespaces; + char serial[20]; + char model[40]; + char firmware_rev[8]; }; /* @@ -979,6 +982,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) { int res, nn, i; struct nvme_ns *ns, *next; + struct nvme_id_ctrl *ctrl; void *id; dma_addr_t dma_addr; struct nvme_command cid, crt; @@ -1003,7 +1007,11 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) goto out_free; } - nn = le32_to_cpup(&((struct nvme_id_ctrl *)id)->nn); + ctrl = id; + nn = le32_to_cpup(&ctrl->nn); + memcpy(dev->serial, ctrl->sn, sizeof(ctrl->sn)); + memcpy(dev->model, ctrl->mn, sizeof(ctrl->mn)); + memcpy(dev->firmware_rev, ctrl->fr, sizeof(ctrl->fr)); cid.identify.cns = 0; memset(&crt, 0, sizeof(crt)); From 8e9f0e71150bf6277d0ea40bc8feb1338ddf13fd Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 31 Jan 2011 10:46:14 -0500 Subject: [PATCH 025/820] NVMe: Remove 'node' from nvme_dev We don't keep a list of nvme_dev any more Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 12e37c1cf057..9377cf32f813 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -48,7 +48,6 @@ module_param(nvme_major, int, 0); * Represents an NVM Express device. Each nvme_dev is a PCI function. */ struct nvme_dev { - struct list_head node; struct nvme_queue **queues; u32 __iomem *dbs; struct pci_dev *pci_dev; From 3f85d50b609e8a5ef151656210203a6e94c19538 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 08:39:04 -0500 Subject: [PATCH 026/820] NVMe: Check returns from nvme_alloc_queue() It can return NULL, so handle that. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 9377cf32f813..dc821776be94 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -619,6 +619,9 @@ static __devinit struct nvme_queue *nvme_create_queue(struct nvme_dev *dev, int result; struct nvme_queue *nvmeq = nvme_alloc_queue(dev, qid, cq_size, vector); + if (!nvmeq) + return NULL; + result = adapter_alloc_cq(dev, qid, nvmeq); if (result < 0) goto free_nvmeq; @@ -655,6 +658,8 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) dev->dbs = ((void __iomem *)dev->bar) + 4096; nvmeq = nvme_alloc_queue(dev, 0, 64, 0); + if (!nvmeq) + return -ENOMEM; aqa = nvmeq->q_depth - 1; aqa |= aqa << 16; From 0ee5a7d7cb9309bd393a25c395f19fb12a842602 Mon Sep 17 00:00:00 2001 From: Shane Michael Matthews Date: Tue, 1 Feb 2011 08:49:30 -0500 Subject: [PATCH 027/820] NVMe: Enable and disable the PCI device Call pci_enable_device_mem() at initialisation and pci_disable_device at exit. Signed-off-by: Shane Michael Matthews Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index dc821776be94..1dda4b5c2302 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1103,6 +1103,9 @@ static int __devinit nvme_probe(struct pci_dev *pdev, if (!dev->queues) goto free; + if (pci_enable_device_mem(pdev)) + goto free; + INIT_LIST_HEAD(&dev->namespaces); dev->pci_dev = pdev; pci_set_drvdata(pdev, dev); @@ -1133,6 +1136,7 @@ static int __devinit nvme_probe(struct pci_dev *pdev, disable: pci_disable_msix(pdev); nvme_release_instance(dev); + pci_disable_device(pdev); free: kfree(dev->queues); kfree(dev->entry); @@ -1147,6 +1151,7 @@ static void __devexit nvme_remove(struct pci_dev *pdev) pci_disable_msix(pdev); iounmap(dev->bar); nvme_release_instance(dev); + pci_disable_device(pdev); kfree(dev->queues); kfree(dev->entry); kfree(dev); From f64d3365a3e5cb46e69db7e2c82a7cb9a5bed1b8 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 09:01:59 -0500 Subject: [PATCH 028/820] NVMe: Enable device DMA Need to call pci_set_master() to enable device DMA Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 1dda4b5c2302..128fd70031a9 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1105,6 +1105,7 @@ static int __devinit nvme_probe(struct pci_dev *pdev, if (pci_enable_device_mem(pdev)) goto free; + pci_set_master(pdev); INIT_LIST_HEAD(&dev->namespaces); dev->pci_dev = pdev; From 2930353f9f2b9e4629e935acd970cb73c1171229 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 16:23:39 -0500 Subject: [PATCH 029/820] NVMe: Allow queues to be allocated above 4GB Need to call dma_set_coherent_mask() to allow queues to be allocated above 4GB. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 128fd70031a9..46f872021369 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1110,7 +1110,8 @@ static int __devinit nvme_probe(struct pci_dev *pdev, INIT_LIST_HEAD(&dev->namespaces); dev->pci_dev = pdev; pci_set_drvdata(pdev, dev); - dma_set_mask(&dev->pci_dev->dev, DMA_BIT_MASK(64)); + dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)); + dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64)); nvme_set_instance(dev); dev->entry[0].vector = pdev->irq; From 574e8b95bc3780e10e9b5e9d51074d503dd3d5d9 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 16:24:35 -0500 Subject: [PATCH 030/820] NVMe: Request I/O regions Calling pci_request_selected_regions() reserves these regions for our use. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 46f872021369..bda91178f475 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1088,7 +1088,7 @@ static void nvme_release_instance(struct nvme_dev *dev) static int __devinit nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) { - int result = -ENOMEM; + int bars, result = -ENOMEM; struct nvme_dev *dev; dev = kzalloc(sizeof(*dev), GFP_KERNEL); @@ -1106,6 +1106,9 @@ static int __devinit nvme_probe(struct pci_dev *pdev, if (pci_enable_device_mem(pdev)) goto free; pci_set_master(pdev); + bars = pci_select_bars(pdev, IORESOURCE_MEM); + if (pci_request_selected_regions(pdev, bars, "nvme")) + goto disable; INIT_LIST_HEAD(&dev->namespaces); dev->pci_dev = pdev; @@ -1118,7 +1121,7 @@ static int __devinit nvme_probe(struct pci_dev *pdev, dev->bar = ioremap(pci_resource_start(pdev, 0), 8192); if (!dev->bar) { result = -ENOMEM; - goto disable; + goto disable_msix; } result = nvme_configure_admin_queue(dev); @@ -1135,10 +1138,12 @@ static int __devinit nvme_probe(struct pci_dev *pdev, nvme_free_queues(dev); unmap: iounmap(dev->bar); - disable: + disable_msix: pci_disable_msix(pdev); nvme_release_instance(dev); + disable: pci_disable_device(pdev); + pci_release_regions(pdev); free: kfree(dev->queues); kfree(dev->entry); @@ -1154,6 +1159,7 @@ static void __devexit nvme_remove(struct pci_dev *pdev) iounmap(dev->bar); nvme_release_instance(dev); pci_disable_device(pdev); + pci_release_regions(pdev); kfree(dev->queues); kfree(dev->entry); kfree(dev); From 5911f20039ce59d7e7834f0c42151cf759b6f786 Mon Sep 17 00:00:00 2001 From: Shane Michael Matthews Date: Tue, 1 Feb 2011 11:31:55 -0500 Subject: [PATCH 031/820] NVMe: Disable the device before we write the admin queues In case the card has been left in a partially-configured state, write 0 to the Enable bit. Signed-off-by: Shane Michael Matthews Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index bda91178f475..e3d921577b94 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -668,6 +668,7 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) dev->ctrl_config |= (PAGE_SHIFT - 12) << NVME_CC_MPS_SHIFT; dev->ctrl_config |= NVME_CC_ARB_RR | NVME_CC_SHN_NONE; + writel(0, &dev->bar->cc); writel(aqa, &dev->bar->aqa); writeq(nvmeq->sq_dma_addr, &dev->bar->asq); writeq(nvmeq->cq_dma_addr, &dev->bar->acq); From 388f037f4e7f0a24bac6b1a24f144f5d939f58cf Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 1 Feb 2011 12:49:38 -0500 Subject: [PATCH 032/820] NVMe: Move sysfs entries to the right place Because I wasn't setting driverfs_dev, the devices were showing up under /sys/devices/virtual/block. Now they appear underneath the PCI device which they belong to. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index e3d921577b94..744db3877c42 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -889,6 +889,7 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int index, disk->fops = &nvme_fops; disk->private_data = ns; disk->queue = ns->queue; + disk->driverfs_dev = &dev->pci_dev->dev; sprintf(disk->disk_name, "nvme%dn%d", dev->instance, index); set_capacity(disk, le64_to_cpup(&id->nsze) << (ns->lba_shift - 9)); From 7a63e07b9a98b77dd075e06b93c1d8dc871ddad5 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 3 Feb 2011 09:20:57 -0500 Subject: [PATCH 033/820] NVMe: Add remaining status codes Signed-off-by: Matthew Wilcox --- include/linux/nvme.h | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 0aaecb059d14..dbbdc126401b 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -321,14 +321,29 @@ enum { NVME_SC_ABORT_QUEUE = 0x8, NVME_SC_FUSED_FAIL = 0x9, NVME_SC_FUSED_MISSING = 0xa, + NVME_SC_INVALID_NS = 0xb, NVME_SC_LBA_RANGE = 0x80, NVME_SC_CAP_EXCEEDED = 0x81, NVME_SC_NS_NOT_READY = 0x82, NVME_SC_CQ_INVALID = 0x100, NVME_SC_QID_INVALID = 0x101, NVME_SC_QUEUE_SIZE = 0x102, + NVME_SC_ABORT_LIMIT = 0x103, + NVME_SC_ABORT_MISSING = 0x104, + NVME_SC_ASYNC_LIMIT = 0x105, + NVME_SC_FIRMWARE_SLOT = 0x106, + NVME_SC_FIRMWARE_IMAGE = 0x107, + NVME_SC_INVALID_VECTOR = 0x108, + NVME_SC_INVALID_LOG_PAGE = 0x109, + NVME_SC_INVALID_FORMAT = 0x10a, + NVME_SC_BAD_ATTRIBUTES = 0x180, NVME_SC_WRITE_FAULT = 0x280, NVME_SC_READ_ERROR = 0x281, + NVME_SC_GUARD_CHECK = 0x282, + NVME_SC_APPTAG_CHECK = 0x283, + NVME_SC_REFTAG_CHECK = 0x284, + NVME_SC_COMPARE_FAILED = 0x285, + NVME_SC_ACCESS_DENIED = 0x286, }; struct nvme_completion { From 6ee44cdced04a53dc4f27eb97067e6cd33784726 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 3 Feb 2011 10:58:26 -0500 Subject: [PATCH 034/820] NVMe: Add download / activate firmware ioctls Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 45 ++++++++++++++++++++++++++++++++++++++++++++ include/linux/nvme.h | 33 ++++++++++++++++++++++++++------ 2 files changed, 72 insertions(+), 6 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 744db3877c42..7cdf7f69cdcd 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -829,6 +829,47 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) return status; } +static int nvme_download_firmware(struct nvme_ns *ns, + struct nvme_dlfw __user *udlfw) +{ + struct nvme_dev *dev = ns->dev; + struct nvme_dlfw dlfw; + struct nvme_command c; + int nents, status; + struct scatterlist *sg; + + if (copy_from_user(&dlfw, udlfw, sizeof(dlfw))) + return -EFAULT; + if (dlfw.length >= (1 << 30)) + return -EINVAL; + + nents = nvme_map_user_pages(dev, 1, dlfw.addr, dlfw.length * 4, &sg); + if (nents < 0) + return nents; + + memset(&c, 0, sizeof(c)); + c.dlfw.opcode = nvme_admin_download_fw; + c.dlfw.numd = cpu_to_le32(dlfw.length); + c.dlfw.offset = cpu_to_le32(dlfw.offset); + nvme_setup_prps(&c.common, sg, dlfw.length * 4); + + status = nvme_submit_admin_cmd(dev, &c, NULL); + nvme_unmap_user_pages(dev, 0, dlfw.addr, dlfw.length * 4, sg, nents); + return status; +} + +static int nvme_activate_firmware(struct nvme_ns *ns, unsigned long arg) +{ + struct nvme_dev *dev = ns->dev; + struct nvme_command c; + + memset(&c, 0, sizeof(c)); + c.common.opcode = nvme_admin_activate_fw; + c.common.rsvd10[0] = cpu_to_le32(arg); + + return nvme_submit_admin_cmd(dev, &c, NULL); +} + static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { @@ -843,6 +884,10 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, return nvme_get_range_type(ns, arg); case NVME_IOCTL_SUBMIT_IO: return nvme_submit_io(ns, (void __user *)arg); + case NVME_IOCTL_DOWNLOAD_FW: + return nvme_download_firmware(ns, (void __user *)arg); + case NVME_IOCTL_ACTIVATE_FW: + return nvme_activate_firmware(ns, arg); default: return -ENOTTY; } diff --git a/include/linux/nvme.h b/include/linux/nvme.h index dbbdc126401b..8eed0e432eef 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -262,7 +262,7 @@ struct nvme_create_cq { __u8 opcode; __u8 flags; __u16 command_id; - __le32 rsvd1[5]; + __u32 rsvd1[5]; __le64 prp1; __u64 rsvd8; __le16 cqid; @@ -276,14 +276,14 @@ struct nvme_create_sq { __u8 opcode; __u8 flags; __u16 command_id; - __le32 rsvd1[5]; + __u32 rsvd1[5]; __le64 prp1; __u64 rsvd8; __le16 sqid; __le16 qsize; __le16 sq_flags; __le16 cqid; - __le32 rsvd12[4]; + __u32 rsvd12[4]; }; struct nvme_delete_queue { @@ -292,8 +292,20 @@ struct nvme_delete_queue { __u16 command_id; __u32 rsvd1[9]; __le16 qid; - __le16 rsvd10; - __le32 rsvd11[5]; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct nvme_download_firmware { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __le64 prp2; + __le32 numd; + __le32 offset; + __u32 rsvd12[4]; }; struct nvme_command { @@ -305,6 +317,7 @@ struct nvme_command { struct nvme_create_cq create_cq; struct nvme_create_sq create_sq; struct nvme_delete_queue delete_queue; + struct nvme_download_firmware dlfw; }; }; @@ -348,7 +361,7 @@ enum { struct nvme_completion { __le32 result; /* Used by admin commands to return data */ - __le32 rsvd; + __u32 rsvd; __le16 sq_head; /* how much of this queue may be reclaimed */ __le16 sq_id; /* submission queue that generated this entry */ __u16 command_id; /* of the command which completed */ @@ -372,9 +385,17 @@ struct nvme_user_io { __u32 result; }; +struct nvme_dlfw { + __u64 addr; + __u32 length; /* In dwords */ + __u32 offset; /* In dwords */ +}; + #define NVME_IOCTL_IDENTIFY_NS _IOW('N', 0x40, struct nvme_id_ns) #define NVME_IOCTL_IDENTIFY_CTRL _IOW('N', 0x41, struct nvme_id_ctrl) #define NVME_IOCTL_GET_RANGE_TYPE _IOW('N', 0x42, struct nvme_lba_range_type) #define NVME_IOCTL_SUBMIT_IO _IOWR('N', 0x43, struct nvme_rw_command) +#define NVME_IOCTL_DOWNLOAD_FW _IOR('N', 0x44, struct nvme_dlfw) +#define NVME_IOCTL_ACTIVATE_FW _IO('N', 0x45) #endif /* _LINUX_NVME_H */ From db5d0c198d673b6a932b449d4db95a2ad50c755e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 3 Feb 2011 14:36:07 -0500 Subject: [PATCH 035/820] NVMe: Release 0.2 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 7cdf7f69cdcd..06a6aeaa827a 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1272,6 +1272,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.1"); +MODULE_VERSION("0.2"); module_init(nvme_init); module_exit(nvme_exit); From 3c0cf138d7789feb3f335f6f1d24ad8fc8b3a23f Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 4 Feb 2011 16:03:56 -0500 Subject: [PATCH 036/820] NVMe: Allow fatal signals to interrupt I/O If the user sends a fatal signal, sleeping in the TASK_KILLABLE state permits the task to be aborted. The only wrinkle is making sure that if/when the command completes later that it doesn't upset anything. Handle this by setting the data pointer to 0, and checking the value isn't NULL in the sync completion path. Eventually, bios can be cancelled through this path too. Note that the cmdid isn't freed to prevent reuse. We should also abort the command in the future, but this is a good start. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 33 +++++++++++++++++++++++++++------ 1 file changed, 27 insertions(+), 6 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 06a6aeaa827a..4bfed59f3629 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -155,7 +155,9 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, } /* If you need more than four handlers, you'll need to change how - * alloc_cmdid and nvme_process_cq work + * alloc_cmdid and nvme_process_cq work. Also, aborted commands take + * the sync_completion path (if they complete), so don't put anything + * else in slot zero. */ enum { sync_completion_id = 0, @@ -172,6 +174,11 @@ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) return data; } +static void clear_cmdid_data(struct nvme_queue *nvmeq, int cmdid) +{ + nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(nvmeq->q_depth)] = 0; +} + static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) { int qid, cpu = get_cpu(); @@ -386,6 +393,8 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, struct nvme_completion *cqe) { struct sync_cmd_info *cmdinfo = ctx; + if (!cmdinfo) + return; /* Command aborted */ cmdinfo->result = le32_to_cpup(&cqe->result); cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; wake_up_process(cmdinfo->task); @@ -446,12 +455,19 @@ static irqreturn_t nvme_irq(int irq, void *data) return nvme_process_cq(data); } +static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid) +{ + spin_lock_irq(&nvmeq->q_lock); + clear_cmdid_data(nvmeq, cmdid); + spin_unlock_irq(&nvmeq->q_lock); +} + /* * Returns 0 on success. If the result is negative, it's a Linux error code; * if the result is positive, it's an NVM Express status code */ -static int nvme_submit_sync_cmd(struct nvme_queue *q, struct nvme_command *cmd, - u32 *result) +static int nvme_submit_sync_cmd(struct nvme_queue *nvmeq, + struct nvme_command *cmd, u32 *result) { int cmdid; struct sync_cmd_info cmdinfo; @@ -459,15 +475,20 @@ static int nvme_submit_sync_cmd(struct nvme_queue *q, struct nvme_command *cmd, cmdinfo.task = current; cmdinfo.status = -EINTR; - cmdid = alloc_cmdid_killable(q, &cmdinfo, sync_completion_id); + cmdid = alloc_cmdid_killable(nvmeq, &cmdinfo, sync_completion_id); if (cmdid < 0) return cmdid; cmd->common.command_id = cmdid; - set_current_state(TASK_UNINTERRUPTIBLE); - nvme_submit_cmd(q, cmd); + set_current_state(TASK_KILLABLE); + nvme_submit_cmd(nvmeq, cmd); schedule(); + if (cmdinfo.status == -EINTR) { + nvme_abort_command(nvmeq, cmdid); + return -EINTR; + } + if (result) *result = cmdinfo.result; From b1ad37efcafe396ac3944853589688dd0ec3c64e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 4 Feb 2011 16:14:30 -0500 Subject: [PATCH 037/820] NVMe: Call put_nvmeq() before calling nvme_submit_sync_cmd() We can't have preemption disabled when we call schedule(). Accept the possibility that we'll get preempted, and it'll cost us some cacheline bounces. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 4bfed59f3629..1c3cd6cc0ad9 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -842,8 +842,13 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) nvme_setup_prps(&c.common, sg, length); nvmeq = get_nvmeq(ns); - status = nvme_submit_sync_cmd(nvmeq, &c, &result); + /* Since nvme_submit_sync_cmd sleeps, we can't keep preemption + * disabled. We may be preempted at any point, and be rescheduled + * to a different CPU. That will cause cacheline bouncing, but no + * additional races since q_lock already protects against other CPUs. + */ put_nvmeq(nvmeq); + status = nvme_submit_sync_cmd(nvmeq, &c, &result); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); put_user(result, &uio->result); From 58ffacb545f76fc2c65d1fbfa5acf5184a2a09e6 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 6 Feb 2011 07:28:06 -0500 Subject: [PATCH 038/820] NVMe: Add a module parameter to use a threaded interrupt We're currently calling bio_endio from hard interrupt context. This is not a good idea for preemptible kernels as it will cause longer latencies. Using a threaded interrupt will run the entire queue processing mechanism (including bio_endio) in a thread, which can be preempted. Unfortuantely, it also adds about 7us of latency to the single-I/O case, so make it a module parameter for the moment. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 27 +++++++++++++++++++++++++++ 1 file changed, 27 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 1c3cd6cc0ad9..60c3786bc787 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -44,6 +44,9 @@ static int nvme_major; module_param(nvme_major, int, 0); +static int use_threaded_interrupts; +module_param(use_threaded_interrupts, int, 0); + /* * Represents an NVM Express device. Each nvme_dev is a PCI function. */ @@ -455,6 +458,25 @@ static irqreturn_t nvme_irq(int irq, void *data) return nvme_process_cq(data); } +static irqreturn_t nvme_irq_thread(int irq, void *data) +{ + irqreturn_t result; + struct nvme_queue *nvmeq = data; + spin_lock(&nvmeq->q_lock); + result = nvme_process_cq(nvmeq); + spin_unlock(&nvmeq->q_lock); + return result; +} + +static irqreturn_t nvme_irq_check(int irq, void *data) +{ + struct nvme_queue *nvmeq = data; + struct nvme_completion cqe = nvmeq->cqes[nvmeq->cq_head]; + if ((le16_to_cpu(cqe.status) & 1) != nvmeq->cq_phase) + return IRQ_NONE; + return IRQ_WAKE_THREAD; +} + static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid) { spin_lock_irq(&nvmeq->q_lock); @@ -630,6 +652,11 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, static int queue_request_irq(struct nvme_dev *dev, struct nvme_queue *nvmeq, const char *name) { + if (use_threaded_interrupts) + return request_threaded_irq(dev->entry[nvmeq->cq_vector].vector, + nvme_irq_check, nvme_irq_thread, + IRQF_DISABLED | IRQF_SHARED, + name, nvmeq); return request_irq(dev->entry[nvmeq->cq_vector].vector, nvme_irq, IRQF_DISABLED | IRQF_SHARED, name, nvmeq); } From be7b62754e097adc0cb16c25c9ee86ee20de62fb Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 6 Feb 2011 07:53:23 -0500 Subject: [PATCH 039/820] NVMe: Use a symbolic name to represent cancelled commands instead of 0 I have plans for other special values in sync_completion. Plus, this is more self-documenting, and lets us detect bogus usages. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 60c3786bc787..802d763d9d06 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include @@ -158,15 +159,17 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, } /* If you need more than four handlers, you'll need to change how - * alloc_cmdid and nvme_process_cq work. Also, aborted commands take - * the sync_completion path (if they complete), so don't put anything - * else in slot zero. + * alloc_cmdid and nvme_process_cq work. Consider using a special + * CMD_CTX value instead, if that works for your situation. */ enum { sync_completion_id = 0, bio_completion_id, }; +#define CMD_CTX_BASE (POISON_POINTER_DELTA + sync_completion_id) +#define CMD_CTX_CANCELLED (0x2008 + CMD_CTX_BASE) + static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { unsigned long data; @@ -177,9 +180,10 @@ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) return data; } -static void clear_cmdid_data(struct nvme_queue *nvmeq, int cmdid) +static void cancel_cmdid_data(struct nvme_queue *nvmeq, int cmdid) { - nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(nvmeq->q_depth)] = 0; + nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(nvmeq->q_depth)] = + CMD_CTX_CANCELLED; } static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) @@ -396,8 +400,8 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, struct nvme_completion *cqe) { struct sync_cmd_info *cmdinfo = ctx; - if (!cmdinfo) - return; /* Command aborted */ + if ((unsigned long)cmdinfo == CMD_CTX_CANCELLED) + return; cmdinfo->result = le32_to_cpup(&cqe->result); cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; wake_up_process(cmdinfo->task); @@ -480,7 +484,7 @@ static irqreturn_t nvme_irq_check(int irq, void *data) static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid) { spin_lock_irq(&nvmeq->q_lock); - clear_cmdid_data(nvmeq, cmdid); + cancel_cmdid_data(nvmeq, cmdid); spin_unlock_irq(&nvmeq->q_lock); } From b36235df01ec4141b4e589571d6789076c346d88 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 6 Feb 2011 08:49:55 -0500 Subject: [PATCH 040/820] NVMe: Detect commands that are completed twice Set the context value to CMD_CTX_COMPLETED, and print a message in the sync_completion handler if we see it. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 802d763d9d06..2dd09e7e142d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -169,12 +169,15 @@ enum { #define CMD_CTX_BASE (POISON_POINTER_DELTA + sync_completion_id) #define CMD_CTX_CANCELLED (0x2008 + CMD_CTX_BASE) +#define CMD_CTX_COMPLETED (0x2010 + CMD_CTX_BASE) static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { unsigned long data; + unsigned offset = cmdid + BITS_TO_LONGS(nvmeq->q_depth); - data = nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(nvmeq->q_depth)]; + data = nvmeq->cmdid_data[offset]; + nvmeq->cmdid_data[offset] = CMD_CTX_COMPLETED; clear_bit(cmdid, nvmeq->cmdid_data); wake_up(&nvmeq->sq_full); return data; @@ -182,8 +185,8 @@ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) static void cancel_cmdid_data(struct nvme_queue *nvmeq, int cmdid) { - nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(nvmeq->q_depth)] = - CMD_CTX_CANCELLED; + unsigned offset = cmdid + BITS_TO_LONGS(nvmeq->q_depth); + nvmeq->cmdid_data[offset] = CMD_CTX_CANCELLED; } static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) @@ -402,6 +405,12 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, struct sync_cmd_info *cmdinfo = ctx; if ((unsigned long)cmdinfo == CMD_CTX_CANCELLED) return; + if (unlikely((unsigned long)cmdinfo == CMD_CTX_COMPLETED)) { + dev_warn(nvmeq->q_dmadev, + "completed id %d twice on queue %d\n", + cqe->command_id, le16_to_cpup(&cqe->sq_id)); + return; + } cmdinfo->result = le32_to_cpup(&cqe->result); cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; wake_up_process(cmdinfo->task); From 48e3d39816416b3bf03dee3a796c0c04427c1a31 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 6 Feb 2011 08:51:15 -0500 Subject: [PATCH 041/820] NVMe: Detect command IDs completing that are out of range If the adapter completes a command ID that is outside the bounds of the array, return CMD_CTX_INVALID instead of random data, and print a message in the sync_completion handler (which is rapidly becoming the misc completion handler :-) Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 2dd09e7e142d..f4085d4fe0f2 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -170,12 +170,15 @@ enum { #define CMD_CTX_BASE (POISON_POINTER_DELTA + sync_completion_id) #define CMD_CTX_CANCELLED (0x2008 + CMD_CTX_BASE) #define CMD_CTX_COMPLETED (0x2010 + CMD_CTX_BASE) +#define CMD_CTX_INVALID (0x2014 + CMD_CTX_BASE) static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { unsigned long data; unsigned offset = cmdid + BITS_TO_LONGS(nvmeq->q_depth); + if (cmdid > nvmeq->q_depth) + return CMD_CTX_INVALID; data = nvmeq->cmdid_data[offset]; nvmeq->cmdid_data[offset] = CMD_CTX_COMPLETED; clear_bit(cmdid, nvmeq->cmdid_data); @@ -411,6 +414,12 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, cqe->command_id, le16_to_cpup(&cqe->sq_id)); return; } + if (unlikely((unsigned long)cmdinfo == CMD_CTX_INVALID)) { + dev_warn(nvmeq->q_dmadev, + "invalid id %d completed on queue %d\n", + cqe->command_id, le16_to_cpup(&cqe->sq_id)); + return; + } cmdinfo->result = le32_to_cpup(&cqe->result); cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; wake_up_process(cmdinfo->task); From ec6ce618d65b5ce1bef83a5509255107a0feac44 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 6 Feb 2011 09:01:00 -0500 Subject: [PATCH 042/820] NVMe: Need to lock queue during interrupt handling If we're sharing a queue between multiple CPUs and we cancel a sync I/O, we must have the queue locked to avoid corrupting the stack of the thread that submitted the I/O. It turns out this is the same locking that's needed for the threaded irq handler, so share that code. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index f4085d4fe0f2..139e6fc1e2a8 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -476,11 +476,6 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) } static irqreturn_t nvme_irq(int irq, void *data) -{ - return nvme_process_cq(data); -} - -static irqreturn_t nvme_irq_thread(int irq, void *data) { irqreturn_t result; struct nvme_queue *nvmeq = data; @@ -676,7 +671,7 @@ static int queue_request_irq(struct nvme_dev *dev, struct nvme_queue *nvmeq, { if (use_threaded_interrupts) return request_threaded_irq(dev->entry[nvmeq->cq_vector].vector, - nvme_irq_check, nvme_irq_thread, + nvme_irq_check, nvme_irq, IRQF_DISABLED | IRQF_SHARED, name, nvmeq); return request_irq(dev->entry[nvmeq->cq_vector].vector, nvme_irq, From e85248e516c550382ba33ca325c272a0ca397e44 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 6 Feb 2011 18:30:16 -0500 Subject: [PATCH 043/820] NVMe: Record the timeout for each command In addition to recording the completion data for each command, record the anticipated completion time. Choose a timeout of 5 seconds for normal I/Os and 60 seconds for admin I/Os. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 49 +++++++++++++++++++++++++++++--------------- 1 file changed, 32 insertions(+), 17 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 139e6fc1e2a8..60c1048dc8bc 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -41,6 +41,8 @@ #define SQ_SIZE(depth) (depth * sizeof(struct nvme_command)) #define CQ_SIZE(depth) (depth * sizeof(struct nvme_completion)) #define NVME_MINORS 64 +#define IO_TIMEOUT (5 * HZ) +#define ADMIN_TIMEOUT (60 * HZ) static int nvme_major; module_param(nvme_major, int, 0); @@ -119,6 +121,16 @@ static inline void _nvme_check_size(void) BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64); } +struct nvme_cmd_info { + unsigned long ctx; + unsigned long timeout; +}; + +static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) +{ + return (void *)&nvmeq->cmdid_data[BITS_TO_LONGS(nvmeq->q_depth)]; +} + /** * alloc_cmdid - Allocate a Command ID * @param nvmeq The queue that will be used for this command @@ -131,10 +143,11 @@ static inline void _nvme_check_size(void) * Passing in a pointer that's not 4-byte aligned will cause a BUG. * We can change this if it becomes a problem. */ -static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler) +static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler, + unsigned timeout) { int depth = nvmeq->q_depth; - unsigned long data = (unsigned long)ctx | handler; + struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); int cmdid; BUG_ON((unsigned long)ctx & 3); @@ -145,16 +158,17 @@ static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler) return -EBUSY; } while (test_and_set_bit(cmdid, nvmeq->cmdid_data)); - nvmeq->cmdid_data[cmdid + BITS_TO_LONGS(depth)] = data; + info[cmdid].ctx = (unsigned long)ctx | handler; + info[cmdid].timeout = jiffies + timeout; return cmdid; } static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, - int handler) + int handler, unsigned timeout) { int cmdid; wait_event_killable(nvmeq->sq_full, - (cmdid = alloc_cmdid(nvmeq, ctx, handler)) >= 0); + (cmdid = alloc_cmdid(nvmeq, ctx, handler, timeout)) >= 0); return (cmdid < 0) ? -EINTR : cmdid; } @@ -175,12 +189,12 @@ enum { static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { unsigned long data; - unsigned offset = cmdid + BITS_TO_LONGS(nvmeq->q_depth); + struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); - if (cmdid > nvmeq->q_depth) + if (cmdid >= nvmeq->q_depth) return CMD_CTX_INVALID; - data = nvmeq->cmdid_data[offset]; - nvmeq->cmdid_data[offset] = CMD_CTX_COMPLETED; + data = info[cmdid].ctx; + info[cmdid].ctx = CMD_CTX_COMPLETED; clear_bit(cmdid, nvmeq->cmdid_data); wake_up(&nvmeq->sq_full); return data; @@ -188,8 +202,8 @@ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) static void cancel_cmdid_data(struct nvme_queue *nvmeq, int cmdid) { - unsigned offset = cmdid + BITS_TO_LONGS(nvmeq->q_depth); - nvmeq->cmdid_data[offset] = CMD_CTX_CANCELLED; + struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); + info[cmdid].ctx = CMD_CTX_CANCELLED; } static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) @@ -327,7 +341,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, goto congestion; info->bio = bio; - cmdid = alloc_cmdid(nvmeq, info, bio_completion_id); + cmdid = alloc_cmdid(nvmeq, info, bio_completion_id, IO_TIMEOUT); if (unlikely(cmdid < 0)) goto free_info; @@ -506,7 +520,7 @@ static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid) * if the result is positive, it's an NVM Express status code */ static int nvme_submit_sync_cmd(struct nvme_queue *nvmeq, - struct nvme_command *cmd, u32 *result) + struct nvme_command *cmd, u32 *result, unsigned timeout) { int cmdid; struct sync_cmd_info cmdinfo; @@ -514,7 +528,8 @@ static int nvme_submit_sync_cmd(struct nvme_queue *nvmeq, cmdinfo.task = current; cmdinfo.status = -EINTR; - cmdid = alloc_cmdid_killable(nvmeq, &cmdinfo, sync_completion_id); + cmdid = alloc_cmdid_killable(nvmeq, &cmdinfo, sync_completion_id, + timeout); if (cmdid < 0) return cmdid; cmd->common.command_id = cmdid; @@ -537,7 +552,7 @@ static int nvme_submit_sync_cmd(struct nvme_queue *nvmeq, static int nvme_submit_admin_cmd(struct nvme_dev *dev, struct nvme_command *cmd, u32 *result) { - return nvme_submit_sync_cmd(dev->queues[0], cmd, result); + return nvme_submit_sync_cmd(dev->queues[0], cmd, result, ADMIN_TIMEOUT); } static int adapter_delete_queue(struct nvme_dev *dev, u8 opcode, u16 id) @@ -630,7 +645,7 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth, int vector) { struct device *dmadev = &dev->pci_dev->dev; - unsigned extra = (depth + BITS_TO_LONGS(depth)) * sizeof(long); + unsigned extra = (depth / 8) + (depth * sizeof(struct nvme_cmd_info)); struct nvme_queue *nvmeq = kzalloc(sizeof(*nvmeq) + extra, GFP_KERNEL); if (!nvmeq) return NULL; @@ -892,7 +907,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) * additional races since q_lock already protects against other CPUs. */ put_nvmeq(nvmeq); - status = nvme_submit_sync_cmd(nvmeq, &c, &result); + status = nvme_submit_sync_cmd(nvmeq, &c, &result, IO_TIMEOUT); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); put_user(result, &uio->result); From 9294bbed78926a895516ec016ba23033f58d1a88 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 7 Feb 2011 12:45:24 -0500 Subject: [PATCH 044/820] NVMe: Handle the congestion list a little better In the bio completion handler, check for bios on the congestion list for this NVM queue. Also, lock the congestion list in the make_request function as the queue may end up being shared between multiple CPUs. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 60c1048dc8bc..2a0dd5e60347 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -105,6 +105,8 @@ struct nvme_queue { unsigned long cmdid_data[]; }; +static void nvme_resubmit_bio(struct nvme_queue *nvmeq, struct bio *bio); + /* * Check we didin't inadvertently grow the command struct */ @@ -274,6 +276,9 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); free_info(info); bio_endio(bio, status ? -EIO : 0); + bio = bio_list_pop(&nvmeq->sq_cong); + if (bio) + nvme_resubmit_bio(nvmeq, bio); } /* length is in bytes */ @@ -392,6 +397,16 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return -EBUSY; } +static void nvme_resubmit_bio(struct nvme_queue *nvmeq, struct bio *bio) +{ + struct nvme_ns *ns = bio->bi_bdev->bd_disk->private_data; + if (nvme_submit_bio_queue(nvmeq, ns, bio)) + bio_list_add_head(&nvmeq->sq_cong, bio); + else if (bio_list_empty(&nvmeq->sq_cong)) + blk_clear_queue_congested(ns->queue, rw_is_sync(bio->bi_rw)); + /* XXX: Need to duplicate the logic from __freed_request here */ +} + /* * NB: return value of non-zero would mean that we were a stacking driver. * make_request must always succeed. @@ -403,7 +418,9 @@ static int nvme_make_request(struct request_queue *q, struct bio *bio) if (nvme_submit_bio_queue(nvmeq, ns, bio)) { blk_set_queue_congested(q, rw_is_sync(bio->bi_rw)); + spin_lock_irq(&nvmeq->q_lock); bio_list_add(&nvmeq->sq_cong, bio); + spin_unlock_irq(&nvmeq->q_lock); } put_nvmeq(nvmeq); From d2d8703481f60d67f49e3177196cbe474b11377c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 7 Feb 2011 15:55:59 -0500 Subject: [PATCH 045/820] NVMe: Renumber the special context values If POISON_POINTER_DELTA isn't defined, ensure they're in page 0 which should never be mapped. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 2a0dd5e60347..71bdf6f2c93b 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -184,9 +184,9 @@ enum { }; #define CMD_CTX_BASE (POISON_POINTER_DELTA + sync_completion_id) -#define CMD_CTX_CANCELLED (0x2008 + CMD_CTX_BASE) -#define CMD_CTX_COMPLETED (0x2010 + CMD_CTX_BASE) -#define CMD_CTX_INVALID (0x2014 + CMD_CTX_BASE) +#define CMD_CTX_CANCELLED (0x30C + CMD_CTX_BASE) +#define CMD_CTX_COMPLETED (0x310 + CMD_CTX_BASE) +#define CMD_CTX_INVALID (0x314 + CMD_CTX_BASE) static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { From 51882d00f07da9601cc962a3596e48aafb4f4163 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 08:49:59 -0500 Subject: [PATCH 046/820] NVMe: Advance the sg pointer when filling in an sg list For multipage BIOs, we were always using sg[0] instead of advancing through the list. Oops :-) Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 71bdf6f2c93b..903e7f15b60d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -321,6 +321,7 @@ static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, sg_init_table(sg, psegs); bio_for_each_segment(bvec, bio, i) { sg_set_page(sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); + sg++; /* XXX: handle non-mergable here */ nsegs++; } From e025344c56e08b155f43ea09647969286c78377c Mon Sep 17 00:00:00 2001 From: Shane Michael Matthews Date: Thu, 10 Feb 2011 08:51:24 -0500 Subject: [PATCH 047/820] NVMe: Initial PRP List support Add a pointer to the nvme_req_info to hold a new data structure (nvme_prps) which contains a list of the pages allocated to this particular request for holding PRP list entries. nvme_setup_prps() now returns this pointer. To allocate and free the memory used for PRP lists, we need a struct device, so we need to pass the nvme_queue pointer to many functions which didn't use to need it. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 106 +++++++++++++++++++++++++++++++++++++------ 1 file changed, 92 insertions(+), 14 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 903e7f15b60d..b1e8445985a2 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -247,21 +247,55 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) return 0; } +static __le64 *alloc_prp_list(struct nvme_queue *nvmeq, int length, + dma_addr_t *addr) +{ + return dma_alloc_coherent(nvmeq->q_dmadev, PAGE_SIZE, addr, GFP_ATOMIC); +} + +struct nvme_prps { + int npages; + dma_addr_t first_dma; + __le64 *list[0]; +}; + +static void nvme_free_prps(struct nvme_queue *nvmeq, struct nvme_prps *prps) +{ + const int last_prp = PAGE_SIZE / 8 - 1; + int i; + dma_addr_t prp_dma; + + if (!prps) + return; + + prp_dma = prps->first_dma; + for (i = 0; i < prps->npages; i++) { + __le64 *prp_list = prps->list[i]; + dma_addr_t next_prp_dma = le64_to_cpu(prp_list[last_prp]); + dma_free_coherent(nvmeq->q_dmadev, PAGE_SIZE, prp_list, + prp_dma); + prp_dma = next_prp_dma; + } + kfree(prps); +} + struct nvme_req_info { struct bio *bio; int nents; + struct nvme_prps *prps; struct scatterlist sg[0]; }; /* XXX: use a mempool */ static struct nvme_req_info *alloc_info(unsigned nseg, gfp_t gfp) { - return kmalloc(sizeof(struct nvme_req_info) + + return kzalloc(sizeof(struct nvme_req_info) + sizeof(struct scatterlist) * nseg, gfp); } -static void free_info(struct nvme_req_info *info) +static void free_info(struct nvme_queue *nvmeq, struct nvme_req_info *info) { + nvme_free_prps(nvmeq, info->prps); kfree(info); } @@ -274,7 +308,7 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, dma_unmap_sg(nvmeq->q_dmadev, info->sg, info->nents, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); - free_info(info); + free_info(nvmeq, info); bio_endio(bio, status ? -EIO : 0); bio = bio_list_pop(&nvmeq->sq_cong); if (bio) @@ -282,17 +316,22 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, } /* length is in bytes */ -static void nvme_setup_prps(struct nvme_common_command *cmd, +static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, + struct nvme_common_command *cmd, struct scatterlist *sg, int length) { int dma_len = sg_dma_len(sg); u64 dma_addr = sg_dma_address(sg); int offset = offset_in_page(dma_addr); + __le64 *prp_list; + dma_addr_t prp_dma; + int nprps, npages, i, prp_page; + struct nvme_prps *prps = NULL; cmd->prp1 = cpu_to_le64(dma_addr); length -= (PAGE_SIZE - offset); if (length <= 0) - return; + return prps; dma_len -= (PAGE_SIZE - offset); if (dma_len) { @@ -305,10 +344,42 @@ static void nvme_setup_prps(struct nvme_common_command *cmd, if (length <= PAGE_SIZE) { cmd->prp2 = cpu_to_le64(dma_addr); - return; + return prps; } - /* XXX: support PRP lists */ + nprps = DIV_ROUND_UP(length, PAGE_SIZE); + npages = DIV_ROUND_UP(8 * nprps, PAGE_SIZE); + prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, GFP_ATOMIC); + prps->npages = npages; + prp_page = 0; + prp_list = alloc_prp_list(nvmeq, length, &prp_dma); + prps->list[prp_page++] = prp_list; + prps->first_dma = prp_dma; + cmd->prp2 = cpu_to_le64(prp_dma); + i = 0; + for (;;) { + if (i == PAGE_SIZE / 8 - 1) { + __le64 *old_prp_list = prp_list; + prp_list = alloc_prp_list(nvmeq, length, &prp_dma); + prps->list[prp_page++] = prp_list; + old_prp_list[i] = cpu_to_le64(prp_dma); + i = 0; + } + prp_list[i++] = cpu_to_le64(dma_addr); + dma_len -= PAGE_SIZE; + dma_addr += PAGE_SIZE; + length -= PAGE_SIZE; + if (length <= 0) + break; + if (dma_len > 0) + continue; + BUG_ON(dma_len < 0); + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + return prps; } static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, @@ -378,7 +449,8 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, cmnd->rw.flags = 1; cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); - nvme_setup_prps(&cmnd->common, info->sg, bio->bi_size); + info->prps = nvme_setup_prps(nvmeq, &cmnd->common, info->sg, + bio->bi_size); cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); cmnd->rw.length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); cmnd->rw.control = cpu_to_le16(control); @@ -393,7 +465,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return 0; free_info: - free_info(info); + free_info(nvmeq, info); congestion: return -EBUSY; } @@ -852,13 +924,15 @@ static int nvme_submit_user_admin_command(struct nvme_dev *dev, { int err, nents; struct scatterlist *sg; + struct nvme_prps *prps; nents = nvme_map_user_pages(dev, 0, addr, length, &sg); if (nents < 0) return nents; - nvme_setup_prps(&cmd->common, sg, length); + prps = nvme_setup_prps(dev->queues[0], &cmd->common, sg, length); err = nvme_submit_admin_cmd(dev, cmd, NULL); nvme_unmap_user_pages(dev, 0, addr, length, sg, nents); + nvme_free_prps(dev->queues[0], prps); return err ? -EIO : 0; } @@ -896,6 +970,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) u32 result; int nents, status; struct scatterlist *sg; + struct nvme_prps *prps; if (copy_from_user(&io, uio, sizeof(io))) return -EFAULT; @@ -915,10 +990,10 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) c.rw.reftag = cpu_to_le32(io.reftag); /* XXX: endian? */ c.rw.apptag = cpu_to_le16(io.apptag); c.rw.appmask = cpu_to_le16(io.appmask); - /* XXX: metadata */ - nvme_setup_prps(&c.common, sg, length); - nvmeq = get_nvmeq(ns); + /* XXX: metadata */ + prps = nvme_setup_prps(nvmeq, &c.common, sg, length); + /* Since nvme_submit_sync_cmd sleeps, we can't keep preemption * disabled. We may be preempted at any point, and be rescheduled * to a different CPU. That will cause cacheline bouncing, but no @@ -928,6 +1003,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) status = nvme_submit_sync_cmd(nvmeq, &c, &result, IO_TIMEOUT); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); + nvme_free_prps(nvmeq, prps); put_user(result, &uio->result); return status; } @@ -940,6 +1016,7 @@ static int nvme_download_firmware(struct nvme_ns *ns, struct nvme_command c; int nents, status; struct scatterlist *sg; + struct nvme_prps *prps; if (copy_from_user(&dlfw, udlfw, sizeof(dlfw))) return -EFAULT; @@ -954,10 +1031,11 @@ static int nvme_download_firmware(struct nvme_ns *ns, c.dlfw.opcode = nvme_admin_download_fw; c.dlfw.numd = cpu_to_le32(dlfw.length); c.dlfw.offset = cpu_to_le32(dlfw.offset); - nvme_setup_prps(&c.common, sg, dlfw.length * 4); + prps = nvme_setup_prps(dev->queues[0], &c.common, sg, dlfw.length * 4); status = nvme_submit_admin_cmd(dev, &c, NULL); nvme_unmap_user_pages(dev, 0, dlfw.addr, dlfw.length * 4, sg, nents); + nvme_free_prps(dev->queues[0], prps); return status; } From d534df3c730af9073a9ddc076d9fd65cbdca22b3 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 09:03:06 -0500 Subject: [PATCH 048/820] NVMe: Rename nvme_req_info to nvme_bio There are too many things called 'info' in this driver. This data structure is auxiliary information for a struct bio, so call it nvme_bio, or nbio when used as a variable. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 48 ++++++++++++++++++++++---------------------- 1 file changed, 24 insertions(+), 24 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b1e8445985a2..11df0e90edad 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -279,7 +279,7 @@ static void nvme_free_prps(struct nvme_queue *nvmeq, struct nvme_prps *prps) kfree(prps); } -struct nvme_req_info { +struct nvme_bio { struct bio *bio; int nents; struct nvme_prps *prps; @@ -287,28 +287,28 @@ struct nvme_req_info { }; /* XXX: use a mempool */ -static struct nvme_req_info *alloc_info(unsigned nseg, gfp_t gfp) +static struct nvme_bio *alloc_nbio(unsigned nseg, gfp_t gfp) { - return kzalloc(sizeof(struct nvme_req_info) + + return kzalloc(sizeof(struct nvme_bio) + sizeof(struct scatterlist) * nseg, gfp); } -static void free_info(struct nvme_queue *nvmeq, struct nvme_req_info *info) +static void free_nbio(struct nvme_queue *nvmeq, struct nvme_bio *nbio) { - nvme_free_prps(nvmeq, info->prps); - kfree(info); + nvme_free_prps(nvmeq, nbio->prps); + kfree(nbio); } static void bio_completion(struct nvme_queue *nvmeq, void *ctx, struct nvme_completion *cqe) { - struct nvme_req_info *info = ctx; - struct bio *bio = info->bio; + struct nvme_bio *nbio = ctx; + struct bio *bio = nbio->bio; u16 status = le16_to_cpup(&cqe->status) >> 1; - dma_unmap_sg(nvmeq->q_dmadev, info->sg, info->nents, + dma_unmap_sg(nvmeq->q_dmadev, nbio->sg, nbio->nents, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); - free_info(nvmeq, info); + free_nbio(nvmeq, nbio); bio_endio(bio, status ? -EIO : 0); bio = bio_list_pop(&nvmeq->sq_cong); if (bio) @@ -382,11 +382,11 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, return prps; } -static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, +static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, struct bio *bio, enum dma_data_direction dma_dir, int psegs) { struct bio_vec *bvec; - struct scatterlist *sg = info->sg; + struct scatterlist *sg = nbio->sg; int i, nsegs; sg_init_table(sg, psegs); @@ -396,16 +396,16 @@ static int nvme_map_bio(struct device *dev, struct nvme_req_info *info, /* XXX: handle non-mergable here */ nsegs++; } - info->nents = nsegs; + nbio->nents = nsegs; - return dma_map_sg(dev, info->sg, info->nents, dma_dir); + return dma_map_sg(dev, nbio->sg, nbio->nents, dma_dir); } static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct bio *bio) { struct nvme_command *cmnd; - struct nvme_req_info *info; + struct nvme_bio *nbio; enum dma_data_direction dma_dir; int cmdid; u16 control; @@ -413,14 +413,14 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, unsigned long flags; int psegs = bio_phys_segments(ns->queue, bio); - info = alloc_info(psegs, GFP_NOIO); - if (!info) + nbio = alloc_nbio(psegs, GFP_NOIO); + if (!nbio) goto congestion; - info->bio = bio; + nbio->bio = bio; - cmdid = alloc_cmdid(nvmeq, info, bio_completion_id, IO_TIMEOUT); + cmdid = alloc_cmdid(nvmeq, nbio, bio_completion_id, IO_TIMEOUT); if (unlikely(cmdid < 0)) - goto free_info; + goto free_nbio; control = 0; if (bio->bi_rw & REQ_FUA) @@ -444,12 +444,12 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, dma_dir = DMA_FROM_DEVICE; } - nvme_map_bio(nvmeq->q_dmadev, info, bio, dma_dir, psegs); + nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs); cmnd->rw.flags = 1; cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); - info->prps = nvme_setup_prps(nvmeq, &cmnd->common, info->sg, + nbio->prps = nvme_setup_prps(nvmeq, &cmnd->common, nbio->sg, bio->bi_size); cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); cmnd->rw.length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); @@ -464,8 +464,8 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return 0; - free_info: - free_info(nvmeq, info); + free_nbio: + free_nbio(nvmeq, nbio); congestion: return -EBUSY; } From 091b609258b8e01cc45b01a41ca5e496f674d989 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 09:56:01 -0500 Subject: [PATCH 049/820] NVMe: Switch to use DMA Pool API Calling dma_free_coherent from interrupt context causes warnings. Using the DMA pools delays freeing until pool destruction, so avoids the problem. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 39 ++++++++++++++++++++++++++++++++------- 1 file changed, 32 insertions(+), 7 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 11df0e90edad..80fe6a7a8163 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -57,6 +57,7 @@ struct nvme_dev { struct nvme_queue **queues; u32 __iomem *dbs; struct pci_dev *pci_dev; + struct dma_pool *prp_page_pool; int instance; int queue_count; u32 ctrl_config; @@ -88,6 +89,7 @@ struct nvme_ns { */ struct nvme_queue { struct device *q_dmadev; + struct nvme_dev *dev; spinlock_t q_lock; struct nvme_command *sq_cmds; volatile struct nvme_completion *cqes; @@ -247,10 +249,9 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) return 0; } -static __le64 *alloc_prp_list(struct nvme_queue *nvmeq, int length, - dma_addr_t *addr) +static __le64 *alloc_prp_list(struct nvme_dev *dev, dma_addr_t *addr) { - return dma_alloc_coherent(nvmeq->q_dmadev, PAGE_SIZE, addr, GFP_ATOMIC); + return dma_pool_alloc(dev->prp_page_pool, GFP_ATOMIC, addr); } struct nvme_prps { @@ -262,6 +263,7 @@ struct nvme_prps { static void nvme_free_prps(struct nvme_queue *nvmeq, struct nvme_prps *prps) { const int last_prp = PAGE_SIZE / 8 - 1; + struct nvme_dev *dev = nvmeq->dev; int i; dma_addr_t prp_dma; @@ -272,8 +274,7 @@ static void nvme_free_prps(struct nvme_queue *nvmeq, struct nvme_prps *prps) for (i = 0; i < prps->npages; i++) { __le64 *prp_list = prps->list[i]; dma_addr_t next_prp_dma = le64_to_cpu(prp_list[last_prp]); - dma_free_coherent(nvmeq->q_dmadev, PAGE_SIZE, prp_list, - prp_dma); + dma_pool_free(dev->prp_page_pool, prp_list, prp_dma); prp_dma = next_prp_dma; } kfree(prps); @@ -320,6 +321,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, struct nvme_common_command *cmd, struct scatterlist *sg, int length) { + struct nvme_dev *dev = nvmeq->dev; int dma_len = sg_dma_len(sg); u64 dma_addr = sg_dma_address(sg); int offset = offset_in_page(dma_addr); @@ -352,7 +354,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, GFP_ATOMIC); prps->npages = npages; prp_page = 0; - prp_list = alloc_prp_list(nvmeq, length, &prp_dma); + prp_list = alloc_prp_list(dev, &prp_dma); prps->list[prp_page++] = prp_list; prps->first_dma = prp_dma; cmd->prp2 = cpu_to_le64(prp_dma); @@ -360,7 +362,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, for (;;) { if (i == PAGE_SIZE / 8 - 1) { __le64 *old_prp_list = prp_list; - prp_list = alloc_prp_list(nvmeq, length, &prp_dma); + prp_list = alloc_prp_list(dev, &prp_dma); prps->list[prp_page++] = prp_list; old_prp_list[i] = cpu_to_le64(prp_dma); i = 0; @@ -752,6 +754,7 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, goto free_cqdma; nvmeq->q_dmadev = dmadev; + nvmeq->dev = dev; spin_lock_init(&nvmeq->q_lock); nvmeq->cq_head = 0; nvmeq->cq_phase = 1; @@ -1302,6 +1305,22 @@ static int nvme_dev_remove(struct nvme_dev *dev) return 0; } +static int nvme_setup_prp_pools(struct nvme_dev *dev) +{ + struct device *dmadev = &dev->pci_dev->dev; + dev->prp_page_pool = dma_pool_create("prp list page", dmadev, + PAGE_SIZE, PAGE_SIZE, 0); + if (!dev->prp_page_pool) + return -ENOMEM; + + return 0; +} + +static void nvme_release_prp_pools(struct nvme_dev *dev) +{ + dma_pool_destroy(dev->prp_page_pool); +} + /* XXX: Use an ida or something to let remove / add work correctly */ static void nvme_set_instance(struct nvme_dev *dev) { @@ -1346,6 +1365,10 @@ static int __devinit nvme_probe(struct pci_dev *pdev, nvme_set_instance(dev); dev->entry[0].vector = pdev->irq; + result = nvme_setup_prp_pools(dev); + if (result) + goto disable_msix; + dev->bar = ioremap(pci_resource_start(pdev, 0), 8192); if (!dev->bar) { result = -ENOMEM; @@ -1369,6 +1392,7 @@ static int __devinit nvme_probe(struct pci_dev *pdev, disable_msix: pci_disable_msix(pdev); nvme_release_instance(dev); + nvme_release_prp_pools(dev); disable: pci_disable_device(pdev); pci_release_regions(pdev); @@ -1386,6 +1410,7 @@ static void __devexit nvme_remove(struct pci_dev *pdev) pci_disable_msix(pdev); iounmap(dev->bar); nvme_release_instance(dev); + nvme_release_prp_pools(dev); pci_disable_device(pdev); pci_release_regions(pdev); kfree(dev->queues); From 99802a7aee2b3dd720e382c52b892cc6a8122b11 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 10:30:34 -0500 Subject: [PATCH 050/820] NVMe: Optimise memory usage for I/Os between 4k and 128k Add a second memory pool for smaller I/Os. We can pack 16 of these on a single page instead of using an entire page for each one. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 31 +++++++++++++++++++++++-------- 1 file changed, 23 insertions(+), 8 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 80fe6a7a8163..cd7aeba8310b 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -58,6 +58,7 @@ struct nvme_dev { u32 __iomem *dbs; struct pci_dev *pci_dev; struct dma_pool *prp_page_pool; + struct dma_pool *prp_small_pool; int instance; int queue_count; u32 ctrl_config; @@ -249,11 +250,6 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) return 0; } -static __le64 *alloc_prp_list(struct nvme_dev *dev, dma_addr_t *addr) -{ - return dma_pool_alloc(dev->prp_page_pool, GFP_ATOMIC, addr); -} - struct nvme_prps { int npages; dma_addr_t first_dma; @@ -271,6 +267,9 @@ static void nvme_free_prps(struct nvme_queue *nvmeq, struct nvme_prps *prps) return; prp_dma = prps->first_dma; + + if (prps->npages == 0) + dma_pool_free(dev->prp_small_pool, prps->list[0], prp_dma); for (i = 0; i < prps->npages; i++) { __le64 *prp_list = prps->list[i]; dma_addr_t next_prp_dma = le64_to_cpu(prp_list[last_prp]); @@ -322,6 +321,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, struct scatterlist *sg, int length) { struct nvme_dev *dev = nvmeq->dev; + struct dma_pool *pool; int dma_len = sg_dma_len(sg); u64 dma_addr = sg_dma_address(sg); int offset = offset_in_page(dma_addr); @@ -352,9 +352,16 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, nprps = DIV_ROUND_UP(length, PAGE_SIZE); npages = DIV_ROUND_UP(8 * nprps, PAGE_SIZE); prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, GFP_ATOMIC); - prps->npages = npages; prp_page = 0; - prp_list = alloc_prp_list(dev, &prp_dma); + if (nprps <= (256 / 8)) { + pool = dev->prp_small_pool; + prps->npages = 0; + } else { + pool = dev->prp_page_pool; + prps->npages = npages; + } + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); prps->list[prp_page++] = prp_list; prps->first_dma = prp_dma; cmd->prp2 = cpu_to_le64(prp_dma); @@ -362,7 +369,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, for (;;) { if (i == PAGE_SIZE / 8 - 1) { __le64 *old_prp_list = prp_list; - prp_list = alloc_prp_list(dev, &prp_dma); + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); prps->list[prp_page++] = prp_list; old_prp_list[i] = cpu_to_le64(prp_dma); i = 0; @@ -1313,12 +1320,20 @@ static int nvme_setup_prp_pools(struct nvme_dev *dev) if (!dev->prp_page_pool) return -ENOMEM; + /* Optimisation for I/Os between 4k and 128k */ + dev->prp_small_pool = dma_pool_create("prp list 256", dmadev, + 256, 256, 0); + if (!dev->prp_small_pool) { + dma_pool_destroy(dev->prp_page_pool); + return -ENOMEM; + } return 0; } static void nvme_release_prp_pools(struct nvme_dev *dev) { dma_pool_destroy(dev->prp_page_pool); + dma_pool_destroy(dev->prp_small_pool); } /* XXX: Use an ida or something to let remove / add work correctly */ From d567760c409f981d35fc755b51d5bf56a99a467b Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 10:47:55 -0500 Subject: [PATCH 051/820] NVMe: Pass the nvme_dev to nvme_free_prps and nvme_setup_prps We were passing the nvme_queue to access the q_dmadev for the dma_alloc_coherent calls, but since we moved to the dma pool API, we really only need the nvme_dev. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index cd7aeba8310b..2948043483fe 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -256,10 +256,9 @@ struct nvme_prps { __le64 *list[0]; }; -static void nvme_free_prps(struct nvme_queue *nvmeq, struct nvme_prps *prps) +static void nvme_free_prps(struct nvme_dev *dev, struct nvme_prps *prps) { const int last_prp = PAGE_SIZE / 8 - 1; - struct nvme_dev *dev = nvmeq->dev; int i; dma_addr_t prp_dma; @@ -295,7 +294,7 @@ static struct nvme_bio *alloc_nbio(unsigned nseg, gfp_t gfp) static void free_nbio(struct nvme_queue *nvmeq, struct nvme_bio *nbio) { - nvme_free_prps(nvmeq, nbio->prps); + nvme_free_prps(nvmeq->dev, nbio->prps); kfree(nbio); } @@ -316,11 +315,10 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, } /* length is in bytes */ -static struct nvme_prps *nvme_setup_prps(struct nvme_queue *nvmeq, +static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, struct nvme_common_command *cmd, struct scatterlist *sg, int length) { - struct nvme_dev *dev = nvmeq->dev; struct dma_pool *pool; int dma_len = sg_dma_len(sg); u64 dma_addr = sg_dma_address(sg); @@ -458,7 +456,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, cmnd->rw.flags = 1; cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); - nbio->prps = nvme_setup_prps(nvmeq, &cmnd->common, nbio->sg, + nbio->prps = nvme_setup_prps(nvmeq->dev, &cmnd->common, nbio->sg, bio->bi_size); cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); cmnd->rw.length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); @@ -939,10 +937,10 @@ static int nvme_submit_user_admin_command(struct nvme_dev *dev, nents = nvme_map_user_pages(dev, 0, addr, length, &sg); if (nents < 0) return nents; - prps = nvme_setup_prps(dev->queues[0], &cmd->common, sg, length); + prps = nvme_setup_prps(dev, &cmd->common, sg, length); err = nvme_submit_admin_cmd(dev, cmd, NULL); nvme_unmap_user_pages(dev, 0, addr, length, sg, nents); - nvme_free_prps(dev->queues[0], prps); + nvme_free_prps(dev, prps); return err ? -EIO : 0; } @@ -1000,10 +998,10 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) c.rw.reftag = cpu_to_le32(io.reftag); /* XXX: endian? */ c.rw.apptag = cpu_to_le16(io.apptag); c.rw.appmask = cpu_to_le16(io.appmask); - nvmeq = get_nvmeq(ns); /* XXX: metadata */ - prps = nvme_setup_prps(nvmeq, &c.common, sg, length); + prps = nvme_setup_prps(dev, &c.common, sg, length); + nvmeq = get_nvmeq(ns); /* Since nvme_submit_sync_cmd sleeps, we can't keep preemption * disabled. We may be preempted at any point, and be rescheduled * to a different CPU. That will cause cacheline bouncing, but no @@ -1013,7 +1011,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) status = nvme_submit_sync_cmd(nvmeq, &c, &result, IO_TIMEOUT); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); - nvme_free_prps(nvmeq, prps); + nvme_free_prps(dev, prps); put_user(result, &uio->result); return status; } @@ -1041,11 +1039,11 @@ static int nvme_download_firmware(struct nvme_ns *ns, c.dlfw.opcode = nvme_admin_download_fw; c.dlfw.numd = cpu_to_le32(dlfw.length); c.dlfw.offset = cpu_to_le32(dlfw.offset); - prps = nvme_setup_prps(dev->queues[0], &c.common, sg, dlfw.length * 4); + prps = nvme_setup_prps(dev, &c.common, sg, dlfw.length * 4); status = nvme_submit_admin_cmd(dev, &c, NULL); nvme_unmap_user_pages(dev, 0, dlfw.addr, dlfw.length * 4, sg, nents); - nvme_free_prps(dev->queues[0], prps); + nvme_free_prps(dev, prps); return status; } From 1974b1ae8852324a75fb8cfecbc7b758fd5a2c3c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 12:01:09 -0500 Subject: [PATCH 052/820] NVMe: Check for DMA mapping failure If dma_map_sg returns 0 (failure), we need to fail the I/O. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 2948043483fe..bfdca3a3a41a 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -451,7 +451,8 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, dma_dir = DMA_FROM_DEVICE; } - nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs); + if (nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs) == 0) + goto mapping_failed; cmnd->rw.flags = 1; cmnd->rw.command_id = cmdid; @@ -471,6 +472,11 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return 0; + mapping_failed: + free_nbio(nvmeq, nbio); + bio_endio(bio, -ENOMEM); + return 0; + free_nbio: free_nbio(nvmeq, nbio); congestion: From 768308400f5b4ce665a072eb976a851978b7706e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 10 Feb 2011 13:55:39 -0500 Subject: [PATCH 053/820] NVMe: Handle physical merging of bvec entries In order to not overrun the sg array, we have to merge physically contiguous pages into a single sg entry. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index bfdca3a3a41a..c0e84b688f50 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -392,19 +392,25 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, struct bio *bio, enum dma_data_direction dma_dir, int psegs) { - struct bio_vec *bvec; - struct scatterlist *sg = nbio->sg; - int i, nsegs; + struct bio_vec *bvec, *bvprv = NULL; + struct scatterlist *sg = NULL; + int i, nsegs = 0; - sg_init_table(sg, psegs); + sg_init_table(nbio->sg, psegs); bio_for_each_segment(bvec, bio, i) { - sg_set_page(sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); - sg++; - /* XXX: handle non-mergable here */ - nsegs++; + if (bvprv && BIOVEC_PHYS_MERGEABLE(bvprv, bvec)) { + sg->length += bvec->bv_len; + } else { + /* Check bvprv && offset == 0 */ + sg = sg ? sg + 1 : nbio->sg; + sg_set_page(sg, bvec->bv_page, bvec->bv_len, + bvec->bv_offset); + nsegs++; + } + bvprv = bvec; } nbio->nents = nsegs; - + sg_mark_end(sg); return dma_map_sg(dev, nbio->sg, nbio->nents, dma_dir); } From 897cfe1ce7db152fa6dde576f4213a6160bf6502 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 14 Feb 2011 12:20:15 -0500 Subject: [PATCH 054/820] NVMe: Update BAR structure to match the current spec Add two reserved registers in the middle of the BAR to match the 1.0 spec plus ECN 0002. Also rename IMC and ISC to INTMC and INTSC to conform with the spec. We still don't need to use them :-) Signed-off-by: Matthew Wilcox --- include/linux/nvme.h | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 8eed0e432eef..757faa71666e 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -24,10 +24,12 @@ struct nvme_bar { __u64 cap; /* Controller Capabilities */ __u32 vs; /* Version */ - __u32 ims; /* Interrupt Mask Set */ - __u32 imc; /* Interrupt Mask Clear */ + __u32 intms; /* Interrupt Mask Set */ + __u32 intmc; /* Interrupt Mask Clear */ __u32 cc; /* Controller Configuration */ + __u32 rsvd1; /* Reserved */ __u32 csts; /* Controller Status */ + __u32 rsvd2; /* Reserved */ __u32 aqa; /* Admin Queue Attributes */ __u64 asq; /* Admin SQ Base Address */ __u64 acq; /* Admin CQ Base Address */ From eeee322647a67c20d9277c5e02c42b2126ea74bc Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 14 Feb 2011 15:55:33 -0500 Subject: [PATCH 055/820] NVMe: Handle failures differently in nvme_submit_bio_queue() Return -EBUSY if the queue is full or -ENOMEM if we failed to allocate memory (or map a scatterlist). Also use GFP_ATOMIC to allocate the nvme_bio and move the locking to the callers of nvme_submit_bio_queue(). In nvme_make_request(), don't permit an I/O to jump the queue -- if the congestion list already has an entry, just add to the tail, rather than trying to submit. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 36 +++++++++++++++++------------------- 1 file changed, 17 insertions(+), 19 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index c0e84b688f50..61a241741ca6 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -420,17 +420,17 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct nvme_command *cmnd; struct nvme_bio *nbio; enum dma_data_direction dma_dir; - int cmdid; + int cmdid, result = -ENOMEM; u16 control; u32 dsmgmt; - unsigned long flags; int psegs = bio_phys_segments(ns->queue, bio); - nbio = alloc_nbio(psegs, GFP_NOIO); + nbio = alloc_nbio(psegs, GFP_ATOMIC); if (!nbio) - goto congestion; + goto nomem; nbio->bio = bio; + result = -EBUSY; cmdid = alloc_cmdid(nvmeq, nbio, bio_completion_id, IO_TIMEOUT); if (unlikely(cmdid < 0)) goto free_nbio; @@ -445,7 +445,6 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, if (bio->bi_rw & REQ_RAHEAD) dsmgmt |= NVME_RW_DSM_FREQ_PREFETCH; - spin_lock_irqsave(&nvmeq->q_lock, flags); cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail]; memset(cmnd, 0, sizeof(*cmnd)); @@ -457,8 +456,9 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, dma_dir = DMA_FROM_DEVICE; } + result = -ENOMEM; if (nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs) == 0) - goto mapping_failed; + goto free_nbio; cmnd->rw.flags = 1; cmnd->rw.command_id = cmdid; @@ -474,19 +474,12 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, if (++nvmeq->sq_tail == nvmeq->q_depth) nvmeq->sq_tail = 0; - spin_unlock_irqrestore(&nvmeq->q_lock, flags); - - return 0; - - mapping_failed: - free_nbio(nvmeq, nbio); - bio_endio(bio, -ENOMEM); return 0; free_nbio: free_nbio(nvmeq, nbio); - congestion: - return -EBUSY; + nomem: + return result; } static void nvme_resubmit_bio(struct nvme_queue *nvmeq, struct bio *bio) @@ -507,13 +500,18 @@ static int nvme_make_request(struct request_queue *q, struct bio *bio) { struct nvme_ns *ns = q->queuedata; struct nvme_queue *nvmeq = get_nvmeq(ns); + int result = -EBUSY; - if (nvme_submit_bio_queue(nvmeq, ns, bio)) { - blk_set_queue_congested(q, rw_is_sync(bio->bi_rw)); - spin_lock_irq(&nvmeq->q_lock); + spin_lock_irq(&nvmeq->q_lock); + if (bio_list_empty(&nvmeq->sq_cong)) + result = nvme_submit_bio_queue(nvmeq, ns, bio); + if (unlikely(result)) { + if (bio_list_empty(&nvmeq->sq_cong)) + add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait); bio_list_add(&nvmeq->sq_cong, bio); - spin_unlock_irq(&nvmeq->q_lock); } + + spin_unlock_irq(&nvmeq->q_lock); put_nvmeq(nvmeq); return 0; From 1fa6aeadf18aeebd7a217d7a3a933856448375b6 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 2 Mar 2011 18:37:18 -0500 Subject: [PATCH 056/820] NVMe: Add a kthread to handle the congestion list Instead of trying to resubmit I/Os in the I/O completion path (in interrupt context), wake up a kthread which will resubmit I/O from user context. This allows mke2fs to run to completion. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 86 ++++++++++++++++++++++++++++++++++---------- 1 file changed, 67 insertions(+), 19 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 61a241741ca6..606371e62905 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include @@ -50,10 +51,15 @@ module_param(nvme_major, int, 0); static int use_threaded_interrupts; module_param(use_threaded_interrupts, int, 0); +static DEFINE_SPINLOCK(dev_list_lock); +static LIST_HEAD(dev_list); +static struct task_struct *nvme_thread; + /* * Represents an NVM Express device. Each nvme_dev is a PCI function. */ struct nvme_dev { + struct list_head node; struct nvme_queue **queues; u32 __iomem *dbs; struct pci_dev *pci_dev; @@ -97,6 +103,7 @@ struct nvme_queue { dma_addr_t sq_dma_addr; dma_addr_t cq_dma_addr; wait_queue_head_t sq_full; + wait_queue_t sq_cong_wait; struct bio_list sq_cong; u32 __iomem *q_db; u16 q_depth; @@ -108,8 +115,6 @@ struct nvme_queue { unsigned long cmdid_data[]; }; -static void nvme_resubmit_bio(struct nvme_queue *nvmeq, struct bio *bio); - /* * Check we didin't inadvertently grow the command struct */ @@ -309,9 +314,6 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); free_nbio(nvmeq, nbio); bio_endio(bio, status ? -EIO : 0); - bio = bio_list_pop(&nvmeq->sq_cong); - if (bio) - nvme_resubmit_bio(nvmeq, bio); } /* length is in bytes */ @@ -482,16 +484,6 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return result; } -static void nvme_resubmit_bio(struct nvme_queue *nvmeq, struct bio *bio) -{ - struct nvme_ns *ns = bio->bi_bdev->bd_disk->private_data; - if (nvme_submit_bio_queue(nvmeq, ns, bio)) - bio_list_add_head(&nvmeq->sq_cong, bio); - else if (bio_list_empty(&nvmeq->sq_cong)) - blk_clear_queue_congested(ns->queue, rw_is_sync(bio->bi_rw)); - /* XXX: Need to duplicate the logic from __freed_request here */ -} - /* * NB: return value of non-zero would mean that we were a stacking driver. * make_request must always succeed. @@ -774,6 +766,7 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, nvmeq->cq_head = 0; nvmeq->cq_phase = 1; init_waitqueue_head(&nvmeq->sq_full); + init_waitqueue_entry(&nvmeq->sq_cong_wait, nvme_thread); bio_list_init(&nvmeq->sq_cong); nvmeq->q_db = &dev->dbs[qid * 2]; nvmeq->q_depth = depth; @@ -1097,6 +1090,43 @@ static const struct block_device_operations nvme_fops = { .ioctl = nvme_ioctl, }; +static void nvme_resubmit_bios(struct nvme_queue *nvmeq) +{ + while (bio_list_peek(&nvmeq->sq_cong)) { + struct bio *bio = bio_list_pop(&nvmeq->sq_cong); + struct nvme_ns *ns = bio->bi_bdev->bd_disk->private_data; + if (nvme_submit_bio_queue(nvmeq, ns, bio)) { + bio_list_add_head(&nvmeq->sq_cong, bio); + break; + } + } +} + +static int nvme_kthread(void *data) +{ + struct nvme_dev *dev; + + while (!kthread_should_stop()) { + __set_current_state(TASK_RUNNING); + spin_lock(&dev_list_lock); + list_for_each_entry(dev, &dev_list, node) { + int i; + for (i = 0; i < dev->queue_count; i++) { + struct nvme_queue *nvmeq = dev->queues[i]; + spin_lock_irq(&nvmeq->q_lock); + if (nvme_process_cq(nvmeq)) + printk("process_cq did something\n"); + nvme_resubmit_bios(nvmeq); + spin_unlock_irq(&nvmeq->q_lock); + } + } + spin_unlock(&dev_list_lock); + set_current_state(TASK_INTERRUPTIBLE); + schedule_timeout(HZ); + } + return 0; +} + static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int index, struct nvme_id_ns *id, struct nvme_lba_range_type *rt) { @@ -1307,6 +1337,10 @@ static int nvme_dev_remove(struct nvme_dev *dev) { struct nvme_ns *ns, *next; + spin_lock(&dev_list_lock); + list_del(&dev->node); + spin_unlock(&dev_list_lock); + /* TODO: wait all I/O finished or cancel them */ list_for_each_entry_safe(ns, next, &dev->namespaces, list) { @@ -1406,6 +1440,11 @@ static int __devinit nvme_probe(struct pci_dev *pdev, result = nvme_dev_add(dev); if (result) goto delete; + + spin_lock(&dev_list_lock); + list_add(&dev->node, &dev_list); + spin_unlock(&dev_list_lock); + return 0; delete: @@ -1479,17 +1518,25 @@ static struct pci_driver nvme_driver = { static int __init nvme_init(void) { - int result; + int result = -EBUSY; + + nvme_thread = kthread_run(nvme_kthread, NULL, "nvme"); + if (IS_ERR(nvme_thread)) + return PTR_ERR(nvme_thread); nvme_major = register_blkdev(nvme_major, "nvme"); if (nvme_major <= 0) - return -EBUSY; + goto kill_kthread; result = pci_register_driver(&nvme_driver); - if (!result) - return 0; + if (result) + goto unregister_blkdev; + return 0; + unregister_blkdev: unregister_blkdev(nvme_major, "nvme"); + kill_kthread: + kthread_stop(nvme_thread); return result; } @@ -1497,6 +1544,7 @@ static void __exit nvme_exit(void) { pci_unregister_driver(&nvme_driver); unregister_blkdev(nvme_major, "nvme"); + kthread_stop(nvme_thread); } MODULE_AUTHOR("Matthew Wilcox "); From ad8a5df97cb060aa4d817af25587c99e2d2fda97 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 14 Feb 2011 17:35:00 -0500 Subject: [PATCH 057/820] NVMe: Release 0.3 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 606371e62905..7554625fb94d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1549,6 +1549,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.2"); +MODULE_VERSION("0.3"); module_init(nvme_init); module_exit(nvme_exit); From ca1615424c9adfdbe7d484771d7a7c5ecc4bb6d2 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 15 Feb 2011 13:44:13 -0500 Subject: [PATCH 058/820] NVMe: Remove setting of 'flags' in rw command This was the data transfer bit until spec rev 0.92 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 7554625fb94d..37cdf0711954 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -462,7 +462,6 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, if (nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs) == 0) goto free_nbio; - cmnd->rw.flags = 1; cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); nbio->prps = nvme_setup_prps(nvmeq->dev, &cmnd->common, nbio->sg, From b348b7d54368c87811907a8e88f0d96713c43009 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 15 Feb 2011 16:16:02 -0500 Subject: [PATCH 059/820] NVMe: Rename nr_queues to nr_io_queues I got confused about whether this included the admin queue or not, and had to resort to reading the spec. It doesn't include the admin queue, so make that clear in the name. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 37cdf0711954..f3aa8097e675 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1202,29 +1202,30 @@ static int set_queue_count(struct nvme_dev *dev, int count) static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) { - int result, cpu, i, nr_queues; + int result, cpu, i, nr_io_queues; - nr_queues = num_online_cpus(); - result = set_queue_count(dev, nr_queues); + nr_io_queues = num_online_cpus(); + result = set_queue_count(dev, nr_io_queues); if (result < 0) return result; - if (result < nr_queues) - nr_queues = result; + if (result < nr_io_queues) + nr_io_queues = result; /* Deregister the admin queue's interrupt */ free_irq(dev->entry[0].vector, dev->queues[0]); - for (i = 0; i < nr_queues; i++) + for (i = 0; i < nr_io_queues; i++) dev->entry[i].entry = i; for (;;) { - result = pci_enable_msix(dev->pci_dev, dev->entry, nr_queues); + result = pci_enable_msix(dev->pci_dev, dev->entry, + nr_io_queues); if (result == 0) { break; } else if (result > 0) { - nr_queues = result; + nr_io_queues = result; continue; } else { - nr_queues = 1; + nr_io_queues = 1; break; } } @@ -1233,12 +1234,12 @@ static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) /* XXX: handle failure here */ cpu = cpumask_first(cpu_online_mask); - for (i = 0; i < nr_queues; i++) { + for (i = 0; i < nr_io_queues; i++) { irq_set_affinity_hint(dev->entry[i].vector, get_cpu_mask(cpu)); cpu = cpumask_next(cpu, cpu_online_mask); } - for (i = 0; i < nr_queues; i++) { + for (i = 0; i < nr_io_queues; i++) { dev->queues[i + 1] = nvme_create_queue(dev, i + 1, NVME_Q_DEPTH, i); if (!dev->queues[i + 1]) From 740216fc59cba54f65187c9ed92f29bce3cf8778 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 15 Feb 2011 16:28:20 -0500 Subject: [PATCH 060/820] NVMe: Let the kthread take care of devices earlier If interrupts are misconfigured, the kthread will be needed to process admin queue completions. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index f3aa8097e675..df1d8bda8c7c 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1112,6 +1112,8 @@ static int nvme_kthread(void *data) int i; for (i = 0; i < dev->queue_count; i++) { struct nvme_queue *nvmeq = dev->queues[i]; + if (!nvmeq) + continue; spin_lock_irq(&nvmeq->q_lock); if (nvme_process_cq(nvmeq)) printk("process_cq did something\n"); @@ -1437,17 +1439,21 @@ static int __devinit nvme_probe(struct pci_dev *pdev, goto unmap; dev->queue_count++; - result = nvme_dev_add(dev); - if (result) - goto delete; - spin_lock(&dev_list_lock); list_add(&dev->node, &dev_list); spin_unlock(&dev_list_lock); + result = nvme_dev_add(dev); + if (result) + goto delete; + return 0; delete: + spin_lock(&dev_list_lock); + list_del(&dev->node); + spin_unlock(&dev_list_lock); + nvme_free_queues(dev); unmap: iounmap(dev->bar); From 7547881d0951384f9833ec3a80fac8f3f16f3b98 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Feb 2011 09:59:59 -0500 Subject: [PATCH 061/820] NVMe: Correct SQ doorbell semantics The value written to the doorbell needs to be the first free index in the queue, not the most recently used index in the queue. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index df1d8bda8c7c..af45e286d5dd 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -246,9 +246,9 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) spin_lock_irqsave(&nvmeq->q_lock, flags); tail = nvmeq->sq_tail; memcpy(&nvmeq->sq_cmds[tail], cmd, sizeof(*cmd)); - writel(tail, nvmeq->q_db); if (++tail == nvmeq->q_depth) tail = 0; + writel(tail, nvmeq->q_db); nvmeq->sq_tail = tail; spin_unlock_irqrestore(&nvmeq->q_lock, flags); @@ -471,9 +471,9 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, cmnd->rw.control = cpu_to_le16(control); cmnd->rw.dsmgmt = cpu_to_le32(dsmgmt); - writel(nvmeq->sq_tail, nvmeq->q_db); if (++nvmeq->sq_tail == nvmeq->q_depth) nvmeq->sq_tail = 0; + writel(nvmeq->sq_tail, nvmeq->q_db); return 0; From c42705592be2a539f3027b6f3907de8e8f9591a8 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 22 Feb 2011 14:15:34 -0500 Subject: [PATCH 062/820] NVMe: Mark CMD_CTX_CANCELLED as being unlikely Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index af45e286d5dd..ce919b49b30d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -518,7 +518,7 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, struct nvme_completion *cqe) { struct sync_cmd_info *cmdinfo = ctx; - if ((unsigned long)cmdinfo == CMD_CTX_CANCELLED) + if (unlikely((unsigned long)cmdinfo == CMD_CTX_CANCELLED)) return; if (unlikely((unsigned long)cmdinfo == CMD_CTX_COMPLETED)) { dev_warn(nvmeq->q_dmadev, From 00df5cb4eb927078850086f8becc3286a69ea12e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 22 Feb 2011 14:18:30 -0500 Subject: [PATCH 063/820] NVMe: Implement Flush Linux implements Flush as a bit in the bio. That means there may also be data associated with the flush; if so the flush should be sent before the data. To avoid completing the bio twice, I add CMD_CTX_FLUSH to indicate the completion routine should do nothing. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 40 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 40 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index ce919b49b30d..d99b400ccd79 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -191,10 +191,12 @@ enum { bio_completion_id, }; +/* Special values must be a multiple of 4, and less than 0x1000 */ #define CMD_CTX_BASE (POISON_POINTER_DELTA + sync_completion_id) #define CMD_CTX_CANCELLED (0x30C + CMD_CTX_BASE) #define CMD_CTX_COMPLETED (0x310 + CMD_CTX_BASE) #define CMD_CTX_INVALID (0x314 + CMD_CTX_BASE) +#define CMD_CTX_FLUSH (0x318 + CMD_CTX_BASE) static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { @@ -416,6 +418,33 @@ static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, return dma_map_sg(dev, nbio->sg, nbio->nents, dma_dir); } +static int nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns, + int cmdid) +{ + struct nvme_command *cmnd = &nvmeq->sq_cmds[nvmeq->sq_tail]; + + memset(cmnd, 0, sizeof(*cmnd)); + cmnd->common.opcode = nvme_cmd_flush; + cmnd->common.command_id = cmdid; + cmnd->common.nsid = cpu_to_le32(ns->ns_id); + + if (++nvmeq->sq_tail == nvmeq->q_depth) + nvmeq->sq_tail = 0; + writel(nvmeq->sq_tail, nvmeq->q_db); + + return 0; +} + +static int nvme_submit_flush_data(struct nvme_queue *nvmeq, struct nvme_ns *ns) +{ + int cmdid = alloc_cmdid(nvmeq, (void *)CMD_CTX_FLUSH, + sync_completion_id, IO_TIMEOUT); + if (unlikely(cmdid < 0)) + return cmdid; + + return nvme_submit_flush(nvmeq, ns, cmdid); +} + static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct bio *bio) { @@ -427,6 +456,12 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, u32 dsmgmt; int psegs = bio_phys_segments(ns->queue, bio); + if ((bio->bi_rw & REQ_FLUSH) && psegs) { + result = nvme_submit_flush_data(nvmeq, ns); + if (result) + return result; + } + nbio = alloc_nbio(psegs, GFP_ATOMIC); if (!nbio) goto nomem; @@ -437,6 +472,9 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, if (unlikely(cmdid < 0)) goto free_nbio; + if ((bio->bi_rw & REQ_FLUSH) && !psegs) + return nvme_submit_flush(nvmeq, ns, cmdid); + control = 0; if (bio->bi_rw & REQ_FUA) control |= NVME_RW_FUA; @@ -520,6 +558,8 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, struct sync_cmd_info *cmdinfo = ctx; if (unlikely((unsigned long)cmdinfo == CMD_CTX_CANCELLED)) return; + if ((unsigned long)cmdinfo == CMD_CTX_FLUSH) + return; if (unlikely((unsigned long)cmdinfo == CMD_CTX_COMPLETED)) { dev_warn(nvmeq->q_dmadev, "completed id %d twice on queue %d\n", From 1ad2f8932a72bf375361727949ced2cb4e8cfcef Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 23 Feb 2011 15:20:00 -0500 Subject: [PATCH 064/820] NVMe: Handle bios that contain non-virtually contiguous addresses NVMe scatterlists must be virtually contiguous, like almost all I/Os. However, when the filesystem lays out files with a hole, it can be that adjacent LBAs map to non-adjacent virtual addresses. Handle this by submitting one NVMe command at a time for each virtually discontiguous range. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 38 +++++++++++++++++++++++++++++--------- 1 file changed, 29 insertions(+), 9 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index d99b400ccd79..240922706a93 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -315,7 +315,14 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, dma_unmap_sg(nvmeq->q_dmadev, nbio->sg, nbio->nents, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); free_nbio(nvmeq, nbio); - bio_endio(bio, status ? -EIO : 0); + if (status) + bio_endio(bio, -EIO); + if (bio->bi_vcnt > bio->bi_idx) { + bio_list_add(&nvmeq->sq_cong, bio); + wake_up_process(nvme_thread); + } else { + bio_endio(bio, 0); + } } /* length is in bytes */ @@ -393,29 +400,41 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, return prps; } +/* NVMe scatterlists require no holes in the virtual address */ +#define BIOVEC_NOT_VIRT_MERGEABLE(vec1, vec2) ((vec2)->bv_offset || \ + (((vec1)->bv_offset + (vec1)->bv_len) % PAGE_SIZE)) + static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, struct bio *bio, enum dma_data_direction dma_dir, int psegs) { struct bio_vec *bvec, *bvprv = NULL; struct scatterlist *sg = NULL; - int i, nsegs = 0; + int i, old_idx, length = 0, nsegs = 0; sg_init_table(nbio->sg, psegs); + old_idx = bio->bi_idx; bio_for_each_segment(bvec, bio, i) { if (bvprv && BIOVEC_PHYS_MERGEABLE(bvprv, bvec)) { sg->length += bvec->bv_len; } else { - /* Check bvprv && offset == 0 */ + if (bvprv && BIOVEC_NOT_VIRT_MERGEABLE(bvprv, bvec)) + break; sg = sg ? sg + 1 : nbio->sg; sg_set_page(sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); nsegs++; } + length += bvec->bv_len; bvprv = bvec; } + bio->bi_idx = i; nbio->nents = nsegs; sg_mark_end(sg); - return dma_map_sg(dev, nbio->sg, nbio->nents, dma_dir); + if (dma_map_sg(dev, nbio->sg, nbio->nents, dma_dir) == 0) { + bio->bi_idx = old_idx; + return -ENOMEM; + } + return length; } static int nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns, @@ -451,7 +470,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct nvme_command *cmnd; struct nvme_bio *nbio; enum dma_data_direction dma_dir; - int cmdid, result = -ENOMEM; + int cmdid, length, result = -ENOMEM; u16 control; u32 dsmgmt; int psegs = bio_phys_segments(ns->queue, bio); @@ -496,16 +515,17 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, dma_dir = DMA_FROM_DEVICE; } - result = -ENOMEM; - if (nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs) == 0) + result = nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs); + if (result < 0) goto free_nbio; + length = result; cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); nbio->prps = nvme_setup_prps(nvmeq->dev, &cmnd->common, nbio->sg, - bio->bi_size); + length); cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); - cmnd->rw.length = cpu_to_le16((bio->bi_size >> ns->lba_shift) - 1); + cmnd->rw.length = cpu_to_le16((length >> ns->lba_shift) - 1); cmnd->rw.control = cpu_to_le16(control); cmnd->rw.dsmgmt = cpu_to_le32(dsmgmt); From d8ee9d69f275769aaad40ef7c944565ff8d2d24f Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 24 Feb 2011 08:46:00 -0500 Subject: [PATCH 065/820] NVMe: Fix discontiguous accesses When we submit subsequent portions of the I/O, we need to access the updated block, not start reading again from the original position. This was showing up as miscompares in the XFS randholes testcase. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 240922706a93..562d75a0fc50 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -529,6 +529,8 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, cmnd->rw.control = cpu_to_le16(control); cmnd->rw.dsmgmt = cpu_to_le32(dsmgmt); + bio->bi_sector += length >> 9; + if (++nvmeq->sq_tail == nvmeq->q_depth) nvmeq->sq_tail = 0; writel(nvmeq->sq_tail, nvmeq->q_db); From e6d15f79f997a98b3a69abbc462fc9041cc1a7b4 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 24 Feb 2011 08:49:41 -0500 Subject: [PATCH 066/820] NVMe: Reduce maximum queue depth by 1 The spec says we're not allowed to completely fill the submission queue. Solve this by reducing the number of allocatable cmdids by 1. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 562d75a0fc50..45bfae1ebd50 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -156,7 +156,7 @@ static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler, unsigned timeout) { - int depth = nvmeq->q_depth; + int depth = nvmeq->q_depth - 1; struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); int cmdid; From b57ab0fada358357571f0eb448cdf2f144785321 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 24 Feb 2011 16:20:14 -0500 Subject: [PATCH 067/820] NVMe: Version 0.4 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 45bfae1ebd50..a8549dff4691 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1617,6 +1617,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.3"); +MODULE_VERSION("0.4"); module_init(nvme_init); module_exit(nvme_exit); From 2ddc4f74d8adcf3e1cdec7f3e72d19b5c878597c Mon Sep 17 00:00:00 2001 From: Krzysztof Wierzbicki Date: Mon, 28 Feb 2011 08:27:13 +0100 Subject: [PATCH 068/820] NVMe: Update admin opcodes to match the 1.0RC spec Signed-off-by: Krzysztof Wierzbicki Signed-off-by: Matthew Wilcox Signed-off-by: Matthew Wilcox --- include/linux/nvme.h | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 757faa71666e..c46a9b7988fb 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -199,19 +199,19 @@ enum { enum nvme_admin_opcode { nvme_admin_delete_sq = 0x00, nvme_admin_create_sq = 0x01, - nvme_admin_get_features = 0x02, + nvme_admin_get_log_page = 0x02, nvme_admin_delete_cq = 0x04, nvme_admin_create_cq = 0x05, nvme_admin_identify = 0x06, nvme_admin_abort_cmd = 0x08, nvme_admin_set_features = 0x09, - nvme_admin_get_log_page = 0x0a, + nvme_admin_get_features = 0x0a, nvme_admin_async_event = 0x0c, - nvme_admin_download_fw = 0x0d, - nvme_admin_security_recv = 0x0e, - nvme_admin_format_nvm = 0x10, - nvme_admin_security_send = 0x11, - nvme_admin_activate_fw = 0x14, + nvme_admin_activate_fw = 0x10, + nvme_admin_download_fw = 0x11, + nvme_admin_format_nvm = 0x80, + nvme_admin_security_send = 0x81, + nvme_admin_security_recv = 0x82, }; enum { From 714a7a22884b74862540bc84955274d86b2f6040 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:28:24 -0400 Subject: [PATCH 069/820] NVMe: Convert comments to kernel-doc notation Reported-by: Randy Dunlap Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index a8549dff4691..e392919e0eac 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -142,10 +142,10 @@ static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) } /** - * alloc_cmdid - Allocate a Command ID - * @param nvmeq The queue that will be used for this command - * @param ctx A pointer that will be passed to the handler - * @param handler The ID of the handler to call + * alloc_cmdid() - Allocate a Command ID + * @nvmeq: The queue that will be used for this command + * @ctx: A pointer that will be passed to the handler + * @handler: The ID of the handler to call * * Allocate a Command ID for a queue. The data passed in will * be passed to the completion handler. This is implemented by using @@ -234,7 +234,7 @@ static void put_nvmeq(struct nvme_queue *nvmeq) } /** - * nvme_submit_cmd: Copy a command into a queue and ring the doorbell + * nvme_submit_cmd() - Copy a command into a queue and ring the doorbell * @nvmeq: The queue to use * @cmd: The command to send * From fa92282149842645931580225647238428374758 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:29:00 -0400 Subject: [PATCH 070/820] NVMe: Fix comment formatting Reported-by: Randy Dunlap Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index e392919e0eac..740a9c1b81aa 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -182,7 +182,8 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, return (cmdid < 0) ? -EINTR : cmdid; } -/* If you need more than four handlers, you'll need to change how +/* + * If you need more than four handlers, you'll need to change how * alloc_cmdid and nvme_process_cq work. Consider using a special * CMD_CTX value instead, if that works for your situation. */ @@ -1066,7 +1067,8 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) prps = nvme_setup_prps(dev, &c.common, sg, length); nvmeq = get_nvmeq(ns); - /* Since nvme_submit_sync_cmd sleeps, we can't keep preemption + /* + * Since nvme_submit_sync_cmd sleeps, we can't keep preemption * disabled. We may be preempted at any point, and be rescheduled * to a different CPU. That will cause cacheline bouncing, but no * additional races since q_lock already protects against other CPUs. From 19e899b2f9f89f4a290dd5c9c24d15987a18ab21 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:29:24 -0400 Subject: [PATCH 071/820] NVMe: Remove outdated comments The head can never overrun the tail since we won't allocate enough command IDs to let that happen. The status codes are in sync with the spec. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 - include/linux/nvme.h | 1 - 2 files changed, 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 740a9c1b81aa..d4f95eb51dc1 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -245,7 +245,6 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) { unsigned long flags; u16 tail; - /* XXX: Need to check tail isn't going to overrun head */ spin_lock_irqsave(&nvmeq->q_lock, flags); tail = nvmeq->sq_tail; memcpy(&nvmeq->sq_cmds[tail], cmd, sizeof(*cmd)); diff --git a/include/linux/nvme.h b/include/linux/nvme.h index c46a9b7988fb..6b5a8d19daf5 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -323,7 +323,6 @@ struct nvme_command { }; }; -/* XXX: Sync with spec */ enum { NVME_SC_SUCCESS = 0x0, NVME_SC_INVALID_OPCODE = 0x1, From ac88c36a385b848cb9efcb877fdfc4153a60bcab Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:29:58 -0400 Subject: [PATCH 072/820] NVMe: Fix interpretation of 'Number of Namespaces' field The spec says this is a 0s based value. We don't need to handle the maximal value because it's reserved to mean "every namespace". Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index d4f95eb51dc1..0d5c918b7d59 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1364,7 +1364,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) crt.features.prp1 = cpu_to_le64(dma_addr + 4096); crt.features.fid = cpu_to_le32(NVME_FEAT_LBA_RANGE); - for (i = 0; i < nn; i++) { + for (i = 0; i <= nn; i++) { cid.identify.nsid = cpu_to_le32(i); res = nvme_submit_admin_cmd(dev, &cid, NULL); if (res) From 7523d834dd1573610078eb1ac0933f6490232f90 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:43:40 -0400 Subject: [PATCH 073/820] NVMe: Fix off-by-one when filling in PRP lists If the last element in the PRP list fits on the end of the page, there's no need to allocate an extra page to put that single element in. It can fit on the end of the page. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 0d5c918b7d59..cf89db8c41ee 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -376,12 +376,13 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, cmd->prp2 = cpu_to_le64(prp_dma); i = 0; for (;;) { - if (i == PAGE_SIZE / 8 - 1) { + if (i == PAGE_SIZE / 8) { __le64 *old_prp_list = prp_list; prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); prps->list[prp_page++] = prp_list; - old_prp_list[i] = cpu_to_le64(prp_dma); - i = 0; + prp_list[0] = old_prp_list[i - 1]; + old_prp_list[i - 1] = cpu_to_le64(prp_dma); + i = 1; } prp_list[i++] = cpu_to_le64(dma_addr); dma_len -= PAGE_SIZE; From 3cb967c03926edd2c414082f4cc0feb7b372edae Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:45:49 -0400 Subject: [PATCH 074/820] NVMe: Remove the kthread from the wait queue Once there are no more bios on the congestion list, we can stop waking up the nvme kthread every time a completion happens. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index cf89db8c41ee..8d3c0b79ac2b 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1162,6 +1162,9 @@ static void nvme_resubmit_bios(struct nvme_queue *nvmeq) bio_list_add_head(&nvmeq->sq_cong, bio); break; } + if (bio_list_empty(&nvmeq->sq_cong)) + remove_wait_queue(&nvmeq->sq_full, + &nvmeq->sq_cong_wait); } } From 9ecdc946212f7cd592986b2c519b470404caa6b8 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 16 Mar 2011 16:52:19 -0400 Subject: [PATCH 075/820] NVMe: Simplify queue lookup Fill in all the num_possible_cpus() entries with duplicate pointers. This reduces the complexity of the frequently-called get_nvmeq(), as well as avoiding a bug in it when there are fewer queues than CPUs. Reported-by: Shane Michael Matthews Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 8d3c0b79ac2b..f94f1731478f 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -221,12 +221,7 @@ static void cancel_cmdid_data(struct nvme_queue *nvmeq, int cmdid) static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) { - int qid, cpu = get_cpu(); - if (cpu < ns->dev->queue_count) - qid = cpu + 1; - else - qid = (cpu % rounddown_pow_of_two(ns->dev->queue_count)) + 1; - return ns->dev->queues[qid]; + return ns->dev->queues[get_cpu() + 1]; } static void put_nvmeq(struct nvme_queue *nvmeq) @@ -1316,6 +1311,11 @@ static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) dev->queue_count++; } + for (; i < num_possible_cpus(); i++) { + int target = i % rounddown_pow_of_two(dev->queue_count - 1); + dev->queues[i + 1] = dev->queues[target + 1]; + } + return 0; } From 4948168280b269a514045766ddd872cfac5968e1 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sat, 19 Mar 2011 14:55:38 -0400 Subject: [PATCH 076/820] NVMe: Add compat_ioctl Make ioctls work for 32-bit applications on 64-bit kernels. The structures are defined to be the same for both 32- and 64-bit applications, so we can use the same handler for both. Reported-by: Arnd Bergmann Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index f94f1731478f..d0b52622e261 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1146,6 +1146,7 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, static const struct block_device_operations nvme_fops = { .owner = THIS_MODULE, .ioctl = nvme_ioctl, + .compat_ioctl = nvme_ioctl, }; static void nvme_resubmit_bios(struct nvme_queue *nvmeq) From 9d4af1b7796ba02b73a79a8694399e5a3cd1c55d Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 20 Mar 2011 07:27:10 -0400 Subject: [PATCH 077/820] NVMe: Correct the definitions of two ioctls NVME_IOCTL_SUBMIT_IO has a struct nvme_user_io, not a struct nvme_rw_command as a parameter, and NVME_IOCTL_DOWNLOAD_FW is a Write, not a Read. Reported-by: Arnd Bergmann Signed-off-by: Matthew Wilcox --- include/linux/nvme.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 6b5a8d19daf5..fd10d597cca7 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -395,8 +395,8 @@ struct nvme_dlfw { #define NVME_IOCTL_IDENTIFY_NS _IOW('N', 0x40, struct nvme_id_ns) #define NVME_IOCTL_IDENTIFY_CTRL _IOW('N', 0x41, struct nvme_id_ctrl) #define NVME_IOCTL_GET_RANGE_TYPE _IOW('N', 0x42, struct nvme_lba_range_type) -#define NVME_IOCTL_SUBMIT_IO _IOWR('N', 0x43, struct nvme_rw_command) -#define NVME_IOCTL_DOWNLOAD_FW _IOR('N', 0x44, struct nvme_dlfw) +#define NVME_IOCTL_SUBMIT_IO _IOWR('N', 0x43, struct nvme_user_io) +#define NVME_IOCTL_DOWNLOAD_FW _IOW('N', 0x44, struct nvme_dlfw) #define NVME_IOCTL_ACTIVATE_FW _IO('N', 0x45) #endif /* _LINUX_NVME_H */ From 6c7d49455ceb63064f992347d9185ff5bf43497a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 21 Mar 2011 09:48:57 -0400 Subject: [PATCH 078/820] NVMe: Change the definition of nvme_user_io The read and write commands don't define a 'result', so there's no need to copy it back to userspace. Remove the ability of the ioctl to submit commands to a different namespace; it's just asking for trouble, and the use case I have in mind will be addressed througha different ioctl in the future. That removes the need for both the block_shift and nsid arguments. Check that the opcode is one of 'read' or 'write'. Future opcodes may be added in the future, but we will need a different structure definition for them. The nblocks field is redefined to be 0-based. This allows the user to request the full 65536 blocks. Don't byteswap the reftag, apptag and appmask. Martin Petersen tells me these are calculated in big-endian and are transmitted to the device in big-endian. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 27 +++++++++++++++++---------- include/linux/nvme.h | 8 +++----- 2 files changed, 20 insertions(+), 15 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index d0b52622e261..90a96ec8a596 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1035,29 +1035,37 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) struct nvme_user_io io; struct nvme_command c; unsigned length; - u32 result; int nents, status; struct scatterlist *sg; struct nvme_prps *prps; if (copy_from_user(&io, uio, sizeof(io))) return -EFAULT; - length = io.nblocks << io.block_shift; - nents = nvme_map_user_pages(dev, io.opcode & 1, io.addr, length, &sg); + length = (io.nblocks + 1) << ns->lba_shift; + + switch (io.opcode) { + case nvme_cmd_write: + case nvme_cmd_read: + nents = nvme_map_user_pages(dev, io.opcode & 1, io.addr, + length, &sg); + default: + return -EFAULT; + } + if (nents < 0) return nents; memset(&c, 0, sizeof(c)); c.rw.opcode = io.opcode; c.rw.flags = io.flags; - c.rw.nsid = cpu_to_le32(io.nsid); + c.rw.nsid = cpu_to_le32(ns->ns_id); c.rw.slba = cpu_to_le64(io.slba); - c.rw.length = cpu_to_le16(io.nblocks - 1); + c.rw.length = cpu_to_le16(io.nblocks); c.rw.control = cpu_to_le16(io.control); c.rw.dsmgmt = cpu_to_le16(io.dsmgmt); - c.rw.reftag = cpu_to_le32(io.reftag); /* XXX: endian? */ - c.rw.apptag = cpu_to_le16(io.apptag); - c.rw.appmask = cpu_to_le16(io.appmask); + c.rw.reftag = io.reftag; + c.rw.apptag = io.apptag; + c.rw.appmask = io.appmask; /* XXX: metadata */ prps = nvme_setup_prps(dev, &c.common, sg, length); @@ -1069,11 +1077,10 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) * additional races since q_lock already protects against other CPUs. */ put_nvmeq(nvmeq); - status = nvme_submit_sync_cmd(nvmeq, &c, &result, IO_TIMEOUT); + status = nvme_submit_sync_cmd(nvmeq, &c, NULL, IO_TIMEOUT); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); nvme_free_prps(dev, prps); - put_user(result, &uio->result); return status; } diff --git a/include/linux/nvme.h b/include/linux/nvme.h index fd10d597cca7..347ad5f9a721 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -373,17 +373,15 @@ struct nvme_user_io { __u8 opcode; __u8 flags; __u16 control; - __u32 nsid; + __u16 nblocks; + __u16 rsvd; __u64 metadata; __u64 addr; __u64 slba; - __u16 nblocks; - __u16 block_shift; __u32 dsmgmt; __u32 reftag; __u16 apptag; __u16 appmask; - __u32 result; }; struct nvme_dlfw { @@ -395,7 +393,7 @@ struct nvme_dlfw { #define NVME_IOCTL_IDENTIFY_NS _IOW('N', 0x40, struct nvme_id_ns) #define NVME_IOCTL_IDENTIFY_CTRL _IOW('N', 0x41, struct nvme_id_ctrl) #define NVME_IOCTL_GET_RANGE_TYPE _IOW('N', 0x42, struct nvme_lba_range_type) -#define NVME_IOCTL_SUBMIT_IO _IOWR('N', 0x43, struct nvme_user_io) +#define NVME_IOCTL_SUBMIT_IO _IOW('N', 0x43, struct nvme_user_io) #define NVME_IOCTL_DOWNLOAD_FW _IOW('N', 0x44, struct nvme_dlfw) #define NVME_IOCTL_ACTIVATE_FW _IO('N', 0x45) From 8ef700678f65e2eef1c3a94cdedb79d757608392 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 21 Mar 2011 10:28:43 -0400 Subject: [PATCH 079/820] NVMe: Version 0.5 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 90a96ec8a596..d3eeca5a3c4c 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1630,6 +1630,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.4"); +MODULE_VERSION("0.5"); module_init(nvme_init); module_exit(nvme_exit); From 7f53f9d2424533256ae86f7df5661a17de743de8 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 22 Mar 2011 15:55:45 -0400 Subject: [PATCH 080/820] NVMe: Correct the Controller Configuration settings The arbitration field was extended by one bit, shifting the shutdown notification bits by one. Also, the SQ/CQ entry size was made configurable for future extensions. Reported-by: Paul Luse Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + include/linux/nvme.h | 10 ++++++---- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index d3eeca5a3c4c..014a7f6e39bc 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -905,6 +905,7 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) dev->ctrl_config = NVME_CC_ENABLE | NVME_CC_CSS_NVM; dev->ctrl_config |= (PAGE_SHIFT - 12) << NVME_CC_MPS_SHIFT; dev->ctrl_config |= NVME_CC_ARB_RR | NVME_CC_SHN_NONE; + dev->ctrl_config |= NVME_CC_IOSQES | NVME_CC_IOCQES; writel(0, &dev->bar->cc); writel(aqa, &dev->bar->aqa); diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 347ad5f9a721..9d6febb91521 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -41,10 +41,12 @@ enum { NVME_CC_MPS_SHIFT = 7, NVME_CC_ARB_RR = 0 << 11, NVME_CC_ARB_WRRU = 1 << 11, - NVME_CC_ARB_VS = 3 << 11, - NVME_CC_SHN_NONE = 0 << 13, - NVME_CC_SHN_NORMAL = 1 << 13, - NVME_CC_SHN_ABRUPT = 2 << 13, + NVME_CC_ARB_VS = 7 << 11, + NVME_CC_SHN_NONE = 0 << 14, + NVME_CC_SHN_NORMAL = 1 << 14, + NVME_CC_SHN_ABRUPT = 2 << 14, + NVME_CC_IOSQES = 6 << 16, + NVME_CC_IOCQES = 4 << 20, NVME_CSTS_RDY = 1 << 0, NVME_CSTS_CFS = 1 << 1, NVME_CSTS_SHST_NORMAL = 0 << 2, From aba2080f3f1639f9202f1a52993669844abcfb80 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sun, 27 Mar 2011 08:52:06 -0400 Subject: [PATCH 081/820] NVMe: Fix warning in free_irq We need to clear the affinity mask before calling free_irq() Reported-by: Shane Michael Matthews Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 014a7f6e39bc..bcc780ac4ec0 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -781,8 +781,10 @@ static int adapter_delete_sq(struct nvme_dev *dev, u16 sqid) static void nvme_free_queue(struct nvme_dev *dev, int qid) { struct nvme_queue *nvmeq = dev->queues[qid]; + int vector = dev->entry[nvmeq->cq_vector].vector; - free_irq(dev->entry[nvmeq->cq_vector].vector, nvmeq); + irq_set_affinity_hint(vector, NULL); + free_irq(vector, nvmeq); /* Don't tell the adapter to delete the admin queue */ if (qid) { From 22605f96810d073eb74051d0295b6577d6a6a563 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 19 Apr 2011 15:04:20 -0400 Subject: [PATCH 082/820] NVMe: Time out initialisation after a few seconds THe device reports (in its capability register) how long it will take to initialise. If that time elapses before the ready bit becomes set, conclude the device is broken and refuse to initialise it. Log a nice error message so the user knows why we did nothing. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 10 ++++++++++ include/linux/nvme.h | 2 ++ 2 files changed, 12 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index bcc780ac4ec0..57f2b33a47dd 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -893,6 +893,8 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) { int result; u32 aqa; + u64 cap; + unsigned long timeout; struct nvme_queue *nvmeq; dev->dbs = ((void __iomem *)dev->bar) + 4096; @@ -915,10 +917,18 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) writeq(nvmeq->cq_dma_addr, &dev->bar->acq); writel(dev->ctrl_config, &dev->bar->cc); + cap = readq(&dev->bar->cap); + timeout = ((NVME_CAP_TIMEOUT(cap) + 1) * HZ / 2) + jiffies; + while (!(readl(&dev->bar->csts) & NVME_CSTS_RDY)) { msleep(100); if (fatal_signal_pending(current)) return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(&dev->pci_dev->dev, + "Device not ready; aborting initialisation\n"); + return -ENODEV; + } } result = queue_request_irq(dev, nvmeq, "nvme admin"); diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 9d6febb91521..a19304fefa7d 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -35,6 +35,8 @@ struct nvme_bar { __u64 acq; /* Admin CQ Base Address */ }; +#define NVME_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) + enum { NVME_CC_ENABLE = 1 << 0, NVME_CC_CSS_NVM = 0 << 4, From 09a58f536436efed02ead722e835cb4ce7674afc Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 28 Apr 2011 23:09:09 -0700 Subject: [PATCH 083/820] NVMe: Fix bug in error handling When an I/O completed with an error, we would call bio_endio twice (once with -EIO and once with 0). Found by inspection. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 57f2b33a47dd..205405e7f6b0 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -310,9 +310,9 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, dma_unmap_sg(nvmeq->q_dmadev, nbio->sg, nbio->nents, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); free_nbio(nvmeq, nbio); - if (status) + if (status) { bio_endio(bio, -EIO); - if (bio->bi_vcnt > bio->bi_idx) { + } else if (bio->bi_vcnt > bio->bi_idx) { bio_list_add(&nvmeq->sq_cong, bio); wake_up_process(nvme_thread); } else { From 21075bdee0a6f56058920d889df4ae561bfed754 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 28 Apr 2011 23:17:36 -0700 Subject: [PATCH 084/820] NVMe: Rename cancel_cmdid_data to cancel_cmdid The trailing '_data' on the end was annoying and inconsistent. Also, make it actually return the data since this is needed for timing out commands. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 205405e7f6b0..9ca9db903ceb 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -213,10 +213,13 @@ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) return data; } -static void cancel_cmdid_data(struct nvme_queue *nvmeq, int cmdid) +static unsigned long cancel_cmdid(struct nvme_queue *nvmeq, int cmdid) { + unsigned long data; struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); + data = info[cmdid].ctx; info[cmdid].ctx = CMD_CTX_CANCELLED; + return data; } static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) @@ -667,7 +670,7 @@ static irqreturn_t nvme_irq_check(int irq, void *data) static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid) { spin_lock_irq(&nvmeq->q_lock); - cancel_cmdid_data(nvmeq, cmdid); + cancel_cmdid(nvmeq, cmdid); spin_unlock_irq(&nvmeq->q_lock); } From 8de055350fbaa96b6563892c195a60be583faa9c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 12 May 2011 13:50:28 -0400 Subject: [PATCH 085/820] NVMe: Add support for timing out I/Os In the kthread, walk the list of outstanding I/Os and check they've not hit the timeout. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 37 +++++++++++++++++++++++++++++++------ 1 file changed, 31 insertions(+), 6 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 9ca9db903ceb..9c0ab2af0fae 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -18,6 +18,7 @@ #include #include +#include #include #include #include @@ -601,15 +602,15 @@ static void sync_completion(struct nvme_queue *nvmeq, void *ctx, typedef void (*completion_fn)(struct nvme_queue *, void *, struct nvme_completion *); +static const completion_fn nvme_completions[4] = { + [sync_completion_id] = sync_completion, + [bio_completion_id] = bio_completion, +}; + static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) { u16 head, phase; - static const completion_fn completions[4] = { - [sync_completion_id] = sync_completion, - [bio_completion_id] = bio_completion, - }; - head = nvmeq->cq_head; phase = nvmeq->cq_phase; @@ -629,7 +630,7 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) data = free_cmdid(nvmeq, cqe.command_id); handler = data & 3; ptr = (void *)(data & ~3UL); - completions[handler](nvmeq, ptr, &cqe); + nvme_completions[handler](nvmeq, ptr, &cqe); } /* If the controller ignores the cq head doorbell and continuously @@ -1172,6 +1173,29 @@ static const struct block_device_operations nvme_fops = { .compat_ioctl = nvme_ioctl, }; +static void nvme_timeout_ios(struct nvme_queue *nvmeq) +{ + int depth = nvmeq->q_depth - 1; + struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); + unsigned long now = jiffies; + int cmdid; + + for_each_set_bit(cmdid, nvmeq->cmdid_data, depth) { + unsigned long data; + void *ptr; + unsigned char handler; + static struct nvme_completion cqe = { .status = cpu_to_le16(NVME_SC_ABORT_REQ) << 1, }; + + if (!time_after(now, info[cmdid].timeout)) + continue; + dev_warn(nvmeq->q_dmadev, "Timing out I/O %d\n", cmdid); + data = cancel_cmdid(nvmeq, cmdid); + handler = data & 3; + ptr = (void *)(data & ~3UL); + nvme_completions[handler](nvmeq, ptr, &cqe); + } +} + static void nvme_resubmit_bios(struct nvme_queue *nvmeq) { while (bio_list_peek(&nvmeq->sq_cong)) { @@ -1203,6 +1227,7 @@ static int nvme_kthread(void *data) spin_lock_irq(&nvmeq->q_lock); if (nvme_process_cq(nvmeq)) printk("process_cq did something\n"); + nvme_timeout_ios(nvmeq); nvme_resubmit_bios(nvmeq); spin_unlock_irq(&nvmeq->q_lock); } From fd63e9ceeeae58cfe877c2d49d41c1bf7532303c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 6 May 2011 08:37:54 -0400 Subject: [PATCH 086/820] NVMe: Add include of delay.h for msleep Previously it was being implicitly included through some other header file Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 9c0ab2af0fae..b285a7e0624d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include #include From 5aff9382ddc8aac6eb0c70ffbb351652d71da69a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 6 May 2011 08:45:47 -0400 Subject: [PATCH 087/820] NVMe: Use an IDA to allocate minor numbers The current approach of using the namespace ID as the minor number doesn't work when there are multiple adapters in the machine. Rather than statically partitioning the number of namespaces between adapters, dynamically allocate minor numbers to namespaces as they are detected. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 38 ++++++++++++++++++++++++++++++++++---- 1 file changed, 34 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b285a7e0624d..79012c53ae9c 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include #include @@ -1240,7 +1241,34 @@ static int nvme_kthread(void *data) return 0; } -static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int index, +static DEFINE_IDA(nvme_index_ida); + +static int nvme_get_ns_idx(void) +{ + int index, error; + + do { + if (!ida_pre_get(&nvme_index_ida, GFP_KERNEL)) + return -1; + + spin_lock(&dev_list_lock); + error = ida_get_new(&nvme_index_ida, &index); + spin_unlock(&dev_list_lock); + } while (error == -EAGAIN); + + if (error) + index = -1; + return index; +} + +static void nvme_put_ns_idx(int index) +{ + spin_lock(&dev_list_lock); + ida_remove(&nvme_index_ida, index); + spin_unlock(&dev_list_lock); +} + +static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int nsid, struct nvme_id_ns *id, struct nvme_lba_range_type *rt) { struct nvme_ns *ns; @@ -1265,19 +1293,19 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int index, disk = alloc_disk(NVME_MINORS); if (!disk) goto out_free_queue; - ns->ns_id = index; + ns->ns_id = nsid; ns->disk = disk; lbaf = id->flbas & 0xf; ns->lba_shift = id->lbaf[lbaf].ds; disk->major = nvme_major; disk->minors = NVME_MINORS; - disk->first_minor = NVME_MINORS * index; + disk->first_minor = NVME_MINORS * nvme_get_ns_idx(); disk->fops = &nvme_fops; disk->private_data = ns; disk->queue = ns->queue; disk->driverfs_dev = &dev->pci_dev->dev; - sprintf(disk->disk_name, "nvme%dn%d", dev->instance, index); + sprintf(disk->disk_name, "nvme%dn%d", dev->instance, nsid); set_capacity(disk, le64_to_cpup(&id->nsze) << (ns->lba_shift - 9)); return ns; @@ -1291,7 +1319,9 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int index, static void nvme_ns_free(struct nvme_ns *ns) { + int index = ns->disk->first_minor / NVME_MINORS; put_disk(ns->disk); + nvme_put_ns_idx(index); blk_cleanup_queue(ns->queue); kfree(ns); } From b77954cbddff28d55a36fad3c16f4daebb0f01df Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 12 May 2011 13:51:41 -0400 Subject: [PATCH 088/820] NVMe: Handle failures from memory allocations in nvme_setup_prps If any of the memory allocations in nvme_setup_prps fail, handle it by modifying the passed-in data length to reflect the number of bytes we are actually able to send. Also allow the caller to specify the GFP flags they need; for user-initiated commands, we can use GFP_KERNEL allocations. The various callers are updated to handle this possibility; the main I/O path is already prepared for this possibility (as it may happen due to nvme_map_bio being unable to map all the segments of the I/O). The other callers return -ENOMEM instead of doing partial I/Os. Reported-by: Andi Kleen Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 56 ++++++++++++++++++++++++++++++++------------ 1 file changed, 41 insertions(+), 15 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 79012c53ae9c..ddc21ba24a70 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -329,9 +329,11 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, /* length is in bytes */ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, struct nvme_common_command *cmd, - struct scatterlist *sg, int length) + struct scatterlist *sg, int *len, + gfp_t gfp) { struct dma_pool *pool; + int length = *len; int dma_len = sg_dma_len(sg); u64 dma_addr = sg_dma_address(sg); int offset = offset_in_page(dma_addr); @@ -361,7 +363,12 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, nprps = DIV_ROUND_UP(length, PAGE_SIZE); npages = DIV_ROUND_UP(8 * nprps, PAGE_SIZE); - prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, GFP_ATOMIC); + prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, gfp); + if (!prps) { + cmd->prp2 = cpu_to_le64(dma_addr); + *len = (*len - length) + PAGE_SIZE; + return prps; + } prp_page = 0; if (nprps <= (256 / 8)) { pool = dev->prp_small_pool; @@ -371,7 +378,13 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, prps->npages = npages; } - prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + prp_list = dma_pool_alloc(pool, gfp, &prp_dma); + if (!prp_list) { + cmd->prp2 = cpu_to_le64(dma_addr); + *len = (*len - length) + PAGE_SIZE; + kfree(prps); + return NULL; + } prps->list[prp_page++] = prp_list; prps->first_dma = prp_dma; cmd->prp2 = cpu_to_le64(prp_dma); @@ -379,7 +392,11 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, for (;;) { if (i == PAGE_SIZE / 8) { __le64 *old_prp_list = prp_list; - prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + prp_list = dma_pool_alloc(pool, gfp, &prp_dma); + if (!prp_list) { + *len = (*len - length); + return prps; + } prps->list[prp_page++] = prp_list; prp_list[0] = old_prp_list[i - 1]; old_prp_list[i - 1] = cpu_to_le64(prp_dma); @@ -525,7 +542,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); nbio->prps = nvme_setup_prps(nvmeq->dev, &cmnd->common, nbio->sg, - length); + &length, GFP_ATOMIC); cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); cmnd->rw.length = cpu_to_le16((length >> ns->lba_shift) - 1); cmnd->rw.control = cpu_to_le16(control); @@ -1009,15 +1026,18 @@ static int nvme_submit_user_admin_command(struct nvme_dev *dev, unsigned long addr, unsigned length, struct nvme_command *cmd) { - int err, nents; + int err, nents, tmplen = length; struct scatterlist *sg; struct nvme_prps *prps; nents = nvme_map_user_pages(dev, 0, addr, length, &sg); if (nents < 0) return nents; - prps = nvme_setup_prps(dev, &cmd->common, sg, length); - err = nvme_submit_admin_cmd(dev, cmd, NULL); + prps = nvme_setup_prps(dev, &cmd->common, sg, &tmplen, GFP_KERNEL); + if (tmplen != length) + err = -ENOMEM; + else + err = nvme_submit_admin_cmd(dev, cmd, NULL); nvme_unmap_user_pages(dev, 0, addr, length, sg, nents); nvme_free_prps(dev, prps); return err ? -EIO : 0; @@ -1086,7 +1106,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) c.rw.apptag = io.apptag; c.rw.appmask = io.appmask; /* XXX: metadata */ - prps = nvme_setup_prps(dev, &c.common, sg, length); + prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); nvmeq = get_nvmeq(ns); /* @@ -1096,7 +1116,10 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) * additional races since q_lock already protects against other CPUs. */ put_nvmeq(nvmeq); - status = nvme_submit_sync_cmd(nvmeq, &c, NULL, IO_TIMEOUT); + if (length != (io.nblocks + 1) << ns->lba_shift) + status = -ENOMEM; + else + status = nvme_submit_sync_cmd(nvmeq, &c, NULL, IO_TIMEOUT); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); nvme_free_prps(dev, prps); @@ -1109,7 +1132,7 @@ static int nvme_download_firmware(struct nvme_ns *ns, struct nvme_dev *dev = ns->dev; struct nvme_dlfw dlfw; struct nvme_command c; - int nents, status; + int nents, status, length; struct scatterlist *sg; struct nvme_prps *prps; @@ -1117,8 +1140,9 @@ static int nvme_download_firmware(struct nvme_ns *ns, return -EFAULT; if (dlfw.length >= (1 << 30)) return -EINVAL; + length = dlfw.length * 4; - nents = nvme_map_user_pages(dev, 1, dlfw.addr, dlfw.length * 4, &sg); + nents = nvme_map_user_pages(dev, 1, dlfw.addr, length, &sg); if (nents < 0) return nents; @@ -1126,9 +1150,11 @@ static int nvme_download_firmware(struct nvme_ns *ns, c.dlfw.opcode = nvme_admin_download_fw; c.dlfw.numd = cpu_to_le32(dlfw.length); c.dlfw.offset = cpu_to_le32(dlfw.offset); - prps = nvme_setup_prps(dev, &c.common, sg, dlfw.length * 4); - - status = nvme_submit_admin_cmd(dev, &c, NULL); + prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); + if (length != dlfw.length * 4) + status = -ENOMEM; + else + status = nvme_submit_admin_cmd(dev, &c, NULL); nvme_unmap_user_pages(dev, 0, dlfw.addr, dlfw.length * 4, sg, nents); nvme_free_prps(dev, prps); return status; From 184d2944cb3b92a2e8e1733c59d1e531ad6e924a Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 11 May 2011 21:36:38 -0400 Subject: [PATCH 089/820] NVMe: Add a few calling convention notes For the benefit of reviewers, add comments to a few functions describing their calling context Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index ddc21ba24a70..12062c108bd9 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -155,6 +155,9 @@ static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) * the bottom two bits of the ctx pointer to store the handler ID. * Passing in a pointer that's not 4-byte aligned will cause a BUG. * We can change this if it becomes a problem. + * + * May be called with local interrupts disabled and the q_lock held, + * or with interrupts enabled and no locks held. */ static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler, unsigned timeout) @@ -202,6 +205,9 @@ enum { #define CMD_CTX_INVALID (0x314 + CMD_CTX_BASE) #define CMD_CTX_FLUSH (0x318 + CMD_CTX_BASE) +/* + * Called with local interrupts disabled and the q_lock held. May not sleep. + */ static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) { unsigned long data; @@ -326,7 +332,7 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, } } -/* length is in bytes */ +/* length is in bytes. gfp flags indicates whether we may sleep. */ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, struct nvme_common_command *cmd, struct scatterlist *sg, int *len, @@ -483,6 +489,9 @@ static int nvme_submit_flush_data(struct nvme_queue *nvmeq, struct nvme_ns *ns) return nvme_submit_flush(nvmeq, ns, cmdid); } +/* + * Called with local interrupts disabled and the q_lock held. May not sleep. + */ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct bio *bio) { From be5e09484078e95af20acb13e215cd8aec705893 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 11 May 2011 21:38:57 -0400 Subject: [PATCH 090/820] NVMe: Version 0.6 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 12062c108bd9..d1cd91becdb0 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1737,6 +1737,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.5"); +MODULE_VERSION("0.6"); module_init(nvme_init); module_exit(nvme_exit); From 6f0f54499f2edf7e25410cdd99e6f030f3485fd1 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 11 May 2011 13:30:59 -0700 Subject: [PATCH 091/820] NVMe: Return real error from nvme_create_queue nvme_setup_io_queues() was assuming that a NULL return from nvme_create_queue() was an out-of-memory error. That's not necessarily true; the adapter might return -EIO, for example. Change the calling convention to return an ERR_PTR on failure instead of NULL. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index d1cd91becdb0..843edbd79c56 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -892,7 +892,7 @@ static __devinit struct nvme_queue *nvme_create_queue(struct nvme_dev *dev, struct nvme_queue *nvmeq = nvme_alloc_queue(dev, qid, cq_size, vector); if (!nvmeq) - return NULL; + return ERR_PTR(-ENOMEM); result = adapter_alloc_cq(dev, qid, nvmeq); if (result < 0) @@ -918,7 +918,7 @@ static __devinit struct nvme_queue *nvme_create_queue(struct nvme_dev *dev, dma_free_coherent(nvmeq->q_dmadev, SQ_SIZE(nvmeq->q_depth), nvmeq->sq_cmds, nvmeq->sq_dma_addr); kfree(nvmeq); - return NULL; + return ERR_PTR(result); } static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) @@ -1421,8 +1421,8 @@ static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) for (i = 0; i < nr_io_queues; i++) { dev->queues[i + 1] = nvme_create_queue(dev, i + 1, NVME_Q_DEPTH, i); - if (!dev->queues[i + 1]) - return -ENOMEM; + if (IS_ERR(dev->queues[i + 1])) + return PTR_ERR(dev->queues[i + 1]); dev->queue_count++; } From eac623ba7a91474a688eb5d0fcd0eaa6a56dc41c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 20 May 2011 09:34:43 -0400 Subject: [PATCH 092/820] NVMe: Add the nvme thread to the wait queue before waking it up If the I/O was not completed by a single NVMe command, we add the bio to the congestion list and wake up the kthread to resubmit it. But the kthread calls remove_wait_queue() unconditionally, which will oops if it's not on the wait queue. So add the kthread to the wait queue before waking it up. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 843edbd79c56..f5e51a6116e3 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -325,6 +325,8 @@ static void bio_completion(struct nvme_queue *nvmeq, void *ctx, if (status) { bio_endio(bio, -EIO); } else if (bio->bi_vcnt > bio->bi_idx) { + if (bio_list_empty(&nvmeq->sq_cong)) + add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait); bio_list_add(&nvmeq->sq_cong, bio); wake_up_process(nvme_thread); } else { From 6bbf1acddeed0bfb345a5578f9fcada16f1e514f Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 20 May 2011 13:03:42 -0400 Subject: [PATCH 093/820] NVMe: Rework ioctls Remove the special-purpose IDENTIFY, GET_RANGE_TYPE, DOWNLOAD_FIRMWARE and ACTIVATE_FIRMWARE commands. Replace them with a generic ADMIN_CMD ioctl that can submit any admin command. Add a new ID ioctl that returns the namespace ID of the queried device. It corresponds to the SCSI Idlun ioctl. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 128 ++++++++++++++----------------------------- include/linux/nvme.h | 34 ++++++++---- 2 files changed, 63 insertions(+), 99 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index f5e51a6116e3..9e3c724b95c3 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1033,51 +1033,6 @@ static void nvme_unmap_user_pages(struct nvme_dev *dev, int write, put_page(sg_page(&sg[i])); } -static int nvme_submit_user_admin_command(struct nvme_dev *dev, - unsigned long addr, unsigned length, - struct nvme_command *cmd) -{ - int err, nents, tmplen = length; - struct scatterlist *sg; - struct nvme_prps *prps; - - nents = nvme_map_user_pages(dev, 0, addr, length, &sg); - if (nents < 0) - return nents; - prps = nvme_setup_prps(dev, &cmd->common, sg, &tmplen, GFP_KERNEL); - if (tmplen != length) - err = -ENOMEM; - else - err = nvme_submit_admin_cmd(dev, cmd, NULL); - nvme_unmap_user_pages(dev, 0, addr, length, sg, nents); - nvme_free_prps(dev, prps); - return err ? -EIO : 0; -} - -static int nvme_identify(struct nvme_ns *ns, unsigned long addr, int cns) -{ - struct nvme_command c; - - memset(&c, 0, sizeof(c)); - c.identify.opcode = nvme_admin_identify; - c.identify.nsid = cns ? 0 : cpu_to_le32(ns->ns_id); - c.identify.cns = cpu_to_le32(cns); - - return nvme_submit_user_admin_command(ns->dev, addr, 4096, &c); -} - -static int nvme_get_range_type(struct nvme_ns *ns, unsigned long addr) -{ - struct nvme_command c; - - memset(&c, 0, sizeof(c)); - c.features.opcode = nvme_admin_get_features; - c.features.nsid = cpu_to_le32(ns->ns_id); - c.features.fid = cpu_to_le32(NVME_FEAT_LBA_RANGE); - - return nvme_submit_user_admin_command(ns->dev, addr, 4096, &c); -} - static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) { struct nvme_dev *dev = ns->dev; @@ -1096,10 +1051,11 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) switch (io.opcode) { case nvme_cmd_write: case nvme_cmd_read: + case nvme_cmd_compare: nents = nvme_map_user_pages(dev, io.opcode & 1, io.addr, length, &sg); default: - return -EFAULT; + return -EINVAL; } if (nents < 0) @@ -1137,70 +1093,66 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) return status; } -static int nvme_download_firmware(struct nvme_ns *ns, - struct nvme_dlfw __user *udlfw) +static int nvme_user_admin_cmd(struct nvme_ns *ns, + struct nvme_admin_cmd __user *ucmd) { struct nvme_dev *dev = ns->dev; - struct nvme_dlfw dlfw; + struct nvme_admin_cmd cmd; struct nvme_command c; - int nents, status, length; + int status, length, nents = 0; struct scatterlist *sg; - struct nvme_prps *prps; + struct nvme_prps *prps = NULL; - if (copy_from_user(&dlfw, udlfw, sizeof(dlfw))) + if (!capable(CAP_SYS_ADMIN)) + return -EACCES; + if (copy_from_user(&cmd, ucmd, sizeof(cmd))) return -EFAULT; - if (dlfw.length >= (1 << 30)) - return -EINVAL; - length = dlfw.length * 4; - - nents = nvme_map_user_pages(dev, 1, dlfw.addr, length, &sg); - if (nents < 0) - return nents; memset(&c, 0, sizeof(c)); - c.dlfw.opcode = nvme_admin_download_fw; - c.dlfw.numd = cpu_to_le32(dlfw.length); - c.dlfw.offset = cpu_to_le32(dlfw.offset); - prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); - if (length != dlfw.length * 4) + c.common.opcode = cmd.opcode; + c.common.flags = cmd.flags; + c.common.nsid = cpu_to_le32(cmd.nsid); + c.common.cdw2[0] = cpu_to_le32(cmd.cdw2); + c.common.cdw2[1] = cpu_to_le32(cmd.cdw3); + c.common.cdw10[0] = cpu_to_le32(cmd.cdw10); + c.common.cdw10[1] = cpu_to_le32(cmd.cdw11); + c.common.cdw10[2] = cpu_to_le32(cmd.cdw12); + c.common.cdw10[3] = cpu_to_le32(cmd.cdw13); + c.common.cdw10[4] = cpu_to_le32(cmd.cdw14); + c.common.cdw10[5] = cpu_to_le32(cmd.cdw15); + + length = cmd.data_len; + if (cmd.data_len) { + nents = nvme_map_user_pages(dev, 1, cmd.addr, length, &sg); + if (nents < 0) + return nents; + prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); + } + + if (length != cmd.data_len) status = -ENOMEM; else status = nvme_submit_admin_cmd(dev, &c, NULL); - nvme_unmap_user_pages(dev, 0, dlfw.addr, dlfw.length * 4, sg, nents); - nvme_free_prps(dev, prps); + if (cmd.data_len) { + nvme_unmap_user_pages(dev, 0, cmd.addr, cmd.data_len, sg, + nents); + nvme_free_prps(dev, prps); + } return status; } -static int nvme_activate_firmware(struct nvme_ns *ns, unsigned long arg) -{ - struct nvme_dev *dev = ns->dev; - struct nvme_command c; - - memset(&c, 0, sizeof(c)); - c.common.opcode = nvme_admin_activate_fw; - c.common.rsvd10[0] = cpu_to_le32(arg); - - return nvme_submit_admin_cmd(dev, &c, NULL); -} - static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { struct nvme_ns *ns = bdev->bd_disk->private_data; switch (cmd) { - case NVME_IOCTL_IDENTIFY_NS: - return nvme_identify(ns, arg, 0); - case NVME_IOCTL_IDENTIFY_CTRL: - return nvme_identify(ns, arg, 1); - case NVME_IOCTL_GET_RANGE_TYPE: - return nvme_get_range_type(ns, arg); + case NVME_IOCTL_ID: + return ns->ns_id; + case NVME_IOCTL_ADMIN_CMD: + return nvme_user_admin_cmd(ns, (void __user *)arg); case NVME_IOCTL_SUBMIT_IO: return nvme_submit_io(ns, (void __user *)arg); - case NVME_IOCTL_DOWNLOAD_FW: - return nvme_download_firmware(ns, (void __user *)arg); - case NVME_IOCTL_ACTIVATE_FW: - return nvme_activate_firmware(ns, arg); default: return -ENOTTY; } diff --git a/include/linux/nvme.h b/include/linux/nvme.h index a19304fefa7d..c96ab0f5ef6f 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -153,11 +153,11 @@ struct nvme_common_command { __u8 flags; __u16 command_id; __le32 nsid; - __u64 rsvd2; + __u32 cdw2[2]; __le64 metadata; __le64 prp1; __le64 prp2; - __u32 rsvd10[6]; + __u32 cdw10[6]; }; struct nvme_rw_command { @@ -388,17 +388,29 @@ struct nvme_user_io { __u16 appmask; }; -struct nvme_dlfw { +struct nvme_admin_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd1; + __u32 nsid; + __u32 cdw2; + __u32 cdw3; + __u64 metadata; __u64 addr; - __u32 length; /* In dwords */ - __u32 offset; /* In dwords */ + __u32 metadata_len; + __u32 data_len; + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result; }; -#define NVME_IOCTL_IDENTIFY_NS _IOW('N', 0x40, struct nvme_id_ns) -#define NVME_IOCTL_IDENTIFY_CTRL _IOW('N', 0x41, struct nvme_id_ctrl) -#define NVME_IOCTL_GET_RANGE_TYPE _IOW('N', 0x42, struct nvme_lba_range_type) -#define NVME_IOCTL_SUBMIT_IO _IOW('N', 0x43, struct nvme_user_io) -#define NVME_IOCTL_DOWNLOAD_FW _IOW('N', 0x44, struct nvme_dlfw) -#define NVME_IOCTL_ACTIVATE_FW _IO('N', 0x45) +#define NVME_IOCTL_ID _IO('N', 0x40) +#define NVME_IOCTL_ADMIN_CMD _IOWR('N', 0x41, struct nvme_admin_cmd) +#define NVME_IOCTL_SUBMIT_IO _IOW('N', 0x42, struct nvme_user_io) #endif /* _LINUX_NVME_H */ From 6413214c5d424fd5aae6567848340f962ad2ce0f Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 9 Aug 2011 12:56:37 -0400 Subject: [PATCH 094/820] Fix bug in NVME_IOCTL_SUBMIT_IO Missing 'break' in the switch statement meant that we'd fall through to the 'return -EINVAL' case. --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 9e3c724b95c3..0956e1241520 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1054,6 +1054,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) case nvme_cmd_compare: nents = nvme_map_user_pages(dev, io.opcode & 1, io.addr, length, &sg); + break; default: return -EINVAL; } From d0ba1e497bca83a3d353eb47c9658afc54d83228 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 13 Sep 2011 17:01:39 -0400 Subject: [PATCH 095/820] NVMe: Correct sg list setup in nvme_map_user_pages Our SG list was constructed to always fill the entire first page, even if that was more than the length of the I/O. This is probably harmless, but some IOMMUs might do something bad. Correcting the first call to sg_set_page() made it look a lot closer to the sg_set_page() in the loop, so fold the first call to sg_set_page() into the loop. Reported-by: Nisheeth Bhat Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 0956e1241520..5843409cac6d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -996,11 +996,11 @@ static int nvme_map_user_pages(struct nvme_dev *dev, int write, sg = kcalloc(count, sizeof(*sg), GFP_KERNEL); sg_init_table(sg, count); - sg_set_page(&sg[0], pages[0], PAGE_SIZE - offset, offset); - length -= (PAGE_SIZE - offset); - for (i = 1; i < count; i++) { - sg_set_page(&sg[i], pages[i], min_t(int, length, PAGE_SIZE), 0); - length -= PAGE_SIZE; + for (i = 0; i < count; i++) { + sg_set_page(&sg[i], pages[i], + min_t(int, length, PAGE_SIZE - offset), offset); + length -= (PAGE_SIZE - offset); + offset = 0; } err = -ENOMEM; From d1a490e026efb22851ed60588b5fad1281d80ec3 Mon Sep 17 00:00:00 2001 From: Nisheeth Bhat Date: Thu, 15 Sep 2011 16:52:24 -0400 Subject: [PATCH 096/820] NVMe: Fix calls to dma_unmap_sg dma_unmap_sg() must be called with the same 'nents' passed to dma_map_sg(), not the number returned from dma_map_sg(). Signed-off-by: Nisheeth Bhat Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 5843409cac6d..a7f82fbdaf87 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1021,13 +1021,12 @@ static int nvme_map_user_pages(struct nvme_dev *dev, int write, } static void nvme_unmap_user_pages(struct nvme_dev *dev, int write, - unsigned long addr, int length, - struct scatterlist *sg, int nents) + unsigned long addr, int length, struct scatterlist *sg) { int i, count; count = DIV_ROUND_UP(offset_in_page(addr) + length, PAGE_SIZE); - dma_unmap_sg(&dev->pci_dev->dev, sg, nents, DMA_FROM_DEVICE); + dma_unmap_sg(&dev->pci_dev->dev, sg, count, DMA_FROM_DEVICE); for (i = 0; i < count; i++) put_page(sg_page(&sg[i])); @@ -1089,7 +1088,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) else status = nvme_submit_sync_cmd(nvmeq, &c, NULL, IO_TIMEOUT); - nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg, nents); + nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg); nvme_free_prps(dev, prps); return status; } @@ -1135,8 +1134,7 @@ static int nvme_user_admin_cmd(struct nvme_ns *ns, else status = nvme_submit_admin_cmd(dev, &c, NULL); if (cmd.data_len) { - nvme_unmap_user_pages(dev, 0, cmd.addr, cmd.data_len, sg, - nents); + nvme_unmap_user_pages(dev, 0, cmd.addr, cmd.data_len, sg); nvme_free_prps(dev, prps); } return status; From 684f5c2025b067a23722e620d0b3b858d8dc5d01 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 19 Sep 2011 17:14:53 -0400 Subject: [PATCH 097/820] NVMe: Fix memory leak in nvme_dev_add() The driver was allocating 8k of memory, then freeing 4k of it. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index a7f82fbdaf87..705f66ebd15f 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1457,7 +1457,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) list_for_each_entry(ns, &dev->namespaces, list) add_disk(ns->disk); - dma_free_coherent(&dev->pci_dev->dev, 4096, id, dma_addr); + dma_free_coherent(&dev->pci_dev->dev, 8192, id, dma_addr); return 0; out_free: @@ -1466,7 +1466,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) nvme_ns_free(ns); } - dma_free_coherent(&dev->pci_dev->dev, 4096, id, dma_addr); + dma_free_coherent(&dev->pci_dev->dev, 8192, mem, dma_addr); return res; } From bc5fc7e4b22ca855902aba02b28c96f09b446407 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Mon, 19 Sep 2011 17:08:14 -0400 Subject: [PATCH 098/820] NVMe: Create nvme_identify and nvme_get_features functions Instead of open-coding calls to nvme_submit_admin_cmd, these small wrappers are simpler to use (the patch removes 14 lines from nvme_dev_add() for example). Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 76 +++++++++++++++++++++++++------------------- 1 file changed, 43 insertions(+), 33 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 705f66ebd15f..b77894a75855 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -812,6 +812,34 @@ static int adapter_delete_sq(struct nvme_dev *dev, u16 sqid) return adapter_delete_queue(dev, nvme_admin_delete_sq, sqid); } +static int nvme_identify(struct nvme_dev *dev, unsigned nsid, unsigned cns, + dma_addr_t dma_addr) +{ + struct nvme_command c; + + memset(&c, 0, sizeof(c)); + c.identify.opcode = nvme_admin_identify; + c.identify.nsid = cpu_to_le32(nsid); + c.identify.prp1 = cpu_to_le64(dma_addr); + c.identify.cns = cpu_to_le32(cns); + + return nvme_submit_admin_cmd(dev, &c, NULL); +} + +static int nvme_get_features(struct nvme_dev *dev, unsigned fid, + unsigned dword11, dma_addr_t dma_addr, u32 *result) +{ + struct nvme_command c; + + memset(&c, 0, sizeof(c)); + c.features.opcode = nvme_admin_get_features; + c.features.prp1 = cpu_to_le64(dma_addr); + c.features.fid = cpu_to_le32(fid); + c.features.dword11 = cpu_to_le32(dword11); + + return nvme_submit_admin_cmd(dev, &c, result); +} + static void nvme_free_queue(struct nvme_dev *dev, int qid) { struct nvme_queue *nvmeq = dev->queues[qid]; @@ -1318,15 +1346,10 @@ static int set_queue_count(struct nvme_dev *dev, int count) { int status; u32 result; - struct nvme_command c; u32 q_count = (count - 1) | ((count - 1) << 16); - memset(&c, 0, sizeof(c)); - c.features.opcode = nvme_admin_get_features; - c.features.fid = cpu_to_le32(NVME_FEAT_NUM_QUEUES); - c.features.dword11 = cpu_to_le32(q_count); - - status = nvme_submit_admin_cmd(dev, &c, &result); + status = nvme_get_features(dev, NVME_FEAT_NUM_QUEUES, q_count, 0, + &result); if (status) return -EIO; return min(result & 0xffff, result >> 16) + 1; @@ -1400,65 +1423,51 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) int res, nn, i; struct nvme_ns *ns, *next; struct nvme_id_ctrl *ctrl; - void *id; + struct nvme_id_ns *id_ns; + void *mem; dma_addr_t dma_addr; - struct nvme_command cid, crt; res = nvme_setup_io_queues(dev); if (res) return res; - /* XXX: Switch to a SG list once prp2 works */ - id = dma_alloc_coherent(&dev->pci_dev->dev, 8192, &dma_addr, + mem = dma_alloc_coherent(&dev->pci_dev->dev, 8192, &dma_addr, GFP_KERNEL); - memset(&cid, 0, sizeof(cid)); - cid.identify.opcode = nvme_admin_identify; - cid.identify.nsid = 0; - cid.identify.prp1 = cpu_to_le64(dma_addr); - cid.identify.cns = cpu_to_le32(1); - - res = nvme_submit_admin_cmd(dev, &cid, NULL); + res = nvme_identify(dev, 0, 1, dma_addr); if (res) { res = -EIO; goto out_free; } - ctrl = id; + ctrl = mem; nn = le32_to_cpup(&ctrl->nn); memcpy(dev->serial, ctrl->sn, sizeof(ctrl->sn)); memcpy(dev->model, ctrl->mn, sizeof(ctrl->mn)); memcpy(dev->firmware_rev, ctrl->fr, sizeof(ctrl->fr)); - cid.identify.cns = 0; - memset(&crt, 0, sizeof(crt)); - crt.features.opcode = nvme_admin_get_features; - crt.features.prp1 = cpu_to_le64(dma_addr + 4096); - crt.features.fid = cpu_to_le32(NVME_FEAT_LBA_RANGE); - + id_ns = mem; for (i = 0; i <= nn; i++) { - cid.identify.nsid = cpu_to_le32(i); - res = nvme_submit_admin_cmd(dev, &cid, NULL); + res = nvme_identify(dev, i, 0, dma_addr); if (res) continue; - if (((struct nvme_id_ns *)id)->ncap == 0) + if (id_ns->ncap == 0) continue; - crt.features.nsid = cpu_to_le32(i); - res = nvme_submit_admin_cmd(dev, &crt, NULL); + res = nvme_get_features(dev, NVME_FEAT_LBA_RANGE, i, + dma_addr + 4096, NULL); if (res) continue; - ns = nvme_alloc_ns(dev, i, id, id + 4096); + ns = nvme_alloc_ns(dev, i, mem, mem + 4096); if (ns) list_add_tail(&ns->list, &dev->namespaces); } list_for_each_entry(ns, &dev->namespaces, list) add_disk(ns->disk); - dma_free_coherent(&dev->pci_dev->dev, 8192, id, dma_addr); - return 0; + goto out; out_free: list_for_each_entry_safe(ns, next, &dev->namespaces, list) { @@ -1466,6 +1475,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) nvme_ns_free(ns); } + out: dma_free_coherent(&dev->pci_dev->dev, 8192, mem, dma_addr); return res; } From 0d1bc9125890426b52ca2de6abedd32e31722e5c Mon Sep 17 00:00:00 2001 From: Nisheeth Bhat Date: Thu, 29 Sep 2011 10:10:10 -0400 Subject: [PATCH 099/820] Fix calculation of number of pages in a PRP List The existing calculation underestimated the number of pages required as it did not take into account the pointer at the end of each page. The replacement calculation may overestimate the number of pages required if the last page in the PRP List is entirely full. By using ->npages as a counter as we fill in the pages, we ensure that we don't try to free a page that was never allocated. Signed-off-by: Nisheeth Bhat Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b77894a75855..3afdc750aaa8 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -265,7 +265,7 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) } struct nvme_prps { - int npages; + int npages; /* 0 means small pool in use */ dma_addr_t first_dma; __le64 *list[0]; }; @@ -347,7 +347,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, int offset = offset_in_page(dma_addr); __le64 *prp_list; dma_addr_t prp_dma; - int nprps, npages, i, prp_page; + int nprps, npages, i; struct nvme_prps *prps = NULL; cmd->prp1 = cpu_to_le64(dma_addr); @@ -370,20 +370,20 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, } nprps = DIV_ROUND_UP(length, PAGE_SIZE); - npages = DIV_ROUND_UP(8 * nprps, PAGE_SIZE); + npages = DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8); prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, gfp); if (!prps) { cmd->prp2 = cpu_to_le64(dma_addr); *len = (*len - length) + PAGE_SIZE; return prps; } - prp_page = 0; + if (nprps <= (256 / 8)) { pool = dev->prp_small_pool; prps->npages = 0; } else { pool = dev->prp_page_pool; - prps->npages = npages; + prps->npages = 1; } prp_list = dma_pool_alloc(pool, gfp, &prp_dma); @@ -393,7 +393,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, kfree(prps); return NULL; } - prps->list[prp_page++] = prp_list; + prps->list[0] = prp_list; prps->first_dma = prp_dma; cmd->prp2 = cpu_to_le64(prp_dma); i = 0; @@ -405,7 +405,7 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, *len = (*len - length); return prps; } - prps->list[prp_page++] = prp_list; + prps->list[prps->npages++] = prp_list; prp_list[0] = old_prp_list[i - 1]; old_prp_list[i - 1] = cpu_to_le64(prp_dma); i = 1; From 2b2c1896871838cdf549442e8ad0264be5fa74e3 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 7 Oct 2011 13:10:13 -0400 Subject: [PATCH 100/820] NVMe: Don't probe namespace 0 ECN 001 documented that namespace 0 is not valid. Sending an Identify with CNS of 0 and Namespace of 0 is an undefined command. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 3afdc750aaa8..660aa5dfe569 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1447,7 +1447,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) memcpy(dev->firmware_rev, ctrl->fr, sizeof(ctrl->fr)); id_ns = mem; - for (i = 0; i <= nn; i++) { + for (i = 1; i <= nn; i++) { res = nvme_identify(dev, i, 0, dma_addr); if (res) continue; From ce38c149576fd0a3360fec3bef4012212d42e736 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 7 Oct 2011 13:20:37 -0400 Subject: [PATCH 101/820] NVMe: Version 0.7 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 660aa5dfe569..cfe5932821d8 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1700,6 +1700,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.6"); +MODULE_VERSION("0.7"); module_init(nvme_init); module_exit(nvme_exit); From f1938f6e1ee1583c87ec74dc406fdd8694e99ac8 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Thu, 20 Oct 2011 17:00:41 -0400 Subject: [PATCH 102/820] NVMe: Implement doorbell stride capability The doorbell stride allows devices to spread out their doorbells instead of packing them tightly. This feature was added as part of ECN 003. This patch also enables support for more than 512 queues :-) Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 17 ++++++++++++++--- include/linux/nvme.h | 1 + 2 files changed, 15 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index cfe5932821d8..a17f80fa3881 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -70,6 +70,7 @@ struct nvme_dev { struct dma_pool *prp_small_pool; int instance; int queue_count; + int db_stride; u32 ctrl_config; struct msix_entry *entry; struct nvme_bar __iomem *bar; @@ -672,7 +673,7 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) if (head == nvmeq->cq_head && phase == nvmeq->cq_phase) return IRQ_NONE; - writel(head, nvmeq->q_db + 1); + writel(head, nvmeq->q_db + (1 << nvmeq->dev->db_stride)); nvmeq->cq_head = head; nvmeq->cq_phase = phase; @@ -889,7 +890,7 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, init_waitqueue_head(&nvmeq->sq_full); init_waitqueue_entry(&nvmeq->sq_cong_wait, nvme_thread); bio_list_init(&nvmeq->sq_cong); - nvmeq->q_db = &dev->dbs[qid * 2]; + nvmeq->q_db = &dev->dbs[qid << (dev->db_stride + 1)]; nvmeq->q_depth = depth; nvmeq->cq_vector = vector; @@ -981,6 +982,7 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) cap = readq(&dev->bar->cap); timeout = ((NVME_CAP_TIMEOUT(cap) + 1) * HZ / 2) + jiffies; + dev->db_stride = NVME_CAP_STRIDE(cap); while (!(readl(&dev->bar->csts) & NVME_CSTS_RDY)) { msleep(100); @@ -1357,7 +1359,7 @@ static int set_queue_count(struct nvme_dev *dev, int count) static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) { - int result, cpu, i, nr_io_queues; + int result, cpu, i, nr_io_queues, db_bar_size; nr_io_queues = num_online_cpus(); result = set_queue_count(dev, nr_io_queues); @@ -1369,6 +1371,15 @@ static int __devinit nvme_setup_io_queues(struct nvme_dev *dev) /* Deregister the admin queue's interrupt */ free_irq(dev->entry[0].vector, dev->queues[0]); + db_bar_size = 4096 + ((nr_io_queues + 1) << (dev->db_stride + 3)); + if (db_bar_size > 8192) { + iounmap(dev->bar); + dev->bar = ioremap(pci_resource_start(dev->pci_dev, 0), + db_bar_size); + dev->dbs = ((void __iomem *)dev->bar) + 4096; + dev->queues[0]->q_db = dev->dbs; + } + for (i = 0; i < nr_io_queues; i++) dev->entry[i].entry = i; for (;;) { diff --git a/include/linux/nvme.h b/include/linux/nvme.h index c96ab0f5ef6f..2a2c535c8345 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -36,6 +36,7 @@ struct nvme_bar { }; #define NVME_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define NVME_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) enum { NVME_CC_ENABLE = 1 << 0, From 010e646ba2fdfc558048a97da746381c35836280 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 4 Nov 2011 16:24:23 -0400 Subject: [PATCH 103/820] NVMe: Update Identify Controller data structure The driver was still using an old definition of Identify Controller which only came to light once we started using the 'number of namespaces' field properly. Reported-by: Nisheeth Bhat Reported-by: Khosrow Panah Signed-off-by: Matthew Wilcox --- include/linux/nvme.h | 27 ++++++++++++++++++++++----- 1 file changed, 22 insertions(+), 5 deletions(-) diff --git a/include/linux/nvme.h b/include/linux/nvme.h index 2a2c535c8345..9490a00529f4 100644 --- a/include/linux/nvme.h +++ b/include/linux/nvme.h @@ -57,6 +57,18 @@ enum { NVME_CSTS_SHST_CMPLT = 2 << 2, }; +struct nvme_id_power_state { + __le16 max_power; /* centiwatts */ + __u16 rsvd2; + __le32 entry_lat; /* microseconds */ + __le32 exit_lat; /* microseconds */ + __u8 read_tput; + __u8 read_lat; + __u8 write_tput; + __u8 write_lat; + __u8 rsvd16[16]; +}; + #define NVME_VS(major, minor) (major << 16 | minor) struct nvme_id_ctrl { @@ -65,9 +77,11 @@ struct nvme_id_ctrl { char sn[20]; char mn[40]; char fr[8]; - __le32 nn; __u8 rab; - __u8 rsvd77[178]; + __u8 ieee[3]; + __u8 mic; + __u8 mdts; + __u8 rsvd78[178]; __le16 oacs; __u8 acl; __u8 aerl; @@ -76,15 +90,18 @@ struct nvme_id_ctrl { __u8 elpe; __u8 npss; __u8 rsvd264[248]; - __le64 psd[32]; + __u8 sqes; + __u8 cqes; + __u8 rsvd514[2]; + __le32 nn; __le16 oncs; __le16 fuses; __u8 fna; __u8 vwc; __le16 awun; __le16 awupf; - __u8 rsvd778[246]; - __u8 cmdset[2048]; + __u8 rsvd530[1518]; + struct nvme_id_power_state psd[32]; __u8 vs[1024]; }; From ed6c038c2167bee886291b62a6865f0da825e8a6 Mon Sep 17 00:00:00 2001 From: Pavel Shilovsky Date: Mon, 7 Nov 2011 16:11:24 +0300 Subject: [PATCH 104/820] CIFS: Add descriptions to the brlock cache functions Signed-off-by: Pavel Shilovsky Signed-off-by: Steve French --- fs/cifs/file.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/fs/cifs/file.c b/fs/cifs/file.c index cf0b1539b321..4dd9283885e7 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -702,6 +702,13 @@ cifs_find_lock_conflict(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock, lock->type, lock->netfid, conf_lock); } +/* + * Check if there is another lock that prevents us to set the lock (mandatory + * style). If such a lock exists, update the flock structure with its + * properties. Otherwise, set the flock type to F_UNLCK if we can cache brlocks + * or leave it the same if we can't. Returns 0 if we don't need to request to + * the server or 1 otherwise. + */ static int cifs_lock_test(struct cifsInodeInfo *cinode, __u64 offset, __u64 length, __u8 type, __u16 netfid, struct file_lock *flock) @@ -739,6 +746,12 @@ cifs_lock_add(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock) mutex_unlock(&cinode->lock_mutex); } +/* + * Set the byte-range lock (mandatory style). Returns: + * 1) 0, if we set the lock and don't need to request to the server; + * 2) 1, if no locks prevent us but we need to request to the server; + * 3) -EACCESS, if there is a lock that prevents us and wait is false. + */ static int cifs_lock_add_if(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock, bool wait) @@ -778,6 +791,13 @@ try_again: return rc; } +/* + * Check if there is another lock that prevents us to set the lock (posix + * style). If such a lock exists, update the flock structure with its + * properties. Otherwise, set the flock type to F_UNLCK if we can cache brlocks + * or leave it the same if we can't. Returns 0 if we don't need to request to + * the server or 1 otherwise. + */ static int cifs_posix_lock_test(struct file *file, struct file_lock *flock) { @@ -800,6 +820,12 @@ cifs_posix_lock_test(struct file *file, struct file_lock *flock) return rc; } +/* + * Set the byte-range lock (posix style). Returns: + * 1) 0, if we set the lock and don't need to request to the server; + * 2) 1, if we need to request to the server; + * 3) <0, if the error occurs while setting the lock. + */ static int cifs_posix_lock_set(struct file *file, struct file_lock *flock) { From 67348450b86cb1b42aa4dd55cf7cde19c2e53461 Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Mon, 17 Oct 2011 14:56:40 +0200 Subject: [PATCH 105/820] dmaengine: at_hdmac: platform data move to use .id_table We remove the use of platform data from DMA controller driver. We now use of .id_table to distinguish between compatible types. The two implementations allow to determine the number of channels and the capabilities of the controller. Signed-off-by: Nicolas Ferre Acked-by: Grant Likely Signed-off-by: Vinod Koul --- drivers/dma/at_hdmac.c | 48 +++++++++++++++++++++++++++++-------- drivers/dma/at_hdmac_regs.h | 8 +++++++ 2 files changed, 46 insertions(+), 10 deletions(-) diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index fcfa0a8b5c59..d1869c597e42 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -1175,6 +1175,18 @@ static void atc_free_chan_resources(struct dma_chan *chan) /*-- Module Management -----------------------------------------------*/ +static struct platform_device_id atdma_devtypes[] = { + { + .name = "at91sam9rl_dma", + .driver_data = ATDMA_DEVTYPE_SAM9RL, + }, { + .name = "at91sam9g45_dma", + .driver_data = ATDMA_DEVTYPE_SAM9G45, + }, { + /* sentinel */ + } +}; + /** * at_dma_off - disable DMA controller * @atdma: the Atmel HDAMC device @@ -1193,18 +1205,32 @@ static void at_dma_off(struct at_dma *atdma) static int __init at_dma_probe(struct platform_device *pdev) { - struct at_dma_platform_data *pdata; struct resource *io; struct at_dma *atdma; size_t size; int irq; int err; int i; + u32 nr_channels; + dma_cap_mask_t cap_mask = {}; + enum atdma_devtype atdmatype; - /* get DMA Controller parameters from platform */ - pdata = pdev->dev.platform_data; - if (!pdata || pdata->nr_channels > AT_DMA_MAX_NR_CHANNELS) + dma_cap_set(DMA_MEMCPY, cap_mask); + + /* get DMA parameters from controller type */ + atdmatype = platform_get_device_id(pdev)->driver_data; + + switch (atdmatype) { + case ATDMA_DEVTYPE_SAM9RL: + nr_channels = 2; + break; + case ATDMA_DEVTYPE_SAM9G45: + nr_channels = 8; + dma_cap_set(DMA_SLAVE, cap_mask); + break; + default: return -EINVAL; + } io = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (!io) @@ -1215,14 +1241,15 @@ static int __init at_dma_probe(struct platform_device *pdev) return irq; size = sizeof(struct at_dma); - size += pdata->nr_channels * sizeof(struct at_dma_chan); + size += nr_channels * sizeof(struct at_dma_chan); atdma = kzalloc(size, GFP_KERNEL); if (!atdma) return -ENOMEM; - /* discover transaction capabilites from the platform data */ - atdma->dma_common.cap_mask = pdata->cap_mask; - atdma->all_chan_mask = (1 << pdata->nr_channels) - 1; + /* discover transaction capabilities */ + atdma->dma_common.cap_mask = cap_mask; + atdma->all_chan_mask = (1 << nr_channels) - 1; + atdma->devtype = atdmatype; size = resource_size(io); if (!request_mem_region(io->start, size, pdev->dev.driver->name)) { @@ -1268,7 +1295,7 @@ static int __init at_dma_probe(struct platform_device *pdev) /* initialize channels related values */ INIT_LIST_HEAD(&atdma->dma_common.channels); - for (i = 0; i < pdata->nr_channels; i++) { + for (i = 0; i < nr_channels; i++) { struct at_dma_chan *atchan = &atdma->chan[i]; atchan->chan_common.device = &atdma->dma_common; @@ -1313,7 +1340,7 @@ static int __init at_dma_probe(struct platform_device *pdev) dev_info(&pdev->dev, "Atmel AHB DMA Controller ( %s%s), %d channels\n", dma_has_cap(DMA_MEMCPY, atdma->dma_common.cap_mask) ? "cpy " : "", dma_has_cap(DMA_SLAVE, atdma->dma_common.cap_mask) ? "slave " : "", - pdata->nr_channels); + nr_channels); dma_async_device_register(&atdma->dma_common); @@ -1495,6 +1522,7 @@ static const struct dev_pm_ops at_dma_dev_pm_ops = { static struct platform_driver at_dma_driver = { .remove = __exit_p(at_dma_remove), .shutdown = at_dma_shutdown, + .id_table = atdma_devtypes, .driver = { .name = "at_hdmac", .pm = &at_dma_dev_pm_ops, diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h index aa4c9aebab7c..d7d67372e4ff 100644 --- a/drivers/dma/at_hdmac_regs.h +++ b/drivers/dma/at_hdmac_regs.h @@ -248,9 +248,16 @@ static inline struct at_dma_chan *to_at_dma_chan(struct dma_chan *dchan) /*-- Controller ------------------------------------------------------*/ +enum atdma_devtype { + ATDMA_DEVTYPE_UNDEFINED = 0, + ATDMA_DEVTYPE_SAM9RL, /* compatible with SAM9RL DMA controller */ + ATDMA_DEVTYPE_SAM9G45, /* compatible with SAM9G45 DMA controller */ +}; + /** * struct at_dma - internal representation of an Atmel HDMA Controller * @chan_common: common dmaengine dma_device object members + * @atdma_devtype: identifier of DMA controller compatibility * @ch_regs: memory mapped register base * @clk: dma controller clock * @save_imr: interrupt mask register that is saved on suspend/resume cycle @@ -260,6 +267,7 @@ static inline struct at_dma_chan *to_at_dma_chan(struct dma_chan *dchan) */ struct at_dma { struct dma_device dma_common; + enum atdma_devtype devtype; void __iomem *regs; struct clk *clk; u32 save_imr; From c511595390a373e19a774246c659b4f563a4c3f3 Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Mon, 17 Oct 2011 14:56:41 +0200 Subject: [PATCH 106/820] dmaengine: at_hdmac: add device tree support Add device tree probe support for atmel at_hdmac DMA driver. Bindings are added to specify DMA controller configuration. Signed-off-by: Nicolas Ferre Acked-by: Grant Likely Signed-off-by: Vinod Koul --- .../devicetree/bindings/dma/atmel-dma.txt | 14 ++++++++ drivers/dma/at_hdmac.c | 32 ++++++++++++++++++- 2 files changed, 45 insertions(+), 1 deletion(-) create mode 100644 Documentation/devicetree/bindings/dma/atmel-dma.txt diff --git a/Documentation/devicetree/bindings/dma/atmel-dma.txt b/Documentation/devicetree/bindings/dma/atmel-dma.txt new file mode 100644 index 000000000000..3c046ee6e8b5 --- /dev/null +++ b/Documentation/devicetree/bindings/dma/atmel-dma.txt @@ -0,0 +1,14 @@ +* Atmel Direct Memory Access Controller (DMA) + +Required properties: +- compatible: Should be "atmel,-dma" +- reg: Should contain DMA registers location and length +- interrupts: Should contain DMA interrupt + +Examples: + +dma@ffffec00 { + compatible = "atmel,at91sam9g45-dma"; + reg = <0xffffec00 0x200>; + interrupts = <21>; +}; diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index d1869c597e42..f3cb4a009e7d 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -23,6 +23,8 @@ #include #include #include +#include +#include #include "at_hdmac_regs.h" @@ -1175,6 +1177,20 @@ static void atc_free_chan_resources(struct dma_chan *chan) /*-- Module Management -----------------------------------------------*/ +#if defined(CONFIG_OF) +static const struct of_device_id atmel_dma_dt_ids[] = { + { + .compatible = "atmel,at91sam9rl-dma", + .data = (void *)ATDMA_DEVTYPE_SAM9RL + }, { + .compatible = "atmel,at91sam9g45-dma", + .data = (void *)ATDMA_DEVTYPE_SAM9G45 + }, { /* sentinel */ } +}; + +MODULE_DEVICE_TABLE(of, atmel_dma_dt_ids); +#endif + static struct platform_device_id atdma_devtypes[] = { { .name = "at91sam9rl_dma", @@ -1187,6 +1203,19 @@ static struct platform_device_id atdma_devtypes[] = { } }; +static inline enum atdma_devtype __init at_dma_get_driver_data( + struct platform_device *pdev) +{ + if (pdev->dev.of_node) { + const struct of_device_id *match; + match = of_match_node(atmel_dma_dt_ids, pdev->dev.of_node); + if (match == NULL) + return ATDMA_DEVTYPE_UNDEFINED; + return (enum atdma_devtype)match->data; + } + return platform_get_device_id(pdev)->driver_data; +} + /** * at_dma_off - disable DMA controller * @atdma: the Atmel HDAMC device @@ -1218,7 +1247,7 @@ static int __init at_dma_probe(struct platform_device *pdev) dma_cap_set(DMA_MEMCPY, cap_mask); /* get DMA parameters from controller type */ - atdmatype = platform_get_device_id(pdev)->driver_data; + atdmatype = at_dma_get_driver_data(pdev); switch (atdmatype) { case ATDMA_DEVTYPE_SAM9RL: @@ -1526,6 +1555,7 @@ static struct platform_driver at_dma_driver = { .driver = { .name = "at_hdmac", .pm = &at_dma_dev_pm_ops, + .of_match_table = of_match_ptr(atmel_dma_dt_ids), }, }; From 035c17dac4ce1f03d6831ff403f5aea7dcb927b4 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 10 Nov 2011 09:54:42 +0100 Subject: [PATCH 107/820] ARM i.MX5: remove unnecessary includes from board files Signed-off-by: Sascha Hauer --- arch/arm/mach-mx5/board-mx53_ard.c | 1 - arch/arm/mach-mx5/board-mx53_evk.c | 1 - arch/arm/mach-mx5/board-mx53_loco.c | 1 - arch/arm/mach-mx5/board-mx53_smd.c | 1 - 4 files changed, 4 deletions(-) diff --git a/arch/arm/mach-mx5/board-mx53_ard.c b/arch/arm/mach-mx5/board-mx53_ard.c index 0d7f0fffb23a..b88a2bc4dc1d 100644 --- a/arch/arm/mach-mx5/board-mx53_ard.c +++ b/arch/arm/mach-mx5/board-mx53_ard.c @@ -32,7 +32,6 @@ #include #include -#include "crm_regs.h" #include "devices-imx53.h" #define ARD_ETHERNET_INT_B IMX_GPIO_NR(2, 31) diff --git a/arch/arm/mach-mx5/board-mx53_evk.c b/arch/arm/mach-mx5/board-mx53_evk.c index 6bea31ab8f85..c69413d7fd6e 100644 --- a/arch/arm/mach-mx5/board-mx53_evk.c +++ b/arch/arm/mach-mx5/board-mx53_evk.c @@ -37,7 +37,6 @@ #define EVK_ECSPI1_CS1 IMX_GPIO_NR(3, 19) #define MX53EVK_LED IMX_GPIO_NR(7, 7) -#include "crm_regs.h" #include "devices-imx53.h" static iomux_v3_cfg_t mx53_evk_pads[] = { diff --git a/arch/arm/mach-mx5/board-mx53_loco.c b/arch/arm/mach-mx5/board-mx53_loco.c index 7678f7734db6..e64a8f74491c 100644 --- a/arch/arm/mach-mx5/board-mx53_loco.c +++ b/arch/arm/mach-mx5/board-mx53_loco.c @@ -32,7 +32,6 @@ #include #include -#include "crm_regs.h" #include "devices-imx53.h" #define MX53_LOCO_POWER IMX_GPIO_NR(1, 8) diff --git a/arch/arm/mach-mx5/board-mx53_smd.c b/arch/arm/mach-mx5/board-mx53_smd.c index 59c0845eb4a6..d498573ca7da 100644 --- a/arch/arm/mach-mx5/board-mx53_smd.c +++ b/arch/arm/mach-mx5/board-mx53_smd.c @@ -31,7 +31,6 @@ #include #include -#include "crm_regs.h" #include "devices-imx53.h" #define SMD_FEC_PHY_RST IMX_GPIO_NR(7, 6) From dab6538e29e68af6206ec1889c847b428f290989 Mon Sep 17 00:00:00 2001 From: Boojin Kim Date: Thu, 3 Nov 2011 15:48:49 +0900 Subject: [PATCH 108/820] DMA: PL330: Fix build warning This patch adds to fix the build warning as following. drivers/dma/pl330.c: In function 'pl330_probe': drivers/dma/pl330.c:859: warning: comparison of distinct pointer types lacks a cast Signed-off-by: Boojin Kim Signed-off-by: Vinod Koul --- drivers/dma/pl330.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c index 571041477ab2..186b8221803e 100644 --- a/drivers/dma/pl330.c +++ b/drivers/dma/pl330.c @@ -856,7 +856,8 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id) INIT_LIST_HEAD(&pd->channels); /* Initialize channel parameters */ - num_chan = max(pdat ? pdat->nr_valid_peri : 0, (u8)pi->pcfg.num_chan); + num_chan = max(pdat ? (int)pdat->nr_valid_peri : 0, + (int)pi->pcfg.num_chan); pdmac->peripherals = kzalloc(num_chan * sizeof(*pch), GFP_KERNEL); for (i = 0; i < num_chan; i++) { From 9c32c63bb70b2fafc3b18bee29959c3bf245ceba Mon Sep 17 00:00:00 2001 From: Steve French Date: Thu, 10 Nov 2011 12:48:20 -0600 Subject: [PATCH 109/820] cifs: Fix sparse warning when calling cifs_strtoUCS Fix sparse endian check warning while calling cifs_strtoUCS CHECK fs/cifs/smbencrypt.c fs/cifs/smbencrypt.c:216:37: warning: incorrect type in argument 1 (different base types) fs/cifs/smbencrypt.c:216:37: expected restricted __le16 [usertype] * fs/cifs/smbencrypt.c:216:37: got unsigned short * Signed-off-by: Steve French Acked-by: Shirish Pargaonkar Date: Mon, 7 Nov 2011 12:36:48 +0100 Subject: [PATCH 110/820] ARM i.MX: Merge i.MX5 support into mach-imx This patch moves the contents of arch/arm/mach-mx5 to arch/arm/mach-imx and adjusts the Makefile/Kconfig entries in a way that it's possible to compile i.MX5 together with i.MX3/6. Signed-off-by: Sascha Hauer Tested-by: Shawn Guo Tested-by: Dirk Behme --- arch/arm/Makefile | 1 - arch/arm/mach-imx/Kconfig | 239 +++++++++++++++++ arch/arm/mach-imx/Makefile | 21 ++ arch/arm/mach-imx/Makefile.boot | 12 + .../{mach-mx5 => mach-imx}/clock-mx51-mx53.c | 2 +- .../{mach-mx5/cpu.c => mach-imx/cpu-imx5.c} | 0 arch/arm/{mach-mx5 => mach-imx}/cpu_op-mx51.c | 0 arch/arm/{mach-mx5 => mach-imx}/cpu_op-mx51.h | 0 .../crm_regs.h => mach-imx/crm-regs-imx5.h} | 0 .../{mach-mx5 => mach-imx}/devices-imx50.h | 0 .../{mach-mx5 => mach-imx}/devices-imx51.h | 0 .../{mach-mx5 => mach-imx}/devices-imx53.h | 0 arch/arm/{mach-mx5 => mach-imx}/efika.h | 0 .../{mach-mx5/ehci.c => mach-imx/ehci-imx5.c} | 0 .../eukrea_mbimx51-baseboard.c | 0 .../eukrea_mbimxsd-baseboard.c | 0 arch/arm/{mach-mx5 => mach-imx}/imx51-dt.c | 0 arch/arm/{mach-mx5 => mach-imx}/imx53-dt.c | 0 .../mach-cpuimx51.c} | 0 .../mach-cpuimx51sd.c} | 0 .../mach-mx50_rdp.c} | 0 .../mach-mx51_3ds.c} | 0 .../mach-mx51_babbage.c} | 0 .../mach-mx51_efikamx.c} | 0 .../mach-mx51_efikasb.c} | 0 .../mach-mx53_ard.c} | 0 .../mach-mx53_evk.c} | 0 .../mach-mx53_loco.c} | 0 .../mach-mx53_smd.c} | 0 .../arm/{mach-mx5/mm.c => mach-imx/mm-imx5.c} | 0 arch/arm/{mach-mx5 => mach-imx}/mx51_efika.c | 0 .../{mach-mx5/system.c => mach-imx/pm-imx5.c} | 89 ++++++- arch/arm/mach-mx5/Kconfig | 244 ------------------ arch/arm/mach-mx5/Makefile | 26 -- arch/arm/mach-mx5/Makefile.boot | 9 - arch/arm/mach-mx5/pm-imx5.c | 83 ------ arch/arm/plat-mxc/Kconfig | 15 +- 37 files changed, 356 insertions(+), 385 deletions(-) rename arch/arm/{mach-mx5 => mach-imx}/clock-mx51-mx53.c (99%) rename arch/arm/{mach-mx5/cpu.c => mach-imx/cpu-imx5.c} (100%) rename arch/arm/{mach-mx5 => mach-imx}/cpu_op-mx51.c (100%) rename arch/arm/{mach-mx5 => mach-imx}/cpu_op-mx51.h (100%) rename arch/arm/{mach-mx5/crm_regs.h => mach-imx/crm-regs-imx5.h} (100%) rename arch/arm/{mach-mx5 => mach-imx}/devices-imx50.h (100%) rename arch/arm/{mach-mx5 => mach-imx}/devices-imx51.h (100%) rename arch/arm/{mach-mx5 => mach-imx}/devices-imx53.h (100%) rename arch/arm/{mach-mx5 => mach-imx}/efika.h (100%) rename arch/arm/{mach-mx5/ehci.c => mach-imx/ehci-imx5.c} (100%) rename arch/arm/{mach-mx5 => mach-imx}/eukrea_mbimx51-baseboard.c (100%) rename arch/arm/{mach-mx5 => mach-imx}/eukrea_mbimxsd-baseboard.c (100%) rename arch/arm/{mach-mx5 => mach-imx}/imx51-dt.c (100%) rename arch/arm/{mach-mx5 => mach-imx}/imx53-dt.c (100%) rename arch/arm/{mach-mx5/board-cpuimx51.c => mach-imx/mach-cpuimx51.c} (100%) rename arch/arm/{mach-mx5/board-cpuimx51sd.c => mach-imx/mach-cpuimx51sd.c} (100%) rename arch/arm/{mach-mx5/board-mx50_rdp.c => mach-imx/mach-mx50_rdp.c} (100%) rename arch/arm/{mach-mx5/board-mx51_3ds.c => mach-imx/mach-mx51_3ds.c} (100%) rename arch/arm/{mach-mx5/board-mx51_babbage.c => mach-imx/mach-mx51_babbage.c} (100%) rename arch/arm/{mach-mx5/board-mx51_efikamx.c => mach-imx/mach-mx51_efikamx.c} (100%) rename arch/arm/{mach-mx5/board-mx51_efikasb.c => mach-imx/mach-mx51_efikasb.c} (100%) rename arch/arm/{mach-mx5/board-mx53_ard.c => mach-imx/mach-mx53_ard.c} (100%) rename arch/arm/{mach-mx5/board-mx53_evk.c => mach-imx/mach-mx53_evk.c} (100%) rename arch/arm/{mach-mx5/board-mx53_loco.c => mach-imx/mach-mx53_loco.c} (100%) rename arch/arm/{mach-mx5/board-mx53_smd.c => mach-imx/mach-mx53_smd.c} (100%) rename arch/arm/{mach-mx5/mm.c => mach-imx/mm-imx5.c} (100%) rename arch/arm/{mach-mx5 => mach-imx}/mx51_efika.c (100%) rename arch/arm/{mach-mx5/system.c => mach-imx/pm-imx5.c} (58%) delete mode 100644 arch/arm/mach-mx5/Kconfig delete mode 100644 arch/arm/mach-mx5/Makefile delete mode 100644 arch/arm/mach-mx5/Makefile.boot delete mode 100644 arch/arm/mach-mx5/pm-imx5.c diff --git a/arch/arm/Makefile b/arch/arm/Makefile index dfcf3b033e10..cf7d467267ad 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -160,7 +160,6 @@ machine-$(CONFIG_ARCH_MSM) := msm machine-$(CONFIG_ARCH_MV78XX0) := mv78xx0 machine-$(CONFIG_ARCH_IMX_V4_V5) := imx machine-$(CONFIG_ARCH_IMX_V6_V7) := imx -machine-$(CONFIG_ARCH_MX5) := mx5 machine-$(CONFIG_ARCH_MXS) := mxs machine-$(CONFIG_ARCH_NETX) := netx machine-$(CONFIG_ARCH_NOMADIK) := nomadik diff --git a/arch/arm/mach-imx/Kconfig b/arch/arm/mach-imx/Kconfig index 5f7f9c2a34ae..ef6a6b8f01ca 100644 --- a/arch/arm/mach-imx/Kconfig +++ b/arch/arm/mach-imx/Kconfig @@ -33,6 +33,18 @@ config ARCH_MX31 config ARCH_MX35 bool +config ARCH_MX5 + bool + +config ARCH_MX50 + bool + +config ARCH_MX51 + bool + +config ARCH_MX53 + bool + config SOC_IMX1 bool select ARCH_MX1 @@ -86,6 +98,32 @@ config SOC_IMX35 select MXC_AVIC select SMP_ON_UP if SMP +config SOC_IMX5 + select CPU_V7 + select ARM_L1_CACHE_SHIFT_6 + select MXC_TZIC + select ARCH_MXC_IOMUX_V3 + select ARCH_MXC_AUDMUX_V2 + select ARCH_HAS_CPUFREQ + select ARCH_MX5 + bool + +config SOC_IMX50 + bool + select SOC_IMX5 + select ARCH_MX50 + +config SOC_IMX51 + bool + select SOC_IMX5 + select ARCH_MX5 + select ARCH_MX51 + +config SOC_IMX53 + bool + select SOC_IMX5 + select ARCH_MX5 + select ARCH_MX53 if ARCH_IMX_V4_V5 @@ -604,6 +642,207 @@ config MACH_VPR200 Include support for VPR200 platform. This includes specific configurations for the board and its peripherals. +comment "i.MX5 platforms:" + +config MACH_MX50_RDP + bool "Support MX50 reference design platform" + depends on BROKEN + select SOC_IMX50 + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_SPI_IMX + help + Include support for MX50 reference design platform (RDP) board. This + includes specific configurations for the board and its peripherals. + +comment "i.MX51 machines:" + +config MACH_IMX51_DT + bool "Support i.MX51 platforms from device tree" + select SOC_IMX51 + select USE_OF + select MACH_MX51_BABBAGE + help + Include support for Freescale i.MX51 based platforms + using the device tree for discovery + +config MACH_MX51_BABBAGE + bool "Support MX51 BABBAGE platforms" + select SOC_IMX51 + select IMX_HAVE_PLATFORM_FSL_USB2_UDC + select IMX_HAVE_PLATFORM_IMX2_WDT + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_MXC_EHCI + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_SPI_IMX + help + Include support for MX51 Babbage platform, also known as MX51EVK in + u-boot. This includes specific configurations for the board and its + peripherals. + +config MACH_MX51_3DS + bool "Support MX51PDK (3DS)" + select SOC_IMX51 + select IMX_HAVE_PLATFORM_IMX2_WDT + select IMX_HAVE_PLATFORM_IMX_KEYPAD + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_SPI_IMX + select MXC_DEBUG_BOARD + help + Include support for MX51PDK (3DS) platform. This includes specific + configurations for the board and its peripherals. + +config MACH_EUKREA_CPUIMX51 + bool "Support Eukrea CPUIMX51 module" + select SOC_IMX51 + select IMX_HAVE_PLATFORM_FSL_USB2_UDC + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_MXC_EHCI + select IMX_HAVE_PLATFORM_MXC_NAND + select IMX_HAVE_PLATFORM_SPI_IMX + help + Include support for Eukrea CPUIMX51 platform. This includes + specific configurations for the module and its peripherals. + +choice + prompt "Baseboard" + depends on MACH_EUKREA_CPUIMX51 + default MACH_EUKREA_MBIMX51_BASEBOARD + +config MACH_EUKREA_MBIMX51_BASEBOARD + prompt "Eukrea MBIMX51 development board" + bool + select IMX_HAVE_PLATFORM_IMX_KEYPAD + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select LEDS_GPIO_REGISTER + help + This adds board specific devices that can be found on Eukrea's + MBIMX51 evaluation board. + +endchoice + +config MACH_EUKREA_CPUIMX51SD + bool "Support Eukrea CPUIMX51SD module" + select SOC_IMX51 + select IMX_HAVE_PLATFORM_FSL_USB2_UDC + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_MXC_EHCI + select IMX_HAVE_PLATFORM_MXC_NAND + select IMX_HAVE_PLATFORM_SPI_IMX + help + Include support for Eukrea CPUIMX51SD platform. This includes + specific configurations for the module and its peripherals. + +choice + prompt "Baseboard" + depends on MACH_EUKREA_CPUIMX51SD + default MACH_EUKREA_MBIMXSD51_BASEBOARD + +config MACH_EUKREA_MBIMXSD51_BASEBOARD + prompt "Eukrea MBIMXSD development board" + bool + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select LEDS_GPIO_REGISTER + help + This adds board specific devices that can be found on Eukrea's + MBIMXSD evaluation board. + +endchoice + +config MX51_EFIKA_COMMON + bool + select SOC_IMX51 + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_MXC_EHCI + select IMX_HAVE_PLATFORM_PATA_IMX + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_SPI_IMX + select MXC_ULPI if USB_ULPI + +config MACH_MX51_EFIKAMX + bool "Support MX51 Genesi Efika MX nettop" + select LEDS_GPIO_REGISTER + select MX51_EFIKA_COMMON + help + Include support for Genesi Efika MX nettop. This includes specific + configurations for the board and its peripherals. + +config MACH_MX51_EFIKASB + bool "Support MX51 Genesi Efika Smartbook" + select LEDS_GPIO_REGISTER + select MX51_EFIKA_COMMON + help + Include support for Genesi Efika Smartbook. This includes specific + configurations for the board and its peripherals. + +comment "i.MX53 machines:" + +config MACH_IMX53_DT + bool "Support i.MX53 platforms from device tree" + select SOC_IMX53 + select USE_OF + select MACH_MX53_ARD + select MACH_MX53_EVK + select MACH_MX53_LOCO + select MACH_MX53_SMD + help + Include support for Freescale i.MX53 based platforms + using the device tree for discovery + +config MACH_MX53_EVK + bool "Support MX53 EVK platforms" + select SOC_IMX53 + select IMX_HAVE_PLATFORM_IMX2_WDT + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_SPI_IMX + select LEDS_GPIO_REGISTER + help + Include support for MX53 EVK platform. This includes specific + configurations for the board and its peripherals. + +config MACH_MX53_SMD + bool "Support MX53 SMD platforms" + select SOC_IMX53 + select IMX_HAVE_PLATFORM_IMX2_WDT + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + help + Include support for MX53 SMD platform. This includes specific + configurations for the board and its peripherals. + +config MACH_MX53_LOCO + bool "Support MX53 LOCO platforms" + select SOC_IMX53 + select IMX_HAVE_PLATFORM_IMX2_WDT + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_GPIO_KEYS + select LEDS_GPIO_REGISTER + help + Include support for MX53 LOCO platform. This includes specific + configurations for the board and its peripherals. + +config MACH_MX53_ARD + bool "Support MX53 ARD platforms" + select SOC_IMX53 + select IMX_HAVE_PLATFORM_IMX2_WDT + select IMX_HAVE_PLATFORM_IMX_I2C + select IMX_HAVE_PLATFORM_IMX_UART + select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX + select IMX_HAVE_PLATFORM_GPIO_KEYS + help + Include support for MX53 ARD platform. This includes specific + configurations for the board and its peripherals. + comment "i.MX6 family:" config SOC_IMX6Q diff --git a/arch/arm/mach-imx/Makefile b/arch/arm/mach-imx/Makefile index aba73214c2a8..9cf630a341e1 100644 --- a/arch/arm/mach-imx/Makefile +++ b/arch/arm/mach-imx/Makefile @@ -11,6 +11,8 @@ obj-$(CONFIG_SOC_IMX27) += clock-imx27.o mm-imx27.o ehci-imx27.o obj-$(CONFIG_SOC_IMX31) += mm-imx3.o cpu-imx31.o clock-imx31.o iomux-imx31.o ehci-imx31.o obj-$(CONFIG_SOC_IMX35) += mm-imx3.o cpu-imx35.o clock-imx35.o ehci-imx35.o +obj-$(CONFIG_SOC_IMX5) += cpu-imx5.o mm-imx5.o clock-mx51-mx53.o ehci-imx5.o pm-imx5.o cpu_op-mx51.o + # Support for CMOS sensor interface obj-$(CONFIG_MX1_VIDEO) += mx1-camera-fiq.o mx1-camera-fiq-ksym.o @@ -71,3 +73,22 @@ obj-$(CONFIG_SMP) += platsmp.o obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o obj-$(CONFIG_LOCAL_TIMERS) += localtimer.o obj-$(CONFIG_SOC_IMX6Q) += clock-imx6q.o mach-imx6q.o pm-imx6q.o + +# i.MX5 based machines +obj-$(CONFIG_MACH_MX51_BABBAGE) += mach-mx51_babbage.o +obj-$(CONFIG_MACH_MX51_3DS) += mach-mx51_3ds.o +obj-$(CONFIG_MACH_MX53_EVK) += mach-mx53_evk.o +obj-$(CONFIG_MACH_MX53_SMD) += mach-mx53_smd.o +obj-$(CONFIG_MACH_MX53_LOCO) += mach-mx53_loco.o +obj-$(CONFIG_MACH_MX53_ARD) += mach-mx53_ard.o +obj-$(CONFIG_MACH_EUKREA_CPUIMX51) += mach-cpuimx51.o +obj-$(CONFIG_MACH_EUKREA_MBIMX51_BASEBOARD) += eukrea_mbimx51-baseboard.o +obj-$(CONFIG_MACH_EUKREA_CPUIMX51SD) += mach-cpuimx51sd.o +obj-$(CONFIG_MACH_EUKREA_MBIMXSD51_BASEBOARD) += eukrea_mbimxsd-baseboard.o +obj-$(CONFIG_MX51_EFIKA_COMMON) += mx51_efika.o +obj-$(CONFIG_MACH_MX51_EFIKAMX) += mach-mx51_efikamx.o +obj-$(CONFIG_MACH_MX51_EFIKASB) += mach-mx51_efikasb.o +obj-$(CONFIG_MACH_MX50_RDP) += mach-mx50_rdp.o + +obj-$(CONFIG_MACH_IMX51_DT) += imx51-dt.o +obj-$(CONFIG_MACH_IMX53_DT) += imx53-dt.o diff --git a/arch/arm/mach-imx/Makefile.boot b/arch/arm/mach-imx/Makefile.boot index 22d85889f622..2c12b7889e97 100644 --- a/arch/arm/mach-imx/Makefile.boot +++ b/arch/arm/mach-imx/Makefile.boot @@ -18,6 +18,18 @@ zreladdr-$(CONFIG_ARCH_MX3) += 0x80008000 params_phys-$(CONFIG_ARCH_MX3) := 0x80000100 initrd_phys-$(CONFIG_ARCH_MX3) := 0x80800000 +zreladdr-$(CONFIG_SOC_IMX50) += 0x70008000 +params_phys-$(CONFIG_SOC_IMX50) := 0x70000100 +initrd_phys-$(CONFIG_SOC_IMX50) := 0x70800000 + +zreladdr-$(CONFIG_SOC_IMX51) += 0x90008000 +params_phys-$(CONFIG_SOC_IMX51) := 0x90000100 +initrd_phys-$(CONFIG_SOC_IMX51) := 0x90800000 + +zreladdr-$(CONFIG_SOC_IMX53) += 0x70008000 +params_phys-$(CONFIG_SOC_IMX53) := 0x70000100 +initrd_phys-$(CONFIG_SOC_IMX53) := 0x70800000 + zreladdr-$(CONFIG_SOC_IMX6Q) += 0x10008000 params_phys-$(CONFIG_SOC_IMX6Q) := 0x10000100 initrd_phys-$(CONFIG_SOC_IMX6Q) := 0x10800000 diff --git a/arch/arm/mach-mx5/clock-mx51-mx53.c b/arch/arm/mach-imx/clock-mx51-mx53.c similarity index 99% rename from arch/arm/mach-mx5/clock-mx51-mx53.c rename to arch/arm/mach-imx/clock-mx51-mx53.c index 2aacf41c48e7..a2c654dce39a 100644 --- a/arch/arm/mach-mx5/clock-mx51-mx53.c +++ b/arch/arm/mach-imx/clock-mx51-mx53.c @@ -23,7 +23,7 @@ #include #include -#include "crm_regs.h" +#include "crm-regs-imx5.h" /* External clock values passed-in by the board code */ static unsigned long external_high_reference, external_low_reference; diff --git a/arch/arm/mach-mx5/cpu.c b/arch/arm/mach-imx/cpu-imx5.c similarity index 100% rename from arch/arm/mach-mx5/cpu.c rename to arch/arm/mach-imx/cpu-imx5.c diff --git a/arch/arm/mach-mx5/cpu_op-mx51.c b/arch/arm/mach-imx/cpu_op-mx51.c similarity index 100% rename from arch/arm/mach-mx5/cpu_op-mx51.c rename to arch/arm/mach-imx/cpu_op-mx51.c diff --git a/arch/arm/mach-mx5/cpu_op-mx51.h b/arch/arm/mach-imx/cpu_op-mx51.h similarity index 100% rename from arch/arm/mach-mx5/cpu_op-mx51.h rename to arch/arm/mach-imx/cpu_op-mx51.h diff --git a/arch/arm/mach-mx5/crm_regs.h b/arch/arm/mach-imx/crm-regs-imx5.h similarity index 100% rename from arch/arm/mach-mx5/crm_regs.h rename to arch/arm/mach-imx/crm-regs-imx5.h diff --git a/arch/arm/mach-mx5/devices-imx50.h b/arch/arm/mach-imx/devices-imx50.h similarity index 100% rename from arch/arm/mach-mx5/devices-imx50.h rename to arch/arm/mach-imx/devices-imx50.h diff --git a/arch/arm/mach-mx5/devices-imx51.h b/arch/arm/mach-imx/devices-imx51.h similarity index 100% rename from arch/arm/mach-mx5/devices-imx51.h rename to arch/arm/mach-imx/devices-imx51.h diff --git a/arch/arm/mach-mx5/devices-imx53.h b/arch/arm/mach-imx/devices-imx53.h similarity index 100% rename from arch/arm/mach-mx5/devices-imx53.h rename to arch/arm/mach-imx/devices-imx53.h diff --git a/arch/arm/mach-mx5/efika.h b/arch/arm/mach-imx/efika.h similarity index 100% rename from arch/arm/mach-mx5/efika.h rename to arch/arm/mach-imx/efika.h diff --git a/arch/arm/mach-mx5/ehci.c b/arch/arm/mach-imx/ehci-imx5.c similarity index 100% rename from arch/arm/mach-mx5/ehci.c rename to arch/arm/mach-imx/ehci-imx5.c diff --git a/arch/arm/mach-mx5/eukrea_mbimx51-baseboard.c b/arch/arm/mach-imx/eukrea_mbimx51-baseboard.c similarity index 100% rename from arch/arm/mach-mx5/eukrea_mbimx51-baseboard.c rename to arch/arm/mach-imx/eukrea_mbimx51-baseboard.c diff --git a/arch/arm/mach-mx5/eukrea_mbimxsd-baseboard.c b/arch/arm/mach-imx/eukrea_mbimxsd-baseboard.c similarity index 100% rename from arch/arm/mach-mx5/eukrea_mbimxsd-baseboard.c rename to arch/arm/mach-imx/eukrea_mbimxsd-baseboard.c diff --git a/arch/arm/mach-mx5/imx51-dt.c b/arch/arm/mach-imx/imx51-dt.c similarity index 100% rename from arch/arm/mach-mx5/imx51-dt.c rename to arch/arm/mach-imx/imx51-dt.c diff --git a/arch/arm/mach-mx5/imx53-dt.c b/arch/arm/mach-imx/imx53-dt.c similarity index 100% rename from arch/arm/mach-mx5/imx53-dt.c rename to arch/arm/mach-imx/imx53-dt.c diff --git a/arch/arm/mach-mx5/board-cpuimx51.c b/arch/arm/mach-imx/mach-cpuimx51.c similarity index 100% rename from arch/arm/mach-mx5/board-cpuimx51.c rename to arch/arm/mach-imx/mach-cpuimx51.c diff --git a/arch/arm/mach-mx5/board-cpuimx51sd.c b/arch/arm/mach-imx/mach-cpuimx51sd.c similarity index 100% rename from arch/arm/mach-mx5/board-cpuimx51sd.c rename to arch/arm/mach-imx/mach-cpuimx51sd.c diff --git a/arch/arm/mach-mx5/board-mx50_rdp.c b/arch/arm/mach-imx/mach-mx50_rdp.c similarity index 100% rename from arch/arm/mach-mx5/board-mx50_rdp.c rename to arch/arm/mach-imx/mach-mx50_rdp.c diff --git a/arch/arm/mach-mx5/board-mx51_3ds.c b/arch/arm/mach-imx/mach-mx51_3ds.c similarity index 100% rename from arch/arm/mach-mx5/board-mx51_3ds.c rename to arch/arm/mach-imx/mach-mx51_3ds.c diff --git a/arch/arm/mach-mx5/board-mx51_babbage.c b/arch/arm/mach-imx/mach-mx51_babbage.c similarity index 100% rename from arch/arm/mach-mx5/board-mx51_babbage.c rename to arch/arm/mach-imx/mach-mx51_babbage.c diff --git a/arch/arm/mach-mx5/board-mx51_efikamx.c b/arch/arm/mach-imx/mach-mx51_efikamx.c similarity index 100% rename from arch/arm/mach-mx5/board-mx51_efikamx.c rename to arch/arm/mach-imx/mach-mx51_efikamx.c diff --git a/arch/arm/mach-mx5/board-mx51_efikasb.c b/arch/arm/mach-imx/mach-mx51_efikasb.c similarity index 100% rename from arch/arm/mach-mx5/board-mx51_efikasb.c rename to arch/arm/mach-imx/mach-mx51_efikasb.c diff --git a/arch/arm/mach-mx5/board-mx53_ard.c b/arch/arm/mach-imx/mach-mx53_ard.c similarity index 100% rename from arch/arm/mach-mx5/board-mx53_ard.c rename to arch/arm/mach-imx/mach-mx53_ard.c diff --git a/arch/arm/mach-mx5/board-mx53_evk.c b/arch/arm/mach-imx/mach-mx53_evk.c similarity index 100% rename from arch/arm/mach-mx5/board-mx53_evk.c rename to arch/arm/mach-imx/mach-mx53_evk.c diff --git a/arch/arm/mach-mx5/board-mx53_loco.c b/arch/arm/mach-imx/mach-mx53_loco.c similarity index 100% rename from arch/arm/mach-mx5/board-mx53_loco.c rename to arch/arm/mach-imx/mach-mx53_loco.c diff --git a/arch/arm/mach-mx5/board-mx53_smd.c b/arch/arm/mach-imx/mach-mx53_smd.c similarity index 100% rename from arch/arm/mach-mx5/board-mx53_smd.c rename to arch/arm/mach-imx/mach-mx53_smd.c diff --git a/arch/arm/mach-mx5/mm.c b/arch/arm/mach-imx/mm-imx5.c similarity index 100% rename from arch/arm/mach-mx5/mm.c rename to arch/arm/mach-imx/mm-imx5.c diff --git a/arch/arm/mach-mx5/mx51_efika.c b/arch/arm/mach-imx/mx51_efika.c similarity index 100% rename from arch/arm/mach-mx5/mx51_efika.c rename to arch/arm/mach-imx/mx51_efika.c diff --git a/arch/arm/mach-mx5/system.c b/arch/arm/mach-imx/pm-imx5.c similarity index 58% rename from arch/arm/mach-mx5/system.c rename to arch/arm/mach-imx/pm-imx5.c index 144ebebc4a61..3d57b5d429fe 100644 --- a/arch/arm/mach-mx5/system.c +++ b/arch/arm/mach-imx/pm-imx5.c @@ -1,8 +1,6 @@ /* - * Copyright (C) 2011 Freescale Semiconductor, Inc. All Rights Reserved. - */ - -/* + * Copyright (C) 2011 Freescale Semiconductor, Inc. All Rights Reserved. + * * The code contained herein is licensed under the GNU General Public * License. You may obtain a copy of the GNU General Public License * Version 2 or later at the following locations: @@ -10,14 +8,22 @@ * http://www.opensource.org/licenses/gpl-license.html * http://www.gnu.org/copyleft/gpl.html */ -#include +#include +#include #include -#include +#include +#include +#include #include -#include "crm_regs.h" +#include +#include "crm-regs-imx5.h" -/* set cpu low power mode before WFI instruction. This function is called - * mx5 because it can be used for mx50, mx51, and mx53.*/ +static struct clk *gpc_dvfs_clk; + +/* + * set cpu low power mode before WFI instruction. This function is called + * mx5 because it can be used for mx50, mx51, and mx53. + */ void mx5_cpu_lp_set(enum mxc_cpu_pwr_mode mode) { u32 plat_lpc, arm_srpgcr, ccm_clpcr; @@ -83,3 +89,68 @@ void mx5_cpu_lp_set(enum mxc_cpu_pwr_mode mode) __raw_writel(empgc1, MXC_SRPG_EMPGC1_SRPGCR); } } + +static int mx5_suspend_prepare(void) +{ + return clk_enable(gpc_dvfs_clk); +} + +static int mx5_suspend_enter(suspend_state_t state) +{ + switch (state) { + case PM_SUSPEND_MEM: + mx5_cpu_lp_set(STOP_POWER_OFF); + break; + case PM_SUSPEND_STANDBY: + mx5_cpu_lp_set(WAIT_UNCLOCKED_POWER_OFF); + break; + default: + return -EINVAL; + } + + if (state == PM_SUSPEND_MEM) { + local_flush_tlb_all(); + flush_cache_all(); + + /*clear the EMPGC0/1 bits */ + __raw_writel(0, MXC_SRPG_EMPGC0_SRPGCR); + __raw_writel(0, MXC_SRPG_EMPGC1_SRPGCR); + } + cpu_do_idle(); + return 0; +} + +static void mx5_suspend_finish(void) +{ + clk_disable(gpc_dvfs_clk); +} + +static int mx5_pm_valid(suspend_state_t state) +{ + return (state > PM_SUSPEND_ON && state <= PM_SUSPEND_MAX); +} + +static const struct platform_suspend_ops mx5_suspend_ops = { + .valid = mx5_pm_valid, + .prepare = mx5_suspend_prepare, + .enter = mx5_suspend_enter, + .finish = mx5_suspend_finish, +}; + +static int __init mx5_pm_init(void) +{ + if (!cpu_is_mx51() && !cpu_is_mx53()) + return 0; + + if (gpc_dvfs_clk == NULL) + gpc_dvfs_clk = clk_get(NULL, "gpc_dvfs"); + + if (!IS_ERR(gpc_dvfs_clk)) { + if (cpu_is_mx51()) + suspend_set_ops(&mx5_suspend_ops); + } else + return -EPERM; + + return 0; +} +device_initcall(mx5_pm_init); diff --git a/arch/arm/mach-mx5/Kconfig b/arch/arm/mach-mx5/Kconfig deleted file mode 100644 index af0c212e3c7b..000000000000 --- a/arch/arm/mach-mx5/Kconfig +++ /dev/null @@ -1,244 +0,0 @@ -if ARCH_MX5 - -# ARCH_MX5/50/53 are left to mark places where prevent multi-soc in single -# image. So for most time, SOC_IMX50/51/53 should be used. - -config ARCH_MX51 - bool - -config ARCH_MX50 - bool - -config ARCH_MX53 - bool - -config SOC_IMX50 - bool - select CPU_V7 - select ARM_L1_CACHE_SHIFT_6 - select MXC_TZIC - select ARCH_MXC_IOMUX_V3 - select ARCH_MXC_AUDMUX_V2 - select ARCH_HAS_CPUFREQ - select ARCH_MX50 - -config SOC_IMX51 - bool - select CPU_V7 - select ARM_L1_CACHE_SHIFT_6 - select MXC_TZIC - select ARCH_MXC_IOMUX_V3 - select ARCH_MXC_AUDMUX_V2 - select ARCH_HAS_CPUFREQ - select ARCH_MX51 - -config SOC_IMX53 - bool - select CPU_V7 - select ARM_L1_CACHE_SHIFT_6 - select MXC_TZIC - select ARCH_MXC_IOMUX_V3 - select ARCH_MX53 - -#comment "i.MX50 machines:" - -config MACH_MX50_RDP - bool "Support MX50 reference design platform" - depends on BROKEN - select SOC_IMX50 - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_SPI_IMX - help - Include support for MX50 reference design platform (RDP) board. This - includes specific configurations for the board and its peripherals. - -comment "i.MX51 machines:" - -config MACH_IMX51_DT - bool "Support i.MX51 platforms from device tree" - select SOC_IMX51 - select USE_OF - select MACH_MX51_BABBAGE - help - Include support for Freescale i.MX51 based platforms - using the device tree for discovery - -config MACH_MX51_BABBAGE - bool "Support MX51 BABBAGE platforms" - select SOC_IMX51 - select IMX_HAVE_PLATFORM_FSL_USB2_UDC - select IMX_HAVE_PLATFORM_IMX2_WDT - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_MXC_EHCI - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_SPI_IMX - help - Include support for MX51 Babbage platform, also known as MX51EVK in - u-boot. This includes specific configurations for the board and its - peripherals. - -config MACH_MX51_3DS - bool "Support MX51PDK (3DS)" - select SOC_IMX51 - select IMX_HAVE_PLATFORM_IMX2_WDT - select IMX_HAVE_PLATFORM_IMX_KEYPAD - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_SPI_IMX - select MXC_DEBUG_BOARD - help - Include support for MX51PDK (3DS) platform. This includes specific - configurations for the board and its peripherals. - -config MACH_EUKREA_CPUIMX51 - bool "Support Eukrea CPUIMX51 module" - select SOC_IMX51 - select IMX_HAVE_PLATFORM_FSL_USB2_UDC - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_MXC_EHCI - select IMX_HAVE_PLATFORM_MXC_NAND - select IMX_HAVE_PLATFORM_SPI_IMX - help - Include support for Eukrea CPUIMX51 platform. This includes - specific configurations for the module and its peripherals. - -choice - prompt "Baseboard" - depends on MACH_EUKREA_CPUIMX51 - default MACH_EUKREA_MBIMX51_BASEBOARD - -config MACH_EUKREA_MBIMX51_BASEBOARD - prompt "Eukrea MBIMX51 development board" - bool - select IMX_HAVE_PLATFORM_IMX_KEYPAD - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select LEDS_GPIO_REGISTER - help - This adds board specific devices that can be found on Eukrea's - MBIMX51 evaluation board. - -endchoice - -config MACH_EUKREA_CPUIMX51SD - bool "Support Eukrea CPUIMX51SD module" - select SOC_IMX51 - select IMX_HAVE_PLATFORM_FSL_USB2_UDC - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_MXC_EHCI - select IMX_HAVE_PLATFORM_MXC_NAND - select IMX_HAVE_PLATFORM_SPI_IMX - help - Include support for Eukrea CPUIMX51SD platform. This includes - specific configurations for the module and its peripherals. - -choice - prompt "Baseboard" - depends on MACH_EUKREA_CPUIMX51SD - default MACH_EUKREA_MBIMXSD51_BASEBOARD - -config MACH_EUKREA_MBIMXSD51_BASEBOARD - prompt "Eukrea MBIMXSD development board" - bool - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select LEDS_GPIO_REGISTER - help - This adds board specific devices that can be found on Eukrea's - MBIMXSD evaluation board. - -endchoice - -config MX51_EFIKA_COMMON - bool - select SOC_IMX51 - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_MXC_EHCI - select IMX_HAVE_PLATFORM_PATA_IMX - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_SPI_IMX - select MXC_ULPI if USB_ULPI - -config MACH_MX51_EFIKAMX - bool "Support MX51 Genesi Efika MX nettop" - select LEDS_GPIO_REGISTER - select MX51_EFIKA_COMMON - help - Include support for Genesi Efika MX nettop. This includes specific - configurations for the board and its peripherals. - -config MACH_MX51_EFIKASB - bool "Support MX51 Genesi Efika Smartbook" - select LEDS_GPIO_REGISTER - select MX51_EFIKA_COMMON - help - Include support for Genesi Efika Smartbook. This includes specific - configurations for the board and its peripherals. - -comment "i.MX53 machines:" - -config MACH_IMX53_DT - bool "Support i.MX53 platforms from device tree" - select SOC_IMX53 - select USE_OF - select MACH_MX53_ARD - select MACH_MX53_EVK - select MACH_MX53_LOCO - select MACH_MX53_SMD - help - Include support for Freescale i.MX53 based platforms - using the device tree for discovery - -config MACH_MX53_EVK - bool "Support MX53 EVK platforms" - select SOC_IMX53 - select IMX_HAVE_PLATFORM_IMX2_WDT - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_SPI_IMX - select LEDS_GPIO_REGISTER - help - Include support for MX53 EVK platform. This includes specific - configurations for the board and its peripherals. - -config MACH_MX53_SMD - bool "Support MX53 SMD platforms" - select SOC_IMX53 - select IMX_HAVE_PLATFORM_IMX2_WDT - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - help - Include support for MX53 SMD platform. This includes specific - configurations for the board and its peripherals. - -config MACH_MX53_LOCO - bool "Support MX53 LOCO platforms" - select SOC_IMX53 - select IMX_HAVE_PLATFORM_IMX2_WDT - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_GPIO_KEYS - select LEDS_GPIO_REGISTER - help - Include support for MX53 LOCO platform. This includes specific - configurations for the board and its peripherals. - -config MACH_MX53_ARD - bool "Support MX53 ARD platforms" - select SOC_IMX53 - select IMX_HAVE_PLATFORM_IMX2_WDT - select IMX_HAVE_PLATFORM_IMX_I2C - select IMX_HAVE_PLATFORM_IMX_UART - select IMX_HAVE_PLATFORM_SDHCI_ESDHC_IMX - select IMX_HAVE_PLATFORM_GPIO_KEYS - help - Include support for MX53 ARD platform. This includes specific - configurations for the board and its peripherals. - -endif diff --git a/arch/arm/mach-mx5/Makefile b/arch/arm/mach-mx5/Makefile deleted file mode 100644 index 0fc60807fa2b..000000000000 --- a/arch/arm/mach-mx5/Makefile +++ /dev/null @@ -1,26 +0,0 @@ -# -# Makefile for the linux kernel. -# - -# Object file lists. -obj-y := cpu.o mm.o clock-mx51-mx53.o ehci.o system.o - -obj-$(CONFIG_PM) += pm-imx5.o -obj-$(CONFIG_CPU_FREQ_IMX) += cpu_op-mx51.o -obj-$(CONFIG_MACH_MX51_BABBAGE) += board-mx51_babbage.o -obj-$(CONFIG_MACH_MX51_3DS) += board-mx51_3ds.o -obj-$(CONFIG_MACH_MX53_EVK) += board-mx53_evk.o -obj-$(CONFIG_MACH_MX53_SMD) += board-mx53_smd.o -obj-$(CONFIG_MACH_MX53_LOCO) += board-mx53_loco.o -obj-$(CONFIG_MACH_MX53_ARD) += board-mx53_ard.o -obj-$(CONFIG_MACH_EUKREA_CPUIMX51) += board-cpuimx51.o -obj-$(CONFIG_MACH_EUKREA_MBIMX51_BASEBOARD) += eukrea_mbimx51-baseboard.o -obj-$(CONFIG_MACH_EUKREA_CPUIMX51SD) += board-cpuimx51sd.o -obj-$(CONFIG_MACH_EUKREA_MBIMXSD51_BASEBOARD) += eukrea_mbimxsd-baseboard.o -obj-$(CONFIG_MX51_EFIKA_COMMON) += mx51_efika.o -obj-$(CONFIG_MACH_MX51_EFIKAMX) += board-mx51_efikamx.o -obj-$(CONFIG_MACH_MX51_EFIKASB) += board-mx51_efikasb.o -obj-$(CONFIG_MACH_MX50_RDP) += board-mx50_rdp.o - -obj-$(CONFIG_MACH_IMX51_DT) += imx51-dt.o -obj-$(CONFIG_MACH_IMX53_DT) += imx53-dt.o diff --git a/arch/arm/mach-mx5/Makefile.boot b/arch/arm/mach-mx5/Makefile.boot deleted file mode 100644 index ca207ca305ec..000000000000 --- a/arch/arm/mach-mx5/Makefile.boot +++ /dev/null @@ -1,9 +0,0 @@ - zreladdr-$(CONFIG_ARCH_MX50) += 0x70008000 -params_phys-$(CONFIG_ARCH_MX50) := 0x70000100 -initrd_phys-$(CONFIG_ARCH_MX50) := 0x70800000 - zreladdr-$(CONFIG_ARCH_MX51) += 0x90008000 -params_phys-$(CONFIG_ARCH_MX51) := 0x90000100 -initrd_phys-$(CONFIG_ARCH_MX51) := 0x90800000 - zreladdr-$(CONFIG_ARCH_MX53) += 0x70008000 -params_phys-$(CONFIG_ARCH_MX53) := 0x70000100 -initrd_phys-$(CONFIG_ARCH_MX53) := 0x70800000 diff --git a/arch/arm/mach-mx5/pm-imx5.c b/arch/arm/mach-mx5/pm-imx5.c deleted file mode 100644 index 98052fc852c7..000000000000 --- a/arch/arm/mach-mx5/pm-imx5.c +++ /dev/null @@ -1,83 +0,0 @@ -/* - * Copyright (C) 2011 Freescale Semiconductor, Inc. All Rights Reserved. - * - * The code contained herein is licensed under the GNU General Public - * License. You may obtain a copy of the GNU General Public License - * Version 2 or later at the following locations: - * - * http://www.opensource.org/licenses/gpl-license.html - * http://www.gnu.org/copyleft/gpl.html - */ -#include -#include -#include -#include -#include -#include -#include -#include -#include "crm_regs.h" - -static struct clk *gpc_dvfs_clk; - -static int mx5_suspend_prepare(void) -{ - return clk_enable(gpc_dvfs_clk); -} - -static int mx5_suspend_enter(suspend_state_t state) -{ - switch (state) { - case PM_SUSPEND_MEM: - mx5_cpu_lp_set(STOP_POWER_OFF); - break; - case PM_SUSPEND_STANDBY: - mx5_cpu_lp_set(WAIT_UNCLOCKED_POWER_OFF); - break; - default: - return -EINVAL; - } - - if (state == PM_SUSPEND_MEM) { - local_flush_tlb_all(); - flush_cache_all(); - - /*clear the EMPGC0/1 bits */ - __raw_writel(0, MXC_SRPG_EMPGC0_SRPGCR); - __raw_writel(0, MXC_SRPG_EMPGC1_SRPGCR); - } - cpu_do_idle(); - return 0; -} - -static void mx5_suspend_finish(void) -{ - clk_disable(gpc_dvfs_clk); -} - -static int mx5_pm_valid(suspend_state_t state) -{ - return (state > PM_SUSPEND_ON && state <= PM_SUSPEND_MAX); -} - -static const struct platform_suspend_ops mx5_suspend_ops = { - .valid = mx5_pm_valid, - .prepare = mx5_suspend_prepare, - .enter = mx5_suspend_enter, - .finish = mx5_suspend_finish, -}; - -static int __init mx5_pm_init(void) -{ - if (gpc_dvfs_clk == NULL) - gpc_dvfs_clk = clk_get(NULL, "gpc_dvfs"); - - if (!IS_ERR(gpc_dvfs_clk)) { - if (cpu_is_mx51()) - suspend_set_ops(&mx5_suspend_ops); - } else - return -EPERM; - - return 0; -} -device_initcall(mx5_pm_init); diff --git a/arch/arm/plat-mxc/Kconfig b/arch/arm/plat-mxc/Kconfig index a08a95107a63..fdde4d13d2e8 100644 --- a/arch/arm/plat-mxc/Kconfig +++ b/arch/arm/plat-mxc/Kconfig @@ -17,25 +17,16 @@ config ARCH_IMX_V4_V5 and ARMv5 SoCs config ARCH_IMX_V6_V7 - bool "i.MX3, i.MX6" + bool "i.MX3, i.MX5, i.MX6" select AUTO_ZRELADDR if !ZBOOT_ROM select ARM_PATCH_PHYS_VIRT help - This enables support for systems based on the Freescale i.MX3 and i.MX6 - family. - -config ARCH_MX5 - bool "i.MX50, i.MX51, i.MX53" - select AUTO_ZRELADDR - select ARM_PATCH_PHYS_VIRT - help - This enables support for machines using Freescale's i.MX50 and i.MX53 - processors. + This enables support for systems based on the Freescale i.MX3, i.MX5 + and i.MX6 family. endchoice source "arch/arm/mach-imx/Kconfig" -source "arch/arm/mach-mx5/Kconfig" endmenu From 3f07f355704bc2ae26399f17c6d5d46c9a3d78d1 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Tue, 8 Nov 2011 16:40:37 +0100 Subject: [PATCH 111/820] ARM i.MX: Update defconfig This updates the defconfig for i.MX3, i.MX5 and i.MX6. - all these can be compiled together now, so remove mx3_defconfig and mx5_defconfig and introduce a imx_v6_v7_defconfig instead. Signed-off-by: Sascha Hauer --- .../{mx5_defconfig => imx_v6_v7_defconfig} | 61 ++++---- arch/arm/configs/mx3_defconfig | 144 ------------------ 2 files changed, 35 insertions(+), 170 deletions(-) rename arch/arm/configs/{mx5_defconfig => imx_v6_v7_defconfig} (80%) delete mode 100644 arch/arm/configs/mx3_defconfig diff --git a/arch/arm/configs/mx5_defconfig b/arch/arm/configs/imx_v6_v7_defconfig similarity index 80% rename from arch/arm/configs/mx5_defconfig rename to arch/arm/configs/imx_v6_v7_defconfig index d0d8dfece37e..3a4fb2e5fc68 100644 --- a/arch/arm/configs/mx5_defconfig +++ b/arch/arm/configs/imx_v6_v7_defconfig @@ -3,6 +3,7 @@ CONFIG_EXPERIMENTAL=y CONFIG_KERNEL_LZO=y CONFIG_SYSVIPC=y CONFIG_LOG_BUF_SHIFT=18 +CONFIG_CGROUPS=y CONFIG_RELAY=y CONFIG_EXPERT=y # CONFIG_SLUB_DEBUG is not set @@ -14,20 +15,31 @@ CONFIG_MODULE_SRCVERSION_ALL=y # CONFIG_LBDAF is not set # CONFIG_BLK_DEV_BSG is not set CONFIG_ARCH_MXC=y -CONFIG_ARCH_MX5=y -CONFIG_MACH_MX51_BABBAGE=y +CONFIG_MACH_MX31LILLY=y +CONFIG_MACH_MX31LITE=y +CONFIG_MACH_PCM037=y +CONFIG_MACH_PCM037_EET=y +CONFIG_MACH_MX31_3DS=y +CONFIG_MACH_MX31MOBOARD=y +CONFIG_MACH_QONG=y +CONFIG_MACH_ARMADILLO5X0=y +CONFIG_MACH_KZM_ARM11_01=y +CONFIG_MACH_PCM043=y +CONFIG_MACH_MX35_3DS=y +CONFIG_MACH_EUKREA_CPUIMX35=y +CONFIG_MACH_VPR200=y +CONFIG_MACH_IMX51_DT=y CONFIG_MACH_MX51_3DS=y CONFIG_MACH_EUKREA_CPUIMX51=y CONFIG_MACH_EUKREA_CPUIMX51SD=y CONFIG_MACH_MX51_EFIKAMX=y CONFIG_MACH_MX51_EFIKASB=y -CONFIG_MACH_MX53_EVK=y -CONFIG_MACH_MX53_SMD=y -CONFIG_MACH_MX53_LOCO=y -CONFIG_MACH_MX53_ARD=y +CONFIG_MACH_IMX53_DT=y +CONFIG_SOC_IMX6Q=y CONFIG_MXC_PWM=y CONFIG_NO_HZ=y CONFIG_HIGH_RES_TIMERS=y +CONFIG_SMP=y CONFIG_VMSPLIT_2G=y CONFIG_PREEMPT_VOLUNTARY=y CONFIG_AEABI=y @@ -49,7 +61,7 @@ CONFIG_IP_PNP_DHCP=y # CONFIG_INET_XFRM_MODE_TUNNEL is not set # CONFIG_INET_XFRM_MODE_BEET is not set # CONFIG_INET_LRO is not set -# CONFIG_IPV6 is not set +CONFIG_IPV6=y # CONFIG_WIRELESS is not set CONFIG_DEVTMPFS=y CONFIG_DEVTMPFS_MOUNT=y @@ -68,24 +80,20 @@ CONFIG_SCSI_SCAN_ASYNC=y CONFIG_ATA=y CONFIG_PATA_IMX=y CONFIG_NETDEVICES=y -CONFIG_MII=m -CONFIG_MARVELL_PHY=y -CONFIG_DAVICOM_PHY=y -CONFIG_QSEMI_PHY=y -CONFIG_LXT_PHY=y -CONFIG_CICADA_PHY=y -CONFIG_VITESSE_PHY=y -CONFIG_SMSC_PHY=y -CONFIG_BROADCOM_PHY=y -CONFIG_ICPLUS_PHY=y -CONFIG_REALTEK_PHY=y -CONFIG_NATIONAL_PHY=y -CONFIG_STE10XP=y -CONFIG_LSI_ET1011C_PHY=y -CONFIG_MICREL_PHY=y -CONFIG_NET_ETHERNET=y -# CONFIG_NETDEV_1000 is not set -# CONFIG_NETDEV_10000 is not set +# CONFIG_NET_VENDOR_BROADCOM is not set +# CONFIG_NET_VENDOR_CHELSIO is not set +# CONFIG_NET_VENDOR_FARADAY is not set +CONFIG_FEC=y +# CONFIG_NET_VENDOR_INTEL is not set +# CONFIG_NET_VENDOR_MARVELL is not set +# CONFIG_NET_VENDOR_MICREL is not set +# CONFIG_NET_VENDOR_MICROCHIP is not set +# CONFIG_NET_VENDOR_NATSEMI is not set +# CONFIG_NET_VENDOR_SEEQ is not set +CONFIG_SMC91X=y +CONFIG_SMC911X=y +CONFIG_SMSC911X=y +# CONFIG_NET_VENDOR_STMICRO is not set # CONFIG_WLAN is not set # CONFIG_INPUT_MOUSEDEV_PSAUX is not set CONFIG_INPUT_EVDEV=y @@ -124,7 +132,6 @@ CONFIG_USB_EHCI_HCD=y CONFIG_USB_EHCI_MXC=y CONFIG_USB_STORAGE=y CONFIG_MMC=y -CONFIG_MMC_BLOCK=m CONFIG_MMC_SDHCI=y CONFIG_MMC_SDHCI_PLTFM=y CONFIG_MMC_SDHCI_ESDHC_IMX=y @@ -133,6 +140,8 @@ CONFIG_LEDS_CLASS=y CONFIG_RTC_CLASS=y CONFIG_RTC_INTF_DEV_UIE_EMUL=y CONFIG_RTC_MXC=y +CONFIG_DMADEVICES=y +CONFIG_IMX_SDMA=y CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_POSIX_ACL=y diff --git a/arch/arm/configs/mx3_defconfig b/arch/arm/configs/mx3_defconfig deleted file mode 100644 index cb0717fbb03d..000000000000 --- a/arch/arm/configs/mx3_defconfig +++ /dev/null @@ -1,144 +0,0 @@ -CONFIG_EXPERIMENTAL=y -CONFIG_SYSVIPC=y -CONFIG_IKCONFIG=y -CONFIG_IKCONFIG_PROC=y -CONFIG_LOG_BUF_SHIFT=14 -CONFIG_EXPERT=y -CONFIG_SLAB=y -CONFIG_MODULES=y -CONFIG_MODULE_UNLOAD=y -CONFIG_MODULE_FORCE_UNLOAD=y -CONFIG_MODVERSIONS=y -# CONFIG_BLK_DEV_BSG is not set -CONFIG_ARCH_MXC=y -CONFIG_MACH_MX31ADS_WM1133_EV1=y -CONFIG_MACH_MX31LILLY=y -CONFIG_MACH_MX31LITE=y -CONFIG_MACH_PCM037=y -CONFIG_MACH_PCM037_EET=y -CONFIG_MACH_MX31_3DS=y -CONFIG_MACH_MX31MOBOARD=y -CONFIG_MACH_QONG=y -CONFIG_MACH_ARMADILLO5X0=y -CONFIG_MACH_KZM_ARM11_01=y -CONFIG_MACH_PCM043=y -CONFIG_MACH_MX35_3DS=y -CONFIG_MACH_EUKREA_CPUIMX35=y -CONFIG_MXC_IRQ_PRIOR=y -CONFIG_MXC_PWM=y -CONFIG_ARM_ERRATA_411920=y -CONFIG_NO_HZ=y -CONFIG_HIGH_RES_TIMERS=y -CONFIG_PREEMPT=y -CONFIG_AEABI=y -CONFIG_ZBOOT_ROM_TEXT=0x0 -CONFIG_ZBOOT_ROM_BSS=0x0 -CONFIG_CMDLINE="noinitrd console=ttymxc0,115200 root=/dev/mtdblock2 rw ip=off" -CONFIG_VFP=y -CONFIG_PM_DEBUG=y -CONFIG_NET=y -CONFIG_PACKET=y -CONFIG_UNIX=y -CONFIG_INET=y -CONFIG_IP_PNP=y -CONFIG_IP_PNP_DHCP=y -# CONFIG_INET_XFRM_MODE_TRANSPORT is not set -# CONFIG_INET_XFRM_MODE_TUNNEL is not set -# CONFIG_INET_XFRM_MODE_BEET is not set -# CONFIG_INET_LRO is not set -# CONFIG_INET_DIAG is not set -# CONFIG_IPV6 is not set -CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" -CONFIG_FW_LOADER=m -CONFIG_MTD=y -CONFIG_MTD_CMDLINE_PARTS=y -CONFIG_MTD_CHAR=y -CONFIG_MTD_BLOCK=y -CONFIG_MTD_CFI=y -CONFIG_MTD_PHYSMAP=y -CONFIG_MTD_NAND=y -CONFIG_MTD_NAND_MXC=y -CONFIG_MTD_UBI=y -# CONFIG_BLK_DEV is not set -CONFIG_MISC_DEVICES=y -CONFIG_EEPROM_AT24=y -CONFIG_NETDEVICES=y -CONFIG_SMSC_PHY=y -CONFIG_NET_ETHERNET=y -CONFIG_SMSC911X=y -CONFIG_DNET=y -# CONFIG_NETDEV_1000 is not set -# CONFIG_NETDEV_10000 is not set -# CONFIG_INPUT_MOUSEDEV is not set -# CONFIG_KEYBOARD_ATKBD is not set -CONFIG_KEYBOARD_IMX=y -# CONFIG_INPUT_MOUSE is not set -# CONFIG_SERIO is not set -# CONFIG_VT is not set -# CONFIG_LEGACY_PTYS is not set -CONFIG_SERIAL_8250=m -CONFIG_SERIAL_8250_EXTENDED=y -CONFIG_SERIAL_8250_SHARE_IRQ=y -CONFIG_SERIAL_IMX=y -CONFIG_SERIAL_IMX_CONSOLE=y -# CONFIG_HW_RANDOM is not set -CONFIG_I2C=y -CONFIG_I2C_CHARDEV=y -CONFIG_I2C_IMX=y -CONFIG_SPI=y -CONFIG_W1=y -CONFIG_W1_MASTER_MXC=y -CONFIG_W1_SLAVE_THERM=y -# CONFIG_HWMON is not set -CONFIG_WATCHDOG=y -CONFIG_IMX2_WDT=y -CONFIG_MFD_WM8350_I2C=y -CONFIG_REGULATOR=y -CONFIG_REGULATOR_WM8350=y -CONFIG_MEDIA_SUPPORT=y -CONFIG_VIDEO_DEV=y -# CONFIG_RC_CORE is not set -# CONFIG_MEDIA_TUNER_CUSTOMISE is not set -CONFIG_SOC_CAMERA=y -CONFIG_SOC_CAMERA_MT9M001=y -CONFIG_SOC_CAMERA_MT9M111=y -CONFIG_SOC_CAMERA_MT9T031=y -CONFIG_SOC_CAMERA_MT9V022=y -CONFIG_SOC_CAMERA_TW9910=y -CONFIG_SOC_CAMERA_OV772X=y -CONFIG_VIDEO_MX3=y -# CONFIG_RADIO_ADAPTERS is not set -CONFIG_FB=y -CONFIG_SOUND=y -CONFIG_SND=y -# CONFIG_SND_ARM is not set -# CONFIG_SND_SPI is not set -CONFIG_SND_SOC=y -CONFIG_SND_IMX_SOC=y -CONFIG_SND_MXC_SOC_WM1133_EV1=y -CONFIG_SND_SOC_PHYCORE_AC97=y -CONFIG_SND_SOC_EUKREA_TLV320=y -CONFIG_USB=y -CONFIG_USB_EHCI_HCD=y -CONFIG_USB_EHCI_MXC=y -CONFIG_USB_GADGET=m -CONFIG_USB_FSL_USB2=m -CONFIG_USB_G_SERIAL=m -CONFIG_USB_ULPI=y -CONFIG_MMC=y -CONFIG_MMC_MXC=y -CONFIG_RTC_CLASS=y -CONFIG_RTC_MXC=y -CONFIG_DMADEVICES=y -# CONFIG_DNOTIFY is not set -CONFIG_TMPFS=y -CONFIG_JFFS2_FS=y -CONFIG_UBIFS_FS=y -CONFIG_NFS_FS=y -CONFIG_NFS_V3=y -CONFIG_NFS_V4=y -CONFIG_ROOT_NFS=y -# CONFIG_ENABLE_WARN_DEPRECATED is not set -# CONFIG_ENABLE_MUST_CHECK is not set -CONFIG_SYSCTL_SYSCALL_CHECK=y -# CONFIG_CRYPTO_ANSI_CPRNG is not set From dbb004294a934d030a9d716692a1dd6f677fc90e Mon Sep 17 00:00:00 2001 From: Ciaran McCormick Date: Mon, 31 Oct 2011 19:29:26 +0000 Subject: [PATCH 112/820] dma: fix spacing for method declaration, coding style issue in iop-adma.c Signed-off-by: Ciaran McCormick Signed-off-by: Vinod Koul --- drivers/dma/iop-adma.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c index e03f811a83dd..686dc7673b19 100644 --- a/drivers/dma/iop-adma.c +++ b/drivers/dma/iop-adma.c @@ -1746,12 +1746,12 @@ static struct platform_driver iop_adma_driver = { }, }; -static int __init iop_adma_init (void) +static int __init iop_adma_init(void) { return platform_driver_register(&iop_adma_driver); } -static void __exit iop_adma_exit (void) +static void __exit iop_adma_exit(void) { platform_driver_unregister(&iop_adma_driver); return; From 8e2d41f8c85613a3739e8b2600ac2c66c08422f3 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Wed, 24 Aug 2011 08:41:09 +0200 Subject: [PATCH 113/820] dma i.MX: remove individual SOC dependency The individual SoC dependency in Kconfig hardly scales anymore. Instead of having such a fine grained dependency just depend on ARCH_MXC and risk that the uninformed user has to look in the help text to figure out which driver is the correct one. Signed-off-by: Sascha Hauer Cc: Vinod Koul Signed-off-by: Vinod Koul --- drivers/dma/Kconfig | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig index ab8f469f5cf8..e514c4560f87 100644 --- a/drivers/dma/Kconfig +++ b/drivers/dma/Kconfig @@ -124,7 +124,7 @@ config MV_XOR config MX3_IPU bool "MX3x Image Processing Unit support" - depends on ARCH_MX3 + depends on ARCH_MXC select DMA_ENGINE default y help @@ -216,11 +216,11 @@ config PCH_DMA config IMX_SDMA tristate "i.MX SDMA support" - depends on ARCH_MX25 || ARCH_MX3 || ARCH_MX5 + depends on ARCH_MXC select DMA_ENGINE help Support the i.MX SDMA engine. This engine is integrated into - Freescale i.MX25/31/35/51 chips. + Freescale i.MX25/31/35/51/53 chips. config IMX_DMA tristate "i.MX DMA support" From e79e72be298dc4726a641686032b8296c2fb473e Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Thu, 17 Nov 2011 16:14:22 +0900 Subject: [PATCH 114/820] pch_dma: Change company name OKI SEMICONDUCTOR to LAPIS Semiconductor On October 1 in 2011, OKI SEMICONDUCTOR Co., Ltd. changed the company name in to LAPIS Semiconductor Co., Ltd. Signed-off-by: Tomoya MORINAGA Signed-off-by: Vinod Koul --- drivers/dma/Kconfig | 4 ++-- drivers/dma/pch_dma.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig index e514c4560f87..606592ecd04f 100644 --- a/drivers/dma/Kconfig +++ b/drivers/dma/Kconfig @@ -201,13 +201,13 @@ config PL330_DMA platform_data for a dma-pl330 device. config PCH_DMA - tristate "Intel EG20T PCH / OKI Semi IOH(ML7213/ML7223) DMA support" + tristate "Intel EG20T PCH / LAPIS Semi IOH(ML7213/ML7223) DMA support" depends on PCI && X86 select DMA_ENGINE help Enable support for Intel EG20T PCH DMA engine. - This driver also can be used for OKI SEMICONDUCTOR IOH(Input/ + This driver also can be used for LAPIS Semiconductor IOH(Input/ Output Hub), ML7213 and ML7223. ML7213 IOH is for IVI(In-Vehicle Infotainment) use and ML7223 IOH is for MP(Media Phone) use. diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c index a6d0e3dbed07..a9b8a7971211 100644 --- a/drivers/dma/pch_dma.c +++ b/drivers/dma/pch_dma.c @@ -1,7 +1,7 @@ /* * Topcliff PCH DMA controller driver * Copyright (c) 2010 Intel Corporation - * Copyright (C) 2011 OKI SEMICONDUCTOR CO., LTD. + * Copyright (C) 2011 LAPIS Semiconductor Co., Ltd. * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -1057,7 +1057,7 @@ static void __exit pch_dma_exit(void) module_init(pch_dma_init); module_exit(pch_dma_exit); -MODULE_DESCRIPTION("Intel EG20T PCH / OKI SEMICONDUCTOR ML7213 IOH " +MODULE_DESCRIPTION("Intel EG20T PCH / LAPIS Semiconductor ML7213/ML7223 IOH " "DMA controller driver"); MODULE_AUTHOR("Yong Wang "); MODULE_LICENSE("GPL v2"); From ca7fe2db892dcf91b2c72ee352eda4ff867903a7 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Thu, 17 Nov 2011 16:14:23 +0900 Subject: [PATCH 115/820] pch_dma: Support new device LAPIS Semiconductor ML7831 IOH ML7831 is companion chip for Intel Atom E6xx series. Signed-off-by: Tomoya MORINAGA Signed-off-by: Vinod Koul --- drivers/dma/Kconfig | 12 ++++++------ drivers/dma/pch_dma.c | 6 +++++- 2 files changed, 11 insertions(+), 7 deletions(-) diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig index 606592ecd04f..7ec0d6cef0c3 100644 --- a/drivers/dma/Kconfig +++ b/drivers/dma/Kconfig @@ -201,18 +201,18 @@ config PL330_DMA platform_data for a dma-pl330 device. config PCH_DMA - tristate "Intel EG20T PCH / LAPIS Semi IOH(ML7213/ML7223) DMA support" + tristate "Intel EG20T PCH / LAPIS Semicon IOH(ML7213/ML7223/ML7831) DMA" depends on PCI && X86 select DMA_ENGINE help Enable support for Intel EG20T PCH DMA engine. This driver also can be used for LAPIS Semiconductor IOH(Input/ - Output Hub), ML7213 and ML7223. - ML7213 IOH is for IVI(In-Vehicle Infotainment) use and ML7223 IOH is - for MP(Media Phone) use. - ML7213/ML7223 is companion chip for Intel Atom E6xx series. - ML7213/ML7223 is completely compatible for Intel EG20T PCH. + Output Hub), ML7213, ML7223 and ML7831. + ML7213 IOH is for IVI(In-Vehicle Infotainment) use, ML7223 IOH is + for MP(Media Phone) use and ML7831 IOH is for general purpose use. + ML7213/ML7223/ML7831 is companion chip for Intel Atom E6xx series. + ML7213/ML7223/ML7831 is completely compatible for Intel EG20T PCH. config IMX_SDMA tristate "i.MX SDMA support" diff --git a/drivers/dma/pch_dma.c b/drivers/dma/pch_dma.c index a9b8a7971211..60aca37e22ea 100644 --- a/drivers/dma/pch_dma.c +++ b/drivers/dma/pch_dma.c @@ -1018,6 +1018,8 @@ static void __devexit pch_dma_remove(struct pci_dev *pdev) #define PCI_DEVICE_ID_ML7223_DMA2_4CH 0x800E #define PCI_DEVICE_ID_ML7223_DMA3_4CH 0x8017 #define PCI_DEVICE_ID_ML7223_DMA4_4CH 0x803B +#define PCI_DEVICE_ID_ML7831_DMA1_8CH 0x8810 +#define PCI_DEVICE_ID_ML7831_DMA2_4CH 0x8815 DEFINE_PCI_DEVICE_TABLE(pch_dma_id_table) = { { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_8CH), 8 }, @@ -1030,6 +1032,8 @@ DEFINE_PCI_DEVICE_TABLE(pch_dma_id_table) = { { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA2_4CH), 4}, /* Video SPI */ { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA3_4CH), 4}, /* Security */ { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA4_4CH), 4}, /* FPGA */ + { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA1_8CH), 8}, /* UART */ + { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA2_4CH), 4}, /* SPI */ { 0, }, }; @@ -1057,7 +1061,7 @@ static void __exit pch_dma_exit(void) module_init(pch_dma_init); module_exit(pch_dma_exit); -MODULE_DESCRIPTION("Intel EG20T PCH / LAPIS Semiconductor ML7213/ML7223 IOH " +MODULE_DESCRIPTION("Intel EG20T PCH / LAPIS Semicon ML7213/ML7223/ML7831 IOH " "DMA controller driver"); MODULE_AUTHOR("Yong Wang "); MODULE_LICENSE("GPL v2"); From 8341544c9d95610f86b560333c7b22dd1096ca4f Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 21:50:31 +0530 Subject: [PATCH 116/820] USB-musb: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves usb/musb driver to use new enum Signed-off-by: Vinod Koul Cc: Felipe Balbi Cc: Greg Kroah-Hartman Acked-by: Felipe Balbi Acked-by: Mian Yousaf Kaukab --- drivers/usb/musb/ux500_dma.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/usb/musb/ux500_dma.c b/drivers/usb/musb/ux500_dma.c index ef4333f4bbe0..f8fd0ddee14a 100644 --- a/drivers/usb/musb/ux500_dma.c +++ b/drivers/usb/musb/ux500_dma.c @@ -112,7 +112,7 @@ static bool ux500_configure_channel(struct dma_channel *channel, struct musb_hw_ep *hw_ep = ux500_channel->hw_ep; struct dma_chan *dma_chan = ux500_channel->dma_chan; struct dma_async_tx_descriptor *dma_desc; - enum dma_data_direction direction; + enum dma_transfer_direction direction; struct scatterlist sg; struct dma_slave_config slave_conf; enum dma_slave_buswidth addr_width; @@ -132,7 +132,7 @@ static bool ux500_configure_channel(struct dma_channel *channel, sg_dma_address(&sg) = dma_addr; sg_dma_len(&sg) = len; - direction = ux500_channel->is_tx ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + direction = ux500_channel->is_tx ? DMA_MEM_TO_DEV : DMA_DEV_TO_MEM; addr_width = (len & 0x3) ? DMA_SLAVE_BUSWIDTH_1_BYTE : DMA_SLAVE_BUSWIDTH_4_BYTES; From 55ba4e5ed4ac57b60fe56acfd324f6a87123cc34 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Fri, 14 Oct 2011 21:57:58 +0530 Subject: [PATCH 117/820] USB-renesas: move to dma_transfer_direction fixup usage of dma direction by introducing dma_transfer_direction, this patch moves usb/renesas driver to use new enum Signed-off-by: Vinod Koul Cc: Felipe Balbi Cc: Greg Kroah-Hartman Acked-by: Felipe Balbi Acked-by: Kuninori Morimoto --- drivers/usb/renesas_usbhs/fifo.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/usb/renesas_usbhs/fifo.c b/drivers/usb/renesas_usbhs/fifo.c index a34430f55fb7..051c9a9337a3 100644 --- a/drivers/usb/renesas_usbhs/fifo.c +++ b/drivers/usb/renesas_usbhs/fifo.c @@ -591,10 +591,10 @@ static void usbhsf_dma_prepare_tasklet(unsigned long data) struct dma_async_tx_descriptor *desc; struct dma_chan *chan = usbhsf_dma_chan_get(fifo, pkt); struct device *dev = usbhs_priv_to_dev(priv); - enum dma_data_direction dir; + enum dma_transfer_direction dir; dma_cookie_t cookie; - dir = usbhs_pipe_is_dir_in(pipe) ? DMA_FROM_DEVICE : DMA_TO_DEVICE; + dir = usbhs_pipe_is_dir_in(pipe) ? DMA_DEV_TO_MEM : DMA_MEM_TO_DEV; sg_init_table(&sg, 1); sg_set_page(&sg, virt_to_page(pkt->dma), From b14dab792dee3245b628e046d80a7fad5573fea6 Mon Sep 17 00:00:00 2001 From: Jassi Brar Date: Thu, 13 Oct 2011 12:33:30 +0530 Subject: [PATCH 118/820] DMAEngine: Define interleaved transfer request api Define a new api that could be used for doing fancy data transfers like interleaved to contiguous copy and vice-versa. Traditional SG_list based transfers tend to be very inefficient in such cases as where the interleave and chunk are only a few bytes, which call for a very condensed api to convey pattern of the transfer. This api supports all 4 variants of scatter-gather and contiguous transfer. Of course, neither can this api help transfers that don't lend to DMA by nature, i.e, scattered tiny read/writes with no periodic pattern. Also since now we support SLAVE channels that might not provide device_prep_slave_sg callback but device_prep_interleaved_dma, remove the BUG_ON check. Signed-off-by: Jassi Brar Acked-by: Barry Song [renamed dmaxfer_template to dma_interleaved_template did fixup after the enum dma_transfer_merge] Signed-off-by: Vinod Koul --- Documentation/dmaengine.txt | 8 ++++ drivers/dma/dmaengine.c | 4 +- include/linux/dmaengine.h | 78 +++++++++++++++++++++++++++++++++++-- 3 files changed, 85 insertions(+), 5 deletions(-) diff --git a/Documentation/dmaengine.txt b/Documentation/dmaengine.txt index 94b7e0f96b38..bbe6cb3d1856 100644 --- a/Documentation/dmaengine.txt +++ b/Documentation/dmaengine.txt @@ -75,6 +75,10 @@ The slave DMA usage consists of following steps: slave_sg - DMA a list of scatter gather buffers from/to a peripheral dma_cyclic - Perform a cyclic DMA operation from/to a peripheral till the operation is explicitly stopped. + interleaved_dma - This is common to Slave as well as M2M clients. For slave + address of devices' fifo could be already known to the driver. + Various types of operations could be expressed by setting + appropriate values to the 'dma_interleaved_template' members. A non-NULL return of this transfer API represents a "descriptor" for the given transaction. @@ -89,6 +93,10 @@ The slave DMA usage consists of following steps: struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, size_t period_len, enum dma_data_direction direction); + struct dma_async_tx_descriptor *(*device_prep_interleaved_dma)( + struct dma_chan *chan, struct dma_interleaved_template *xt, + unsigned long flags); + The peripheral driver is expected to have mapped the scatterlist for the DMA operation prior to calling device_prep_slave_sg, and must keep the scatterlist mapped until the DMA operation has completed. diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index b48967b499da..a6c6051ec858 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -693,12 +693,12 @@ int dma_async_device_register(struct dma_device *device) !device->device_prep_dma_interrupt); BUG_ON(dma_has_cap(DMA_SG, device->cap_mask) && !device->device_prep_dma_sg); - BUG_ON(dma_has_cap(DMA_SLAVE, device->cap_mask) && - !device->device_prep_slave_sg); BUG_ON(dma_has_cap(DMA_CYCLIC, device->cap_mask) && !device->device_prep_dma_cyclic); BUG_ON(dma_has_cap(DMA_SLAVE, device->cap_mask) && !device->device_control); + BUG_ON(dma_has_cap(DMA_INTERLEAVE, device->cap_mask) && + !device->device_prep_interleaved_dma); BUG_ON(!device->device_alloc_chan_resources); BUG_ON(!device->device_free_chan_resources); diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index a865b3a354cd..5532bb8b500c 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -71,10 +71,10 @@ enum dma_transaction_type { DMA_ASYNC_TX, DMA_SLAVE, DMA_CYCLIC, -}; - + DMA_INTERLEAVE, /* last transaction type for creation of the capabilities mask */ -#define DMA_TX_TYPE_END (DMA_CYCLIC + 1) + DMA_TX_TYPE_END, +}; /** * enum dma_transfer_direction - dma transfer mode and direction indicator @@ -90,6 +90,74 @@ enum dma_transfer_direction { DMA_DEV_TO_DEV, }; +/** + * Interleaved Transfer Request + * ---------------------------- + * A chunk is collection of contiguous bytes to be transfered. + * The gap(in bytes) between two chunks is called inter-chunk-gap(ICG). + * ICGs may or maynot change between chunks. + * A FRAME is the smallest series of contiguous {chunk,icg} pairs, + * that when repeated an integral number of times, specifies the transfer. + * A transfer template is specification of a Frame, the number of times + * it is to be repeated and other per-transfer attributes. + * + * Practically, a client driver would have ready a template for each + * type of transfer it is going to need during its lifetime and + * set only 'src_start' and 'dst_start' before submitting the requests. + * + * + * | Frame-1 | Frame-2 | ~ | Frame-'numf' | + * |====....==.===...=...|====....==.===...=...| ~ |====....==.===...=...| + * + * == Chunk size + * ... ICG + */ + +/** + * struct data_chunk - Element of scatter-gather list that makes a frame. + * @size: Number of bytes to read from source. + * size_dst := fn(op, size_src), so doesn't mean much for destination. + * @icg: Number of bytes to jump after last src/dst address of this + * chunk and before first src/dst address for next chunk. + * Ignored for dst(assumed 0), if dst_inc is true and dst_sgl is false. + * Ignored for src(assumed 0), if src_inc is true and src_sgl is false. + */ +struct data_chunk { + size_t size; + size_t icg; +}; + +/** + * struct dma_interleaved_template - Template to convey DMAC the transfer pattern + * and attributes. + * @src_start: Bus address of source for the first chunk. + * @dst_start: Bus address of destination for the first chunk. + * @dir: Specifies the type of Source and Destination. + * @src_inc: If the source address increments after reading from it. + * @dst_inc: If the destination address increments after writing to it. + * @src_sgl: If the 'icg' of sgl[] applies to Source (scattered read). + * Otherwise, source is read contiguously (icg ignored). + * Ignored if src_inc is false. + * @dst_sgl: If the 'icg' of sgl[] applies to Destination (scattered write). + * Otherwise, destination is filled contiguously (icg ignored). + * Ignored if dst_inc is false. + * @numf: Number of frames in this template. + * @frame_size: Number of chunks in a frame i.e, size of sgl[]. + * @sgl: Array of {chunk,icg} pairs that make up a frame. + */ +struct dma_interleaved_template { + dma_addr_t src_start; + dma_addr_t dst_start; + enum dma_transfer_direction dir; + bool src_inc; + bool dst_inc; + bool src_sgl; + bool dst_sgl; + size_t numf; + size_t frame_size; + struct data_chunk sgl[0]; +}; + /** * enum dma_ctrl_flags - DMA flags to augment operation preparation, * control completion, and communicate status. @@ -445,6 +513,7 @@ struct dma_tx_state { * @device_prep_dma_cyclic: prepare a cyclic dma operation suitable for audio. * The function takes a buffer of size buf_len. The callback function will * be called after period_len bytes have been transferred. + * @device_prep_interleaved_dma: Transfer expression in a generic way. * @device_control: manipulate all pending operations on a channel, returns * zero or error code * @device_tx_status: poll for transaction completion, the optional @@ -509,6 +578,9 @@ struct dma_device { struct dma_async_tx_descriptor *(*device_prep_dma_cyclic)( struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, size_t period_len, enum dma_transfer_direction direction); + struct dma_async_tx_descriptor *(*device_prep_interleaved_dma)( + struct dma_chan *chan, struct dma_interleaved_template *xt, + unsigned long flags); int (*device_control)(struct dma_chan *chan, enum dma_ctrl_cmd cmd, unsigned long arg); From ca21a146a45a179a2a7bc86d938a2fbf571a7510 Mon Sep 17 00:00:00 2001 From: Rongjun Ying Date: Thu, 27 Oct 2011 19:22:39 -0700 Subject: [PATCH 119/820] dmaengine: add CSR SiRFprimaII DMAC driver Cc: Jassi Brar Cc: Arnd Bergmann Cc: Linus Walleij Signed-off-by: Rongjun Ying Signed-off-by: Barry Song [fixed direction enums and cyclic api based on changes already merged] Signed-off-by: Vinod Koul --- MAINTAINERS | 1 + drivers/dma/Kconfig | 7 + drivers/dma/Makefile | 1 + drivers/dma/sirf-dma.c | 717 ++++++++++++++++++++++++++++++++++++ include/linux/sirfsoc_dma.h | 6 + 5 files changed, 732 insertions(+) create mode 100644 drivers/dma/sirf-dma.c create mode 100644 include/linux/sirfsoc_dma.h diff --git a/MAINTAINERS b/MAINTAINERS index 4808256446f2..1b141d71ea13 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -749,6 +749,7 @@ M: Barry Song L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-prima2/ +F: drivers/dma/sirf-dma* ARM/EBSA110 MACHINE SUPPORT M: Russell King diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig index 7ec0d6cef0c3..f1a274994bb1 100644 --- a/drivers/dma/Kconfig +++ b/drivers/dma/Kconfig @@ -187,6 +187,13 @@ config TIMB_DMA help Enable support for the Timberdale FPGA DMA engine. +config SIRF_DMA + tristate "CSR SiRFprimaII DMA support" + depends on ARCH_PRIMA2 + select DMA_ENGINE + help + Enable support for the CSR SiRFprimaII DMA engine. + config ARCH_HAS_ASYNC_TX_FIND_CHANNEL bool diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile index 30cf3b1f0c5c..009a222e8283 100644 --- a/drivers/dma/Makefile +++ b/drivers/dma/Makefile @@ -21,6 +21,7 @@ obj-$(CONFIG_IMX_SDMA) += imx-sdma.o obj-$(CONFIG_IMX_DMA) += imx-dma.o obj-$(CONFIG_MXS_DMA) += mxs-dma.o obj-$(CONFIG_TIMB_DMA) += timb_dma.o +obj-$(CONFIG_SIRF_DMA) += sirf-dma.o obj-$(CONFIG_STE_DMA40) += ste_dma40.o ste_dma40_ll.o obj-$(CONFIG_PL330_DMA) += pl330.o obj-$(CONFIG_PCH_DMA) += pch_dma.o diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c new file mode 100644 index 000000000000..55ec67997670 --- /dev/null +++ b/drivers/dma/sirf-dma.c @@ -0,0 +1,717 @@ +/* + * DMA controller driver for CSR SiRFprimaII + * + * Copyright (c) 2011 Cambridge Silicon Radio Limited, a CSR plc group company. + * + * Licensed under GPLv2 or later. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define SIRFSOC_DMA_DESCRIPTORS 16 +#define SIRFSOC_DMA_CHANNELS 16 + +#define SIRFSOC_DMA_CH_ADDR 0x00 +#define SIRFSOC_DMA_CH_XLEN 0x04 +#define SIRFSOC_DMA_CH_YLEN 0x08 +#define SIRFSOC_DMA_CH_CTRL 0x0C + +#define SIRFSOC_DMA_WIDTH_0 0x100 +#define SIRFSOC_DMA_CH_VALID 0x140 +#define SIRFSOC_DMA_CH_INT 0x144 +#define SIRFSOC_DMA_INT_EN 0x148 +#define SIRFSOC_DMA_CH_LOOP_CTRL 0x150 + +#define SIRFSOC_DMA_MODE_CTRL_BIT 4 +#define SIRFSOC_DMA_DIR_CTRL_BIT 5 + +/* xlen and dma_width register is in 4 bytes boundary */ +#define SIRFSOC_DMA_WORD_LEN 4 + +struct sirfsoc_dma_desc { + struct dma_async_tx_descriptor desc; + struct list_head node; + + /* SiRFprimaII 2D-DMA parameters */ + + int xlen; /* DMA xlen */ + int ylen; /* DMA ylen */ + int width; /* DMA width */ + int dir; + bool cyclic; /* is loop DMA? */ + u32 addr; /* DMA buffer address */ +}; + +struct sirfsoc_dma_chan { + struct dma_chan chan; + struct list_head free; + struct list_head prepared; + struct list_head queued; + struct list_head active; + struct list_head completed; + dma_cookie_t completed_cookie; + unsigned long happened_cyclic; + unsigned long completed_cyclic; + + /* Lock for this structure */ + spinlock_t lock; + + int mode; +}; + +struct sirfsoc_dma { + struct dma_device dma; + struct tasklet_struct tasklet; + struct sirfsoc_dma_chan channels[SIRFSOC_DMA_CHANNELS]; + void __iomem *base; + int irq; +}; + +#define DRV_NAME "sirfsoc_dma" + +/* Convert struct dma_chan to struct sirfsoc_dma_chan */ +static inline +struct sirfsoc_dma_chan *dma_chan_to_sirfsoc_dma_chan(struct dma_chan *c) +{ + return container_of(c, struct sirfsoc_dma_chan, chan); +} + +/* Convert struct dma_chan to struct sirfsoc_dma */ +static inline struct sirfsoc_dma *dma_chan_to_sirfsoc_dma(struct dma_chan *c) +{ + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(c); + return container_of(schan, struct sirfsoc_dma, channels[c->chan_id]); +} + +/* Execute all queued DMA descriptors */ +static void sirfsoc_dma_execute(struct sirfsoc_dma_chan *schan) +{ + struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan); + int cid = schan->chan.chan_id; + struct sirfsoc_dma_desc *sdesc = NULL; + + /* + * lock has been held by functions calling this, so we don't hold + * lock again + */ + + sdesc = list_first_entry(&schan->queued, struct sirfsoc_dma_desc, + node); + /* Move the first queued descriptor to active list */ + list_move_tail(&schan->queued, &schan->active); + + /* Start the DMA transfer */ + writel_relaxed(sdesc->width, sdma->base + SIRFSOC_DMA_WIDTH_0 + + cid * 4); + writel_relaxed(cid | (schan->mode << SIRFSOC_DMA_MODE_CTRL_BIT) | + (sdesc->dir << SIRFSOC_DMA_DIR_CTRL_BIT), + sdma->base + cid * 0x10 + SIRFSOC_DMA_CH_CTRL); + writel_relaxed(sdesc->xlen, sdma->base + cid * 0x10 + + SIRFSOC_DMA_CH_XLEN); + writel_relaxed(sdesc->ylen, sdma->base + cid * 0x10 + + SIRFSOC_DMA_CH_YLEN); + writel_relaxed(readl_relaxed(sdma->base + SIRFSOC_DMA_INT_EN) | + (1 << cid), sdma->base + SIRFSOC_DMA_INT_EN); + + /* + * writel has an implict memory write barrier to make sure data is + * flushed into memory before starting DMA + */ + writel(sdesc->addr >> 2, sdma->base + cid * 0x10 + SIRFSOC_DMA_CH_ADDR); + + if (sdesc->cyclic) { + writel((1 << cid) | 1 << (cid + 16) | + readl_relaxed(sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL), + sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL); + schan->happened_cyclic = schan->completed_cyclic = 0; + } +} + +/* Interrupt handler */ +static irqreturn_t sirfsoc_dma_irq(int irq, void *data) +{ + struct sirfsoc_dma *sdma = data; + struct sirfsoc_dma_chan *schan; + struct sirfsoc_dma_desc *sdesc = NULL; + u32 is; + int ch; + + is = readl(sdma->base + SIRFSOC_DMA_CH_INT); + while ((ch = fls(is) - 1) >= 0) { + is &= ~(1 << ch); + writel_relaxed(1 << ch, sdma->base + SIRFSOC_DMA_CH_INT); + schan = &sdma->channels[ch]; + + spin_lock(&schan->lock); + + sdesc = list_first_entry(&schan->active, struct sirfsoc_dma_desc, + node); + if (!sdesc->cyclic) { + /* Execute queued descriptors */ + list_splice_tail_init(&schan->active, &schan->completed); + if (!list_empty(&schan->queued)) + sirfsoc_dma_execute(schan); + } else + schan->happened_cyclic++; + + spin_unlock(&schan->lock); + } + + /* Schedule tasklet */ + tasklet_schedule(&sdma->tasklet); + + return IRQ_HANDLED; +} + +/* process completed descriptors */ +static void sirfsoc_dma_process_completed(struct sirfsoc_dma *sdma) +{ + dma_cookie_t last_cookie = 0; + struct sirfsoc_dma_chan *schan; + struct sirfsoc_dma_desc *sdesc; + struct dma_async_tx_descriptor *desc; + unsigned long flags; + unsigned long happened_cyclic; + LIST_HEAD(list); + int i; + + for (i = 0; i < sdma->dma.chancnt; i++) { + schan = &sdma->channels[i]; + + /* Get all completed descriptors */ + spin_lock_irqsave(&schan->lock, flags); + if (!list_empty(&schan->completed)) { + list_splice_tail_init(&schan->completed, &list); + spin_unlock_irqrestore(&schan->lock, flags); + + /* Execute callbacks and run dependencies */ + list_for_each_entry(sdesc, &list, node) { + desc = &sdesc->desc; + + if (desc->callback) + desc->callback(desc->callback_param); + + last_cookie = desc->cookie; + dma_run_dependencies(desc); + } + + /* Free descriptors */ + spin_lock_irqsave(&schan->lock, flags); + list_splice_tail_init(&list, &schan->free); + schan->completed_cookie = last_cookie; + spin_unlock_irqrestore(&schan->lock, flags); + } else { + /* for cyclic channel, desc is always in active list */ + sdesc = list_first_entry(&schan->active, struct sirfsoc_dma_desc, + node); + + if (!sdesc || (sdesc && !sdesc->cyclic)) { + /* without active cyclic DMA */ + spin_unlock_irqrestore(&schan->lock, flags); + continue; + } + + /* cyclic DMA */ + happened_cyclic = schan->happened_cyclic; + spin_unlock_irqrestore(&schan->lock, flags); + + desc = &sdesc->desc; + while (happened_cyclic != schan->completed_cyclic) { + if (desc->callback) + desc->callback(desc->callback_param); + schan->completed_cyclic++; + } + } + } +} + +/* DMA Tasklet */ +static void sirfsoc_dma_tasklet(unsigned long data) +{ + struct sirfsoc_dma *sdma = (void *)data; + + sirfsoc_dma_process_completed(sdma); +} + +/* Submit descriptor to hardware */ +static dma_cookie_t sirfsoc_dma_tx_submit(struct dma_async_tx_descriptor *txd) +{ + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(txd->chan); + struct sirfsoc_dma_desc *sdesc; + unsigned long flags; + dma_cookie_t cookie; + + sdesc = container_of(txd, struct sirfsoc_dma_desc, desc); + + spin_lock_irqsave(&schan->lock, flags); + + /* Move descriptor to queue */ + list_move_tail(&sdesc->node, &schan->queued); + + /* Update cookie */ + cookie = schan->chan.cookie + 1; + if (cookie <= 0) + cookie = 1; + + schan->chan.cookie = cookie; + sdesc->desc.cookie = cookie; + + spin_unlock_irqrestore(&schan->lock, flags); + + return cookie; +} + +static int sirfsoc_dma_slave_config(struct sirfsoc_dma_chan *schan, + struct dma_slave_config *config) +{ + unsigned long flags; + + if ((config->src_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES) || + (config->dst_addr_width != DMA_SLAVE_BUSWIDTH_4_BYTES)) + return -EINVAL; + + spin_lock_irqsave(&schan->lock, flags); + schan->mode = (config->src_maxburst == 4 ? 1 : 0); + spin_unlock_irqrestore(&schan->lock, flags); + + return 0; +} + +static int sirfsoc_dma_terminate_all(struct sirfsoc_dma_chan *schan) +{ + struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(&schan->chan); + int cid = schan->chan.chan_id; + unsigned long flags; + + writel_relaxed(readl_relaxed(sdma->base + SIRFSOC_DMA_INT_EN) & + ~(1 << cid), sdma->base + SIRFSOC_DMA_INT_EN); + writel_relaxed(1 << cid, sdma->base + SIRFSOC_DMA_CH_VALID); + + writel_relaxed(readl_relaxed(sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL) + & ~((1 << cid) | 1 << (cid + 16)), + sdma->base + SIRFSOC_DMA_CH_LOOP_CTRL); + + spin_lock_irqsave(&schan->lock, flags); + list_splice_tail_init(&schan->active, &schan->free); + list_splice_tail_init(&schan->queued, &schan->free); + spin_unlock_irqrestore(&schan->lock, flags); + + return 0; +} + +static int sirfsoc_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, + unsigned long arg) +{ + struct dma_slave_config *config; + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + + switch (cmd) { + case DMA_TERMINATE_ALL: + return sirfsoc_dma_terminate_all(schan); + case DMA_SLAVE_CONFIG: + config = (struct dma_slave_config *)arg; + return sirfsoc_dma_slave_config(schan, config); + + default: + break; + } + + return -ENOSYS; +} + +/* Alloc channel resources */ +static int sirfsoc_dma_alloc_chan_resources(struct dma_chan *chan) +{ + struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(chan); + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + struct sirfsoc_dma_desc *sdesc; + unsigned long flags; + LIST_HEAD(descs); + int i; + + /* Alloc descriptors for this channel */ + for (i = 0; i < SIRFSOC_DMA_DESCRIPTORS; i++) { + sdesc = kzalloc(sizeof(*sdesc), GFP_KERNEL); + if (!sdesc) { + dev_notice(sdma->dma.dev, "Memory allocation error. " + "Allocated only %u descriptors\n", i); + break; + } + + dma_async_tx_descriptor_init(&sdesc->desc, chan); + sdesc->desc.flags = DMA_CTRL_ACK; + sdesc->desc.tx_submit = sirfsoc_dma_tx_submit; + + list_add_tail(&sdesc->node, &descs); + } + + /* Return error only if no descriptors were allocated */ + if (i == 0) + return -ENOMEM; + + spin_lock_irqsave(&schan->lock, flags); + + list_splice_tail_init(&descs, &schan->free); + spin_unlock_irqrestore(&schan->lock, flags); + + return i; +} + +/* Free channel resources */ +static void sirfsoc_dma_free_chan_resources(struct dma_chan *chan) +{ + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + struct sirfsoc_dma_desc *sdesc, *tmp; + unsigned long flags; + LIST_HEAD(descs); + + spin_lock_irqsave(&schan->lock, flags); + + /* Channel must be idle */ + BUG_ON(!list_empty(&schan->prepared)); + BUG_ON(!list_empty(&schan->queued)); + BUG_ON(!list_empty(&schan->active)); + BUG_ON(!list_empty(&schan->completed)); + + /* Move data */ + list_splice_tail_init(&schan->free, &descs); + + spin_unlock_irqrestore(&schan->lock, flags); + + /* Free descriptors */ + list_for_each_entry_safe(sdesc, tmp, &descs, node) + kfree(sdesc); +} + +/* Send pending descriptor to hardware */ +static void sirfsoc_dma_issue_pending(struct dma_chan *chan) +{ + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + unsigned long flags; + + spin_lock_irqsave(&schan->lock, flags); + + if (list_empty(&schan->active) && !list_empty(&schan->queued)) + sirfsoc_dma_execute(schan); + + spin_unlock_irqrestore(&schan->lock, flags); +} + +/* Check request completion status */ +static enum dma_status +sirfsoc_dma_tx_status(struct dma_chan *chan, dma_cookie_t cookie, + struct dma_tx_state *txstate) +{ + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + unsigned long flags; + dma_cookie_t last_used; + dma_cookie_t last_complete; + + spin_lock_irqsave(&schan->lock, flags); + last_used = schan->chan.cookie; + last_complete = schan->completed_cookie; + spin_unlock_irqrestore(&schan->lock, flags); + + dma_set_tx_state(txstate, last_complete, last_used, 0); + return dma_async_is_complete(cookie, last_complete, last_used); +} + +static struct dma_async_tx_descriptor *sirfsoc_dma_prep_interleaved( + struct dma_chan *chan, struct dma_interleaved_template *xt, + unsigned long flags) +{ + struct sirfsoc_dma *sdma = dma_chan_to_sirfsoc_dma(chan); + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + struct sirfsoc_dma_desc *sdesc = NULL; + unsigned long iflags; + int ret; + + if ((xt->dir != DMA_MEM_TO_DEV) || (xt->dir != DMA_DEV_TO_MEM)) { + ret = -EINVAL; + goto err_dir; + } + + /* Get free descriptor */ + spin_lock_irqsave(&schan->lock, iflags); + if (!list_empty(&schan->free)) { + sdesc = list_first_entry(&schan->free, struct sirfsoc_dma_desc, + node); + list_del(&sdesc->node); + } + spin_unlock_irqrestore(&schan->lock, iflags); + + if (!sdesc) { + /* try to free completed descriptors */ + sirfsoc_dma_process_completed(sdma); + ret = 0; + goto no_desc; + } + + /* Place descriptor in prepared list */ + spin_lock_irqsave(&schan->lock, iflags); + + /* + * Number of chunks in a frame can only be 1 for prima2 + * and ylen (number of frame - 1) must be at least 0 + */ + if ((xt->frame_size == 1) && (xt->numf > 0)) { + sdesc->cyclic = 0; + sdesc->xlen = xt->sgl[0].size / SIRFSOC_DMA_WORD_LEN; + sdesc->width = (xt->sgl[0].size + xt->sgl[0].icg) / + SIRFSOC_DMA_WORD_LEN; + sdesc->ylen = xt->numf - 1; + if (xt->dir == DMA_MEM_TO_DEV) { + sdesc->addr = xt->src_start; + sdesc->dir = 1; + } else { + sdesc->addr = xt->dst_start; + sdesc->dir = 0; + } + + list_add_tail(&sdesc->node, &schan->prepared); + } else { + pr_err("sirfsoc DMA Invalid xfer\n"); + ret = -EINVAL; + goto err_xfer; + } + spin_unlock_irqrestore(&schan->lock, iflags); + + return &sdesc->desc; +err_xfer: + spin_unlock_irqrestore(&schan->lock, iflags); +no_desc: +err_dir: + return ERR_PTR(ret); +} + +static struct dma_async_tx_descriptor * +sirfsoc_dma_prep_cyclic(struct dma_chan *chan, dma_addr_t addr, + size_t buf_len, size_t period_len, + enum dma_transfer_direction direction) +{ + struct sirfsoc_dma_chan *schan = dma_chan_to_sirfsoc_dma_chan(chan); + struct sirfsoc_dma_desc *sdesc = NULL; + unsigned long iflags; + + /* + * we only support cycle transfer with 2 period + * If the X-length is set to 0, it would be the loop mode. + * The DMA address keeps increasing until reaching the end of a loop + * area whose size is defined by (DMA_WIDTH x (Y_LENGTH + 1)). Then + * the DMA address goes back to the beginning of this area. + * In loop mode, the DMA data region is divided into two parts, BUFA + * and BUFB. DMA controller generates interrupts twice in each loop: + * when the DMA address reaches the end of BUFA or the end of the + * BUFB + */ + if (buf_len != 2 * period_len) + return ERR_PTR(-EINVAL); + + /* Get free descriptor */ + spin_lock_irqsave(&schan->lock, iflags); + if (!list_empty(&schan->free)) { + sdesc = list_first_entry(&schan->free, struct sirfsoc_dma_desc, + node); + list_del(&sdesc->node); + } + spin_unlock_irqrestore(&schan->lock, iflags); + + if (!sdesc) + return 0; + + /* Place descriptor in prepared list */ + spin_lock_irqsave(&schan->lock, iflags); + sdesc->addr = addr; + sdesc->cyclic = 1; + sdesc->xlen = 0; + sdesc->ylen = buf_len / SIRFSOC_DMA_WORD_LEN - 1; + sdesc->width = 1; + list_add_tail(&sdesc->node, &schan->prepared); + spin_unlock_irqrestore(&schan->lock, iflags); + + return &sdesc->desc; +} + +/* + * The DMA controller consists of 16 independent DMA channels. + * Each channel is allocated to a different function + */ +bool sirfsoc_dma_filter_id(struct dma_chan *chan, void *chan_id) +{ + unsigned int ch_nr = (unsigned int) chan_id; + + if (ch_nr == chan->chan_id + + chan->device->dev_id * SIRFSOC_DMA_CHANNELS) + return true; + + return false; +} +EXPORT_SYMBOL(sirfsoc_dma_filter_id); + +static int __devinit sirfsoc_dma_probe(struct platform_device *op) +{ + struct device_node *dn = op->dev.of_node; + struct device *dev = &op->dev; + struct dma_device *dma; + struct sirfsoc_dma *sdma; + struct sirfsoc_dma_chan *schan; + struct resource res; + ulong regs_start, regs_size; + u32 id; + int ret, i; + + sdma = devm_kzalloc(dev, sizeof(*sdma), GFP_KERNEL); + if (!sdma) { + dev_err(dev, "Memory exhausted!\n"); + return -ENOMEM; + } + + if (of_property_read_u32(dn, "cell-index", &id)) { + dev_err(dev, "Fail to get DMAC index\n"); + ret = -ENODEV; + goto free_mem; + } + + sdma->irq = irq_of_parse_and_map(dn, 0); + if (sdma->irq == NO_IRQ) { + dev_err(dev, "Error mapping IRQ!\n"); + ret = -EINVAL; + goto free_mem; + } + + ret = of_address_to_resource(dn, 0, &res); + if (ret) { + dev_err(dev, "Error parsing memory region!\n"); + goto free_mem; + } + + regs_start = res.start; + regs_size = resource_size(&res); + + sdma->base = devm_ioremap(dev, regs_start, regs_size); + if (!sdma->base) { + dev_err(dev, "Error mapping memory region!\n"); + ret = -ENOMEM; + goto irq_dispose; + } + + ret = devm_request_irq(dev, sdma->irq, &sirfsoc_dma_irq, 0, DRV_NAME, + sdma); + if (ret) { + dev_err(dev, "Error requesting IRQ!\n"); + ret = -EINVAL; + goto unmap_mem; + } + + dma = &sdma->dma; + dma->dev = dev; + dma->chancnt = SIRFSOC_DMA_CHANNELS; + + dma->device_alloc_chan_resources = sirfsoc_dma_alloc_chan_resources; + dma->device_free_chan_resources = sirfsoc_dma_free_chan_resources; + dma->device_issue_pending = sirfsoc_dma_issue_pending; + dma->device_control = sirfsoc_dma_control; + dma->device_tx_status = sirfsoc_dma_tx_status; + dma->device_prep_interleaved_dma = sirfsoc_dma_prep_interleaved; + dma->device_prep_dma_cyclic = sirfsoc_dma_prep_cyclic; + + INIT_LIST_HEAD(&dma->channels); + dma_cap_set(DMA_SLAVE, dma->cap_mask); + dma_cap_set(DMA_CYCLIC, dma->cap_mask); + dma_cap_set(DMA_INTERLEAVE, dma->cap_mask); + dma_cap_set(DMA_PRIVATE, dma->cap_mask); + + for (i = 0; i < dma->chancnt; i++) { + schan = &sdma->channels[i]; + + schan->chan.device = dma; + schan->chan.cookie = 1; + schan->completed_cookie = schan->chan.cookie; + + INIT_LIST_HEAD(&schan->free); + INIT_LIST_HEAD(&schan->prepared); + INIT_LIST_HEAD(&schan->queued); + INIT_LIST_HEAD(&schan->active); + INIT_LIST_HEAD(&schan->completed); + + spin_lock_init(&schan->lock); + list_add_tail(&schan->chan.device_node, &dma->channels); + } + + tasklet_init(&sdma->tasklet, sirfsoc_dma_tasklet, (unsigned long)sdma); + + /* Register DMA engine */ + dev_set_drvdata(dev, sdma); + ret = dma_async_device_register(dma); + if (ret) + goto free_irq; + + dev_info(dev, "initialized SIRFSOC DMAC driver\n"); + + return 0; + +free_irq: + devm_free_irq(dev, sdma->irq, sdma); +irq_dispose: + irq_dispose_mapping(sdma->irq); +unmap_mem: + iounmap(sdma->base); +free_mem: + devm_kfree(dev, sdma); + return ret; +} + +static int __devexit sirfsoc_dma_remove(struct platform_device *op) +{ + struct device *dev = &op->dev; + struct sirfsoc_dma *sdma = dev_get_drvdata(dev); + + dma_async_device_unregister(&sdma->dma); + devm_free_irq(dev, sdma->irq, sdma); + irq_dispose_mapping(sdma->irq); + iounmap(sdma->base); + devm_kfree(dev, sdma); + return 0; +} + +static struct of_device_id sirfsoc_dma_match[] = { + { .compatible = "sirf,prima2-dmac", }, + {}, +}; + +static struct platform_driver sirfsoc_dma_driver = { + .probe = sirfsoc_dma_probe, + .remove = __devexit_p(sirfsoc_dma_remove), + .driver = { + .name = DRV_NAME, + .owner = THIS_MODULE, + .of_match_table = sirfsoc_dma_match, + }, +}; + +static int __init sirfsoc_dma_init(void) +{ + return platform_driver_register(&sirfsoc_dma_driver); +} +module_init(sirfsoc_dma_init); + +static void __exit sirfsoc_dma_exit(void) +{ + platform_driver_unregister(&sirfsoc_dma_driver); +} +module_exit(sirfsoc_dma_exit); + +MODULE_AUTHOR("Rongjun Ying , " + "Barry Song "); +MODULE_DESCRIPTION("SIRFSOC DMA control driver"); +MODULE_LICENSE("GPL v2"); diff --git a/include/linux/sirfsoc_dma.h b/include/linux/sirfsoc_dma.h new file mode 100644 index 000000000000..29d959333d81 --- /dev/null +++ b/include/linux/sirfsoc_dma.h @@ -0,0 +1,6 @@ +#ifndef _SIRFSOC_DMA_H_ +#define _SIRFSOC_DMA_H_ + +bool sirfsoc_dma_filter_id(struct dma_chan *chan, void *chan_id); + +#endif From 553575f1ae048aa44682b46b3c51929a0b3ad337 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Fri, 18 Nov 2011 03:32:01 -0500 Subject: [PATCH 120/820] tools turbostat: recognize and run properly on IVB Signed-off-by: Len Brown --- tools/power/x86/turbostat/turbostat.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 8b2d37b59c9e..4b05b445969e 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -809,6 +809,8 @@ int has_nehalem_turbo_ratio_limit(unsigned int family, unsigned int model) case 0x2C: /* Westmere EP - Gulftown */ case 0x2A: /* SNB */ case 0x2D: /* SNB Xeon */ + case 0x3A: /* IVB */ + case 0x3D: /* IVB Xeon */ return 1; case 0x2E: /* Nehalem-EX Xeon - Beckton */ case 0x2F: /* Westmere-EX Xeon - Eagleton */ From 7fb3e75e1833743d5faf3adbae46b63f503c6fdf Mon Sep 17 00:00:00 2001 From: Narayanan G Date: Thu, 17 Nov 2011 17:26:41 +0530 Subject: [PATCH 121/820] dmaengine/ste_dma40: support pm in dma40 This patch adds power management support to the dma40 driver. The DMA registers are backed up and restored, during suspend/resume. Also flags to track the dma usage have been introduced to facilitate this. Patch also includes few other minor changes, related to formatting, comments. Signed-off-by: Narayanan G Acked-by: Linus Walleij Signed-off-by: Vinod Koul --- drivers/dma/ste_dma40.c | 267 ++++++++++++++++++++++++++++++++++--- drivers/dma/ste_dma40_ll.h | 11 ++ 2 files changed, 259 insertions(+), 19 deletions(-) diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c index 15b311d54b74..c2cf8cfaf7d4 100644 --- a/drivers/dma/ste_dma40.c +++ b/drivers/dma/ste_dma40.c @@ -14,6 +14,8 @@ #include #include #include +#include +#include #include #include @@ -32,6 +34,9 @@ /* Maximum iterations taken before giving up suspending a channel */ #define D40_SUSPEND_MAX_IT 500 +/* Milliseconds */ +#define DMA40_AUTOSUSPEND_DELAY 100 + /* Hardware requirement on LCLA alignment */ #define LCLA_ALIGNMENT 0x40000 @@ -62,6 +67,55 @@ enum d40_command { D40_DMA_SUSPENDED = 3 }; +/* + * These are the registers that has to be saved and later restored + * when the DMA hw is powered off. + * TODO: Add save/restore of D40_DREG_GCC on dma40 v3 or later, if that works. + */ +static u32 d40_backup_regs[] = { + D40_DREG_LCPA, + D40_DREG_LCLA, + D40_DREG_PRMSE, + D40_DREG_PRMSO, + D40_DREG_PRMOE, + D40_DREG_PRMOO, +}; + +#define BACKUP_REGS_SZ ARRAY_SIZE(d40_backup_regs) + +/* TODO: Check if all these registers have to be saved/restored on dma40 v3 */ +static u32 d40_backup_regs_v3[] = { + D40_DREG_PSEG1, + D40_DREG_PSEG2, + D40_DREG_PSEG3, + D40_DREG_PSEG4, + D40_DREG_PCEG1, + D40_DREG_PCEG2, + D40_DREG_PCEG3, + D40_DREG_PCEG4, + D40_DREG_RSEG1, + D40_DREG_RSEG2, + D40_DREG_RSEG3, + D40_DREG_RSEG4, + D40_DREG_RCEG1, + D40_DREG_RCEG2, + D40_DREG_RCEG3, + D40_DREG_RCEG4, +}; + +#define BACKUP_REGS_SZ_V3 ARRAY_SIZE(d40_backup_regs_v3) + +static u32 d40_backup_regs_chan[] = { + D40_CHAN_REG_SSCFG, + D40_CHAN_REG_SSELT, + D40_CHAN_REG_SSPTR, + D40_CHAN_REG_SSLNK, + D40_CHAN_REG_SDCFG, + D40_CHAN_REG_SDELT, + D40_CHAN_REG_SDPTR, + D40_CHAN_REG_SDLNK, +}; + /** * struct d40_lli_pool - Structure for keeping LLIs in memory * @@ -96,7 +150,7 @@ struct d40_lli_pool { * during a transfer. * @node: List entry. * @is_in_client_list: true if the client owns this descriptor. - * the previous one. + * @cyclic: true if this is a cyclic job * * This descriptor is used for both logical and physical transfers. */ @@ -143,6 +197,7 @@ struct d40_lcla_pool { * channels. * * @lock: A lock protection this entity. + * @reserved: True if used by secure world or otherwise. * @num: The physical channel number of this entity. * @allocated_src: Bit mapped to show which src event line's are mapped to * this physical channel. Can also be free or physically allocated. @@ -152,6 +207,7 @@ struct d40_lcla_pool { */ struct d40_phy_res { spinlock_t lock; + bool reserved; int num; u32 allocated_src; u32 allocated_dst; @@ -185,7 +241,6 @@ struct d40_base; * @src_def_cfg: Default cfg register setting for src. * @dst_def_cfg: Default cfg register setting for dst. * @log_def: Default logical channel settings. - * @lcla: Space for one dst src pair for logical channel transfers. * @lcpa: Pointer to dst and src lcpa settings. * @runtime_addr: runtime configured address. * @runtime_direction: runtime configured direction. @@ -241,6 +296,7 @@ struct d40_chan { * @dma_both: dma_device channels that can do both memcpy and slave transfers. * @dma_slave: dma_device channels that can do only do slave transfers. * @dma_memcpy: dma_device channels that can do only do memcpy transfers. + * @phy_chans: Room for all possible physical channels in system. * @log_chans: Room for all possible logical channels in system. * @lookup_log_chans: Used to map interrupt number to logical channel. Points * to log_chans entries. @@ -254,6 +310,13 @@ struct d40_chan { * @phy_lcpa: The physical address of the LCPA. * @lcpa_size: The size of the LCPA area. * @desc_slab: cache for descriptors. + * @reg_val_backup: Here the values of some hardware registers are stored + * before the DMA is powered off. They are restored when the power is back on. + * @reg_val_backup_v3: Backup of registers that only exits on dma40 v3 and + * later. + * @reg_val_backup_chan: Backup data for standard channel parameter registers. + * @gcc_pwr_off_mask: Mask to maintain the channels that can be turned off. + * @initialized: true if the dma has been initialized */ struct d40_base { spinlock_t interrupt_lock; @@ -282,6 +345,11 @@ struct d40_base { dma_addr_t phy_lcpa; resource_size_t lcpa_size; struct kmem_cache *desc_slab; + u32 reg_val_backup[BACKUP_REGS_SZ]; + u32 reg_val_backup_v3[BACKUP_REGS_SZ_V3]; + u32 *reg_val_backup_chan; + u16 gcc_pwr_off_mask; + bool initialized; }; /** @@ -479,13 +547,14 @@ static struct d40_desc *d40_desc_get(struct d40_chan *d40c) struct d40_desc *d; struct d40_desc *_d; - list_for_each_entry_safe(d, _d, &d40c->client, node) + list_for_each_entry_safe(d, _d, &d40c->client, node) { if (async_tx_test_ack(&d->txd)) { d40_desc_remove(d); desc = d; memset(desc, 0, sizeof(*desc)); break; } + } } if (!desc) @@ -740,7 +809,61 @@ static int d40_sg_2_dmalen(struct scatterlist *sgl, int sg_len, return len; } -/* Support functions for logical channels */ + +#ifdef CONFIG_PM +static void dma40_backup(void __iomem *baseaddr, u32 *backup, + u32 *regaddr, int num, bool save) +{ + int i; + + for (i = 0; i < num; i++) { + void __iomem *addr = baseaddr + regaddr[i]; + + if (save) + backup[i] = readl_relaxed(addr); + else + writel_relaxed(backup[i], addr); + } +} + +static void d40_save_restore_registers(struct d40_base *base, bool save) +{ + int i; + + /* Save/Restore channel specific registers */ + for (i = 0; i < base->num_phy_chans; i++) { + void __iomem *addr; + int idx; + + if (base->phy_res[i].reserved) + continue; + + addr = base->virtbase + D40_DREG_PCBASE + i * D40_DREG_PCDELTA; + idx = i * ARRAY_SIZE(d40_backup_regs_chan); + + dma40_backup(addr, &base->reg_val_backup_chan[idx], + d40_backup_regs_chan, + ARRAY_SIZE(d40_backup_regs_chan), + save); + } + + /* Save/Restore global registers */ + dma40_backup(base->virtbase, base->reg_val_backup, + d40_backup_regs, ARRAY_SIZE(d40_backup_regs), + save); + + /* Save/Restore registers only existing on dma40 v3 and later */ + if (base->rev >= 3) + dma40_backup(base->virtbase, base->reg_val_backup_v3, + d40_backup_regs_v3, + ARRAY_SIZE(d40_backup_regs_v3), + save); +} +#else +static void d40_save_restore_registers(struct d40_base *base, bool save) +{ +} +#endif static int d40_channel_execute_command(struct d40_chan *d40c, enum d40_command command) @@ -1013,6 +1136,7 @@ static int d40_pause(struct d40_chan *d40c) if (!d40c->busy) return 0; + pm_runtime_get_sync(d40c->base->dev); spin_lock_irqsave(&d40c->lock, flags); res = d40_channel_execute_command(d40c, D40_DMA_SUSPEND_REQ); @@ -1025,7 +1149,8 @@ static int d40_pause(struct d40_chan *d40c) D40_DMA_RUN); } } - + pm_runtime_mark_last_busy(d40c->base->dev); + pm_runtime_put_autosuspend(d40c->base->dev); spin_unlock_irqrestore(&d40c->lock, flags); return res; } @@ -1039,7 +1164,7 @@ static int d40_resume(struct d40_chan *d40c) return 0; spin_lock_irqsave(&d40c->lock, flags); - + pm_runtime_get_sync(d40c->base->dev); if (d40c->base->rev == 0) if (chan_is_logical(d40c)) { res = d40_channel_execute_command(d40c, @@ -1057,6 +1182,8 @@ static int d40_resume(struct d40_chan *d40c) } no_suspend: + pm_runtime_mark_last_busy(d40c->base->dev); + pm_runtime_put_autosuspend(d40c->base->dev); spin_unlock_irqrestore(&d40c->lock, flags); return res; } @@ -1129,7 +1256,10 @@ static struct d40_desc *d40_queue_start(struct d40_chan *d40c) d40d = d40_first_queued(d40c); if (d40d != NULL) { - d40c->busy = true; + if (!d40c->busy) + d40c->busy = true; + + pm_runtime_get_sync(d40c->base->dev); /* Remove from queue */ d40_desc_remove(d40d); @@ -1190,6 +1320,8 @@ static void dma_tc_handle(struct d40_chan *d40c) if (d40_queue_start(d40c) == NULL) d40c->busy = false; + pm_runtime_mark_last_busy(d40c->base->dev); + pm_runtime_put_autosuspend(d40c->base->dev); } d40c->pending_tx++; @@ -1643,10 +1775,11 @@ static int d40_free_dma(struct d40_chan *d40c) return -EINVAL; } + pm_runtime_get_sync(d40c->base->dev); res = d40_channel_execute_command(d40c, D40_DMA_SUSPEND_REQ); if (res) { chan_err(d40c, "suspend failed\n"); - return res; + goto out; } if (chan_is_logical(d40c)) { @@ -1664,13 +1797,11 @@ static int d40_free_dma(struct d40_chan *d40c) if (d40_chan_has_events(d40c)) { res = d40_channel_execute_command(d40c, D40_DMA_RUN); - if (res) { + if (res) chan_err(d40c, "Executing RUN command\n"); - return res; - } } - return 0; + goto out; } } else { (void) d40_alloc_mask_free(phy, is_src, 0); @@ -1680,13 +1811,23 @@ static int d40_free_dma(struct d40_chan *d40c) res = d40_channel_execute_command(d40c, D40_DMA_STOP); if (res) { chan_err(d40c, "Failed to stop channel\n"); - return res; + goto out; } + + if (d40c->busy) { + pm_runtime_mark_last_busy(d40c->base->dev); + pm_runtime_put_autosuspend(d40c->base->dev); + } + + d40c->busy = false; d40c->phy_chan = NULL; d40c->configured = false; d40c->base->lookup_phy_chans[phy->num] = NULL; +out: - return 0; + pm_runtime_mark_last_busy(d40c->base->dev); + pm_runtime_put_autosuspend(d40c->base->dev); + return res; } static bool d40_is_paused(struct d40_chan *d40c) @@ -2016,9 +2157,11 @@ static int d40_alloc_chan_resources(struct dma_chan *chan) err = d40_allocate_channel(d40c); if (err) { chan_err(d40c, "Failed to allocate channel\n"); + d40c->configured = false; goto fail; } + pm_runtime_get_sync(d40c->base->dev); /* Fill in basic CFG register values */ d40_phy_cfg(&d40c->dma_cfg, &d40c->src_def_cfg, &d40c->dst_def_cfg, chan_is_logical(d40c)); @@ -2046,6 +2189,8 @@ static int d40_alloc_chan_resources(struct dma_chan *chan) if (is_free_phy) d40_config_write(d40c); fail: + pm_runtime_mark_last_busy(d40c->base->dev); + pm_runtime_put_autosuspend(d40c->base->dev); spin_unlock_irqrestore(&d40c->lock, flags); return err; } @@ -2519,6 +2664,55 @@ failure1: return err; } +/* Suspend resume functionality */ +#ifdef CONFIG_PM +static int dma40_pm_suspend(struct device *dev) +{ + if (!pm_runtime_suspended(dev)) + return -EBUSY; + + return 0; +} + +static int dma40_runtime_suspend(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct d40_base *base = platform_get_drvdata(pdev); + + d40_save_restore_registers(base, true); + + /* Don't disable/enable clocks for v1 due to HW bugs */ + if (base->rev != 1) + writel_relaxed(base->gcc_pwr_off_mask, + base->virtbase + D40_DREG_GCC); + + return 0; +} + +static int dma40_runtime_resume(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct d40_base *base = platform_get_drvdata(pdev); + + if (base->initialized) + d40_save_restore_registers(base, false); + + writel_relaxed(D40_DREG_GCC_ENABLE_ALL, + base->virtbase + D40_DREG_GCC); + return 0; +} + + +static const struct dev_pm_ops dma40_pm_ops = { + .suspend = dma40_pm_suspend, + .runtime_suspend = dma40_runtime_suspend, + .runtime_resume = dma40_runtime_resume, +}; +#define DMA40_PM_OPS (&dma40_pm_ops) +#else +#define DMA40_PM_OPS NULL +#endif + /* Initialization functions. */ static int __init d40_phy_res_init(struct d40_base *base) @@ -2527,6 +2721,7 @@ static int __init d40_phy_res_init(struct d40_base *base) int num_phy_chans_avail = 0; u32 val[2]; int odd_even_bit = -2; + int gcc = D40_DREG_GCC_ENA; val[0] = readl(base->virtbase + D40_DREG_PRSME); val[1] = readl(base->virtbase + D40_DREG_PRSMO); @@ -2538,9 +2733,17 @@ static int __init d40_phy_res_init(struct d40_base *base) /* Mark security only channels as occupied */ base->phy_res[i].allocated_src = D40_ALLOC_PHY; base->phy_res[i].allocated_dst = D40_ALLOC_PHY; + base->phy_res[i].reserved = true; + gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(i), + D40_DREG_GCC_SRC); + gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(i), + D40_DREG_GCC_DST); + + } else { base->phy_res[i].allocated_src = D40_ALLOC_FREE; base->phy_res[i].allocated_dst = D40_ALLOC_FREE; + base->phy_res[i].reserved = false; num_phy_chans_avail++; } spin_lock_init(&base->phy_res[i].lock); @@ -2552,6 +2755,11 @@ static int __init d40_phy_res_init(struct d40_base *base) base->phy_res[chan].allocated_src = D40_ALLOC_PHY; base->phy_res[chan].allocated_dst = D40_ALLOC_PHY; + base->phy_res[chan].reserved = true; + gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(chan), + D40_DREG_GCC_SRC); + gcc |= D40_DREG_GCC_EVTGRP_ENA(D40_PHYS_TO_GROUP(chan), + D40_DREG_GCC_DST); num_phy_chans_avail--; } @@ -2572,6 +2780,15 @@ static int __init d40_phy_res_init(struct d40_base *base) val[0] = val[0] >> 2; } + /* + * To keep things simple, Enable all clocks initially. + * The clocks will get managed later post channel allocation. + * The clocks for the event lines on which reserved channels exists + * are not managed here. + */ + writel(D40_DREG_GCC_ENABLE_ALL, base->virtbase + D40_DREG_GCC); + base->gcc_pwr_off_mask = gcc; + return num_phy_chans_avail; } @@ -2699,10 +2916,15 @@ static struct d40_base * __init d40_hw_detect_init(struct platform_device *pdev) goto failure; } - base->lcla_pool.alloc_map = kzalloc(num_phy_chans * - sizeof(struct d40_desc *) * - D40_LCLA_LINK_PER_EVENT_GRP, + base->reg_val_backup_chan = kmalloc(base->num_phy_chans * + sizeof(d40_backup_regs_chan), GFP_KERNEL); + if (!base->reg_val_backup_chan) + goto failure; + + base->lcla_pool.alloc_map = + kzalloc(num_phy_chans * sizeof(struct d40_desc *) + * D40_LCLA_LINK_PER_EVENT_GRP, GFP_KERNEL); if (!base->lcla_pool.alloc_map) goto failure; @@ -2741,9 +2963,9 @@ failure: static void __init d40_hw_init(struct d40_base *base) { - static const struct d40_reg_val dma_init_reg[] = { + static struct d40_reg_val dma_init_reg[] = { /* Clock every part of the DMA block from start */ - { .reg = D40_DREG_GCC, .val = 0x0000ff01}, + { .reg = D40_DREG_GCC, .val = D40_DREG_GCC_ENABLE_ALL}, /* Interrupts on all logical channels */ { .reg = D40_DREG_LCMIS0, .val = 0xFFFFFFFF}, @@ -2960,6 +3182,12 @@ static int __init d40_probe(struct platform_device *pdev) goto failure; } + pm_runtime_irq_safe(base->dev); + pm_runtime_set_autosuspend_delay(base->dev, DMA40_AUTOSUSPEND_DELAY); + pm_runtime_use_autosuspend(base->dev); + pm_runtime_enable(base->dev); + pm_runtime_resume(base->dev); + base->initialized = true; err = d40_dmaengine_init(base, num_reserved_chans); if (err) goto failure; @@ -3013,6 +3241,7 @@ static struct platform_driver d40_driver = { .driver = { .owner = THIS_MODULE, .name = D40_NAME, + .pm = DMA40_PM_OPS, }, }; diff --git a/drivers/dma/ste_dma40_ll.h b/drivers/dma/ste_dma40_ll.h index b44c455158de..8d3d490968a3 100644 --- a/drivers/dma/ste_dma40_ll.h +++ b/drivers/dma/ste_dma40_ll.h @@ -16,6 +16,8 @@ #define D40_TYPE_TO_GROUP(type) (type / 16) #define D40_TYPE_TO_EVENT(type) (type % 16) +#define D40_GROUP_SIZE 8 +#define D40_PHYS_TO_GROUP(phys) ((phys & (D40_GROUP_SIZE - 1)) / 2) /* Most bits of the CFG register are the same in log as in phy mode */ #define D40_SREG_CFG_MST_POS 15 @@ -123,6 +125,15 @@ /* DMA Register Offsets */ #define D40_DREG_GCC 0x000 +#define D40_DREG_GCC_ENA 0x1 +/* This assumes that there are only 4 event groups */ +#define D40_DREG_GCC_ENABLE_ALL 0xff01 +#define D40_DREG_GCC_EVTGRP_POS 8 +#define D40_DREG_GCC_SRC 0 +#define D40_DREG_GCC_DST 1 +#define D40_DREG_GCC_EVTGRP_ENA(x, y) \ + (1 << (D40_DREG_GCC_EVTGRP_POS + 2 * x + y)) + #define D40_DREG_PRTYP 0x004 #define D40_DREG_PRSME 0x008 #define D40_DREG_PRSMO 0x00C From e69664336d6400cc1685716ee5f04ef74e85703e Mon Sep 17 00:00:00 2001 From: Huang Shijie Date: Fri, 18 Nov 2011 16:38:02 +0800 Subject: [PATCH 122/820] IMX/DMA : set the DMA direction in the sdma_control() Set the right DMA direction in the sdma_control(), else we will get the wrong log when enable the DYNAMIC_DEBUG. Signed-off-by: Huang Shijie Signed-off-by: Vinod Koul --- drivers/dma/imx-sdma.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c index 2e0357178d19..d9e5933d91eb 100644 --- a/drivers/dma/imx-sdma.c +++ b/drivers/dma/imx-sdma.c @@ -1102,6 +1102,7 @@ static int sdma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, sdmac->watermark_level = dmaengine_cfg->dst_maxburst; sdmac->word_size = dmaengine_cfg->dst_addr_width; } + sdmac->direction = dmaengine_cfg->direction; return sdma_config_channel(sdmac); default: return -ENOSYS; From 61e183f8306934a9f66557f69f1f0f56f18dca06 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 17 Nov 2011 16:01:29 +0530 Subject: [PATCH 123/820] dmaengine/dw_dmac: Reconfigure interrupt and chan_cfg register on resume In S2R all DMA registers are reset by hardware and thus they are required to be reprogrammed. The channels which aren't reprogrammed are channel configuration and interrupt enable registers, which are currently programmed at chan_alloc time. This patch creates another routine to initialize a channel. It will try to initialize channel on every dwc_dostart() call. If channel is already initialised then it simply returns, otherwise it configures registers. This routine will also initialize registers on wakeup from S2R, as we mark channels as uninitialized on suspend. Signed-off-by: Viresh Kumar Signed-off-by: Vinod Koul --- drivers/dma/dw_dmac.c | 69 ++++++++++++++++++++++---------------- drivers/dma/dw_dmac_regs.h | 1 + 2 files changed, 42 insertions(+), 28 deletions(-) diff --git a/drivers/dma/dw_dmac.c b/drivers/dma/dw_dmac.c index decca1c3c83d..9b592b02b5f4 100644 --- a/drivers/dma/dw_dmac.c +++ b/drivers/dma/dw_dmac.c @@ -166,6 +166,38 @@ dwc_assign_cookie(struct dw_dma_chan *dwc, struct dw_desc *desc) return cookie; } +static void dwc_initialize(struct dw_dma_chan *dwc) +{ + struct dw_dma *dw = to_dw_dma(dwc->chan.device); + struct dw_dma_slave *dws = dwc->chan.private; + u32 cfghi = DWC_CFGH_FIFO_MODE; + u32 cfglo = DWC_CFGL_CH_PRIOR(dwc->priority); + + if (dwc->initialized == true) + return; + + if (dws) { + /* + * We need controller-specific data to set up slave + * transfers. + */ + BUG_ON(!dws->dma_dev || dws->dma_dev != dw->dma.dev); + + cfghi = dws->cfg_hi; + cfglo |= dws->cfg_lo & ~DWC_CFGL_CH_PRIOR_MASK; + } + + channel_writel(dwc, CFG_LO, cfglo); + channel_writel(dwc, CFG_HI, cfghi); + + /* Enable interrupts */ + channel_set_bit(dw, MASK.XFER, dwc->mask); + channel_set_bit(dw, MASK.BLOCK, dwc->mask); + channel_set_bit(dw, MASK.ERROR, dwc->mask); + + dwc->initialized = true; +} + /*----------------------------------------------------------------------*/ /* Called with dwc->lock held and bh disabled */ @@ -189,6 +221,8 @@ static void dwc_dostart(struct dw_dma_chan *dwc, struct dw_desc *first) return; } + dwc_initialize(dwc); + channel_writel(dwc, LLP, first->txd.phys); channel_writel(dwc, CTL_LO, DWC_CTLL_LLP_D_EN | DWC_CTLL_LLP_S_EN); @@ -959,10 +993,7 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan) struct dw_dma_chan *dwc = to_dw_dma_chan(chan); struct dw_dma *dw = to_dw_dma(chan->device); struct dw_desc *desc; - struct dw_dma_slave *dws; int i; - u32 cfghi; - u32 cfglo; unsigned long flags; dev_vdbg(chan2dev(chan), "alloc_chan_resources\n"); @@ -975,26 +1006,6 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan) dwc->completed = chan->cookie = 1; - cfghi = DWC_CFGH_FIFO_MODE; - cfglo = 0; - - dws = chan->private; - if (dws) { - /* - * We need controller-specific data to set up slave - * transfers. - */ - BUG_ON(!dws->dma_dev || dws->dma_dev != dw->dma.dev); - - cfghi = dws->cfg_hi; - cfglo = dws->cfg_lo & ~DWC_CFGL_CH_PRIOR_MASK; - } - - cfglo |= DWC_CFGL_CH_PRIOR(dwc->priority); - - channel_writel(dwc, CFG_LO, cfglo); - channel_writel(dwc, CFG_HI, cfghi); - /* * NOTE: some controllers may have additional features that we * need to initialize here, like "scatter-gather" (which @@ -1026,11 +1037,6 @@ static int dwc_alloc_chan_resources(struct dma_chan *chan) i = ++dwc->descs_allocated; } - /* Enable interrupts */ - channel_set_bit(dw, MASK.XFER, dwc->mask); - channel_set_bit(dw, MASK.BLOCK, dwc->mask); - channel_set_bit(dw, MASK.ERROR, dwc->mask); - spin_unlock_irqrestore(&dwc->lock, flags); dev_dbg(chan2dev(chan), @@ -1058,6 +1064,7 @@ static void dwc_free_chan_resources(struct dma_chan *chan) spin_lock_irqsave(&dwc->lock, flags); list_splice_init(&dwc->free_list, &list); dwc->descs_allocated = 0; + dwc->initialized = false; /* Disable interrupts */ channel_clear_bit(dw, MASK.XFER, dwc->mask); @@ -1335,6 +1342,8 @@ EXPORT_SYMBOL(dw_dma_cyclic_free); static void dw_dma_off(struct dw_dma *dw) { + int i; + dma_writel(dw, CFG, 0); channel_clear_bit(dw, MASK.XFER, dw->all_chan_mask); @@ -1345,6 +1354,9 @@ static void dw_dma_off(struct dw_dma *dw) while (dma_readl(dw, CFG) & DW_CFG_DMA_EN) cpu_relax(); + + for (i = 0; i < dw->dma.chancnt; i++) + dw->chan[i].initialized = false; } static int __init dw_probe(struct platform_device *pdev) @@ -1533,6 +1545,7 @@ static int dw_suspend_noirq(struct device *dev) dw_dma_off(platform_get_drvdata(pdev)); clk_disable(dw->clk); + return 0; } diff --git a/drivers/dma/dw_dmac_regs.h b/drivers/dma/dw_dmac_regs.h index c3419518d701..5eef6946a367 100644 --- a/drivers/dma/dw_dmac_regs.h +++ b/drivers/dma/dw_dmac_regs.h @@ -140,6 +140,7 @@ struct dw_dma_chan { u8 mask; u8 priority; bool paused; + bool initialized; spinlock_t lock; From d5613947addb1e1096a3be29dfe817905e385469 Mon Sep 17 00:00:00 2001 From: Vinod Koul Date: Mon, 28 Nov 2011 08:51:16 +0530 Subject: [PATCH 124/820] dmaengine: timb_dma: fix the spare warinings Sparse warns: drivers/dma/timb_dma.c:168:17: warning: mixing different enum types drivers/dma/timb_dma.c:168:17: int enum dma_transfer_direction versus drivers/dma/timb_dma.c:168:17: int enum dma_data_direction drivers/dma/timb_dma.c:172:32: warning: mixing different enum types drivers/dma/timb_dma.c:172:32: int enum dma_transfer_direction versus drivers/dma/timb_dma.c:172:32: int enum dma_data_direction Reported-by: Dan Carpenter Signed-off-by: Vinod Koul --- drivers/dma/timb_dma.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c index 8c880729b094..d142ae1fce52 100644 --- a/drivers/dma/timb_dma.c +++ b/drivers/dma/timb_dma.c @@ -166,10 +166,10 @@ static void __td_unmap_desc(struct timb_dma_chan *td_chan, const u8 *dma_desc, if (single) dma_unmap_single(chan2dev(&td_chan->chan), addr, len, - td_chan->direction); + DMA_TO_DEVICE); else dma_unmap_page(chan2dev(&td_chan->chan), addr, len, - td_chan->direction); + DMA_TO_DEVICE); } static void __td_unmap_descs(struct timb_dma_desc *td_desc, bool single) @@ -398,7 +398,7 @@ static struct timb_dma_desc *td_alloc_init_desc(struct timb_dma_chan *td_chan) td_desc->txd.flags = DMA_CTRL_ACK; td_desc->txd.phys = dma_map_single(chan2dmadev(chan), - td_desc->desc_list, td_desc->desc_list_len, DMA_MEM_TO_DEV); + td_desc->desc_list, td_desc->desc_list_len, DMA_TO_DEVICE); err = dma_mapping_error(chan2dmadev(chan), td_desc->txd.phys); if (err) { @@ -419,7 +419,7 @@ static void td_free_desc(struct timb_dma_desc *td_desc) { dev_dbg(chan2dev(td_desc->txd.chan), "Freeing desc: %p\n", td_desc); dma_unmap_single(chan2dmadev(td_desc->txd.chan), td_desc->txd.phys, - td_desc->desc_list_len, DMA_MEM_TO_DEV); + td_desc->desc_list_len, DMA_TO_DEVICE); kfree(td_desc->desc_list); kfree(td_desc); From 28c7a19d230228ab9ae61c300c5003a2400fadd3 Mon Sep 17 00:00:00 2001 From: Narayanan G Date: Tue, 22 Nov 2011 13:56:55 +0530 Subject: [PATCH 125/820] dmaengine/ste_dma40: Add support to use lcla area from esram This patch provides an option of having the lcla (link address) in ESRAM instead of allocating it. The bool value (use_esram_lcla) in the stedma40_platform_data if set to true, then the lcla address would be taken from platform resources. Also, the corresponding esram regulator is managed in the suspend/resume functions. Signed-off-by: Narayanan G Signed-off-by: Vinod Koul --- .../arm/plat-nomadik/include/plat/ste_dma40.h | 1 + drivers/dma/ste_dma40.c | 95 +++++++++++++++++-- 2 files changed, 86 insertions(+), 10 deletions(-) diff --git a/arch/arm/plat-nomadik/include/plat/ste_dma40.h b/arch/arm/plat-nomadik/include/plat/ste_dma40.h index 38b041a40db4..3177bed253df 100644 --- a/arch/arm/plat-nomadik/include/plat/ste_dma40.h +++ b/arch/arm/plat-nomadik/include/plat/ste_dma40.h @@ -153,6 +153,7 @@ struct stedma40_platform_data { struct stedma40_chan_cfg *memcpy_conf_phy; struct stedma40_chan_cfg *memcpy_conf_log; int disabled_channels[STEDMA40_MAX_PHYS]; + bool use_esram_lcla; }; #ifdef CONFIG_STE_DMA40 diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c index c2cf8cfaf7d4..aff128a39a4b 100644 --- a/drivers/dma/ste_dma40.c +++ b/drivers/dma/ste_dma40.c @@ -304,6 +304,7 @@ struct d40_chan { * to phy_chans entries. * @plat_data: Pointer to provided platform_data which is the driver * configuration. + * @lcpa_regulator: Pointer to hold the regulator for the esram bank for lcla. * @phy_res: Vector containing all physical channels. * @lcla_pool: lcla pool settings and data. * @lcpa_base: The virtual mapped address of LCPA. @@ -338,6 +339,7 @@ struct d40_base { struct d40_chan **lookup_log_chans; struct d40_chan **lookup_phy_chans; struct stedma40_platform_data *plat_data; + struct regulator *lcpa_regulator; /* Physical half channels */ struct d40_phy_res *phy_res; struct d40_lcla_pool lcla_pool; @@ -605,6 +607,7 @@ static void d40_log_lli_to_lcxa(struct d40_chan *chan, struct d40_desc *desc) bool cyclic = desc->cyclic; int curr_lcla = -EINVAL; int first_lcla = 0; + bool use_esram_lcla = chan->base->plat_data->use_esram_lcla; bool linkback; /* @@ -677,11 +680,16 @@ static void d40_log_lli_to_lcxa(struct d40_chan *chan, struct d40_desc *desc) &lli->src[lli_current], next_lcla, flags); - dma_sync_single_range_for_device(chan->base->dev, - pool->dma_addr, lcla_offset, - 2 * sizeof(struct d40_log_lli), - DMA_TO_DEVICE); - + /* + * Cache maintenance is not needed if lcla is + * mapped in esram + */ + if (!use_esram_lcla) { + dma_sync_single_range_for_device(chan->base->dev, + pool->dma_addr, lcla_offset, + 2 * sizeof(struct d40_log_lli), + DMA_TO_DEVICE); + } curr_lcla = next_lcla; if (curr_lcla == -EINVAL || curr_lcla == first_lcla) { @@ -2668,10 +2676,15 @@ failure1: #ifdef CONFIG_PM static int dma40_pm_suspend(struct device *dev) { + struct platform_device *pdev = to_platform_device(dev); + struct d40_base *base = platform_get_drvdata(pdev); + int ret = 0; if (!pm_runtime_suspended(dev)) return -EBUSY; - return 0; + if (base->lcpa_regulator) + ret = regulator_disable(base->lcpa_regulator); + return ret; } static int dma40_runtime_suspend(struct device *dev) @@ -2702,11 +2715,23 @@ static int dma40_runtime_resume(struct device *dev) return 0; } +static int dma40_resume(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct d40_base *base = platform_get_drvdata(pdev); + int ret = 0; + + if (base->lcpa_regulator) + ret = regulator_enable(base->lcpa_regulator); + + return ret; +} static const struct dev_pm_ops dma40_pm_ops = { .suspend = dma40_pm_suspend, .runtime_suspend = dma40_runtime_suspend, .runtime_resume = dma40_runtime_resume, + .resume = dma40_resume, }; #define DMA40_PM_OPS (&dma40_pm_ops) #else @@ -3165,11 +3190,31 @@ static int __init d40_probe(struct platform_device *pdev) d40_err(&pdev->dev, "Failed to ioremap LCPA region\n"); goto failure; } + /* If lcla has to be located in ESRAM we don't need to allocate */ + if (base->plat_data->use_esram_lcla) { + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, + "lcla_esram"); + if (!res) { + ret = -ENOENT; + d40_err(&pdev->dev, + "No \"lcla_esram\" memory resource\n"); + goto failure; + } + base->lcla_pool.base = ioremap(res->start, + resource_size(res)); + if (!base->lcla_pool.base) { + ret = -ENOMEM; + d40_err(&pdev->dev, "Failed to ioremap LCLA region\n"); + goto failure; + } + writel(res->start, base->virtbase + D40_DREG_LCLA); - ret = d40_lcla_allocate(base); - if (ret) { - d40_err(&pdev->dev, "Failed to allocate LCLA area\n"); - goto failure; + } else { + ret = d40_lcla_allocate(base); + if (ret) { + d40_err(&pdev->dev, "Failed to allocate LCLA area\n"); + goto failure; + } } spin_lock_init(&base->lcla_pool.lock); @@ -3187,6 +3232,26 @@ static int __init d40_probe(struct platform_device *pdev) pm_runtime_use_autosuspend(base->dev); pm_runtime_enable(base->dev); pm_runtime_resume(base->dev); + + if (base->plat_data->use_esram_lcla) { + + base->lcpa_regulator = regulator_get(base->dev, "lcla_esram"); + if (IS_ERR(base->lcpa_regulator)) { + d40_err(&pdev->dev, "Failed to get lcpa_regulator\n"); + base->lcpa_regulator = NULL; + goto failure; + } + + ret = regulator_enable(base->lcpa_regulator); + if (ret) { + d40_err(&pdev->dev, + "Failed to enable lcpa_regulator\n"); + regulator_put(base->lcpa_regulator); + base->lcpa_regulator = NULL; + goto failure; + } + } + base->initialized = true; err = d40_dmaengine_init(base, num_reserved_chans); if (err) @@ -3204,6 +3269,11 @@ failure: if (base->virtbase) iounmap(base->virtbase); + if (base->lcla_pool.base && base->plat_data->use_esram_lcla) { + iounmap(base->lcla_pool.base); + base->lcla_pool.base = NULL; + } + if (base->lcla_pool.dma_addr) dma_unmap_single(base->dev, base->lcla_pool.dma_addr, SZ_1K * base->num_phy_chans, @@ -3226,6 +3296,11 @@ failure: clk_put(base->clk); } + if (base->lcpa_regulator) { + regulator_disable(base->lcpa_regulator); + regulator_put(base->lcpa_regulator); + } + kfree(base->lcla_pool.alloc_map); kfree(base->lookup_log_chans); kfree(base->lookup_phy_chans); From 0ab88a01832d529c48e2a51f3d24662c143df39a Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Tue, 22 Nov 2011 11:55:52 +0100 Subject: [PATCH 126/820] dmaengine: at_hdmac: add const to struct platform_device_id definition Signed-off-by: Nicolas Ferre Signed-off-by: Vinod Koul --- drivers/dma/at_hdmac.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index 3b0c28ff4ad6..3b44f24888ea 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -1191,7 +1191,7 @@ static const struct of_device_id atmel_dma_dt_ids[] = { MODULE_DEVICE_TABLE(of, atmel_dma_dt_ids); #endif -static struct platform_device_id atdma_devtypes[] = { +static const struct platform_device_id atdma_devtypes[] = { { .name = "at91sam9rl_dma", .driver_data = ATDMA_DEVTYPE_SAM9RL, From dcc817346d311bbd9dac6fc9d6bddb552134851c Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Tue, 22 Nov 2011 11:55:53 +0100 Subject: [PATCH 127/820] dmaengine: at_hdmac: empty entry shape modification Trivial form modification to unify structure look. Signed-off-by: Nicolas Ferre Signed-off-by: Vinod Koul --- drivers/dma/at_hdmac.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index 3b44f24888ea..30ec030456cf 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -1185,7 +1185,9 @@ static const struct of_device_id atmel_dma_dt_ids[] = { }, { .compatible = "atmel,at91sam9g45-dma", .data = (void *)ATDMA_DEVTYPE_SAM9G45 - }, { /* sentinel */ } + }, { + /* sentinel */ + } }; MODULE_DEVICE_TABLE(of, atmel_dma_dt_ids); From 02f88be9488a3d831f073c1161b1e5feacb9d3ec Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Tue, 22 Nov 2011 11:55:54 +0100 Subject: [PATCH 128/820] dmaengine: at_hdmac: simplify device selection from platform data or DT Using a configuration structure simplify the finding of SoC dependent parameters. Both platform data and device tree ids are using these structures. This will separate data from code and remove the need for an enum. Idea from Grant Likely. Signed-off-by: Nicolas Ferre Signed-off-by: Vinod Koul --- drivers/dma/at_hdmac.c | 63 ++++++++++++++++++------------------- drivers/dma/at_hdmac_regs.h | 7 ----- 2 files changed, 31 insertions(+), 39 deletions(-) diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index 30ec030456cf..97f87b29b9f3 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -1177,14 +1177,22 @@ static void atc_free_chan_resources(struct dma_chan *chan) /*-- Module Management -----------------------------------------------*/ +/* cap_mask is a multi-u32 bitfield, fill it with proper C code. */ +static struct at_dma_platform_data at91sam9rl_config = { + .nr_channels = 2, +}; +static struct at_dma_platform_data at91sam9g45_config = { + .nr_channels = 8, +}; + #if defined(CONFIG_OF) static const struct of_device_id atmel_dma_dt_ids[] = { { .compatible = "atmel,at91sam9rl-dma", - .data = (void *)ATDMA_DEVTYPE_SAM9RL + .data = &at91sam9rl_config, }, { .compatible = "atmel,at91sam9g45-dma", - .data = (void *)ATDMA_DEVTYPE_SAM9G45 + .data = &at91sam9g45_config, }, { /* sentinel */ } @@ -1196,26 +1204,27 @@ MODULE_DEVICE_TABLE(of, atmel_dma_dt_ids); static const struct platform_device_id atdma_devtypes[] = { { .name = "at91sam9rl_dma", - .driver_data = ATDMA_DEVTYPE_SAM9RL, + .driver_data = (unsigned long) &at91sam9rl_config, }, { .name = "at91sam9g45_dma", - .driver_data = ATDMA_DEVTYPE_SAM9G45, + .driver_data = (unsigned long) &at91sam9g45_config, }, { /* sentinel */ } }; -static inline enum atdma_devtype __init at_dma_get_driver_data( - struct platform_device *pdev) +static inline struct at_dma_platform_data * __init at_dma_get_driver_data( + struct platform_device *pdev) { if (pdev->dev.of_node) { const struct of_device_id *match; match = of_match_node(atmel_dma_dt_ids, pdev->dev.of_node); if (match == NULL) - return ATDMA_DEVTYPE_UNDEFINED; - return (enum atdma_devtype)match->data; + return NULL; + return match->data; } - return platform_get_device_id(pdev)->driver_data; + return (struct at_dma_platform_data *) + platform_get_device_id(pdev)->driver_data; } /** @@ -1242,26 +1251,17 @@ static int __init at_dma_probe(struct platform_device *pdev) int irq; int err; int i; - u32 nr_channels; - dma_cap_mask_t cap_mask = {}; - enum atdma_devtype atdmatype; + struct at_dma_platform_data *plat_dat; - dma_cap_set(DMA_MEMCPY, cap_mask); + /* setup platform data for each SoC */ + dma_cap_set(DMA_MEMCPY, at91sam9rl_config.cap_mask); + dma_cap_set(DMA_MEMCPY, at91sam9g45_config.cap_mask); + dma_cap_set(DMA_SLAVE, at91sam9g45_config.cap_mask); /* get DMA parameters from controller type */ - atdmatype = at_dma_get_driver_data(pdev); - - switch (atdmatype) { - case ATDMA_DEVTYPE_SAM9RL: - nr_channels = 2; - break; - case ATDMA_DEVTYPE_SAM9G45: - nr_channels = 8; - dma_cap_set(DMA_SLAVE, cap_mask); - break; - default: - return -EINVAL; - } + plat_dat = at_dma_get_driver_data(pdev); + if (!plat_dat) + return -ENODEV; io = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (!io) @@ -1272,15 +1272,14 @@ static int __init at_dma_probe(struct platform_device *pdev) return irq; size = sizeof(struct at_dma); - size += nr_channels * sizeof(struct at_dma_chan); + size += plat_dat->nr_channels * sizeof(struct at_dma_chan); atdma = kzalloc(size, GFP_KERNEL); if (!atdma) return -ENOMEM; /* discover transaction capabilities */ - atdma->dma_common.cap_mask = cap_mask; - atdma->all_chan_mask = (1 << nr_channels) - 1; - atdma->devtype = atdmatype; + atdma->dma_common.cap_mask = plat_dat->cap_mask; + atdma->all_chan_mask = (1 << plat_dat->nr_channels) - 1; size = resource_size(io); if (!request_mem_region(io->start, size, pdev->dev.driver->name)) { @@ -1326,7 +1325,7 @@ static int __init at_dma_probe(struct platform_device *pdev) /* initialize channels related values */ INIT_LIST_HEAD(&atdma->dma_common.channels); - for (i = 0; i < nr_channels; i++) { + for (i = 0; i < plat_dat->nr_channels; i++) { struct at_dma_chan *atchan = &atdma->chan[i]; atchan->chan_common.device = &atdma->dma_common; @@ -1371,7 +1370,7 @@ static int __init at_dma_probe(struct platform_device *pdev) dev_info(&pdev->dev, "Atmel AHB DMA Controller ( %s%s), %d channels\n", dma_has_cap(DMA_MEMCPY, atdma->dma_common.cap_mask) ? "cpy " : "", dma_has_cap(DMA_SLAVE, atdma->dma_common.cap_mask) ? "slave " : "", - nr_channels); + plat_dat->nr_channels); dma_async_device_register(&atdma->dma_common); diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h index d7d67372e4ff..dcaedfc181cf 100644 --- a/drivers/dma/at_hdmac_regs.h +++ b/drivers/dma/at_hdmac_regs.h @@ -248,12 +248,6 @@ static inline struct at_dma_chan *to_at_dma_chan(struct dma_chan *dchan) /*-- Controller ------------------------------------------------------*/ -enum atdma_devtype { - ATDMA_DEVTYPE_UNDEFINED = 0, - ATDMA_DEVTYPE_SAM9RL, /* compatible with SAM9RL DMA controller */ - ATDMA_DEVTYPE_SAM9G45, /* compatible with SAM9G45 DMA controller */ -}; - /** * struct at_dma - internal representation of an Atmel HDMA Controller * @chan_common: common dmaengine dma_device object members @@ -267,7 +261,6 @@ enum atdma_devtype { */ struct at_dma { struct dma_device dma_common; - enum atdma_devtype devtype; void __iomem *regs; struct clk *clk; u32 save_imr; From b62cfc5e0e59d1e0192105144b724b0c332720a0 Mon Sep 17 00:00:00 2001 From: Rafal Prylowski Date: Sat, 26 Nov 2011 12:54:07 +0200 Subject: [PATCH 129/820] dma/ep93xx_dma: fix initialization of M2M control register Setting the flags in case of IDE didn't have any effect since the control register is overwritten few lines below. So move these to be after the control register is initialized. Signed-off-by: Rafal Prylowski Signed-off-by: Mika Westerberg Acked-by: H Hartley Sweeten Signed-off-by: Vinod Koul --- drivers/dma/ep93xx_dma.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c index 009851b2aeea..526a7424f6a9 100644 --- a/drivers/dma/ep93xx_dma.c +++ b/drivers/dma/ep93xx_dma.c @@ -459,10 +459,6 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac) * This IDE part is totally untested. Values below are taken * from the EP93xx Users's Guide and might not be correct. */ - control |= M2M_CONTROL_NO_HDSK; - control |= M2M_CONTROL_RSS_IDE; - control |= M2M_CONTROL_PW_16; - if (data->direction == DMA_MEM_TO_DEV) { /* Worst case from the UG */ control = (3 << M2M_CONTROL_PWSC_SHIFT); @@ -473,6 +469,10 @@ static int m2m_hw_setup(struct ep93xx_dma_chan *edmac) control |= M2M_CONTROL_SAH; control |= M2M_CONTROL_TM_RX; } + + control |= M2M_CONTROL_NO_HDSK; + control |= M2M_CONTROL_RSS_IDE; + control |= M2M_CONTROL_PW_16; break; default: From 6d0709d2000ae7dbead759715f57ba381b7057bb Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Sat, 26 Nov 2011 12:54:08 +0200 Subject: [PATCH 130/820] dma/ep93xx_dma: prevent ep93xx_dma_tasklet() to reference an empty list If dma_terminate_all() is called before the ep93xx_dma_tasklet() gets to run, it tries to access an empty ->active list which results following OOPS: Internal error: Oops - undefined instruction: 0 [#1] CPU: 0 Not tainted (3.2.0-rc1EP-1+ #1008) PC is at 0xc184c868 LR is at ep93xx_dma_tasklet+0xec/0x164 pc : [] lr : [] psr: 00000013 sp : c02b7e70 ip : ffffffff fp : c02b7ea4 r10: 00000100 r9 : 80000013 r8 : c02b7e50 r7 : c02b7e70 r6 : c02b7ea4 r5 : 000000a4 r4 : c02b7e70 r3 : c02b751d r2 : 8ae34598 r1 : c184c6e0 r0 : c02b7ea4 Flags: nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel Control: c000717f Table: c0004000 DAC: 00000017 Process swapper (pid: 0, stack limit = 0xc02b6270) Stack: (0xc02b7e70 to 0xc02b8000) 7e60: c02b7ea4 c02b7e70 c0008b64 c02bd5c4 7e80: c02d60e0 00000000 00000000 c02bd44c c02d60e0 00000100 c02b7ec4 c02b7ea8 7ea0: c001c49c c012b44c 00000018 00000001 c02d60e0 c02b6000 c02b7f04 c02b7ec8 7ec0: c001cbc0 c001c3e4 c02b7eec c02b7ed8 00000006 0000000a c02bf674 c02c458c 7ee0: 00000011 00000000 c02b7f7c c0004000 41129200 c02b0c80 c02b7f14 c02b7f08 7f00: c001cdd0 c001cb38 c02b7f34 c02b7f18 c000983c c001cd98 c0009a60 60000013 7f20: fefb0001 c02b7f7c c02b7f44 c02b7f38 c0008190 c0009810 c02b7f9c c02b7f48 7f40: c0008b64 c0008190 c02c2bf8 00000002 c02b7f90 60000013 c02b6000 c02d1504 7f60: c02baa88 c02baa80 c0004000 41129200 c02b0c80 c02b7f9c c02b7fa0 c02b7f90 7f80: c0009a54 c0009a60 60000013 ffffffff c02b7fbc c02b7fa0 c000a03c c0009a40 7fa0: c02b80b0 c02b19dc c02b19d8 c02baa80 c02b7fcc c02b7fc0 c02384e4 c0009fd4 7fc0: c02b7ff4 c02b7fd0 c029d924 c0238494 c029d49c 00000000 00000000 c02b19dc 7fe0: c0007175 c02b803c 00000000 c02b7ff8 c000803c c029d700 00000000 00000000 Backtrace: [] (ep93xx_dma_tasklet+0x0/0x164) from [] (tasklet_action+0xc8/0xdc) [] (tasklet_action+0x0/0xdc) from [] (__do_softirq+0x98/0x154) r7:c02b6000 r6:c02d60e0 r5:00000001 r4:00000018 [] (__do_softirq+0x0/0x154) from [] (irq_exit+0x48/0x50) [] (irq_exit+0x0/0x50) from [] (handle_IRQ+0x3c/0x8c) [] (handle_IRQ+0x0/0x8c) from [] (asm_do_IRQ+0x10/0x14) r7:c02b7f7c r6:fefb0001 r5:60000013 r4:c0009a60 [] (asm_do_IRQ+0x0/0x14) from [] (__irq_svc+0x24/0xc0) Exception stack(0xc02b7f48 to 0xc02b7f90) 7f40: c02c2bf8 00000002 c02b7f90 60000013 c02b6000 c02d1504 7f60: c02baa88 c02baa80 c0004000 41129200 c02b0c80 c02b7f9c c02b7fa0 c02b7f90 7f80: c0009a54 c0009a60 60000013 ffffffff [] (default_idle+0x0/0x34) from [] (cpu_idle+0x78/0xb0) [] (cpu_idle+0x0/0xb0) from [] (rest_init+0x60/0x78) r7:c02baa80 r6:c02b19d8 r5:c02b19dc r4:c02b80b0 [] (rest_init+0x0/0x78) from [] (start_kernel+0x234/0x278) [] (start_kernel+0x0/0x278) from [] (0xc000803c) r5:c02b803c r4:c0007175 Code: 42555300 54535953 643d4d45 65766972 (53007372) To make the code a bit more robust against things like these, we modify ep93xx_dma_get_active() to return NULL in case of empty ->active list and make sure that callers handle this correctly. Reported-by: Rafal Prylowski Signed-off-by: Mika Westerberg Acked-by: H Hartley Sweeten Signed-off-by: Vinod Koul --- drivers/dma/ep93xx_dma.c | 60 ++++++++++++++++++++++++++++++++-------- 1 file changed, 48 insertions(+), 12 deletions(-) diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c index 526a7424f6a9..59e7a965772b 100644 --- a/drivers/dma/ep93xx_dma.c +++ b/drivers/dma/ep93xx_dma.c @@ -246,6 +246,9 @@ static void ep93xx_dma_set_active(struct ep93xx_dma_chan *edmac, static struct ep93xx_dma_desc * ep93xx_dma_get_active(struct ep93xx_dma_chan *edmac) { + if (list_empty(&edmac->active)) + return NULL; + return list_first_entry(&edmac->active, struct ep93xx_dma_desc, node); } @@ -263,16 +266,22 @@ ep93xx_dma_get_active(struct ep93xx_dma_chan *edmac) */ static bool ep93xx_dma_advance_active(struct ep93xx_dma_chan *edmac) { + struct ep93xx_dma_desc *desc; + list_rotate_left(&edmac->active); if (test_bit(EP93XX_DMA_IS_CYCLIC, &edmac->flags)) return true; + desc = ep93xx_dma_get_active(edmac); + if (!desc) + return false; + /* * If txd.cookie is set it means that we are back in the first * descriptor in the chain and hence done with it. */ - return !ep93xx_dma_get_active(edmac)->txd.cookie; + return !desc->txd.cookie; } /* @@ -327,9 +336,15 @@ static void m2p_hw_shutdown(struct ep93xx_dma_chan *edmac) static void m2p_fill_desc(struct ep93xx_dma_chan *edmac) { - struct ep93xx_dma_desc *desc = ep93xx_dma_get_active(edmac); + struct ep93xx_dma_desc *desc; u32 bus_addr; + desc = ep93xx_dma_get_active(edmac); + if (!desc) { + dev_warn(chan2dev(edmac), "M2P: empty descriptor list\n"); + return; + } + if (ep93xx_dma_chan_direction(&edmac->chan) == DMA_MEM_TO_DEV) bus_addr = desc->src_addr; else @@ -491,7 +506,13 @@ static void m2m_hw_shutdown(struct ep93xx_dma_chan *edmac) static void m2m_fill_desc(struct ep93xx_dma_chan *edmac) { - struct ep93xx_dma_desc *desc = ep93xx_dma_get_active(edmac); + struct ep93xx_dma_desc *desc; + + desc = ep93xx_dma_get_active(edmac); + if (!desc) { + dev_warn(chan2dev(edmac), "M2M: empty descriptor list\n"); + return; + } if (edmac->buffer == 0) { writel(desc->src_addr, edmac->regs + M2M_SAR_BASE0); @@ -669,24 +690,30 @@ static void ep93xx_dma_tasklet(unsigned long data) { struct ep93xx_dma_chan *edmac = (struct ep93xx_dma_chan *)data; struct ep93xx_dma_desc *desc, *d; - dma_async_tx_callback callback; - void *callback_param; + dma_async_tx_callback callback = NULL; + void *callback_param = NULL; LIST_HEAD(list); spin_lock_irq(&edmac->lock); + /* + * If dma_terminate_all() was called before we get to run, the active + * list has become empty. If that happens we aren't supposed to do + * anything more than call ep93xx_dma_advance_work(). + */ desc = ep93xx_dma_get_active(edmac); - if (desc->complete) { - edmac->last_completed = desc->txd.cookie; - list_splice_init(&edmac->active, &list); + if (desc) { + if (desc->complete) { + edmac->last_completed = desc->txd.cookie; + list_splice_init(&edmac->active, &list); + } + callback = desc->txd.callback; + callback_param = desc->txd.callback_param; } spin_unlock_irq(&edmac->lock); /* Pick up the next descriptor from the queue */ ep93xx_dma_advance_work(edmac); - callback = desc->txd.callback; - callback_param = desc->txd.callback_param; - /* Now we can release all the chained descriptors */ list_for_each_entry_safe(desc, d, &list, node) { /* @@ -706,13 +733,22 @@ static void ep93xx_dma_tasklet(unsigned long data) static irqreturn_t ep93xx_dma_interrupt(int irq, void *dev_id) { struct ep93xx_dma_chan *edmac = dev_id; + struct ep93xx_dma_desc *desc; irqreturn_t ret = IRQ_HANDLED; spin_lock(&edmac->lock); + desc = ep93xx_dma_get_active(edmac); + if (!desc) { + dev_warn(chan2dev(edmac), + "got interrupt while active list is empty\n"); + spin_unlock(&edmac->lock); + return IRQ_NONE; + } + switch (edmac->edma->hw_interrupt(edmac)) { case INTERRUPT_DONE: - ep93xx_dma_get_active(edmac)->complete = true; + desc->complete = true; tasklet_schedule(&edmac->tasklet); break; From ab59a510c6ad6b3add5125df64843be754782de6 Mon Sep 17 00:00:00 2001 From: Huang Shijie Date: Fri, 2 Dec 2011 10:16:25 +0800 Subject: [PATCH 131/820] IMX/SDMA : save the real count for one DMA transaction. When we use the SDMA in the UART driver(such as imx6q), we will meet one situation: Assume we set 64 bytes for the RX DMA buffer. The RX DMA buffer has received some data, but not full. An Aging DMA request will be received by the SDMA controller if we enable the IDDMAEN(UCR4[6]) in this case. So the UART driver needs to know the count of the real received bytes, and push them to upper layer. Add two new fields to sdmac, and update the `residue` in sdma_tx_status(). Signed-off-by: Huang Shijie Signed-off-by: Vinod Koul --- drivers/dma/imx-sdma.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c index d9e5933d91eb..f59fd8fffa88 100644 --- a/drivers/dma/imx-sdma.c +++ b/drivers/dma/imx-sdma.c @@ -268,6 +268,8 @@ struct sdma_channel { struct dma_async_tx_descriptor desc; dma_cookie_t last_completed; enum dma_status status; + unsigned int chn_count; + unsigned int chn_real_count; }; #define IMX_DMA_SG_LOOP (1 << 0) @@ -503,6 +505,7 @@ static void mxc_sdma_handle_channel_normal(struct sdma_channel *sdmac) struct sdma_buffer_descriptor *bd; int i, error = 0; + sdmac->chn_real_count = 0; /* * non loop mode. Iterate over all descriptors, collect * errors and call callback function @@ -512,6 +515,7 @@ static void mxc_sdma_handle_channel_normal(struct sdma_channel *sdmac) if (bd->mode.status & (BD_DONE | BD_RROR)) error = -EIO; + sdmac->chn_real_count += bd->mode.count; } if (error) @@ -519,9 +523,9 @@ static void mxc_sdma_handle_channel_normal(struct sdma_channel *sdmac) else sdmac->status = DMA_SUCCESS; + sdmac->last_completed = sdmac->desc.cookie; if (sdmac->desc.callback) sdmac->desc.callback(sdmac->desc.callback_param); - sdmac->last_completed = sdmac->desc.cookie; } static void mxc_sdma_handle_channel(struct sdma_channel *sdmac) @@ -941,6 +945,7 @@ static struct dma_async_tx_descriptor *sdma_prep_slave_sg( goto err_out; } + sdmac->chn_count = 0; for_each_sg(sgl, sg, sg_len, i) { struct sdma_buffer_descriptor *bd = &sdmac->bd[i]; int param; @@ -957,6 +962,7 @@ static struct dma_async_tx_descriptor *sdma_prep_slave_sg( } bd->mode.count = count; + sdmac->chn_count += count; if (sdmac->word_size > DMA_SLAVE_BUSWIDTH_4_BYTES) { ret = -EINVAL; @@ -1120,7 +1126,8 @@ static enum dma_status sdma_tx_status(struct dma_chan *chan, last_used = chan->cookie; - dma_set_tx_state(txstate, sdmac->last_completed, last_used, 0); + dma_set_tx_state(txstate, sdmac->last_completed, last_used, + sdmac->chn_count - sdmac->chn_real_count); return sdmac->status; } From c94e910535be72f0c6ac0c69e6acd8d44414e80d Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 26 Nov 2011 15:11:12 +0800 Subject: [PATCH 132/820] dmaengine: convert drivers/dma/* to use module_platform_driver() This patch converts the drivers in drivers/dma/* to use the module_platform_driver() macro which makes the code smaller and a bit simpler. Cc: Dan Williams Cc: Piotr Ziecik Cc: Rongjun Ying Cc: Barry Song <21cnbao@gmail.com> Cc: Pelagicore AB Signed-off-by: Axel Lin Signed-off-by: Vinod Koul --- drivers/dma/iop-adma.c | 16 ++-------------- drivers/dma/mpc512x_dma.c | 12 +----------- drivers/dma/sirf-dma.c | 12 +----------- drivers/dma/timb_dma.c | 12 +----------- 4 files changed, 5 insertions(+), 47 deletions(-) diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c index 686dc7673b19..04be90b645b8 100644 --- a/drivers/dma/iop-adma.c +++ b/drivers/dma/iop-adma.c @@ -1735,8 +1735,6 @@ static void iop_chan_start_null_xor(struct iop_adma_chan *iop_chan) spin_unlock_bh(&iop_chan->lock); } -MODULE_ALIAS("platform:iop-adma"); - static struct platform_driver iop_adma_driver = { .probe = iop_adma_probe, .remove = __devexit_p(iop_adma_remove), @@ -1746,19 +1744,9 @@ static struct platform_driver iop_adma_driver = { }, }; -static int __init iop_adma_init(void) -{ - return platform_driver_register(&iop_adma_driver); -} - -static void __exit iop_adma_exit(void) -{ - platform_driver_unregister(&iop_adma_driver); - return; -} -module_exit(iop_adma_exit); -module_init(iop_adma_init); +module_platform_driver(iop_adma_driver); MODULE_AUTHOR("Intel Corporation"); MODULE_DESCRIPTION("IOP ADMA Engine Driver"); MODULE_LICENSE("GPL"); +MODULE_ALIAS("platform:iop-adma"); diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c index 8ba4edc6185e..4d6d4cf66949 100644 --- a/drivers/dma/mpc512x_dma.c +++ b/drivers/dma/mpc512x_dma.c @@ -835,17 +835,7 @@ static struct platform_driver mpc_dma_driver = { }, }; -static int __init mpc_dma_init(void) -{ - return platform_driver_register(&mpc_dma_driver); -} -module_init(mpc_dma_init); - -static void __exit mpc_dma_exit(void) -{ - platform_driver_unregister(&mpc_dma_driver); -} -module_exit(mpc_dma_exit); +module_platform_driver(mpc_dma_driver); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Piotr Ziecik "); diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c index 55ec67997670..2333810d1688 100644 --- a/drivers/dma/sirf-dma.c +++ b/drivers/dma/sirf-dma.c @@ -699,17 +699,7 @@ static struct platform_driver sirfsoc_dma_driver = { }, }; -static int __init sirfsoc_dma_init(void) -{ - return platform_driver_register(&sirfsoc_dma_driver); -} -module_init(sirfsoc_dma_init); - -static void __exit sirfsoc_dma_exit(void) -{ - platform_driver_unregister(&sirfsoc_dma_driver); -} -module_exit(sirfsoc_dma_exit); +module_platform_driver(sirfsoc_dma_driver); MODULE_AUTHOR("Rongjun Ying , " "Barry Song "); diff --git a/drivers/dma/timb_dma.c b/drivers/dma/timb_dma.c index d142ae1fce52..a6f9c1684a0f 100644 --- a/drivers/dma/timb_dma.c +++ b/drivers/dma/timb_dma.c @@ -841,17 +841,7 @@ static struct platform_driver td_driver = { .remove = __exit_p(td_remove), }; -static int __init td_init(void) -{ - return platform_driver_register(&td_driver); -} -module_init(td_init); - -static void __exit td_exit(void) -{ - platform_driver_unregister(&td_driver); -} -module_exit(td_exit); +module_platform_driver(td_driver); MODULE_LICENSE("GPL v2"); MODULE_DESCRIPTION("Timberdale DMA controller driver"); From 8f0d30f9ee333f4fb3458a9a33f7ead5269ea8f3 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 29 Nov 2011 12:56:50 +0530 Subject: [PATCH 133/820] dmaengine/amba-pl08x: Config ccfg and increment phychan_hold if phychan is true Currently, if plchan->phychan is true, we return immediately from prep_phy_chan(). We must configure txd->ccfg and increment phychan_hold before returning. Otherwise, request line number wouldn't be configured in this txd. Reported-by: Rajeev Kumar Signed-off-by: Viresh Kumar Acked-by: Linus Walleij Signed-off-by: Vinod Koul --- drivers/dma/amba-pl08x.c | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-) diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c index 41c62fd0680d..9ebceca17530 100644 --- a/drivers/dma/amba-pl08x.c +++ b/drivers/dma/amba-pl08x.c @@ -854,8 +854,10 @@ static int prep_phy_channel(struct pl08x_dma_chan *plchan, int ret; /* Check if we already have a channel */ - if (plchan->phychan) - return 0; + if (plchan->phychan) { + ch = plchan->phychan; + goto got_channel; + } ch = pl08x_get_phy_channel(pl08x, plchan); if (!ch) { @@ -880,21 +882,22 @@ static int prep_phy_channel(struct pl08x_dma_chan *plchan, return -EBUSY; } ch->signal = ret; - - /* Assign the flow control signal to this channel */ - if (txd->direction == DMA_MEM_TO_DEV) - txd->ccfg |= ch->signal << PL080_CONFIG_DST_SEL_SHIFT; - else if (txd->direction == DMA_DEV_TO_MEM) - txd->ccfg |= ch->signal << PL080_CONFIG_SRC_SEL_SHIFT; } + plchan->phychan = ch; dev_dbg(&pl08x->adev->dev, "allocated physical channel %d and signal %d for xfer on %s\n", ch->id, ch->signal, plchan->name); +got_channel: + /* Assign the flow control signal to this channel */ + if (txd->direction == DMA_MEM_TO_DEV) + txd->ccfg |= ch->signal << PL080_CONFIG_DST_SEL_SHIFT; + else if (txd->direction == DMA_DEV_TO_MEM) + txd->ccfg |= ch->signal << PL080_CONFIG_SRC_SEL_SHIFT; + plchan->phychan_hold++; - plchan->phychan = ch; return 0; } From 5cd326fd27da347925019fcc041b79bad8dd55ed Mon Sep 17 00:00:00 2001 From: Narayanan G Date: Wed, 30 Nov 2011 19:20:42 +0530 Subject: [PATCH 134/820] dmaengine/ste_dma40: allow fixed physical channel Allow logical channels to specify the physical channel they want to use. This is needed to avoid two peripherals operating on the same physical channel during some special use-cases. (like mmc and usb during a usb mass storage case). Signed-off-by: Rabin Vincent Signed-off-by: Narayanan G Reviewed-by: Linus Walleij Signed-off-by: Vinod Koul --- .../arm/plat-nomadik/include/plat/ste_dma40.h | 6 ++- drivers/dma/ste_dma40.c | 51 +++++++++++++++---- 2 files changed, 47 insertions(+), 10 deletions(-) diff --git a/arch/arm/plat-nomadik/include/plat/ste_dma40.h b/arch/arm/plat-nomadik/include/plat/ste_dma40.h index 3177bed253df..fd0ee84c45d1 100644 --- a/arch/arm/plat-nomadik/include/plat/ste_dma40.h +++ b/arch/arm/plat-nomadik/include/plat/ste_dma40.h @@ -113,7 +113,8 @@ struct stedma40_half_channel_info { * @dst_dev_type: Dst device type * @src_info: Parameters for dst half channel * @dst_info: Parameters for dst half channel - * + * @use_fixed_channel: if true, use physical channel specified by phy_channel + * @phy_channel: physical channel to use, only if use_fixed_channel is true * * This structure has to be filled by the client drivers. * It is recommended to do all dma configurations for clients in the machine. @@ -129,6 +130,9 @@ struct stedma40_chan_cfg { int dst_dev_type; struct stedma40_half_channel_info src_info; struct stedma40_half_channel_info dst_info; + + bool use_fixed_channel; + int phy_channel; }; /** diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c index aff128a39a4b..972dc35770f9 100644 --- a/drivers/dma/ste_dma40.c +++ b/drivers/dma/ste_dma40.c @@ -1545,11 +1545,16 @@ static int d40_validate_conf(struct d40_chan *d40c, return res; } -static bool d40_alloc_mask_set(struct d40_phy_res *phy, bool is_src, - int log_event_line, bool is_log) +static bool d40_alloc_mask_set(struct d40_phy_res *phy, + bool is_src, int log_event_line, bool is_log, + bool *first_user) { unsigned long flags; spin_lock_irqsave(&phy->lock, flags); + + *first_user = ((phy->allocated_src | phy->allocated_dst) + == D40_ALLOC_FREE); + if (!is_log) { /* Physical interrupts are masked per physical full channel */ if (phy->allocated_src == D40_ALLOC_FREE && @@ -1630,7 +1635,7 @@ out: return is_free; } -static int d40_allocate_channel(struct d40_chan *d40c) +static int d40_allocate_channel(struct d40_chan *d40c, bool *first_phy_user) { int dev_type; int event_group; @@ -1666,7 +1671,8 @@ static int d40_allocate_channel(struct d40_chan *d40c) for (i = 0; i < d40c->base->num_phy_chans; i++) { if (d40_alloc_mask_set(&phys[i], is_src, - 0, is_log)) + 0, is_log, + first_phy_user)) goto found_phy; } } else @@ -1676,7 +1682,8 @@ static int d40_allocate_channel(struct d40_chan *d40c) if (d40_alloc_mask_set(&phys[i], is_src, 0, - is_log)) + is_log, + first_phy_user)) goto found_phy; } } @@ -1692,6 +1699,25 @@ found_phy: /* Find logical channel */ for (j = 0; j < d40c->base->num_phy_chans; j += 8) { int phy_num = j + event_group * 2; + + if (d40c->dma_cfg.use_fixed_channel) { + i = d40c->dma_cfg.phy_channel; + + if ((i != phy_num) && (i != phy_num + 1)) { + dev_err(chan2dev(d40c), + "invalid fixed phy channel %d\n", i); + return -EINVAL; + } + + if (d40_alloc_mask_set(&phys[i], is_src, event_line, + is_log, first_phy_user)) + goto found_log; + + dev_err(chan2dev(d40c), + "could not allocate fixed phy channel %d\n", i); + return -EINVAL; + } + /* * Spread logical channels across all available physical rather * than pack every logical channel at the first available phy @@ -1700,13 +1726,15 @@ found_phy: if (is_src) { for (i = phy_num; i < phy_num + 2; i++) { if (d40_alloc_mask_set(&phys[i], is_src, - event_line, is_log)) + event_line, is_log, + first_phy_user)) goto found_log; } } else { for (i = phy_num + 1; i >= phy_num; i--) { if (d40_alloc_mask_set(&phys[i], is_src, - event_line, is_log)) + event_line, is_log, + first_phy_user)) goto found_log; } } @@ -2160,9 +2188,8 @@ static int d40_alloc_chan_resources(struct dma_chan *chan) goto fail; } } - is_free_phy = (d40c->phy_chan == NULL); - err = d40_allocate_channel(d40c); + err = d40_allocate_channel(d40c, &is_free_phy); if (err) { chan_err(d40c, "Failed to allocate channel\n"); d40c->configured = false; @@ -2189,6 +2216,12 @@ static int d40_alloc_chan_resources(struct dma_chan *chan) D40_LCPA_CHAN_SIZE + D40_LCPA_CHAN_DST_DELTA; } + dev_dbg(chan2dev(d40c), "allocated %s channel (phy %d%s)\n", + chan_is_logical(d40c) ? "logical" : "physical", + d40c->phy_chan->num, + d40c->dma_cfg.use_fixed_channel ? ", fixed" : ""); + + /* * Only write channel configuration to the DMA if the physical * resource is free. In case of multiple logical channels From c99e78435342a65a6a0bf9b86f87fd05831858d2 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 1 Dec 2011 14:58:51 +0100 Subject: [PATCH 135/820] i.MX IPU DMA: Fix wrong burstsize settings The burstsize (npb in struct chan_param_mem) is set in ipu_ch_param_set_size() once. The number of allowed pixels in a burst depend on the pixel format and the rotation mode. For 16bit formats 16 pixels are allowed whereas for 32bit formats only 8 pixels are allowed. Set these values correctly in ipu_ch_param_set_size() and do not overwrite them afterwards. We do not support rotation right now, so ignore this case. This patch fixes the wrong burstsize setting of 16 pixels for 32bpp. Signed-off-by: Sascha Hauer Signed-off-by: Vinod Koul --- drivers/dma/ipu/ipu_idmac.c | 25 +------------------------ 1 file changed, 1 insertion(+), 24 deletions(-) diff --git a/drivers/dma/ipu/ipu_idmac.c b/drivers/dma/ipu/ipu_idmac.c index e16e280b26eb..6212b16e8cf2 100644 --- a/drivers/dma/ipu/ipu_idmac.c +++ b/drivers/dma/ipu/ipu_idmac.c @@ -312,7 +312,7 @@ static void ipu_ch_param_set_size(union chan_param_mem *params, case IPU_PIX_FMT_RGB565: params->ip.bpp = 2; params->ip.pfs = 4; - params->ip.npb = 7; + params->ip.npb = 15; params->ip.sat = 2; /* SAT = 32-bit access */ params->ip.ofs0 = 0; /* Red bit offset */ params->ip.ofs1 = 5; /* Green bit offset */ @@ -422,12 +422,6 @@ static void ipu_ch_param_set_size(union chan_param_mem *params, params->pp.nsb = 1; } -static void ipu_ch_param_set_burst_size(union chan_param_mem *params, - uint16_t burst_pixels) -{ - params->pp.npb = burst_pixels - 1; -} - static void ipu_ch_param_set_buffer(union chan_param_mem *params, dma_addr_t buf0, dma_addr_t buf1) { @@ -690,23 +684,6 @@ static int ipu_init_channel_buffer(struct idmac_channel *ichan, ipu_ch_param_set_size(¶ms, pixel_fmt, width, height, stride_bytes); ipu_ch_param_set_buffer(¶ms, phyaddr_0, phyaddr_1); ipu_ch_param_set_rotation(¶ms, rot_mode); - /* Some channels (rotation) have restriction on burst length */ - switch (channel) { - case IDMAC_IC_7: /* Hangs with burst 8, 16, other values - invalid - Table 44-30 */ -/* - ipu_ch_param_set_burst_size(¶ms, 8); - */ - break; - case IDMAC_SDC_0: - case IDMAC_SDC_1: - /* In original code only IPU_PIX_FMT_RGB565 was setting burst */ - ipu_ch_param_set_burst_size(¶ms, 16); - break; - case IDMAC_IC_0: - default: - break; - } spin_lock_irqsave(&ipu->lock, flags); From f910fb8fcd1c97788f2291c8646597bcd87ee061 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 1 Dec 2011 14:58:52 +0100 Subject: [PATCH 136/820] video i.MX IPU: Fix display connections The IPU internally works on 32bit colors. It can arbitrarily map between pixel formats and internal representation and also between internal representation and the physical connection to the display. The driver used to change the mapping between internal representation and display connection depending on the user selected bpp which is wrong. Instead, the mapping is specified by the hardware, so an additional field in platform data is added to describe the connection between i.MX and the display. The default for this field is RGB666 which seems to be the only configuration which works without this patch, so I assumed that all in Kernel boards are connected this way. This patch has been tested on a RGB666 connected display and a RGB888 connected display in both 16bpp and 32bpp modes. Signed-off-by: Sascha Hauer Signed-off-by: Vinod Koul --- arch/arm/plat-mxc/include/mach/mx3fb.h | 15 +++++++ drivers/video/mx3fb.c | 61 ++++++++++---------------- 2 files changed, 37 insertions(+), 39 deletions(-) diff --git a/arch/arm/plat-mxc/include/mach/mx3fb.h b/arch/arm/plat-mxc/include/mach/mx3fb.h index ac24c5c4bc83..fdbe60001542 100644 --- a/arch/arm/plat-mxc/include/mach/mx3fb.h +++ b/arch/arm/plat-mxc/include/mach/mx3fb.h @@ -22,6 +22,20 @@ #define FB_SYNC_SWAP_RGB 0x04000000 #define FB_SYNC_CLK_SEL_EN 0x02000000 +/* + * Specify the way your display is connected. The IPU can arbitrarily + * map the internal colors to the external data lines. We only support + * the following mappings at the moment. + */ +enum disp_data_mapping { + /* blue -> d[0..5], green -> d[6..11], red -> d[12..17] */ + IPU_DISP_DATA_MAPPING_RGB666, + /* blue -> d[0..4], green -> d[5..10], red -> d[11..15] */ + IPU_DISP_DATA_MAPPING_RGB565, + /* blue -> d[0..7], green -> d[8..15], red -> d[16..23] */ + IPU_DISP_DATA_MAPPING_RGB888, +}; + /** * struct mx3fb_platform_data - mx3fb platform data * @@ -33,6 +47,7 @@ struct mx3fb_platform_data { const char *name; const struct fb_videomode *mode; int num_modes; + enum disp_data_mapping disp_data_fmt; }; #endif diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c index 1981e3665d07..727a5149d818 100644 --- a/drivers/video/mx3fb.c +++ b/drivers/video/mx3fb.c @@ -245,6 +245,7 @@ struct mx3fb_data { uint32_t h_start_width; uint32_t v_start_width; + enum disp_data_mapping disp_data_fmt; }; struct dma_chan_request { @@ -287,11 +288,14 @@ static void mx3fb_write_reg(struct mx3fb_data *mx3fb, u32 value, unsigned long r __raw_writel(value, mx3fb->reg_base + reg); } -static const uint32_t di_mappings[] = { - 0x1600AAAA, 0x00E05555, 0x00070000, 3, /* RGB888 */ - 0x0005000F, 0x000B000F, 0x0011000F, 1, /* RGB666 */ - 0x0011000F, 0x000B000F, 0x0005000F, 1, /* BGR666 */ - 0x0004003F, 0x000A000F, 0x000F003F, 1 /* RGB565 */ +struct di_mapping { + uint32_t b0, b1, b2; +}; + +static const struct di_mapping di_mappings[] = { + [IPU_DISP_DATA_MAPPING_RGB666] = { 0x0005000f, 0x000b000f, 0x0011000f }, + [IPU_DISP_DATA_MAPPING_RGB565] = { 0x0004003f, 0x000a000f, 0x000f003f }, + [IPU_DISP_DATA_MAPPING_RGB888] = { 0x00070000, 0x000f0000, 0x00170000 }, }; static void sdc_fb_init(struct mx3fb_info *fbi) @@ -425,7 +429,6 @@ static int sdc_set_window_pos(struct mx3fb_data *mx3fb, enum ipu_channel channel * @pixel_clk: desired pixel clock frequency in Hz. * @width: width of panel in pixels. * @height: height of panel in pixels. - * @pixel_fmt: pixel format of buffer as FOURCC ASCII code. * @h_start_width: number of pixel clocks between the HSYNC signal pulse * and the start of valid data. * @h_sync_width: width of the HSYNC signal in units of pixel clocks. @@ -442,7 +445,6 @@ static int sdc_set_window_pos(struct mx3fb_data *mx3fb, enum ipu_channel channel static int sdc_init_panel(struct mx3fb_data *mx3fb, enum ipu_panel panel, uint32_t pixel_clk, uint16_t width, uint16_t height, - enum pixel_fmt pixel_fmt, uint16_t h_start_width, uint16_t h_sync_width, uint16_t h_end_width, uint16_t v_start_width, uint16_t v_sync_width, uint16_t v_end_width, @@ -453,6 +455,7 @@ static int sdc_init_panel(struct mx3fb_data *mx3fb, enum ipu_panel panel, uint32_t old_conf; uint32_t div; struct clk *ipu_clk; + const struct di_mapping *map; dev_dbg(mx3fb->dev, "panel size = %d x %d", width, height); @@ -540,36 +543,10 @@ static int sdc_init_panel(struct mx3fb_data *mx3fb, enum ipu_panel panel, sig.Vsync_pol << DI_D3_VSYNC_POL_SHIFT; mx3fb_write_reg(mx3fb, old_conf, DI_DISP_SIG_POL); - switch (pixel_fmt) { - case IPU_PIX_FMT_RGB24: - mx3fb_write_reg(mx3fb, di_mappings[0], DI_DISP3_B0_MAP); - mx3fb_write_reg(mx3fb, di_mappings[1], DI_DISP3_B1_MAP); - mx3fb_write_reg(mx3fb, di_mappings[2], DI_DISP3_B2_MAP); - mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) | - ((di_mappings[3] - 1) << 12), DI_DISP_ACC_CC); - break; - case IPU_PIX_FMT_RGB666: - mx3fb_write_reg(mx3fb, di_mappings[4], DI_DISP3_B0_MAP); - mx3fb_write_reg(mx3fb, di_mappings[5], DI_DISP3_B1_MAP); - mx3fb_write_reg(mx3fb, di_mappings[6], DI_DISP3_B2_MAP); - mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) | - ((di_mappings[7] - 1) << 12), DI_DISP_ACC_CC); - break; - case IPU_PIX_FMT_BGR666: - mx3fb_write_reg(mx3fb, di_mappings[8], DI_DISP3_B0_MAP); - mx3fb_write_reg(mx3fb, di_mappings[9], DI_DISP3_B1_MAP); - mx3fb_write_reg(mx3fb, di_mappings[10], DI_DISP3_B2_MAP); - mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) | - ((di_mappings[11] - 1) << 12), DI_DISP_ACC_CC); - break; - default: - mx3fb_write_reg(mx3fb, di_mappings[12], DI_DISP3_B0_MAP); - mx3fb_write_reg(mx3fb, di_mappings[13], DI_DISP3_B1_MAP); - mx3fb_write_reg(mx3fb, di_mappings[14], DI_DISP3_B2_MAP); - mx3fb_write_reg(mx3fb, mx3fb_read_reg(mx3fb, DI_DISP_ACC_CC) | - ((di_mappings[15] - 1) << 12), DI_DISP_ACC_CC); - break; - } + map = &di_mappings[mx3fb->disp_data_fmt]; + mx3fb_write_reg(mx3fb, map->b0, DI_DISP3_B0_MAP); + mx3fb_write_reg(mx3fb, map->b1, DI_DISP3_B1_MAP); + mx3fb_write_reg(mx3fb, map->b2, DI_DISP3_B2_MAP); spin_unlock_irqrestore(&mx3fb->lock, lock_flags); @@ -780,8 +757,6 @@ static int __set_par(struct fb_info *fbi, bool lock) if (sdc_init_panel(mx3fb, mode, (PICOS2KHZ(fbi->var.pixclock)) * 1000UL, fbi->var.xres, fbi->var.yres, - (fbi->var.sync & FB_SYNC_SWAP_RGB) ? - IPU_PIX_FMT_BGR666 : IPU_PIX_FMT_RGB666, fbi->var.left_margin, fbi->var.hsync_len, fbi->var.right_margin + @@ -1349,6 +1324,12 @@ static int init_fb_chan(struct mx3fb_data *mx3fb, struct idmac_channel *ichan) const struct fb_videomode *mode; int ret, num_modes; + if (mx3fb_pdata->disp_data_fmt >= ARRAY_SIZE(di_mappings)) { + dev_err(dev, "Illegal display data format %d\n", + mx3fb_pdata->disp_data_fmt); + return -EINVAL; + } + ichan->client = mx3fb; irq = ichan->eof_irq; @@ -1402,6 +1383,8 @@ static int init_fb_chan(struct mx3fb_data *mx3fb, struct idmac_channel *ichan) mx3fbi->mx3fb = mx3fb; mx3fbi->blank = FB_BLANK_NORMAL; + mx3fb->disp_data_fmt = mx3fb_pdata->disp_data_fmt; + init_completion(&mx3fbi->flip_cmpl); disable_irq(ichan->eof_irq); dev_dbg(mx3fb->dev, "disabling irq %d\n", ichan->eof_irq); From 3506c0d507144d9b0f19efd5a56d289f70611179 Mon Sep 17 00:00:00 2001 From: Tushar Behera Date: Tue, 6 Dec 2011 16:15:54 +0530 Subject: [PATCH 137/820] DMA: PL330: Remove pm_runtime_xxx calls from pl330 probe/remove amba_probe() now calls pm_runtime_get_noresume() and pm_runtime_enable() for the devices before the device probe is called. Hence we don't need to call pm_runtime_get_xxx and pm_runtime_enable() in device probe again. In the same way, since amba_remove() calls the respective pm_runtime functions, those functions need not be called from device remove. This patch fixes following run time error with pl330 driver. dma-pl330 dma-pl330.0: Unbalanced pm_runtime_enable! dma-pl330 dma-pl330.0: failed to get runtime pm Signed-off-by: Giridhar Maruthy Signed-off-by: Tushar Behera Signed-off-by: Vinod Koul --- drivers/dma/pl330.c | 17 ++--------------- 1 file changed, 2 insertions(+), 15 deletions(-) diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c index e0da795bdcb3..a5737575d236 100644 --- a/drivers/dma/pl330.c +++ b/drivers/dma/pl330.c @@ -820,17 +820,7 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id) amba_set_drvdata(adev, pdmac); -#ifdef CONFIG_PM_RUNTIME - /* to use the runtime PM helper functions */ - pm_runtime_enable(&adev->dev); - - /* enable the power domain */ - if (pm_runtime_get_sync(&adev->dev)) { - dev_err(&adev->dev, "failed to get runtime pm\n"); - ret = -ENODEV; - goto probe_err1; - } -#else +#ifndef CONFIG_PM_RUNTIME /* enable dma clk */ clk_enable(pdmac->clk); #endif @@ -971,10 +961,7 @@ static int __devexit pl330_remove(struct amba_device *adev) res = &adev->res; release_mem_region(res->start, resource_size(res)); -#ifdef CONFIG_PM_RUNTIME - pm_runtime_put(&adev->dev); - pm_runtime_disable(&adev->dev); -#else +#ifndef CONFIG_PM_RUNTIME clk_disable(pdmac->clk); #endif From e26d99aed42ec6cdc9540d19c77ac5d4dd2c5b00 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 14 Nov 2011 12:30:30 -0500 Subject: [PATCH 138/820] target: reshuffle headers Create a new headers, drivers/target/target_core_internal.h that is supposed to hold all target_core-internal prototypes. Move all non-exported includes from include/target to it, and merge the smaller prototype-only includes inside drivers/target into it as well. Mark functions that were found to not be called outside their implementation file static. Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_alua.c | 2 +- drivers/target/target_core_cdb.c | 3 +- drivers/target/target_core_cdb.h | 14 --- drivers/target/target_core_configfs.c | 3 +- drivers/target/target_core_device.c | 2 +- drivers/target/target_core_fabric_configfs.c | 3 +- drivers/target/target_core_fabric_lib.c | 2 +- drivers/target/target_core_hba.c | 2 +- drivers/target/target_core_hba.h | 7 -- drivers/target/target_core_internal.h | 123 +++++++++++++++++++ drivers/target/target_core_pr.c | 2 +- drivers/target/target_core_stat.c | 2 +- drivers/target/target_core_stat.h | 8 -- drivers/target/target_core_tmr.c | 1 + drivers/target/target_core_tpg.c | 3 +- drivers/target/target_core_transport.c | 67 +++++----- drivers/target/target_core_ua.c | 2 +- include/target/target_core_device.h | 58 +-------- include/target/target_core_tmr.h | 4 - include/target/target_core_tpg.h | 13 -- include/target/target_core_transport.h | 35 +----- 21 files changed, 170 insertions(+), 186 deletions(-) delete mode 100644 drivers/target/target_core_cdb.h delete mode 100644 drivers/target/target_core_hba.h create mode 100644 drivers/target/target_core_internal.h delete mode 100644 drivers/target/target_core_stat.h diff --git a/drivers/target/target_core_alua.c b/drivers/target/target_core_alua.c index 1dcbef499d6a..097eb5a49648 100644 --- a/drivers/target/target_core_alua.c +++ b/drivers/target/target_core_alua.c @@ -37,8 +37,8 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" -#include "target_core_hba.h" #include "target_core_ua.h" static int core_alua_check_transition(int state, int *primary); diff --git a/drivers/target/target_core_cdb.c b/drivers/target/target_core_cdb.c index 831468b3163d..8f2934fcc68c 100644 --- a/drivers/target/target_core_cdb.c +++ b/drivers/target/target_core_cdb.c @@ -31,8 +31,9 @@ #include #include #include + +#include "target_core_internal.h" #include "target_core_ua.h" -#include "target_core_cdb.h" static void target_fill_alua_data(struct se_port *port, unsigned char *buf) diff --git a/drivers/target/target_core_cdb.h b/drivers/target/target_core_cdb.h deleted file mode 100644 index ad6b1e393001..000000000000 --- a/drivers/target/target_core_cdb.h +++ /dev/null @@ -1,14 +0,0 @@ -#ifndef TARGET_CORE_CDB_H -#define TARGET_CORE_CDB_H - -int target_emulate_inquiry(struct se_task *task); -int target_emulate_readcapacity(struct se_task *task); -int target_emulate_readcapacity_16(struct se_task *task); -int target_emulate_modesense(struct se_task *task); -int target_emulate_request_sense(struct se_task *task); -int target_emulate_unmap(struct se_task *task); -int target_emulate_write_same(struct se_task *task); -int target_emulate_synchronize_cache(struct se_task *task); -int target_emulate_noop(struct se_task *task); - -#endif /* TARGET_CORE_CDB_H */ diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index 93d4f6a1b798..ac55af56ba12 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -46,11 +46,10 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" -#include "target_core_hba.h" #include "target_core_pr.h" #include "target_core_rd.h" -#include "target_core_stat.h" extern struct t10_alua_lu_gp *default_lu_gp; diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index 9b8639425472..28ffe0b52bde 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -47,8 +47,8 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" -#include "target_core_hba.h" #include "target_core_pr.h" #include "target_core_ua.h" diff --git a/drivers/target/target_core_fabric_configfs.c b/drivers/target/target_core_fabric_configfs.c index 09b6f8729f91..fa1e4eb01bc7 100644 --- a/drivers/target/target_core_fabric_configfs.c +++ b/drivers/target/target_core_fabric_configfs.c @@ -44,10 +44,9 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" -#include "target_core_hba.h" #include "target_core_pr.h" -#include "target_core_stat.h" #define TF_CIT_SETUP(_name, _item_ops, _group_ops, _attrs) \ static void target_fabric_setup_##_name##_cit(struct target_fabric_configfs *tf) \ diff --git a/drivers/target/target_core_fabric_lib.c b/drivers/target/target_core_fabric_lib.c index ec4249be617e..dd9ae929dda2 100644 --- a/drivers/target/target_core_fabric_lib.c +++ b/drivers/target/target_core_fabric_lib.c @@ -40,7 +40,7 @@ #include #include -#include "target_core_hba.h" +#include "target_core_internal.h" #include "target_core_pr.h" /* diff --git a/drivers/target/target_core_hba.c b/drivers/target/target_core_hba.c index c68019d6c406..26c1238976cb 100644 --- a/drivers/target/target_core_hba.c +++ b/drivers/target/target_core_hba.c @@ -41,7 +41,7 @@ #include #include -#include "target_core_hba.h" +#include "target_core_internal.h" static LIST_HEAD(subsystem_list); static DEFINE_MUTEX(subsystem_mutex); diff --git a/drivers/target/target_core_hba.h b/drivers/target/target_core_hba.h deleted file mode 100644 index bb0fea5f730c..000000000000 --- a/drivers/target/target_core_hba.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef TARGET_CORE_HBA_H -#define TARGET_CORE_HBA_H - -extern struct se_hba *core_alloc_hba(const char *, u32, u32); -extern int core_delete_hba(struct se_hba *); - -#endif /* TARGET_CORE_HBA_H */ diff --git a/drivers/target/target_core_internal.h b/drivers/target/target_core_internal.h new file mode 100644 index 000000000000..26f135e94f6e --- /dev/null +++ b/drivers/target/target_core_internal.h @@ -0,0 +1,123 @@ +#ifndef TARGET_CORE_INTERNAL_H +#define TARGET_CORE_INTERNAL_H + +/* target_core_alua.c */ +extern struct t10_alua_lu_gp *default_lu_gp; + +/* target_core_cdb.c */ +int target_emulate_inquiry(struct se_task *task); +int target_emulate_readcapacity(struct se_task *task); +int target_emulate_readcapacity_16(struct se_task *task); +int target_emulate_modesense(struct se_task *task); +int target_emulate_request_sense(struct se_task *task); +int target_emulate_unmap(struct se_task *task); +int target_emulate_write_same(struct se_task *task); +int target_emulate_synchronize_cache(struct se_task *task); +int target_emulate_noop(struct se_task *task); + +/* target_core_device.c */ +struct se_dev_entry *core_get_se_deve_from_rtpi(struct se_node_acl *, u16); +int core_free_device_list_for_node(struct se_node_acl *, + struct se_portal_group *); +void core_dec_lacl_count(struct se_node_acl *, struct se_cmd *); +void core_update_device_list_access(u32, u32, struct se_node_acl *); +int core_update_device_list_for_node(struct se_lun *, struct se_lun_acl *, + u32, u32, struct se_node_acl *, struct se_portal_group *, int); +void core_clear_lun_from_tpg(struct se_lun *, struct se_portal_group *); +int core_dev_export(struct se_device *, struct se_portal_group *, + struct se_lun *); +void core_dev_unexport(struct se_device *, struct se_portal_group *, + struct se_lun *); +int target_report_luns(struct se_task *); +void se_release_device_for_hba(struct se_device *); +void se_release_vpd_for_dev(struct se_device *); +int se_free_virtual_device(struct se_device *, struct se_hba *); +int se_dev_check_online(struct se_device *); +int se_dev_check_shutdown(struct se_device *); +void se_dev_set_default_attribs(struct se_device *, struct se_dev_limits *); +int se_dev_set_task_timeout(struct se_device *, u32); +int se_dev_set_max_unmap_lba_count(struct se_device *, u32); +int se_dev_set_max_unmap_block_desc_count(struct se_device *, u32); +int se_dev_set_unmap_granularity(struct se_device *, u32); +int se_dev_set_unmap_granularity_alignment(struct se_device *, u32); +int se_dev_set_emulate_dpo(struct se_device *, int); +int se_dev_set_emulate_fua_write(struct se_device *, int); +int se_dev_set_emulate_fua_read(struct se_device *, int); +int se_dev_set_emulate_write_cache(struct se_device *, int); +int se_dev_set_emulate_ua_intlck_ctrl(struct se_device *, int); +int se_dev_set_emulate_tas(struct se_device *, int); +int se_dev_set_emulate_tpu(struct se_device *, int); +int se_dev_set_emulate_tpws(struct se_device *, int); +int se_dev_set_enforce_pr_isids(struct se_device *, int); +int se_dev_set_is_nonrot(struct se_device *, int); +int se_dev_set_emulate_rest_reord(struct se_device *dev, int); +int se_dev_set_queue_depth(struct se_device *, u32); +int se_dev_set_max_sectors(struct se_device *, u32); +int se_dev_set_optimal_sectors(struct se_device *, u32); +int se_dev_set_block_size(struct se_device *, u32); +struct se_lun *core_dev_add_lun(struct se_portal_group *, struct se_hba *, + struct se_device *, u32); +int core_dev_del_lun(struct se_portal_group *, u32); +struct se_lun *core_get_lun_from_tpg(struct se_portal_group *, u32); +struct se_lun_acl *core_dev_init_initiator_node_lun_acl(struct se_portal_group *, + u32, char *, int *); +int core_dev_add_initiator_node_lun_acl(struct se_portal_group *, + struct se_lun_acl *, u32, u32); +int core_dev_del_initiator_node_lun_acl(struct se_portal_group *, + struct se_lun *, struct se_lun_acl *); +void core_dev_free_initiator_node_lun_acl(struct se_portal_group *, + struct se_lun_acl *lacl); +int core_dev_setup_virtual_lun0(void); +void core_dev_release_virtual_lun0(void); + +/* target_core_hba.c */ +struct se_hba *core_alloc_hba(const char *, u32, u32); +int core_delete_hba(struct se_hba *); + +/* target_core_tmr.c */ +int core_tmr_lun_reset(struct se_device *, struct se_tmr_req *, + struct list_head *, struct se_cmd *); + +/* target_core_tpg.c */ +extern struct se_device *g_lun0_dev; + +struct se_node_acl *__core_tpg_get_initiator_node_acl(struct se_portal_group *tpg, + const char *); +struct se_node_acl *core_tpg_get_initiator_node_acl(struct se_portal_group *tpg, + unsigned char *); +void core_tpg_add_node_to_devs(struct se_node_acl *, struct se_portal_group *); +void core_tpg_wait_for_nacl_pr_ref(struct se_node_acl *); +struct se_lun *core_tpg_pre_addlun(struct se_portal_group *, u32); +int core_tpg_post_addlun(struct se_portal_group *, struct se_lun *, + u32, void *); +struct se_lun *core_tpg_pre_dellun(struct se_portal_group *, u32, int *); +int core_tpg_post_dellun(struct se_portal_group *, struct se_lun *); + +/* target_core_transport.c */ +extern struct kmem_cache *se_tmr_req_cache; + +int init_se_kmem_caches(void); +void release_se_kmem_caches(void); +u32 scsi_get_new_index(scsi_index_t); +void transport_subsystem_check_init(void); +void transport_cmd_finish_abort(struct se_cmd *, int); +void __transport_remove_task_from_execute_queue(struct se_task *, + struct se_device *); +unsigned char *transport_dump_cmd_direction(struct se_cmd *); +void transport_dump_dev_state(struct se_device *, char *, int *); +void transport_dump_dev_info(struct se_device *, struct se_lun *, + unsigned long long, char *, int *); +void transport_dump_vpd_proto_id(struct t10_vpd *, unsigned char *, int); +int transport_dump_vpd_assoc(struct t10_vpd *, unsigned char *, int); +int transport_dump_vpd_ident_type(struct t10_vpd *, unsigned char *, int); +int transport_dump_vpd_ident(struct t10_vpd *, unsigned char *, int); +bool target_stop_task(struct se_task *task, unsigned long *flags); +int transport_clear_lun_from_sessions(struct se_lun *); +void transport_send_task_abort(struct se_cmd *); + +/* target_core_stat.c */ +void target_stat_setup_dev_default_groups(struct se_subsystem_dev *); +void target_stat_setup_port_default_groups(struct se_lun *); +void target_stat_setup_mappedlun_default_groups(struct se_lun_acl *); + +#endif /* TARGET_CORE_INTERNAL_H */ diff --git a/drivers/target/target_core_pr.c b/drivers/target/target_core_pr.c index 95dee7074aeb..36fa9a5f1ba6 100644 --- a/drivers/target/target_core_pr.c +++ b/drivers/target/target_core_pr.c @@ -40,7 +40,7 @@ #include #include -#include "target_core_hba.h" +#include "target_core_internal.h" #include "target_core_pr.h" #include "target_core_ua.h" diff --git a/drivers/target/target_core_stat.c b/drivers/target/target_core_stat.c index 874152aed94a..d2ba9266ff4d 100644 --- a/drivers/target/target_core_stat.c +++ b/drivers/target/target_core_stat.c @@ -48,7 +48,7 @@ #include #include -#include "target_core_hba.h" +#include "target_core_internal.h" #ifndef INITIAL_JIFFIES #define INITIAL_JIFFIES ((unsigned long)(unsigned int) (-300*HZ)) diff --git a/drivers/target/target_core_stat.h b/drivers/target/target_core_stat.h deleted file mode 100644 index 86c252f9ea47..000000000000 --- a/drivers/target/target_core_stat.h +++ /dev/null @@ -1,8 +0,0 @@ -#ifndef TARGET_CORE_STAT_H -#define TARGET_CORE_STAT_H - -extern void target_stat_setup_dev_default_groups(struct se_subsystem_dev *); -extern void target_stat_setup_port_default_groups(struct se_lun *); -extern void target_stat_setup_mappedlun_default_groups(struct se_lun_acl *); - -#endif /*** TARGET_CORE_STAT_H ***/ diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c index 684522805a1f..47bc26ba7865 100644 --- a/drivers/target/target_core_tmr.c +++ b/drivers/target/target_core_tmr.c @@ -38,6 +38,7 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" #include "target_core_pr.h" diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c index 8ddd133025b9..0a936fae3d23 100644 --- a/drivers/target/target_core_tpg.c +++ b/drivers/target/target_core_tpg.c @@ -44,8 +44,7 @@ #include #include -#include "target_core_hba.h" -#include "target_core_stat.h" +#include "target_core_internal.h" extern struct se_device *g_lun0_dev; diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 0257658e2e3e..e0ea63dedde3 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -52,9 +52,8 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" -#include "target_core_cdb.h" -#include "target_core_hba.h" #include "target_core_pr.h" #include "target_core_ua.h" @@ -212,14 +211,13 @@ u32 scsi_get_new_index(scsi_index_t type) return new_index; } -void transport_init_queue_obj(struct se_queue_obj *qobj) +static void transport_init_queue_obj(struct se_queue_obj *qobj) { atomic_set(&qobj->queue_cnt, 0); INIT_LIST_HEAD(&qobj->qobj_list); init_waitqueue_head(&qobj->thread_wq); spin_lock_init(&qobj->cmd_queue_lock); } -EXPORT_SYMBOL(transport_init_queue_obj); void transport_subsystem_check_init(void) { @@ -896,7 +894,7 @@ void __transport_remove_task_from_execute_queue(struct se_task *task, atomic_dec(&dev->execute_tasks); } -void transport_remove_task_from_execute_queue( +static void transport_remove_task_from_execute_queue( struct se_task *task, struct se_device *dev) { @@ -3345,6 +3343,32 @@ static inline void transport_free_pages(struct se_cmd *cmd) cmd->t_bidi_data_nents = 0; } +/** + * transport_release_cmd - free a command + * @cmd: command to free + * + * This routine unconditionally frees a command, and reference counting + * or list removal must be done in the caller. + */ +static void transport_release_cmd(struct se_cmd *cmd) +{ + BUG_ON(!cmd->se_tfo); + + if (cmd->se_tmr_req) + core_tmr_release_req(cmd->se_tmr_req); + if (cmd->t_task_cdb != cmd->__t_task_cdb) + kfree(cmd->t_task_cdb); + /* + * Check if target_wait_for_sess_cmds() is expecting to + * release se_cmd directly here.. + */ + if (cmd->check_release != 0 && cmd->se_tfo->check_release_cmd) + if (cmd->se_tfo->check_release_cmd(cmd) != 0) + return; + + cmd->se_tfo->release_cmd(cmd); +} + /** * transport_put_cmd - release a reference to a command * @cmd: command to release @@ -3870,33 +3894,6 @@ queue_full: return 0; } -/** - * transport_release_cmd - free a command - * @cmd: command to free - * - * This routine unconditionally frees a command, and reference counting - * or list removal must be done in the caller. - */ -void transport_release_cmd(struct se_cmd *cmd) -{ - BUG_ON(!cmd->se_tfo); - - if (cmd->se_tmr_req) - core_tmr_release_req(cmd->se_tmr_req); - if (cmd->t_task_cdb != cmd->__t_task_cdb) - kfree(cmd->t_task_cdb); - /* - * Check if target_wait_for_sess_cmds() is expecting to - * release se_cmd directly here.. - */ - if (cmd->check_release != 0 && cmd->se_tfo->check_release_cmd) - if (cmd->se_tfo->check_release_cmd(cmd) != 0) - return; - - cmd->se_tfo->release_cmd(cmd); -} -EXPORT_SYMBOL(transport_release_cmd); - void transport_generic_free_cmd(struct se_cmd *cmd, int wait_for_tasks) { if (!(cmd->se_cmd_flags & SCF_SE_LUN_CMD)) { @@ -4545,11 +4542,7 @@ void transport_send_task_abort(struct se_cmd *cmd) cmd->se_tfo->queue_status(cmd); } -/* transport_generic_do_tmr(): - * - * - */ -int transport_generic_do_tmr(struct se_cmd *cmd) +static int transport_generic_do_tmr(struct se_cmd *cmd) { struct se_device *dev = cmd->se_dev; struct se_tmr_req *tmr = cmd->se_tmr_req; diff --git a/drivers/target/target_core_ua.c b/drivers/target/target_core_ua.c index 50a480db7a66..e2ae53c7e088 100644 --- a/drivers/target/target_core_ua.c +++ b/drivers/target/target_core_ua.c @@ -35,8 +35,8 @@ #include #include +#include "target_core_internal.h" #include "target_core_alua.h" -#include "target_core_hba.h" #include "target_core_pr.h" #include "target_core_ua.h" diff --git a/include/target/target_core_device.h b/include/target/target_core_device.h index 2be31ff8763b..c6123a49014d 100644 --- a/include/target/target_core_device.h +++ b/include/target/target_core_device.h @@ -1,63 +1,9 @@ #ifndef TARGET_CORE_DEVICE_H #define TARGET_CORE_DEVICE_H + +// external extern int transport_lookup_cmd_lun(struct se_cmd *, u32); extern int transport_lookup_tmr_lun(struct se_cmd *, u32); -extern struct se_dev_entry *core_get_se_deve_from_rtpi( - struct se_node_acl *, u16); -extern int core_free_device_list_for_node(struct se_node_acl *, - struct se_portal_group *); -extern void core_dec_lacl_count(struct se_node_acl *, struct se_cmd *); -extern void core_update_device_list_access(u32, u32, struct se_node_acl *); -extern int core_update_device_list_for_node(struct se_lun *, struct se_lun_acl *, u32, - u32, struct se_node_acl *, - struct se_portal_group *, int); -extern void core_clear_lun_from_tpg(struct se_lun *, struct se_portal_group *); -extern int core_dev_export(struct se_device *, struct se_portal_group *, - struct se_lun *); -extern void core_dev_unexport(struct se_device *, struct se_portal_group *, - struct se_lun *); -extern int target_report_luns(struct se_task *); -extern void se_release_device_for_hba(struct se_device *); -extern void se_release_vpd_for_dev(struct se_device *); -extern void se_clear_dev_ports(struct se_device *); -extern int se_free_virtual_device(struct se_device *, struct se_hba *); -extern int se_dev_check_online(struct se_device *); -extern int se_dev_check_shutdown(struct se_device *); -extern void se_dev_set_default_attribs(struct se_device *, struct se_dev_limits *); -extern int se_dev_set_task_timeout(struct se_device *, u32); -extern int se_dev_set_max_unmap_lba_count(struct se_device *, u32); -extern int se_dev_set_max_unmap_block_desc_count(struct se_device *, u32); -extern int se_dev_set_unmap_granularity(struct se_device *, u32); -extern int se_dev_set_unmap_granularity_alignment(struct se_device *, u32); -extern int se_dev_set_emulate_dpo(struct se_device *, int); -extern int se_dev_set_emulate_fua_write(struct se_device *, int); -extern int se_dev_set_emulate_fua_read(struct se_device *, int); -extern int se_dev_set_emulate_write_cache(struct se_device *, int); -extern int se_dev_set_emulate_ua_intlck_ctrl(struct se_device *, int); -extern int se_dev_set_emulate_tas(struct se_device *, int); -extern int se_dev_set_emulate_tpu(struct se_device *, int); -extern int se_dev_set_emulate_tpws(struct se_device *, int); -extern int se_dev_set_enforce_pr_isids(struct se_device *, int); -extern int se_dev_set_is_nonrot(struct se_device *, int); -extern int se_dev_set_emulate_rest_reord(struct se_device *dev, int); -extern int se_dev_set_queue_depth(struct se_device *, u32); -extern int se_dev_set_max_sectors(struct se_device *, u32); -extern int se_dev_set_optimal_sectors(struct se_device *, u32); -extern int se_dev_set_block_size(struct se_device *, u32); -extern struct se_lun *core_dev_add_lun(struct se_portal_group *, struct se_hba *, - struct se_device *, u32); -extern int core_dev_del_lun(struct se_portal_group *, u32); -extern struct se_lun *core_get_lun_from_tpg(struct se_portal_group *, u32); -extern struct se_lun_acl *core_dev_init_initiator_node_lun_acl(struct se_portal_group *, - u32, char *, int *); -extern int core_dev_add_initiator_node_lun_acl(struct se_portal_group *, - struct se_lun_acl *, u32, u32); -extern int core_dev_del_initiator_node_lun_acl(struct se_portal_group *, - struct se_lun *, struct se_lun_acl *); -extern void core_dev_free_initiator_node_lun_acl(struct se_portal_group *, - struct se_lun_acl *lacl); -extern int core_dev_setup_virtual_lun0(void); -extern void core_dev_release_virtual_lun0(void); #endif /* TARGET_CORE_DEVICE_H */ diff --git a/include/target/target_core_tmr.h b/include/target/target_core_tmr.h index d5876e17d3fb..36b3e0de99b4 100644 --- a/include/target/target_core_tmr.h +++ b/include/target/target_core_tmr.h @@ -25,11 +25,7 @@ enum tcm_tmrsp_table { TMR_FUNCTION_REJECTED = 255, }; -extern struct kmem_cache *se_tmr_req_cache; - extern struct se_tmr_req *core_tmr_alloc_req(struct se_cmd *, void *, u8, gfp_t); extern void core_tmr_release_req(struct se_tmr_req *); -extern int core_tmr_lun_reset(struct se_device *, struct se_tmr_req *, - struct list_head *, struct se_cmd *); #endif /* TARGET_CORE_TMR_H */ diff --git a/include/target/target_core_tpg.h b/include/target/target_core_tpg.h index 77e18729c4c1..80ae59fdada3 100644 --- a/include/target/target_core_tpg.h +++ b/include/target/target_core_tpg.h @@ -1,17 +1,9 @@ #ifndef TARGET_CORE_TPG_H #define TARGET_CORE_TPG_H -extern struct se_node_acl *__core_tpg_get_initiator_node_acl(struct se_portal_group *tpg, - const char *); -extern struct se_node_acl *core_tpg_get_initiator_node_acl(struct se_portal_group *tpg, - unsigned char *); -extern void core_tpg_add_node_to_devs(struct se_node_acl *, - struct se_portal_group *); extern struct se_node_acl *core_tpg_check_initiator_node_acl( struct se_portal_group *, unsigned char *); -extern void core_tpg_wait_for_nacl_pr_ref(struct se_node_acl *); -extern void core_tpg_wait_for_mib_ref(struct se_node_acl *); extern void core_tpg_clear_object_luns(struct se_portal_group *); extern struct se_node_acl *core_tpg_add_initiator_node_acl( struct se_portal_group *, @@ -26,10 +18,5 @@ extern int core_tpg_register(struct target_core_fabric_ops *, struct se_portal_group *, void *, int); extern int core_tpg_deregister(struct se_portal_group *); -extern struct se_lun *core_tpg_pre_addlun(struct se_portal_group *, u32); -extern int core_tpg_post_addlun(struct se_portal_group *, struct se_lun *, u32, - void *); -extern struct se_lun *core_tpg_pre_dellun(struct se_portal_group *, u32, int *); -extern int core_tpg_post_dellun(struct se_portal_group *, struct se_lun *); #endif /* TARGET_CORE_TPG_H */ diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h index dac4f2d859fd..68fe1f2c081c 100644 --- a/include/target/target_core_transport.h +++ b/include/target/target_core_transport.h @@ -76,14 +76,8 @@ struct se_subsystem_api; -extern int init_se_kmem_caches(void); -extern void release_se_kmem_caches(void); -extern u32 scsi_get_new_index(scsi_index_t); -extern void transport_init_queue_obj(struct se_queue_obj *); -extern void transport_subsystem_check_init(void); extern int transport_subsystem_register(struct se_subsystem_api *); extern void transport_subsystem_release(struct se_subsystem_api *); -extern void transport_load_plugins(void); extern struct se_session *transport_init_session(void); extern void __transport_register_session(struct se_portal_group *, struct se_node_acl *, @@ -94,32 +88,14 @@ extern void transport_register_session(struct se_portal_group *, extern void transport_free_session(struct se_session *); extern void transport_deregister_session_configfs(struct se_session *); extern void transport_deregister_session(struct se_session *); -extern void transport_cmd_finish_abort(struct se_cmd *, int); extern void transport_complete_sync_cache(struct se_cmd *, int); extern void transport_complete_task(struct se_task *, int); -extern void transport_add_task_to_execute_queue(struct se_task *, - struct se_task *, - struct se_device *); -extern void transport_remove_task_from_execute_queue(struct se_task *, - struct se_device *); -extern void __transport_remove_task_from_execute_queue(struct se_task *, - struct se_device *); -unsigned char *transport_dump_cmd_direction(struct se_cmd *); -extern void transport_dump_dev_state(struct se_device *, char *, int *); -extern void transport_dump_dev_info(struct se_device *, struct se_lun *, - unsigned long long, char *, int *); -extern void transport_dump_vpd_proto_id(struct t10_vpd *, - unsigned char *, int); + extern void transport_set_vpd_proto_id(struct t10_vpd *, unsigned char *); -extern int transport_dump_vpd_assoc(struct t10_vpd *, - unsigned char *, int); extern int transport_set_vpd_assoc(struct t10_vpd *, unsigned char *); -extern int transport_dump_vpd_ident_type(struct t10_vpd *, - unsigned char *, int); extern int transport_set_vpd_ident_type(struct t10_vpd *, unsigned char *); -extern int transport_dump_vpd_ident(struct t10_vpd *, - unsigned char *, int); extern int transport_set_vpd_ident(struct t10_vpd *, unsigned char *); + extern struct se_device *transport_add_device_to_core_hba(struct se_hba *, struct se_subsystem_api *, struct se_subsystem_dev *, u32, @@ -136,29 +112,22 @@ extern int transport_handle_cdb_direct(struct se_cmd *); extern int transport_generic_handle_cdb_map(struct se_cmd *); extern int transport_generic_handle_data(struct se_cmd *); extern int transport_generic_handle_tmr(struct se_cmd *); -extern bool target_stop_task(struct se_task *task, unsigned long *flags); extern int transport_generic_map_mem_to_cmd(struct se_cmd *cmd, struct scatterlist *, u32, struct scatterlist *, u32); -extern int transport_clear_lun_from_sessions(struct se_lun *); extern bool transport_wait_for_tasks(struct se_cmd *); extern int transport_check_aborted_status(struct se_cmd *, int); extern int transport_send_check_condition_and_sense(struct se_cmd *, u8, int); -extern void transport_send_task_abort(struct se_cmd *); -extern void transport_release_cmd(struct se_cmd *); extern void transport_generic_free_cmd(struct se_cmd *, int); extern void target_get_sess_cmd(struct se_session *, struct se_cmd *); extern int target_put_sess_cmd(struct se_session *, struct se_cmd *); extern void target_splice_sess_cmd_list(struct se_session *); extern void target_wait_for_sess_cmds(struct se_session *, int); -extern void transport_generic_wait_for_cmds(struct se_cmd *, int); extern void transport_do_task_sg_chain(struct se_cmd *); extern void transport_generic_process_write(struct se_cmd *); extern int transport_generic_new_cmd(struct se_cmd *); -extern int transport_generic_do_tmr(struct se_cmd *); /* From target_core_alua.c */ extern int core_alua_check_nonop_delay(struct se_cmd *); /* From target_core_cdb.c */ -extern int transport_emulate_control_cdb(struct se_task *); extern void target_get_task_cdb(struct se_task *task, unsigned char *cdb); /* From c4795fb20edf2fe2c862c8fe9f8b681edeb79ac1 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 16 Nov 2011 09:46:48 -0500 Subject: [PATCH 139/820] target: header reshuffle, part2 This reorganized the headers under include/target into: - target_core_base.h stays as is with all target-wide data stuctures and defines - target_core_backend.h contains the whole interface to I/O backends - target_core_fabric.h contains the whole interface to fabric modules Except for those only the various configfs macro headers stay around. Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- Documentation/target/tcm_mod_builder.py | 15 +- drivers/target/iscsi/iscsi_target.c | 3 +- drivers/target/iscsi/iscsi_target_configfs.c | 6 +- drivers/target/iscsi/iscsi_target_device.c | 3 +- drivers/target/iscsi/iscsi_target_erl0.c | 2 +- drivers/target/iscsi/iscsi_target_erl1.c | 2 +- drivers/target/iscsi/iscsi_target_erl2.c | 2 +- drivers/target/iscsi/iscsi_target_login.c | 2 +- drivers/target/iscsi/iscsi_target_nego.c | 2 +- .../target/iscsi/iscsi_target_nodeattrib.c | 1 - drivers/target/iscsi/iscsi_target_stat.c | 1 - drivers/target/iscsi/iscsi_target_tmr.c | 2 +- drivers/target/iscsi/iscsi_target_tpg.c | 4 +- drivers/target/iscsi/iscsi_target_util.c | 4 +- drivers/target/loopback/tcm_loop.c | 7 +- drivers/target/target_core_alua.c | 5 +- drivers/target/target_core_cdb.c | 4 +- drivers/target/target_core_configfs.c | 5 +- drivers/target/target_core_device.c | 6 +- drivers/target/target_core_fabric_configfs.c | 5 +- drivers/target/target_core_fabric_lib.c | 5 +- drivers/target/target_core_file.c | 3 +- drivers/target/target_core_hba.c | 5 +- drivers/target/target_core_iblock.c | 3 +- drivers/target/target_core_pr.c | 7 +- drivers/target/target_core_pscsi.c | 3 +- drivers/target/target_core_rd.c | 4 +- drivers/target/target_core_stat.c | 4 +- drivers/target/target_core_tmr.c | 6 +- drivers/target/target_core_tpg.c | 6 +- drivers/target/target_core_transport.c | 7 +- drivers/target/target_core_ua.c | 4 +- drivers/target/tcm_fc/tfc_cmd.c | 6 +- drivers/target/tcm_fc/tfc_conf.c | 6 +- drivers/target/tcm_fc/tfc_io.c | 5 +- drivers/target/tcm_fc/tfc_sess.c | 5 +- include/target/target_core_backend.h | 65 +++++ include/target/target_core_base.h | 91 +++++++ include/target/target_core_device.h | 9 - ...core_fabric_ops.h => target_core_fabric.h} | 88 +++++- include/target/target_core_fabric_lib.h | 28 -- include/target/target_core_tmr.h | 31 --- include/target/target_core_tpg.h | 22 -- include/target/target_core_transport.h | 256 ------------------ 44 files changed, 288 insertions(+), 462 deletions(-) create mode 100644 include/target/target_core_backend.h delete mode 100644 include/target/target_core_device.h rename include/target/{target_core_fabric_ops.h => target_core_fabric.h} (51%) delete mode 100644 include/target/target_core_fabric_lib.h delete mode 100644 include/target/target_core_tmr.h delete mode 100644 include/target/target_core_tpg.h delete mode 100644 include/target/target_core_transport.h diff --git a/Documentation/target/tcm_mod_builder.py b/Documentation/target/tcm_mod_builder.py index 7ef9b843d529..e2abb29fb993 100755 --- a/Documentation/target/tcm_mod_builder.py +++ b/Documentation/target/tcm_mod_builder.py @@ -230,14 +230,9 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += "#include \n" buf += "#include \n\n" buf += "#include \n" - buf += "#include \n" - buf += "#include \n" + buf += "#include \n" buf += "#include \n" - buf += "#include \n" - buf += "#include \n" - buf += "#include \n" buf += "#include \n" - buf += "#include \n" buf += "#include \n\n" buf += "#include \"" + fabric_mod_name + "_base.h\"\n" buf += "#include \"" + fabric_mod_name + "_fabric.h\"\n\n" @@ -514,7 +509,7 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): def tcm_mod_scan_fabric_ops(tcm_dir): - fabric_ops_api = tcm_dir + "include/target/target_core_fabric_ops.h" + fabric_ops_api = tcm_dir + "include/target/target_core_fabric.h" print "Using tcm_mod_scan_fabric_ops: " + fabric_ops_api process_fo = 0; @@ -579,11 +574,7 @@ def tcm_mod_dump_fabric_ops(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += "#include \n" buf += "#include \n\n" buf += "#include \n" - buf += "#include \n" - buf += "#include \n" - buf += "#include \n" - buf += "#include \n" - buf += "#include \n" + buf += "#include \n" buf += "#include \n\n" buf += "#include \"" + fabric_mod_name + "_base.h\"\n" buf += "#include \"" + fabric_mod_name + "_fabric.h\"\n\n" diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c index 8599545cdf9e..f74a62d46eea 100644 --- a/drivers/target/iscsi/iscsi_target.c +++ b/drivers/target/iscsi/iscsi_target.c @@ -27,8 +27,7 @@ #include #include #include -#include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_parameters.h" diff --git a/drivers/target/iscsi/iscsi_target_configfs.c b/drivers/target/iscsi/iscsi_target_configfs.c index db327845e46b..c04d389a2ff2 100644 --- a/drivers/target/iscsi/iscsi_target_configfs.c +++ b/drivers/target/iscsi/iscsi_target_configfs.c @@ -22,12 +22,8 @@ #include #include #include -#include -#include +#include #include -#include -#include -#include #include #include diff --git a/drivers/target/iscsi/iscsi_target_device.c b/drivers/target/iscsi/iscsi_target_device.c index a19fa5eea88e..f63ea35bc4ae 100644 --- a/drivers/target/iscsi/iscsi_target_device.c +++ b/drivers/target/iscsi/iscsi_target_device.c @@ -21,8 +21,7 @@ #include #include -#include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_device.h" diff --git a/drivers/target/iscsi/iscsi_target_erl0.c b/drivers/target/iscsi/iscsi_target_erl0.c index b7ffc3cd40cc..478451167b62 100644 --- a/drivers/target/iscsi/iscsi_target_erl0.c +++ b/drivers/target/iscsi/iscsi_target_erl0.c @@ -21,7 +21,7 @@ #include #include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_seq_pdu_list.h" diff --git a/drivers/target/iscsi/iscsi_target_erl1.c b/drivers/target/iscsi/iscsi_target_erl1.c index 101b1beb3bca..255c0d67e898 100644 --- a/drivers/target/iscsi/iscsi_target_erl1.c +++ b/drivers/target/iscsi/iscsi_target_erl1.c @@ -21,7 +21,7 @@ #include #include #include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_seq_pdu_list.h" diff --git a/drivers/target/iscsi/iscsi_target_erl2.c b/drivers/target/iscsi/iscsi_target_erl2.c index 0b8404c30125..1af1f21af21f 100644 --- a/drivers/target/iscsi/iscsi_target_erl2.c +++ b/drivers/target/iscsi/iscsi_target_erl2.c @@ -21,7 +21,7 @@ #include #include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_datain_values.h" diff --git a/drivers/target/iscsi/iscsi_target_login.c b/drivers/target/iscsi/iscsi_target_login.c index d734bdec24f9..d74e278ec8f7 100644 --- a/drivers/target/iscsi/iscsi_target_login.c +++ b/drivers/target/iscsi/iscsi_target_login.c @@ -23,7 +23,7 @@ #include #include #include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_tq.h" diff --git a/drivers/target/iscsi/iscsi_target_nego.c b/drivers/target/iscsi/iscsi_target_nego.c index 98936cb7c294..53eea5c827f0 100644 --- a/drivers/target/iscsi/iscsi_target_nego.c +++ b/drivers/target/iscsi/iscsi_target_nego.c @@ -21,7 +21,7 @@ #include #include #include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_parameters.h" diff --git a/drivers/target/iscsi/iscsi_target_nodeattrib.c b/drivers/target/iscsi/iscsi_target_nodeattrib.c index aeafbe0cd7d1..ac047d8a7a13 100644 --- a/drivers/target/iscsi/iscsi_target_nodeattrib.c +++ b/drivers/target/iscsi/iscsi_target_nodeattrib.c @@ -19,7 +19,6 @@ ******************************************************************************/ #include -#include #include "iscsi_target_core.h" #include "iscsi_target_device.h" diff --git a/drivers/target/iscsi/iscsi_target_stat.c b/drivers/target/iscsi/iscsi_target_stat.c index f1db83077e0a..5a1aac0ec7de 100644 --- a/drivers/target/iscsi/iscsi_target_stat.c +++ b/drivers/target/iscsi/iscsi_target_stat.c @@ -23,7 +23,6 @@ #include #include #include -#include #include #include "iscsi_target_core.h" diff --git a/drivers/target/iscsi/iscsi_target_tmr.c b/drivers/target/iscsi/iscsi_target_tmr.c index 490207eacde9..255ed35da815 100644 --- a/drivers/target/iscsi/iscsi_target_tmr.c +++ b/drivers/target/iscsi/iscsi_target_tmr.c @@ -21,7 +21,7 @@ #include #include #include -#include +#include #include "iscsi_target_core.h" #include "iscsi_target_seq_pdu_list.h" diff --git a/drivers/target/iscsi/iscsi_target_tpg.c b/drivers/target/iscsi/iscsi_target_tpg.c index d4cf2cd25c44..cfbcf6e4877d 100644 --- a/drivers/target/iscsi/iscsi_target_tpg.c +++ b/drivers/target/iscsi/iscsi_target_tpg.c @@ -19,10 +19,8 @@ ******************************************************************************/ #include -#include -#include +#include #include -#include #include "iscsi_target_core.h" #include "iscsi_target_erl0.h" diff --git a/drivers/target/iscsi/iscsi_target_util.c b/drivers/target/iscsi/iscsi_target_util.c index 02348f727bd4..c1daea190859 100644 --- a/drivers/target/iscsi/iscsi_target_util.c +++ b/drivers/target/iscsi/iscsi_target_util.c @@ -22,9 +22,7 @@ #include #include #include -#include -#include -#include +#include #include #include "iscsi_target_core.h" diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c index 81d5832fbbd5..d427a290e184 100644 --- a/drivers/target/loopback/tcm_loop.c +++ b/drivers/target/loopback/tcm_loop.c @@ -33,14 +33,9 @@ #include #include -#include -#include +#include #include -#include #include -#include -#include -#include #include "tcm_loop.h" diff --git a/drivers/target/target_core_alua.c b/drivers/target/target_core_alua.c index 097eb5a49648..1b1edd14f4bf 100644 --- a/drivers/target/target_core_alua.c +++ b/drivers/target/target_core_alua.c @@ -32,9 +32,8 @@ #include #include -#include -#include -#include +#include +#include #include #include "target_core_internal.h" diff --git a/drivers/target/target_core_cdb.c b/drivers/target/target_core_cdb.c index 8f2934fcc68c..59219509d886 100644 --- a/drivers/target/target_core_cdb.c +++ b/drivers/target/target_core_cdb.c @@ -29,8 +29,8 @@ #include #include -#include -#include +#include +#include #include "target_core_internal.h" #include "target_core_ua.h" diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index ac55af56ba12..3c358907a4a6 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -39,9 +39,8 @@ #include #include -#include -#include -#include +#include +#include #include #include #include diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index 28ffe0b52bde..1f74de25a92f 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -42,10 +42,8 @@ #include #include -#include -#include -#include -#include +#include +#include #include "target_core_internal.h" #include "target_core_alua.h" diff --git a/drivers/target/target_core_fabric_configfs.c b/drivers/target/target_core_fabric_configfs.c index fa1e4eb01bc7..4f77cce22646 100644 --- a/drivers/target/target_core_fabric_configfs.c +++ b/drivers/target/target_core_fabric_configfs.c @@ -36,10 +36,7 @@ #include #include -#include -#include -#include -#include +#include #include #include #include diff --git a/drivers/target/target_core_fabric_lib.c b/drivers/target/target_core_fabric_lib.c index dd9ae929dda2..32d7feb405ec 100644 --- a/drivers/target/target_core_fabric_lib.c +++ b/drivers/target/target_core_fabric_lib.c @@ -34,10 +34,7 @@ #include #include -#include -#include -#include -#include +#include #include #include "target_core_internal.h" diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c index b4864fba4ef0..ed7e05353d62 100644 --- a/drivers/target/target_core_file.c +++ b/drivers/target/target_core_file.c @@ -37,8 +37,7 @@ #include #include -#include -#include +#include #include "target_core_file.h" diff --git a/drivers/target/target_core_hba.c b/drivers/target/target_core_hba.c index 26c1238976cb..3dd1bd4b6f71 100644 --- a/drivers/target/target_core_hba.c +++ b/drivers/target/target_core_hba.c @@ -37,9 +37,8 @@ #include #include -#include -#include -#include +#include +#include #include "target_core_internal.h" diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c index 4aa992204438..feebbfe6cbd0 100644 --- a/drivers/target/target_core_iblock.c +++ b/drivers/target/target_core_iblock.c @@ -42,8 +42,7 @@ #include #include -#include -#include +#include #include "target_core_iblock.h" diff --git a/drivers/target/target_core_pr.c b/drivers/target/target_core_pr.c index 36fa9a5f1ba6..d2b73b3f4e6c 100644 --- a/drivers/target/target_core_pr.c +++ b/drivers/target/target_core_pr.c @@ -33,11 +33,8 @@ #include #include -#include -#include -#include -#include -#include +#include +#include #include #include "target_core_internal.h" diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c index 8b15e56b0384..5d8851de3aee 100644 --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -44,8 +44,7 @@ #include #include -#include -#include +#include #include "target_core_pscsi.h" diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c index 02e51faa2f4e..94b4087dadd5 100644 --- a/drivers/target/target_core_rd.c +++ b/drivers/target/target_core_rd.c @@ -37,9 +37,7 @@ #include #include -#include -#include -#include +#include #include "target_core_rd.h" diff --git a/drivers/target/target_core_stat.c b/drivers/target/target_core_stat.c index d2ba9266ff4d..6d8a90464f2e 100644 --- a/drivers/target/target_core_stat.c +++ b/drivers/target/target_core_stat.c @@ -43,8 +43,8 @@ #include #include -#include -#include +#include +#include #include #include diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c index 47bc26ba7865..fc9d74890319 100644 --- a/drivers/target/target_core_tmr.c +++ b/drivers/target/target_core_tmr.c @@ -32,10 +32,8 @@ #include #include -#include -#include -#include -#include +#include +#include #include #include "target_core_internal.h" diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c index 0a936fae3d23..b7668029bb31 100644 --- a/drivers/target/target_core_tpg.c +++ b/drivers/target/target_core_tpg.c @@ -39,10 +39,8 @@ #include #include -#include -#include -#include -#include +#include +#include #include "target_core_internal.h" diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index e0ea63dedde3..72c13b7a349f 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -45,11 +45,8 @@ #include #include -#include -#include -#include -#include -#include +#include +#include #include #include "target_core_internal.h" diff --git a/drivers/target/target_core_ua.c b/drivers/target/target_core_ua.c index e2ae53c7e088..3e12f6bcfa10 100644 --- a/drivers/target/target_core_ua.c +++ b/drivers/target/target_core_ua.c @@ -30,9 +30,7 @@ #include #include -#include -#include -#include +#include #include #include "target_core_internal.h" diff --git a/drivers/target/tcm_fc/tfc_cmd.c b/drivers/target/tcm_fc/tfc_cmd.c index 71fc9cea5dc9..dbbbc8376d01 100644 --- a/drivers/target/tcm_fc/tfc_cmd.c +++ b/drivers/target/tcm_fc/tfc_cmd.c @@ -39,12 +39,8 @@ #include #include -#include -#include -#include -#include +#include #include -#include #include #include "tcm_fc.h" diff --git a/drivers/target/tcm_fc/tfc_conf.c b/drivers/target/tcm_fc/tfc_conf.c index 9402b7387cac..73852fbc857b 100644 --- a/drivers/target/tcm_fc/tfc_conf.c +++ b/drivers/target/tcm_fc/tfc_conf.c @@ -41,12 +41,8 @@ #include #include -#include -#include +#include #include -#include -#include -#include #include #include diff --git a/drivers/target/tcm_fc/tfc_io.c b/drivers/target/tcm_fc/tfc_io.c index 1369b1cb103d..d8cabc21036d 100644 --- a/drivers/target/tcm_fc/tfc_io.c +++ b/drivers/target/tcm_fc/tfc_io.c @@ -48,10 +48,7 @@ #include #include -#include -#include -#include -#include +#include #include #include diff --git a/drivers/target/tcm_fc/tfc_sess.c b/drivers/target/tcm_fc/tfc_sess.c index 326921385aff..4c0507cf808c 100644 --- a/drivers/target/tcm_fc/tfc_sess.c +++ b/drivers/target/tcm_fc/tfc_sess.c @@ -40,10 +40,7 @@ #include #include -#include -#include -#include -#include +#include #include #include diff --git a/include/target/target_core_backend.h b/include/target/target_core_backend.h new file mode 100644 index 000000000000..4866499bdeeb --- /dev/null +++ b/include/target/target_core_backend.h @@ -0,0 +1,65 @@ +#ifndef TARGET_CORE_BACKEND_H +#define TARGET_CORE_BACKEND_H + +#define TRANSPORT_PLUGIN_PHBA_PDEV 1 +#define TRANSPORT_PLUGIN_VHBA_PDEV 2 +#define TRANSPORT_PLUGIN_VHBA_VDEV 3 + +struct se_subsystem_api { + struct list_head sub_api_list; + + char name[16]; + struct module *owner; + + u8 transport_type; + + unsigned int fua_write_emulated : 1; + unsigned int write_cache_emulated : 1; + + int (*attach_hba)(struct se_hba *, u32); + void (*detach_hba)(struct se_hba *); + int (*pmode_enable_hba)(struct se_hba *, unsigned long); + void *(*allocate_virtdevice)(struct se_hba *, const char *); + struct se_device *(*create_virtdevice)(struct se_hba *, + struct se_subsystem_dev *, void *); + void (*free_device)(void *); + int (*transport_complete)(struct se_task *task); + struct se_task *(*alloc_task)(unsigned char *cdb); + int (*do_task)(struct se_task *); + int (*do_discard)(struct se_device *, sector_t, u32); + void (*do_sync_cache)(struct se_task *); + void (*free_task)(struct se_task *); + ssize_t (*check_configfs_dev_params)(struct se_hba *, + struct se_subsystem_dev *); + ssize_t (*set_configfs_dev_params)(struct se_hba *, + struct se_subsystem_dev *, const char *, ssize_t); + ssize_t (*show_configfs_dev_params)(struct se_hba *, + struct se_subsystem_dev *, char *); + u32 (*get_device_rev)(struct se_device *); + u32 (*get_device_type)(struct se_device *); + sector_t (*get_blocks)(struct se_device *); + unsigned char *(*get_sense_buffer)(struct se_task *); +}; + +int transport_subsystem_register(struct se_subsystem_api *); +void transport_subsystem_release(struct se_subsystem_api *); + +struct se_device *transport_add_device_to_core_hba(struct se_hba *, + struct se_subsystem_api *, struct se_subsystem_dev *, u32, + void *, struct se_dev_limits *, const char *, const char *); + +void transport_complete_sync_cache(struct se_cmd *, int); +void transport_complete_task(struct se_task *, int); + +void target_get_task_cdb(struct se_task *, unsigned char *); + +void transport_set_vpd_proto_id(struct t10_vpd *, unsigned char *); +int transport_set_vpd_assoc(struct t10_vpd *, unsigned char *); +int transport_set_vpd_ident_type(struct t10_vpd *, unsigned char *); +int transport_set_vpd_ident(struct t10_vpd *, unsigned char *); + +/* core helpers also used by command snooping in pscsi */ +void *transport_kmap_first_data_page(struct se_cmd *); +void transport_kunmap_first_data_page(struct se_cmd *); + +#endif /* TARGET_CORE_BACKEND_H */ diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 6873c7dd9145..1d90fb33e60b 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -10,6 +10,7 @@ #include #define TARGET_CORE_MOD_VERSION "v4.1.0-rc1-ml" +#define TARGET_CORE_VERSION TARGET_CORE_MOD_VERSION /* Maximum Number of LUNs per Target Portal Group */ /* Don't raise above 511 or REPORT_LUNS needs to handle >1 page */ @@ -53,6 +54,72 @@ /* Used by transport_get_inquiry_vpd_device_ident() */ #define INQUIRY_VPD_DEVICE_IDENTIFIER_LEN 254 +/* Attempts before moving from SHORT to LONG */ +#define PYX_TRANSPORT_WINDOW_CLOSED_THRESHOLD 3 +#define PYX_TRANSPORT_WINDOW_CLOSED_WAIT_SHORT 3 /* In milliseconds */ +#define PYX_TRANSPORT_WINDOW_CLOSED_WAIT_LONG 10 /* In milliseconds */ + +#define PYX_TRANSPORT_STATUS_INTERVAL 5 /* In seconds */ + +/* + * struct se_subsystem_dev->su_dev_flags +*/ +#define SDF_FIRMWARE_VPD_UNIT_SERIAL 0x00000001 +#define SDF_EMULATED_VPD_UNIT_SERIAL 0x00000002 +#define SDF_USING_UDEV_PATH 0x00000004 +#define SDF_USING_ALIAS 0x00000008 + +/* + * struct se_device->dev_flags + */ +#define DF_READ_ONLY 0x00000001 +#define DF_SPC2_RESERVATIONS 0x00000002 +#define DF_SPC2_RESERVATIONS_WITH_ISID 0x00000004 + +/* struct se_dev_attrib sanity values */ +/* Default max_unmap_lba_count */ +#define DA_MAX_UNMAP_LBA_COUNT 0 +/* Default max_unmap_block_desc_count */ +#define DA_MAX_UNMAP_BLOCK_DESC_COUNT 0 +/* Default unmap_granularity */ +#define DA_UNMAP_GRANULARITY_DEFAULT 0 +/* Default unmap_granularity_alignment */ +#define DA_UNMAP_GRANULARITY_ALIGNMENT_DEFAULT 0 +/* Emulation for Direct Page Out */ +#define DA_EMULATE_DPO 0 +/* Emulation for Forced Unit Access WRITEs */ +#define DA_EMULATE_FUA_WRITE 1 +/* Emulation for Forced Unit Access READs */ +#define DA_EMULATE_FUA_READ 0 +/* Emulation for WriteCache and SYNCHRONIZE_CACHE */ +#define DA_EMULATE_WRITE_CACHE 0 +/* Emulation for UNIT ATTENTION Interlock Control */ +#define DA_EMULATE_UA_INTLLCK_CTRL 0 +/* Emulation for TASK_ABORTED status (TAS) by default */ +#define DA_EMULATE_TAS 1 +/* Emulation for Thin Provisioning UNMAP using block/blk-lib.c:blkdev_issue_discard() */ +#define DA_EMULATE_TPU 0 +/* + * Emulation for Thin Provisioning WRITE_SAME w/ UNMAP=1 bit using + * block/blk-lib.c:blkdev_issue_discard() + */ +#define DA_EMULATE_TPWS 0 +/* No Emulation for PSCSI by default */ +#define DA_EMULATE_RESERVATIONS 0 +/* No Emulation for PSCSI by default */ +#define DA_EMULATE_ALUA 0 +/* Enforce SCSI Initiator Port TransportID with 'ISID' for PR */ +#define DA_ENFORCE_PR_ISIDS 1 +#define DA_STATUS_MAX_SECTORS_MIN 16 +#define DA_STATUS_MAX_SECTORS_MAX 8192 +/* By default don't report non-rotating (solid state) medium */ +#define DA_IS_NONROT 0 +/* Queue Algorithm Modifier default for restricted reordering in control mode page */ +#define DA_EMULATE_REST_REORD 0 + +#define SE_MODE_PAGE_BUF 512 + + /* struct se_hba->hba_flags */ enum hba_flags_table { HBA_FLAGS_INTERNAL_USE = 0x01, @@ -158,6 +225,30 @@ enum tcm_sense_reason_table { TCM_RESERVATION_CONFLICT = 0x10, }; +/* fabric independent task management function values */ +enum tcm_tmreq_table { + TMR_ABORT_TASK = 1, + TMR_ABORT_TASK_SET = 2, + TMR_CLEAR_ACA = 3, + TMR_CLEAR_TASK_SET = 4, + TMR_LUN_RESET = 5, + TMR_TARGET_WARM_RESET = 6, + TMR_TARGET_COLD_RESET = 7, + TMR_FABRIC_TMR = 255, +}; + +/* fabric independent task management response values */ +enum tcm_tmrsp_table { + TMR_FUNCTION_COMPLETE = 0, + TMR_TASK_DOES_NOT_EXIST = 1, + TMR_LUN_DOES_NOT_EXIST = 2, + TMR_TASK_STILL_ALLEGIANT = 3, + TMR_TASK_FAILOVER_NOT_SUPPORTED = 4, + TMR_TASK_MGMT_FUNCTION_NOT_SUPPORTED = 5, + TMR_FUNCTION_AUTHORIZATION_FAILED = 6, + TMR_FUNCTION_REJECTED = 255, +}; + struct se_obj { atomic_t obj_access_count; } ____cacheline_aligned; diff --git a/include/target/target_core_device.h b/include/target/target_core_device.h deleted file mode 100644 index c6123a49014d..000000000000 --- a/include/target/target_core_device.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef TARGET_CORE_DEVICE_H -#define TARGET_CORE_DEVICE_H - - -// external -extern int transport_lookup_cmd_lun(struct se_cmd *, u32); -extern int transport_lookup_tmr_lun(struct se_cmd *, u32); - -#endif /* TARGET_CORE_DEVICE_H */ diff --git a/include/target/target_core_fabric_ops.h b/include/target/target_core_fabric.h similarity index 51% rename from include/target/target_core_fabric_ops.h rename to include/target/target_core_fabric.h index 0256825f923d..d035d865dbef 100644 --- a/include/target/target_core_fabric_ops.h +++ b/include/target/target_core_fabric.h @@ -1,5 +1,5 @@ -/* Defined in target_core_configfs.h */ -struct target_fabric_configfs; +#ifndef TARGET_CORE_FABRIC_H +#define TARGET_CORE_FABRIC_H struct target_core_fabric_ops { struct configfs_subsystem *tf_subsys; @@ -103,3 +103,87 @@ struct target_core_fabric_ops { struct config_group *, const char *); void (*fabric_drop_nodeacl)(struct se_node_acl *); }; + +struct se_session *transport_init_session(void); +void __transport_register_session(struct se_portal_group *, + struct se_node_acl *, struct se_session *, void *); +void transport_register_session(struct se_portal_group *, + struct se_node_acl *, struct se_session *, void *); +void transport_free_session(struct se_session *); +void transport_deregister_session_configfs(struct se_session *); +void transport_deregister_session(struct se_session *); + + +void transport_init_se_cmd(struct se_cmd *, struct target_core_fabric_ops *, + struct se_session *, u32, int, int, unsigned char *); +int transport_lookup_cmd_lun(struct se_cmd *, u32); +int transport_generic_allocate_tasks(struct se_cmd *, unsigned char *); +int transport_handle_cdb_direct(struct se_cmd *); +int transport_generic_handle_cdb_map(struct se_cmd *); +int transport_generic_handle_data(struct se_cmd *); +int transport_generic_map_mem_to_cmd(struct se_cmd *cmd, + struct scatterlist *, u32, struct scatterlist *, u32); +void transport_do_task_sg_chain(struct se_cmd *); +int transport_generic_new_cmd(struct se_cmd *); + +void transport_generic_process_write(struct se_cmd *); + +void transport_generic_free_cmd(struct se_cmd *, int); + +bool transport_wait_for_tasks(struct se_cmd *); +int transport_check_aborted_status(struct se_cmd *, int); +int transport_send_check_condition_and_sense(struct se_cmd *, u8, int); + +void target_get_sess_cmd(struct se_session *, struct se_cmd *); +int target_put_sess_cmd(struct se_session *, struct se_cmd *); +void target_splice_sess_cmd_list(struct se_session *); +void target_wait_for_sess_cmds(struct se_session *, int); + +int core_alua_check_nonop_delay(struct se_cmd *); + +struct se_tmr_req *core_tmr_alloc_req(struct se_cmd *, void *, u8, gfp_t); +void core_tmr_release_req(struct se_tmr_req *); +int transport_generic_handle_tmr(struct se_cmd *); +int transport_lookup_tmr_lun(struct se_cmd *, u32); + +struct se_node_acl *core_tpg_check_initiator_node_acl(struct se_portal_group *, + unsigned char *); +void core_tpg_clear_object_luns(struct se_portal_group *); +struct se_node_acl *core_tpg_add_initiator_node_acl(struct se_portal_group *, + struct se_node_acl *, const char *, u32); +int core_tpg_del_initiator_node_acl(struct se_portal_group *, + struct se_node_acl *, int); +int core_tpg_set_initiator_node_queue_depth(struct se_portal_group *, + unsigned char *, u32, int); +int core_tpg_register(struct target_core_fabric_ops *, struct se_wwn *, + struct se_portal_group *, void *, int); +int core_tpg_deregister(struct se_portal_group *); + +/* SAS helpers */ +u8 sas_get_fabric_proto_ident(struct se_portal_group *); +u32 sas_get_pr_transport_id(struct se_portal_group *, struct se_node_acl *, + struct t10_pr_registration *, int *, unsigned char *); +u32 sas_get_pr_transport_id_len(struct se_portal_group *, struct se_node_acl *, + struct t10_pr_registration *, int *); +char *sas_parse_pr_out_transport_id(struct se_portal_group *, const char *, + u32 *, char **); + +/* FC helpers */ +u8 fc_get_fabric_proto_ident(struct se_portal_group *); +u32 fc_get_pr_transport_id(struct se_portal_group *, struct se_node_acl *, + struct t10_pr_registration *, int *, unsigned char *); +u32 fc_get_pr_transport_id_len(struct se_portal_group *, struct se_node_acl *, + struct t10_pr_registration *, int *); +char *fc_parse_pr_out_transport_id(struct se_portal_group *, const char *, + u32 *, char **); + +/* iSCSI helpers */ +u8 iscsi_get_fabric_proto_ident(struct se_portal_group *); +u32 iscsi_get_pr_transport_id(struct se_portal_group *, struct se_node_acl *, + struct t10_pr_registration *, int *, unsigned char *); +u32 iscsi_get_pr_transport_id_len(struct se_portal_group *, struct se_node_acl *, + struct t10_pr_registration *, int *); +char *iscsi_parse_pr_out_transport_id(struct se_portal_group *, const char *, + u32 *, char **); + +#endif /* TARGET_CORE_FABRICH */ diff --git a/include/target/target_core_fabric_lib.h b/include/target/target_core_fabric_lib.h deleted file mode 100644 index c2f8d0e3a03b..000000000000 --- a/include/target/target_core_fabric_lib.h +++ /dev/null @@ -1,28 +0,0 @@ -#ifndef TARGET_CORE_FABRIC_LIB_H -#define TARGET_CORE_FABRIC_LIB_H - -extern u8 sas_get_fabric_proto_ident(struct se_portal_group *); -extern u32 sas_get_pr_transport_id(struct se_portal_group *, struct se_node_acl *, - struct t10_pr_registration *, int *, unsigned char *); -extern u32 sas_get_pr_transport_id_len(struct se_portal_group *, struct se_node_acl *, - struct t10_pr_registration *, int *); -extern char *sas_parse_pr_out_transport_id(struct se_portal_group *, - const char *, u32 *, char **); - -extern u8 fc_get_fabric_proto_ident(struct se_portal_group *); -extern u32 fc_get_pr_transport_id(struct se_portal_group *, struct se_node_acl *, - struct t10_pr_registration *, int *, unsigned char *); -extern u32 fc_get_pr_transport_id_len(struct se_portal_group *, struct se_node_acl *, - struct t10_pr_registration *, int *); -extern char *fc_parse_pr_out_transport_id(struct se_portal_group *, - const char *, u32 *, char **); - -extern u8 iscsi_get_fabric_proto_ident(struct se_portal_group *); -extern u32 iscsi_get_pr_transport_id(struct se_portal_group *, struct se_node_acl *, - struct t10_pr_registration *, int *, unsigned char *); -extern u32 iscsi_get_pr_transport_id_len(struct se_portal_group *, struct se_node_acl *, - struct t10_pr_registration *, int *); -extern char *iscsi_parse_pr_out_transport_id(struct se_portal_group *, - const char *, u32 *, char **); - -#endif /* TARGET_CORE_FABRIC_LIB_H */ diff --git a/include/target/target_core_tmr.h b/include/target/target_core_tmr.h deleted file mode 100644 index 36b3e0de99b4..000000000000 --- a/include/target/target_core_tmr.h +++ /dev/null @@ -1,31 +0,0 @@ -#ifndef TARGET_CORE_TMR_H -#define TARGET_CORE_TMR_H - -/* fabric independent task management function values */ -enum tcm_tmreq_table { - TMR_ABORT_TASK = 1, - TMR_ABORT_TASK_SET = 2, - TMR_CLEAR_ACA = 3, - TMR_CLEAR_TASK_SET = 4, - TMR_LUN_RESET = 5, - TMR_TARGET_WARM_RESET = 6, - TMR_TARGET_COLD_RESET = 7, - TMR_FABRIC_TMR = 255, -}; - -/* fabric independent task management response values */ -enum tcm_tmrsp_table { - TMR_FUNCTION_COMPLETE = 0, - TMR_TASK_DOES_NOT_EXIST = 1, - TMR_LUN_DOES_NOT_EXIST = 2, - TMR_TASK_STILL_ALLEGIANT = 3, - TMR_TASK_FAILOVER_NOT_SUPPORTED = 4, - TMR_TASK_MGMT_FUNCTION_NOT_SUPPORTED = 5, - TMR_FUNCTION_AUTHORIZATION_FAILED = 6, - TMR_FUNCTION_REJECTED = 255, -}; - -extern struct se_tmr_req *core_tmr_alloc_req(struct se_cmd *, void *, u8, gfp_t); -extern void core_tmr_release_req(struct se_tmr_req *); - -#endif /* TARGET_CORE_TMR_H */ diff --git a/include/target/target_core_tpg.h b/include/target/target_core_tpg.h deleted file mode 100644 index 80ae59fdada3..000000000000 --- a/include/target/target_core_tpg.h +++ /dev/null @@ -1,22 +0,0 @@ -#ifndef TARGET_CORE_TPG_H -#define TARGET_CORE_TPG_H - -extern struct se_node_acl *core_tpg_check_initiator_node_acl( - struct se_portal_group *, - unsigned char *); -extern void core_tpg_clear_object_luns(struct se_portal_group *); -extern struct se_node_acl *core_tpg_add_initiator_node_acl( - struct se_portal_group *, - struct se_node_acl *, - const char *, u32); -extern int core_tpg_del_initiator_node_acl(struct se_portal_group *, - struct se_node_acl *, int); -extern int core_tpg_set_initiator_node_queue_depth(struct se_portal_group *, - unsigned char *, u32, int); -extern int core_tpg_register(struct target_core_fabric_ops *, - struct se_wwn *, - struct se_portal_group *, void *, - int); -extern int core_tpg_deregister(struct se_portal_group *); - -#endif /* TARGET_CORE_TPG_H */ diff --git a/include/target/target_core_transport.h b/include/target/target_core_transport.h deleted file mode 100644 index 68fe1f2c081c..000000000000 --- a/include/target/target_core_transport.h +++ /dev/null @@ -1,256 +0,0 @@ -#ifndef TARGET_CORE_TRANSPORT_H -#define TARGET_CORE_TRANSPORT_H - -#define TARGET_CORE_VERSION TARGET_CORE_MOD_VERSION - -/* Attempts before moving from SHORT to LONG */ -#define PYX_TRANSPORT_WINDOW_CLOSED_THRESHOLD 3 -#define PYX_TRANSPORT_WINDOW_CLOSED_WAIT_SHORT 3 /* In milliseconds */ -#define PYX_TRANSPORT_WINDOW_CLOSED_WAIT_LONG 10 /* In milliseconds */ - -#define PYX_TRANSPORT_STATUS_INTERVAL 5 /* In seconds */ - -#define TRANSPORT_PLUGIN_PHBA_PDEV 1 -#define TRANSPORT_PLUGIN_VHBA_PDEV 2 -#define TRANSPORT_PLUGIN_VHBA_VDEV 3 - -/* - * struct se_subsystem_dev->su_dev_flags -*/ -#define SDF_FIRMWARE_VPD_UNIT_SERIAL 0x00000001 -#define SDF_EMULATED_VPD_UNIT_SERIAL 0x00000002 -#define SDF_USING_UDEV_PATH 0x00000004 -#define SDF_USING_ALIAS 0x00000008 - -/* - * struct se_device->dev_flags - */ -#define DF_READ_ONLY 0x00000001 -#define DF_SPC2_RESERVATIONS 0x00000002 -#define DF_SPC2_RESERVATIONS_WITH_ISID 0x00000004 - -/* struct se_dev_attrib sanity values */ -/* Default max_unmap_lba_count */ -#define DA_MAX_UNMAP_LBA_COUNT 0 -/* Default max_unmap_block_desc_count */ -#define DA_MAX_UNMAP_BLOCK_DESC_COUNT 0 -/* Default unmap_granularity */ -#define DA_UNMAP_GRANULARITY_DEFAULT 0 -/* Default unmap_granularity_alignment */ -#define DA_UNMAP_GRANULARITY_ALIGNMENT_DEFAULT 0 -/* Emulation for Direct Page Out */ -#define DA_EMULATE_DPO 0 -/* Emulation for Forced Unit Access WRITEs */ -#define DA_EMULATE_FUA_WRITE 1 -/* Emulation for Forced Unit Access READs */ -#define DA_EMULATE_FUA_READ 0 -/* Emulation for WriteCache and SYNCHRONIZE_CACHE */ -#define DA_EMULATE_WRITE_CACHE 0 -/* Emulation for UNIT ATTENTION Interlock Control */ -#define DA_EMULATE_UA_INTLLCK_CTRL 0 -/* Emulation for TASK_ABORTED status (TAS) by default */ -#define DA_EMULATE_TAS 1 -/* Emulation for Thin Provisioning UNMAP using block/blk-lib.c:blkdev_issue_discard() */ -#define DA_EMULATE_TPU 0 -/* - * Emulation for Thin Provisioning WRITE_SAME w/ UNMAP=1 bit using - * block/blk-lib.c:blkdev_issue_discard() - */ -#define DA_EMULATE_TPWS 0 -/* No Emulation for PSCSI by default */ -#define DA_EMULATE_RESERVATIONS 0 -/* No Emulation for PSCSI by default */ -#define DA_EMULATE_ALUA 0 -/* Enforce SCSI Initiator Port TransportID with 'ISID' for PR */ -#define DA_ENFORCE_PR_ISIDS 1 -#define DA_STATUS_MAX_SECTORS_MIN 16 -#define DA_STATUS_MAX_SECTORS_MAX 8192 -/* By default don't report non-rotating (solid state) medium */ -#define DA_IS_NONROT 0 -/* Queue Algorithm Modifier default for restricted reordering in control mode page */ -#define DA_EMULATE_REST_REORD 0 - -#define SE_MODE_PAGE_BUF 512 - -#define MOD_MAX_SECTORS(ms, bs) (ms % (PAGE_SIZE / bs)) - -struct se_subsystem_api; - -extern int transport_subsystem_register(struct se_subsystem_api *); -extern void transport_subsystem_release(struct se_subsystem_api *); -extern struct se_session *transport_init_session(void); -extern void __transport_register_session(struct se_portal_group *, - struct se_node_acl *, - struct se_session *, void *); -extern void transport_register_session(struct se_portal_group *, - struct se_node_acl *, - struct se_session *, void *); -extern void transport_free_session(struct se_session *); -extern void transport_deregister_session_configfs(struct se_session *); -extern void transport_deregister_session(struct se_session *); -extern void transport_complete_sync_cache(struct se_cmd *, int); -extern void transport_complete_task(struct se_task *, int); - -extern void transport_set_vpd_proto_id(struct t10_vpd *, unsigned char *); -extern int transport_set_vpd_assoc(struct t10_vpd *, unsigned char *); -extern int transport_set_vpd_ident_type(struct t10_vpd *, unsigned char *); -extern int transport_set_vpd_ident(struct t10_vpd *, unsigned char *); - -extern struct se_device *transport_add_device_to_core_hba(struct se_hba *, - struct se_subsystem_api *, - struct se_subsystem_dev *, u32, - void *, struct se_dev_limits *, - const char *, const char *); -extern void transport_init_se_cmd(struct se_cmd *, - struct target_core_fabric_ops *, - struct se_session *, u32, int, int, - unsigned char *); -void *transport_kmap_first_data_page(struct se_cmd *cmd); -void transport_kunmap_first_data_page(struct se_cmd *cmd); -extern int transport_generic_allocate_tasks(struct se_cmd *, unsigned char *); -extern int transport_handle_cdb_direct(struct se_cmd *); -extern int transport_generic_handle_cdb_map(struct se_cmd *); -extern int transport_generic_handle_data(struct se_cmd *); -extern int transport_generic_handle_tmr(struct se_cmd *); -extern int transport_generic_map_mem_to_cmd(struct se_cmd *cmd, struct scatterlist *, u32, - struct scatterlist *, u32); -extern bool transport_wait_for_tasks(struct se_cmd *); -extern int transport_check_aborted_status(struct se_cmd *, int); -extern int transport_send_check_condition_and_sense(struct se_cmd *, u8, int); -extern void transport_generic_free_cmd(struct se_cmd *, int); -extern void target_get_sess_cmd(struct se_session *, struct se_cmd *); -extern int target_put_sess_cmd(struct se_session *, struct se_cmd *); -extern void target_splice_sess_cmd_list(struct se_session *); -extern void target_wait_for_sess_cmds(struct se_session *, int); -extern void transport_do_task_sg_chain(struct se_cmd *); -extern void transport_generic_process_write(struct se_cmd *); -extern int transport_generic_new_cmd(struct se_cmd *); -/* From target_core_alua.c */ -extern int core_alua_check_nonop_delay(struct se_cmd *); -/* From target_core_cdb.c */ -extern void target_get_task_cdb(struct se_task *task, unsigned char *cdb); - -/* - * Each se_transport_task_t can have N number of possible struct se_task's - * for the storage transport(s) to possibly execute. - * Used primarily for splitting up CDBs that exceed the physical storage - * HBA's maximum sector count per task. - */ -struct se_mem { - struct page *se_page; - u32 se_len; - u32 se_off; - struct list_head se_list; -} ____cacheline_aligned; - -/* - * Each type of disk transport supported MUST have a template defined - * within its .h file. - */ -struct se_subsystem_api { - /* - * The Name. :-) - */ - char name[16]; - /* - * Transport Type. - */ - u8 transport_type; - - unsigned int fua_write_emulated : 1; - unsigned int write_cache_emulated : 1; - - /* - * struct module for struct se_hba references - */ - struct module *owner; - /* - * Used for global se_subsystem_api list_head - */ - struct list_head sub_api_list; - /* - * attach_hba(): - */ - int (*attach_hba)(struct se_hba *, u32); - /* - * detach_hba(): - */ - void (*detach_hba)(struct se_hba *); - /* - * pmode_hba(): Used for TCM/pSCSI subsystem plugin HBA -> - * Linux/SCSI struct Scsi_Host passthrough - */ - int (*pmode_enable_hba)(struct se_hba *, unsigned long); - /* - * allocate_virtdevice(): - */ - void *(*allocate_virtdevice)(struct se_hba *, const char *); - /* - * create_virtdevice(): Only for Virtual HBAs - */ - struct se_device *(*create_virtdevice)(struct se_hba *, - struct se_subsystem_dev *, void *); - /* - * free_device(): - */ - void (*free_device)(void *); - - /* - * transport_complete(): - * - * Use transport_generic_complete() for majority of DAS transport - * drivers. Provided out of convenience. - */ - int (*transport_complete)(struct se_task *task); - struct se_task *(*alloc_task)(unsigned char *cdb); - /* - * do_task(): - */ - int (*do_task)(struct se_task *); - /* - * Used by virtual subsystem plugins IBLOCK and FILEIO to emulate - * UNMAP and WRITE_SAME_* w/ UNMAP=1 <-> Linux/Block Discard - */ - int (*do_discard)(struct se_device *, sector_t, u32); - /* - * Used by virtual subsystem plugins IBLOCK and FILEIO to emulate - * SYNCHRONIZE_CACHE_* <-> Linux/Block blkdev_issue_flush() - */ - void (*do_sync_cache)(struct se_task *); - /* - * free_task(): - */ - void (*free_task)(struct se_task *); - /* - * check_configfs_dev_params(): - */ - ssize_t (*check_configfs_dev_params)(struct se_hba *, struct se_subsystem_dev *); - /* - * set_configfs_dev_params(): - */ - ssize_t (*set_configfs_dev_params)(struct se_hba *, struct se_subsystem_dev *, - const char *, ssize_t); - /* - * show_configfs_dev_params(): - */ - ssize_t (*show_configfs_dev_params)(struct se_hba *, struct se_subsystem_dev *, - char *); - /* - * get_device_rev(): - */ - u32 (*get_device_rev)(struct se_device *); - /* - * get_device_type(): - */ - u32 (*get_device_type)(struct se_device *); - /* - * Get the sector_t from a subsystem backstore.. - */ - sector_t (*get_blocks)(struct se_device *); - /* - * get_sense_buffer(): - */ - unsigned char *(*get_sense_buffer)(struct se_task *); -} ____cacheline_aligned; - -#endif /* TARGET_CORE_TRANSPORT_H */ From ef804a849ffae058a783e9dddd24cc1a555bbdb4 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 23 Nov 2011 06:53:58 -0500 Subject: [PATCH 140/820] target: fold se_task.task_sense into task_flags Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 4 ++-- include/target/target_core_base.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 72c13b7a349f..e13fafa6f526 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -709,7 +709,7 @@ void transport_complete_task(struct se_task *task, int success) if (dev && dev->transport->transport_complete) { if (dev->transport->transport_complete(task) != 0) { cmd->se_cmd_flags |= SCF_TRANSPORT_TASK_SENSE; - task->task_sense = 1; + task->task_flags |= TF_HAS_SENSE; success = 1; } } @@ -2346,7 +2346,7 @@ static int transport_get_sense_data(struct se_cmd *cmd) list_for_each_entry_safe(task, task_tmp, &cmd->t_task_list, t_list) { - if (!task->task_sense) + if (!(task->task_flags & TF_HAS_SENSE)) continue; if (!dev->transport->get_sense_buffer) { diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 1d90fb33e60b..7f3acdde97de 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -138,11 +138,12 @@ enum transport_tpg_type_table { TRANSPORT_TPG_TYPE_DISCOVERY = 1, }; -/* Used for generate timer flags */ +/* struct se_task->task_flags */ enum se_task_flags { TF_ACTIVE = (1 << 0), TF_SENT = (1 << 1), TF_REQUEST_STOP = (1 << 2), + TF_HAS_SENSE = (1 << 3), }; /* Special transport agnostic struct se_cmd->t_states */ @@ -488,7 +489,6 @@ struct se_task { struct scatterlist *task_sg; u32 task_sg_nents; u16 task_flags; - u8 task_sense; u8 task_scsi_status; int task_error_status; enum dma_data_direction task_data_direction; From 41e16e981679124c78c30f046d4f0b71d86ff1b2 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 23 Nov 2011 06:54:15 -0500 Subject: [PATCH 141/820] target: remove the se_task task_error_status field We only reach transport_complete_task once per task, so the test and set on task_error_status is never going to have an effect. Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 8 +------- include/target/target_core_base.h | 1 - 2 files changed, 1 insertion(+), 8 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index e13fafa6f526..ae112ac91c47 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -738,13 +738,7 @@ void transport_complete_task(struct se_task *task, int success) } if (cmd->t_tasks_failed) { - if (!task->task_error_status) { - task->task_error_status = - TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE; - cmd->scsi_sense_reason = - TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE; - } - + cmd->scsi_sense_reason = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE; INIT_WORK(&cmd->work, target_complete_failure_work); } else { atomic_set(&cmd->t_transport_complete, 1); diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 7f3acdde97de..e2473e6770b3 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -490,7 +490,6 @@ struct se_task { u32 task_sg_nents; u16 task_flags; u8 task_scsi_status; - int task_error_status; enum dma_data_direction task_data_direction; atomic_t task_state_active; struct list_head t_list; From 1880807adb21d741f08b747956c90bf4a6f95fbf Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 23 Nov 2011 06:54:36 -0500 Subject: [PATCH 142/820] target: make the se_task task_state_active a normal bool There is no need to make task_state_active an atomic_t given that it is always set under the execute_task_lock so we can make it a simple bool. Also rename it to t_state_active to be closer to the list it guards, and make sure all checks before the list addion/removal actually happen under execute_task_lock. Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_tmr.c | 2 +- drivers/target/target_core_transport.c | 39 +++++++++++++------------- include/target/target_core_base.h | 2 +- 3 files changed, 21 insertions(+), 22 deletions(-) diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c index fc9d74890319..b4c9bb783c46 100644 --- a/drivers/target/target_core_tmr.c +++ b/drivers/target/target_core_tmr.c @@ -221,7 +221,7 @@ static void core_tmr_drain_task_list( continue; list_move_tail(&task->t_state_list, &drain_task_list); - atomic_set(&task->task_state_active, 0); + task->t_state_active = false; /* * Remove from task execute list before processing drain_task_list */ diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index ae112ac91c47..511836eb2953 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -421,18 +421,18 @@ static void transport_all_task_dev_remove_state(struct se_cmd *cmd) if (task->task_flags & TF_ACTIVE) continue; - if (!atomic_read(&task->task_state_active)) - continue; - spin_lock_irqsave(&dev->execute_task_lock, flags); - list_del(&task->t_state_list); - pr_debug("Removed ITT: 0x%08x dev: %p task[%p]\n", - cmd->se_tfo->get_task_tag(cmd), dev, task); - spin_unlock_irqrestore(&dev->execute_task_lock, flags); + if (task->t_state_active) { + pr_debug("Removed ITT: 0x%08x dev: %p task[%p]\n", + cmd->se_tfo->get_task_tag(cmd), dev, task); - atomic_set(&task->task_state_active, 0); - atomic_dec(&cmd->t_task_cdbs_ex_left); + list_del(&task->t_state_list); + atomic_dec(&cmd->t_task_cdbs_ex_left); + task->t_state_active = false; + } + spin_unlock_irqrestore(&dev->execute_task_lock, flags); } + } /* transport_cmd_check_stop(): @@ -813,7 +813,7 @@ static void __transport_add_task_to_execute_queue( head_of_queue = transport_add_task_check_sam_attr(task, task_prev, dev); atomic_inc(&dev->execute_tasks); - if (atomic_read(&task->task_state_active)) + if (task->t_state_active) return; /* * Determine if this task needs to go to HEAD_OF_QUEUE for the @@ -827,7 +827,7 @@ static void __transport_add_task_to_execute_queue( else list_add_tail(&task->t_state_list, &dev->state_task_list); - atomic_set(&task->task_state_active, 1); + task->t_state_active = true; pr_debug("Added ITT: 0x%08x task[%p] to dev: %p\n", task->task_se_cmd->se_tfo->get_task_tag(task->task_se_cmd), @@ -842,17 +842,16 @@ static void transport_add_tasks_to_state_queue(struct se_cmd *cmd) spin_lock_irqsave(&cmd->t_state_lock, flags); list_for_each_entry(task, &cmd->t_task_list, t_list) { - if (atomic_read(&task->task_state_active)) - continue; - spin_lock(&dev->execute_task_lock); - list_add_tail(&task->t_state_list, &dev->state_task_list); - atomic_set(&task->task_state_active, 1); - - pr_debug("Added ITT: 0x%08x task[%p] to dev: %p\n", - task->task_se_cmd->se_tfo->get_task_tag( - task->task_se_cmd), task, dev); + if (!task->t_state_active) { + list_add_tail(&task->t_state_list, + &dev->state_task_list); + task->t_state_active = true; + pr_debug("Added ITT: 0x%08x task[%p] to dev: %p\n", + task->task_se_cmd->se_tfo->get_task_tag( + task->task_se_cmd), task, dev); + } spin_unlock(&dev->execute_task_lock); } spin_unlock_irqrestore(&cmd->t_state_lock, flags); diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index e2473e6770b3..311f5fc607e6 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -491,10 +491,10 @@ struct se_task { u16 task_flags; u8 task_scsi_status; enum dma_data_direction task_data_direction; - atomic_t task_state_active; struct list_head t_list; struct list_head t_execute_list; struct list_head t_state_list; + bool t_state_active; struct completion task_stop_comp; } ____cacheline_aligned; From 90c161b643d9531d271110876a14e68b49172d8a Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Wed, 23 Nov 2011 20:53:17 +0100 Subject: [PATCH 143/820] target: use \n as a separator for configuration The command | echo rd_pages=32768 > ramdisk/control Does not work because it writes "rd_pages=32768\n" and the parser which matches for "rd_pages=%d" does not recognize it due to the \n. One way of fixing this would be using "echo -n" instead. This patch adds \n to the list of separators so we don't have to use the -n argument which I find is more convinient. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_configfs.c | 2 +- drivers/target/target_core_file.c | 2 +- drivers/target/target_core_iblock.c | 2 +- drivers/target/target_core_pscsi.c | 2 +- drivers/target/target_core_rd.c | 2 +- 5 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index 3c358907a4a6..37b2b3d03c65 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -1450,7 +1450,7 @@ static ssize_t target_core_dev_pr_store_attr_res_aptpl_metadata( return -ENOMEM; orig = opts; - while ((ptr = strsep(&opts, ",")) != NULL) { + while ((ptr = strsep(&opts, ",\n")) != NULL) { if (!*ptr) continue; diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c index ed7e05353d62..91bf11a0ab16 100644 --- a/drivers/target/target_core_file.c +++ b/drivers/target/target_core_file.c @@ -497,7 +497,7 @@ static ssize_t fd_set_configfs_dev_params( orig = opts; - while ((ptr = strsep(&opts, ",")) != NULL) { + while ((ptr = strsep(&opts, ",\n")) != NULL) { if (!*ptr) continue; diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c index feebbfe6cbd0..3c1a7aa658a2 100644 --- a/drivers/target/target_core_iblock.c +++ b/drivers/target/target_core_iblock.c @@ -390,7 +390,7 @@ static ssize_t iblock_set_configfs_dev_params(struct se_hba *hba, orig = opts; - while ((ptr = strsep(&opts, ",")) != NULL) { + while ((ptr = strsep(&opts, ",\n")) != NULL) { if (!*ptr) continue; diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c index 5d8851de3aee..bb44ed4d0272 100644 --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -817,7 +817,7 @@ static ssize_t pscsi_set_configfs_dev_params(struct se_hba *hba, orig = opts; - while ((ptr = strsep(&opts, ",")) != NULL) { + while ((ptr = strsep(&opts, ",\n")) != NULL) { if (!*ptr) continue; diff --git a/drivers/target/target_core_rd.c b/drivers/target/target_core_rd.c index 94b4087dadd5..8b68f7b82631 100644 --- a/drivers/target/target_core_rd.c +++ b/drivers/target/target_core_rd.c @@ -472,7 +472,7 @@ static ssize_t rd_set_configfs_dev_params( orig = opts; - while ((ptr = strsep(&opts, ",")) != NULL) { + while ((ptr = strsep(&opts, ",\n")) != NULL) { if (!*ptr) continue; From c165f69c2ccbbb4ecf16c5d29d65922286bf62ff Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=B6rn=20Engel?= Date: Thu, 24 Nov 2011 02:04:42 +0100 Subject: [PATCH 144/820] target: Move core_scsi3_check_cdb_abort_and_preempt And make it static afterwards. Signed-off-by: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_pr.c | 15 --------------- drivers/target/target_core_pr.h | 2 -- drivers/target/target_core_tmr.c | 15 +++++++++++++++ 3 files changed, 15 insertions(+), 17 deletions(-) diff --git a/drivers/target/target_core_pr.c b/drivers/target/target_core_pr.c index d2b73b3f4e6c..429ad7291664 100644 --- a/drivers/target/target_core_pr.c +++ b/drivers/target/target_core_pr.c @@ -2981,21 +2981,6 @@ static void core_scsi3_release_preempt_and_abort( } } -int core_scsi3_check_cdb_abort_and_preempt( - struct list_head *preempt_and_abort_list, - struct se_cmd *cmd) -{ - struct t10_pr_registration *pr_reg, *pr_reg_tmp; - - list_for_each_entry_safe(pr_reg, pr_reg_tmp, preempt_and_abort_list, - pr_reg_abort_list) { - if (pr_reg->pr_res_key == cmd->pr_res_key) - return 0; - } - - return 1; -} - static int core_scsi3_pro_preempt( struct se_cmd *cmd, int type, diff --git a/drivers/target/target_core_pr.h b/drivers/target/target_core_pr.h index b97f6940dd05..7a233feb7e99 100644 --- a/drivers/target/target_core_pr.h +++ b/drivers/target/target_core_pr.h @@ -60,8 +60,6 @@ extern void core_scsi3_free_pr_reg_from_nacl(struct se_device *, struct se_node_acl *); extern void core_scsi3_free_all_registrations(struct se_device *); extern unsigned char *core_scsi3_pr_dump_type(int); -extern int core_scsi3_check_cdb_abort_and_preempt(struct list_head *, - struct se_cmd *); extern int target_scsi3_emulate_pr_in(struct se_task *task); extern int target_scsi3_emulate_pr_out(struct se_task *task); diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c index b4c9bb783c46..22cefeb544bd 100644 --- a/drivers/target/target_core_tmr.c +++ b/drivers/target/target_core_tmr.c @@ -100,6 +100,21 @@ static void core_tmr_handle_tas_abort( transport_cmd_finish_abort(cmd, 0); } +static int core_scsi3_check_cdb_abort_and_preempt( + struct list_head *preempt_and_abort_list, + struct se_cmd *cmd) +{ + struct t10_pr_registration *pr_reg, *pr_reg_tmp; + + list_for_each_entry_safe(pr_reg, pr_reg_tmp, preempt_and_abort_list, + pr_reg_abort_list) { + if (pr_reg->pr_res_key == cmd->pr_res_key) + return 0; + } + + return 1; +} + static void core_tmr_drain_tmr_list( struct se_device *dev, struct se_tmr_req *tmr, From feae85644f1460b3373ef5141183ee43e6191b58 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=B6rn=20Engel?= Date: Thu, 24 Nov 2011 02:05:12 +0100 Subject: [PATCH 145/820] target: simplify target_check_cdb_and_preempt - rename to target_check_cdb_and_preempt - use non-safe list_for_each_entry - move common check into callee (simplifying callers) Signed-off-by: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_tmr.c | 26 ++++++++++---------------- 1 file changed, 10 insertions(+), 16 deletions(-) diff --git a/drivers/target/target_core_tmr.c b/drivers/target/target_core_tmr.c index 22cefeb544bd..dcb0618c9388 100644 --- a/drivers/target/target_core_tmr.c +++ b/drivers/target/target_core_tmr.c @@ -100,15 +100,15 @@ static void core_tmr_handle_tas_abort( transport_cmd_finish_abort(cmd, 0); } -static int core_scsi3_check_cdb_abort_and_preempt( - struct list_head *preempt_and_abort_list, - struct se_cmd *cmd) +static int target_check_cdb_and_preempt(struct list_head *list, + struct se_cmd *cmd) { - struct t10_pr_registration *pr_reg, *pr_reg_tmp; + struct t10_pr_registration *reg; - list_for_each_entry_safe(pr_reg, pr_reg_tmp, preempt_and_abort_list, - pr_reg_abort_list) { - if (pr_reg->pr_res_key == cmd->pr_res_key) + if (!list) + return 0; + list_for_each_entry(reg, list, pr_reg_abort_list) { + if (reg->pr_res_key == cmd->pr_res_key) return 0; } @@ -146,9 +146,7 @@ static void core_tmr_drain_tmr_list( * parameter (eg: for PROUT PREEMPT_AND_ABORT service action * skip non regisration key matching TMRs. */ - if (preempt_and_abort_list && - (core_scsi3_check_cdb_abort_and_preempt( - preempt_and_abort_list, cmd) != 0)) + if (target_check_cdb_and_preempt(preempt_and_abort_list, cmd)) continue; spin_lock(&cmd->t_state_lock); @@ -225,9 +223,7 @@ static void core_tmr_drain_task_list( * For PREEMPT_AND_ABORT usage, only process commands * with a matching reservation key. */ - if (preempt_and_abort_list && - (core_scsi3_check_cdb_abort_and_preempt( - preempt_and_abort_list, cmd) != 0)) + if (target_check_cdb_and_preempt(preempt_and_abort_list, cmd)) continue; /* * Not aborting PROUT PREEMPT_AND_ABORT CDB.. @@ -335,9 +331,7 @@ static void core_tmr_drain_cmd_list( * For PREEMPT_AND_ABORT usage, only process commands * with a matching reservation key. */ - if (preempt_and_abort_list && - (core_scsi3_check_cdb_abort_and_preempt( - preempt_and_abort_list, cmd) != 0)) + if (target_check_cdb_and_preempt(preempt_and_abort_list, cmd)) continue; /* * Not aborting PROUT PREEMPT_AND_ABORT CDB.. From 8359cf43b9dccddeebb0d247146719a14ce6371a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=B6rn=20Engel?= Date: Thu, 24 Nov 2011 02:05:51 +0100 Subject: [PATCH 146/820] target: remove useless casts A reader should spend an extra moment whenever noticing a cast, because either something special is going on that deserves extra attention or, as is all too often the case, the code is wrong. These casts, afaics, have all been useless. They cast a foo* to a foo*, cast a void* to the assigned type, cast a foo* to void*, before assigning it to a void* variable, etc. In a few cases I also removed an additional &...[0], which is equally useless. Lastly I added three FIXMEs where, to the best of my judgement, the code appears to have a bug. It would be good if someone could check these. Signed-off-by: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/iscsi/iscsi_target.c | 16 +++++------ drivers/target/iscsi/iscsi_target_auth.c | 28 +++++++++---------- drivers/target/iscsi/iscsi_target_configfs.c | 5 ++-- drivers/target/iscsi/iscsi_target_login.c | 21 +++++++------- drivers/target/iscsi/iscsi_target_nego.c | 2 +- .../target/iscsi/iscsi_target_nodeattrib.c | 2 +- drivers/target/iscsi/iscsi_target_stat.c | 16 +++++------ drivers/target/iscsi/iscsi_target_tpg.c | 2 +- drivers/target/iscsi/iscsi_target_util.c | 4 +-- drivers/target/loopback/tcm_loop.c | 18 ++++-------- drivers/target/target_core_cdb.c | 20 ++++++------- drivers/target/target_core_configfs.c | 18 ++++++------ drivers/target/target_core_fabric_lib.c | 6 ++-- drivers/target/target_core_file.c | 10 +++---- drivers/target/target_core_iblock.c | 2 +- drivers/target/target_core_pscsi.c | 14 +++++----- drivers/target/target_core_stat.c | 3 +- drivers/target/target_core_transport.c | 4 +-- 18 files changed, 90 insertions(+), 101 deletions(-) diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c index f74a62d46eea..ac44af165b27 100644 --- a/drivers/target/iscsi/iscsi_target.c +++ b/drivers/target/iscsi/iscsi_target.c @@ -283,8 +283,8 @@ static struct iscsi_np *iscsit_get_np( sock_in6 = (struct sockaddr_in6 *)sockaddr; sock_in6_e = (struct sockaddr_in6 *)&np->np_sockaddr; - if (!memcmp((void *)&sock_in6->sin6_addr.in6_u, - (void *)&sock_in6_e->sin6_addr.in6_u, + if (!memcmp(&sock_in6->sin6_addr.in6_u, + &sock_in6_e->sin6_addr.in6_u, sizeof(struct in6_addr))) ip_match = 1; @@ -1224,7 +1224,7 @@ static void iscsit_do_crypto_hash_buf( crypto_hash_init(hash); - sg_init_one(&sg, (u8 *)buf, payload_length); + sg_init_one(&sg, buf, payload_length); crypto_hash_update(hash, &sg, payload_length); if (padding) { @@ -1602,7 +1602,7 @@ static int iscsit_handle_nop_out( /* * Attach ping data to struct iscsi_cmd->buf_ptr. */ - cmd->buf_ptr = (void *)ping_data; + cmd->buf_ptr = ping_data; cmd->buf_ptr_size = payload_length; pr_debug("Got %u bytes of NOPOUT ping" @@ -3196,7 +3196,7 @@ static int iscsit_build_sendtargets_response(struct iscsi_cmd *cmd) end_of_buf = 1; goto eob; } - memcpy((void *)payload + payload_len, buf, len); + memcpy(payload + payload_len, buf, len); payload_len += len; spin_lock(&tiqn->tiqn_tpg_lock); @@ -3228,7 +3228,7 @@ static int iscsit_build_sendtargets_response(struct iscsi_cmd *cmd) end_of_buf = 1; goto eob; } - memcpy((void *)payload + payload_len, buf, len); + memcpy(payload + payload_len, buf, len); payload_len += len; } spin_unlock(&tpg->tpg_np_lock); @@ -3485,7 +3485,7 @@ int iscsi_target_tx_thread(void *arg) struct iscsi_conn *conn; struct iscsi_queue_req *qr = NULL; struct se_cmd *se_cmd; - struct iscsi_thread_set *ts = (struct iscsi_thread_set *)arg; + struct iscsi_thread_set *ts = arg; /* * Allow ourselves to be interrupted by SIGINT so that a * connection recovery / failure event can be triggered externally. @@ -3774,7 +3774,7 @@ int iscsi_target_rx_thread(void *arg) u8 buffer[ISCSI_HDR_LEN], opcode; u32 checksum = 0, digest = 0; struct iscsi_conn *conn = NULL; - struct iscsi_thread_set *ts = (struct iscsi_thread_set *)arg; + struct iscsi_thread_set *ts = arg; struct kvec iov; /* * Allow ourselves to be interrupted by SIGINT so that a diff --git a/drivers/target/iscsi/iscsi_target_auth.c b/drivers/target/iscsi/iscsi_target_auth.c index 1cd6ce373b83..92a2526f0a20 100644 --- a/drivers/target/iscsi/iscsi_target_auth.c +++ b/drivers/target/iscsi/iscsi_target_auth.c @@ -82,7 +82,7 @@ static void chap_gen_challenge( unsigned int *c_len) { unsigned char challenge_asciihex[CHAP_CHALLENGE_LENGTH * 2 + 1]; - struct iscsi_chap *chap = (struct iscsi_chap *) conn->auth_protocol; + struct iscsi_chap *chap = conn->auth_protocol; memset(challenge_asciihex, 0, CHAP_CHALLENGE_LENGTH * 2 + 1); @@ -120,7 +120,7 @@ static struct iscsi_chap *chap_server_open( if (!conn->auth_protocol) return NULL; - chap = (struct iscsi_chap *) conn->auth_protocol; + chap = conn->auth_protocol; /* * We only support MD5 MDA presently. */ @@ -172,7 +172,7 @@ static int chap_server_compute_md5( unsigned char client_digest[MD5_SIGNATURE_SIZE]; unsigned char server_digest[MD5_SIGNATURE_SIZE]; unsigned char chap_n[MAX_CHAP_N_SIZE], chap_r[MAX_RESPONSE_LENGTH]; - struct iscsi_chap *chap = (struct iscsi_chap *) conn->auth_protocol; + struct iscsi_chap *chap = conn->auth_protocol; struct crypto_hash *tfm; struct hash_desc desc; struct scatterlist sg; @@ -246,7 +246,7 @@ static int chap_server_compute_md5( goto out; } - sg_init_one(&sg, (void *)&chap->id, 1); + sg_init_one(&sg, &chap->id, 1); ret = crypto_hash_update(&desc, &sg, 1); if (ret < 0) { pr_err("crypto_hash_update() failed for id\n"); @@ -254,7 +254,7 @@ static int chap_server_compute_md5( goto out; } - sg_init_one(&sg, (void *)&auth->password, strlen(auth->password)); + sg_init_one(&sg, &auth->password, strlen(auth->password)); ret = crypto_hash_update(&desc, &sg, strlen(auth->password)); if (ret < 0) { pr_err("crypto_hash_update() failed for password\n"); @@ -262,7 +262,7 @@ static int chap_server_compute_md5( goto out; } - sg_init_one(&sg, (void *)chap->challenge, CHAP_CHALLENGE_LENGTH); + sg_init_one(&sg, chap->challenge, CHAP_CHALLENGE_LENGTH); ret = crypto_hash_update(&desc, &sg, CHAP_CHALLENGE_LENGTH); if (ret < 0) { pr_err("crypto_hash_update() failed for challenge\n"); @@ -304,11 +304,11 @@ static int chap_server_compute_md5( goto out; } + /* FIXME: What happens when simple_strtoul() return 256, 257, etc.? */ if (type == HEX) - id = (unsigned char)simple_strtoul((char *)&identifier[2], - &endptr, 0); + id = simple_strtoul(&identifier[2], &endptr, 0); else - id = (unsigned char)simple_strtoul(identifier, &endptr, 0); + id = simple_strtoul(identifier, &endptr, 0); /* * RFC 1994 says Identifier is no more than octet (8 bits). */ @@ -351,7 +351,7 @@ static int chap_server_compute_md5( goto out; } - sg_init_one(&sg, (void *)&id, 1); + sg_init_one(&sg, &id, 1); ret = crypto_hash_update(&desc, &sg, 1); if (ret < 0) { pr_err("crypto_hash_update() failed for id\n"); @@ -359,7 +359,7 @@ static int chap_server_compute_md5( goto out; } - sg_init_one(&sg, (void *)auth->password_mutual, + sg_init_one(&sg, auth->password_mutual, strlen(auth->password_mutual)); ret = crypto_hash_update(&desc, &sg, strlen(auth->password_mutual)); if (ret < 0) { @@ -371,7 +371,7 @@ static int chap_server_compute_md5( /* * Convert received challenge to binary hex. */ - sg_init_one(&sg, (void *)challenge_binhex, challenge_len); + sg_init_one(&sg, challenge_binhex, challenge_len); ret = crypto_hash_update(&desc, &sg, challenge_len); if (ret < 0) { pr_err("crypto_hash_update() failed for ma challenge\n"); @@ -414,7 +414,7 @@ static int chap_got_response( char *nr_out_ptr, unsigned int *nr_out_len) { - struct iscsi_chap *chap = (struct iscsi_chap *) conn->auth_protocol; + struct iscsi_chap *chap = conn->auth_protocol; switch (chap->digest_type) { case CHAP_DIGEST_MD5: @@ -437,7 +437,7 @@ u32 chap_main_loop( int *in_len, int *out_len) { - struct iscsi_chap *chap = (struct iscsi_chap *) conn->auth_protocol; + struct iscsi_chap *chap = conn->auth_protocol; if (!chap) { chap = chap_server_open(conn, auth, in_text, out_text, out_len); diff --git a/drivers/target/iscsi/iscsi_target_configfs.c b/drivers/target/iscsi/iscsi_target_configfs.c index c04d389a2ff2..3468caab47a2 100644 --- a/drivers/target/iscsi/iscsi_target_configfs.c +++ b/drivers/target/iscsi/iscsi_target_configfs.c @@ -52,8 +52,7 @@ struct iscsi_portal_group *lio_get_tpg_from_tpg_item( { struct se_portal_group *se_tpg = container_of(to_config_group(item), struct se_portal_group, tpg_group); - struct iscsi_portal_group *tpg = - (struct iscsi_portal_group *)se_tpg->se_tpg_fabric_ptr; + struct iscsi_portal_group *tpg = se_tpg->se_tpg_fabric_ptr; int ret; if (!tpg) { @@ -1221,7 +1220,7 @@ struct se_portal_group *lio_target_tiqn_addtpg( ret = core_tpg_register( &lio_target_fabric_configfs->tf_ops, - wwn, &tpg->tpg_se_tpg, (void *)tpg, + wwn, &tpg->tpg_se_tpg, tpg, TRANSPORT_TPG_TYPE_NORMAL); if (ret < 0) return NULL; diff --git a/drivers/target/iscsi/iscsi_target_login.c b/drivers/target/iscsi/iscsi_target_login.c index d74e278ec8f7..373b0cc6abd8 100644 --- a/drivers/target/iscsi/iscsi_target_login.c +++ b/drivers/target/iscsi/iscsi_target_login.c @@ -143,7 +143,7 @@ int iscsi_check_for_session_reinstatement(struct iscsi_conn *conn) list_for_each_entry_safe(se_sess, se_sess_tmp, &se_tpg->tpg_sess_list, sess_list) { - sess_p = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess_p = se_sess->fabric_sess_ptr; spin_lock(&sess_p->conn_lock); if (atomic_read(&sess_p->session_fall_back_to_erl0) || atomic_read(&sess_p->session_logout) || @@ -151,9 +151,9 @@ int iscsi_check_for_session_reinstatement(struct iscsi_conn *conn) spin_unlock(&sess_p->conn_lock); continue; } - if (!memcmp((void *)sess_p->isid, (void *)conn->sess->isid, 6) && - (!strcmp((void *)sess_p->sess_ops->InitiatorName, - (void *)initiatorname_param->value) && + if (!memcmp(sess_p->isid, conn->sess->isid, 6) && + (!strcmp(sess_p->sess_ops->InitiatorName, + initiatorname_param->value) && (sess_p->sess_ops->SessionType == sessiontype))) { atomic_set(&sess_p->session_reinstatement, 1); spin_unlock(&sess_p->conn_lock); @@ -229,7 +229,7 @@ static int iscsi_login_zero_tsih_s1( iscsi_login_set_conn_values(sess, conn, pdu->cid); sess->init_task_tag = pdu->itt; - memcpy((void *)&sess->isid, (void *)pdu->isid, 6); + memcpy(&sess->isid, pdu->isid, 6); sess->exp_cmd_sn = pdu->cmdsn; INIT_LIST_HEAD(&sess->sess_conn_list); INIT_LIST_HEAD(&sess->sess_ooo_cmdsn_list); @@ -440,8 +440,7 @@ static int iscsi_login_non_zero_tsih_s2( atomic_read(&sess_p->session_logout) || (sess_p->time2retain_timer_flags & ISCSI_TF_EXPIRED)) continue; - if (!memcmp((const void *)sess_p->isid, - (const void *)pdu->isid, 6) && + if (!memcmp(sess_p->isid, pdu->isid, 6) && (sess_p->tsih == pdu->tsih)) { iscsit_inc_session_usage_count(sess_p); iscsit_stop_time2retain_timer(sess_p); @@ -654,7 +653,7 @@ static int iscsi_post_login_handler( spin_lock_bh(&se_tpg->session_lock); __transport_register_session(&sess->tpg->tpg_se_tpg, - se_sess->se_node_acl, se_sess, (void *)sess); + se_sess->se_node_acl, se_sess, sess); pr_debug("Moving to TARG_SESS_STATE_LOGGED_IN.\n"); sess->session_state = TARG_SESS_STATE_LOGGED_IN; @@ -811,7 +810,7 @@ int iscsi_target_setup_login_socket( * Setup the np->np_sockaddr from the passed sockaddr setup * in iscsi_target_configfs.c code.. */ - memcpy((void *)&np->np_sockaddr, (void *)sockaddr, + memcpy(&np->np_sockaddr, sockaddr, sizeof(struct __kernel_sockaddr_storage)); if (sockaddr->ss_family == AF_INET6) @@ -821,6 +820,7 @@ int iscsi_target_setup_login_socket( /* * Set SO_REUSEADDR, and disable Nagel Algorithm with TCP_NODELAY. */ + /* FIXME: Someone please explain why this is endian-safe */ opt = 1; if (np->np_network_transport == ISCSI_TCP) { ret = kernel_setsockopt(sock, IPPROTO_TCP, TCP_NODELAY, @@ -832,6 +832,7 @@ int iscsi_target_setup_login_socket( } } + /* FIXME: Someone please explain why this is endian-safe */ ret = kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEADDR, (char *)&opt, sizeof(opt)); if (ret < 0) { @@ -1206,7 +1207,7 @@ out: int iscsi_target_login_thread(void *arg) { - struct iscsi_np *np = (struct iscsi_np *)arg; + struct iscsi_np *np = arg; int ret; allow_signal(SIGINT); diff --git a/drivers/target/iscsi/iscsi_target_nego.c b/drivers/target/iscsi/iscsi_target_nego.c index 53eea5c827f0..e89fa7457254 100644 --- a/drivers/target/iscsi/iscsi_target_nego.c +++ b/drivers/target/iscsi/iscsi_target_nego.c @@ -732,7 +732,7 @@ static void iscsi_initiatorname_tolower( u32 iqn_size = strlen(param_buf), i; for (i = 0; i < iqn_size; i++) { - c = (char *)¶m_buf[i]; + c = ¶m_buf[i]; if (!isupper(*c)) continue; diff --git a/drivers/target/iscsi/iscsi_target_nodeattrib.c b/drivers/target/iscsi/iscsi_target_nodeattrib.c index ac047d8a7a13..b3c699c4fe8c 100644 --- a/drivers/target/iscsi/iscsi_target_nodeattrib.c +++ b/drivers/target/iscsi/iscsi_target_nodeattrib.c @@ -134,7 +134,7 @@ extern int iscsit_na_nopin_timeout( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; spin_lock(&sess->conn_lock); list_for_each_entry(conn, &sess->sess_conn_list, diff --git a/drivers/target/iscsi/iscsi_target_stat.c b/drivers/target/iscsi/iscsi_target_stat.c index 5a1aac0ec7de..421d6947dc64 100644 --- a/drivers/target/iscsi/iscsi_target_stat.c +++ b/drivers/target/iscsi/iscsi_target_stat.c @@ -745,7 +745,7 @@ static ssize_t iscsi_stat_sess_show_attr_node( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%u\n", sess->sess_ops->SessionType ? 0 : ISCSI_NODE_INDEX); @@ -769,7 +769,7 @@ static ssize_t iscsi_stat_sess_show_attr_indx( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%u\n", sess->session_index); @@ -793,7 +793,7 @@ static ssize_t iscsi_stat_sess_show_attr_cmd_pdus( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%u\n", sess->cmd_pdus); } @@ -816,7 +816,7 @@ static ssize_t iscsi_stat_sess_show_attr_rsp_pdus( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%u\n", sess->rsp_pdus); } @@ -839,7 +839,7 @@ static ssize_t iscsi_stat_sess_show_attr_txdata_octs( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%llu\n", (unsigned long long)sess->tx_data_octets); @@ -863,7 +863,7 @@ static ssize_t iscsi_stat_sess_show_attr_rxdata_octs( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%llu\n", (unsigned long long)sess->rx_data_octets); @@ -887,7 +887,7 @@ static ssize_t iscsi_stat_sess_show_attr_conn_digest_errors( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%u\n", sess->conn_digest_errors); @@ -911,7 +911,7 @@ static ssize_t iscsi_stat_sess_show_attr_conn_timeout_errors( spin_lock_bh(&se_nacl->nacl_sess_lock); se_sess = se_nacl->nacl_sess; if (se_sess) { - sess = (struct iscsi_session *)se_sess->fabric_sess_ptr; + sess = se_sess->fabric_sess_ptr; if (sess) ret = snprintf(page, PAGE_SIZE, "%u\n", sess->conn_timeout_errors); diff --git a/drivers/target/iscsi/iscsi_target_tpg.c b/drivers/target/iscsi/iscsi_target_tpg.c index cfbcf6e4877d..879d8d0fa3fe 100644 --- a/drivers/target/iscsi/iscsi_target_tpg.c +++ b/drivers/target/iscsi/iscsi_target_tpg.c @@ -70,7 +70,7 @@ int iscsit_load_discovery_tpg(void) ret = core_tpg_register( &lio_target_fabric_configfs->tf_ops, - NULL, &tpg->tpg_se_tpg, (void *)tpg, + NULL, &tpg->tpg_se_tpg, tpg, TRANSPORT_TPG_TYPE_DISCOVERY); if (ret < 0) { kfree(tpg); diff --git a/drivers/target/iscsi/iscsi_target_util.c b/drivers/target/iscsi/iscsi_target_util.c index c1daea190859..a05ca1c4f01c 100644 --- a/drivers/target/iscsi/iscsi_target_util.c +++ b/drivers/target/iscsi/iscsi_target_util.c @@ -287,7 +287,7 @@ struct iscsi_cmd *iscsit_allocate_se_cmd_for_tmr( } se_cmd->se_tmr_req = core_tmr_alloc_req(se_cmd, - (void *)cmd->tmr_req, tcm_function, + cmd->tmr_req, tcm_function, GFP_KERNEL); if (!se_cmd->se_tmr_req) goto out; @@ -1064,7 +1064,7 @@ static void iscsit_handle_nopin_response_timeout(unsigned long data) if (tiqn) { spin_lock_bh(&tiqn->sess_err_stats.lock); strcpy(tiqn->sess_err_stats.last_sess_fail_rem_name, - (void *)conn->sess->sess_ops->InitiatorName); + conn->sess->sess_ops->InitiatorName); tiqn->sess_err_stats.last_sess_failure_type = ISCSI_SESS_ERR_CXN_TIMEOUT; tiqn->sess_err_stats.cxn_timeout_errors++; diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c index d427a290e184..572d27a6b7ae 100644 --- a/drivers/target/loopback/tcm_loop.c +++ b/drivers/target/loopback/tcm_loop.c @@ -559,8 +559,7 @@ static char *tcm_loop_get_fabric_name(void) static u8 tcm_loop_get_fabric_proto_ident(struct se_portal_group *se_tpg) { - struct tcm_loop_tpg *tl_tpg = - (struct tcm_loop_tpg *)se_tpg->se_tpg_fabric_ptr; + struct tcm_loop_tpg *tl_tpg = se_tpg->se_tpg_fabric_ptr; struct tcm_loop_hba *tl_hba = tl_tpg->tl_hba; /* * tl_proto_id is set at tcm_loop_configfs.c:tcm_loop_make_scsi_hba() @@ -587,8 +586,7 @@ static u8 tcm_loop_get_fabric_proto_ident(struct se_portal_group *se_tpg) static char *tcm_loop_get_endpoint_wwn(struct se_portal_group *se_tpg) { - struct tcm_loop_tpg *tl_tpg = - (struct tcm_loop_tpg *)se_tpg->se_tpg_fabric_ptr; + struct tcm_loop_tpg *tl_tpg = se_tpg->se_tpg_fabric_ptr; /* * Return the passed NAA identifier for the SAS Target Port */ @@ -597,8 +595,7 @@ static char *tcm_loop_get_endpoint_wwn(struct se_portal_group *se_tpg) static u16 tcm_loop_get_tag(struct se_portal_group *se_tpg) { - struct tcm_loop_tpg *tl_tpg = - (struct tcm_loop_tpg *)se_tpg->se_tpg_fabric_ptr; + struct tcm_loop_tpg *tl_tpg = se_tpg->se_tpg_fabric_ptr; /* * This Tag is used when forming SCSI Name identifier in EVPD=1 0x83 * to represent the SCSI Target Port. @@ -618,8 +615,7 @@ static u32 tcm_loop_get_pr_transport_id( int *format_code, unsigned char *buf) { - struct tcm_loop_tpg *tl_tpg = - (struct tcm_loop_tpg *)se_tpg->se_tpg_fabric_ptr; + struct tcm_loop_tpg *tl_tpg = se_tpg->se_tpg_fabric_ptr; struct tcm_loop_hba *tl_hba = tl_tpg->tl_hba; switch (tl_hba->tl_proto_id) { @@ -648,8 +644,7 @@ static u32 tcm_loop_get_pr_transport_id_len( struct t10_pr_registration *pr_reg, int *format_code) { - struct tcm_loop_tpg *tl_tpg = - (struct tcm_loop_tpg *)se_tpg->se_tpg_fabric_ptr; + struct tcm_loop_tpg *tl_tpg = se_tpg->se_tpg_fabric_ptr; struct tcm_loop_hba *tl_hba = tl_tpg->tl_hba; switch (tl_hba->tl_proto_id) { @@ -682,8 +677,7 @@ static char *tcm_loop_parse_pr_out_transport_id( u32 *out_tid_len, char **port_nexus_ptr) { - struct tcm_loop_tpg *tl_tpg = - (struct tcm_loop_tpg *)se_tpg->se_tpg_fabric_ptr; + struct tcm_loop_tpg *tl_tpg = se_tpg->se_tpg_fabric_ptr; struct tcm_loop_hba *tl_hba = tl_tpg->tl_hba; switch (tl_hba->tl_proto_id) { diff --git a/drivers/target/target_core_cdb.c b/drivers/target/target_core_cdb.c index 59219509d886..5c60d0a97a50 100644 --- a/drivers/target/target_core_cdb.c +++ b/drivers/target/target_core_cdb.c @@ -116,11 +116,9 @@ target_emulate_inquiry_std(struct se_cmd *cmd) goto out; } - snprintf((unsigned char *)&buf[8], 8, "LIO-ORG"); - snprintf((unsigned char *)&buf[16], 16, "%s", - &dev->se_sub_dev->t10_wwn.model[0]); - snprintf((unsigned char *)&buf[32], 4, "%s", - &dev->se_sub_dev->t10_wwn.revision[0]); + snprintf(&buf[8], 8, "LIO-ORG"); + snprintf(&buf[16], 16, "%s", dev->se_sub_dev->t10_wwn.model); + snprintf(&buf[32], 4, "%s", dev->se_sub_dev->t10_wwn.revision); buf[4] = 31; /* Set additional length to 31 */ out: @@ -139,8 +137,7 @@ target_emulate_evpd_80(struct se_cmd *cmd, unsigned char *buf) SDF_EMULATED_VPD_UNIT_SERIAL) { u32 unit_serial_len; - unit_serial_len = - strlen(&dev->se_sub_dev->t10_wwn.unit_serial[0]); + unit_serial_len = strlen(dev->se_sub_dev->t10_wwn.unit_serial); unit_serial_len++; /* For NULL Terminator */ if (((len + 4) + unit_serial_len) > cmd->data_length) { @@ -149,8 +146,8 @@ target_emulate_evpd_80(struct se_cmd *cmd, unsigned char *buf) buf[3] = (len & 0xff); return 0; } - len += sprintf((unsigned char *)&buf[4], "%s", - &dev->se_sub_dev->t10_wwn.unit_serial[0]); + len += sprintf(&buf[4], "%s", + dev->se_sub_dev->t10_wwn.unit_serial); len++; /* Extra Byte for NULL Terminator */ buf[3] = len; } @@ -280,14 +277,13 @@ check_t10_vend_desc: len += (prod_len + unit_serial_len); goto check_port; } - id_len += sprintf((unsigned char *)&buf[off+12], - "%s:%s", prod, + id_len += sprintf(&buf[off+12], "%s:%s", prod, &dev->se_sub_dev->t10_wwn.unit_serial[0]); } buf[off] = 0x2; /* ASCII */ buf[off+1] = 0x1; /* T10 Vendor ID */ buf[off+2] = 0x0; - memcpy((unsigned char *)&buf[off+4], "LIO-ORG", 8); + memcpy(&buf[off+4], "LIO-ORG", 8); /* Extra Byte for NULL Terminator */ id_len++; /* Identifier Length */ diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index 37b2b3d03c65..eb62371611f6 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -1629,7 +1629,7 @@ static struct config_item_type target_core_dev_pr_cit = { static ssize_t target_core_show_dev_info(void *p, char *page) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; struct se_hba *hba = se_dev->se_dev_hba; struct se_subsystem_api *t = hba->transport; int bl = 0; @@ -1657,7 +1657,7 @@ static ssize_t target_core_store_dev_control( const char *page, size_t count) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; struct se_hba *hba = se_dev->se_dev_hba; struct se_subsystem_api *t = hba->transport; @@ -1680,7 +1680,7 @@ static struct target_core_configfs_attribute target_core_attr_dev_control = { static ssize_t target_core_show_dev_alias(void *p, char *page) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; if (!(se_dev->su_dev_flags & SDF_USING_ALIAS)) return 0; @@ -1693,7 +1693,7 @@ static ssize_t target_core_store_dev_alias( const char *page, size_t count) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; struct se_hba *hba = se_dev->se_dev_hba; ssize_t read_bytes; @@ -1726,7 +1726,7 @@ static struct target_core_configfs_attribute target_core_attr_dev_alias = { static ssize_t target_core_show_dev_udev_path(void *p, char *page) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; if (!(se_dev->su_dev_flags & SDF_USING_UDEV_PATH)) return 0; @@ -1739,7 +1739,7 @@ static ssize_t target_core_store_dev_udev_path( const char *page, size_t count) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; struct se_hba *hba = se_dev->se_dev_hba; ssize_t read_bytes; @@ -1775,7 +1775,7 @@ static ssize_t target_core_store_dev_enable( const char *page, size_t count) { - struct se_subsystem_dev *se_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *se_dev = p; struct se_device *dev; struct se_hba *hba = se_dev->se_dev_hba; struct se_subsystem_api *t = hba->transport; @@ -1820,7 +1820,7 @@ static struct target_core_configfs_attribute target_core_attr_dev_enable = { static ssize_t target_core_show_alua_lu_gp(void *p, char *page) { struct se_device *dev; - struct se_subsystem_dev *su_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *su_dev = p; struct config_item *lu_ci; struct t10_alua_lu_gp *lu_gp; struct t10_alua_lu_gp_member *lu_gp_mem; @@ -1858,7 +1858,7 @@ static ssize_t target_core_store_alua_lu_gp( size_t count) { struct se_device *dev; - struct se_subsystem_dev *su_dev = (struct se_subsystem_dev *)p; + struct se_subsystem_dev *su_dev = p; struct se_hba *hba = su_dev->se_dev_hba; struct t10_alua_lu_gp *lu_gp = NULL, *lu_gp_new = NULL; struct t10_alua_lu_gp_member *lu_gp_mem; diff --git a/drivers/target/target_core_fabric_lib.c b/drivers/target/target_core_fabric_lib.c index 32d7feb405ec..283a36e464e6 100644 --- a/drivers/target/target_core_fabric_lib.c +++ b/drivers/target/target_core_fabric_lib.c @@ -399,7 +399,7 @@ char *iscsi_parse_pr_out_transport_id( add_len = ((buf[2] >> 8) & 0xff); add_len |= (buf[3] & 0xff); - tid_len = strlen((char *)&buf[4]); + tid_len = strlen(&buf[4]); tid_len += 4; /* Add four bytes for iSCSI Transport ID header */ tid_len += 1; /* Add one byte for NULL terminator */ padding = ((-tid_len) & 3); @@ -420,11 +420,11 @@ char *iscsi_parse_pr_out_transport_id( * format. */ if (format_code == 0x40) { - p = strstr((char *)&buf[4], ",i,0x"); + p = strstr(&buf[4], ",i,0x"); if (!p) { pr_err("Unable to locate \",i,0x\" seperator" " for Initiator port identifier: %s\n", - (char *)&buf[4]); + &buf[4]); return NULL; } *p = '\0'; /* Terminate iSCSI Name */ diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c index 91bf11a0ab16..7ed58e2df791 100644 --- a/drivers/target/target_core_file.c +++ b/drivers/target/target_core_file.c @@ -85,7 +85,7 @@ static void fd_detach_hba(struct se_hba *hba) static void *fd_allocate_virtdevice(struct se_hba *hba, const char *name) { struct fd_dev *fd_dev; - struct fd_host *fd_host = (struct fd_host *) hba->hba_ptr; + struct fd_host *fd_host = hba->hba_ptr; fd_dev = kzalloc(sizeof(struct fd_dev), GFP_KERNEL); if (!fd_dev) { @@ -113,8 +113,8 @@ static struct se_device *fd_create_virtdevice( struct se_device *dev; struct se_dev_limits dev_limits; struct queue_limits *limits; - struct fd_dev *fd_dev = (struct fd_dev *) p; - struct fd_host *fd_host = (struct fd_host *) hba->hba_ptr; + struct fd_dev *fd_dev = p; + struct fd_host *fd_host = hba->hba_ptr; mm_segment_t old_fs; struct file *file; struct inode *inode = NULL; @@ -239,7 +239,7 @@ fail: */ static void fd_free_device(void *p) { - struct fd_dev *fd_dev = (struct fd_dev *) p; + struct fd_dev *fd_dev = p; if (fd_dev->fd_file) { filp_close(fd_dev->fd_file, NULL); @@ -558,7 +558,7 @@ out: static ssize_t fd_check_configfs_dev_params(struct se_hba *hba, struct se_subsystem_dev *se_dev) { - struct fd_dev *fd_dev = (struct fd_dev *) se_dev->se_dev_su_ptr; + struct fd_dev *fd_dev = se_dev->se_dev_su_ptr; if (!(fd_dev->fbd_flags & FBDF_HAS_PATH)) { pr_err("Missing fd_dev_name=\n"); diff --git a/drivers/target/target_core_iblock.c b/drivers/target/target_core_iblock.c index 3c1a7aa658a2..cc8e6b58ef20 100644 --- a/drivers/target/target_core_iblock.c +++ b/drivers/target/target_core_iblock.c @@ -464,7 +464,7 @@ static ssize_t iblock_show_configfs_dev_params( if (bd) { bl += sprintf(b + bl, "Major: %d Minor: %d %s\n", MAJOR(bd->bd_dev), MINOR(bd->bd_dev), (!bd->bd_contains) ? - "" : (bd->bd_holder == (struct iblock_dev *)ibd) ? + "" : (bd->bd_holder == ibd) ? "CLAIMED: IBLOCK" : "CLAIMED: OS"); } else { bl += sprintf(b + bl, "Major: 0 Minor: 0\n"); diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c index bb44ed4d0272..b73a399cdd54 100644 --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -104,7 +104,7 @@ static void pscsi_detach_hba(struct se_hba *hba) static int pscsi_pmode_enable_hba(struct se_hba *hba, unsigned long mode_flag) { - struct pscsi_hba_virt *phv = (struct pscsi_hba_virt *)hba->hba_ptr; + struct pscsi_hba_virt *phv = hba->hba_ptr; struct Scsi_Host *sh = phv->phv_lld_host; /* * Release the struct Scsi_Host @@ -405,7 +405,7 @@ static struct se_device *pscsi_create_type_disk( __releases(sh->host_lock) { struct se_device *dev; - struct pscsi_hba_virt *phv = (struct pscsi_hba_virt *)pdv->pdv_se_hba->hba_ptr; + struct pscsi_hba_virt *phv = pdv->pdv_se_hba->hba_ptr; struct Scsi_Host *sh = sd->host; struct block_device *bd; u32 dev_flags = 0; @@ -453,7 +453,7 @@ static struct se_device *pscsi_create_type_rom( __releases(sh->host_lock) { struct se_device *dev; - struct pscsi_hba_virt *phv = (struct pscsi_hba_virt *)pdv->pdv_se_hba->hba_ptr; + struct pscsi_hba_virt *phv = pdv->pdv_se_hba->hba_ptr; struct Scsi_Host *sh = sd->host; u32 dev_flags = 0; @@ -488,7 +488,7 @@ static struct se_device *pscsi_create_type_other( __releases(sh->host_lock) { struct se_device *dev; - struct pscsi_hba_virt *phv = (struct pscsi_hba_virt *)pdv->pdv_se_hba->hba_ptr; + struct pscsi_hba_virt *phv = pdv->pdv_se_hba->hba_ptr; struct Scsi_Host *sh = sd->host; u32 dev_flags = 0; @@ -509,10 +509,10 @@ static struct se_device *pscsi_create_virtdevice( struct se_subsystem_dev *se_dev, void *p) { - struct pscsi_dev_virt *pdv = (struct pscsi_dev_virt *)p; + struct pscsi_dev_virt *pdv = p; struct se_device *dev; struct scsi_device *sd; - struct pscsi_hba_virt *phv = (struct pscsi_hba_virt *)hba->hba_ptr; + struct pscsi_hba_virt *phv = hba->hba_ptr; struct Scsi_Host *sh = phv->phv_lld_host; int legacy_mode_enable = 0; @@ -1143,7 +1143,7 @@ static unsigned char *pscsi_get_sense_buffer(struct se_task *task) { struct pscsi_plugin_task *pt = PSCSI_TASK(task); - return (unsigned char *)&pt->pscsi_sense[0]; + return pt->pscsi_sense; } /* pscsi_get_device_rev(): diff --git a/drivers/target/target_core_stat.c b/drivers/target/target_core_stat.c index 6d8a90464f2e..f8c2d2cc3431 100644 --- a/drivers/target/target_core_stat.c +++ b/drivers/target/target_core_stat.c @@ -1755,8 +1755,7 @@ static ssize_t target_stat_scsi_att_intr_port_show_attr_port_ident( /* scsiAttIntrPortName+scsiAttIntrPortIdentifier */ memset(buf, 0, 64); if (tpg->se_tpg_tfo->sess_get_initiator_sid != NULL) - tpg->se_tpg_tfo->sess_get_initiator_sid(se_sess, - (unsigned char *)&buf[0], 64); + tpg->se_tpg_tfo->sess_get_initiator_sid(se_sess, buf, 64); ret = snprintf(page, PAGE_SIZE, "%s+i+%s\n", nacl->initiatorname, buf); spin_unlock_irq(&nacl->nacl_sess_lock); diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 511836eb2953..122925686de2 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -4161,7 +4161,7 @@ check_cond: static int transport_clear_lun_thread(void *p) { - struct se_lun *lun = (struct se_lun *)p; + struct se_lun *lun = p; __transport_clear_lun_from_sessions(lun); complete(&lun->lun_shutdown_comp); @@ -4580,7 +4580,7 @@ static int transport_processing_thread(void *param) { int ret; struct se_cmd *cmd; - struct se_device *dev = (struct se_device *) param; + struct se_device *dev = param; while (!kthread_should_stop()) { ret = wait_event_interruptible(dev->dev_queue_obj.thread_wq, From bc704fb58f507deea84a4c047a3a71351f0c55aa Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Mon, 28 Nov 2011 01:02:07 -0800 Subject: [PATCH 147/820] iscsi-target: fix chap identifier simple_strtoul usage This patch makes chap_server_compute_md5() use proper unsigned long usage for the CHAP_I (identifier) and check for values beyond 255 as per RFC-1994. Reported-by: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/iscsi/iscsi_target_auth.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/target/iscsi/iscsi_target_auth.c b/drivers/target/iscsi/iscsi_target_auth.c index 92a2526f0a20..db0cf7c8adde 100644 --- a/drivers/target/iscsi/iscsi_target_auth.c +++ b/drivers/target/iscsi/iscsi_target_auth.c @@ -165,7 +165,8 @@ static int chap_server_compute_md5( unsigned int *nr_out_len) { char *endptr; - unsigned char id, digest[MD5_SIGNATURE_SIZE]; + unsigned long id; + unsigned char digest[MD5_SIGNATURE_SIZE]; unsigned char type, response[MD5_SIGNATURE_SIZE * 2 + 2]; unsigned char identifier[10], *challenge = NULL; unsigned char *challenge_binhex = NULL; @@ -304,15 +305,18 @@ static int chap_server_compute_md5( goto out; } - /* FIXME: What happens when simple_strtoul() return 256, 257, etc.? */ if (type == HEX) id = simple_strtoul(&identifier[2], &endptr, 0); else id = simple_strtoul(identifier, &endptr, 0); + if (id > 255) { + pr_err("chap identifier: %lu greater than 255\n", id); + goto out; + } /* * RFC 1994 says Identifier is no more than octet (8 bits). */ - pr_debug("[server] Got CHAP_I=%d\n", id); + pr_debug("[server] Got CHAP_I=%lu\n", id); /* * Get CHAP_C. */ From 0877eafd1618ef683e2ff2273acdeca5f1b077b9 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 28 Nov 2011 13:57:25 +0100 Subject: [PATCH 148/820] target/configs: remove trailing newline from udev_path and alias This patch strips the trailing newline from backend device udev_path and alias attributes. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_configfs.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/target/target_core_configfs.c b/drivers/target/target_core_configfs.c index eb62371611f6..0955bb8979fb 100644 --- a/drivers/target/target_core_configfs.c +++ b/drivers/target/target_core_configfs.c @@ -1708,6 +1708,9 @@ static ssize_t target_core_store_dev_alias( read_bytes = snprintf(&se_dev->se_dev_alias[0], SE_DEV_ALIAS_LEN, "%s", page); + if (se_dev->se_dev_alias[read_bytes - 1] == '\n') + se_dev->se_dev_alias[read_bytes - 1] = '\0'; + pr_debug("Target_Core_ConfigFS: %s/%s set alias: %s\n", config_item_name(&hba->hba_group.cg_item), config_item_name(&se_dev->se_dev_group.cg_item), @@ -1754,6 +1757,9 @@ static ssize_t target_core_store_dev_udev_path( read_bytes = snprintf(&se_dev->se_dev_udev_path[0], SE_UDEV_PATH_LEN, "%s", page); + if (se_dev->se_dev_udev_path[read_bytes - 1] == '\n') + se_dev->se_dev_udev_path[read_bytes - 1] = '\0'; + pr_debug("Target_Core_ConfigFS: %s/%s set udev_path: %s\n", config_item_name(&hba->hba_group.cg_item), config_item_name(&se_dev->se_dev_group.cg_item), From 2e88efd3aaafa0df05593cdfa5fd17afe922781d Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 29 Nov 2011 03:20:41 -0500 Subject: [PATCH 149/820] tcm_loop: bump max_sectors There is not reason to artifically limit max_sectors in tcm_loop, set it to UINT_MAX to allow stressing the large I/O handling in the target core using the loopback driver. Also remove various superflous defines hiding the values set in the host template. Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- drivers/target/loopback/tcm_loop.c | 8 ++++---- drivers/target/loopback/tcm_loop.h | 11 +---------- 2 files changed, 5 insertions(+), 14 deletions(-) diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c index 572d27a6b7ae..c47ff7f59e57 100644 --- a/drivers/target/loopback/tcm_loop.c +++ b/drivers/target/loopback/tcm_loop.c @@ -416,11 +416,11 @@ static struct scsi_host_template tcm_loop_driver_template = { .queuecommand = tcm_loop_queuecommand, .change_queue_depth = tcm_loop_change_queue_depth, .eh_device_reset_handler = tcm_loop_device_reset, - .can_queue = TL_SCSI_CAN_QUEUE, + .can_queue = 1024, .this_id = -1, - .sg_tablesize = TL_SCSI_SG_TABLESIZE, - .cmd_per_lun = TL_SCSI_CMD_PER_LUN, - .max_sectors = TL_SCSI_MAX_SECTORS, + .sg_tablesize = 256, + .cmd_per_lun = 1024, + .max_sectors = 0xFFFF, .use_clustering = DISABLE_CLUSTERING, .slave_alloc = tcm_loop_slave_alloc, .slave_configure = tcm_loop_slave_configure, diff --git a/drivers/target/loopback/tcm_loop.h b/drivers/target/loopback/tcm_loop.h index 6b76c7a22bb0..15a036441471 100644 --- a/drivers/target/loopback/tcm_loop.h +++ b/drivers/target/loopback/tcm_loop.h @@ -1,16 +1,7 @@ #define TCM_LOOP_VERSION "v2.1-rc1" #define TL_WWN_ADDR_LEN 256 #define TL_TPGS_PER_HBA 32 -/* - * Defaults for struct scsi_host_template tcm_loop_driver_template - * - * We use large can_queue and cmd_per_lun here and let TCM enforce - * the underlying se_device_t->queue_depth. - */ -#define TL_SCSI_CAN_QUEUE 1024 -#define TL_SCSI_CMD_PER_LUN 1024 -#define TL_SCSI_MAX_SECTORS 1024 -#define TL_SCSI_SG_TABLESIZE 256 + /* * Used in tcm_loop_driver_probe() for struct Scsi_Host->max_cmd_len */ From e0a53e70e8d519579d92c150606ece22333645cb Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 29 Nov 2011 03:29:38 -0500 Subject: [PATCH 150/820] target: remove overagressive ____cacheline_aligned annoations If we want dynamically allocated objects to be cacheline aligned we need to tell that to the slab allocator by using the proper flags and not by liberally sprinkling annotations onto all structures. Signed-off-by: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- include/target/target_core_base.h | 62 +++++++++++++++---------------- 1 file changed, 31 insertions(+), 31 deletions(-) diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 311f5fc607e6..cadd4f6e57e1 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -252,7 +252,7 @@ enum tcm_tmrsp_table { struct se_obj { atomic_t obj_access_count; -} ____cacheline_aligned; +}; /* * Used by TCM Core internally to signal if ALUA emulation is enabled or @@ -299,7 +299,7 @@ struct t10_alua { struct config_group alua_tg_pt_gps_group; int (*alua_state_check)(struct se_cmd *, unsigned char *, u8 *); struct list_head tg_pt_gps_list; -} ____cacheline_aligned; +}; struct t10_alua_lu_gp { u16 lu_gp_id; @@ -310,7 +310,7 @@ struct t10_alua_lu_gp { struct config_group lu_gp_group; struct list_head lu_gp_node; struct list_head lu_gp_mem_list; -} ____cacheline_aligned; +}; struct t10_alua_lu_gp_member { bool lu_gp_assoc; @@ -319,7 +319,7 @@ struct t10_alua_lu_gp_member { struct t10_alua_lu_gp *lu_gp; struct se_device *lu_gp_mem_dev; struct list_head lu_gp_mem_list; -} ____cacheline_aligned; +}; struct t10_alua_tg_pt_gp { u16 tg_pt_gp_id; @@ -342,7 +342,7 @@ struct t10_alua_tg_pt_gp { struct config_group tg_pt_gp_group; struct list_head tg_pt_gp_list; struct list_head tg_pt_gp_mem_list; -} ____cacheline_aligned; +}; struct t10_alua_tg_pt_gp_member { bool tg_pt_gp_assoc; @@ -351,7 +351,7 @@ struct t10_alua_tg_pt_gp_member { struct t10_alua_tg_pt_gp *tg_pt_gp; struct se_port *tg_pt; struct list_head tg_pt_gp_mem_list; -} ____cacheline_aligned; +}; struct t10_vpd { unsigned char device_identifier[INQUIRY_VPD_DEVICE_IDENTIFIER_LEN]; @@ -361,7 +361,7 @@ struct t10_vpd { u32 association; u32 device_identifier_type; struct list_head vpd_list; -} ____cacheline_aligned; +}; struct t10_wwn { char vendor[8]; @@ -372,7 +372,7 @@ struct t10_wwn { struct se_subsystem_dev *t10_sub_dev; struct config_group t10_wwn_group; struct list_head t10_vpd_list; -} ____cacheline_aligned; +}; /* @@ -425,7 +425,7 @@ struct t10_pr_registration { struct list_head pr_reg_aptpl_list; struct list_head pr_reg_atp_list; struct list_head pr_reg_atp_mem_list; -} ____cacheline_aligned; +}; /* * This set of function pointer ops is set based upon SPC3_PERSISTENT_RESERVATIONS, @@ -466,20 +466,20 @@ struct t10_reservation { struct list_head registration_list; struct list_head aptpl_reg_list; struct t10_reservation_ops pr_ops; -} ____cacheline_aligned; +}; struct se_queue_req { int state; struct se_cmd *cmd; struct list_head qr_list; -} ____cacheline_aligned; +}; struct se_queue_obj { atomic_t queue_cnt; spinlock_t cmd_queue_lock; struct list_head qobj_list; wait_queue_head_t thread_wq; -} ____cacheline_aligned; +}; struct se_task { unsigned long long task_lba; @@ -496,7 +496,7 @@ struct se_task { struct list_head t_state_list; bool t_state_active; struct completion task_stop_comp; -} ____cacheline_aligned; +}; struct se_cmd { /* SAM response code being sent to initiator */ @@ -582,7 +582,7 @@ struct se_cmd { struct list_head t_task_list; u32 t_task_list_num; -} ____cacheline_aligned; +}; struct se_tmr_req { /* Task Management function to be preformed */ @@ -600,7 +600,7 @@ struct se_tmr_req { struct se_device *tmr_dev; struct se_lun *tmr_lun; struct list_head tmr_list; -} ____cacheline_aligned; +}; struct se_ua { u8 ua_asc; @@ -608,7 +608,7 @@ struct se_ua { struct se_node_acl *ua_nacl; struct list_head ua_dev_list; struct list_head ua_nacl_list; -} ____cacheline_aligned; +}; struct se_node_acl { char initiatorname[TRANSPORT_IQN_LEN]; @@ -635,7 +635,7 @@ struct se_node_acl { struct config_group *acl_default_groups[5]; struct list_head acl_list; struct list_head acl_sess_list; -} ____cacheline_aligned; +}; struct se_session { unsigned sess_tearing_down:1; @@ -648,7 +648,7 @@ struct se_session { struct list_head sess_cmd_list; struct list_head sess_wait_list; spinlock_t sess_cmd_lock; -} ____cacheline_aligned; +}; struct se_device; struct se_transform_info; @@ -668,7 +668,7 @@ struct se_lun_acl { struct list_head lacl_list; struct config_group se_lun_group; struct se_ml_stat_grps ml_stat_grps; -} ____cacheline_aligned; +}; struct se_dev_entry { bool def_pr_registered; @@ -693,7 +693,7 @@ struct se_dev_entry { struct se_lun *se_lun; struct list_head alua_port_list; struct list_head ua_list; -} ____cacheline_aligned; +}; struct se_dev_limits { /* Max supported HW queue depth */ @@ -702,7 +702,7 @@ struct se_dev_limits { u32 queue_depth; /* From include/linux/blkdev.h for the other HW/SW limits. */ struct queue_limits limits; -} ____cacheline_aligned; +}; struct se_dev_attrib { int emulate_dpo; @@ -731,7 +731,7 @@ struct se_dev_attrib { u32 unmap_granularity_alignment; struct se_subsystem_dev *da_sub_dev; struct config_group da_group; -} ____cacheline_aligned; +}; struct se_dev_stat_grps { struct config_group stat_group; @@ -764,7 +764,7 @@ struct se_subsystem_dev { struct config_group se_dev_pr_group; /* For target_core_stat.c groups */ struct se_dev_stat_grps dev_stat_grps; -} ____cacheline_aligned; +}; struct se_device { /* RELATIVE TARGET PORT IDENTIFER Counter */ @@ -830,7 +830,7 @@ struct se_device { struct se_subsystem_api *transport; /* Linked list for struct se_hba struct se_device list */ struct list_head dev_list; -} ____cacheline_aligned; +}; struct se_hba { u16 hba_tpgt; @@ -849,7 +849,7 @@ struct se_hba { struct config_group hba_group; struct mutex hba_access_mutex; struct se_subsystem_api *transport; -} ____cacheline_aligned; +}; struct se_port_stat_grps { struct config_group stat_group; @@ -875,13 +875,13 @@ struct se_lun { struct se_port *lun_sep; struct config_group lun_group; struct se_port_stat_grps port_stat_grps; -} ____cacheline_aligned; +}; struct scsi_port_stats { u64 cmd_pdus; u64 tx_data_octets; u64 rx_data_octets; -} ____cacheline_aligned; +}; struct se_port { /* RELATIVE TARGET PORT IDENTIFER */ @@ -901,12 +901,12 @@ struct se_port { struct se_portal_group *sep_tpg; struct list_head sep_alua_list; struct list_head sep_list; -} ____cacheline_aligned; +}; struct se_tpg_np { struct se_portal_group *tpg_np_parent; struct config_group tpg_np_group; -} ____cacheline_aligned; +}; struct se_portal_group { /* Type of target portal group, see transport_tpg_type_table */ @@ -939,13 +939,13 @@ struct se_portal_group { struct config_group tpg_acl_group; struct config_group tpg_attrib_group; struct config_group tpg_param_group; -} ____cacheline_aligned; +}; struct se_wwn { struct target_fabric_configfs *wwn_tf; struct config_group wwn_group; struct config_group *wwn_default_groups[2]; struct config_group fabric_stat_group; -} ____cacheline_aligned; +}; #endif /* TARGET_CORE_BASE_H */ From 77271de645631cb1dc4563d564f9dd5c214b5eed Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Mon, 28 Nov 2011 17:54:28 -0800 Subject: [PATCH 151/820] Documentation/target: Fix tcm_mod_builder.py build breakage This patch fixes TFO->release_cmd() and removes legacy pack_lun() usage and new_cmd_failure when generating new TCM fabric skeleton from the tcm_mod_builder.py script. Reported-by: Stefan Bergstrand Signed-off-by: Nicholas Bellinger --- Documentation/target/tcm_mod_builder.py | 23 ++--------------------- 1 file changed, 2 insertions(+), 21 deletions(-) diff --git a/Documentation/target/tcm_mod_builder.py b/Documentation/target/tcm_mod_builder.py index e2abb29fb993..520a923d7ac4 100755 --- a/Documentation/target/tcm_mod_builder.py +++ b/Documentation/target/tcm_mod_builder.py @@ -386,8 +386,7 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " .tpg_alloc_fabric_acl = " + fabric_mod_name + "_alloc_fabric_acl,\n" buf += " .tpg_release_fabric_acl = " + fabric_mod_name + "_release_fabric_acl,\n" buf += " .tpg_get_inst_index = " + fabric_mod_name + "_tpg_get_inst_index,\n" - buf += " .release_cmd_to_pool = " + fabric_mod_name + "_release_cmd,\n" - buf += " .release_cmd_direct = " + fabric_mod_name + "_release_cmd,\n" + buf += " .release_cmd = " + fabric_mod_name + "_release_cmd,\n" buf += " .shutdown_session = " + fabric_mod_name + "_shutdown_session,\n" buf += " .close_session = " + fabric_mod_name + "_close_session,\n" buf += " .stop_session = " + fabric_mod_name + "_stop_session,\n" @@ -400,14 +399,12 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " .set_default_node_attributes = " + fabric_mod_name + "_set_default_node_attrs,\n" buf += " .get_task_tag = " + fabric_mod_name + "_get_task_tag,\n" buf += " .get_cmd_state = " + fabric_mod_name + "_get_cmd_state,\n" - buf += " .new_cmd_failure = " + fabric_mod_name + "_new_cmd_failure,\n" buf += " .queue_data_in = " + fabric_mod_name + "_queue_data_in,\n" buf += " .queue_status = " + fabric_mod_name + "_queue_status,\n" buf += " .queue_tm_rsp = " + fabric_mod_name + "_queue_tm_rsp,\n" buf += " .get_fabric_sense_len = " + fabric_mod_name + "_get_fabric_sense_len,\n" buf += " .set_fabric_sense_len = " + fabric_mod_name + "_set_fabric_sense_len,\n" buf += " .is_state_remove = " + fabric_mod_name + "_is_state_remove,\n" - buf += " .pack_lun = " + fabric_mod_name + "_pack_lun,\n" buf += " /*\n" buf += " * Setup function pointers for generic logic in target_core_fabric_configfs.c\n" buf += " */\n" @@ -806,7 +803,7 @@ def tcm_mod_dump_fabric_ops(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += "}\n\n" bufi += "u32 " + fabric_mod_name + "_tpg_get_inst_index(struct se_portal_group *);\n" - if re.search('release_cmd_to_pool', fo): + if re.search('\*release_cmd\)\(', fo): buf += "void " + fabric_mod_name + "_release_cmd(struct se_cmd *se_cmd)\n" buf += "{\n" buf += " return;\n" @@ -890,13 +887,6 @@ def tcm_mod_dump_fabric_ops(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += "}\n\n" bufi += "int " + fabric_mod_name + "_get_cmd_state(struct se_cmd *);\n" - if re.search('new_cmd_failure\)\(', fo): - buf += "void " + fabric_mod_name + "_new_cmd_failure(struct se_cmd *se_cmd)\n" - buf += "{\n" - buf += " return;\n" - buf += "}\n\n" - bufi += "void " + fabric_mod_name + "_new_cmd_failure(struct se_cmd *);\n" - if re.search('queue_data_in\)\(', fo): buf += "int " + fabric_mod_name + "_queue_data_in(struct se_cmd *se_cmd)\n" buf += "{\n" @@ -939,15 +929,6 @@ def tcm_mod_dump_fabric_ops(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += "}\n\n" bufi += "int " + fabric_mod_name + "_is_state_remove(struct se_cmd *);\n" - if re.search('pack_lun\)\(', fo): - buf += "u64 " + fabric_mod_name + "_pack_lun(unsigned int lun)\n" - buf += "{\n" - buf += " WARN_ON(lun >= 256);\n" - buf += " /* Caller wants this byte-swapped */\n" - buf += " return cpu_to_le64((lun & 0xff) << 8);\n" - buf += "}\n\n" - bufi += "u64 " + fabric_mod_name + "_pack_lun(unsigned int);\n" - ret = p.write(buf) if ret: From ced38c8c00962c9b6c1fcd330572406162d150b9 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Tue, 6 Dec 2011 11:20:31 +0100 Subject: [PATCH 152/820] target: tcm_mod_builder: small fixups This includes: - remove on _ in "__NAMELEN" in $fabric _make_tport - target_fabric_configfs_init() returns an error pointer and not NULL anymore. Consider that. - replace (!(var_name)) with (!var_name). The extra () are not required - remove #ifdef MODULE. If the code is builtin it needs an init function or the code is useless - put exit/clean functions into __exit Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Nicholas Bellinger --- Documentation/target/tcm_mod_builder.py | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) diff --git a/Documentation/target/tcm_mod_builder.py b/Documentation/target/tcm_mod_builder.py index 520a923d7ac4..6e21b8b52638 100755 --- a/Documentation/target/tcm_mod_builder.py +++ b/Documentation/target/tcm_mod_builder.py @@ -255,7 +255,7 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " /* " + fabric_mod_name + "_parse_wwn(name, &wwpn, 1) < 0)\n" buf += " return ERR_PTR(-EINVAL); */\n" buf += " se_nacl_new = " + fabric_mod_name + "_alloc_fabric_acl(se_tpg);\n" - buf += " if (!(se_nacl_new))\n" + buf += " if (!se_nacl_new)\n" buf += " return ERR_PTR(-ENOMEM);\n" buf += "//#warning FIXME: Hardcoded nexus depth in " + fabric_mod_name + "_make_nodeacl()\n" buf += " nexus_depth = 1;\n" @@ -303,7 +303,7 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " if (strict_strtoul(name + 5, 10, &tpgt) || tpgt > UINT_MAX)\n" buf += " return ERR_PTR(-EINVAL);\n\n" buf += " tpg = kzalloc(sizeof(struct " + fabric_mod_name + "_tpg), GFP_KERNEL);\n" - buf += " if (!(tpg)) {\n" + buf += " if (!tpg) {\n" buf += " printk(KERN_ERR \"Unable to allocate struct " + fabric_mod_name + "_tpg\");\n" buf += " return ERR_PTR(-ENOMEM);\n" buf += " }\n" @@ -339,7 +339,7 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " /* if (" + fabric_mod_name + "_parse_wwn(name, &wwpn, 1) < 0)\n" buf += " return ERR_PTR(-EINVAL); */\n\n" buf += " " + fabric_mod_port + " = kzalloc(sizeof(struct " + fabric_mod_name + "_" + fabric_mod_port + "), GFP_KERNEL);\n" - buf += " if (!(" + fabric_mod_port + ")) {\n" + buf += " if (!" + fabric_mod_port + ") {\n" buf += " printk(KERN_ERR \"Unable to allocate struct " + fabric_mod_name + "_" + fabric_mod_port + "\");\n" buf += " return ERR_PTR(-ENOMEM);\n" buf += " }\n" @@ -347,7 +347,7 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): if proto_ident == "FC" or proto_ident == "SAS": buf += " " + fabric_mod_port + "->" + fabric_mod_port + "_wwpn = wwpn;\n" - buf += " /* " + fabric_mod_name + "_format_wwn(&" + fabric_mod_port + "->" + fabric_mod_port + "_name[0], " + fabric_mod_name.upper() + "__NAMELEN, wwpn); */\n\n" + buf += " /* " + fabric_mod_name + "_format_wwn(&" + fabric_mod_port + "->" + fabric_mod_port + "_name[0], " + fabric_mod_name.upper() + "_NAMELEN, wwpn); */\n\n" buf += " return &" + fabric_mod_port + "->" + fabric_mod_port + "_wwn;\n" buf += "}\n\n" buf += "static void " + fabric_mod_name + "_drop_" + fabric_mod_port + "(struct se_wwn *wwn)\n" @@ -431,9 +431,9 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " * Register the top level struct config_item_type with TCM core\n" buf += " */\n" buf += " fabric = target_fabric_configfs_init(THIS_MODULE, \"" + fabric_mod_name[4:] + "\");\n" - buf += " if (!(fabric)) {\n" + buf += " if (IS_ERR(fabric)) {\n" buf += " printk(KERN_ERR \"target_fabric_configfs_init() failed\\n\");\n" - buf += " return -ENOMEM;\n" + buf += " return PTR_ERR(fabric);\n" buf += " }\n" buf += " /*\n" buf += " * Setup fabric->tf_ops from our local " + fabric_mod_name + "_ops\n" @@ -467,9 +467,9 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " printk(KERN_INFO \"" + fabric_mod_name.upper() + "[0] - Set fabric -> " + fabric_mod_name + "_fabric_configfs\\n\");\n" buf += " return 0;\n" buf += "};\n\n" - buf += "static void " + fabric_mod_name + "_deregister_configfs(void)\n" + buf += "static void __exit " + fabric_mod_name + "_deregister_configfs(void)\n" buf += "{\n" - buf += " if (!(" + fabric_mod_name + "_fabric_configfs))\n" + buf += " if (!" + fabric_mod_name + "_fabric_configfs)\n" buf += " return;\n\n" buf += " target_fabric_configfs_deregister(" + fabric_mod_name + "_fabric_configfs);\n" buf += " " + fabric_mod_name + "_fabric_configfs = NULL;\n" @@ -484,17 +484,15 @@ def tcm_mod_build_configfs(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += " return ret;\n\n" buf += " return 0;\n" buf += "};\n\n" - buf += "static void " + fabric_mod_name + "_exit(void)\n" + buf += "static void __exit " + fabric_mod_name + "_exit(void)\n" buf += "{\n" buf += " " + fabric_mod_name + "_deregister_configfs();\n" buf += "};\n\n" - buf += "#ifdef MODULE\n" buf += "MODULE_DESCRIPTION(\"" + fabric_mod_name.upper() + " series fabric driver\");\n" buf += "MODULE_LICENSE(\"GPL\");\n" buf += "module_init(" + fabric_mod_name + "_init);\n" buf += "module_exit(" + fabric_mod_name + "_exit);\n" - buf += "#endif\n" ret = p.write(buf) if ret: @@ -776,7 +774,7 @@ def tcm_mod_dump_fabric_ops(proto_ident, fabric_mod_dir_var, fabric_mod_name): buf += "{\n" buf += " struct " + fabric_mod_name + "_nacl *nacl;\n\n" buf += " nacl = kzalloc(sizeof(struct " + fabric_mod_name + "_nacl), GFP_KERNEL);\n" - buf += " if (!(nacl)) {\n" + buf += " if (!nacl) {\n" buf += " printk(KERN_ERR \"Unable to alocate struct " + fabric_mod_name + "_nacl\\n\");\n" buf += " return NULL;\n" buf += " }\n\n" From ce136176fea522fc8f4c16dcae7e8ed1d890ca39 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 6 Dec 2011 10:02:09 -0800 Subject: [PATCH 153/820] target: Set response format in INQUIRY response Current SCSI specs say that the "response format" field in the standard INQUIRY response should be set to 2, and all the real SCSI devices I have do put 2 here. So let's do that too. Signed-off-by: Roland Dreier Cc: stable@kernel.org Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_cdb.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/target/target_core_cdb.c b/drivers/target/target_core_cdb.c index 5c60d0a97a50..2f2235edefff 100644 --- a/drivers/target/target_core_cdb.c +++ b/drivers/target/target_core_cdb.c @@ -94,6 +94,18 @@ target_emulate_inquiry_std(struct se_cmd *cmd) } buf[2] = dev->transport->get_device_rev(dev); + /* + * NORMACA and HISUP = 0, RESPONSE DATA FORMAT = 2 + * + * SPC4 says: + * A RESPONSE DATA FORMAT field set to 2h indicates that the + * standard INQUIRY data is in the format defined in this + * standard. Response data format values less than 2h are + * obsolete. Response data format values greater than 2h are + * reserved. + */ + buf[3] = 2; + /* * Enable SCCS and TPGS fields for Emulated ALUA */ From 7481deb413be132a22193e8a0bce88b311ecb3c2 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Sat, 12 Nov 2011 00:32:17 -0800 Subject: [PATCH 154/820] target: Make target_put_sess_cmd use target_release_cmd_kref This patch moves target_put_sess_cmd() to use a se_cmd->cmd_kref callback target_release_cmd_kref when performing driver release of fabric->se_cmd descriptor memory. It sets the default cmd_kref count value to '2' within target_get_sess_cmd() setup, and currently assumes TFO->check_stop_free() usage. It drops se_tfo->check_release_cmd() usage in the main transport_release_cmd codepath. Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 39 ++++++++++++++++---------- include/target/target_core_base.h | 1 + 2 files changed, 25 insertions(+), 15 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 122925686de2..176f956665e8 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -3349,13 +3349,13 @@ static void transport_release_cmd(struct se_cmd *cmd) if (cmd->t_task_cdb != cmd->__t_task_cdb) kfree(cmd->t_task_cdb); /* - * Check if target_wait_for_sess_cmds() is expecting to - * release se_cmd directly here.. + * If this cmd has been setup with target_get_sess_cmd(), drop + * the kref and call ->release_cmd() in kref callback. */ - if (cmd->check_release != 0 && cmd->se_tfo->check_release_cmd) - if (cmd->se_tfo->check_release_cmd(cmd) != 0) - return; - + if (cmd->check_release != 0) { + target_put_sess_cmd(cmd->se_sess, cmd); + return; + } cmd->se_tfo->release_cmd(cmd); } @@ -3915,6 +3915,9 @@ void target_get_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd) { unsigned long flags; + kref_init(&se_cmd->cmd_kref); + kref_get(&se_cmd->cmd_kref); + spin_lock_irqsave(&se_sess->sess_cmd_lock, flags); list_add_tail(&se_cmd->se_cmd_list, &se_sess->sess_cmd_list); se_cmd->check_release = 1; @@ -3922,30 +3925,36 @@ void target_get_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd) } EXPORT_SYMBOL(target_get_sess_cmd); -/* target_put_sess_cmd - Check for active I/O shutdown or list delete - * @se_sess: session to reference - * @se_cmd: command descriptor to drop - */ -int target_put_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd) +static void target_release_cmd_kref(struct kref *kref) { + struct se_cmd *se_cmd = container_of(kref, struct se_cmd, cmd_kref); + struct se_session *se_sess = se_cmd->se_sess; unsigned long flags; spin_lock_irqsave(&se_sess->sess_cmd_lock, flags); if (list_empty(&se_cmd->se_cmd_list)) { spin_unlock_irqrestore(&se_sess->sess_cmd_lock, flags); WARN_ON(1); - return 0; + return; } - if (se_sess->sess_tearing_down && se_cmd->cmd_wait_set) { spin_unlock_irqrestore(&se_sess->sess_cmd_lock, flags); complete(&se_cmd->cmd_wait_comp); - return 1; + return; } list_del(&se_cmd->se_cmd_list); spin_unlock_irqrestore(&se_sess->sess_cmd_lock, flags); - return 0; + se_cmd->se_tfo->release_cmd(se_cmd); +} + +/* target_put_sess_cmd - Check for active I/O shutdown via kref_put + * @se_sess: session to reference + * @se_cmd: command descriptor to drop + */ +int target_put_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd) +{ + return kref_put(&se_cmd->cmd_kref, target_release_cmd_kref); } EXPORT_SYMBOL(target_put_sess_cmd); diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index cadd4f6e57e1..af088a9e4905 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -541,6 +541,7 @@ struct se_cmd { struct list_head se_queue_node; struct list_head se_cmd_list; struct completion cmd_wait_comp; + struct kref cmd_kref; struct target_core_fabric_ops *se_tfo; int (*execute_task)(struct se_task *); void (*transport_complete_callback)(struct se_cmd *); From a63607855224702ea17e6016ecf3f7d544e83625 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Fri, 18 Nov 2011 20:36:22 -0800 Subject: [PATCH 155/820] target: Add target_submit_cmd() for process context fabric submission This patch adds a target_submit_cmd() caller that can be used by fabrics to submit an uninitialized se_cmd descriptor to an struct se_session + unpacked_lun from workqueue process context. This call will invoke the following steps: - transport_init_se_cmd() to setup se_cmd specific pointers - Obtain se_cmd->cmd_kref references with target_get_sess_cmd() - set se_cmd->t_tasks_bidi - transport_lookup_cmd_lun() to setup struct se_cmd->se_lun from the passed unpacked_lun - transport_generic_allocate_tasks() to setup the passed *cdb, and - transport_handle_cdb_direct() handle READ dispatch or WRITE ready-to-transfer callback to fabric v2 changes from hch feedback: *) Add target_sc_flags_table for target_submit_cmd flags *) Rename bidi parameter to flags, add TARGET_SCF_BIDI_OP *) Convert checks to BUG_ON *) Add out_check_cond for transport_send_check_condition_and_sense usage v3 changes: *) Add TARGET_SCF_ACK_KREF for target_submit_cmd into target_get_sess_cmd to determine when the fabric caller is expecting a second kref_put() from fabric packet acknowledgement. Cc: Christoph Hellwig Cc: Roland Dreier Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 86 +++++++++++++++++++++++++- include/target/target_core_base.h | 5 ++ include/target/target_core_fabric.h | 4 +- 3 files changed, 92 insertions(+), 3 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 176f956665e8..e4389d40c057 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -1642,6 +1642,80 @@ int transport_handle_cdb_direct( } EXPORT_SYMBOL(transport_handle_cdb_direct); +/** + * target_submit_cmd - lookup unpacked lun and submit uninitialized se_cmd + * + * @se_cmd: command descriptor to submit + * @se_sess: associated se_sess for endpoint + * @cdb: pointer to SCSI CDB + * @sense: pointer to SCSI sense buffer + * @unpacked_lun: unpacked LUN to reference for struct se_lun + * @data_length: fabric expected data transfer length + * @task_addr: SAM task attribute + * @data_dir: DMA data direction + * @flags: flags for command submission from target_sc_flags_tables + * + * This may only be called from process context, and also currently + * assumes internal allocation of fabric payload buffer by target-core. + **/ +int target_submit_cmd(struct se_cmd *se_cmd, struct se_session *se_sess, + unsigned char *cdb, unsigned char *sense, u32 unpacked_lun, + u32 data_length, int task_attr, int data_dir, int flags) +{ + struct se_portal_group *se_tpg; + int rc; + + se_tpg = se_sess->se_tpg; + BUG_ON(!se_tpg); + BUG_ON(se_cmd->se_tfo || se_cmd->se_sess); + BUG_ON(in_interrupt()); + /* + * Initialize se_cmd for target operation. From this point + * exceptions are handled by sending exception status via + * target_core_fabric_ops->queue_status() callback + */ + transport_init_se_cmd(se_cmd, se_tpg->se_tpg_tfo, se_sess, + data_length, data_dir, task_attr, sense); + /* + * Obtain struct se_cmd->cmd_kref reference and add new cmd to + * se_sess->sess_cmd_list. A second kref_get here is necessary + * for fabrics using TARGET_SCF_ACK_KREF that expect a second + * kref_put() to happen during fabric packet acknowledgement. + */ + target_get_sess_cmd(se_sess, se_cmd, (flags & TARGET_SCF_ACK_KREF)); + /* + * Signal bidirectional data payloads to target-core + */ + if (flags & TARGET_SCF_BIDI_OP) + se_cmd->se_cmd_flags |= SCF_BIDI; + /* + * Locate se_lun pointer and attach it to struct se_cmd + */ + if (transport_lookup_cmd_lun(se_cmd, unpacked_lun) < 0) + goto out_check_cond; + /* + * Sanitize CDBs via transport_generic_cmd_sequencer() and + * allocate the necessary tasks to complete the received CDB+data + */ + rc = transport_generic_allocate_tasks(se_cmd, cdb); + if (rc != 0) + goto out_check_cond; + /* + * Dispatch se_cmd descriptor to se_lun->lun_se_dev backend + * for immediate execution of READs, otherwise wait for + * transport_generic_handle_data() to be called for WRITEs + * when fabric has filled the incoming buffer. + */ + transport_handle_cdb_direct(se_cmd); + return 0; + +out_check_cond: + transport_send_check_condition_and_sense(se_cmd, + se_cmd->scsi_sense_reason, 0); + return 0; +} +EXPORT_SYMBOL(target_submit_cmd); + /* * Used by fabric module frontends defining a TFO->new_cmd_map() caller * to queue up a newly setup se_cmd w/ TRANSPORT_NEW_CMD_MAP in order to @@ -3910,13 +3984,21 @@ EXPORT_SYMBOL(transport_generic_free_cmd); /* target_get_sess_cmd - Add command to active ->sess_cmd_list * @se_sess: session to reference * @se_cmd: command descriptor to add + * @ack_kref: Signal that fabric will perform an ack target_put_sess_cmd() */ -void target_get_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd) +void target_get_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd, + bool ack_kref) { unsigned long flags; kref_init(&se_cmd->cmd_kref); - kref_get(&se_cmd->cmd_kref); + /* + * Add a second kref if the fabric caller is expecting to handle + * fabric acknowledgement that requires two target_put_sess_cmd() + * invocations before se_cmd descriptor release. + */ + if (ack_kref == true) + kref_get(&se_cmd->cmd_kref); spin_lock_irqsave(&se_sess->sess_cmd_lock, flags); list_add_tail(&se_cmd->se_cmd_list, &se_sess->sess_cmd_list); diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index af088a9e4905..28190dc10eef 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -226,6 +226,11 @@ enum tcm_sense_reason_table { TCM_RESERVATION_CONFLICT = 0x10, }; +enum target_sc_flags_table { + TARGET_SCF_BIDI_OP = 0x01, + TARGET_SCF_ACK_KREF = 0x02, +}; + /* fabric independent task management function values */ enum tcm_tmreq_table { TMR_ABORT_TASK = 1, diff --git a/include/target/target_core_fabric.h b/include/target/target_core_fabric.h index d035d865dbef..aaa26da5a527 100644 --- a/include/target/target_core_fabric.h +++ b/include/target/target_core_fabric.h @@ -118,6 +118,8 @@ void transport_init_se_cmd(struct se_cmd *, struct target_core_fabric_ops *, struct se_session *, u32, int, int, unsigned char *); int transport_lookup_cmd_lun(struct se_cmd *, u32); int transport_generic_allocate_tasks(struct se_cmd *, unsigned char *); +int target_submit_cmd(struct se_cmd *, struct se_session *, unsigned char *, + unsigned char *, u32, u32, int, int, int); int transport_handle_cdb_direct(struct se_cmd *); int transport_generic_handle_cdb_map(struct se_cmd *); int transport_generic_handle_data(struct se_cmd *); @@ -134,7 +136,7 @@ bool transport_wait_for_tasks(struct se_cmd *); int transport_check_aborted_status(struct se_cmd *, int); int transport_send_check_condition_and_sense(struct se_cmd *, u8, int); -void target_get_sess_cmd(struct se_session *, struct se_cmd *); +void target_get_sess_cmd(struct se_session *, struct se_cmd *, bool); int target_put_sess_cmd(struct se_session *, struct se_cmd *); void target_splice_sess_cmd_list(struct se_session *); void target_wait_for_sess_cmds(struct se_session *, int); From 4355a9110eeb2eaf1dd44fcab16ccbd1c8c5fad4 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Tue, 22 Nov 2011 21:41:54 -0800 Subject: [PATCH 156/820] tcm_fc: Convert ft_send_work to use target_submit_cmd This patch converts the main ft_send_work() I/O path to use target_submit_cmd() with a single se_cmd->cmd_kref reference that is released via the existing ft_check_stop_free() response path callback. It also makes ft_send_tm() use transport_init_se_cmd() and target_get_sess_cmd() to also use single se_cmd->cmd_kref reference. Cc: Christoph Hellwig Cc: Kiran Patil Signed-off-by: Nicholas Bellinger --- drivers/target/tcm_fc/tfc_cmd.c | 51 +++++++++------------------------ 1 file changed, 13 insertions(+), 38 deletions(-) diff --git a/drivers/target/tcm_fc/tfc_cmd.c b/drivers/target/tcm_fc/tfc_cmd.c index dbbbc8376d01..addc18f727ea 100644 --- a/drivers/target/tcm_fc/tfc_cmd.c +++ b/drivers/target/tcm_fc/tfc_cmd.c @@ -363,6 +363,11 @@ static void ft_send_tm(struct ft_cmd *cmd) struct ft_sess *sess; u8 tm_func; + transport_init_se_cmd(&cmd->se_cmd, &ft_configfs->tf_ops, + cmd->sess->se_sess, 0, DMA_NONE, 0, + &cmd->ft_sense_buffer[0]); + target_get_sess_cmd(cmd->sess->se_sess, &cmd->se_cmd, false); + fcp = fc_frame_payload_get(cmd->req_frame, sizeof(*fcp)); switch (fcp->fc_tm_flags) { @@ -416,7 +421,6 @@ static void ft_send_tm(struct ft_cmd *cmd) sess = cmd->sess; transport_send_check_condition_and_sense(&cmd->se_cmd, cmd->se_cmd.scsi_sense_reason, 0); - transport_generic_free_cmd(&cmd->se_cmd, 0); ft_sess_put(sess); return; } @@ -532,7 +536,6 @@ static void ft_send_work(struct work_struct *work) { struct ft_cmd *cmd = container_of(work, struct ft_cmd, work); struct fc_frame_header *fh = fc_frame_header_get(cmd->req_frame); - struct se_cmd *se_cmd; struct fcp_cmnd *fcp; int data_dir = 0; u32 data_len; @@ -587,15 +590,6 @@ static void ft_send_work(struct work_struct *work) data_len = ntohl(fcp->fc_dl); cmd->cdb = fcp->fc_cdb; } - - se_cmd = &cmd->se_cmd; - /* - * Initialize struct se_cmd descriptor from target_core_mod - * infrastructure - */ - transport_init_se_cmd(se_cmd, &ft_configfs->tf_ops, cmd->sess->se_sess, - data_len, data_dir, task_attr, - &cmd->ft_sense_buffer[0]); /* * Check for FCP task management flags */ @@ -603,39 +597,20 @@ static void ft_send_work(struct work_struct *work) ft_send_tm(cmd); return; } - fc_seq_exch(cmd->seq)->lp->tt.seq_set_resp(cmd->seq, ft_recv_seq, cmd); - cmd->lun = scsilun_to_int((struct scsi_lun *)fcp->fc_lun); - ret = transport_lookup_cmd_lun(&cmd->se_cmd, cmd->lun); + /* + * Use a single se_cmd->cmd_kref as we expect to release se_cmd + * directly from ft_check_stop_free callback in response path. + */ + ret = target_submit_cmd(&cmd->se_cmd, cmd->sess->se_sess, cmd->cdb, + &cmd->ft_sense_buffer[0], cmd->lun, data_len, + task_attr, data_dir, 0); + pr_debug("r_ctl %x alloc target_submit_cmd %d\n", fh->fh_r_ctl, ret); if (ret < 0) { ft_dump_cmd(cmd, __func__); - transport_send_check_condition_and_sense(&cmd->se_cmd, - cmd->se_cmd.scsi_sense_reason, 0); return; } - - ret = transport_generic_allocate_tasks(se_cmd, cmd->cdb); - - pr_debug("r_ctl %x alloc task ret %d\n", fh->fh_r_ctl, ret); - ft_dump_cmd(cmd, __func__); - - if (ret == -ENOMEM) { - transport_send_check_condition_and_sense(se_cmd, - TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE, 0); - transport_generic_free_cmd(se_cmd, 0); - return; - } - if (ret == -EINVAL) { - if (se_cmd->se_cmd_flags & SCF_SCSI_RESERVATION_CONFLICT) - ft_queue_status(se_cmd); - else - transport_send_check_condition_and_sense(se_cmd, - se_cmd->scsi_sense_reason, 0); - transport_generic_free_cmd(se_cmd, 0); - return; - } - transport_handle_cdb_direct(se_cmd); return; err: From ec54cc081ead14e85736c6543e36ec59be8a7d3d Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 14 Dec 2011 10:50:16 +0000 Subject: [PATCH 157/820] target: Remove TFO->check_release_cmd() fabric API caller Remove the now unused target_core_fabric_ops->check_release_cmd() as target_core handles this directly for se_cmd->cmd_kref objects now. Cc: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- include/target/target_core_fabric.h | 4 ---- 1 file changed, 4 deletions(-) diff --git a/include/target/target_core_fabric.h b/include/target/target_core_fabric.h index aaa26da5a527..523e8bc104d4 100644 --- a/include/target/target_core_fabric.h +++ b/include/target/target_core_fabric.h @@ -52,10 +52,6 @@ struct target_core_fabric_ops { * Returning 0 will signal a descriptor has not been released. */ int (*check_stop_free)(struct se_cmd *); - /* - * Optional check for active I/O shutdown - */ - int (*check_release_cmd)(struct se_cmd *); void (*release_cmd)(struct se_cmd *); /* * Called with spin_lock_bh(struct se_portal_group->session_lock held. From 40be67f4c588fe2f3e2dbd60ae1f470abc5b6ad8 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 30 Nov 2011 00:41:20 -0800 Subject: [PATCH 158/820] target: Fix possible NULL pointer with __transport_execute_tasks This patch makes __transport_execute_tasks() use a local *se_dev reference to prevent direct se_cmd->se_dev access after transport_cmd_check_stop() -> transport_add_tasks_from_cmd() has been called, as in the current implementation we can expect __transport_execute_tasks() may be called from another context that may have already completed the I/O. Cc: Christoph Hellwig Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index e4389d40c057..1cc7e920ab0b 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -2076,6 +2076,7 @@ static inline int transport_execute_task_attr(struct se_cmd *cmd) static int transport_execute_tasks(struct se_cmd *cmd) { int add_tasks; + struct se_device *se_dev = cmd->se_dev; if (se_dev_check_online(cmd->se_dev) != 0) { cmd->scsi_sense_reason = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE; @@ -2108,7 +2109,7 @@ static int transport_execute_tasks(struct se_cmd *cmd) * storage object. */ execute_tasks: - __transport_execute_tasks(cmd->se_dev); + __transport_execute_tasks(se_dev); return 0; } From 65586d51e0986be574118286c3d0007e903a2add Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 30 Nov 2011 01:25:21 -0800 Subject: [PATCH 159/820] target: Drop se_device TCQ queue_depth usage from I/O path Historically, pSCSI devices have been the ones that required target-core to enforce a per se_device->depth_left. This patch changes target-core to no longer (by default) enforce a per se_device->depth_left or sleep in transport_tcq_window_closed() when we out of queue slots for all backend export cases. Cc: Christoph Hellwig Cc: Roland Dreier Cc: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_device.c | 7 ----- drivers/target/target_core_pscsi.c | 1 - drivers/target/target_core_transport.c | 39 ++------------------------ include/target/target_core_base.h | 1 - 4 files changed, 2 insertions(+), 46 deletions(-) diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index 1f74de25a92f..0c5992f0d946 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -1132,8 +1132,6 @@ int se_dev_set_emulate_rest_reord(struct se_device *dev, int flag) */ int se_dev_set_queue_depth(struct se_device *dev, u32 queue_depth) { - u32 orig_queue_depth = dev->queue_depth; - if (atomic_read(&dev->dev_export_obj.obj_access_count)) { pr_err("dev[%p]: Unable to change SE Device TCQ while" " dev_export_obj: %d count exists\n", dev, @@ -1167,11 +1165,6 @@ int se_dev_set_queue_depth(struct se_device *dev, u32 queue_depth) } dev->se_sub_dev->se_dev_attrib.queue_depth = dev->queue_depth = queue_depth; - if (queue_depth > orig_queue_depth) - atomic_add(queue_depth - orig_queue_depth, &dev->depth_left); - else if (queue_depth < orig_queue_depth) - atomic_sub(orig_queue_depth - queue_depth, &dev->depth_left); - pr_debug("dev[%p]: SE Device TCQ Depth changed to: %u\n", dev, queue_depth); return 0; diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c index b73a399cdd54..d35467d42e12 100644 --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -350,7 +350,6 @@ static struct se_device *pscsi_add_device_to_list( * scsi_device_put() and the pdv->pdv_sd cleared. */ pdv->pdv_sd = sd; - dev = transport_add_device_to_core_hba(hba, &pscsi_template, se_dev, dev_flags, pdv, &dev_limits, NULL, NULL); diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 1cc7e920ab0b..7c2def7e2593 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -691,12 +691,6 @@ void transport_complete_task(struct se_task *task, int success) struct se_cmd *cmd = task->task_se_cmd; struct se_device *dev = cmd->se_dev; unsigned long flags; -#if 0 - pr_debug("task: %p CDB: 0x%02x obj_ptr: %p\n", task, - cmd->t_task_cdb[0], dev); -#endif - if (dev) - atomic_inc(&dev->depth_left); spin_lock_irqsave(&cmd->t_state_lock, flags); task->task_flags &= ~TF_ACTIVE; @@ -971,9 +965,8 @@ void transport_dump_dev_state( break; } - *bl += sprintf(b + *bl, " Execute/Left/Max Queue Depth: %d/%d/%d", - atomic_read(&dev->execute_tasks), atomic_read(&dev->depth_left), - dev->queue_depth); + *bl += sprintf(b + *bl, " Execute/Max Queue Depth: %d/%d", + atomic_read(&dev->execute_tasks), dev->queue_depth); *bl += sprintf(b + *bl, " SectorSize: %u MaxSectors: %u\n", dev->se_sub_dev->se_dev_attrib.block_size, dev->se_sub_dev->se_dev_attrib.max_sectors); *bl += sprintf(b + *bl, " "); @@ -1328,9 +1321,6 @@ struct se_device *transport_add_device_to_core_hba( spin_lock_init(&dev->se_port_lock); spin_lock_init(&dev->se_tmr_lock); spin_lock_init(&dev->qf_cmd_lock); - - dev->queue_depth = dev_limits->queue_depth; - atomic_set(&dev->depth_left, dev->queue_depth); atomic_set(&dev->dev_ordered_id, 0); se_dev_set_default_attribs(dev, dev_limits); @@ -1982,18 +1972,6 @@ static void transport_set_supported_SAM_opcode(struct se_cmd *se_cmd) spin_unlock_irqrestore(&se_cmd->t_state_lock, flags); } -static inline int transport_tcq_window_closed(struct se_device *dev) -{ - if (dev->dev_tcq_window_closed++ < - PYX_TRANSPORT_WINDOW_CLOSED_THRESHOLD) { - msleep(PYX_TRANSPORT_WINDOW_CLOSED_WAIT_SHORT); - } else - msleep(PYX_TRANSPORT_WINDOW_CLOSED_WAIT_LONG); - - wake_up_interruptible(&dev->dev_queue_obj.thread_wq); - return 0; -} - /* * Called from Fabric Module context from transport_execute_tasks() * @@ -2126,16 +2104,7 @@ static int __transport_execute_tasks(struct se_device *dev) struct se_task *task = NULL; unsigned long flags; - /* - * Check if there is enough room in the device and HBA queue to send - * struct se_tasks to the selected transport. - */ check_depth: - if (!atomic_read(&dev->depth_left)) - return transport_tcq_window_closed(dev); - - dev->dev_tcq_window_closed = 0; - spin_lock_irq(&dev->execute_task_lock); if (list_empty(&dev->execute_task_list)) { spin_unlock_irq(&dev->execute_task_lock); @@ -2146,10 +2115,7 @@ check_depth: __transport_remove_task_from_execute_queue(task, dev); spin_unlock_irq(&dev->execute_task_lock); - atomic_dec(&dev->depth_left); - cmd = task->task_se_cmd; - spin_lock_irqsave(&cmd->t_state_lock, flags); task->task_flags |= (TF_ACTIVE | TF_SENT); atomic_inc(&cmd->t_task_cdbs_sent); @@ -2170,7 +2136,6 @@ check_depth: spin_unlock_irqrestore(&cmd->t_state_lock, flags); atomic_set(&cmd->t_transport_sent, 0); transport_stop_tasks_for_cmd(cmd); - atomic_inc(&dev->depth_left); transport_generic_request_failure(cmd); } diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index 28190dc10eef..cd4caf3a598f 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -798,7 +798,6 @@ struct se_device { spinlock_t stats_lock; /* Active commands on this virtual SE device */ atomic_t simple_cmds; - atomic_t depth_left; atomic_t dev_ordered_id; atomic_t execute_tasks; atomic_t dev_ordered_sync; From 4d2300ccffd22d1d0213b6a8e4d685eb6ca069c0 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 30 Nov 2011 18:18:33 -0800 Subject: [PATCH 160/820] target: Remove extra se_device->execute_task_lock access in fast path This patch makes __transport_execute_tasks() perform the addition of tasks to dev->execute_task_list via __transport_add_tasks_from_cmd() while holding dev->execute_task_lock during normal I/O fast path submission. It effectively removes the unnecessary re-acquire of dev->execute_task_lock during transport_execute_tasks() -> transport_add_tasks_from_cmd() ahead of calling __transport_execute_tasks() to queue tasks for the passed *se_cmd descriptor. (v2: Re-add goto check_depth usage for multi-task submission for now..) Cc: Christoph Hellwig Cc: Roland Dreier Cc: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 40 +++++++++++++++----------- include/target/target_core_base.h | 1 - 2 files changed, 24 insertions(+), 17 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 7c2def7e2593..46e25118ab25 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -68,7 +68,7 @@ struct kmem_cache *t10_alua_tg_pt_gp_mem_cache; static int transport_generic_write_pending(struct se_cmd *); static int transport_processing_thread(void *param); -static int __transport_execute_tasks(struct se_device *dev); +static int __transport_execute_tasks(struct se_device *dev, struct se_cmd *); static void transport_complete_task_attr(struct se_cmd *cmd); static void transport_handle_queue_full(struct se_cmd *cmd, struct se_device *dev); @@ -851,13 +851,11 @@ static void transport_add_tasks_to_state_queue(struct se_cmd *cmd) spin_unlock_irqrestore(&cmd->t_state_lock, flags); } -static void transport_add_tasks_from_cmd(struct se_cmd *cmd) +static void __transport_add_tasks_from_cmd(struct se_cmd *cmd) { struct se_device *dev = cmd->se_dev; struct se_task *task, *task_prev = NULL; - unsigned long flags; - spin_lock_irqsave(&dev->execute_task_lock, flags); list_for_each_entry(task, &cmd->t_task_list, t_list) { if (!list_empty(&task->t_execute_list)) continue; @@ -868,6 +866,15 @@ static void transport_add_tasks_from_cmd(struct se_cmd *cmd) __transport_add_task_to_execute_queue(task, task_prev, dev); task_prev = task; } +} + +static void transport_add_tasks_from_cmd(struct se_cmd *cmd) +{ + unsigned long flags; + struct se_device *dev = cmd->se_dev; + + spin_lock_irqsave(&dev->execute_task_lock, flags); + __transport_add_tasks_from_cmd(cmd); spin_unlock_irqrestore(&dev->execute_task_lock, flags); } @@ -2075,19 +2082,16 @@ static int transport_execute_tasks(struct se_cmd *cmd) if (!add_tasks) goto execute_tasks; /* - * This calls transport_add_tasks_from_cmd() to handle - * HEAD_OF_QUEUE ordering for SAM Task Attribute emulation - * (if enabled) in __transport_add_task_to_execute_queue() and - * transport_add_task_check_sam_attr(). + * __transport_execute_tasks() -> __transport_add_tasks_from_cmd() + * adds associated se_tasks while holding dev->execute_task_lock + * before I/O dispath to avoid a double spinlock access. */ - transport_add_tasks_from_cmd(cmd); + __transport_execute_tasks(se_dev, cmd); + return 0; } - /* - * Kick the execution queue for the cmd associated struct se_device - * storage object. - */ + execute_tasks: - __transport_execute_tasks(se_dev); + __transport_execute_tasks(se_dev, NULL); return 0; } @@ -2097,7 +2101,7 @@ execute_tasks: * * Called from transport_processing_thread() */ -static int __transport_execute_tasks(struct se_device *dev) +static int __transport_execute_tasks(struct se_device *dev, struct se_cmd *new_cmd) { int error; struct se_cmd *cmd = NULL; @@ -2106,6 +2110,9 @@ static int __transport_execute_tasks(struct se_device *dev) check_depth: spin_lock_irq(&dev->execute_task_lock); + if (new_cmd != NULL) + __transport_add_tasks_from_cmd(new_cmd); + if (list_empty(&dev->execute_task_list)) { spin_unlock_irq(&dev->execute_task_lock); return 0; @@ -2139,6 +2146,7 @@ check_depth: transport_generic_request_failure(cmd); } + new_cmd = NULL; goto check_depth; return 0; @@ -4647,7 +4655,7 @@ static int transport_processing_thread(void *param) goto out; get_cmd: - __transport_execute_tasks(dev); + __transport_execute_tasks(dev, NULL); cmd = transport_get_cmd_from_queue(&dev->dev_queue_obj); if (!cmd) diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index cd4caf3a598f..a74cb2be74de 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -781,7 +781,6 @@ struct se_device { u32 dev_port_count; /* See transport_device_status_table */ u32 dev_status; - u32 dev_tcq_window_closed; /* Physical device queue depth */ u32 queue_depth; /* Used for SPC-2 reservations enforce of ISIDs */ From beb55a0cc15a43be43a646b588fcf83822f0c44f Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 30 Nov 2011 18:21:24 -0800 Subject: [PATCH 161/820] target: Remove __transport_execute_tasks() for each processing context This patch removes the original usage of __transport_execute_tasks() ahead of every transport_get_cmd_from_queue() call in transport_processing_thread(). This helps reduce se_device->execute_task_lock contention between qla2xxx wq with target_submit_cmd() for READs and transport_processing_thread() context servicing WRITEs with full payloads for I/O submission. It also adds a __transport_execute_tasks() to kick the task queue again without a *se_cmd descriptor with existing queue full logic, but this may end up not being necessary. Cc: Christoph Hellwig Cc: Roland Dreier Cc: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 46e25118ab25..373e01f523dd 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -4655,8 +4655,6 @@ static int transport_processing_thread(void *param) goto out; get_cmd: - __transport_execute_tasks(dev, NULL); - cmd = transport_get_cmd_from_queue(&dev->dev_queue_obj); if (!cmd) continue; From 6d5b59756033c2a029bde7262bb5f8d45f4ca952 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 30 Nov 2011 18:24:03 -0800 Subject: [PATCH 162/820] target: Remove legacy device status check from transport_execute_tasks This patch removes a legacy se_dev_check_online() check from within transport_execute_tasks() that should no longer be necessary as transport_lookup_cmd_lun() is already making this call. Using transport_cmd_check_stop() from transport_execute_tasks() should already be checking per se_cmd context for each descriptor upon active I/O shutdown, so no need to acquire dev->dev_status_lock again while executing se_task submission. Cc: Christoph Hellwig Cc: Roland Dreier Cc: Joern Engel Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 373e01f523dd..b5fdacce0671 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -2062,13 +2062,6 @@ static int transport_execute_tasks(struct se_cmd *cmd) { int add_tasks; struct se_device *se_dev = cmd->se_dev; - - if (se_dev_check_online(cmd->se_dev) != 0) { - cmd->scsi_sense_reason = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE; - transport_generic_request_failure(cmd); - return 0; - } - /* * Call transport_cmd_check_stop() to see if a fabric exception * has occurred that prevents execution. From 895f3022523361e9b383cf48f51feb1f7d5e7e53 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 13 Dec 2011 14:55:33 -0800 Subject: [PATCH 163/820] target: Set additional sense length field in sense data The target code was not setting the additional sense length field in the sense data it returned, which meant that at least the Linux stack ignored the ASC/ASCQ fields. For example, without this patch, on a tcm_loop device: # sg_raw -v /dev/sda 2 0 0 0 0 0 gives cdb to send: 02 00 00 00 00 00 SCSI Status: Check Condition Sense Information: Fixed format, current; Sense key: Illegal Request Raw sense data (in hex): 70 00 05 00 00 00 00 00 while after the patch we correctly get the following (which matches what a regular disk returns): cdb to send: 02 00 00 00 00 00 SCSI Status: Check Condition Sense Information: Fixed format, current; Sense key: Illegal Request Additional sense: Invalid command operation code Raw sense data (in hex): 70 00 05 00 00 00 00 0a 00 00 00 00 20 00 00 00 00 00 Signed-off-by: Roland Dreier Cc: stable@kernel.org Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 14 ++++++++++++++ include/target/target_core_base.h | 1 + 2 files changed, 15 insertions(+) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index b5fdacce0671..d3ddd1361949 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -4398,6 +4398,7 @@ int transport_send_check_condition_and_sense( case TCM_NON_EXISTENT_LUN: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ILLEGAL REQUEST */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ILLEGAL_REQUEST; /* LOGICAL UNIT NOT SUPPORTED */ @@ -4407,6 +4408,7 @@ int transport_send_check_condition_and_sense( case TCM_SECTOR_COUNT_TOO_MANY: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ILLEGAL REQUEST */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ILLEGAL_REQUEST; /* INVALID COMMAND OPERATION CODE */ @@ -4415,6 +4417,7 @@ int transport_send_check_condition_and_sense( case TCM_UNKNOWN_MODE_PAGE: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ILLEGAL REQUEST */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ILLEGAL_REQUEST; /* INVALID FIELD IN CDB */ @@ -4423,6 +4426,7 @@ int transport_send_check_condition_and_sense( case TCM_CHECK_CONDITION_ABORT_CMD: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* BUS DEVICE RESET FUNCTION OCCURRED */ @@ -4432,6 +4436,7 @@ int transport_send_check_condition_and_sense( case TCM_INCORRECT_AMOUNT_OF_DATA: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* WRITE ERROR */ @@ -4442,6 +4447,7 @@ int transport_send_check_condition_and_sense( case TCM_INVALID_CDB_FIELD: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* INVALID FIELD IN CDB */ @@ -4450,6 +4456,7 @@ int transport_send_check_condition_and_sense( case TCM_INVALID_PARAMETER_LIST: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* INVALID FIELD IN PARAMETER LIST */ @@ -4458,6 +4465,7 @@ int transport_send_check_condition_and_sense( case TCM_UNEXPECTED_UNSOLICITED_DATA: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* WRITE ERROR */ @@ -4468,6 +4476,7 @@ int transport_send_check_condition_and_sense( case TCM_SERVICE_CRC_ERROR: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* PROTOCOL SERVICE CRC ERROR */ @@ -4478,6 +4487,7 @@ int transport_send_check_condition_and_sense( case TCM_SNACK_REJECTED: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ABORTED COMMAND */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ABORTED_COMMAND; /* READ ERROR */ @@ -4488,6 +4498,7 @@ int transport_send_check_condition_and_sense( case TCM_WRITE_PROTECTED: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* DATA PROTECT */ buffer[offset+SPC_SENSE_KEY_OFFSET] = DATA_PROTECT; /* WRITE PROTECTED */ @@ -4496,6 +4507,7 @@ int transport_send_check_condition_and_sense( case TCM_CHECK_CONDITION_UNIT_ATTENTION: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* UNIT ATTENTION */ buffer[offset+SPC_SENSE_KEY_OFFSET] = UNIT_ATTENTION; core_scsi3_ua_for_check_condition(cmd, &asc, &ascq); @@ -4505,6 +4517,7 @@ int transport_send_check_condition_and_sense( case TCM_CHECK_CONDITION_NOT_READY: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* Not Ready */ buffer[offset+SPC_SENSE_KEY_OFFSET] = NOT_READY; transport_get_sense_codes(cmd, &asc, &ascq); @@ -4515,6 +4528,7 @@ int transport_send_check_condition_and_sense( default: /* CURRENT ERROR */ buffer[offset] = 0x70; + buffer[offset+SPC_ADD_SENSE_LEN_OFFSET] = 10; /* ILLEGAL REQUEST */ buffer[offset+SPC_SENSE_KEY_OFFSET] = ILLEGAL_REQUEST; /* LOGICAL UNIT COMMUNICATION FAILURE */ diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h index a74cb2be74de..daf532bc721a 100644 --- a/include/target/target_core_base.h +++ b/include/target/target_core_base.h @@ -35,6 +35,7 @@ #define TRANSPORT_SENSE_BUFFER SCSI_SENSE_BUFFERSIZE /* Used by transport_send_check_condition_and_sense() */ #define SPC_SENSE_KEY_OFFSET 2 +#define SPC_ADD_SENSE_LEN_OFFSET 7 #define SPC_ASC_KEY_OFFSET 12 #define SPC_ASCQ_KEY_OFFSET 13 #define TRANSPORT_IQN_LEN 224 From a42d985bd5b234da8b61347a78dc3057bf7bb94d Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Fri, 14 Oct 2011 01:30:46 +0000 Subject: [PATCH 164/820] ib_srpt: Initial SRP Target merge for v3.3-rc1 This patch adds the kernel module ib_srpt SCSI RDMA Protocol (SRP) target implementation conforming to the SRP r16a specification for the mainline drivers/target infrastructure. This driver was originally developed by Vu Pham and has been optimized by Bart Van Assche and merged into upstream LIO based on his srpt-lio-4.1 branch here: https://github.com/bvanassche/srpt-lio/commits/srpt-lio-4.1/ This updated patch also contains the following two changes from lio-core-2.6.git/master. One is to fix a bug with 1 >= task->task_sg[] chained mappings in ib_srpt, and the other to convert the configfs control plane to reference IB Port GUID and struct srpt_port directly following mainline v4.x target_core_fabric_configfs.c convertion for ib_srpt to work with rtslib/rtsadmin v2 code. These seperate patches can be found here: ib_srpt: Fix bug with chainged SGLs in srpt_map_sg_to_ib_sge http://www.risingtidesystems.com/git/?p=lio-core-2.6.git;a=commitdiff;h=ea485147563b6555a97dbf811825fbb586519252 ib_srpt: Convert se_wwn endpoint reference to struct srpt_port->port_wwn http://www.risingtidesystems.com/git/?p=lio-core-2.6.git;a=commitdiff;h=4e544a210acb227df1bb4ca5086e65bdf4e648ea This also includes the following recent v1 -> v2 review changes: ib_srpt: Fix potential out-of-bounds array access ib_srpt: Avoid failed multipart RDMA transfers ib_srpt: Fix srpt_alloc_fabric_acl failure case return value ib_srpt: Update comments to reference $driver/$port layout ib_srpt: Fix sport->port_guid formatting code ib_srpt: Remove legacy use_port_guid_in_session_name module parameter ib_srpt: Convert srp_max_rdma_size into per port configfs attribute ib_srpt: Convert srp_max_rsp_size into per port configfs attribute ib_srpt: Convert srpt_sq_size into per port configfs attribute and v2 -> v3 review changes: ib_srpt: Fix possible race with srp_sq_size in srpt_create_ch_ib ib_srpt: Fix possible race with srp_max_rsp_size in srpt_release_channel_work ib_srpt: Fix up MAX_SRPT_RDMA_SIZE define ib_srpt: Make srpt_map_sg_to_ib_sge() failure case return -EAGAIN ib_srpt: Convert port_guid to use subnet_prefix + interface_id formatting ib_srpt: Make srpt_check_stop_free return kref_put status ib_srpt: Make compilation with BUG=n proceed` ib_srpt: Use new target_core_fabric.h include ib_srpt: Check hex2bin() return code to silence build warning Cc: Bart Van Assche Cc: Roland Dreier Cc: Christoph Hellwig Cc: Vu Pham Cc: David Dillow Signed-off-by: Nicholas A. Bellinger --- drivers/infiniband/Kconfig | 1 + drivers/infiniband/Makefile | 1 + drivers/infiniband/ulp/srpt/Kconfig | 12 + drivers/infiniband/ulp/srpt/Makefile | 2 + drivers/infiniband/ulp/srpt/ib_dm_mad.h | 139 + drivers/infiniband/ulp/srpt/ib_srpt.c | 4073 +++++++++++++++++++++++ drivers/infiniband/ulp/srpt/ib_srpt.h | 444 +++ 7 files changed, 4672 insertions(+) create mode 100644 drivers/infiniband/ulp/srpt/Kconfig create mode 100644 drivers/infiniband/ulp/srpt/Makefile create mode 100644 drivers/infiniband/ulp/srpt/ib_dm_mad.h create mode 100644 drivers/infiniband/ulp/srpt/ib_srpt.c create mode 100644 drivers/infiniband/ulp/srpt/ib_srpt.h diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig index 0f9a84c1046a..eb0add311dc8 100644 --- a/drivers/infiniband/Kconfig +++ b/drivers/infiniband/Kconfig @@ -55,6 +55,7 @@ source "drivers/infiniband/hw/nes/Kconfig" source "drivers/infiniband/ulp/ipoib/Kconfig" source "drivers/infiniband/ulp/srp/Kconfig" +source "drivers/infiniband/ulp/srpt/Kconfig" source "drivers/infiniband/ulp/iser/Kconfig" diff --git a/drivers/infiniband/Makefile b/drivers/infiniband/Makefile index 9cc7a47d3e67..a3b2d8eac86e 100644 --- a/drivers/infiniband/Makefile +++ b/drivers/infiniband/Makefile @@ -10,4 +10,5 @@ obj-$(CONFIG_MLX4_INFINIBAND) += hw/mlx4/ obj-$(CONFIG_INFINIBAND_NES) += hw/nes/ obj-$(CONFIG_INFINIBAND_IPOIB) += ulp/ipoib/ obj-$(CONFIG_INFINIBAND_SRP) += ulp/srp/ +obj-$(CONFIG_INFINIBAND_SRPT) += ulp/srpt/ obj-$(CONFIG_INFINIBAND_ISER) += ulp/iser/ diff --git a/drivers/infiniband/ulp/srpt/Kconfig b/drivers/infiniband/ulp/srpt/Kconfig new file mode 100644 index 000000000000..31ee83d528d9 --- /dev/null +++ b/drivers/infiniband/ulp/srpt/Kconfig @@ -0,0 +1,12 @@ +config INFINIBAND_SRPT + tristate "InfiniBand SCSI RDMA Protocol target support" + depends on INFINIBAND && TARGET_CORE + ---help--- + + Support for the SCSI RDMA Protocol (SRP) Target driver. The + SRP protocol is a protocol that allows an initiator to access + a block storage device on another host (target) over a network + that supports the RDMA protocol. Currently the RDMA protocol is + supported by InfiniBand and by iWarp network hardware. More + information about the SRP protocol can be found on the website + of the INCITS T10 technical committee (http://www.t10.org/). diff --git a/drivers/infiniband/ulp/srpt/Makefile b/drivers/infiniband/ulp/srpt/Makefile new file mode 100644 index 000000000000..e3ee4bdfffa5 --- /dev/null +++ b/drivers/infiniband/ulp/srpt/Makefile @@ -0,0 +1,2 @@ +ccflags-y := -Idrivers/target +obj-$(CONFIG_INFINIBAND_SRPT) += ib_srpt.o diff --git a/drivers/infiniband/ulp/srpt/ib_dm_mad.h b/drivers/infiniband/ulp/srpt/ib_dm_mad.h new file mode 100644 index 000000000000..fb1de1f6f297 --- /dev/null +++ b/drivers/infiniband/ulp/srpt/ib_dm_mad.h @@ -0,0 +1,139 @@ +/* + * Copyright (c) 2006 - 2009 Mellanox Technology Inc. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef IB_DM_MAD_H +#define IB_DM_MAD_H + +#include + +#include + +enum { + /* + * See also section 13.4.7 Status Field, table 115 MAD Common Status + * Field Bit Values and also section 16.3.1.1 Status Field in the + * InfiniBand Architecture Specification. + */ + DM_MAD_STATUS_UNSUP_METHOD = 0x0008, + DM_MAD_STATUS_UNSUP_METHOD_ATTR = 0x000c, + DM_MAD_STATUS_INVALID_FIELD = 0x001c, + DM_MAD_STATUS_NO_IOC = 0x0100, + + /* + * See also the Device Management chapter, section 16.3.3 Attributes, + * table 279 Device Management Attributes in the InfiniBand + * Architecture Specification. + */ + DM_ATTR_CLASS_PORT_INFO = 0x01, + DM_ATTR_IOU_INFO = 0x10, + DM_ATTR_IOC_PROFILE = 0x11, + DM_ATTR_SVC_ENTRIES = 0x12 +}; + +struct ib_dm_hdr { + u8 reserved[28]; +}; + +/* + * Structure of management datagram sent by the SRP target implementation. + * Contains a management datagram header, reliable multi-packet transaction + * protocol (RMPP) header and ib_dm_hdr. Notes: + * - The SRP target implementation does not use RMPP or ib_dm_hdr when sending + * management datagrams. + * - The header size must be exactly 64 bytes (IB_MGMT_DEVICE_HDR), since this + * is the header size that is passed to ib_create_send_mad() in ib_srpt.c. + * - The maximum supported size for a management datagram when not using RMPP + * is 256 bytes -- 64 bytes header and 192 (IB_MGMT_DEVICE_DATA) bytes data. + */ +struct ib_dm_mad { + struct ib_mad_hdr mad_hdr; + struct ib_rmpp_hdr rmpp_hdr; + struct ib_dm_hdr dm_hdr; + u8 data[IB_MGMT_DEVICE_DATA]; +}; + +/* + * IOUnitInfo as defined in section 16.3.3.3 IOUnitInfo of the InfiniBand + * Architecture Specification. + */ +struct ib_dm_iou_info { + __be16 change_id; + u8 max_controllers; + u8 op_rom; + u8 controller_list[128]; +}; + +/* + * IOControllerprofile as defined in section 16.3.3.4 IOControllerProfile of + * the InfiniBand Architecture Specification. + */ +struct ib_dm_ioc_profile { + __be64 guid; + __be32 vendor_id; + __be32 device_id; + __be16 device_version; + __be16 reserved1; + __be32 subsys_vendor_id; + __be32 subsys_device_id; + __be16 io_class; + __be16 io_subclass; + __be16 protocol; + __be16 protocol_version; + __be16 service_conn; + __be16 initiators_supported; + __be16 send_queue_depth; + u8 reserved2; + u8 rdma_read_depth; + __be32 send_size; + __be32 rdma_size; + u8 op_cap_mask; + u8 svc_cap_mask; + u8 num_svc_entries; + u8 reserved3[9]; + u8 id_string[64]; +}; + +struct ib_dm_svc_entry { + u8 name[40]; + __be64 id; +}; + +/* + * See also section 16.3.3.5 ServiceEntries in the InfiniBand Architecture + * Specification. See also section B.7, table B.8 in the T10 SRP r16a document. + */ +struct ib_dm_svc_entries { + struct ib_dm_svc_entry service_entries[4]; +}; + +#endif diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c new file mode 100644 index 000000000000..cd5d05e22a77 --- /dev/null +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -0,0 +1,4073 @@ +/* + * Copyright (c) 2006 - 2009 Mellanox Technology Inc. All rights reserved. + * Copyright (C) 2008 - 2011 Bart Van Assche . + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "ib_srpt.h" + +/* Name of this kernel module. */ +#define DRV_NAME "ib_srpt" +#define DRV_VERSION "2.0.0" +#define DRV_RELDATE "2011-02-14" + +#define SRPT_ID_STRING "Linux SRP target" + +#undef pr_fmt +#define pr_fmt(fmt) DRV_NAME " " fmt + +MODULE_AUTHOR("Vu Pham and Bart Van Assche"); +MODULE_DESCRIPTION("InfiniBand SCSI RDMA Protocol target " + "v" DRV_VERSION " (" DRV_RELDATE ")"); +MODULE_LICENSE("Dual BSD/GPL"); + +/* + * Global Variables + */ + +static u64 srpt_service_guid; +static spinlock_t srpt_dev_lock; /* Protects srpt_dev_list. */ +static struct list_head srpt_dev_list; /* List of srpt_device structures. */ + +static unsigned srp_max_req_size = DEFAULT_MAX_REQ_SIZE; +module_param(srp_max_req_size, int, 0444); +MODULE_PARM_DESC(srp_max_req_size, + "Maximum size of SRP request messages in bytes."); + +static int srpt_srq_size = DEFAULT_SRPT_SRQ_SIZE; +module_param(srpt_srq_size, int, 0444); +MODULE_PARM_DESC(srpt_srq_size, + "Shared receive queue (SRQ) size."); + +static int srpt_get_u64_x(char *buffer, struct kernel_param *kp) +{ + return sprintf(buffer, "0x%016llx", *(u64 *)kp->arg); +} +module_param_call(srpt_service_guid, NULL, srpt_get_u64_x, &srpt_service_guid, + 0444); +MODULE_PARM_DESC(srpt_service_guid, + "Using this value for ioc_guid, id_ext, and cm_listen_id" + " instead of using the node_guid of the first HCA."); + +static struct ib_client srpt_client; +static struct target_fabric_configfs *srpt_target; +static void srpt_release_channel(struct srpt_rdma_ch *ch); +static int srpt_queue_status(struct se_cmd *cmd); + +/** + * opposite_dma_dir() - Swap DMA_TO_DEVICE and DMA_FROM_DEVICE. + */ +static inline +enum dma_data_direction opposite_dma_dir(enum dma_data_direction dir) +{ + switch (dir) { + case DMA_TO_DEVICE: return DMA_FROM_DEVICE; + case DMA_FROM_DEVICE: return DMA_TO_DEVICE; + default: return dir; + } +} + +/** + * srpt_sdev_name() - Return the name associated with the HCA. + * + * Examples are ib0, ib1, ... + */ +static inline const char *srpt_sdev_name(struct srpt_device *sdev) +{ + return sdev->device->name; +} + +static enum rdma_ch_state srpt_get_ch_state(struct srpt_rdma_ch *ch) +{ + unsigned long flags; + enum rdma_ch_state state; + + spin_lock_irqsave(&ch->spinlock, flags); + state = ch->state; + spin_unlock_irqrestore(&ch->spinlock, flags); + return state; +} + +static enum rdma_ch_state +srpt_set_ch_state(struct srpt_rdma_ch *ch, enum rdma_ch_state new_state) +{ + unsigned long flags; + enum rdma_ch_state prev; + + spin_lock_irqsave(&ch->spinlock, flags); + prev = ch->state; + ch->state = new_state; + spin_unlock_irqrestore(&ch->spinlock, flags); + return prev; +} + +/** + * srpt_test_and_set_ch_state() - Test and set the channel state. + * + * Returns true if and only if the channel state has been set to the new state. + */ +static bool +srpt_test_and_set_ch_state(struct srpt_rdma_ch *ch, enum rdma_ch_state old, + enum rdma_ch_state new) +{ + unsigned long flags; + enum rdma_ch_state prev; + + spin_lock_irqsave(&ch->spinlock, flags); + prev = ch->state; + if (prev == old) + ch->state = new; + spin_unlock_irqrestore(&ch->spinlock, flags); + return prev == old; +} + +/** + * srpt_event_handler() - Asynchronous IB event callback function. + * + * Callback function called by the InfiniBand core when an asynchronous IB + * event occurs. This callback may occur in interrupt context. See also + * section 11.5.2, Set Asynchronous Event Handler in the InfiniBand + * Architecture Specification. + */ +static void srpt_event_handler(struct ib_event_handler *handler, + struct ib_event *event) +{ + struct srpt_device *sdev; + struct srpt_port *sport; + + sdev = ib_get_client_data(event->device, &srpt_client); + if (!sdev || sdev->device != event->device) + return; + + pr_debug("ASYNC event= %d on device= %s\n", event->event, + srpt_sdev_name(sdev)); + + switch (event->event) { + case IB_EVENT_PORT_ERR: + if (event->element.port_num <= sdev->device->phys_port_cnt) { + sport = &sdev->port[event->element.port_num - 1]; + sport->lid = 0; + sport->sm_lid = 0; + } + break; + case IB_EVENT_PORT_ACTIVE: + case IB_EVENT_LID_CHANGE: + case IB_EVENT_PKEY_CHANGE: + case IB_EVENT_SM_CHANGE: + case IB_EVENT_CLIENT_REREGISTER: + /* Refresh port data asynchronously. */ + if (event->element.port_num <= sdev->device->phys_port_cnt) { + sport = &sdev->port[event->element.port_num - 1]; + if (!sport->lid && !sport->sm_lid) + schedule_work(&sport->work); + } + break; + default: + printk(KERN_ERR "received unrecognized IB event %d\n", + event->event); + break; + } +} + +/** + * srpt_srq_event() - SRQ event callback function. + */ +static void srpt_srq_event(struct ib_event *event, void *ctx) +{ + printk(KERN_INFO "SRQ event %d\n", event->event); +} + +/** + * srpt_qp_event() - QP event callback function. + */ +static void srpt_qp_event(struct ib_event *event, struct srpt_rdma_ch *ch) +{ + pr_debug("QP event %d on cm_id=%p sess_name=%s state=%d\n", + event->event, ch->cm_id, ch->sess_name, srpt_get_ch_state(ch)); + + switch (event->event) { + case IB_EVENT_COMM_EST: + ib_cm_notify(ch->cm_id, event->event); + break; + case IB_EVENT_QP_LAST_WQE_REACHED: + if (srpt_test_and_set_ch_state(ch, CH_DRAINING, + CH_RELEASING)) + srpt_release_channel(ch); + else + pr_debug("%s: state %d - ignored LAST_WQE.\n", + ch->sess_name, srpt_get_ch_state(ch)); + break; + default: + printk(KERN_ERR "received unrecognized IB QP event %d\n", + event->event); + break; + } +} + +/** + * srpt_set_ioc() - Helper function for initializing an IOUnitInfo structure. + * + * @slot: one-based slot number. + * @value: four-bit value. + * + * Copies the lowest four bits of value in element slot of the array of four + * bit elements called c_list (controller list). The index slot is one-based. + */ +static void srpt_set_ioc(u8 *c_list, u32 slot, u8 value) +{ + u16 id; + u8 tmp; + + id = (slot - 1) / 2; + if (slot & 0x1) { + tmp = c_list[id] & 0xf; + c_list[id] = (value << 4) | tmp; + } else { + tmp = c_list[id] & 0xf0; + c_list[id] = (value & 0xf) | tmp; + } +} + +/** + * srpt_get_class_port_info() - Copy ClassPortInfo to a management datagram. + * + * See also section 16.3.3.1 ClassPortInfo in the InfiniBand Architecture + * Specification. + */ +static void srpt_get_class_port_info(struct ib_dm_mad *mad) +{ + struct ib_class_port_info *cif; + + cif = (struct ib_class_port_info *)mad->data; + memset(cif, 0, sizeof *cif); + cif->base_version = 1; + cif->class_version = 1; + cif->resp_time_value = 20; + + mad->mad_hdr.status = 0; +} + +/** + * srpt_get_iou() - Write IOUnitInfo to a management datagram. + * + * See also section 16.3.3.3 IOUnitInfo in the InfiniBand Architecture + * Specification. See also section B.7, table B.6 in the SRP r16a document. + */ +static void srpt_get_iou(struct ib_dm_mad *mad) +{ + struct ib_dm_iou_info *ioui; + u8 slot; + int i; + + ioui = (struct ib_dm_iou_info *)mad->data; + ioui->change_id = __constant_cpu_to_be16(1); + ioui->max_controllers = 16; + + /* set present for slot 1 and empty for the rest */ + srpt_set_ioc(ioui->controller_list, 1, 1); + for (i = 1, slot = 2; i < 16; i++, slot++) + srpt_set_ioc(ioui->controller_list, slot, 0); + + mad->mad_hdr.status = 0; +} + +/** + * srpt_get_ioc() - Write IOControllerprofile to a management datagram. + * + * See also section 16.3.3.4 IOControllerProfile in the InfiniBand + * Architecture Specification. See also section B.7, table B.7 in the SRP + * r16a document. + */ +static void srpt_get_ioc(struct srpt_port *sport, u32 slot, + struct ib_dm_mad *mad) +{ + struct srpt_device *sdev = sport->sdev; + struct ib_dm_ioc_profile *iocp; + + iocp = (struct ib_dm_ioc_profile *)mad->data; + + if (!slot || slot > 16) { + mad->mad_hdr.status + = __constant_cpu_to_be16(DM_MAD_STATUS_INVALID_FIELD); + return; + } + + if (slot > 2) { + mad->mad_hdr.status + = __constant_cpu_to_be16(DM_MAD_STATUS_NO_IOC); + return; + } + + memset(iocp, 0, sizeof *iocp); + strcpy(iocp->id_string, SRPT_ID_STRING); + iocp->guid = cpu_to_be64(srpt_service_guid); + iocp->vendor_id = cpu_to_be32(sdev->dev_attr.vendor_id); + iocp->device_id = cpu_to_be32(sdev->dev_attr.vendor_part_id); + iocp->device_version = cpu_to_be16(sdev->dev_attr.hw_ver); + iocp->subsys_vendor_id = cpu_to_be32(sdev->dev_attr.vendor_id); + iocp->subsys_device_id = 0x0; + iocp->io_class = __constant_cpu_to_be16(SRP_REV16A_IB_IO_CLASS); + iocp->io_subclass = __constant_cpu_to_be16(SRP_IO_SUBCLASS); + iocp->protocol = __constant_cpu_to_be16(SRP_PROTOCOL); + iocp->protocol_version = __constant_cpu_to_be16(SRP_PROTOCOL_VERSION); + iocp->send_queue_depth = cpu_to_be16(sdev->srq_size); + iocp->rdma_read_depth = 4; + iocp->send_size = cpu_to_be32(srp_max_req_size); + iocp->rdma_size = cpu_to_be32(min(sport->port_attrib.srp_max_rdma_size, + 1U << 24)); + iocp->num_svc_entries = 1; + iocp->op_cap_mask = SRP_SEND_TO_IOC | SRP_SEND_FROM_IOC | + SRP_RDMA_READ_FROM_IOC | SRP_RDMA_WRITE_FROM_IOC; + + mad->mad_hdr.status = 0; +} + +/** + * srpt_get_svc_entries() - Write ServiceEntries to a management datagram. + * + * See also section 16.3.3.5 ServiceEntries in the InfiniBand Architecture + * Specification. See also section B.7, table B.8 in the SRP r16a document. + */ +static void srpt_get_svc_entries(u64 ioc_guid, + u16 slot, u8 hi, u8 lo, struct ib_dm_mad *mad) +{ + struct ib_dm_svc_entries *svc_entries; + + WARN_ON(!ioc_guid); + + if (!slot || slot > 16) { + mad->mad_hdr.status + = __constant_cpu_to_be16(DM_MAD_STATUS_INVALID_FIELD); + return; + } + + if (slot > 2 || lo > hi || hi > 1) { + mad->mad_hdr.status + = __constant_cpu_to_be16(DM_MAD_STATUS_NO_IOC); + return; + } + + svc_entries = (struct ib_dm_svc_entries *)mad->data; + memset(svc_entries, 0, sizeof *svc_entries); + svc_entries->service_entries[0].id = cpu_to_be64(ioc_guid); + snprintf(svc_entries->service_entries[0].name, + sizeof(svc_entries->service_entries[0].name), + "%s%016llx", + SRP_SERVICE_NAME_PREFIX, + ioc_guid); + + mad->mad_hdr.status = 0; +} + +/** + * srpt_mgmt_method_get() - Process a received management datagram. + * @sp: source port through which the MAD has been received. + * @rq_mad: received MAD. + * @rsp_mad: response MAD. + */ +static void srpt_mgmt_method_get(struct srpt_port *sp, struct ib_mad *rq_mad, + struct ib_dm_mad *rsp_mad) +{ + u16 attr_id; + u32 slot; + u8 hi, lo; + + attr_id = be16_to_cpu(rq_mad->mad_hdr.attr_id); + switch (attr_id) { + case DM_ATTR_CLASS_PORT_INFO: + srpt_get_class_port_info(rsp_mad); + break; + case DM_ATTR_IOU_INFO: + srpt_get_iou(rsp_mad); + break; + case DM_ATTR_IOC_PROFILE: + slot = be32_to_cpu(rq_mad->mad_hdr.attr_mod); + srpt_get_ioc(sp, slot, rsp_mad); + break; + case DM_ATTR_SVC_ENTRIES: + slot = be32_to_cpu(rq_mad->mad_hdr.attr_mod); + hi = (u8) ((slot >> 8) & 0xff); + lo = (u8) (slot & 0xff); + slot = (u16) ((slot >> 16) & 0xffff); + srpt_get_svc_entries(srpt_service_guid, + slot, hi, lo, rsp_mad); + break; + default: + rsp_mad->mad_hdr.status = + __constant_cpu_to_be16(DM_MAD_STATUS_UNSUP_METHOD_ATTR); + break; + } +} + +/** + * srpt_mad_send_handler() - Post MAD-send callback function. + */ +static void srpt_mad_send_handler(struct ib_mad_agent *mad_agent, + struct ib_mad_send_wc *mad_wc) +{ + ib_destroy_ah(mad_wc->send_buf->ah); + ib_free_send_mad(mad_wc->send_buf); +} + +/** + * srpt_mad_recv_handler() - MAD reception callback function. + */ +static void srpt_mad_recv_handler(struct ib_mad_agent *mad_agent, + struct ib_mad_recv_wc *mad_wc) +{ + struct srpt_port *sport = (struct srpt_port *)mad_agent->context; + struct ib_ah *ah; + struct ib_mad_send_buf *rsp; + struct ib_dm_mad *dm_mad; + + if (!mad_wc || !mad_wc->recv_buf.mad) + return; + + ah = ib_create_ah_from_wc(mad_agent->qp->pd, mad_wc->wc, + mad_wc->recv_buf.grh, mad_agent->port_num); + if (IS_ERR(ah)) + goto err; + + BUILD_BUG_ON(offsetof(struct ib_dm_mad, data) != IB_MGMT_DEVICE_HDR); + + rsp = ib_create_send_mad(mad_agent, mad_wc->wc->src_qp, + mad_wc->wc->pkey_index, 0, + IB_MGMT_DEVICE_HDR, IB_MGMT_DEVICE_DATA, + GFP_KERNEL); + if (IS_ERR(rsp)) + goto err_rsp; + + rsp->ah = ah; + + dm_mad = rsp->mad; + memcpy(dm_mad, mad_wc->recv_buf.mad, sizeof *dm_mad); + dm_mad->mad_hdr.method = IB_MGMT_METHOD_GET_RESP; + dm_mad->mad_hdr.status = 0; + + switch (mad_wc->recv_buf.mad->mad_hdr.method) { + case IB_MGMT_METHOD_GET: + srpt_mgmt_method_get(sport, mad_wc->recv_buf.mad, dm_mad); + break; + case IB_MGMT_METHOD_SET: + dm_mad->mad_hdr.status = + __constant_cpu_to_be16(DM_MAD_STATUS_UNSUP_METHOD_ATTR); + break; + default: + dm_mad->mad_hdr.status = + __constant_cpu_to_be16(DM_MAD_STATUS_UNSUP_METHOD); + break; + } + + if (!ib_post_send_mad(rsp, NULL)) { + ib_free_recv_mad(mad_wc); + /* will destroy_ah & free_send_mad in send completion */ + return; + } + + ib_free_send_mad(rsp); + +err_rsp: + ib_destroy_ah(ah); +err: + ib_free_recv_mad(mad_wc); +} + +/** + * srpt_refresh_port() - Configure a HCA port. + * + * Enable InfiniBand management datagram processing, update the cached sm_lid, + * lid and gid values, and register a callback function for processing MADs + * on the specified port. + * + * Note: It is safe to call this function more than once for the same port. + */ +static int srpt_refresh_port(struct srpt_port *sport) +{ + struct ib_mad_reg_req reg_req; + struct ib_port_modify port_modify; + struct ib_port_attr port_attr; + int ret; + + memset(&port_modify, 0, sizeof port_modify); + port_modify.set_port_cap_mask = IB_PORT_DEVICE_MGMT_SUP; + port_modify.clr_port_cap_mask = 0; + + ret = ib_modify_port(sport->sdev->device, sport->port, 0, &port_modify); + if (ret) + goto err_mod_port; + + ret = ib_query_port(sport->sdev->device, sport->port, &port_attr); + if (ret) + goto err_query_port; + + sport->sm_lid = port_attr.sm_lid; + sport->lid = port_attr.lid; + + ret = ib_query_gid(sport->sdev->device, sport->port, 0, &sport->gid); + if (ret) + goto err_query_port; + + if (!sport->mad_agent) { + memset(®_req, 0, sizeof reg_req); + reg_req.mgmt_class = IB_MGMT_CLASS_DEVICE_MGMT; + reg_req.mgmt_class_version = IB_MGMT_BASE_VERSION; + set_bit(IB_MGMT_METHOD_GET, reg_req.method_mask); + set_bit(IB_MGMT_METHOD_SET, reg_req.method_mask); + + sport->mad_agent = ib_register_mad_agent(sport->sdev->device, + sport->port, + IB_QPT_GSI, + ®_req, 0, + srpt_mad_send_handler, + srpt_mad_recv_handler, + sport); + if (IS_ERR(sport->mad_agent)) { + ret = PTR_ERR(sport->mad_agent); + sport->mad_agent = NULL; + goto err_query_port; + } + } + + return 0; + +err_query_port: + + port_modify.set_port_cap_mask = 0; + port_modify.clr_port_cap_mask = IB_PORT_DEVICE_MGMT_SUP; + ib_modify_port(sport->sdev->device, sport->port, 0, &port_modify); + +err_mod_port: + + return ret; +} + +/** + * srpt_unregister_mad_agent() - Unregister MAD callback functions. + * + * Note: It is safe to call this function more than once for the same device. + */ +static void srpt_unregister_mad_agent(struct srpt_device *sdev) +{ + struct ib_port_modify port_modify = { + .clr_port_cap_mask = IB_PORT_DEVICE_MGMT_SUP, + }; + struct srpt_port *sport; + int i; + + for (i = 1; i <= sdev->device->phys_port_cnt; i++) { + sport = &sdev->port[i - 1]; + WARN_ON(sport->port != i); + if (ib_modify_port(sdev->device, i, 0, &port_modify) < 0) + printk(KERN_ERR "disabling MAD processing failed.\n"); + if (sport->mad_agent) { + ib_unregister_mad_agent(sport->mad_agent); + sport->mad_agent = NULL; + } + } +} + +/** + * srpt_alloc_ioctx() - Allocate an SRPT I/O context structure. + */ +static struct srpt_ioctx *srpt_alloc_ioctx(struct srpt_device *sdev, + int ioctx_size, int dma_size, + enum dma_data_direction dir) +{ + struct srpt_ioctx *ioctx; + + ioctx = kmalloc(ioctx_size, GFP_KERNEL); + if (!ioctx) + goto err; + + ioctx->buf = kmalloc(dma_size, GFP_KERNEL); + if (!ioctx->buf) + goto err_free_ioctx; + + ioctx->dma = ib_dma_map_single(sdev->device, ioctx->buf, dma_size, dir); + if (ib_dma_mapping_error(sdev->device, ioctx->dma)) + goto err_free_buf; + + return ioctx; + +err_free_buf: + kfree(ioctx->buf); +err_free_ioctx: + kfree(ioctx); +err: + return NULL; +} + +/** + * srpt_free_ioctx() - Free an SRPT I/O context structure. + */ +static void srpt_free_ioctx(struct srpt_device *sdev, struct srpt_ioctx *ioctx, + int dma_size, enum dma_data_direction dir) +{ + if (!ioctx) + return; + + ib_dma_unmap_single(sdev->device, ioctx->dma, dma_size, dir); + kfree(ioctx->buf); + kfree(ioctx); +} + +/** + * srpt_alloc_ioctx_ring() - Allocate a ring of SRPT I/O context structures. + * @sdev: Device to allocate the I/O context ring for. + * @ring_size: Number of elements in the I/O context ring. + * @ioctx_size: I/O context size. + * @dma_size: DMA buffer size. + * @dir: DMA data direction. + */ +static struct srpt_ioctx **srpt_alloc_ioctx_ring(struct srpt_device *sdev, + int ring_size, int ioctx_size, + int dma_size, enum dma_data_direction dir) +{ + struct srpt_ioctx **ring; + int i; + + WARN_ON(ioctx_size != sizeof(struct srpt_recv_ioctx) + && ioctx_size != sizeof(struct srpt_send_ioctx)); + + ring = kmalloc(ring_size * sizeof(ring[0]), GFP_KERNEL); + if (!ring) + goto out; + for (i = 0; i < ring_size; ++i) { + ring[i] = srpt_alloc_ioctx(sdev, ioctx_size, dma_size, dir); + if (!ring[i]) + goto err; + ring[i]->index = i; + } + goto out; + +err: + while (--i >= 0) + srpt_free_ioctx(sdev, ring[i], dma_size, dir); + kfree(ring); +out: + return ring; +} + +/** + * srpt_free_ioctx_ring() - Free the ring of SRPT I/O context structures. + */ +static void srpt_free_ioctx_ring(struct srpt_ioctx **ioctx_ring, + struct srpt_device *sdev, int ring_size, + int dma_size, enum dma_data_direction dir) +{ + int i; + + for (i = 0; i < ring_size; ++i) + srpt_free_ioctx(sdev, ioctx_ring[i], dma_size, dir); + kfree(ioctx_ring); +} + +/** + * srpt_get_cmd_state() - Get the state of a SCSI command. + */ +static enum srpt_command_state srpt_get_cmd_state(struct srpt_send_ioctx *ioctx) +{ + enum srpt_command_state state; + unsigned long flags; + + BUG_ON(!ioctx); + + spin_lock_irqsave(&ioctx->spinlock, flags); + state = ioctx->state; + spin_unlock_irqrestore(&ioctx->spinlock, flags); + return state; +} + +/** + * srpt_set_cmd_state() - Set the state of a SCSI command. + * + * Does not modify the state of aborted commands. Returns the previous command + * state. + */ +static enum srpt_command_state srpt_set_cmd_state(struct srpt_send_ioctx *ioctx, + enum srpt_command_state new) +{ + enum srpt_command_state previous; + unsigned long flags; + + BUG_ON(!ioctx); + + spin_lock_irqsave(&ioctx->spinlock, flags); + previous = ioctx->state; + if (previous != SRPT_STATE_DONE) + ioctx->state = new; + spin_unlock_irqrestore(&ioctx->spinlock, flags); + + return previous; +} + +/** + * srpt_test_and_set_cmd_state() - Test and set the state of a command. + * + * Returns true if and only if the previous command state was equal to 'old'. + */ +static bool srpt_test_and_set_cmd_state(struct srpt_send_ioctx *ioctx, + enum srpt_command_state old, + enum srpt_command_state new) +{ + enum srpt_command_state previous; + unsigned long flags; + + WARN_ON(!ioctx); + WARN_ON(old == SRPT_STATE_DONE); + WARN_ON(new == SRPT_STATE_NEW); + + spin_lock_irqsave(&ioctx->spinlock, flags); + previous = ioctx->state; + if (previous == old) + ioctx->state = new; + spin_unlock_irqrestore(&ioctx->spinlock, flags); + return previous == old; +} + +/** + * srpt_post_recv() - Post an IB receive request. + */ +static int srpt_post_recv(struct srpt_device *sdev, + struct srpt_recv_ioctx *ioctx) +{ + struct ib_sge list; + struct ib_recv_wr wr, *bad_wr; + + BUG_ON(!sdev); + wr.wr_id = encode_wr_id(SRPT_RECV, ioctx->ioctx.index); + + list.addr = ioctx->ioctx.dma; + list.length = srp_max_req_size; + list.lkey = sdev->mr->lkey; + + wr.next = NULL; + wr.sg_list = &list; + wr.num_sge = 1; + + return ib_post_srq_recv(sdev->srq, &wr, &bad_wr); +} + +/** + * srpt_post_send() - Post an IB send request. + * + * Returns zero upon success and a non-zero value upon failure. + */ +static int srpt_post_send(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx, int len) +{ + struct ib_sge list; + struct ib_send_wr wr, *bad_wr; + struct srpt_device *sdev = ch->sport->sdev; + int ret; + + atomic_inc(&ch->req_lim); + + ret = -ENOMEM; + if (unlikely(atomic_dec_return(&ch->sq_wr_avail) < 0)) { + printk(KERN_WARNING "IB send queue full (needed 1)\n"); + goto out; + } + + ib_dma_sync_single_for_device(sdev->device, ioctx->ioctx.dma, len, + DMA_TO_DEVICE); + + list.addr = ioctx->ioctx.dma; + list.length = len; + list.lkey = sdev->mr->lkey; + + wr.next = NULL; + wr.wr_id = encode_wr_id(SRPT_SEND, ioctx->ioctx.index); + wr.sg_list = &list; + wr.num_sge = 1; + wr.opcode = IB_WR_SEND; + wr.send_flags = IB_SEND_SIGNALED; + + ret = ib_post_send(ch->qp, &wr, &bad_wr); + +out: + if (ret < 0) { + atomic_inc(&ch->sq_wr_avail); + atomic_dec(&ch->req_lim); + } + return ret; +} + +/** + * srpt_get_desc_tbl() - Parse the data descriptors of an SRP_CMD request. + * @ioctx: Pointer to the I/O context associated with the request. + * @srp_cmd: Pointer to the SRP_CMD request data. + * @dir: Pointer to the variable to which the transfer direction will be + * written. + * @data_len: Pointer to the variable to which the total data length of all + * descriptors in the SRP_CMD request will be written. + * + * This function initializes ioctx->nrbuf and ioctx->r_bufs. + * + * Returns -EINVAL when the SRP_CMD request contains inconsistent descriptors; + * -ENOMEM when memory allocation fails and zero upon success. + */ +static int srpt_get_desc_tbl(struct srpt_send_ioctx *ioctx, + struct srp_cmd *srp_cmd, + enum dma_data_direction *dir, u64 *data_len) +{ + struct srp_indirect_buf *idb; + struct srp_direct_buf *db; + unsigned add_cdb_offset; + int ret; + + /* + * The pointer computations below will only be compiled correctly + * if srp_cmd::add_data is declared as s8*, u8*, s8[] or u8[], so check + * whether srp_cmd::add_data has been declared as a byte pointer. + */ + BUILD_BUG_ON(!__same_type(srp_cmd->add_data[0], (s8)0) + && !__same_type(srp_cmd->add_data[0], (u8)0)); + + BUG_ON(!dir); + BUG_ON(!data_len); + + ret = 0; + *data_len = 0; + + /* + * The lower four bits of the buffer format field contain the DATA-IN + * buffer descriptor format, and the highest four bits contain the + * DATA-OUT buffer descriptor format. + */ + *dir = DMA_NONE; + if (srp_cmd->buf_fmt & 0xf) + /* DATA-IN: transfer data from target to initiator (read). */ + *dir = DMA_FROM_DEVICE; + else if (srp_cmd->buf_fmt >> 4) + /* DATA-OUT: transfer data from initiator to target (write). */ + *dir = DMA_TO_DEVICE; + + /* + * According to the SRP spec, the lower two bits of the 'ADDITIONAL + * CDB LENGTH' field are reserved and the size in bytes of this field + * is four times the value specified in bits 3..7. Hence the "& ~3". + */ + add_cdb_offset = srp_cmd->add_cdb_len & ~3; + if (((srp_cmd->buf_fmt & 0xf) == SRP_DATA_DESC_DIRECT) || + ((srp_cmd->buf_fmt >> 4) == SRP_DATA_DESC_DIRECT)) { + ioctx->n_rbuf = 1; + ioctx->rbufs = &ioctx->single_rbuf; + + db = (struct srp_direct_buf *)(srp_cmd->add_data + + add_cdb_offset); + memcpy(ioctx->rbufs, db, sizeof *db); + *data_len = be32_to_cpu(db->len); + } else if (((srp_cmd->buf_fmt & 0xf) == SRP_DATA_DESC_INDIRECT) || + ((srp_cmd->buf_fmt >> 4) == SRP_DATA_DESC_INDIRECT)) { + idb = (struct srp_indirect_buf *)(srp_cmd->add_data + + add_cdb_offset); + + ioctx->n_rbuf = be32_to_cpu(idb->table_desc.len) / sizeof *db; + + if (ioctx->n_rbuf > + (srp_cmd->data_out_desc_cnt + srp_cmd->data_in_desc_cnt)) { + printk(KERN_ERR "received unsupported SRP_CMD request" + " type (%u out + %u in != %u / %zu)\n", + srp_cmd->data_out_desc_cnt, + srp_cmd->data_in_desc_cnt, + be32_to_cpu(idb->table_desc.len), + sizeof(*db)); + ioctx->n_rbuf = 0; + ret = -EINVAL; + goto out; + } + + if (ioctx->n_rbuf == 1) + ioctx->rbufs = &ioctx->single_rbuf; + else { + ioctx->rbufs = + kmalloc(ioctx->n_rbuf * sizeof *db, GFP_ATOMIC); + if (!ioctx->rbufs) { + ioctx->n_rbuf = 0; + ret = -ENOMEM; + goto out; + } + } + + db = idb->desc_list; + memcpy(ioctx->rbufs, db, ioctx->n_rbuf * sizeof *db); + *data_len = be32_to_cpu(idb->len); + } +out: + return ret; +} + +/** + * srpt_init_ch_qp() - Initialize queue pair attributes. + * + * Initialized the attributes of queue pair 'qp' by allowing local write, + * remote read and remote write. Also transitions 'qp' to state IB_QPS_INIT. + */ +static int srpt_init_ch_qp(struct srpt_rdma_ch *ch, struct ib_qp *qp) +{ + struct ib_qp_attr *attr; + int ret; + + attr = kzalloc(sizeof *attr, GFP_KERNEL); + if (!attr) + return -ENOMEM; + + attr->qp_state = IB_QPS_INIT; + attr->qp_access_flags = IB_ACCESS_LOCAL_WRITE | IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_WRITE; + attr->port_num = ch->sport->port; + attr->pkey_index = 0; + + ret = ib_modify_qp(qp, attr, + IB_QP_STATE | IB_QP_ACCESS_FLAGS | IB_QP_PORT | + IB_QP_PKEY_INDEX); + + kfree(attr); + return ret; +} + +/** + * srpt_ch_qp_rtr() - Change the state of a channel to 'ready to receive' (RTR). + * @ch: channel of the queue pair. + * @qp: queue pair to change the state of. + * + * Returns zero upon success and a negative value upon failure. + * + * Note: currently a struct ib_qp_attr takes 136 bytes on a 64-bit system. + * If this structure ever becomes larger, it might be necessary to allocate + * it dynamically instead of on the stack. + */ +static int srpt_ch_qp_rtr(struct srpt_rdma_ch *ch, struct ib_qp *qp) +{ + struct ib_qp_attr qp_attr; + int attr_mask; + int ret; + + qp_attr.qp_state = IB_QPS_RTR; + ret = ib_cm_init_qp_attr(ch->cm_id, &qp_attr, &attr_mask); + if (ret) + goto out; + + qp_attr.max_dest_rd_atomic = 4; + + ret = ib_modify_qp(qp, &qp_attr, attr_mask); + +out: + return ret; +} + +/** + * srpt_ch_qp_rts() - Change the state of a channel to 'ready to send' (RTS). + * @ch: channel of the queue pair. + * @qp: queue pair to change the state of. + * + * Returns zero upon success and a negative value upon failure. + * + * Note: currently a struct ib_qp_attr takes 136 bytes on a 64-bit system. + * If this structure ever becomes larger, it might be necessary to allocate + * it dynamically instead of on the stack. + */ +static int srpt_ch_qp_rts(struct srpt_rdma_ch *ch, struct ib_qp *qp) +{ + struct ib_qp_attr qp_attr; + int attr_mask; + int ret; + + qp_attr.qp_state = IB_QPS_RTS; + ret = ib_cm_init_qp_attr(ch->cm_id, &qp_attr, &attr_mask); + if (ret) + goto out; + + qp_attr.max_rd_atomic = 4; + + ret = ib_modify_qp(qp, &qp_attr, attr_mask); + +out: + return ret; +} + +/** + * srpt_ch_qp_err() - Set the channel queue pair state to 'error'. + */ +static int srpt_ch_qp_err(struct srpt_rdma_ch *ch) +{ + struct ib_qp_attr qp_attr; + + qp_attr.qp_state = IB_QPS_ERR; + return ib_modify_qp(ch->qp, &qp_attr, IB_QP_STATE); +} + +/** + * srpt_unmap_sg_to_ib_sge() - Unmap an IB SGE list. + */ +static void srpt_unmap_sg_to_ib_sge(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx) +{ + struct scatterlist *sg; + enum dma_data_direction dir; + + BUG_ON(!ch); + BUG_ON(!ioctx); + BUG_ON(ioctx->n_rdma && !ioctx->rdma_ius); + + while (ioctx->n_rdma) + kfree(ioctx->rdma_ius[--ioctx->n_rdma].sge); + + kfree(ioctx->rdma_ius); + ioctx->rdma_ius = NULL; + + if (ioctx->mapped_sg_count) { + sg = ioctx->sg; + WARN_ON(!sg); + dir = ioctx->cmd.data_direction; + BUG_ON(dir == DMA_NONE); + ib_dma_unmap_sg(ch->sport->sdev->device, sg, ioctx->sg_cnt, + opposite_dma_dir(dir)); + ioctx->mapped_sg_count = 0; + } +} + +/** + * srpt_map_sg_to_ib_sge() - Map an SG list to an IB SGE list. + */ +static int srpt_map_sg_to_ib_sge(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx) +{ + struct se_cmd *cmd; + struct scatterlist *sg, *sg_orig; + int sg_cnt; + enum dma_data_direction dir; + struct rdma_iu *riu; + struct srp_direct_buf *db; + dma_addr_t dma_addr; + struct ib_sge *sge; + u64 raddr; + u32 rsize; + u32 tsize; + u32 dma_len; + int count, nrdma; + int i, j, k; + + BUG_ON(!ch); + BUG_ON(!ioctx); + cmd = &ioctx->cmd; + dir = cmd->data_direction; + BUG_ON(dir == DMA_NONE); + + transport_do_task_sg_chain(cmd); + ioctx->sg = sg = sg_orig = cmd->t_tasks_sg_chained; + ioctx->sg_cnt = sg_cnt = cmd->t_tasks_sg_chained_no; + + count = ib_dma_map_sg(ch->sport->sdev->device, sg, sg_cnt, + opposite_dma_dir(dir)); + if (unlikely(!count)) + return -EAGAIN; + + ioctx->mapped_sg_count = count; + + if (ioctx->rdma_ius && ioctx->n_rdma_ius) + nrdma = ioctx->n_rdma_ius; + else { + nrdma = (count + SRPT_DEF_SG_PER_WQE - 1) / SRPT_DEF_SG_PER_WQE + + ioctx->n_rbuf; + + ioctx->rdma_ius = kzalloc(nrdma * sizeof *riu, GFP_KERNEL); + if (!ioctx->rdma_ius) + goto free_mem; + + ioctx->n_rdma_ius = nrdma; + } + + db = ioctx->rbufs; + tsize = cmd->data_length; + dma_len = sg_dma_len(&sg[0]); + riu = ioctx->rdma_ius; + + /* + * For each remote desc - calculate the #ib_sge. + * If #ib_sge < SRPT_DEF_SG_PER_WQE per rdma operation then + * each remote desc rdma_iu is required a rdma wr; + * else + * we need to allocate extra rdma_iu to carry extra #ib_sge in + * another rdma wr + */ + for (i = 0, j = 0; + j < count && i < ioctx->n_rbuf && tsize > 0; ++i, ++riu, ++db) { + rsize = be32_to_cpu(db->len); + raddr = be64_to_cpu(db->va); + riu->raddr = raddr; + riu->rkey = be32_to_cpu(db->key); + riu->sge_cnt = 0; + + /* calculate how many sge required for this remote_buf */ + while (rsize > 0 && tsize > 0) { + + if (rsize >= dma_len) { + tsize -= dma_len; + rsize -= dma_len; + raddr += dma_len; + + if (tsize > 0) { + ++j; + if (j < count) { + sg = sg_next(sg); + dma_len = sg_dma_len(sg); + } + } + } else { + tsize -= rsize; + dma_len -= rsize; + rsize = 0; + } + + ++riu->sge_cnt; + + if (rsize > 0 && riu->sge_cnt == SRPT_DEF_SG_PER_WQE) { + ++ioctx->n_rdma; + riu->sge = + kmalloc(riu->sge_cnt * sizeof *riu->sge, + GFP_KERNEL); + if (!riu->sge) + goto free_mem; + + ++riu; + riu->sge_cnt = 0; + riu->raddr = raddr; + riu->rkey = be32_to_cpu(db->key); + } + } + + ++ioctx->n_rdma; + riu->sge = kmalloc(riu->sge_cnt * sizeof *riu->sge, + GFP_KERNEL); + if (!riu->sge) + goto free_mem; + } + + db = ioctx->rbufs; + tsize = cmd->data_length; + riu = ioctx->rdma_ius; + sg = sg_orig; + dma_len = sg_dma_len(&sg[0]); + dma_addr = sg_dma_address(&sg[0]); + + /* this second loop is really mapped sg_addres to rdma_iu->ib_sge */ + for (i = 0, j = 0; + j < count && i < ioctx->n_rbuf && tsize > 0; ++i, ++riu, ++db) { + rsize = be32_to_cpu(db->len); + sge = riu->sge; + k = 0; + + while (rsize > 0 && tsize > 0) { + sge->addr = dma_addr; + sge->lkey = ch->sport->sdev->mr->lkey; + + if (rsize >= dma_len) { + sge->length = + (tsize < dma_len) ? tsize : dma_len; + tsize -= dma_len; + rsize -= dma_len; + + if (tsize > 0) { + ++j; + if (j < count) { + sg = sg_next(sg); + dma_len = sg_dma_len(sg); + dma_addr = sg_dma_address(sg); + } + } + } else { + sge->length = (tsize < rsize) ? tsize : rsize; + tsize -= rsize; + dma_len -= rsize; + dma_addr += rsize; + rsize = 0; + } + + ++k; + if (k == riu->sge_cnt && rsize > 0 && tsize > 0) { + ++riu; + sge = riu->sge; + k = 0; + } else if (rsize > 0 && tsize > 0) + ++sge; + } + } + + return 0; + +free_mem: + srpt_unmap_sg_to_ib_sge(ch, ioctx); + + return -ENOMEM; +} + +/** + * srpt_get_send_ioctx() - Obtain an I/O context for sending to the initiator. + */ +static struct srpt_send_ioctx *srpt_get_send_ioctx(struct srpt_rdma_ch *ch) +{ + struct srpt_send_ioctx *ioctx; + unsigned long flags; + + BUG_ON(!ch); + + ioctx = NULL; + spin_lock_irqsave(&ch->spinlock, flags); + if (!list_empty(&ch->free_list)) { + ioctx = list_first_entry(&ch->free_list, + struct srpt_send_ioctx, free_list); + list_del(&ioctx->free_list); + } + spin_unlock_irqrestore(&ch->spinlock, flags); + + if (!ioctx) + return ioctx; + + BUG_ON(ioctx->ch != ch); + kref_init(&ioctx->kref); + spin_lock_init(&ioctx->spinlock); + ioctx->state = SRPT_STATE_NEW; + ioctx->n_rbuf = 0; + ioctx->rbufs = NULL; + ioctx->n_rdma = 0; + ioctx->n_rdma_ius = 0; + ioctx->rdma_ius = NULL; + ioctx->mapped_sg_count = 0; + init_completion(&ioctx->tx_done); + ioctx->queue_status_only = false; + /* + * transport_init_se_cmd() does not initialize all fields, so do it + * here. + */ + memset(&ioctx->cmd, 0, sizeof(ioctx->cmd)); + memset(&ioctx->sense_data, 0, sizeof(ioctx->sense_data)); + + return ioctx; +} + +/** + * srpt_put_send_ioctx() - Free up resources. + */ +static void srpt_put_send_ioctx(struct srpt_send_ioctx *ioctx) +{ + struct srpt_rdma_ch *ch; + unsigned long flags; + + BUG_ON(!ioctx); + ch = ioctx->ch; + BUG_ON(!ch); + + WARN_ON(srpt_get_cmd_state(ioctx) != SRPT_STATE_DONE); + + srpt_unmap_sg_to_ib_sge(ioctx->ch, ioctx); + transport_generic_free_cmd(&ioctx->cmd, 0); + + if (ioctx->n_rbuf > 1) { + kfree(ioctx->rbufs); + ioctx->rbufs = NULL; + ioctx->n_rbuf = 0; + } + + spin_lock_irqsave(&ch->spinlock, flags); + list_add(&ioctx->free_list, &ch->free_list); + spin_unlock_irqrestore(&ch->spinlock, flags); +} + +static void srpt_put_send_ioctx_kref(struct kref *kref) +{ + srpt_put_send_ioctx(container_of(kref, struct srpt_send_ioctx, kref)); +} + +/** + * srpt_abort_cmd() - Abort a SCSI command. + * @ioctx: I/O context associated with the SCSI command. + * @context: Preferred execution context. + */ +static int srpt_abort_cmd(struct srpt_send_ioctx *ioctx) +{ + enum srpt_command_state state; + unsigned long flags; + + BUG_ON(!ioctx); + + /* + * If the command is in a state where the target core is waiting for + * the ib_srpt driver, change the state to the next state. Changing + * the state of the command from SRPT_STATE_NEED_DATA to + * SRPT_STATE_DATA_IN ensures that srpt_xmit_response() will call this + * function a second time. + */ + + spin_lock_irqsave(&ioctx->spinlock, flags); + state = ioctx->state; + switch (state) { + case SRPT_STATE_NEED_DATA: + ioctx->state = SRPT_STATE_DATA_IN; + break; + case SRPT_STATE_DATA_IN: + case SRPT_STATE_CMD_RSP_SENT: + case SRPT_STATE_MGMT_RSP_SENT: + ioctx->state = SRPT_STATE_DONE; + break; + default: + break; + } + spin_unlock_irqrestore(&ioctx->spinlock, flags); + + if (state == SRPT_STATE_DONE) + goto out; + + pr_debug("Aborting cmd with state %d and tag %lld\n", state, + ioctx->tag); + + switch (state) { + case SRPT_STATE_NEW: + case SRPT_STATE_DATA_IN: + case SRPT_STATE_MGMT: + /* + * Do nothing - defer abort processing until + * srpt_queue_response() is invoked. + */ + WARN_ON(!transport_check_aborted_status(&ioctx->cmd, false)); + break; + case SRPT_STATE_NEED_DATA: + /* DMA_TO_DEVICE (write) - RDMA read error. */ + atomic_set(&ioctx->cmd.transport_lun_stop, 1); + transport_generic_handle_data(&ioctx->cmd); + break; + case SRPT_STATE_CMD_RSP_SENT: + /* + * SRP_RSP sending failed or the SRP_RSP send completion has + * not been received in time. + */ + srpt_unmap_sg_to_ib_sge(ioctx->ch, ioctx); + atomic_set(&ioctx->cmd.transport_lun_stop, 1); + kref_put(&ioctx->kref, srpt_put_send_ioctx_kref); + break; + case SRPT_STATE_MGMT_RSP_SENT: + srpt_set_cmd_state(ioctx, SRPT_STATE_DONE); + kref_put(&ioctx->kref, srpt_put_send_ioctx_kref); + break; + default: + WARN_ON("ERROR: unexpected command state"); + break; + } + +out: + return state; +} + +/** + * srpt_handle_send_err_comp() - Process an IB_WC_SEND error completion. + */ +static void srpt_handle_send_err_comp(struct srpt_rdma_ch *ch, u64 wr_id) +{ + struct srpt_send_ioctx *ioctx; + enum srpt_command_state state; + struct se_cmd *cmd; + u32 index; + + atomic_inc(&ch->sq_wr_avail); + + index = idx_from_wr_id(wr_id); + ioctx = ch->ioctx_ring[index]; + state = srpt_get_cmd_state(ioctx); + cmd = &ioctx->cmd; + + WARN_ON(state != SRPT_STATE_CMD_RSP_SENT + && state != SRPT_STATE_MGMT_RSP_SENT + && state != SRPT_STATE_NEED_DATA + && state != SRPT_STATE_DONE); + + /* If SRP_RSP sending failed, undo the ch->req_lim change. */ + if (state == SRPT_STATE_CMD_RSP_SENT + || state == SRPT_STATE_MGMT_RSP_SENT) + atomic_dec(&ch->req_lim); + + srpt_abort_cmd(ioctx); +} + +/** + * srpt_handle_send_comp() - Process an IB send completion notification. + */ +static void srpt_handle_send_comp(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx) +{ + enum srpt_command_state state; + + atomic_inc(&ch->sq_wr_avail); + + state = srpt_set_cmd_state(ioctx, SRPT_STATE_DONE); + + if (WARN_ON(state != SRPT_STATE_CMD_RSP_SENT + && state != SRPT_STATE_MGMT_RSP_SENT + && state != SRPT_STATE_DONE)) + pr_debug("state = %d\n", state); + + if (state != SRPT_STATE_DONE) + kref_put(&ioctx->kref, srpt_put_send_ioctx_kref); + else + printk(KERN_ERR "IB completion has been received too late for" + " wr_id = %u.\n", ioctx->ioctx.index); +} + +/** + * srpt_handle_rdma_comp() - Process an IB RDMA completion notification. + * + * Note: transport_generic_handle_data() is asynchronous so unmapping the + * data that has been transferred via IB RDMA must be postponed until the + * check_stop_free() callback. + */ +static void srpt_handle_rdma_comp(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx, + enum srpt_opcode opcode) +{ + WARN_ON(ioctx->n_rdma <= 0); + atomic_add(ioctx->n_rdma, &ch->sq_wr_avail); + + if (opcode == SRPT_RDMA_READ_LAST) { + if (srpt_test_and_set_cmd_state(ioctx, SRPT_STATE_NEED_DATA, + SRPT_STATE_DATA_IN)) + transport_generic_handle_data(&ioctx->cmd); + else + printk(KERN_ERR "%s[%d]: wrong state = %d\n", __func__, + __LINE__, srpt_get_cmd_state(ioctx)); + } else if (opcode == SRPT_RDMA_ABORT) { + ioctx->rdma_aborted = true; + } else { + WARN(true, "unexpected opcode %d\n", opcode); + } +} + +/** + * srpt_handle_rdma_err_comp() - Process an IB RDMA error completion. + */ +static void srpt_handle_rdma_err_comp(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx, + enum srpt_opcode opcode) +{ + struct se_cmd *cmd; + enum srpt_command_state state; + + cmd = &ioctx->cmd; + state = srpt_get_cmd_state(ioctx); + switch (opcode) { + case SRPT_RDMA_READ_LAST: + if (ioctx->n_rdma <= 0) { + printk(KERN_ERR "Received invalid RDMA read" + " error completion with idx %d\n", + ioctx->ioctx.index); + break; + } + atomic_add(ioctx->n_rdma, &ch->sq_wr_avail); + if (state == SRPT_STATE_NEED_DATA) + srpt_abort_cmd(ioctx); + else + printk(KERN_ERR "%s[%d]: wrong state = %d\n", + __func__, __LINE__, state); + break; + case SRPT_RDMA_WRITE_LAST: + atomic_set(&ioctx->cmd.transport_lun_stop, 1); + break; + default: + printk(KERN_ERR "%s[%d]: opcode = %u\n", __func__, + __LINE__, opcode); + break; + } +} + +/** + * srpt_build_cmd_rsp() - Build an SRP_RSP response. + * @ch: RDMA channel through which the request has been received. + * @ioctx: I/O context associated with the SRP_CMD request. The response will + * be built in the buffer ioctx->buf points at and hence this function will + * overwrite the request data. + * @tag: tag of the request for which this response is being generated. + * @status: value for the STATUS field of the SRP_RSP information unit. + * + * Returns the size in bytes of the SRP_RSP response. + * + * An SRP_RSP response contains a SCSI status or service response. See also + * section 6.9 in the SRP r16a document for the format of an SRP_RSP + * response. See also SPC-2 for more information about sense data. + */ +static int srpt_build_cmd_rsp(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx, u64 tag, + int status) +{ + struct srp_rsp *srp_rsp; + const u8 *sense_data; + int sense_data_len, max_sense_len; + + /* + * The lowest bit of all SAM-3 status codes is zero (see also + * paragraph 5.3 in SAM-3). + */ + WARN_ON(status & 1); + + srp_rsp = ioctx->ioctx.buf; + BUG_ON(!srp_rsp); + + sense_data = ioctx->sense_data; + sense_data_len = ioctx->cmd.scsi_sense_length; + WARN_ON(sense_data_len > sizeof(ioctx->sense_data)); + + memset(srp_rsp, 0, sizeof *srp_rsp); + srp_rsp->opcode = SRP_RSP; + srp_rsp->req_lim_delta = + __constant_cpu_to_be32(1 + atomic_xchg(&ch->req_lim_delta, 0)); + srp_rsp->tag = tag; + srp_rsp->status = status; + + if (sense_data_len) { + BUILD_BUG_ON(MIN_MAX_RSP_SIZE <= sizeof(*srp_rsp)); + max_sense_len = ch->max_ti_iu_len - sizeof(*srp_rsp); + if (sense_data_len > max_sense_len) { + printk(KERN_WARNING "truncated sense data from %d to %d" + " bytes\n", sense_data_len, max_sense_len); + sense_data_len = max_sense_len; + } + + srp_rsp->flags |= SRP_RSP_FLAG_SNSVALID; + srp_rsp->sense_data_len = cpu_to_be32(sense_data_len); + memcpy(srp_rsp + 1, sense_data, sense_data_len); + } + + return sizeof(*srp_rsp) + sense_data_len; +} + +/** + * srpt_build_tskmgmt_rsp() - Build a task management response. + * @ch: RDMA channel through which the request has been received. + * @ioctx: I/O context in which the SRP_RSP response will be built. + * @rsp_code: RSP_CODE that will be stored in the response. + * @tag: Tag of the request for which this response is being generated. + * + * Returns the size in bytes of the SRP_RSP response. + * + * An SRP_RSP response contains a SCSI status or service response. See also + * section 6.9 in the SRP r16a document for the format of an SRP_RSP + * response. + */ +static int srpt_build_tskmgmt_rsp(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx, + u8 rsp_code, u64 tag) +{ + struct srp_rsp *srp_rsp; + int resp_data_len; + int resp_len; + + resp_data_len = (rsp_code == SRP_TSK_MGMT_SUCCESS) ? 0 : 4; + resp_len = sizeof(*srp_rsp) + resp_data_len; + + srp_rsp = ioctx->ioctx.buf; + BUG_ON(!srp_rsp); + memset(srp_rsp, 0, sizeof *srp_rsp); + + srp_rsp->opcode = SRP_RSP; + srp_rsp->req_lim_delta = __constant_cpu_to_be32(1 + + atomic_xchg(&ch->req_lim_delta, 0)); + srp_rsp->tag = tag; + + if (rsp_code != SRP_TSK_MGMT_SUCCESS) { + srp_rsp->flags |= SRP_RSP_FLAG_RSPVALID; + srp_rsp->resp_data_len = cpu_to_be32(resp_data_len); + srp_rsp->data[3] = rsp_code; + } + + return resp_len; +} + +#define NO_SUCH_LUN ((uint64_t)-1LL) + +/* + * SCSI LUN addressing method. See also SAM-2 and the section about + * eight byte LUNs. + */ +enum scsi_lun_addr_method { + SCSI_LUN_ADDR_METHOD_PERIPHERAL = 0, + SCSI_LUN_ADDR_METHOD_FLAT = 1, + SCSI_LUN_ADDR_METHOD_LUN = 2, + SCSI_LUN_ADDR_METHOD_EXTENDED_LUN = 3, +}; + +/* + * srpt_unpack_lun() - Convert from network LUN to linear LUN. + * + * Convert an 2-byte, 4-byte, 6-byte or 8-byte LUN structure in network byte + * order (big endian) to a linear LUN. Supports three LUN addressing methods: + * peripheral, flat and logical unit. See also SAM-2, section 4.9.4 (page 40). + */ +static uint64_t srpt_unpack_lun(const uint8_t *lun, int len) +{ + uint64_t res = NO_SUCH_LUN; + int addressing_method; + + if (unlikely(len < 2)) { + printk(KERN_ERR "Illegal LUN length %d, expected 2 bytes or " + "more", len); + goto out; + } + + switch (len) { + case 8: + if ((*((__be64 *)lun) & + __constant_cpu_to_be64(0x0000FFFFFFFFFFFFLL)) != 0) + goto out_err; + break; + case 4: + if (*((__be16 *)&lun[2]) != 0) + goto out_err; + break; + case 6: + if (*((__be32 *)&lun[2]) != 0) + goto out_err; + break; + case 2: + break; + default: + goto out_err; + } + + addressing_method = (*lun) >> 6; /* highest two bits of byte 0 */ + switch (addressing_method) { + case SCSI_LUN_ADDR_METHOD_PERIPHERAL: + case SCSI_LUN_ADDR_METHOD_FLAT: + case SCSI_LUN_ADDR_METHOD_LUN: + res = *(lun + 1) | (((*lun) & 0x3f) << 8); + break; + + case SCSI_LUN_ADDR_METHOD_EXTENDED_LUN: + default: + printk(KERN_ERR "Unimplemented LUN addressing method %u", + addressing_method); + break; + } + +out: + return res; + +out_err: + printk(KERN_ERR "Support for multi-level LUNs has not yet been" + " implemented"); + goto out; +} + +static int srpt_check_stop_free(struct se_cmd *cmd) +{ + struct srpt_send_ioctx *ioctx; + + ioctx = container_of(cmd, struct srpt_send_ioctx, cmd); + return kref_put(&ioctx->kref, srpt_put_send_ioctx_kref); +} + +/** + * srpt_handle_cmd() - Process SRP_CMD. + */ +static int srpt_handle_cmd(struct srpt_rdma_ch *ch, + struct srpt_recv_ioctx *recv_ioctx, + struct srpt_send_ioctx *send_ioctx) +{ + struct se_cmd *cmd; + struct srp_cmd *srp_cmd; + uint64_t unpacked_lun; + u64 data_len; + enum dma_data_direction dir; + int ret; + + BUG_ON(!send_ioctx); + + srp_cmd = recv_ioctx->ioctx.buf; + kref_get(&send_ioctx->kref); + cmd = &send_ioctx->cmd; + send_ioctx->tag = srp_cmd->tag; + + switch (srp_cmd->task_attr) { + case SRP_CMD_SIMPLE_Q: + cmd->sam_task_attr = MSG_SIMPLE_TAG; + break; + case SRP_CMD_ORDERED_Q: + default: + cmd->sam_task_attr = MSG_ORDERED_TAG; + break; + case SRP_CMD_HEAD_OF_Q: + cmd->sam_task_attr = MSG_HEAD_TAG; + break; + case SRP_CMD_ACA: + cmd->sam_task_attr = MSG_ACA_TAG; + break; + } + + ret = srpt_get_desc_tbl(send_ioctx, srp_cmd, &dir, &data_len); + if (ret) { + printk(KERN_ERR "0x%llx: parsing SRP descriptor table failed.\n", + srp_cmd->tag); + cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION; + cmd->scsi_sense_reason = TCM_INVALID_CDB_FIELD; + goto send_sense; + } + + cmd->data_length = data_len; + cmd->data_direction = dir; + unpacked_lun = srpt_unpack_lun((uint8_t *)&srp_cmd->lun, + sizeof(srp_cmd->lun)); + if (transport_lookup_cmd_lun(cmd, unpacked_lun) < 0) + goto send_sense; + ret = transport_generic_allocate_tasks(cmd, srp_cmd->cdb); + if (cmd->se_cmd_flags & SCF_SCSI_RESERVATION_CONFLICT) + srpt_queue_status(cmd); + else if (cmd->se_cmd_flags & SCF_SCSI_CDB_EXCEPTION) + goto send_sense; + else + WARN_ON_ONCE(ret); + + transport_handle_cdb_direct(cmd); + return 0; + +send_sense: + transport_send_check_condition_and_sense(cmd, cmd->scsi_sense_reason, + 0); + return -1; +} + +/** + * srpt_rx_mgmt_fn_tag() - Process a task management function by tag. + * @ch: RDMA channel of the task management request. + * @fn: Task management function to perform. + * @req_tag: Tag of the SRP task management request. + * @mgmt_ioctx: I/O context of the task management request. + * + * Returns zero if the target core will process the task management + * request asynchronously. + * + * Note: It is assumed that the initiator serializes tag-based task management + * requests. + */ +static int srpt_rx_mgmt_fn_tag(struct srpt_send_ioctx *ioctx, u64 tag) +{ + struct srpt_device *sdev; + struct srpt_rdma_ch *ch; + struct srpt_send_ioctx *target; + int ret, i; + + ret = -EINVAL; + ch = ioctx->ch; + BUG_ON(!ch); + BUG_ON(!ch->sport); + sdev = ch->sport->sdev; + BUG_ON(!sdev); + spin_lock_irq(&sdev->spinlock); + for (i = 0; i < ch->rq_size; ++i) { + target = ch->ioctx_ring[i]; + if (target->cmd.se_lun == ioctx->cmd.se_lun && + target->tag == tag && + srpt_get_cmd_state(target) != SRPT_STATE_DONE) { + ret = 0; + /* now let the target core abort &target->cmd; */ + break; + } + } + spin_unlock_irq(&sdev->spinlock); + return ret; +} + +static int srp_tmr_to_tcm(int fn) +{ + switch (fn) { + case SRP_TSK_ABORT_TASK: + return TMR_ABORT_TASK; + case SRP_TSK_ABORT_TASK_SET: + return TMR_ABORT_TASK_SET; + case SRP_TSK_CLEAR_TASK_SET: + return TMR_CLEAR_TASK_SET; + case SRP_TSK_LUN_RESET: + return TMR_LUN_RESET; + case SRP_TSK_CLEAR_ACA: + return TMR_CLEAR_ACA; + default: + return -1; + } +} + +/** + * srpt_handle_tsk_mgmt() - Process an SRP_TSK_MGMT information unit. + * + * Returns 0 if and only if the request will be processed by the target core. + * + * For more information about SRP_TSK_MGMT information units, see also section + * 6.7 in the SRP r16a document. + */ +static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch, + struct srpt_recv_ioctx *recv_ioctx, + struct srpt_send_ioctx *send_ioctx) +{ + struct srp_tsk_mgmt *srp_tsk; + struct se_cmd *cmd; + uint64_t unpacked_lun; + int tcm_tmr; + int res; + + BUG_ON(!send_ioctx); + + srp_tsk = recv_ioctx->ioctx.buf; + cmd = &send_ioctx->cmd; + + pr_debug("recv tsk_mgmt fn %d for task_tag %lld and cmd tag %lld" + " cm_id %p sess %p\n", srp_tsk->tsk_mgmt_func, + srp_tsk->task_tag, srp_tsk->tag, ch->cm_id, ch->sess); + + srpt_set_cmd_state(send_ioctx, SRPT_STATE_MGMT); + send_ioctx->tag = srp_tsk->tag; + tcm_tmr = srp_tmr_to_tcm(srp_tsk->tsk_mgmt_func); + if (tcm_tmr < 0) { + send_ioctx->cmd.se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION; + send_ioctx->cmd.se_tmr_req->response = + TMR_TASK_MGMT_FUNCTION_NOT_SUPPORTED; + goto process_tmr; + } + cmd->se_tmr_req = core_tmr_alloc_req(cmd, NULL, tcm_tmr, GFP_KERNEL); + if (!cmd->se_tmr_req) { + send_ioctx->cmd.se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION; + send_ioctx->cmd.se_tmr_req->response = TMR_FUNCTION_REJECTED; + goto process_tmr; + } + + unpacked_lun = srpt_unpack_lun((uint8_t *)&srp_tsk->lun, + sizeof(srp_tsk->lun)); + res = transport_lookup_tmr_lun(&send_ioctx->cmd, unpacked_lun); + if (res) { + pr_debug("rejecting TMR for LUN %lld\n", unpacked_lun); + send_ioctx->cmd.se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION; + send_ioctx->cmd.se_tmr_req->response = TMR_LUN_DOES_NOT_EXIST; + goto process_tmr; + } + + if (srp_tsk->tsk_mgmt_func == SRP_TSK_ABORT_TASK) + srpt_rx_mgmt_fn_tag(send_ioctx, srp_tsk->task_tag); + +process_tmr: + kref_get(&send_ioctx->kref); + if (!(send_ioctx->cmd.se_cmd_flags & SCF_SCSI_CDB_EXCEPTION)) + transport_generic_handle_tmr(&send_ioctx->cmd); + else + transport_send_check_condition_and_sense(cmd, + cmd->scsi_sense_reason, 0); + +} + +/** + * srpt_handle_new_iu() - Process a newly received information unit. + * @ch: RDMA channel through which the information unit has been received. + * @ioctx: SRPT I/O context associated with the information unit. + */ +static void srpt_handle_new_iu(struct srpt_rdma_ch *ch, + struct srpt_recv_ioctx *recv_ioctx, + struct srpt_send_ioctx *send_ioctx) +{ + struct srp_cmd *srp_cmd; + enum rdma_ch_state ch_state; + + BUG_ON(!ch); + BUG_ON(!recv_ioctx); + + ib_dma_sync_single_for_cpu(ch->sport->sdev->device, + recv_ioctx->ioctx.dma, srp_max_req_size, + DMA_FROM_DEVICE); + + ch_state = srpt_get_ch_state(ch); + if (unlikely(ch_state == CH_CONNECTING)) { + list_add_tail(&recv_ioctx->wait_list, &ch->cmd_wait_list); + goto out; + } + + if (unlikely(ch_state != CH_LIVE)) + goto out; + + srp_cmd = recv_ioctx->ioctx.buf; + if (srp_cmd->opcode == SRP_CMD || srp_cmd->opcode == SRP_TSK_MGMT) { + if (!send_ioctx) + send_ioctx = srpt_get_send_ioctx(ch); + if (unlikely(!send_ioctx)) { + list_add_tail(&recv_ioctx->wait_list, + &ch->cmd_wait_list); + goto out; + } + } + + transport_init_se_cmd(&send_ioctx->cmd, &srpt_target->tf_ops, ch->sess, + 0, DMA_NONE, MSG_SIMPLE_TAG, + send_ioctx->sense_data); + + switch (srp_cmd->opcode) { + case SRP_CMD: + srpt_handle_cmd(ch, recv_ioctx, send_ioctx); + break; + case SRP_TSK_MGMT: + srpt_handle_tsk_mgmt(ch, recv_ioctx, send_ioctx); + break; + case SRP_I_LOGOUT: + printk(KERN_ERR "Not yet implemented: SRP_I_LOGOUT\n"); + break; + case SRP_CRED_RSP: + pr_debug("received SRP_CRED_RSP\n"); + break; + case SRP_AER_RSP: + pr_debug("received SRP_AER_RSP\n"); + break; + case SRP_RSP: + printk(KERN_ERR "Received SRP_RSP\n"); + break; + default: + printk(KERN_ERR "received IU with unknown opcode 0x%x\n", + srp_cmd->opcode); + break; + } + + srpt_post_recv(ch->sport->sdev, recv_ioctx); +out: + return; +} + +static void srpt_process_rcv_completion(struct ib_cq *cq, + struct srpt_rdma_ch *ch, + struct ib_wc *wc) +{ + struct srpt_device *sdev = ch->sport->sdev; + struct srpt_recv_ioctx *ioctx; + u32 index; + + index = idx_from_wr_id(wc->wr_id); + if (wc->status == IB_WC_SUCCESS) { + int req_lim; + + req_lim = atomic_dec_return(&ch->req_lim); + if (unlikely(req_lim < 0)) + printk(KERN_ERR "req_lim = %d < 0\n", req_lim); + ioctx = sdev->ioctx_ring[index]; + srpt_handle_new_iu(ch, ioctx, NULL); + } else { + printk(KERN_INFO "receiving failed for idx %u with status %d\n", + index, wc->status); + } +} + +/** + * srpt_process_send_completion() - Process an IB send completion. + * + * Note: Although this has not yet been observed during tests, at least in + * theory it is possible that the srpt_get_send_ioctx() call invoked by + * srpt_handle_new_iu() fails. This is possible because the req_lim_delta + * value in each response is set to one, and it is possible that this response + * makes the initiator send a new request before the send completion for that + * response has been processed. This could e.g. happen if the call to + * srpt_put_send_iotcx() is delayed because of a higher priority interrupt or + * if IB retransmission causes generation of the send completion to be + * delayed. Incoming information units for which srpt_get_send_ioctx() fails + * are queued on cmd_wait_list. The code below processes these delayed + * requests one at a time. + */ +static void srpt_process_send_completion(struct ib_cq *cq, + struct srpt_rdma_ch *ch, + struct ib_wc *wc) +{ + struct srpt_send_ioctx *send_ioctx; + uint32_t index; + enum srpt_opcode opcode; + + index = idx_from_wr_id(wc->wr_id); + opcode = opcode_from_wr_id(wc->wr_id); + send_ioctx = ch->ioctx_ring[index]; + if (wc->status == IB_WC_SUCCESS) { + if (opcode == SRPT_SEND) + srpt_handle_send_comp(ch, send_ioctx); + else { + WARN_ON(opcode != SRPT_RDMA_ABORT && + wc->opcode != IB_WC_RDMA_READ); + srpt_handle_rdma_comp(ch, send_ioctx, opcode); + } + } else { + if (opcode == SRPT_SEND) { + printk(KERN_INFO "sending response for idx %u failed" + " with status %d\n", index, wc->status); + srpt_handle_send_err_comp(ch, wc->wr_id); + } else if (opcode != SRPT_RDMA_MID) { + printk(KERN_INFO "RDMA t %d for idx %u failed with" + " status %d", opcode, index, wc->status); + srpt_handle_rdma_err_comp(ch, send_ioctx, opcode); + } + } + + while (unlikely(opcode == SRPT_SEND + && !list_empty(&ch->cmd_wait_list) + && srpt_get_ch_state(ch) == CH_LIVE + && (send_ioctx = srpt_get_send_ioctx(ch)) != NULL)) { + struct srpt_recv_ioctx *recv_ioctx; + + recv_ioctx = list_first_entry(&ch->cmd_wait_list, + struct srpt_recv_ioctx, + wait_list); + list_del(&recv_ioctx->wait_list); + srpt_handle_new_iu(ch, recv_ioctx, send_ioctx); + } +} + +static void srpt_process_completion(struct ib_cq *cq, struct srpt_rdma_ch *ch) +{ + struct ib_wc *const wc = ch->wc; + int i, n; + + WARN_ON(cq != ch->cq); + + ib_req_notify_cq(cq, IB_CQ_NEXT_COMP); + while ((n = ib_poll_cq(cq, ARRAY_SIZE(ch->wc), wc)) > 0) { + for (i = 0; i < n; i++) { + if (opcode_from_wr_id(wc[i].wr_id) == SRPT_RECV) + srpt_process_rcv_completion(cq, ch, &wc[i]); + else + srpt_process_send_completion(cq, ch, &wc[i]); + } + } +} + +/** + * srpt_completion() - IB completion queue callback function. + * + * Notes: + * - It is guaranteed that a completion handler will never be invoked + * concurrently on two different CPUs for the same completion queue. See also + * Documentation/infiniband/core_locking.txt and the implementation of + * handle_edge_irq() in kernel/irq/chip.c. + * - When threaded IRQs are enabled, completion handlers are invoked in thread + * context instead of interrupt context. + */ +static void srpt_completion(struct ib_cq *cq, void *ctx) +{ + struct srpt_rdma_ch *ch = ctx; + + wake_up_interruptible(&ch->wait_queue); +} + +static int srpt_compl_thread(void *arg) +{ + struct srpt_rdma_ch *ch; + + /* Hibernation / freezing of the SRPT kernel thread is not supported. */ + current->flags |= PF_NOFREEZE; + + ch = arg; + BUG_ON(!ch); + printk(KERN_INFO "Session %s: kernel thread %s (PID %d) started\n", + ch->sess_name, ch->thread->comm, current->pid); + while (!kthread_should_stop()) { + wait_event_interruptible(ch->wait_queue, + (srpt_process_completion(ch->cq, ch), + kthread_should_stop())); + } + printk(KERN_INFO "Session %s: kernel thread %s (PID %d) stopped\n", + ch->sess_name, ch->thread->comm, current->pid); + return 0; +} + +/** + * srpt_create_ch_ib() - Create receive and send completion queues. + */ +static int srpt_create_ch_ib(struct srpt_rdma_ch *ch) +{ + struct ib_qp_init_attr *qp_init; + struct srpt_port *sport = ch->sport; + struct srpt_device *sdev = sport->sdev; + u32 srp_sq_size = sport->port_attrib.srp_sq_size; + int ret; + + WARN_ON(ch->rq_size < 1); + + ret = -ENOMEM; + qp_init = kzalloc(sizeof *qp_init, GFP_KERNEL); + if (!qp_init) + goto out; + + ch->cq = ib_create_cq(sdev->device, srpt_completion, NULL, ch, + ch->rq_size + srp_sq_size, 0); + if (IS_ERR(ch->cq)) { + ret = PTR_ERR(ch->cq); + printk(KERN_ERR "failed to create CQ cqe= %d ret= %d\n", + ch->rq_size + srp_sq_size, ret); + goto out; + } + + qp_init->qp_context = (void *)ch; + qp_init->event_handler + = (void(*)(struct ib_event *, void*))srpt_qp_event; + qp_init->send_cq = ch->cq; + qp_init->recv_cq = ch->cq; + qp_init->srq = sdev->srq; + qp_init->sq_sig_type = IB_SIGNAL_REQ_WR; + qp_init->qp_type = IB_QPT_RC; + qp_init->cap.max_send_wr = srp_sq_size; + qp_init->cap.max_send_sge = SRPT_DEF_SG_PER_WQE; + + ch->qp = ib_create_qp(sdev->pd, qp_init); + if (IS_ERR(ch->qp)) { + ret = PTR_ERR(ch->qp); + printk(KERN_ERR "failed to create_qp ret= %d\n", ret); + goto err_destroy_cq; + } + + atomic_set(&ch->sq_wr_avail, qp_init->cap.max_send_wr); + + pr_debug("%s: max_cqe= %d max_sge= %d sq_size = %d cm_id= %p\n", + __func__, ch->cq->cqe, qp_init->cap.max_send_sge, + qp_init->cap.max_send_wr, ch->cm_id); + + ret = srpt_init_ch_qp(ch, ch->qp); + if (ret) + goto err_destroy_qp; + + init_waitqueue_head(&ch->wait_queue); + + pr_debug("creating thread for session %s\n", ch->sess_name); + + ch->thread = kthread_run(srpt_compl_thread, ch, "ib_srpt_compl"); + if (IS_ERR(ch->thread)) { + printk(KERN_ERR "failed to create kernel thread %ld\n", + PTR_ERR(ch->thread)); + ch->thread = NULL; + goto err_destroy_qp; + } + +out: + kfree(qp_init); + return ret; + +err_destroy_qp: + ib_destroy_qp(ch->qp); +err_destroy_cq: + ib_destroy_cq(ch->cq); + goto out; +} + +static void srpt_destroy_ch_ib(struct srpt_rdma_ch *ch) +{ + if (ch->thread) + kthread_stop(ch->thread); + + ib_destroy_qp(ch->qp); + ib_destroy_cq(ch->cq); +} + +/** + * __srpt_close_ch() - Close an RDMA channel by setting the QP error state. + * + * Reset the QP and make sure all resources associated with the channel will + * be deallocated at an appropriate time. + * + * Note: The caller must hold ch->sport->sdev->spinlock. + */ +static void __srpt_close_ch(struct srpt_rdma_ch *ch) +{ + struct srpt_device *sdev; + enum rdma_ch_state prev_state; + unsigned long flags; + + sdev = ch->sport->sdev; + + spin_lock_irqsave(&ch->spinlock, flags); + prev_state = ch->state; + switch (prev_state) { + case CH_CONNECTING: + case CH_LIVE: + ch->state = CH_DISCONNECTING; + break; + default: + break; + } + spin_unlock_irqrestore(&ch->spinlock, flags); + + switch (prev_state) { + case CH_CONNECTING: + ib_send_cm_rej(ch->cm_id, IB_CM_REJ_NO_RESOURCES, NULL, 0, + NULL, 0); + /* fall through */ + case CH_LIVE: + if (ib_send_cm_dreq(ch->cm_id, NULL, 0) < 0) + printk(KERN_ERR "sending CM DREQ failed.\n"); + break; + case CH_DISCONNECTING: + break; + case CH_DRAINING: + case CH_RELEASING: + break; + } +} + +/** + * srpt_close_ch() - Close an RDMA channel. + */ +static void srpt_close_ch(struct srpt_rdma_ch *ch) +{ + struct srpt_device *sdev; + + sdev = ch->sport->sdev; + spin_lock_irq(&sdev->spinlock); + __srpt_close_ch(ch); + spin_unlock_irq(&sdev->spinlock); +} + +/** + * srpt_drain_channel() - Drain a channel by resetting the IB queue pair. + * @cm_id: Pointer to the CM ID of the channel to be drained. + * + * Note: Must be called from inside srpt_cm_handler to avoid a race between + * accessing sdev->spinlock and the call to kfree(sdev) in srpt_remove_one() + * (the caller of srpt_cm_handler holds the cm_id spinlock; srpt_remove_one() + * waits until all target sessions for the associated IB device have been + * unregistered and target session registration involves a call to + * ib_destroy_cm_id(), which locks the cm_id spinlock and hence waits until + * this function has finished). + */ +static void srpt_drain_channel(struct ib_cm_id *cm_id) +{ + struct srpt_device *sdev; + struct srpt_rdma_ch *ch; + int ret; + bool do_reset = false; + + WARN_ON_ONCE(irqs_disabled()); + + sdev = cm_id->context; + BUG_ON(!sdev); + spin_lock_irq(&sdev->spinlock); + list_for_each_entry(ch, &sdev->rch_list, list) { + if (ch->cm_id == cm_id) { + do_reset = srpt_test_and_set_ch_state(ch, + CH_CONNECTING, CH_DRAINING) || + srpt_test_and_set_ch_state(ch, + CH_LIVE, CH_DRAINING) || + srpt_test_and_set_ch_state(ch, + CH_DISCONNECTING, CH_DRAINING); + break; + } + } + spin_unlock_irq(&sdev->spinlock); + + if (do_reset) { + ret = srpt_ch_qp_err(ch); + if (ret < 0) + printk(KERN_ERR "Setting queue pair in error state" + " failed: %d\n", ret); + } +} + +/** + * srpt_find_channel() - Look up an RDMA channel. + * @cm_id: Pointer to the CM ID of the channel to be looked up. + * + * Return NULL if no matching RDMA channel has been found. + */ +static struct srpt_rdma_ch *srpt_find_channel(struct srpt_device *sdev, + struct ib_cm_id *cm_id) +{ + struct srpt_rdma_ch *ch; + bool found; + + WARN_ON_ONCE(irqs_disabled()); + BUG_ON(!sdev); + + found = false; + spin_lock_irq(&sdev->spinlock); + list_for_each_entry(ch, &sdev->rch_list, list) { + if (ch->cm_id == cm_id) { + found = true; + break; + } + } + spin_unlock_irq(&sdev->spinlock); + + return found ? ch : NULL; +} + +/** + * srpt_release_channel() - Release channel resources. + * + * Schedules the actual release because: + * - Calling the ib_destroy_cm_id() call from inside an IB CM callback would + * trigger a deadlock. + * - It is not safe to call TCM transport_* functions from interrupt context. + */ +static void srpt_release_channel(struct srpt_rdma_ch *ch) +{ + schedule_work(&ch->release_work); +} + +static void srpt_release_channel_work(struct work_struct *w) +{ + struct srpt_rdma_ch *ch; + struct srpt_device *sdev; + + ch = container_of(w, struct srpt_rdma_ch, release_work); + pr_debug("ch = %p; ch->sess = %p; release_done = %p\n", ch, ch->sess, + ch->release_done); + + sdev = ch->sport->sdev; + BUG_ON(!sdev); + + transport_deregister_session_configfs(ch->sess); + transport_deregister_session(ch->sess); + ch->sess = NULL; + + srpt_destroy_ch_ib(ch); + + srpt_free_ioctx_ring((struct srpt_ioctx **)ch->ioctx_ring, + ch->sport->sdev, ch->rq_size, + ch->rsp_size, DMA_TO_DEVICE); + + spin_lock_irq(&sdev->spinlock); + list_del(&ch->list); + spin_unlock_irq(&sdev->spinlock); + + ib_destroy_cm_id(ch->cm_id); + + if (ch->release_done) + complete(ch->release_done); + + wake_up(&sdev->ch_releaseQ); + + kfree(ch); +} + +static struct srpt_node_acl *__srpt_lookup_acl(struct srpt_port *sport, + u8 i_port_id[16]) +{ + struct srpt_node_acl *nacl; + + list_for_each_entry(nacl, &sport->port_acl_list, list) + if (memcmp(nacl->i_port_id, i_port_id, + sizeof(nacl->i_port_id)) == 0) + return nacl; + + return NULL; +} + +static struct srpt_node_acl *srpt_lookup_acl(struct srpt_port *sport, + u8 i_port_id[16]) +{ + struct srpt_node_acl *nacl; + + spin_lock_irq(&sport->port_acl_lock); + nacl = __srpt_lookup_acl(sport, i_port_id); + spin_unlock_irq(&sport->port_acl_lock); + + return nacl; +} + +/** + * srpt_cm_req_recv() - Process the event IB_CM_REQ_RECEIVED. + * + * Ownership of the cm_id is transferred to the target session if this + * functions returns zero. Otherwise the caller remains the owner of cm_id. + */ +static int srpt_cm_req_recv(struct ib_cm_id *cm_id, + struct ib_cm_req_event_param *param, + void *private_data) +{ + struct srpt_device *sdev = cm_id->context; + struct srpt_port *sport = &sdev->port[param->port - 1]; + struct srp_login_req *req; + struct srp_login_rsp *rsp; + struct srp_login_rej *rej; + struct ib_cm_rep_param *rep_param; + struct srpt_rdma_ch *ch, *tmp_ch; + struct srpt_node_acl *nacl; + u32 it_iu_len; + int i; + int ret = 0; + + WARN_ON_ONCE(irqs_disabled()); + + if (WARN_ON(!sdev || !private_data)) + return -EINVAL; + + req = (struct srp_login_req *)private_data; + + it_iu_len = be32_to_cpu(req->req_it_iu_len); + + printk(KERN_INFO "Received SRP_LOGIN_REQ with i_port_id 0x%llx:0x%llx," + " t_port_id 0x%llx:0x%llx and it_iu_len %d on port %d" + " (guid=0x%llx:0x%llx)\n", + be64_to_cpu(*(__be64 *)&req->initiator_port_id[0]), + be64_to_cpu(*(__be64 *)&req->initiator_port_id[8]), + be64_to_cpu(*(__be64 *)&req->target_port_id[0]), + be64_to_cpu(*(__be64 *)&req->target_port_id[8]), + it_iu_len, + param->port, + be64_to_cpu(*(__be64 *)&sdev->port[param->port - 1].gid.raw[0]), + be64_to_cpu(*(__be64 *)&sdev->port[param->port - 1].gid.raw[8])); + + rsp = kzalloc(sizeof *rsp, GFP_KERNEL); + rej = kzalloc(sizeof *rej, GFP_KERNEL); + rep_param = kzalloc(sizeof *rep_param, GFP_KERNEL); + + if (!rsp || !rej || !rep_param) { + ret = -ENOMEM; + goto out; + } + + if (it_iu_len > srp_max_req_size || it_iu_len < 64) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_REQ_IT_IU_LENGTH_TOO_LARGE); + ret = -EINVAL; + printk(KERN_ERR "rejected SRP_LOGIN_REQ because its" + " length (%d bytes) is out of range (%d .. %d)\n", + it_iu_len, 64, srp_max_req_size); + goto reject; + } + + if (!sport->enabled) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_INSUFFICIENT_RESOURCES); + ret = -EINVAL; + printk(KERN_ERR "rejected SRP_LOGIN_REQ because the target port" + " has not yet been enabled\n"); + goto reject; + } + + if ((req->req_flags & SRP_MTCH_ACTION) == SRP_MULTICHAN_SINGLE) { + rsp->rsp_flags = SRP_LOGIN_RSP_MULTICHAN_NO_CHAN; + + spin_lock_irq(&sdev->spinlock); + + list_for_each_entry_safe(ch, tmp_ch, &sdev->rch_list, list) { + if (!memcmp(ch->i_port_id, req->initiator_port_id, 16) + && !memcmp(ch->t_port_id, req->target_port_id, 16) + && param->port == ch->sport->port + && param->listen_id == ch->sport->sdev->cm_id + && ch->cm_id) { + enum rdma_ch_state ch_state; + + ch_state = srpt_get_ch_state(ch); + if (ch_state != CH_CONNECTING + && ch_state != CH_LIVE) + continue; + + /* found an existing channel */ + pr_debug("Found existing channel %s" + " cm_id= %p state= %d\n", + ch->sess_name, ch->cm_id, ch_state); + + __srpt_close_ch(ch); + + rsp->rsp_flags = + SRP_LOGIN_RSP_MULTICHAN_TERMINATED; + } + } + + spin_unlock_irq(&sdev->spinlock); + + } else + rsp->rsp_flags = SRP_LOGIN_RSP_MULTICHAN_MAINTAINED; + + if (*(__be64 *)req->target_port_id != cpu_to_be64(srpt_service_guid) + || *(__be64 *)(req->target_port_id + 8) != + cpu_to_be64(srpt_service_guid)) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_UNABLE_ASSOCIATE_CHANNEL); + ret = -ENOMEM; + printk(KERN_ERR "rejected SRP_LOGIN_REQ because it" + " has an invalid target port identifier.\n"); + goto reject; + } + + ch = kzalloc(sizeof *ch, GFP_KERNEL); + if (!ch) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_INSUFFICIENT_RESOURCES); + printk(KERN_ERR "rejected SRP_LOGIN_REQ because no memory.\n"); + ret = -ENOMEM; + goto reject; + } + + INIT_WORK(&ch->release_work, srpt_release_channel_work); + memcpy(ch->i_port_id, req->initiator_port_id, 16); + memcpy(ch->t_port_id, req->target_port_id, 16); + ch->sport = &sdev->port[param->port - 1]; + ch->cm_id = cm_id; + /* + * Avoid QUEUE_FULL conditions by limiting the number of buffers used + * for the SRP protocol to the command queue size. + */ + ch->rq_size = SRPT_RQ_SIZE; + spin_lock_init(&ch->spinlock); + ch->state = CH_CONNECTING; + INIT_LIST_HEAD(&ch->cmd_wait_list); + ch->rsp_size = ch->sport->port_attrib.srp_max_rsp_size; + + ch->ioctx_ring = (struct srpt_send_ioctx **) + srpt_alloc_ioctx_ring(ch->sport->sdev, ch->rq_size, + sizeof(*ch->ioctx_ring[0]), + ch->rsp_size, DMA_TO_DEVICE); + if (!ch->ioctx_ring) + goto free_ch; + + INIT_LIST_HEAD(&ch->free_list); + for (i = 0; i < ch->rq_size; i++) { + ch->ioctx_ring[i]->ch = ch; + list_add_tail(&ch->ioctx_ring[i]->free_list, &ch->free_list); + } + + ret = srpt_create_ch_ib(ch); + if (ret) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_INSUFFICIENT_RESOURCES); + printk(KERN_ERR "rejected SRP_LOGIN_REQ because creating" + " a new RDMA channel failed.\n"); + goto free_ring; + } + + ret = srpt_ch_qp_rtr(ch, ch->qp); + if (ret) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_INSUFFICIENT_RESOURCES); + printk(KERN_ERR "rejected SRP_LOGIN_REQ because enabling" + " RTR failed (error code = %d)\n", ret); + goto destroy_ib; + } + /* + * Use the initator port identifier as the session name. + */ + snprintf(ch->sess_name, sizeof(ch->sess_name), "0x%016llx%016llx", + be64_to_cpu(*(__be64 *)ch->i_port_id), + be64_to_cpu(*(__be64 *)(ch->i_port_id + 8))); + + pr_debug("registering session %s\n", ch->sess_name); + + nacl = srpt_lookup_acl(sport, ch->i_port_id); + if (!nacl) { + printk(KERN_INFO "Rejected login because no ACL has been" + " configured yet for initiator %s.\n", ch->sess_name); + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_CHANNEL_LIMIT_REACHED); + goto destroy_ib; + } + + ch->sess = transport_init_session(); + if (!ch->sess) { + rej->reason = __constant_cpu_to_be32( + SRP_LOGIN_REJ_INSUFFICIENT_RESOURCES); + pr_debug("Failed to create session\n"); + goto deregister_session; + } + ch->sess->se_node_acl = &nacl->nacl; + transport_register_session(&sport->port_tpg_1, &nacl->nacl, ch->sess, ch); + + pr_debug("Establish connection sess=%p name=%s cm_id=%p\n", ch->sess, + ch->sess_name, ch->cm_id); + + /* create srp_login_response */ + rsp->opcode = SRP_LOGIN_RSP; + rsp->tag = req->tag; + rsp->max_it_iu_len = req->req_it_iu_len; + rsp->max_ti_iu_len = req->req_it_iu_len; + ch->max_ti_iu_len = it_iu_len; + rsp->buf_fmt = __constant_cpu_to_be16(SRP_BUF_FORMAT_DIRECT + | SRP_BUF_FORMAT_INDIRECT); + rsp->req_lim_delta = cpu_to_be32(ch->rq_size); + atomic_set(&ch->req_lim, ch->rq_size); + atomic_set(&ch->req_lim_delta, 0); + + /* create cm reply */ + rep_param->qp_num = ch->qp->qp_num; + rep_param->private_data = (void *)rsp; + rep_param->private_data_len = sizeof *rsp; + rep_param->rnr_retry_count = 7; + rep_param->flow_control = 1; + rep_param->failover_accepted = 0; + rep_param->srq = 1; + rep_param->responder_resources = 4; + rep_param->initiator_depth = 4; + + ret = ib_send_cm_rep(cm_id, rep_param); + if (ret) { + printk(KERN_ERR "sending SRP_LOGIN_REQ response failed" + " (error code = %d)\n", ret); + goto release_channel; + } + + spin_lock_irq(&sdev->spinlock); + list_add_tail(&ch->list, &sdev->rch_list); + spin_unlock_irq(&sdev->spinlock); + + goto out; + +release_channel: + srpt_set_ch_state(ch, CH_RELEASING); + transport_deregister_session_configfs(ch->sess); + +deregister_session: + transport_deregister_session(ch->sess); + ch->sess = NULL; + +destroy_ib: + srpt_destroy_ch_ib(ch); + +free_ring: + srpt_free_ioctx_ring((struct srpt_ioctx **)ch->ioctx_ring, + ch->sport->sdev, ch->rq_size, + ch->rsp_size, DMA_TO_DEVICE); +free_ch: + kfree(ch); + +reject: + rej->opcode = SRP_LOGIN_REJ; + rej->tag = req->tag; + rej->buf_fmt = __constant_cpu_to_be16(SRP_BUF_FORMAT_DIRECT + | SRP_BUF_FORMAT_INDIRECT); + + ib_send_cm_rej(cm_id, IB_CM_REJ_CONSUMER_DEFINED, NULL, 0, + (void *)rej, sizeof *rej); + +out: + kfree(rep_param); + kfree(rsp); + kfree(rej); + + return ret; +} + +static void srpt_cm_rej_recv(struct ib_cm_id *cm_id) +{ + printk(KERN_INFO "Received IB REJ for cm_id %p.\n", cm_id); + srpt_drain_channel(cm_id); +} + +/** + * srpt_cm_rtu_recv() - Process an IB_CM_RTU_RECEIVED or USER_ESTABLISHED event. + * + * An IB_CM_RTU_RECEIVED message indicates that the connection is established + * and that the recipient may begin transmitting (RTU = ready to use). + */ +static void srpt_cm_rtu_recv(struct ib_cm_id *cm_id) +{ + struct srpt_rdma_ch *ch; + int ret; + + ch = srpt_find_channel(cm_id->context, cm_id); + BUG_ON(!ch); + + if (srpt_test_and_set_ch_state(ch, CH_CONNECTING, CH_LIVE)) { + struct srpt_recv_ioctx *ioctx, *ioctx_tmp; + + ret = srpt_ch_qp_rts(ch, ch->qp); + + list_for_each_entry_safe(ioctx, ioctx_tmp, &ch->cmd_wait_list, + wait_list) { + list_del(&ioctx->wait_list); + srpt_handle_new_iu(ch, ioctx, NULL); + } + if (ret) + srpt_close_ch(ch); + } +} + +static void srpt_cm_timewait_exit(struct ib_cm_id *cm_id) +{ + printk(KERN_INFO "Received IB TimeWait exit for cm_id %p.\n", cm_id); + srpt_drain_channel(cm_id); +} + +static void srpt_cm_rep_error(struct ib_cm_id *cm_id) +{ + printk(KERN_INFO "Received IB REP error for cm_id %p.\n", cm_id); + srpt_drain_channel(cm_id); +} + +/** + * srpt_cm_dreq_recv() - Process reception of a DREQ message. + */ +static void srpt_cm_dreq_recv(struct ib_cm_id *cm_id) +{ + struct srpt_rdma_ch *ch; + unsigned long flags; + bool send_drep = false; + + ch = srpt_find_channel(cm_id->context, cm_id); + BUG_ON(!ch); + + pr_debug("cm_id= %p ch->state= %d\n", cm_id, srpt_get_ch_state(ch)); + + spin_lock_irqsave(&ch->spinlock, flags); + switch (ch->state) { + case CH_CONNECTING: + case CH_LIVE: + send_drep = true; + ch->state = CH_DISCONNECTING; + break; + case CH_DISCONNECTING: + case CH_DRAINING: + case CH_RELEASING: + WARN(true, "unexpected channel state %d\n", ch->state); + break; + } + spin_unlock_irqrestore(&ch->spinlock, flags); + + if (send_drep) { + if (ib_send_cm_drep(ch->cm_id, NULL, 0) < 0) + printk(KERN_ERR "Sending IB DREP failed.\n"); + printk(KERN_INFO "Received DREQ and sent DREP for session %s.\n", + ch->sess_name); + } +} + +/** + * srpt_cm_drep_recv() - Process reception of a DREP message. + */ +static void srpt_cm_drep_recv(struct ib_cm_id *cm_id) +{ + printk(KERN_INFO "Received InfiniBand DREP message for cm_id %p.\n", + cm_id); + srpt_drain_channel(cm_id); +} + +/** + * srpt_cm_handler() - IB connection manager callback function. + * + * A non-zero return value will cause the caller destroy the CM ID. + * + * Note: srpt_cm_handler() must only return a non-zero value when transferring + * ownership of the cm_id to a channel by srpt_cm_req_recv() failed. Returning + * a non-zero value in any other case will trigger a race with the + * ib_destroy_cm_id() call in srpt_release_channel(). + */ +static int srpt_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event) +{ + int ret; + + ret = 0; + switch (event->event) { + case IB_CM_REQ_RECEIVED: + ret = srpt_cm_req_recv(cm_id, &event->param.req_rcvd, + event->private_data); + break; + case IB_CM_REJ_RECEIVED: + srpt_cm_rej_recv(cm_id); + break; + case IB_CM_RTU_RECEIVED: + case IB_CM_USER_ESTABLISHED: + srpt_cm_rtu_recv(cm_id); + break; + case IB_CM_DREQ_RECEIVED: + srpt_cm_dreq_recv(cm_id); + break; + case IB_CM_DREP_RECEIVED: + srpt_cm_drep_recv(cm_id); + break; + case IB_CM_TIMEWAIT_EXIT: + srpt_cm_timewait_exit(cm_id); + break; + case IB_CM_REP_ERROR: + srpt_cm_rep_error(cm_id); + break; + case IB_CM_DREQ_ERROR: + printk(KERN_INFO "Received IB DREQ ERROR event.\n"); + break; + case IB_CM_MRA_RECEIVED: + printk(KERN_INFO "Received IB MRA event\n"); + break; + default: + printk(KERN_ERR "received unrecognized IB CM event %d\n", + event->event); + break; + } + + return ret; +} + +/** + * srpt_perform_rdmas() - Perform IB RDMA. + * + * Returns zero upon success or a negative number upon failure. + */ +static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx) +{ + struct ib_send_wr wr; + struct ib_send_wr *bad_wr; + struct rdma_iu *riu; + int i; + int ret; + int sq_wr_avail; + enum dma_data_direction dir; + const int n_rdma = ioctx->n_rdma; + + dir = ioctx->cmd.data_direction; + if (dir == DMA_TO_DEVICE) { + /* write */ + ret = -ENOMEM; + sq_wr_avail = atomic_sub_return(n_rdma, &ch->sq_wr_avail); + if (sq_wr_avail < 0) { + printk(KERN_WARNING "IB send queue full (needed %d)\n", + n_rdma); + goto out; + } + } + + ioctx->rdma_aborted = false; + ret = 0; + riu = ioctx->rdma_ius; + memset(&wr, 0, sizeof wr); + + for (i = 0; i < n_rdma; ++i, ++riu) { + if (dir == DMA_FROM_DEVICE) { + wr.opcode = IB_WR_RDMA_WRITE; + wr.wr_id = encode_wr_id(i == n_rdma - 1 ? + SRPT_RDMA_WRITE_LAST : + SRPT_RDMA_MID, + ioctx->ioctx.index); + } else { + wr.opcode = IB_WR_RDMA_READ; + wr.wr_id = encode_wr_id(i == n_rdma - 1 ? + SRPT_RDMA_READ_LAST : + SRPT_RDMA_MID, + ioctx->ioctx.index); + } + wr.next = NULL; + wr.wr.rdma.remote_addr = riu->raddr; + wr.wr.rdma.rkey = riu->rkey; + wr.num_sge = riu->sge_cnt; + wr.sg_list = riu->sge; + + /* only get completion event for the last rdma write */ + if (i == (n_rdma - 1) && dir == DMA_TO_DEVICE) + wr.send_flags = IB_SEND_SIGNALED; + + ret = ib_post_send(ch->qp, &wr, &bad_wr); + if (ret) + break; + } + + if (ret) + printk(KERN_ERR "%s[%d]: ib_post_send() returned %d for %d/%d", + __func__, __LINE__, ret, i, n_rdma); + if (ret && i > 0) { + wr.num_sge = 0; + wr.wr_id = encode_wr_id(SRPT_RDMA_ABORT, ioctx->ioctx.index); + wr.send_flags = IB_SEND_SIGNALED; + while (ch->state == CH_LIVE && + ib_post_send(ch->qp, &wr, &bad_wr) != 0) { + printk(KERN_INFO "Trying to abort failed RDMA transfer [%d]", + ioctx->ioctx.index); + msleep(1000); + } + while (ch->state != CH_RELEASING && !ioctx->rdma_aborted) { + printk(KERN_INFO "Waiting until RDMA abort finished [%d]", + ioctx->ioctx.index); + msleep(1000); + } + } +out: + if (unlikely(dir == DMA_TO_DEVICE && ret < 0)) + atomic_add(n_rdma, &ch->sq_wr_avail); + return ret; +} + +/** + * srpt_xfer_data() - Start data transfer from initiator to target. + */ +static int srpt_xfer_data(struct srpt_rdma_ch *ch, + struct srpt_send_ioctx *ioctx) +{ + int ret; + + ret = srpt_map_sg_to_ib_sge(ch, ioctx); + if (ret) { + printk(KERN_ERR "%s[%d] ret=%d\n", __func__, __LINE__, ret); + goto out; + } + + ret = srpt_perform_rdmas(ch, ioctx); + if (ret) { + if (ret == -EAGAIN || ret == -ENOMEM) + printk(KERN_INFO "%s[%d] queue full -- ret=%d\n", + __func__, __LINE__, ret); + else + printk(KERN_ERR "%s[%d] fatal error -- ret=%d\n", + __func__, __LINE__, ret); + goto out_unmap; + } + +out: + return ret; +out_unmap: + srpt_unmap_sg_to_ib_sge(ch, ioctx); + goto out; +} + +static int srpt_write_pending_status(struct se_cmd *se_cmd) +{ + struct srpt_send_ioctx *ioctx; + + ioctx = container_of(se_cmd, struct srpt_send_ioctx, cmd); + return srpt_get_cmd_state(ioctx) == SRPT_STATE_NEED_DATA; +} + +/* + * srpt_write_pending() - Start data transfer from initiator to target (write). + */ +static int srpt_write_pending(struct se_cmd *se_cmd) +{ + struct srpt_rdma_ch *ch; + struct srpt_send_ioctx *ioctx; + enum srpt_command_state new_state; + enum rdma_ch_state ch_state; + int ret; + + ioctx = container_of(se_cmd, struct srpt_send_ioctx, cmd); + + new_state = srpt_set_cmd_state(ioctx, SRPT_STATE_NEED_DATA); + WARN_ON(new_state == SRPT_STATE_DONE); + + ch = ioctx->ch; + BUG_ON(!ch); + + ch_state = srpt_get_ch_state(ch); + switch (ch_state) { + case CH_CONNECTING: + WARN(true, "unexpected channel state %d\n", ch_state); + ret = -EINVAL; + goto out; + case CH_LIVE: + break; + case CH_DISCONNECTING: + case CH_DRAINING: + case CH_RELEASING: + pr_debug("cmd with tag %lld: channel disconnecting\n", + ioctx->tag); + srpt_set_cmd_state(ioctx, SRPT_STATE_DATA_IN); + ret = -EINVAL; + goto out; + } + ret = srpt_xfer_data(ch, ioctx); + +out: + return ret; +} + +static u8 tcm_to_srp_tsk_mgmt_status(const int tcm_mgmt_status) +{ + switch (tcm_mgmt_status) { + case TMR_FUNCTION_COMPLETE: + return SRP_TSK_MGMT_SUCCESS; + case TMR_FUNCTION_REJECTED: + return SRP_TSK_MGMT_FUNC_NOT_SUPP; + } + return SRP_TSK_MGMT_FAILED; +} + +/** + * srpt_queue_response() - Transmits the response to a SCSI command. + * + * Callback function called by the TCM core. Must not block since it can be + * invoked on the context of the IB completion handler. + */ +static int srpt_queue_response(struct se_cmd *cmd) +{ + struct srpt_rdma_ch *ch; + struct srpt_send_ioctx *ioctx; + enum srpt_command_state state; + unsigned long flags; + int ret; + enum dma_data_direction dir; + int resp_len; + u8 srp_tm_status; + + ret = 0; + + ioctx = container_of(cmd, struct srpt_send_ioctx, cmd); + ch = ioctx->ch; + BUG_ON(!ch); + + spin_lock_irqsave(&ioctx->spinlock, flags); + state = ioctx->state; + switch (state) { + case SRPT_STATE_NEW: + case SRPT_STATE_DATA_IN: + ioctx->state = SRPT_STATE_CMD_RSP_SENT; + break; + case SRPT_STATE_MGMT: + ioctx->state = SRPT_STATE_MGMT_RSP_SENT; + break; + default: + WARN(true, "ch %p; cmd %d: unexpected command state %d\n", + ch, ioctx->ioctx.index, ioctx->state); + break; + } + spin_unlock_irqrestore(&ioctx->spinlock, flags); + + if (unlikely(transport_check_aborted_status(&ioctx->cmd, false) + || WARN_ON_ONCE(state == SRPT_STATE_CMD_RSP_SENT))) { + atomic_inc(&ch->req_lim_delta); + srpt_abort_cmd(ioctx); + goto out; + } + + dir = ioctx->cmd.data_direction; + + /* For read commands, transfer the data to the initiator. */ + if (dir == DMA_FROM_DEVICE && ioctx->cmd.data_length && + !ioctx->queue_status_only) { + ret = srpt_xfer_data(ch, ioctx); + if (ret) { + printk(KERN_ERR "xfer_data failed for tag %llu\n", + ioctx->tag); + goto out; + } + } + + if (state != SRPT_STATE_MGMT) + resp_len = srpt_build_cmd_rsp(ch, ioctx, ioctx->tag, + cmd->scsi_status); + else { + srp_tm_status + = tcm_to_srp_tsk_mgmt_status(cmd->se_tmr_req->response); + resp_len = srpt_build_tskmgmt_rsp(ch, ioctx, srp_tm_status, + ioctx->tag); + } + ret = srpt_post_send(ch, ioctx, resp_len); + if (ret) { + printk(KERN_ERR "sending cmd response failed for tag %llu\n", + ioctx->tag); + srpt_unmap_sg_to_ib_sge(ch, ioctx); + srpt_set_cmd_state(ioctx, SRPT_STATE_DONE); + kref_put(&ioctx->kref, srpt_put_send_ioctx_kref); + } + +out: + return ret; +} + +static int srpt_queue_status(struct se_cmd *cmd) +{ + struct srpt_send_ioctx *ioctx; + + ioctx = container_of(cmd, struct srpt_send_ioctx, cmd); + BUG_ON(ioctx->sense_data != cmd->sense_buffer); + if (cmd->se_cmd_flags & + (SCF_TRANSPORT_TASK_SENSE | SCF_EMULATED_TASK_SENSE)) + WARN_ON(cmd->scsi_status != SAM_STAT_CHECK_CONDITION); + ioctx->queue_status_only = true; + return srpt_queue_response(cmd); +} + +static void srpt_refresh_port_work(struct work_struct *work) +{ + struct srpt_port *sport = container_of(work, struct srpt_port, work); + + srpt_refresh_port(sport); +} + +static int srpt_ch_list_empty(struct srpt_device *sdev) +{ + int res; + + spin_lock_irq(&sdev->spinlock); + res = list_empty(&sdev->rch_list); + spin_unlock_irq(&sdev->spinlock); + + return res; +} + +/** + * srpt_release_sdev() - Free the channel resources associated with a target. + */ +static int srpt_release_sdev(struct srpt_device *sdev) +{ + struct srpt_rdma_ch *ch, *tmp_ch; + int res; + + WARN_ON_ONCE(irqs_disabled()); + + BUG_ON(!sdev); + + spin_lock_irq(&sdev->spinlock); + list_for_each_entry_safe(ch, tmp_ch, &sdev->rch_list, list) + __srpt_close_ch(ch); + spin_unlock_irq(&sdev->spinlock); + + res = wait_event_interruptible(sdev->ch_releaseQ, + srpt_ch_list_empty(sdev)); + if (res) + printk(KERN_ERR "%s: interrupted.\n", __func__); + + return 0; +} + +static struct srpt_port *__srpt_lookup_port(const char *name) +{ + struct ib_device *dev; + struct srpt_device *sdev; + struct srpt_port *sport; + int i; + + list_for_each_entry(sdev, &srpt_dev_list, list) { + dev = sdev->device; + if (!dev) + continue; + + for (i = 0; i < dev->phys_port_cnt; i++) { + sport = &sdev->port[i]; + + if (!strcmp(sport->port_guid, name)) + return sport; + } + } + + return NULL; +} + +static struct srpt_port *srpt_lookup_port(const char *name) +{ + struct srpt_port *sport; + + spin_lock(&srpt_dev_lock); + sport = __srpt_lookup_port(name); + spin_unlock(&srpt_dev_lock); + + return sport; +} + +/** + * srpt_add_one() - Infiniband device addition callback function. + */ +static void srpt_add_one(struct ib_device *device) +{ + struct srpt_device *sdev; + struct srpt_port *sport; + struct ib_srq_init_attr srq_attr; + int i; + + pr_debug("device = %p, device->dma_ops = %p\n", device, + device->dma_ops); + + sdev = kzalloc(sizeof *sdev, GFP_KERNEL); + if (!sdev) + goto err; + + sdev->device = device; + INIT_LIST_HEAD(&sdev->rch_list); + init_waitqueue_head(&sdev->ch_releaseQ); + spin_lock_init(&sdev->spinlock); + + if (ib_query_device(device, &sdev->dev_attr)) + goto free_dev; + + sdev->pd = ib_alloc_pd(device); + if (IS_ERR(sdev->pd)) + goto free_dev; + + sdev->mr = ib_get_dma_mr(sdev->pd, IB_ACCESS_LOCAL_WRITE); + if (IS_ERR(sdev->mr)) + goto err_pd; + + sdev->srq_size = min(srpt_srq_size, sdev->dev_attr.max_srq_wr); + + srq_attr.event_handler = srpt_srq_event; + srq_attr.srq_context = (void *)sdev; + srq_attr.attr.max_wr = sdev->srq_size; + srq_attr.attr.max_sge = 1; + srq_attr.attr.srq_limit = 0; + + sdev->srq = ib_create_srq(sdev->pd, &srq_attr); + if (IS_ERR(sdev->srq)) + goto err_mr; + + pr_debug("%s: create SRQ #wr= %d max_allow=%d dev= %s\n", + __func__, sdev->srq_size, sdev->dev_attr.max_srq_wr, + device->name); + + if (!srpt_service_guid) + srpt_service_guid = be64_to_cpu(device->node_guid); + + sdev->cm_id = ib_create_cm_id(device, srpt_cm_handler, sdev); + if (IS_ERR(sdev->cm_id)) + goto err_srq; + + /* print out target login information */ + pr_debug("Target login info: id_ext=%016llx,ioc_guid=%016llx," + "pkey=ffff,service_id=%016llx\n", srpt_service_guid, + srpt_service_guid, srpt_service_guid); + + /* + * We do not have a consistent service_id (ie. also id_ext of target_id) + * to identify this target. We currently use the guid of the first HCA + * in the system as service_id; therefore, the target_id will change + * if this HCA is gone bad and replaced by different HCA + */ + if (ib_cm_listen(sdev->cm_id, cpu_to_be64(srpt_service_guid), 0, NULL)) + goto err_cm; + + INIT_IB_EVENT_HANDLER(&sdev->event_handler, sdev->device, + srpt_event_handler); + if (ib_register_event_handler(&sdev->event_handler)) + goto err_cm; + + sdev->ioctx_ring = (struct srpt_recv_ioctx **) + srpt_alloc_ioctx_ring(sdev, sdev->srq_size, + sizeof(*sdev->ioctx_ring[0]), + srp_max_req_size, DMA_FROM_DEVICE); + if (!sdev->ioctx_ring) + goto err_event; + + for (i = 0; i < sdev->srq_size; ++i) + srpt_post_recv(sdev, sdev->ioctx_ring[i]); + + WARN_ON(sdev->device->phys_port_cnt + > sizeof(sdev->port)/sizeof(sdev->port[0])); + + for (i = 1; i <= sdev->device->phys_port_cnt; i++) { + sport = &sdev->port[i - 1]; + sport->sdev = sdev; + sport->port = i; + sport->port_attrib.srp_max_rdma_size = DEFAULT_MAX_RDMA_SIZE; + sport->port_attrib.srp_max_rsp_size = DEFAULT_MAX_RSP_SIZE; + sport->port_attrib.srp_sq_size = DEF_SRPT_SQ_SIZE; + INIT_WORK(&sport->work, srpt_refresh_port_work); + INIT_LIST_HEAD(&sport->port_acl_list); + spin_lock_init(&sport->port_acl_lock); + + if (srpt_refresh_port(sport)) { + printk(KERN_ERR "MAD registration failed for %s-%d.\n", + srpt_sdev_name(sdev), i); + goto err_ring; + } + snprintf(sport->port_guid, sizeof(sport->port_guid), + "0x%016llx%016llx", + be64_to_cpu(sport->gid.global.subnet_prefix), + be64_to_cpu(sport->gid.global.interface_id)); + } + + spin_lock(&srpt_dev_lock); + list_add_tail(&sdev->list, &srpt_dev_list); + spin_unlock(&srpt_dev_lock); + +out: + ib_set_client_data(device, &srpt_client, sdev); + pr_debug("added %s.\n", device->name); + return; + +err_ring: + srpt_free_ioctx_ring((struct srpt_ioctx **)sdev->ioctx_ring, sdev, + sdev->srq_size, srp_max_req_size, + DMA_FROM_DEVICE); +err_event: + ib_unregister_event_handler(&sdev->event_handler); +err_cm: + ib_destroy_cm_id(sdev->cm_id); +err_srq: + ib_destroy_srq(sdev->srq); +err_mr: + ib_dereg_mr(sdev->mr); +err_pd: + ib_dealloc_pd(sdev->pd); +free_dev: + kfree(sdev); +err: + sdev = NULL; + printk(KERN_INFO "%s(%s) failed.\n", __func__, device->name); + goto out; +} + +/** + * srpt_remove_one() - InfiniBand device removal callback function. + */ +static void srpt_remove_one(struct ib_device *device) +{ + struct srpt_device *sdev; + int i; + + sdev = ib_get_client_data(device, &srpt_client); + if (!sdev) { + printk(KERN_INFO "%s(%s): nothing to do.\n", __func__, + device->name); + return; + } + + srpt_unregister_mad_agent(sdev); + + ib_unregister_event_handler(&sdev->event_handler); + + /* Cancel any work queued by the just unregistered IB event handler. */ + for (i = 0; i < sdev->device->phys_port_cnt; i++) + cancel_work_sync(&sdev->port[i].work); + + ib_destroy_cm_id(sdev->cm_id); + + /* + * Unregistering a target must happen after destroying sdev->cm_id + * such that no new SRP_LOGIN_REQ information units can arrive while + * destroying the target. + */ + spin_lock(&srpt_dev_lock); + list_del(&sdev->list); + spin_unlock(&srpt_dev_lock); + srpt_release_sdev(sdev); + + ib_destroy_srq(sdev->srq); + ib_dereg_mr(sdev->mr); + ib_dealloc_pd(sdev->pd); + + srpt_free_ioctx_ring((struct srpt_ioctx **)sdev->ioctx_ring, sdev, + sdev->srq_size, srp_max_req_size, DMA_FROM_DEVICE); + sdev->ioctx_ring = NULL; + kfree(sdev); +} + +static struct ib_client srpt_client = { + .name = DRV_NAME, + .add = srpt_add_one, + .remove = srpt_remove_one +}; + +static int srpt_check_true(struct se_portal_group *se_tpg) +{ + return 1; +} + +static int srpt_check_false(struct se_portal_group *se_tpg) +{ + return 0; +} + +static char *srpt_get_fabric_name(void) +{ + return "srpt"; +} + +static u8 srpt_get_fabric_proto_ident(struct se_portal_group *se_tpg) +{ + return SCSI_TRANSPORTID_PROTOCOLID_SRP; +} + +static char *srpt_get_fabric_wwn(struct se_portal_group *tpg) +{ + struct srpt_port *sport = container_of(tpg, struct srpt_port, port_tpg_1); + + return sport->port_guid; +} + +static u16 srpt_get_tag(struct se_portal_group *tpg) +{ + return 1; +} + +static u32 srpt_get_default_depth(struct se_portal_group *se_tpg) +{ + return 1; +} + +static u32 srpt_get_pr_transport_id(struct se_portal_group *se_tpg, + struct se_node_acl *se_nacl, + struct t10_pr_registration *pr_reg, + int *format_code, unsigned char *buf) +{ + struct srpt_node_acl *nacl; + struct spc_rdma_transport_id *tr_id; + + nacl = container_of(se_nacl, struct srpt_node_acl, nacl); + tr_id = (void *)buf; + tr_id->protocol_identifier = SCSI_TRANSPORTID_PROTOCOLID_SRP; + memcpy(tr_id->i_port_id, nacl->i_port_id, sizeof(tr_id->i_port_id)); + return sizeof(*tr_id); +} + +static u32 srpt_get_pr_transport_id_len(struct se_portal_group *se_tpg, + struct se_node_acl *se_nacl, + struct t10_pr_registration *pr_reg, + int *format_code) +{ + *format_code = 0; + return sizeof(struct spc_rdma_transport_id); +} + +static char *srpt_parse_pr_out_transport_id(struct se_portal_group *se_tpg, + const char *buf, u32 *out_tid_len, + char **port_nexus_ptr) +{ + struct spc_rdma_transport_id *tr_id; + + *port_nexus_ptr = NULL; + *out_tid_len = sizeof(struct spc_rdma_transport_id); + tr_id = (void *)buf; + return (char *)tr_id->i_port_id; +} + +static struct se_node_acl *srpt_alloc_fabric_acl(struct se_portal_group *se_tpg) +{ + struct srpt_node_acl *nacl; + + nacl = kzalloc(sizeof(struct srpt_node_acl), GFP_KERNEL); + if (!nacl) { + printk(KERN_ERR "Unable to alocate struct srpt_node_acl\n"); + return NULL; + } + + return &nacl->nacl; +} + +static void srpt_release_fabric_acl(struct se_portal_group *se_tpg, + struct se_node_acl *se_nacl) +{ + struct srpt_node_acl *nacl; + + nacl = container_of(se_nacl, struct srpt_node_acl, nacl); + kfree(nacl); +} + +static u32 srpt_tpg_get_inst_index(struct se_portal_group *se_tpg) +{ + return 1; +} + +static void srpt_release_cmd(struct se_cmd *se_cmd) +{ +} + +/** + * srpt_shutdown_session() - Whether or not a session may be shut down. + */ +static int srpt_shutdown_session(struct se_session *se_sess) +{ + return true; +} + +/** + * srpt_close_session() - Forcibly close a session. + * + * Callback function invoked by the TCM core to clean up sessions associated + * with a node ACL when the user invokes + * rmdir /sys/kernel/config/target/$driver/$port/$tpg/acls/$i_port_id + */ +static void srpt_close_session(struct se_session *se_sess) +{ + DECLARE_COMPLETION_ONSTACK(release_done); + struct srpt_rdma_ch *ch; + struct srpt_device *sdev; + int res; + + ch = se_sess->fabric_sess_ptr; + WARN_ON(ch->sess != se_sess); + + pr_debug("ch %p state %d\n", ch, srpt_get_ch_state(ch)); + + sdev = ch->sport->sdev; + spin_lock_irq(&sdev->spinlock); + BUG_ON(ch->release_done); + ch->release_done = &release_done; + __srpt_close_ch(ch); + spin_unlock_irq(&sdev->spinlock); + + res = wait_for_completion_timeout(&release_done, 60 * HZ); + WARN_ON(res <= 0); +} + +/** + * To do: Find out whether stop_session() has a meaning for transports + * other than iSCSI. + */ +static void srpt_stop_session(struct se_session *se_sess, int sess_sleep, + int conn_sleep) +{ +} + +static void srpt_reset_nexus(struct se_session *sess) +{ + printk(KERN_ERR "This is the SRP protocol, not iSCSI\n"); +} + +static int srpt_sess_logged_in(struct se_session *se_sess) +{ + return true; +} + +/** + * srpt_sess_get_index() - Return the value of scsiAttIntrPortIndex (SCSI-MIB). + * + * A quote from RFC 4455 (SCSI-MIB) about this MIB object: + * This object represents an arbitrary integer used to uniquely identify a + * particular attached remote initiator port to a particular SCSI target port + * within a particular SCSI target device within a particular SCSI instance. + */ +static u32 srpt_sess_get_index(struct se_session *se_sess) +{ + return 0; +} + +static void srpt_set_default_node_attrs(struct se_node_acl *nacl) +{ +} + +static u32 srpt_get_task_tag(struct se_cmd *se_cmd) +{ + struct srpt_send_ioctx *ioctx; + + ioctx = container_of(se_cmd, struct srpt_send_ioctx, cmd); + return ioctx->tag; +} + +/* Note: only used from inside debug printk's by the TCM core. */ +static int srpt_get_tcm_cmd_state(struct se_cmd *se_cmd) +{ + struct srpt_send_ioctx *ioctx; + + ioctx = container_of(se_cmd, struct srpt_send_ioctx, cmd); + return srpt_get_cmd_state(ioctx); +} + +static u16 srpt_set_fabric_sense_len(struct se_cmd *cmd, u32 sense_length) +{ + return 0; +} + +static u16 srpt_get_fabric_sense_len(void) +{ + return 0; +} + +static int srpt_is_state_remove(struct se_cmd *se_cmd) +{ + return 0; +} + +/** + * srpt_parse_i_port_id() - Parse an initiator port ID. + * @name: ASCII representation of a 128-bit initiator port ID. + * @i_port_id: Binary 128-bit port ID. + */ +static int srpt_parse_i_port_id(u8 i_port_id[16], const char *name) +{ + const char *p; + unsigned len, count, leading_zero_bytes; + int ret, rc; + + p = name; + if (strnicmp(p, "0x", 2) == 0) + p += 2; + ret = -EINVAL; + len = strlen(p); + if (len % 2) + goto out; + count = min(len / 2, 16U); + leading_zero_bytes = 16 - count; + memset(i_port_id, 0, leading_zero_bytes); + rc = hex2bin(i_port_id + leading_zero_bytes, p, count); + if (rc < 0) + pr_debug("hex2bin failed for srpt_parse_i_port_id: %d\n", rc); + ret = 0; +out: + return ret; +} + +/* + * configfs callback function invoked for + * mkdir /sys/kernel/config/target/$driver/$port/$tpg/acls/$i_port_id + */ +static struct se_node_acl *srpt_make_nodeacl(struct se_portal_group *tpg, + struct config_group *group, + const char *name) +{ + struct srpt_port *sport = container_of(tpg, struct srpt_port, port_tpg_1); + struct se_node_acl *se_nacl, *se_nacl_new; + struct srpt_node_acl *nacl; + int ret = 0; + u32 nexus_depth = 1; + u8 i_port_id[16]; + + if (srpt_parse_i_port_id(i_port_id, name) < 0) { + printk(KERN_ERR "invalid initiator port ID %s\n", name); + ret = -EINVAL; + goto err; + } + + se_nacl_new = srpt_alloc_fabric_acl(tpg); + if (!se_nacl_new) { + ret = -ENOMEM; + goto err; + } + /* + * nacl_new may be released by core_tpg_add_initiator_node_acl() + * when converting a node ACL from demo mode to explict + */ + se_nacl = core_tpg_add_initiator_node_acl(tpg, se_nacl_new, name, + nexus_depth); + if (IS_ERR(se_nacl)) { + ret = PTR_ERR(se_nacl); + goto err; + } + /* Locate our struct srpt_node_acl and set sdev and i_port_id. */ + nacl = container_of(se_nacl, struct srpt_node_acl, nacl); + memcpy(&nacl->i_port_id[0], &i_port_id[0], 16); + nacl->sport = sport; + + spin_lock_irq(&sport->port_acl_lock); + list_add_tail(&nacl->list, &sport->port_acl_list); + spin_unlock_irq(&sport->port_acl_lock); + + return se_nacl; +err: + return ERR_PTR(ret); +} + +/* + * configfs callback function invoked for + * rmdir /sys/kernel/config/target/$driver/$port/$tpg/acls/$i_port_id + */ +static void srpt_drop_nodeacl(struct se_node_acl *se_nacl) +{ + struct srpt_node_acl *nacl; + struct srpt_device *sdev; + struct srpt_port *sport; + + nacl = container_of(se_nacl, struct srpt_node_acl, nacl); + sport = nacl->sport; + sdev = sport->sdev; + spin_lock_irq(&sport->port_acl_lock); + list_del(&nacl->list); + spin_unlock_irq(&sport->port_acl_lock); + core_tpg_del_initiator_node_acl(&sport->port_tpg_1, se_nacl, 1); + srpt_release_fabric_acl(NULL, se_nacl); +} + +static ssize_t srpt_tpg_attrib_show_srp_max_rdma_size( + struct se_portal_group *se_tpg, + char *page) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + + return sprintf(page, "%u\n", sport->port_attrib.srp_max_rdma_size); +} + +static ssize_t srpt_tpg_attrib_store_srp_max_rdma_size( + struct se_portal_group *se_tpg, + const char *page, + size_t count) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + unsigned long val; + int ret; + + ret = strict_strtoul(page, 0, &val); + if (ret < 0) { + pr_err("strict_strtoul() failed with ret: %d\n", ret); + return -EINVAL; + } + if (val > MAX_SRPT_RDMA_SIZE) { + pr_err("val: %lu exceeds MAX_SRPT_RDMA_SIZE: %d\n", val, + MAX_SRPT_RDMA_SIZE); + return -EINVAL; + } + if (val < DEFAULT_MAX_RDMA_SIZE) { + pr_err("val: %lu smaller than DEFAULT_MAX_RDMA_SIZE: %d\n", + val, DEFAULT_MAX_RDMA_SIZE); + return -EINVAL; + } + sport->port_attrib.srp_max_rdma_size = val; + + return count; +} + +TF_TPG_ATTRIB_ATTR(srpt, srp_max_rdma_size, S_IRUGO | S_IWUSR); + +static ssize_t srpt_tpg_attrib_show_srp_max_rsp_size( + struct se_portal_group *se_tpg, + char *page) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + + return sprintf(page, "%u\n", sport->port_attrib.srp_max_rsp_size); +} + +static ssize_t srpt_tpg_attrib_store_srp_max_rsp_size( + struct se_portal_group *se_tpg, + const char *page, + size_t count) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + unsigned long val; + int ret; + + ret = strict_strtoul(page, 0, &val); + if (ret < 0) { + pr_err("strict_strtoul() failed with ret: %d\n", ret); + return -EINVAL; + } + if (val > MAX_SRPT_RSP_SIZE) { + pr_err("val: %lu exceeds MAX_SRPT_RSP_SIZE: %d\n", val, + MAX_SRPT_RSP_SIZE); + return -EINVAL; + } + if (val < MIN_MAX_RSP_SIZE) { + pr_err("val: %lu smaller than MIN_MAX_RSP_SIZE: %d\n", val, + MIN_MAX_RSP_SIZE); + return -EINVAL; + } + sport->port_attrib.srp_max_rsp_size = val; + + return count; +} + +TF_TPG_ATTRIB_ATTR(srpt, srp_max_rsp_size, S_IRUGO | S_IWUSR); + +static ssize_t srpt_tpg_attrib_show_srp_sq_size( + struct se_portal_group *se_tpg, + char *page) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + + return sprintf(page, "%u\n", sport->port_attrib.srp_sq_size); +} + +static ssize_t srpt_tpg_attrib_store_srp_sq_size( + struct se_portal_group *se_tpg, + const char *page, + size_t count) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + unsigned long val; + int ret; + + ret = strict_strtoul(page, 0, &val); + if (ret < 0) { + pr_err("strict_strtoul() failed with ret: %d\n", ret); + return -EINVAL; + } + if (val > MAX_SRPT_SRQ_SIZE) { + pr_err("val: %lu exceeds MAX_SRPT_SRQ_SIZE: %d\n", val, + MAX_SRPT_SRQ_SIZE); + return -EINVAL; + } + if (val < MIN_SRPT_SRQ_SIZE) { + pr_err("val: %lu smaller than MIN_SRPT_SRQ_SIZE: %d\n", val, + MIN_SRPT_SRQ_SIZE); + return -EINVAL; + } + sport->port_attrib.srp_sq_size = val; + + return count; +} + +TF_TPG_ATTRIB_ATTR(srpt, srp_sq_size, S_IRUGO | S_IWUSR); + +static struct configfs_attribute *srpt_tpg_attrib_attrs[] = { + &srpt_tpg_attrib_srp_max_rdma_size.attr, + &srpt_tpg_attrib_srp_max_rsp_size.attr, + &srpt_tpg_attrib_srp_sq_size.attr, + NULL, +}; + +static ssize_t srpt_tpg_show_enable( + struct se_portal_group *se_tpg, + char *page) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + + return snprintf(page, PAGE_SIZE, "%d\n", (sport->enabled) ? 1: 0); +} + +static ssize_t srpt_tpg_store_enable( + struct se_portal_group *se_tpg, + const char *page, + size_t count) +{ + struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + unsigned long tmp; + int ret; + + ret = strict_strtoul(page, 0, &tmp); + if (ret < 0) { + printk(KERN_ERR "Unable to extract srpt_tpg_store_enable\n"); + return -EINVAL; + } + + if ((tmp != 0) && (tmp != 1)) { + printk(KERN_ERR "Illegal value for srpt_tpg_store_enable: %lu\n", tmp); + return -EINVAL; + } + if (tmp == 1) + sport->enabled = true; + else + sport->enabled = false; + + return count; +} + +TF_TPG_BASE_ATTR(srpt, enable, S_IRUGO | S_IWUSR); + +static struct configfs_attribute *srpt_tpg_attrs[] = { + &srpt_tpg_enable.attr, + NULL, +}; + +/** + * configfs callback invoked for + * mkdir /sys/kernel/config/target/$driver/$port/$tpg + */ +static struct se_portal_group *srpt_make_tpg(struct se_wwn *wwn, + struct config_group *group, + const char *name) +{ + struct srpt_port *sport = container_of(wwn, struct srpt_port, port_wwn); + int res; + + /* Initialize sport->port_wwn and sport->port_tpg_1 */ + res = core_tpg_register(&srpt_target->tf_ops, &sport->port_wwn, + &sport->port_tpg_1, sport, TRANSPORT_TPG_TYPE_NORMAL); + if (res) + return ERR_PTR(res); + + return &sport->port_tpg_1; +} + +/** + * configfs callback invoked for + * rmdir /sys/kernel/config/target/$driver/$port/$tpg + */ +static void srpt_drop_tpg(struct se_portal_group *tpg) +{ + struct srpt_port *sport = container_of(tpg, + struct srpt_port, port_tpg_1); + + sport->enabled = false; + core_tpg_deregister(&sport->port_tpg_1); +} + +/** + * configfs callback invoked for + * mkdir /sys/kernel/config/target/$driver/$port + */ +static struct se_wwn *srpt_make_tport(struct target_fabric_configfs *tf, + struct config_group *group, + const char *name) +{ + struct srpt_port *sport; + int ret; + + sport = srpt_lookup_port(name); + pr_debug("make_tport(%s)\n", name); + ret = -EINVAL; + if (!sport) + goto err; + + return &sport->port_wwn; + +err: + return ERR_PTR(ret); +} + +/** + * configfs callback invoked for + * rmdir /sys/kernel/config/target/$driver/$port + */ +static void srpt_drop_tport(struct se_wwn *wwn) +{ + struct srpt_port *sport = container_of(wwn, struct srpt_port, port_wwn); + + pr_debug("drop_tport(%s\n", config_item_name(&sport->port_wwn.wwn_group.cg_item)); +} + +static ssize_t srpt_wwn_show_attr_version(struct target_fabric_configfs *tf, + char *buf) +{ + return scnprintf(buf, PAGE_SIZE, "%s\n", DRV_VERSION); +} + +TF_WWN_ATTR_RO(srpt, version); + +static struct configfs_attribute *srpt_wwn_attrs[] = { + &srpt_wwn_version.attr, + NULL, +}; + +static struct target_core_fabric_ops srpt_template = { + .get_fabric_name = srpt_get_fabric_name, + .get_fabric_proto_ident = srpt_get_fabric_proto_ident, + .tpg_get_wwn = srpt_get_fabric_wwn, + .tpg_get_tag = srpt_get_tag, + .tpg_get_default_depth = srpt_get_default_depth, + .tpg_get_pr_transport_id = srpt_get_pr_transport_id, + .tpg_get_pr_transport_id_len = srpt_get_pr_transport_id_len, + .tpg_parse_pr_out_transport_id = srpt_parse_pr_out_transport_id, + .tpg_check_demo_mode = srpt_check_false, + .tpg_check_demo_mode_cache = srpt_check_true, + .tpg_check_demo_mode_write_protect = srpt_check_true, + .tpg_check_prod_mode_write_protect = srpt_check_false, + .tpg_alloc_fabric_acl = srpt_alloc_fabric_acl, + .tpg_release_fabric_acl = srpt_release_fabric_acl, + .tpg_get_inst_index = srpt_tpg_get_inst_index, + .release_cmd = srpt_release_cmd, + .check_stop_free = srpt_check_stop_free, + .shutdown_session = srpt_shutdown_session, + .close_session = srpt_close_session, + .stop_session = srpt_stop_session, + .fall_back_to_erl0 = srpt_reset_nexus, + .sess_logged_in = srpt_sess_logged_in, + .sess_get_index = srpt_sess_get_index, + .sess_get_initiator_sid = NULL, + .write_pending = srpt_write_pending, + .write_pending_status = srpt_write_pending_status, + .set_default_node_attributes = srpt_set_default_node_attrs, + .get_task_tag = srpt_get_task_tag, + .get_cmd_state = srpt_get_tcm_cmd_state, + .queue_data_in = srpt_queue_response, + .queue_status = srpt_queue_status, + .queue_tm_rsp = srpt_queue_response, + .get_fabric_sense_len = srpt_get_fabric_sense_len, + .set_fabric_sense_len = srpt_set_fabric_sense_len, + .is_state_remove = srpt_is_state_remove, + /* + * Setup function pointers for generic logic in + * target_core_fabric_configfs.c + */ + .fabric_make_wwn = srpt_make_tport, + .fabric_drop_wwn = srpt_drop_tport, + .fabric_make_tpg = srpt_make_tpg, + .fabric_drop_tpg = srpt_drop_tpg, + .fabric_post_link = NULL, + .fabric_pre_unlink = NULL, + .fabric_make_np = NULL, + .fabric_drop_np = NULL, + .fabric_make_nodeacl = srpt_make_nodeacl, + .fabric_drop_nodeacl = srpt_drop_nodeacl, +}; + +/** + * srpt_init_module() - Kernel module initialization. + * + * Note: Since ib_register_client() registers callback functions, and since at + * least one of these callback functions (srpt_add_one()) calls target core + * functions, this driver must be registered with the target core before + * ib_register_client() is called. + */ +static int __init srpt_init_module(void) +{ + int ret; + + ret = -EINVAL; + if (srp_max_req_size < MIN_MAX_REQ_SIZE) { + printk(KERN_ERR "invalid value %d for kernel module parameter" + " srp_max_req_size -- must be at least %d.\n", + srp_max_req_size, MIN_MAX_REQ_SIZE); + goto out; + } + + if (srpt_srq_size < MIN_SRPT_SRQ_SIZE + || srpt_srq_size > MAX_SRPT_SRQ_SIZE) { + printk(KERN_ERR "invalid value %d for kernel module parameter" + " srpt_srq_size -- must be in the range [%d..%d].\n", + srpt_srq_size, MIN_SRPT_SRQ_SIZE, MAX_SRPT_SRQ_SIZE); + goto out; + } + + spin_lock_init(&srpt_dev_lock); + INIT_LIST_HEAD(&srpt_dev_list); + + ret = -ENODEV; + srpt_target = target_fabric_configfs_init(THIS_MODULE, "srpt"); + if (!srpt_target) { + printk(KERN_ERR "couldn't register\n"); + goto out; + } + + srpt_target->tf_ops = srpt_template; + + /* Enable SG chaining */ + srpt_target->tf_ops.task_sg_chaining = true; + + /* + * Set up default attribute lists. + */ + srpt_target->tf_cit_tmpl.tfc_wwn_cit.ct_attrs = srpt_wwn_attrs; + srpt_target->tf_cit_tmpl.tfc_tpg_base_cit.ct_attrs = srpt_tpg_attrs; + srpt_target->tf_cit_tmpl.tfc_tpg_attrib_cit.ct_attrs = srpt_tpg_attrib_attrs; + srpt_target->tf_cit_tmpl.tfc_tpg_param_cit.ct_attrs = NULL; + srpt_target->tf_cit_tmpl.tfc_tpg_np_base_cit.ct_attrs = NULL; + srpt_target->tf_cit_tmpl.tfc_tpg_nacl_base_cit.ct_attrs = NULL; + srpt_target->tf_cit_tmpl.tfc_tpg_nacl_attrib_cit.ct_attrs = NULL; + srpt_target->tf_cit_tmpl.tfc_tpg_nacl_auth_cit.ct_attrs = NULL; + srpt_target->tf_cit_tmpl.tfc_tpg_nacl_param_cit.ct_attrs = NULL; + + ret = target_fabric_configfs_register(srpt_target); + if (ret < 0) { + printk(KERN_ERR "couldn't register\n"); + goto out_free_target; + } + + ret = ib_register_client(&srpt_client); + if (ret) { + printk(KERN_ERR "couldn't register IB client\n"); + goto out_unregister_target; + } + + return 0; + +out_unregister_target: + target_fabric_configfs_deregister(srpt_target); + srpt_target = NULL; +out_free_target: + if (srpt_target) + target_fabric_configfs_free(srpt_target); +out: + return ret; +} + +static void __exit srpt_cleanup_module(void) +{ + ib_unregister_client(&srpt_client); + target_fabric_configfs_deregister(srpt_target); + srpt_target = NULL; +} + +module_init(srpt_init_module); +module_exit(srpt_cleanup_module); diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.h b/drivers/infiniband/ulp/srpt/ib_srpt.h new file mode 100644 index 000000000000..b4b4bbcd7f16 --- /dev/null +++ b/drivers/infiniband/ulp/srpt/ib_srpt.h @@ -0,0 +1,444 @@ +/* + * Copyright (c) 2006 - 2009 Mellanox Technology Inc. All rights reserved. + * Copyright (C) 2009 - 2010 Bart Van Assche . + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + */ + +#ifndef IB_SRPT_H +#define IB_SRPT_H + +#include +#include +#include +#include + +#include +#include +#include + +#include + +#include "ib_dm_mad.h" + +/* + * The prefix the ServiceName field must start with in the device management + * ServiceEntries attribute pair. See also the SRP specification. + */ +#define SRP_SERVICE_NAME_PREFIX "SRP.T10:" + +enum { + /* + * SRP IOControllerProfile attributes for SRP target ports that have + * not been defined in . Source: section B.7, table B.7 + * in the SRP specification. + */ + SRP_PROTOCOL = 0x0108, + SRP_PROTOCOL_VERSION = 0x0001, + SRP_IO_SUBCLASS = 0x609e, + SRP_SEND_TO_IOC = 0x01, + SRP_SEND_FROM_IOC = 0x02, + SRP_RDMA_READ_FROM_IOC = 0x08, + SRP_RDMA_WRITE_FROM_IOC = 0x20, + + /* + * srp_login_cmd.req_flags bitmasks. See also table 9 in the SRP + * specification. + */ + SRP_MTCH_ACTION = 0x03, /* MULTI-CHANNEL ACTION */ + SRP_LOSOLNT = 0x10, /* logout solicited notification */ + SRP_CRSOLNT = 0x20, /* credit request solicited notification */ + SRP_AESOLNT = 0x40, /* asynchronous event solicited notification */ + + /* + * srp_cmd.sol_nt / srp_tsk_mgmt.sol_not bitmasks. See also tables + * 18 and 20 in the SRP specification. + */ + SRP_SCSOLNT = 0x02, /* SCSOLNT = successful solicited notification */ + SRP_UCSOLNT = 0x04, /* UCSOLNT = unsuccessful solicited notification */ + + /* + * srp_rsp.sol_not / srp_t_logout.sol_not bitmasks. See also tables + * 16 and 22 in the SRP specification. + */ + SRP_SOLNT = 0x01, /* SOLNT = solicited notification */ + + /* See also table 24 in the SRP specification. */ + SRP_TSK_MGMT_SUCCESS = 0x00, + SRP_TSK_MGMT_FUNC_NOT_SUPP = 0x04, + SRP_TSK_MGMT_FAILED = 0x05, + + /* See also table 21 in the SRP specification. */ + SRP_CMD_SIMPLE_Q = 0x0, + SRP_CMD_HEAD_OF_Q = 0x1, + SRP_CMD_ORDERED_Q = 0x2, + SRP_CMD_ACA = 0x4, + + SRP_LOGIN_RSP_MULTICHAN_NO_CHAN = 0x0, + SRP_LOGIN_RSP_MULTICHAN_TERMINATED = 0x1, + SRP_LOGIN_RSP_MULTICHAN_MAINTAINED = 0x2, + + SRPT_DEF_SG_TABLESIZE = 128, + SRPT_DEF_SG_PER_WQE = 16, + + MIN_SRPT_SQ_SIZE = 16, + DEF_SRPT_SQ_SIZE = 4096, + SRPT_RQ_SIZE = 128, + MIN_SRPT_SRQ_SIZE = 4, + DEFAULT_SRPT_SRQ_SIZE = 4095, + MAX_SRPT_SRQ_SIZE = 65535, + MAX_SRPT_RDMA_SIZE = 1U << 24, + MAX_SRPT_RSP_SIZE = 1024, + + MIN_MAX_REQ_SIZE = 996, + DEFAULT_MAX_REQ_SIZE + = sizeof(struct srp_cmd)/*48*/ + + sizeof(struct srp_indirect_buf)/*20*/ + + 128 * sizeof(struct srp_direct_buf)/*16*/, + + MIN_MAX_RSP_SIZE = sizeof(struct srp_rsp)/*36*/ + 4, + DEFAULT_MAX_RSP_SIZE = 256, /* leaves 220 bytes for sense data */ + + DEFAULT_MAX_RDMA_SIZE = 65536, +}; + +enum srpt_opcode { + SRPT_RECV, + SRPT_SEND, + SRPT_RDMA_MID, + SRPT_RDMA_ABORT, + SRPT_RDMA_READ_LAST, + SRPT_RDMA_WRITE_LAST, +}; + +static inline u64 encode_wr_id(u8 opcode, u32 idx) +{ + return ((u64)opcode << 32) | idx; +} +static inline enum srpt_opcode opcode_from_wr_id(u64 wr_id) +{ + return wr_id >> 32; +} +static inline u32 idx_from_wr_id(u64 wr_id) +{ + return (u32)wr_id; +} + +struct rdma_iu { + u64 raddr; + u32 rkey; + struct ib_sge *sge; + u32 sge_cnt; + int mem_id; +}; + +/** + * enum srpt_command_state - SCSI command state managed by SRPT. + * @SRPT_STATE_NEW: New command arrived and is being processed. + * @SRPT_STATE_NEED_DATA: Processing a write or bidir command and waiting + * for data arrival. + * @SRPT_STATE_DATA_IN: Data for the write or bidir command arrived and is + * being processed. + * @SRPT_STATE_CMD_RSP_SENT: SRP_RSP for SRP_CMD has been sent. + * @SRPT_STATE_MGMT: Processing a SCSI task management command. + * @SRPT_STATE_MGMT_RSP_SENT: SRP_RSP for SRP_TSK_MGMT has been sent. + * @SRPT_STATE_DONE: Command processing finished successfully, command + * processing has been aborted or command processing + * failed. + */ +enum srpt_command_state { + SRPT_STATE_NEW = 0, + SRPT_STATE_NEED_DATA = 1, + SRPT_STATE_DATA_IN = 2, + SRPT_STATE_CMD_RSP_SENT = 3, + SRPT_STATE_MGMT = 4, + SRPT_STATE_MGMT_RSP_SENT = 5, + SRPT_STATE_DONE = 6, +}; + +/** + * struct srpt_ioctx - Shared SRPT I/O context information. + * @buf: Pointer to the buffer. + * @dma: DMA address of the buffer. + * @index: Index of the I/O context in its ioctx_ring array. + */ +struct srpt_ioctx { + void *buf; + dma_addr_t dma; + uint32_t index; +}; + +/** + * struct srpt_recv_ioctx - SRPT receive I/O context. + * @ioctx: See above. + * @wait_list: Node for insertion in srpt_rdma_ch.cmd_wait_list. + */ +struct srpt_recv_ioctx { + struct srpt_ioctx ioctx; + struct list_head wait_list; +}; + +/** + * struct srpt_send_ioctx - SRPT send I/O context. + * @ioctx: See above. + * @ch: Channel pointer. + * @free_list: Node in srpt_rdma_ch.free_list. + * @n_rbuf: Number of data buffers in the received SRP command. + * @rbufs: Pointer to SRP data buffer array. + * @single_rbuf: SRP data buffer if the command has only a single buffer. + * @sg: Pointer to sg-list associated with this I/O context. + * @sg_cnt: SG-list size. + * @mapped_sg_count: ib_dma_map_sg() return value. + * @n_rdma_ius: Number of elements in the rdma_ius array. + * @rdma_ius: Array with information about the RDMA mapping. + * @tag: Tag of the received SRP information unit. + * @spinlock: Protects 'state'. + * @state: I/O context state. + * @rdma_aborted: If initiating a multipart RDMA transfer failed, whether + * the already initiated transfers have finished. + * @cmd: Target core command data structure. + * @sense_data: SCSI sense data. + */ +struct srpt_send_ioctx { + struct srpt_ioctx ioctx; + struct srpt_rdma_ch *ch; + struct kref kref; + struct rdma_iu *rdma_ius; + struct srp_direct_buf *rbufs; + struct srp_direct_buf single_rbuf; + struct scatterlist *sg; + struct list_head free_list; + spinlock_t spinlock; + enum srpt_command_state state; + bool rdma_aborted; + struct se_cmd cmd; + struct completion tx_done; + u64 tag; + int sg_cnt; + int mapped_sg_count; + u16 n_rdma_ius; + u8 n_rdma; + u8 n_rbuf; + bool queue_status_only; + u8 sense_data[SCSI_SENSE_BUFFERSIZE]; +}; + +/** + * enum rdma_ch_state - SRP channel state. + * @CH_CONNECTING: QP is in RTR state; waiting for RTU. + * @CH_LIVE: QP is in RTS state. + * @CH_DISCONNECTING: DREQ has been received; waiting for DREP + * or DREQ has been send and waiting for DREP + * or . + * @CH_DRAINING: QP is in ERR state; waiting for last WQE event. + * @CH_RELEASING: Last WQE event has been received; releasing resources. + */ +enum rdma_ch_state { + CH_CONNECTING, + CH_LIVE, + CH_DISCONNECTING, + CH_DRAINING, + CH_RELEASING +}; + +/** + * struct srpt_rdma_ch - RDMA channel. + * @wait_queue: Allows the kernel thread to wait for more work. + * @thread: Kernel thread that processes the IB queues associated with + * the channel. + * @cm_id: IB CM ID associated with the channel. + * @qp: IB queue pair used for communicating over this channel. + * @cq: IB completion queue for this channel. + * @rq_size: IB receive queue size. + * @rsp_size IB response message size in bytes. + * @sq_wr_avail: number of work requests available in the send queue. + * @sport: pointer to the information of the HCA port used by this + * channel. + * @i_port_id: 128-bit initiator port identifier copied from SRP_LOGIN_REQ. + * @t_port_id: 128-bit target port identifier copied from SRP_LOGIN_REQ. + * @max_ti_iu_len: maximum target-to-initiator information unit length. + * @req_lim: request limit: maximum number of requests that may be sent + * by the initiator without having received a response. + * @req_lim_delta: Number of credits not yet sent back to the initiator. + * @spinlock: Protects free_list and state. + * @free_list: Head of list with free send I/O contexts. + * @state: channel state. See also enum rdma_ch_state. + * @ioctx_ring: Send ring. + * @wc: IB work completion array for srpt_process_completion(). + * @list: Node for insertion in the srpt_device.rch_list list. + * @cmd_wait_list: List of SCSI commands that arrived before the RTU event. This + * list contains struct srpt_ioctx elements and is protected + * against concurrent modification by the cm_id spinlock. + * @sess: Session information associated with this SRP channel. + * @sess_name: Session name. + * @release_work: Allows scheduling of srpt_release_channel(). + * @release_done: Enables waiting for srpt_release_channel() completion. + */ +struct srpt_rdma_ch { + wait_queue_head_t wait_queue; + struct task_struct *thread; + struct ib_cm_id *cm_id; + struct ib_qp *qp; + struct ib_cq *cq; + int rq_size; + u32 rsp_size; + atomic_t sq_wr_avail; + struct srpt_port *sport; + u8 i_port_id[16]; + u8 t_port_id[16]; + int max_ti_iu_len; + atomic_t req_lim; + atomic_t req_lim_delta; + spinlock_t spinlock; + struct list_head free_list; + enum rdma_ch_state state; + struct srpt_send_ioctx **ioctx_ring; + struct ib_wc wc[16]; + struct list_head list; + struct list_head cmd_wait_list; + struct se_session *sess; + u8 sess_name[36]; + struct work_struct release_work; + struct completion *release_done; +}; + +/** + * struct srpt_port_attib - Attributes for SRPT port + * @srp_max_rdma_size: Maximum size of SRP RDMA transfers for new connections. + * @srp_max_rsp_size: Maximum size of SRP response messages in bytes. + * @srp_sq_size: Shared receive queue (SRQ) size. + */ +struct srpt_port_attrib { + u32 srp_max_rdma_size; + u32 srp_max_rsp_size; + u32 srp_sq_size; +}; + +/** + * struct srpt_port - Information associated by SRPT with a single IB port. + * @sdev: backpointer to the HCA information. + * @mad_agent: per-port management datagram processing information. + * @enabled: Whether or not this target port is enabled. + * @port_guid: ASCII representation of Port GUID + * @port: one-based port number. + * @sm_lid: cached value of the port's sm_lid. + * @lid: cached value of the port's lid. + * @gid: cached value of the port's gid. + * @port_acl_lock spinlock for port_acl_list: + * @work: work structure for refreshing the aforementioned cached values. + * @port_tpg_1 Target portal group = 1 data. + * @port_wwn: Target core WWN data. + * @port_acl_list: Head of the list with all node ACLs for this port. + */ +struct srpt_port { + struct srpt_device *sdev; + struct ib_mad_agent *mad_agent; + bool enabled; + u8 port_guid[64]; + u8 port; + u16 sm_lid; + u16 lid; + union ib_gid gid; + spinlock_t port_acl_lock; + struct work_struct work; + struct se_portal_group port_tpg_1; + struct se_wwn port_wwn; + struct list_head port_acl_list; + struct srpt_port_attrib port_attrib; +}; + +/** + * struct srpt_device - Information associated by SRPT with a single HCA. + * @device: Backpointer to the struct ib_device managed by the IB core. + * @pd: IB protection domain. + * @mr: L_Key (local key) with write access to all local memory. + * @srq: Per-HCA SRQ (shared receive queue). + * @cm_id: Connection identifier. + * @dev_attr: Attributes of the InfiniBand device as obtained during the + * ib_client.add() callback. + * @srq_size: SRQ size. + * @ioctx_ring: Per-HCA SRQ. + * @rch_list: Per-device channel list -- see also srpt_rdma_ch.list. + * @ch_releaseQ: Enables waiting for removal from rch_list. + * @spinlock: Protects rch_list and tpg. + * @port: Information about the ports owned by this HCA. + * @event_handler: Per-HCA asynchronous IB event handler. + * @list: Node in srpt_dev_list. + */ +struct srpt_device { + struct ib_device *device; + struct ib_pd *pd; + struct ib_mr *mr; + struct ib_srq *srq; + struct ib_cm_id *cm_id; + struct ib_device_attr dev_attr; + int srq_size; + struct srpt_recv_ioctx **ioctx_ring; + struct list_head rch_list; + wait_queue_head_t ch_releaseQ; + spinlock_t spinlock; + struct srpt_port port[2]; + struct ib_event_handler event_handler; + struct list_head list; +}; + +/** + * struct srpt_node_acl - Per-initiator ACL data (managed via configfs). + * @i_port_id: 128-bit SRP initiator port ID. + * @sport: port information. + * @nacl: Target core node ACL information. + * @list: Element of the per-HCA ACL list. + */ +struct srpt_node_acl { + u8 i_port_id[16]; + struct srpt_port *sport; + struct se_node_acl nacl; + struct list_head list; +}; + +/* + * SRP-releated SCSI persistent reservation definitions. + * + * See also SPC4r28, section 7.6.1 (Protocol specific parameters introduction). + * See also SPC4r28, section 7.6.4.5 (TransportID for initiator ports using + * SCSI over an RDMA interface). + */ + +enum { + SCSI_TRANSPORTID_PROTOCOLID_SRP = 4, +}; + +struct spc_rdma_transport_id { + uint8_t protocol_identifier; + uint8_t reserved[7]; + uint8_t i_port_id[16]; +}; + +#endif /* IB_SRPT_H */ From 5db0276014b80484689eb6c1bf7b94af1c7d5b1a Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Tue, 1 Nov 2011 17:04:16 +0100 Subject: [PATCH 165/820] Btrfs: add optional integrity check code The two files added in this patch contain all the code that is required to implement the integrity checks. Signed-off-by: Stefan Behrens --- fs/btrfs/check-integrity.c | 3068 ++++++++++++++++++++++++++++++++++++ fs/btrfs/check-integrity.h | 36 + 2 files changed, 3104 insertions(+) create mode 100644 fs/btrfs/check-integrity.c create mode 100644 fs/btrfs/check-integrity.h diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c new file mode 100644 index 000000000000..ad0b3ba735b7 --- /dev/null +++ b/fs/btrfs/check-integrity.c @@ -0,0 +1,3068 @@ +/* + * Copyright (C) STRATO AG 2011. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ + +/* + * This module can be used to catch cases when the btrfs kernel + * code executes write requests to the disk that bring the file + * system in an inconsistent state. In such a state, a power-loss + * or kernel panic event would cause that the data on disk is + * lost or at least damaged. + * + * Code is added that examines all block write requests during + * runtime (including writes of the super block). Three rules + * are verified and an error is printed on violation of the + * rules: + * 1. It is not allowed to write a disk block which is + * currently referenced by the super block (either directly + * or indirectly). + * 2. When a super block is written, it is verified that all + * referenced (directly or indirectly) blocks fulfill the + * following requirements: + * 2a. All referenced blocks have either been present when + * the file system was mounted, (i.e., they have been + * referenced by the super block) or they have been + * written since then and the write completion callback + * was called and a FLUSH request to the device where + * these blocks are located was received and completed. + * 2b. All referenced blocks need to have a generation + * number which is equal to the parent's number. + * + * One issue that was found using this module was that the log + * tree on disk became temporarily corrupted because disk blocks + * that had been in use for the log tree had been freed and + * reused too early, while being referenced by the written super + * block. + * + * The search term in the kernel log that can be used to filter + * on the existence of detected integrity issues is + * "btrfs: attempt". + * + * The integrity check is enabled via mount options. These + * mount options are only supported if the integrity check + * tool is compiled by defining BTRFS_FS_CHECK_INTEGRITY. + * + * Example #1, apply integrity checks to all metadata: + * mount /dev/sdb1 /mnt -o check_int + * + * Example #2, apply integrity checks to all metadata and + * to data extents: + * mount /dev/sdb1 /mnt -o check_int_data + * + * Example #3, apply integrity checks to all metadata and dump + * the tree that the super block references to kernel messages + * each time after a super block was written: + * mount /dev/sdb1 /mnt -o check_int,check_int_print_mask=263 + * + * If the integrity check tool is included and activated in + * the mount options, plenty of kernel memory is used, and + * plenty of additional CPU cycles are spent. Enabling this + * functionality is not intended for normal use. In most + * cases, unless you are a btrfs developer who needs to verify + * the integrity of (super)-block write requests, do not + * enable the config option BTRFS_FS_CHECK_INTEGRITY to + * include and compile the integrity check tool. + */ + +#include +#include +#include +#include +#include +#include +#include +#include "ctree.h" +#include "disk-io.h" +#include "transaction.h" +#include "extent_io.h" +#include "disk-io.h" +#include "volumes.h" +#include "print-tree.h" +#include "locking.h" +#include "check-integrity.h" + +#define BTRFSIC_BLOCK_HASHTABLE_SIZE 0x10000 +#define BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE 0x10000 +#define BTRFSIC_DEV2STATE_HASHTABLE_SIZE 0x100 +#define BTRFSIC_BLOCK_MAGIC_NUMBER 0x14491051 +#define BTRFSIC_BLOCK_LINK_MAGIC_NUMBER 0x11070807 +#define BTRFSIC_DEV2STATE_MAGIC_NUMBER 0x20111530 +#define BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER 20111300 +#define BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL (200 - 6) /* in characters, + * excluding " [...]" */ +#define BTRFSIC_BLOCK_SIZE PAGE_SIZE + +#define BTRFSIC_GENERATION_UNKNOWN ((u64)-1) + +/* + * The definition of the bitmask fields for the print_mask. + * They are specified with the mount option check_integrity_print_mask. + */ +#define BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE 0x00000001 +#define BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION 0x00000002 +#define BTRFSIC_PRINT_MASK_TREE_AFTER_SB_WRITE 0x00000004 +#define BTRFSIC_PRINT_MASK_TREE_BEFORE_SB_WRITE 0x00000008 +#define BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH 0x00000010 +#define BTRFSIC_PRINT_MASK_END_IO_BIO_BH 0x00000020 +#define BTRFSIC_PRINT_MASK_VERBOSE 0x00000040 +#define BTRFSIC_PRINT_MASK_VERY_VERBOSE 0x00000080 +#define BTRFSIC_PRINT_MASK_INITIAL_TREE 0x00000100 +#define BTRFSIC_PRINT_MASK_INITIAL_ALL_TREES 0x00000200 +#define BTRFSIC_PRINT_MASK_INITIAL_DATABASE 0x00000400 +#define BTRFSIC_PRINT_MASK_NUM_COPIES 0x00000800 +#define BTRFSIC_PRINT_MASK_TREE_WITH_ALL_MIRRORS 0x00001000 + +struct btrfsic_dev_state; +struct btrfsic_state; + +struct btrfsic_block { + u32 magic_num; /* only used for debug purposes */ + unsigned int is_metadata:1; /* if it is meta-data, not data-data */ + unsigned int is_superblock:1; /* if it is one of the superblocks */ + unsigned int is_iodone:1; /* if is done by lower subsystem */ + unsigned int iodone_w_error:1; /* error was indicated to endio */ + unsigned int never_written:1; /* block was added because it was + * referenced, not because it was + * written */ + unsigned int mirror_num:2; /* large enough to hold + * BTRFS_SUPER_MIRROR_MAX */ + struct btrfsic_dev_state *dev_state; + u64 dev_bytenr; /* key, physical byte num on disk */ + u64 logical_bytenr; /* logical byte num on disk */ + u64 generation; + struct btrfs_disk_key disk_key; /* extra info to print in case of + * issues, will not always be correct */ + struct list_head collision_resolving_node; /* list node */ + struct list_head all_blocks_node; /* list node */ + + /* the following two lists contain block_link items */ + struct list_head ref_to_list; /* list */ + struct list_head ref_from_list; /* list */ + struct btrfsic_block *next_in_same_bio; + void *orig_bio_bh_private; + union { + bio_end_io_t *bio; + bh_end_io_t *bh; + } orig_bio_bh_end_io; + int submit_bio_bh_rw; + u64 flush_gen; /* only valid if !never_written */ +}; + +/* + * Elements of this type are allocated dynamically and required because + * each block object can refer to and can be ref from multiple blocks. + * The key to lookup them in the hashtable is the dev_bytenr of + * the block ref to plus the one from the block refered from. + * The fact that they are searchable via a hashtable and that a + * ref_cnt is maintained is not required for the btrfs integrity + * check algorithm itself, it is only used to make the output more + * beautiful in case that an error is detected (an error is defined + * as a write operation to a block while that block is still referenced). + */ +struct btrfsic_block_link { + u32 magic_num; /* only used for debug purposes */ + u32 ref_cnt; + struct list_head node_ref_to; /* list node */ + struct list_head node_ref_from; /* list node */ + struct list_head collision_resolving_node; /* list node */ + struct btrfsic_block *block_ref_to; + struct btrfsic_block *block_ref_from; + u64 parent_generation; +}; + +struct btrfsic_dev_state { + u32 magic_num; /* only used for debug purposes */ + struct block_device *bdev; + struct btrfsic_state *state; + struct list_head collision_resolving_node; /* list node */ + struct btrfsic_block dummy_block_for_bio_bh_flush; + u64 last_flush_gen; + char name[BDEVNAME_SIZE]; +}; + +struct btrfsic_block_hashtable { + struct list_head table[BTRFSIC_BLOCK_HASHTABLE_SIZE]; +}; + +struct btrfsic_block_link_hashtable { + struct list_head table[BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE]; +}; + +struct btrfsic_dev_state_hashtable { + struct list_head table[BTRFSIC_DEV2STATE_HASHTABLE_SIZE]; +}; + +struct btrfsic_block_data_ctx { + u64 start; /* virtual bytenr */ + u64 dev_bytenr; /* physical bytenr on device */ + u32 len; + struct btrfsic_dev_state *dev; + char *data; + struct buffer_head *bh; /* do not use if set to NULL */ +}; + +/* This structure is used to implement recursion without occupying + * any stack space, refer to btrfsic_process_metablock() */ +struct btrfsic_stack_frame { + u32 magic; + u32 nr; + int error; + int i; + int limit_nesting; + int num_copies; + int mirror_num; + struct btrfsic_block *block; + struct btrfsic_block_data_ctx *block_ctx; + struct btrfsic_block *next_block; + struct btrfsic_block_data_ctx next_block_ctx; + struct btrfs_header *hdr; + struct btrfsic_stack_frame *prev; +}; + +/* Some state per mounted filesystem */ +struct btrfsic_state { + u32 print_mask; + int include_extent_data; + int csum_size; + struct list_head all_blocks_list; + struct btrfsic_block_hashtable block_hashtable; + struct btrfsic_block_link_hashtable block_link_hashtable; + struct btrfs_root *root; + u64 max_superblock_generation; + struct btrfsic_block *latest_superblock; +}; + +static void btrfsic_block_init(struct btrfsic_block *b); +static struct btrfsic_block *btrfsic_block_alloc(void); +static void btrfsic_block_free(struct btrfsic_block *b); +static void btrfsic_block_link_init(struct btrfsic_block_link *n); +static struct btrfsic_block_link *btrfsic_block_link_alloc(void); +static void btrfsic_block_link_free(struct btrfsic_block_link *n); +static void btrfsic_dev_state_init(struct btrfsic_dev_state *ds); +static struct btrfsic_dev_state *btrfsic_dev_state_alloc(void); +static void btrfsic_dev_state_free(struct btrfsic_dev_state *ds); +static void btrfsic_block_hashtable_init(struct btrfsic_block_hashtable *h); +static void btrfsic_block_hashtable_add(struct btrfsic_block *b, + struct btrfsic_block_hashtable *h); +static void btrfsic_block_hashtable_remove(struct btrfsic_block *b); +static struct btrfsic_block *btrfsic_block_hashtable_lookup( + struct block_device *bdev, + u64 dev_bytenr, + struct btrfsic_block_hashtable *h); +static void btrfsic_block_link_hashtable_init( + struct btrfsic_block_link_hashtable *h); +static void btrfsic_block_link_hashtable_add( + struct btrfsic_block_link *l, + struct btrfsic_block_link_hashtable *h); +static void btrfsic_block_link_hashtable_remove(struct btrfsic_block_link *l); +static struct btrfsic_block_link *btrfsic_block_link_hashtable_lookup( + struct block_device *bdev_ref_to, + u64 dev_bytenr_ref_to, + struct block_device *bdev_ref_from, + u64 dev_bytenr_ref_from, + struct btrfsic_block_link_hashtable *h); +static void btrfsic_dev_state_hashtable_init( + struct btrfsic_dev_state_hashtable *h); +static void btrfsic_dev_state_hashtable_add( + struct btrfsic_dev_state *ds, + struct btrfsic_dev_state_hashtable *h); +static void btrfsic_dev_state_hashtable_remove(struct btrfsic_dev_state *ds); +static struct btrfsic_dev_state *btrfsic_dev_state_hashtable_lookup( + struct block_device *bdev, + struct btrfsic_dev_state_hashtable *h); +static struct btrfsic_stack_frame *btrfsic_stack_frame_alloc(void); +static void btrfsic_stack_frame_free(struct btrfsic_stack_frame *sf); +static int btrfsic_process_superblock(struct btrfsic_state *state, + struct btrfs_fs_devices *fs_devices); +static int btrfsic_process_metablock(struct btrfsic_state *state, + struct btrfsic_block *block, + struct btrfsic_block_data_ctx *block_ctx, + struct btrfs_header *hdr, + int limit_nesting, int force_iodone_flag); +static int btrfsic_create_link_to_next_block( + struct btrfsic_state *state, + struct btrfsic_block *block, + struct btrfsic_block_data_ctx + *block_ctx, u64 next_bytenr, + int limit_nesting, + struct btrfsic_block_data_ctx *next_block_ctx, + struct btrfsic_block **next_blockp, + int force_iodone_flag, + int *num_copiesp, int *mirror_nump, + struct btrfs_disk_key *disk_key, + u64 parent_generation); +static int btrfsic_handle_extent_data(struct btrfsic_state *state, + struct btrfsic_block *block, + struct btrfsic_block_data_ctx *block_ctx, + u32 item_offset, int force_iodone_flag); +static int btrfsic_map_block(struct btrfsic_state *state, u64 bytenr, u32 len, + struct btrfsic_block_data_ctx *block_ctx_out, + int mirror_num); +static int btrfsic_map_superblock(struct btrfsic_state *state, u64 bytenr, + u32 len, struct block_device *bdev, + struct btrfsic_block_data_ctx *block_ctx_out); +static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx); +static int btrfsic_read_block(struct btrfsic_state *state, + struct btrfsic_block_data_ctx *block_ctx); +static void btrfsic_dump_database(struct btrfsic_state *state); +static int btrfsic_test_for_metadata(struct btrfsic_state *state, + const u8 *data, unsigned int size); +static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state, + u64 dev_bytenr, u8 *mapped_data, + unsigned int len, struct bio *bio, + int *bio_is_patched, + struct buffer_head *bh, + int submit_bio_bh_rw); +static int btrfsic_process_written_superblock( + struct btrfsic_state *state, + struct btrfsic_block *const block, + struct btrfs_super_block *const super_hdr); +static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status); +static void btrfsic_bh_end_io(struct buffer_head *bh, int uptodate); +static int btrfsic_is_block_ref_by_superblock(const struct btrfsic_state *state, + const struct btrfsic_block *block, + int recursion_level); +static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state, + struct btrfsic_block *const block, + int recursion_level); +static void btrfsic_print_add_link(const struct btrfsic_state *state, + const struct btrfsic_block_link *l); +static void btrfsic_print_rem_link(const struct btrfsic_state *state, + const struct btrfsic_block_link *l); +static char btrfsic_get_block_type(const struct btrfsic_state *state, + const struct btrfsic_block *block); +static void btrfsic_dump_tree(const struct btrfsic_state *state); +static void btrfsic_dump_tree_sub(const struct btrfsic_state *state, + const struct btrfsic_block *block, + int indent_level); +static struct btrfsic_block_link *btrfsic_block_link_lookup_or_add( + struct btrfsic_state *state, + struct btrfsic_block_data_ctx *next_block_ctx, + struct btrfsic_block *next_block, + struct btrfsic_block *from_block, + u64 parent_generation); +static struct btrfsic_block *btrfsic_block_lookup_or_add( + struct btrfsic_state *state, + struct btrfsic_block_data_ctx *block_ctx, + const char *additional_string, + int is_metadata, + int is_iodone, + int never_written, + int mirror_num, + int *was_created); +static int btrfsic_process_superblock_dev_mirror( + struct btrfsic_state *state, + struct btrfsic_dev_state *dev_state, + struct btrfs_device *device, + int superblock_mirror_num, + struct btrfsic_dev_state **selected_dev_state, + struct btrfs_super_block *selected_super); +static struct btrfsic_dev_state *btrfsic_dev_state_lookup( + struct block_device *bdev); +static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state, + u64 bytenr, + struct btrfsic_dev_state *dev_state, + u64 dev_bytenr, char *data); + +static struct mutex btrfsic_mutex; +static int btrfsic_is_initialized; +static struct btrfsic_dev_state_hashtable btrfsic_dev_state_hashtable; + + +static void btrfsic_block_init(struct btrfsic_block *b) +{ + b->magic_num = BTRFSIC_BLOCK_MAGIC_NUMBER; + b->dev_state = NULL; + b->dev_bytenr = 0; + b->logical_bytenr = 0; + b->generation = BTRFSIC_GENERATION_UNKNOWN; + b->disk_key.objectid = 0; + b->disk_key.type = 0; + b->disk_key.offset = 0; + b->is_metadata = 0; + b->is_superblock = 0; + b->is_iodone = 0; + b->iodone_w_error = 0; + b->never_written = 0; + b->mirror_num = 0; + b->next_in_same_bio = NULL; + b->orig_bio_bh_private = NULL; + b->orig_bio_bh_end_io.bio = NULL; + INIT_LIST_HEAD(&b->collision_resolving_node); + INIT_LIST_HEAD(&b->all_blocks_node); + INIT_LIST_HEAD(&b->ref_to_list); + INIT_LIST_HEAD(&b->ref_from_list); + b->submit_bio_bh_rw = 0; + b->flush_gen = 0; +} + +static struct btrfsic_block *btrfsic_block_alloc(void) +{ + struct btrfsic_block *b; + + b = kzalloc(sizeof(*b), GFP_NOFS); + if (NULL != b) + btrfsic_block_init(b); + + return b; +} + +static void btrfsic_block_free(struct btrfsic_block *b) +{ + BUG_ON(!(NULL == b || BTRFSIC_BLOCK_MAGIC_NUMBER == b->magic_num)); + kfree(b); +} + +static void btrfsic_block_link_init(struct btrfsic_block_link *l) +{ + l->magic_num = BTRFSIC_BLOCK_LINK_MAGIC_NUMBER; + l->ref_cnt = 1; + INIT_LIST_HEAD(&l->node_ref_to); + INIT_LIST_HEAD(&l->node_ref_from); + INIT_LIST_HEAD(&l->collision_resolving_node); + l->block_ref_to = NULL; + l->block_ref_from = NULL; +} + +static struct btrfsic_block_link *btrfsic_block_link_alloc(void) +{ + struct btrfsic_block_link *l; + + l = kzalloc(sizeof(*l), GFP_NOFS); + if (NULL != l) + btrfsic_block_link_init(l); + + return l; +} + +static void btrfsic_block_link_free(struct btrfsic_block_link *l) +{ + BUG_ON(!(NULL == l || BTRFSIC_BLOCK_LINK_MAGIC_NUMBER == l->magic_num)); + kfree(l); +} + +static void btrfsic_dev_state_init(struct btrfsic_dev_state *ds) +{ + ds->magic_num = BTRFSIC_DEV2STATE_MAGIC_NUMBER; + ds->bdev = NULL; + ds->state = NULL; + ds->name[0] = '\0'; + INIT_LIST_HEAD(&ds->collision_resolving_node); + ds->last_flush_gen = 0; + btrfsic_block_init(&ds->dummy_block_for_bio_bh_flush); + ds->dummy_block_for_bio_bh_flush.is_iodone = 1; + ds->dummy_block_for_bio_bh_flush.dev_state = ds; +} + +static struct btrfsic_dev_state *btrfsic_dev_state_alloc(void) +{ + struct btrfsic_dev_state *ds; + + ds = kzalloc(sizeof(*ds), GFP_NOFS); + if (NULL != ds) + btrfsic_dev_state_init(ds); + + return ds; +} + +static void btrfsic_dev_state_free(struct btrfsic_dev_state *ds) +{ + BUG_ON(!(NULL == ds || + BTRFSIC_DEV2STATE_MAGIC_NUMBER == ds->magic_num)); + kfree(ds); +} + +static void btrfsic_block_hashtable_init(struct btrfsic_block_hashtable *h) +{ + int i; + + for (i = 0; i < BTRFSIC_BLOCK_HASHTABLE_SIZE; i++) + INIT_LIST_HEAD(h->table + i); +} + +static void btrfsic_block_hashtable_add(struct btrfsic_block *b, + struct btrfsic_block_hashtable *h) +{ + const unsigned int hashval = + (((unsigned int)(b->dev_bytenr >> 16)) ^ + ((unsigned int)((uintptr_t)b->dev_state->bdev))) & + (BTRFSIC_BLOCK_HASHTABLE_SIZE - 1); + + list_add(&b->collision_resolving_node, h->table + hashval); +} + +static void btrfsic_block_hashtable_remove(struct btrfsic_block *b) +{ + list_del(&b->collision_resolving_node); +} + +static struct btrfsic_block *btrfsic_block_hashtable_lookup( + struct block_device *bdev, + u64 dev_bytenr, + struct btrfsic_block_hashtable *h) +{ + const unsigned int hashval = + (((unsigned int)(dev_bytenr >> 16)) ^ + ((unsigned int)((uintptr_t)bdev))) & + (BTRFSIC_BLOCK_HASHTABLE_SIZE - 1); + struct list_head *elem; + + list_for_each(elem, h->table + hashval) { + struct btrfsic_block *const b = + list_entry(elem, struct btrfsic_block, + collision_resolving_node); + + if (b->dev_state->bdev == bdev && b->dev_bytenr == dev_bytenr) + return b; + } + + return NULL; +} + +static void btrfsic_block_link_hashtable_init( + struct btrfsic_block_link_hashtable *h) +{ + int i; + + for (i = 0; i < BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE; i++) + INIT_LIST_HEAD(h->table + i); +} + +static void btrfsic_block_link_hashtable_add( + struct btrfsic_block_link *l, + struct btrfsic_block_link_hashtable *h) +{ + const unsigned int hashval = + (((unsigned int)(l->block_ref_to->dev_bytenr >> 16)) ^ + ((unsigned int)(l->block_ref_from->dev_bytenr >> 16)) ^ + ((unsigned int)((uintptr_t)l->block_ref_to->dev_state->bdev)) ^ + ((unsigned int)((uintptr_t)l->block_ref_from->dev_state->bdev))) + & (BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE - 1); + + BUG_ON(NULL == l->block_ref_to); + BUG_ON(NULL == l->block_ref_from); + list_add(&l->collision_resolving_node, h->table + hashval); +} + +static void btrfsic_block_link_hashtable_remove(struct btrfsic_block_link *l) +{ + list_del(&l->collision_resolving_node); +} + +static struct btrfsic_block_link *btrfsic_block_link_hashtable_lookup( + struct block_device *bdev_ref_to, + u64 dev_bytenr_ref_to, + struct block_device *bdev_ref_from, + u64 dev_bytenr_ref_from, + struct btrfsic_block_link_hashtable *h) +{ + const unsigned int hashval = + (((unsigned int)(dev_bytenr_ref_to >> 16)) ^ + ((unsigned int)(dev_bytenr_ref_from >> 16)) ^ + ((unsigned int)((uintptr_t)bdev_ref_to)) ^ + ((unsigned int)((uintptr_t)bdev_ref_from))) & + (BTRFSIC_BLOCK_LINK_HASHTABLE_SIZE - 1); + struct list_head *elem; + + list_for_each(elem, h->table + hashval) { + struct btrfsic_block_link *const l = + list_entry(elem, struct btrfsic_block_link, + collision_resolving_node); + + BUG_ON(NULL == l->block_ref_to); + BUG_ON(NULL == l->block_ref_from); + if (l->block_ref_to->dev_state->bdev == bdev_ref_to && + l->block_ref_to->dev_bytenr == dev_bytenr_ref_to && + l->block_ref_from->dev_state->bdev == bdev_ref_from && + l->block_ref_from->dev_bytenr == dev_bytenr_ref_from) + return l; + } + + return NULL; +} + +static void btrfsic_dev_state_hashtable_init( + struct btrfsic_dev_state_hashtable *h) +{ + int i; + + for (i = 0; i < BTRFSIC_DEV2STATE_HASHTABLE_SIZE; i++) + INIT_LIST_HEAD(h->table + i); +} + +static void btrfsic_dev_state_hashtable_add( + struct btrfsic_dev_state *ds, + struct btrfsic_dev_state_hashtable *h) +{ + const unsigned int hashval = + (((unsigned int)((uintptr_t)ds->bdev)) & + (BTRFSIC_DEV2STATE_HASHTABLE_SIZE - 1)); + + list_add(&ds->collision_resolving_node, h->table + hashval); +} + +static void btrfsic_dev_state_hashtable_remove(struct btrfsic_dev_state *ds) +{ + list_del(&ds->collision_resolving_node); +} + +static struct btrfsic_dev_state *btrfsic_dev_state_hashtable_lookup( + struct block_device *bdev, + struct btrfsic_dev_state_hashtable *h) +{ + const unsigned int hashval = + (((unsigned int)((uintptr_t)bdev)) & + (BTRFSIC_DEV2STATE_HASHTABLE_SIZE - 1)); + struct list_head *elem; + + list_for_each(elem, h->table + hashval) { + struct btrfsic_dev_state *const ds = + list_entry(elem, struct btrfsic_dev_state, + collision_resolving_node); + + if (ds->bdev == bdev) + return ds; + } + + return NULL; +} + +static int btrfsic_process_superblock(struct btrfsic_state *state, + struct btrfs_fs_devices *fs_devices) +{ + int ret; + struct btrfs_super_block *selected_super; + struct list_head *dev_head = &fs_devices->devices; + struct btrfs_device *device; + struct btrfsic_dev_state *selected_dev_state = NULL; + int pass; + + BUG_ON(NULL == state); + selected_super = kmalloc(sizeof(*selected_super), GFP_NOFS); + if (NULL == selected_super) { + printk(KERN_INFO "btrfsic: error, kmalloc failed!\n"); + return -1; + } + + list_for_each_entry(device, dev_head, dev_list) { + int i; + struct btrfsic_dev_state *dev_state; + + if (!device->bdev || !device->name) + continue; + + dev_state = btrfsic_dev_state_lookup(device->bdev); + BUG_ON(NULL == dev_state); + for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { + ret = btrfsic_process_superblock_dev_mirror( + state, dev_state, device, i, + &selected_dev_state, selected_super); + if (0 != ret && 0 == i) { + kfree(selected_super); + return ret; + } + } + } + + if (NULL == state->latest_superblock) { + printk(KERN_INFO "btrfsic: no superblock found!\n"); + kfree(selected_super); + return -1; + } + + state->csum_size = btrfs_super_csum_size(selected_super); + + for (pass = 0; pass < 3; pass++) { + int num_copies; + int mirror_num; + u64 next_bytenr; + + switch (pass) { + case 0: + next_bytenr = btrfs_super_root(selected_super); + if (state->print_mask & + BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION) + printk(KERN_INFO "root@%llu\n", + (unsigned long long)next_bytenr); + break; + case 1: + next_bytenr = btrfs_super_chunk_root(selected_super); + if (state->print_mask & + BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION) + printk(KERN_INFO "chunk@%llu\n", + (unsigned long long)next_bytenr); + break; + case 2: + next_bytenr = btrfs_super_log_root(selected_super); + if (0 == next_bytenr) + continue; + if (state->print_mask & + BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION) + printk(KERN_INFO "log@%llu\n", + (unsigned long long)next_bytenr); + break; + } + + num_copies = + btrfs_num_copies(&state->root->fs_info->mapping_tree, + next_bytenr, PAGE_SIZE); + if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES) + printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n", + (unsigned long long)next_bytenr, num_copies); + + for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) { + struct btrfsic_block *next_block; + struct btrfsic_block_data_ctx tmp_next_block_ctx; + struct btrfsic_block_link *l; + struct btrfs_header *hdr; + + ret = btrfsic_map_block(state, next_bytenr, PAGE_SIZE, + &tmp_next_block_ctx, + mirror_num); + if (ret) { + printk(KERN_INFO "btrfsic:" + " btrfsic_map_block(root @%llu," + " mirror %d) failed!\n", + (unsigned long long)next_bytenr, + mirror_num); + kfree(selected_super); + return -1; + } + + next_block = btrfsic_block_hashtable_lookup( + tmp_next_block_ctx.dev->bdev, + tmp_next_block_ctx.dev_bytenr, + &state->block_hashtable); + BUG_ON(NULL == next_block); + + l = btrfsic_block_link_hashtable_lookup( + tmp_next_block_ctx.dev->bdev, + tmp_next_block_ctx.dev_bytenr, + state->latest_superblock->dev_state-> + bdev, + state->latest_superblock->dev_bytenr, + &state->block_link_hashtable); + BUG_ON(NULL == l); + + ret = btrfsic_read_block(state, &tmp_next_block_ctx); + if (ret < (int)BTRFSIC_BLOCK_SIZE) { + printk(KERN_INFO + "btrfsic: read @logical %llu failed!\n", + (unsigned long long) + tmp_next_block_ctx.start); + btrfsic_release_block_ctx(&tmp_next_block_ctx); + kfree(selected_super); + return -1; + } + + hdr = (struct btrfs_header *)tmp_next_block_ctx.data; + ret = btrfsic_process_metablock(state, + next_block, + &tmp_next_block_ctx, + hdr, + BTRFS_MAX_LEVEL + 3, 1); + btrfsic_release_block_ctx(&tmp_next_block_ctx); + } + } + + kfree(selected_super); + return ret; +} + +static int btrfsic_process_superblock_dev_mirror( + struct btrfsic_state *state, + struct btrfsic_dev_state *dev_state, + struct btrfs_device *device, + int superblock_mirror_num, + struct btrfsic_dev_state **selected_dev_state, + struct btrfs_super_block *selected_super) +{ + struct btrfs_super_block *super_tmp; + u64 dev_bytenr; + struct buffer_head *bh; + struct btrfsic_block *superblock_tmp; + int pass; + struct block_device *const superblock_bdev = device->bdev; + + /* super block bytenr is always the unmapped device bytenr */ + dev_bytenr = btrfs_sb_offset(superblock_mirror_num); + bh = __bread(superblock_bdev, dev_bytenr / 4096, 4096); + if (NULL == bh) + return -1; + super_tmp = (struct btrfs_super_block *) + (bh->b_data + (dev_bytenr & 4095)); + + if (btrfs_super_bytenr(super_tmp) != dev_bytenr || + strncmp((char *)(&(super_tmp->magic)), BTRFS_MAGIC, + sizeof(super_tmp->magic)) || + memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE)) { + brelse(bh); + return 0; + } + + superblock_tmp = + btrfsic_block_hashtable_lookup(superblock_bdev, + dev_bytenr, + &state->block_hashtable); + if (NULL == superblock_tmp) { + superblock_tmp = btrfsic_block_alloc(); + if (NULL == superblock_tmp) { + printk(KERN_INFO "btrfsic: error, kmalloc failed!\n"); + brelse(bh); + return -1; + } + /* for superblock, only the dev_bytenr makes sense */ + superblock_tmp->dev_bytenr = dev_bytenr; + superblock_tmp->dev_state = dev_state; + superblock_tmp->logical_bytenr = dev_bytenr; + superblock_tmp->generation = btrfs_super_generation(super_tmp); + superblock_tmp->is_metadata = 1; + superblock_tmp->is_superblock = 1; + superblock_tmp->is_iodone = 1; + superblock_tmp->never_written = 0; + superblock_tmp->mirror_num = 1 + superblock_mirror_num; + if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE) + printk(KERN_INFO "New initial S-block (bdev %p, %s)" + " @%llu (%s/%llu/%d)\n", + superblock_bdev, device->name, + (unsigned long long)dev_bytenr, + dev_state->name, + (unsigned long long)dev_bytenr, + superblock_mirror_num); + list_add(&superblock_tmp->all_blocks_node, + &state->all_blocks_list); + btrfsic_block_hashtable_add(superblock_tmp, + &state->block_hashtable); + } + + /* select the one with the highest generation field */ + if (btrfs_super_generation(super_tmp) > + state->max_superblock_generation || + 0 == state->max_superblock_generation) { + memcpy(selected_super, super_tmp, sizeof(*selected_super)); + *selected_dev_state = dev_state; + state->max_superblock_generation = + btrfs_super_generation(super_tmp); + state->latest_superblock = superblock_tmp; + } + + for (pass = 0; pass < 3; pass++) { + u64 next_bytenr; + int num_copies; + int mirror_num; + const char *additional_string = NULL; + struct btrfs_disk_key tmp_disk_key; + + tmp_disk_key.type = BTRFS_ROOT_ITEM_KEY; + tmp_disk_key.offset = 0; + switch (pass) { + case 0: + tmp_disk_key.objectid = + cpu_to_le64(BTRFS_ROOT_TREE_OBJECTID); + additional_string = "initial root "; + next_bytenr = btrfs_super_root(super_tmp); + break; + case 1: + tmp_disk_key.objectid = + cpu_to_le64(BTRFS_CHUNK_TREE_OBJECTID); + additional_string = "initial chunk "; + next_bytenr = btrfs_super_chunk_root(super_tmp); + break; + case 2: + tmp_disk_key.objectid = + cpu_to_le64(BTRFS_TREE_LOG_OBJECTID); + additional_string = "initial log "; + next_bytenr = btrfs_super_log_root(super_tmp); + if (0 == next_bytenr) + continue; + break; + } + + num_copies = + btrfs_num_copies(&state->root->fs_info->mapping_tree, + next_bytenr, PAGE_SIZE); + if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES) + printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n", + (unsigned long long)next_bytenr, num_copies); + for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) { + struct btrfsic_block *next_block; + struct btrfsic_block_data_ctx tmp_next_block_ctx; + struct btrfsic_block_link *l; + + if (btrfsic_map_block(state, next_bytenr, PAGE_SIZE, + &tmp_next_block_ctx, + mirror_num)) { + printk(KERN_INFO "btrfsic: btrfsic_map_block(" + "bytenr @%llu, mirror %d) failed!\n", + (unsigned long long)next_bytenr, + mirror_num); + brelse(bh); + return -1; + } + + next_block = btrfsic_block_lookup_or_add( + state, &tmp_next_block_ctx, + additional_string, 1, 1, 0, + mirror_num, NULL); + if (NULL == next_block) { + btrfsic_release_block_ctx(&tmp_next_block_ctx); + brelse(bh); + return -1; + } + + next_block->disk_key = tmp_disk_key; + next_block->generation = BTRFSIC_GENERATION_UNKNOWN; + l = btrfsic_block_link_lookup_or_add( + state, &tmp_next_block_ctx, + next_block, superblock_tmp, + BTRFSIC_GENERATION_UNKNOWN); + btrfsic_release_block_ctx(&tmp_next_block_ctx); + if (NULL == l) { + brelse(bh); + return -1; + } + } + } + if (state->print_mask & BTRFSIC_PRINT_MASK_INITIAL_ALL_TREES) + btrfsic_dump_tree_sub(state, superblock_tmp, 0); + + brelse(bh); + return 0; +} + +static struct btrfsic_stack_frame *btrfsic_stack_frame_alloc(void) +{ + struct btrfsic_stack_frame *sf; + + sf = kzalloc(sizeof(*sf), GFP_NOFS); + if (NULL == sf) + printk(KERN_INFO "btrfsic: alloc memory failed!\n"); + else + sf->magic = BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER; + return sf; +} + +static void btrfsic_stack_frame_free(struct btrfsic_stack_frame *sf) +{ + BUG_ON(!(NULL == sf || + BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER == sf->magic)); + kfree(sf); +} + +static int btrfsic_process_metablock( + struct btrfsic_state *state, + struct btrfsic_block *const first_block, + struct btrfsic_block_data_ctx *const first_block_ctx, + struct btrfs_header *const first_hdr, + int first_limit_nesting, int force_iodone_flag) +{ + struct btrfsic_stack_frame initial_stack_frame = { 0 }; + struct btrfsic_stack_frame *sf; + struct btrfsic_stack_frame *next_stack; + + sf = &initial_stack_frame; + sf->error = 0; + sf->i = -1; + sf->limit_nesting = first_limit_nesting; + sf->block = first_block; + sf->block_ctx = first_block_ctx; + sf->next_block = NULL; + sf->hdr = first_hdr; + sf->prev = NULL; + +continue_with_new_stack_frame: + sf->block->generation = le64_to_cpu(sf->hdr->generation); + if (0 == sf->hdr->level) { + struct btrfs_leaf *const leafhdr = + (struct btrfs_leaf *)sf->hdr; + + if (-1 == sf->i) { + sf->nr = le32_to_cpu(leafhdr->header.nritems); + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "leaf %llu items %d generation %llu" + " owner %llu\n", + (unsigned long long) + sf->block_ctx->start, + sf->nr, + (unsigned long long) + le64_to_cpu(leafhdr->header.generation), + (unsigned long long) + le64_to_cpu(leafhdr->header.owner)); + } + +continue_with_current_leaf_stack_frame: + if (0 == sf->num_copies || sf->mirror_num > sf->num_copies) { + sf->i++; + sf->num_copies = 0; + } + + if (sf->i < sf->nr) { + struct btrfs_item *disk_item = leafhdr->items + sf->i; + struct btrfs_disk_key *disk_key = &disk_item->key; + u8 type; + const u32 item_offset = le32_to_cpu(disk_item->offset); + + type = disk_key->type; + + if (BTRFS_ROOT_ITEM_KEY == type) { + const struct btrfs_root_item *const root_item = + (struct btrfs_root_item *) + (sf->block_ctx->data + + offsetof(struct btrfs_leaf, items) + + item_offset); + const u64 next_bytenr = + le64_to_cpu(root_item->bytenr); + + sf->error = + btrfsic_create_link_to_next_block( + state, + sf->block, + sf->block_ctx, + next_bytenr, + sf->limit_nesting, + &sf->next_block_ctx, + &sf->next_block, + force_iodone_flag, + &sf->num_copies, + &sf->mirror_num, + disk_key, + le64_to_cpu(root_item-> + generation)); + if (sf->error) + goto one_stack_frame_backwards; + + if (NULL != sf->next_block) { + struct btrfs_header *const next_hdr = + (struct btrfs_header *) + sf->next_block_ctx.data; + + next_stack = + btrfsic_stack_frame_alloc(); + if (NULL == next_stack) { + btrfsic_release_block_ctx( + &sf-> + next_block_ctx); + goto one_stack_frame_backwards; + } + + next_stack->i = -1; + next_stack->block = sf->next_block; + next_stack->block_ctx = + &sf->next_block_ctx; + next_stack->next_block = NULL; + next_stack->hdr = next_hdr; + next_stack->limit_nesting = + sf->limit_nesting - 1; + next_stack->prev = sf; + sf = next_stack; + goto continue_with_new_stack_frame; + } + } else if (BTRFS_EXTENT_DATA_KEY == type && + state->include_extent_data) { + sf->error = btrfsic_handle_extent_data( + state, + sf->block, + sf->block_ctx, + item_offset, + force_iodone_flag); + if (sf->error) + goto one_stack_frame_backwards; + } + + goto continue_with_current_leaf_stack_frame; + } + } else { + struct btrfs_node *const nodehdr = (struct btrfs_node *)sf->hdr; + + if (-1 == sf->i) { + sf->nr = le32_to_cpu(nodehdr->header.nritems); + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO "node %llu level %d items %d" + " generation %llu owner %llu\n", + (unsigned long long) + sf->block_ctx->start, + nodehdr->header.level, sf->nr, + (unsigned long long) + le64_to_cpu(nodehdr->header.generation), + (unsigned long long) + le64_to_cpu(nodehdr->header.owner)); + } + +continue_with_current_node_stack_frame: + if (0 == sf->num_copies || sf->mirror_num > sf->num_copies) { + sf->i++; + sf->num_copies = 0; + } + + if (sf->i < sf->nr) { + struct btrfs_key_ptr *disk_key_ptr = + nodehdr->ptrs + sf->i; + const u64 next_bytenr = + le64_to_cpu(disk_key_ptr->blockptr); + + sf->error = btrfsic_create_link_to_next_block( + state, + sf->block, + sf->block_ctx, + next_bytenr, + sf->limit_nesting, + &sf->next_block_ctx, + &sf->next_block, + force_iodone_flag, + &sf->num_copies, + &sf->mirror_num, + &disk_key_ptr->key, + le64_to_cpu(disk_key_ptr->generation)); + if (sf->error) + goto one_stack_frame_backwards; + + if (NULL != sf->next_block) { + struct btrfs_header *const next_hdr = + (struct btrfs_header *) + sf->next_block_ctx.data; + + next_stack = btrfsic_stack_frame_alloc(); + if (NULL == next_stack) + goto one_stack_frame_backwards; + + next_stack->i = -1; + next_stack->block = sf->next_block; + next_stack->block_ctx = &sf->next_block_ctx; + next_stack->next_block = NULL; + next_stack->hdr = next_hdr; + next_stack->limit_nesting = + sf->limit_nesting - 1; + next_stack->prev = sf; + sf = next_stack; + goto continue_with_new_stack_frame; + } + + goto continue_with_current_node_stack_frame; + } + } + +one_stack_frame_backwards: + if (NULL != sf->prev) { + struct btrfsic_stack_frame *const prev = sf->prev; + + /* the one for the initial block is freed in the caller */ + btrfsic_release_block_ctx(sf->block_ctx); + + if (sf->error) { + prev->error = sf->error; + btrfsic_stack_frame_free(sf); + sf = prev; + goto one_stack_frame_backwards; + } + + btrfsic_stack_frame_free(sf); + sf = prev; + goto continue_with_new_stack_frame; + } else { + BUG_ON(&initial_stack_frame != sf); + } + + return sf->error; +} + +static int btrfsic_create_link_to_next_block( + struct btrfsic_state *state, + struct btrfsic_block *block, + struct btrfsic_block_data_ctx *block_ctx, + u64 next_bytenr, + int limit_nesting, + struct btrfsic_block_data_ctx *next_block_ctx, + struct btrfsic_block **next_blockp, + int force_iodone_flag, + int *num_copiesp, int *mirror_nump, + struct btrfs_disk_key *disk_key, + u64 parent_generation) +{ + struct btrfsic_block *next_block = NULL; + int ret; + struct btrfsic_block_link *l; + int did_alloc_block_link; + int block_was_created; + + *next_blockp = NULL; + if (0 == *num_copiesp) { + *num_copiesp = + btrfs_num_copies(&state->root->fs_info->mapping_tree, + next_bytenr, PAGE_SIZE); + if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES) + printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n", + (unsigned long long)next_bytenr, *num_copiesp); + *mirror_nump = 1; + } + + if (*mirror_nump > *num_copiesp) + return 0; + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "btrfsic_create_link_to_next_block(mirror_num=%d)\n", + *mirror_nump); + ret = btrfsic_map_block(state, next_bytenr, + BTRFSIC_BLOCK_SIZE, + next_block_ctx, *mirror_nump); + if (ret) { + printk(KERN_INFO + "btrfsic: btrfsic_map_block(@%llu, mirror=%d) failed!\n", + (unsigned long long)next_bytenr, *mirror_nump); + btrfsic_release_block_ctx(next_block_ctx); + *next_blockp = NULL; + return -1; + } + + next_block = btrfsic_block_lookup_or_add(state, + next_block_ctx, "referenced ", + 1, force_iodone_flag, + !force_iodone_flag, + *mirror_nump, + &block_was_created); + if (NULL == next_block) { + btrfsic_release_block_ctx(next_block_ctx); + *next_blockp = NULL; + return -1; + } + if (block_was_created) { + l = NULL; + next_block->generation = BTRFSIC_GENERATION_UNKNOWN; + } else { + if (next_block->logical_bytenr != next_bytenr && + !(!next_block->is_metadata && + 0 == next_block->logical_bytenr)) { + printk(KERN_INFO + "Referenced block @%llu (%s/%llu/%d)" + " found in hash table, %c," + " bytenr mismatch (!= stored %llu).\n", + (unsigned long long)next_bytenr, + next_block_ctx->dev->name, + (unsigned long long)next_block_ctx->dev_bytenr, + *mirror_nump, + btrfsic_get_block_type(state, next_block), + (unsigned long long)next_block->logical_bytenr); + } else if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "Referenced block @%llu (%s/%llu/%d)" + " found in hash table, %c.\n", + (unsigned long long)next_bytenr, + next_block_ctx->dev->name, + (unsigned long long)next_block_ctx->dev_bytenr, + *mirror_nump, + btrfsic_get_block_type(state, next_block)); + next_block->logical_bytenr = next_bytenr; + + next_block->mirror_num = *mirror_nump; + l = btrfsic_block_link_hashtable_lookup( + next_block_ctx->dev->bdev, + next_block_ctx->dev_bytenr, + block_ctx->dev->bdev, + block_ctx->dev_bytenr, + &state->block_link_hashtable); + } + + next_block->disk_key = *disk_key; + if (NULL == l) { + l = btrfsic_block_link_alloc(); + if (NULL == l) { + printk(KERN_INFO "btrfsic: error, kmalloc failed!\n"); + btrfsic_release_block_ctx(next_block_ctx); + *next_blockp = NULL; + return -1; + } + + did_alloc_block_link = 1; + l->block_ref_to = next_block; + l->block_ref_from = block; + l->ref_cnt = 1; + l->parent_generation = parent_generation; + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + btrfsic_print_add_link(state, l); + + list_add(&l->node_ref_to, &block->ref_to_list); + list_add(&l->node_ref_from, &next_block->ref_from_list); + + btrfsic_block_link_hashtable_add(l, + &state->block_link_hashtable); + } else { + did_alloc_block_link = 0; + if (0 == limit_nesting) { + l->ref_cnt++; + l->parent_generation = parent_generation; + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + btrfsic_print_add_link(state, l); + } + } + + if (limit_nesting > 0 && did_alloc_block_link) { + ret = btrfsic_read_block(state, next_block_ctx); + if (ret < (int)BTRFSIC_BLOCK_SIZE) { + printk(KERN_INFO + "btrfsic: read block @logical %llu failed!\n", + (unsigned long long)next_bytenr); + btrfsic_release_block_ctx(next_block_ctx); + *next_blockp = NULL; + return -1; + } + + *next_blockp = next_block; + } else { + *next_blockp = NULL; + } + (*mirror_nump)++; + + return 0; +} + +static int btrfsic_handle_extent_data( + struct btrfsic_state *state, + struct btrfsic_block *block, + struct btrfsic_block_data_ctx *block_ctx, + u32 item_offset, int force_iodone_flag) +{ + int ret; + struct btrfs_file_extent_item *file_extent_item = + (struct btrfs_file_extent_item *)(block_ctx->data + + offsetof(struct btrfs_leaf, + items) + item_offset); + u64 next_bytenr = + le64_to_cpu(file_extent_item->disk_bytenr) + + le64_to_cpu(file_extent_item->offset); + u64 num_bytes = le64_to_cpu(file_extent_item->num_bytes); + u64 generation = le64_to_cpu(file_extent_item->generation); + struct btrfsic_block_link *l; + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE) + printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu," + " offset = %llu, num_bytes = %llu\n", + file_extent_item->type, + (unsigned long long) + le64_to_cpu(file_extent_item->disk_bytenr), + (unsigned long long) + le64_to_cpu(file_extent_item->offset), + (unsigned long long) + le64_to_cpu(file_extent_item->num_bytes)); + if (BTRFS_FILE_EXTENT_REG != file_extent_item->type || + ((u64)0) == le64_to_cpu(file_extent_item->disk_bytenr)) + return 0; + while (num_bytes > 0) { + u32 chunk_len; + int num_copies; + int mirror_num; + + if (num_bytes > BTRFSIC_BLOCK_SIZE) + chunk_len = BTRFSIC_BLOCK_SIZE; + else + chunk_len = num_bytes; + + num_copies = + btrfs_num_copies(&state->root->fs_info->mapping_tree, + next_bytenr, PAGE_SIZE); + if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES) + printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n", + (unsigned long long)next_bytenr, num_copies); + for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) { + struct btrfsic_block_data_ctx next_block_ctx; + struct btrfsic_block *next_block; + int block_was_created; + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO "btrfsic_handle_extent_data(" + "mirror_num=%d)\n", mirror_num); + if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE) + printk(KERN_INFO + "\tdisk_bytenr = %llu, num_bytes %u\n", + (unsigned long long)next_bytenr, + chunk_len); + ret = btrfsic_map_block(state, next_bytenr, + chunk_len, &next_block_ctx, + mirror_num); + if (ret) { + printk(KERN_INFO + "btrfsic: btrfsic_map_block(@%llu," + " mirror=%d) failed!\n", + (unsigned long long)next_bytenr, + mirror_num); + return -1; + } + + next_block = btrfsic_block_lookup_or_add( + state, + &next_block_ctx, + "referenced ", + 0, + force_iodone_flag, + !force_iodone_flag, + mirror_num, + &block_was_created); + if (NULL == next_block) { + printk(KERN_INFO + "btrfsic: error, kmalloc failed!\n"); + btrfsic_release_block_ctx(&next_block_ctx); + return -1; + } + if (!block_was_created) { + if (next_block->logical_bytenr != next_bytenr && + !(!next_block->is_metadata && + 0 == next_block->logical_bytenr)) { + printk(KERN_INFO + "Referenced block" + " @%llu (%s/%llu/%d)" + " found in hash table, D," + " bytenr mismatch" + " (!= stored %llu).\n", + (unsigned long long)next_bytenr, + next_block_ctx.dev->name, + (unsigned long long) + next_block_ctx.dev_bytenr, + mirror_num, + (unsigned long long) + next_block->logical_bytenr); + } + next_block->logical_bytenr = next_bytenr; + next_block->mirror_num = mirror_num; + } + + l = btrfsic_block_link_lookup_or_add(state, + &next_block_ctx, + next_block, block, + generation); + btrfsic_release_block_ctx(&next_block_ctx); + if (NULL == l) + return -1; + } + + next_bytenr += chunk_len; + num_bytes -= chunk_len; + } + + return 0; +} + +static int btrfsic_map_block(struct btrfsic_state *state, u64 bytenr, u32 len, + struct btrfsic_block_data_ctx *block_ctx_out, + int mirror_num) +{ + int ret; + u64 length; + struct btrfs_bio *multi = NULL; + struct btrfs_device *device; + + length = len; + ret = btrfs_map_block(&state->root->fs_info->mapping_tree, READ, + bytenr, &length, &multi, mirror_num); + + device = multi->stripes[0].dev; + block_ctx_out->dev = btrfsic_dev_state_lookup(device->bdev); + block_ctx_out->dev_bytenr = multi->stripes[0].physical; + block_ctx_out->start = bytenr; + block_ctx_out->len = len; + block_ctx_out->data = NULL; + block_ctx_out->bh = NULL; + + if (0 == ret) + kfree(multi); + if (NULL == block_ctx_out->dev) { + ret = -ENXIO; + printk(KERN_INFO "btrfsic: error, cannot lookup dev (#1)!\n"); + } + + return ret; +} + +static int btrfsic_map_superblock(struct btrfsic_state *state, u64 bytenr, + u32 len, struct block_device *bdev, + struct btrfsic_block_data_ctx *block_ctx_out) +{ + block_ctx_out->dev = btrfsic_dev_state_lookup(bdev); + block_ctx_out->dev_bytenr = bytenr; + block_ctx_out->start = bytenr; + block_ctx_out->len = len; + block_ctx_out->data = NULL; + block_ctx_out->bh = NULL; + if (NULL != block_ctx_out->dev) { + return 0; + } else { + printk(KERN_INFO "btrfsic: error, cannot lookup dev (#2)!\n"); + return -ENXIO; + } +} + +static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx) +{ + if (NULL != block_ctx->bh) { + brelse(block_ctx->bh); + block_ctx->bh = NULL; + } +} + +static int btrfsic_read_block(struct btrfsic_state *state, + struct btrfsic_block_data_ctx *block_ctx) +{ + block_ctx->bh = NULL; + if (block_ctx->dev_bytenr & 4095) { + printk(KERN_INFO + "btrfsic: read_block() with unaligned bytenr %llu\n", + (unsigned long long)block_ctx->dev_bytenr); + return -1; + } + if (block_ctx->len > 4096) { + printk(KERN_INFO + "btrfsic: read_block() with too huge size %d\n", + block_ctx->len); + return -1; + } + + block_ctx->bh = __bread(block_ctx->dev->bdev, + block_ctx->dev_bytenr >> 12, 4096); + if (NULL == block_ctx->bh) + return -1; + block_ctx->data = block_ctx->bh->b_data; + + return block_ctx->len; +} + +static void btrfsic_dump_database(struct btrfsic_state *state) +{ + struct list_head *elem_all; + + BUG_ON(NULL == state); + + printk(KERN_INFO "all_blocks_list:\n"); + list_for_each(elem_all, &state->all_blocks_list) { + const struct btrfsic_block *const b_all = + list_entry(elem_all, struct btrfsic_block, + all_blocks_node); + struct list_head *elem_ref_to; + struct list_head *elem_ref_from; + + printk(KERN_INFO "%c-block @%llu (%s/%llu/%d)\n", + btrfsic_get_block_type(state, b_all), + (unsigned long long)b_all->logical_bytenr, + b_all->dev_state->name, + (unsigned long long)b_all->dev_bytenr, + b_all->mirror_num); + + list_for_each(elem_ref_to, &b_all->ref_to_list) { + const struct btrfsic_block_link *const l = + list_entry(elem_ref_to, + struct btrfsic_block_link, + node_ref_to); + + printk(KERN_INFO " %c @%llu (%s/%llu/%d)" + " refers %u* to" + " %c @%llu (%s/%llu/%d)\n", + btrfsic_get_block_type(state, b_all), + (unsigned long long)b_all->logical_bytenr, + b_all->dev_state->name, + (unsigned long long)b_all->dev_bytenr, + b_all->mirror_num, + l->ref_cnt, + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long) + l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num); + } + + list_for_each(elem_ref_from, &b_all->ref_from_list) { + const struct btrfsic_block_link *const l = + list_entry(elem_ref_from, + struct btrfsic_block_link, + node_ref_from); + + printk(KERN_INFO " %c @%llu (%s/%llu/%d)" + " is ref %u* from" + " %c @%llu (%s/%llu/%d)\n", + btrfsic_get_block_type(state, b_all), + (unsigned long long)b_all->logical_bytenr, + b_all->dev_state->name, + (unsigned long long)b_all->dev_bytenr, + b_all->mirror_num, + l->ref_cnt, + btrfsic_get_block_type(state, l->block_ref_from), + (unsigned long long) + l->block_ref_from->logical_bytenr, + l->block_ref_from->dev_state->name, + (unsigned long long) + l->block_ref_from->dev_bytenr, + l->block_ref_from->mirror_num); + } + + printk(KERN_INFO "\n"); + } +} + +/* + * Test whether the disk block contains a tree block (leaf or node) + * (note that this test fails for the super block) + */ +static int btrfsic_test_for_metadata(struct btrfsic_state *state, + const u8 *data, unsigned int size) +{ + struct btrfs_header *h; + u8 csum[BTRFS_CSUM_SIZE]; + u32 crc = ~(u32)0; + int fail = 0; + int crc_fail = 0; + + h = (struct btrfs_header *)data; + + if (memcmp(h->fsid, state->root->fs_info->fsid, BTRFS_UUID_SIZE)) + fail++; + + crc = crc32c(crc, data + BTRFS_CSUM_SIZE, PAGE_SIZE - BTRFS_CSUM_SIZE); + btrfs_csum_final(crc, csum); + if (memcmp(csum, h->csum, state->csum_size)) + crc_fail++; + + return fail || crc_fail; +} + +static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state, + u64 dev_bytenr, + u8 *mapped_data, unsigned int len, + struct bio *bio, + int *bio_is_patched, + struct buffer_head *bh, + int submit_bio_bh_rw) +{ + int is_metadata; + struct btrfsic_block *block; + struct btrfsic_block_data_ctx block_ctx; + int ret; + struct btrfsic_state *state = dev_state->state; + struct block_device *bdev = dev_state->bdev; + + WARN_ON(len > PAGE_SIZE); + is_metadata = (0 == btrfsic_test_for_metadata(state, mapped_data, len)); + if (NULL != bio_is_patched) + *bio_is_patched = 0; + + block = btrfsic_block_hashtable_lookup(bdev, dev_bytenr, + &state->block_hashtable); + if (NULL != block) { + u64 bytenr; + struct list_head *elem_ref_to; + struct list_head *tmp_ref_to; + + if (block->is_superblock) { + bytenr = le64_to_cpu(((struct btrfs_super_block *) + mapped_data)->bytenr); + is_metadata = 1; + if (state->print_mask & + BTRFSIC_PRINT_MASK_TREE_BEFORE_SB_WRITE) { + printk(KERN_INFO + "[before new superblock is written]:\n"); + btrfsic_dump_tree_sub(state, block, 0); + } + } + if (is_metadata) { + if (!block->is_superblock) { + bytenr = le64_to_cpu(((struct btrfs_header *) + mapped_data)->bytenr); + btrfsic_cmp_log_and_dev_bytenr(state, bytenr, + dev_state, + dev_bytenr, + mapped_data); + } + if (block->logical_bytenr != bytenr) { + printk(KERN_INFO + "Written block @%llu (%s/%llu/%d)" + " found in hash table, %c," + " bytenr mismatch" + " (!= stored %llu).\n", + (unsigned long long)bytenr, + dev_state->name, + (unsigned long long)dev_bytenr, + block->mirror_num, + btrfsic_get_block_type(state, block), + (unsigned long long) + block->logical_bytenr); + block->logical_bytenr = bytenr; + } else if (state->print_mask & + BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "Written block @%llu (%s/%llu/%d)" + " found in hash table, %c.\n", + (unsigned long long)bytenr, + dev_state->name, + (unsigned long long)dev_bytenr, + block->mirror_num, + btrfsic_get_block_type(state, block)); + } else { + bytenr = block->logical_bytenr; + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "Written block @%llu (%s/%llu/%d)" + " found in hash table, %c.\n", + (unsigned long long)bytenr, + dev_state->name, + (unsigned long long)dev_bytenr, + block->mirror_num, + btrfsic_get_block_type(state, block)); + } + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "ref_to_list: %cE, ref_from_list: %cE\n", + list_empty(&block->ref_to_list) ? ' ' : '!', + list_empty(&block->ref_from_list) ? ' ' : '!'); + if (btrfsic_is_block_ref_by_superblock(state, block, 0)) { + printk(KERN_INFO "btrfs: attempt to overwrite %c-block" + " @%llu (%s/%llu/%d), old(gen=%llu," + " objectid=%llu, type=%d, offset=%llu)," + " new(gen=%llu)," + " which is referenced by most recent superblock" + " (superblockgen=%llu)!\n", + btrfsic_get_block_type(state, block), + (unsigned long long)bytenr, + dev_state->name, + (unsigned long long)dev_bytenr, + block->mirror_num, + (unsigned long long)block->generation, + (unsigned long long) + le64_to_cpu(block->disk_key.objectid), + block->disk_key.type, + (unsigned long long) + le64_to_cpu(block->disk_key.offset), + (unsigned long long) + le64_to_cpu(((struct btrfs_header *) + mapped_data)->generation), + (unsigned long long) + state->max_superblock_generation); + btrfsic_dump_tree(state); + } + + if (!block->is_iodone && !block->never_written) { + printk(KERN_INFO "btrfs: attempt to overwrite %c-block" + " @%llu (%s/%llu/%d), oldgen=%llu, newgen=%llu," + " which is not yet iodone!\n", + btrfsic_get_block_type(state, block), + (unsigned long long)bytenr, + dev_state->name, + (unsigned long long)dev_bytenr, + block->mirror_num, + (unsigned long long)block->generation, + (unsigned long long) + le64_to_cpu(((struct btrfs_header *) + mapped_data)->generation)); + /* it would not be safe to go on */ + btrfsic_dump_tree(state); + return; + } + + /* + * Clear all references of this block. Do not free + * the block itself even if is not referenced anymore + * because it still carries valueable information + * like whether it was ever written and IO completed. + */ + list_for_each_safe(elem_ref_to, tmp_ref_to, + &block->ref_to_list) { + struct btrfsic_block_link *const l = + list_entry(elem_ref_to, + struct btrfsic_block_link, + node_ref_to); + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + btrfsic_print_rem_link(state, l); + l->ref_cnt--; + if (0 == l->ref_cnt) { + list_del(&l->node_ref_to); + list_del(&l->node_ref_from); + btrfsic_block_link_hashtable_remove(l); + btrfsic_block_link_free(l); + } + } + + if (block->is_superblock) + ret = btrfsic_map_superblock(state, bytenr, len, + bdev, &block_ctx); + else + ret = btrfsic_map_block(state, bytenr, len, + &block_ctx, 0); + if (ret) { + printk(KERN_INFO + "btrfsic: btrfsic_map_block(root @%llu)" + " failed!\n", (unsigned long long)bytenr); + return; + } + block_ctx.data = mapped_data; + /* the following is required in case of writes to mirrors, + * use the same that was used for the lookup */ + block_ctx.dev = dev_state; + block_ctx.dev_bytenr = dev_bytenr; + + if (is_metadata || state->include_extent_data) { + block->never_written = 0; + block->iodone_w_error = 0; + if (NULL != bio) { + block->is_iodone = 0; + BUG_ON(NULL == bio_is_patched); + if (!*bio_is_patched) { + block->orig_bio_bh_private = + bio->bi_private; + block->orig_bio_bh_end_io.bio = + bio->bi_end_io; + block->next_in_same_bio = NULL; + bio->bi_private = block; + bio->bi_end_io = btrfsic_bio_end_io; + *bio_is_patched = 1; + } else { + struct btrfsic_block *chained_block = + (struct btrfsic_block *) + bio->bi_private; + + BUG_ON(NULL == chained_block); + block->orig_bio_bh_private = + chained_block->orig_bio_bh_private; + block->orig_bio_bh_end_io.bio = + chained_block->orig_bio_bh_end_io. + bio; + block->next_in_same_bio = chained_block; + bio->bi_private = block; + } + } else if (NULL != bh) { + block->is_iodone = 0; + block->orig_bio_bh_private = bh->b_private; + block->orig_bio_bh_end_io.bh = bh->b_end_io; + block->next_in_same_bio = NULL; + bh->b_private = block; + bh->b_end_io = btrfsic_bh_end_io; + } else { + block->is_iodone = 1; + block->orig_bio_bh_private = NULL; + block->orig_bio_bh_end_io.bio = NULL; + block->next_in_same_bio = NULL; + } + } + + block->flush_gen = dev_state->last_flush_gen + 1; + block->submit_bio_bh_rw = submit_bio_bh_rw; + if (is_metadata) { + block->logical_bytenr = bytenr; + block->is_metadata = 1; + if (block->is_superblock) { + ret = btrfsic_process_written_superblock( + state, + block, + (struct btrfs_super_block *) + mapped_data); + if (state->print_mask & + BTRFSIC_PRINT_MASK_TREE_AFTER_SB_WRITE) { + printk(KERN_INFO + "[after new superblock is written]:\n"); + btrfsic_dump_tree_sub(state, block, 0); + } + } else { + block->mirror_num = 0; /* unknown */ + ret = btrfsic_process_metablock( + state, + block, + &block_ctx, + (struct btrfs_header *) + block_ctx.data, + 0, 0); + } + if (ret) + printk(KERN_INFO + "btrfsic: btrfsic_process_metablock" + "(root @%llu) failed!\n", + (unsigned long long)dev_bytenr); + } else { + block->is_metadata = 0; + block->mirror_num = 0; /* unknown */ + block->generation = BTRFSIC_GENERATION_UNKNOWN; + if (!state->include_extent_data + && list_empty(&block->ref_from_list)) { + /* + * disk block is overwritten with extent + * data (not meta data) and we are configured + * to not include extent data: take the + * chance and free the block's memory + */ + btrfsic_block_hashtable_remove(block); + list_del(&block->all_blocks_node); + btrfsic_block_free(block); + } + } + btrfsic_release_block_ctx(&block_ctx); + } else { + /* block has not been found in hash table */ + u64 bytenr; + + if (!is_metadata) { + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO "Written block (%s/%llu/?)" + " !found in hash table, D.\n", + dev_state->name, + (unsigned long long)dev_bytenr); + if (!state->include_extent_data) + return; /* ignore that written D block */ + + /* this is getting ugly for the + * include_extent_data case... */ + bytenr = 0; /* unknown */ + block_ctx.start = bytenr; + block_ctx.len = len; + block_ctx.bh = NULL; + } else { + bytenr = le64_to_cpu(((struct btrfs_header *) + mapped_data)->bytenr); + btrfsic_cmp_log_and_dev_bytenr(state, bytenr, dev_state, + dev_bytenr, + mapped_data); + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "Written block @%llu (%s/%llu/?)" + " !found in hash table, M.\n", + (unsigned long long)bytenr, + dev_state->name, + (unsigned long long)dev_bytenr); + + ret = btrfsic_map_block(state, bytenr, len, &block_ctx, + 0); + if (ret) { + printk(KERN_INFO + "btrfsic: btrfsic_map_block(root @%llu)" + " failed!\n", + (unsigned long long)dev_bytenr); + return; + } + } + block_ctx.data = mapped_data; + /* the following is required in case of writes to mirrors, + * use the same that was used for the lookup */ + block_ctx.dev = dev_state; + block_ctx.dev_bytenr = dev_bytenr; + + block = btrfsic_block_alloc(); + if (NULL == block) { + printk(KERN_INFO "btrfsic: error, kmalloc failed!\n"); + btrfsic_release_block_ctx(&block_ctx); + return; + } + block->dev_state = dev_state; + block->dev_bytenr = dev_bytenr; + block->logical_bytenr = bytenr; + block->is_metadata = is_metadata; + block->never_written = 0; + block->iodone_w_error = 0; + block->mirror_num = 0; /* unknown */ + block->flush_gen = dev_state->last_flush_gen + 1; + block->submit_bio_bh_rw = submit_bio_bh_rw; + if (NULL != bio) { + block->is_iodone = 0; + BUG_ON(NULL == bio_is_patched); + if (!*bio_is_patched) { + block->orig_bio_bh_private = bio->bi_private; + block->orig_bio_bh_end_io.bio = bio->bi_end_io; + block->next_in_same_bio = NULL; + bio->bi_private = block; + bio->bi_end_io = btrfsic_bio_end_io; + *bio_is_patched = 1; + } else { + struct btrfsic_block *chained_block = + (struct btrfsic_block *) + bio->bi_private; + + BUG_ON(NULL == chained_block); + block->orig_bio_bh_private = + chained_block->orig_bio_bh_private; + block->orig_bio_bh_end_io.bio = + chained_block->orig_bio_bh_end_io.bio; + block->next_in_same_bio = chained_block; + bio->bi_private = block; + } + } else if (NULL != bh) { + block->is_iodone = 0; + block->orig_bio_bh_private = bh->b_private; + block->orig_bio_bh_end_io.bh = bh->b_end_io; + block->next_in_same_bio = NULL; + bh->b_private = block; + bh->b_end_io = btrfsic_bh_end_io; + } else { + block->is_iodone = 1; + block->orig_bio_bh_private = NULL; + block->orig_bio_bh_end_io.bio = NULL; + block->next_in_same_bio = NULL; + } + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "New written %c-block @%llu (%s/%llu/%d)\n", + is_metadata ? 'M' : 'D', + (unsigned long long)block->logical_bytenr, + block->dev_state->name, + (unsigned long long)block->dev_bytenr, + block->mirror_num); + list_add(&block->all_blocks_node, &state->all_blocks_list); + btrfsic_block_hashtable_add(block, &state->block_hashtable); + + if (is_metadata) { + ret = btrfsic_process_metablock(state, block, + &block_ctx, + (struct btrfs_header *) + block_ctx.data, 0, 0); + if (ret) + printk(KERN_INFO + "btrfsic: process_metablock(root @%llu)" + " failed!\n", + (unsigned long long)dev_bytenr); + } + btrfsic_release_block_ctx(&block_ctx); + } +} + +static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status) +{ + struct btrfsic_block *block = (struct btrfsic_block *)bp->bi_private; + int iodone_w_error; + + /* mutex is not held! This is not save if IO is not yet completed + * on umount */ + iodone_w_error = 0; + if (bio_error_status) + iodone_w_error = 1; + + BUG_ON(NULL == block); + bp->bi_private = block->orig_bio_bh_private; + bp->bi_end_io = block->orig_bio_bh_end_io.bio; + + do { + struct btrfsic_block *next_block; + struct btrfsic_dev_state *const dev_state = block->dev_state; + + if ((dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_END_IO_BIO_BH)) + printk(KERN_INFO + "bio_end_io(err=%d) for %c @%llu (%s/%llu/%d)\n", + bio_error_status, + btrfsic_get_block_type(dev_state->state, block), + (unsigned long long)block->logical_bytenr, + dev_state->name, + (unsigned long long)block->dev_bytenr, + block->mirror_num); + next_block = block->next_in_same_bio; + block->iodone_w_error = iodone_w_error; + if (block->submit_bio_bh_rw & REQ_FLUSH) { + dev_state->last_flush_gen++; + if ((dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_END_IO_BIO_BH)) + printk(KERN_INFO + "bio_end_io() new %s flush_gen=%llu\n", + dev_state->name, + (unsigned long long) + dev_state->last_flush_gen); + } + if (block->submit_bio_bh_rw & REQ_FUA) + block->flush_gen = 0; /* FUA completed means block is + * on disk */ + block->is_iodone = 1; /* for FLUSH, this releases the block */ + block = next_block; + } while (NULL != block); + + bp->bi_end_io(bp, bio_error_status); +} + +static void btrfsic_bh_end_io(struct buffer_head *bh, int uptodate) +{ + struct btrfsic_block *block = (struct btrfsic_block *)bh->b_private; + int iodone_w_error = !uptodate; + struct btrfsic_dev_state *dev_state; + + BUG_ON(NULL == block); + dev_state = block->dev_state; + if ((dev_state->state->print_mask & BTRFSIC_PRINT_MASK_END_IO_BIO_BH)) + printk(KERN_INFO + "bh_end_io(error=%d) for %c @%llu (%s/%llu/%d)\n", + iodone_w_error, + btrfsic_get_block_type(dev_state->state, block), + (unsigned long long)block->logical_bytenr, + block->dev_state->name, + (unsigned long long)block->dev_bytenr, + block->mirror_num); + + block->iodone_w_error = iodone_w_error; + if (block->submit_bio_bh_rw & REQ_FLUSH) { + dev_state->last_flush_gen++; + if ((dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_END_IO_BIO_BH)) + printk(KERN_INFO + "bh_end_io() new %s flush_gen=%llu\n", + dev_state->name, + (unsigned long long)dev_state->last_flush_gen); + } + if (block->submit_bio_bh_rw & REQ_FUA) + block->flush_gen = 0; /* FUA completed means block is on disk */ + + bh->b_private = block->orig_bio_bh_private; + bh->b_end_io = block->orig_bio_bh_end_io.bh; + block->is_iodone = 1; /* for FLUSH, this releases the block */ + bh->b_end_io(bh, uptodate); +} + +static int btrfsic_process_written_superblock( + struct btrfsic_state *state, + struct btrfsic_block *const superblock, + struct btrfs_super_block *const super_hdr) +{ + int pass; + + superblock->generation = btrfs_super_generation(super_hdr); + if (!(superblock->generation > state->max_superblock_generation || + 0 == state->max_superblock_generation)) { + if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE) + printk(KERN_INFO + "btrfsic: superblock @%llu (%s/%llu/%d)" + " with old gen %llu <= %llu\n", + (unsigned long long)superblock->logical_bytenr, + superblock->dev_state->name, + (unsigned long long)superblock->dev_bytenr, + superblock->mirror_num, + (unsigned long long) + btrfs_super_generation(super_hdr), + (unsigned long long) + state->max_superblock_generation); + } else { + if (state->print_mask & BTRFSIC_PRINT_MASK_SUPERBLOCK_WRITE) + printk(KERN_INFO + "btrfsic: got new superblock @%llu (%s/%llu/%d)" + " with new gen %llu > %llu\n", + (unsigned long long)superblock->logical_bytenr, + superblock->dev_state->name, + (unsigned long long)superblock->dev_bytenr, + superblock->mirror_num, + (unsigned long long) + btrfs_super_generation(super_hdr), + (unsigned long long) + state->max_superblock_generation); + + state->max_superblock_generation = + btrfs_super_generation(super_hdr); + state->latest_superblock = superblock; + } + + for (pass = 0; pass < 3; pass++) { + int ret; + u64 next_bytenr; + struct btrfsic_block *next_block; + struct btrfsic_block_data_ctx tmp_next_block_ctx; + struct btrfsic_block_link *l; + int num_copies; + int mirror_num; + const char *additional_string = NULL; + struct btrfs_disk_key tmp_disk_key; + + tmp_disk_key.type = BTRFS_ROOT_ITEM_KEY; + tmp_disk_key.offset = 0; + + switch (pass) { + case 0: + tmp_disk_key.objectid = + cpu_to_le64(BTRFS_ROOT_TREE_OBJECTID); + additional_string = "root "; + next_bytenr = btrfs_super_root(super_hdr); + if (state->print_mask & + BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION) + printk(KERN_INFO "root@%llu\n", + (unsigned long long)next_bytenr); + break; + case 1: + tmp_disk_key.objectid = + cpu_to_le64(BTRFS_CHUNK_TREE_OBJECTID); + additional_string = "chunk "; + next_bytenr = btrfs_super_chunk_root(super_hdr); + if (state->print_mask & + BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION) + printk(KERN_INFO "chunk@%llu\n", + (unsigned long long)next_bytenr); + break; + case 2: + tmp_disk_key.objectid = + cpu_to_le64(BTRFS_TREE_LOG_OBJECTID); + additional_string = "log "; + next_bytenr = btrfs_super_log_root(super_hdr); + if (0 == next_bytenr) + continue; + if (state->print_mask & + BTRFSIC_PRINT_MASK_ROOT_CHUNK_LOG_TREE_LOCATION) + printk(KERN_INFO "log@%llu\n", + (unsigned long long)next_bytenr); + break; + } + + num_copies = + btrfs_num_copies(&state->root->fs_info->mapping_tree, + next_bytenr, PAGE_SIZE); + if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES) + printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n", + (unsigned long long)next_bytenr, num_copies); + for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) { + int was_created; + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "btrfsic_process_written_superblock(" + "mirror_num=%d)\n", mirror_num); + ret = btrfsic_map_block(state, next_bytenr, PAGE_SIZE, + &tmp_next_block_ctx, + mirror_num); + if (ret) { + printk(KERN_INFO + "btrfsic: btrfsic_map_block(@%llu," + " mirror=%d) failed!\n", + (unsigned long long)next_bytenr, + mirror_num); + return -1; + } + + next_block = btrfsic_block_lookup_or_add( + state, + &tmp_next_block_ctx, + additional_string, + 1, 0, 1, + mirror_num, + &was_created); + if (NULL == next_block) { + printk(KERN_INFO + "btrfsic: error, kmalloc failed!\n"); + btrfsic_release_block_ctx(&tmp_next_block_ctx); + return -1; + } + + next_block->disk_key = tmp_disk_key; + if (was_created) + next_block->generation = + BTRFSIC_GENERATION_UNKNOWN; + l = btrfsic_block_link_lookup_or_add( + state, + &tmp_next_block_ctx, + next_block, + superblock, + BTRFSIC_GENERATION_UNKNOWN); + btrfsic_release_block_ctx(&tmp_next_block_ctx); + if (NULL == l) + return -1; + } + } + + if (-1 == btrfsic_check_all_ref_blocks(state, superblock, 0)) { + WARN_ON(1); + btrfsic_dump_tree(state); + } + + return 0; +} + +static int btrfsic_check_all_ref_blocks(struct btrfsic_state *state, + struct btrfsic_block *const block, + int recursion_level) +{ + struct list_head *elem_ref_to; + int ret = 0; + + if (recursion_level >= 3 + BTRFS_MAX_LEVEL) { + /* + * Note that this situation can happen and does not + * indicate an error in regular cases. It happens + * when disk blocks are freed and later reused. + * The check-integrity module is not aware of any + * block free operations, it just recognizes block + * write operations. Therefore it keeps the linkage + * information for a block until a block is + * rewritten. This can temporarily cause incorrect + * and even circular linkage informations. This + * causes no harm unless such blocks are referenced + * by the most recent super block. + */ + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "btrfsic: abort cyclic linkage (case 1).\n"); + + return ret; + } + + /* + * This algorithm is recursive because the amount of used stack + * space is very small and the max recursion depth is limited. + */ + list_for_each(elem_ref_to, &block->ref_to_list) { + const struct btrfsic_block_link *const l = + list_entry(elem_ref_to, struct btrfsic_block_link, + node_ref_to); + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "rl=%d, %c @%llu (%s/%llu/%d)" + " %u* refers to %c @%llu (%s/%llu/%d)\n", + recursion_level, + btrfsic_get_block_type(state, block), + (unsigned long long)block->logical_bytenr, + block->dev_state->name, + (unsigned long long)block->dev_bytenr, + block->mirror_num, + l->ref_cnt, + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long) + l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num); + if (l->block_ref_to->never_written) { + printk(KERN_INFO "btrfs: attempt to write superblock" + " which references block %c @%llu (%s/%llu/%d)" + " which is never written!\n", + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long) + l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num); + ret = -1; + } else if (!l->block_ref_to->is_iodone) { + printk(KERN_INFO "btrfs: attempt to write superblock" + " which references block %c @%llu (%s/%llu/%d)" + " which is not yet iodone!\n", + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long) + l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num); + ret = -1; + } else if (l->parent_generation != + l->block_ref_to->generation && + BTRFSIC_GENERATION_UNKNOWN != + l->parent_generation && + BTRFSIC_GENERATION_UNKNOWN != + l->block_ref_to->generation) { + printk(KERN_INFO "btrfs: attempt to write superblock" + " which references block %c @%llu (%s/%llu/%d)" + " with generation %llu !=" + " parent generation %llu!\n", + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long) + l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num, + (unsigned long long)l->block_ref_to->generation, + (unsigned long long)l->parent_generation); + ret = -1; + } else if (l->block_ref_to->flush_gen > + l->block_ref_to->dev_state->last_flush_gen) { + printk(KERN_INFO "btrfs: attempt to write superblock" + " which references block %c @%llu (%s/%llu/%d)" + " which is not flushed out of disk's write cache" + " (block flush_gen=%llu," + " dev->flush_gen=%llu)!\n", + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long) + l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num, + (unsigned long long)block->flush_gen, + (unsigned long long) + l->block_ref_to->dev_state->last_flush_gen); + ret = -1; + } else if (-1 == btrfsic_check_all_ref_blocks(state, + l->block_ref_to, + recursion_level + + 1)) { + ret = -1; + } + } + + return ret; +} + +static int btrfsic_is_block_ref_by_superblock( + const struct btrfsic_state *state, + const struct btrfsic_block *block, + int recursion_level) +{ + struct list_head *elem_ref_from; + + if (recursion_level >= 3 + BTRFS_MAX_LEVEL) { + /* refer to comment at "abort cyclic linkage (case 1)" */ + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "btrfsic: abort cyclic linkage (case 2).\n"); + + return 0; + } + + /* + * This algorithm is recursive because the amount of used stack space + * is very small and the max recursion depth is limited. + */ + list_for_each(elem_ref_from, &block->ref_from_list) { + const struct btrfsic_block_link *const l = + list_entry(elem_ref_from, struct btrfsic_block_link, + node_ref_from); + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "rl=%d, %c @%llu (%s/%llu/%d)" + " is ref %u* from %c @%llu (%s/%llu/%d)\n", + recursion_level, + btrfsic_get_block_type(state, block), + (unsigned long long)block->logical_bytenr, + block->dev_state->name, + (unsigned long long)block->dev_bytenr, + block->mirror_num, + l->ref_cnt, + btrfsic_get_block_type(state, l->block_ref_from), + (unsigned long long) + l->block_ref_from->logical_bytenr, + l->block_ref_from->dev_state->name, + (unsigned long long) + l->block_ref_from->dev_bytenr, + l->block_ref_from->mirror_num); + if (l->block_ref_from->is_superblock && + state->latest_superblock->dev_bytenr == + l->block_ref_from->dev_bytenr && + state->latest_superblock->dev_state->bdev == + l->block_ref_from->dev_state->bdev) + return 1; + else if (btrfsic_is_block_ref_by_superblock(state, + l->block_ref_from, + recursion_level + + 1)) + return 1; + } + + return 0; +} + +static void btrfsic_print_add_link(const struct btrfsic_state *state, + const struct btrfsic_block_link *l) +{ + printk(KERN_INFO + "Add %u* link from %c @%llu (%s/%llu/%d)" + " to %c @%llu (%s/%llu/%d).\n", + l->ref_cnt, + btrfsic_get_block_type(state, l->block_ref_from), + (unsigned long long)l->block_ref_from->logical_bytenr, + l->block_ref_from->dev_state->name, + (unsigned long long)l->block_ref_from->dev_bytenr, + l->block_ref_from->mirror_num, + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long)l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num); +} + +static void btrfsic_print_rem_link(const struct btrfsic_state *state, + const struct btrfsic_block_link *l) +{ + printk(KERN_INFO + "Rem %u* link from %c @%llu (%s/%llu/%d)" + " to %c @%llu (%s/%llu/%d).\n", + l->ref_cnt, + btrfsic_get_block_type(state, l->block_ref_from), + (unsigned long long)l->block_ref_from->logical_bytenr, + l->block_ref_from->dev_state->name, + (unsigned long long)l->block_ref_from->dev_bytenr, + l->block_ref_from->mirror_num, + btrfsic_get_block_type(state, l->block_ref_to), + (unsigned long long)l->block_ref_to->logical_bytenr, + l->block_ref_to->dev_state->name, + (unsigned long long)l->block_ref_to->dev_bytenr, + l->block_ref_to->mirror_num); +} + +static char btrfsic_get_block_type(const struct btrfsic_state *state, + const struct btrfsic_block *block) +{ + if (block->is_superblock && + state->latest_superblock->dev_bytenr == block->dev_bytenr && + state->latest_superblock->dev_state->bdev == block->dev_state->bdev) + return 'S'; + else if (block->is_superblock) + return 's'; + else if (block->is_metadata) + return 'M'; + else + return 'D'; +} + +static void btrfsic_dump_tree(const struct btrfsic_state *state) +{ + btrfsic_dump_tree_sub(state, state->latest_superblock, 0); +} + +static void btrfsic_dump_tree_sub(const struct btrfsic_state *state, + const struct btrfsic_block *block, + int indent_level) +{ + struct list_head *elem_ref_to; + int indent_add; + static char buf[80]; + int cursor_position; + + /* + * Should better fill an on-stack buffer with a complete line and + * dump it at once when it is time to print a newline character. + */ + + /* + * This algorithm is recursive because the amount of used stack space + * is very small and the max recursion depth is limited. + */ + indent_add = sprintf(buf, "%c-%llu(%s/%llu/%d)", + btrfsic_get_block_type(state, block), + (unsigned long long)block->logical_bytenr, + block->dev_state->name, + (unsigned long long)block->dev_bytenr, + block->mirror_num); + if (indent_level + indent_add > BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL) { + printk("[...]\n"); + return; + } + printk(buf); + indent_level += indent_add; + if (list_empty(&block->ref_to_list)) { + printk("\n"); + return; + } + if (block->mirror_num > 1 && + !(state->print_mask & BTRFSIC_PRINT_MASK_TREE_WITH_ALL_MIRRORS)) { + printk(" [...]\n"); + return; + } + + cursor_position = indent_level; + list_for_each(elem_ref_to, &block->ref_to_list) { + const struct btrfsic_block_link *const l = + list_entry(elem_ref_to, struct btrfsic_block_link, + node_ref_to); + + while (cursor_position < indent_level) { + printk(" "); + cursor_position++; + } + if (l->ref_cnt > 1) + indent_add = sprintf(buf, " %d*--> ", l->ref_cnt); + else + indent_add = sprintf(buf, " --> "); + if (indent_level + indent_add > + BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL) { + printk("[...]\n"); + cursor_position = 0; + continue; + } + + printk(buf); + + btrfsic_dump_tree_sub(state, l->block_ref_to, + indent_level + indent_add); + cursor_position = 0; + } +} + +static struct btrfsic_block_link *btrfsic_block_link_lookup_or_add( + struct btrfsic_state *state, + struct btrfsic_block_data_ctx *next_block_ctx, + struct btrfsic_block *next_block, + struct btrfsic_block *from_block, + u64 parent_generation) +{ + struct btrfsic_block_link *l; + + l = btrfsic_block_link_hashtable_lookup(next_block_ctx->dev->bdev, + next_block_ctx->dev_bytenr, + from_block->dev_state->bdev, + from_block->dev_bytenr, + &state->block_link_hashtable); + if (NULL == l) { + l = btrfsic_block_link_alloc(); + if (NULL == l) { + printk(KERN_INFO + "btrfsic: error, kmalloc" " failed!\n"); + return NULL; + } + + l->block_ref_to = next_block; + l->block_ref_from = from_block; + l->ref_cnt = 1; + l->parent_generation = parent_generation; + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + btrfsic_print_add_link(state, l); + + list_add(&l->node_ref_to, &from_block->ref_to_list); + list_add(&l->node_ref_from, &next_block->ref_from_list); + + btrfsic_block_link_hashtable_add(l, + &state->block_link_hashtable); + } else { + l->ref_cnt++; + l->parent_generation = parent_generation; + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + btrfsic_print_add_link(state, l); + } + + return l; +} + +static struct btrfsic_block *btrfsic_block_lookup_or_add( + struct btrfsic_state *state, + struct btrfsic_block_data_ctx *block_ctx, + const char *additional_string, + int is_metadata, + int is_iodone, + int never_written, + int mirror_num, + int *was_created) +{ + struct btrfsic_block *block; + + block = btrfsic_block_hashtable_lookup(block_ctx->dev->bdev, + block_ctx->dev_bytenr, + &state->block_hashtable); + if (NULL == block) { + struct btrfsic_dev_state *dev_state; + + block = btrfsic_block_alloc(); + if (NULL == block) { + printk(KERN_INFO "btrfsic: error, kmalloc failed!\n"); + return NULL; + } + dev_state = btrfsic_dev_state_lookup(block_ctx->dev->bdev); + if (NULL == dev_state) { + printk(KERN_INFO + "btrfsic: error, lookup dev_state failed!\n"); + btrfsic_block_free(block); + return NULL; + } + block->dev_state = dev_state; + block->dev_bytenr = block_ctx->dev_bytenr; + block->logical_bytenr = block_ctx->start; + block->is_metadata = is_metadata; + block->is_iodone = is_iodone; + block->never_written = never_written; + block->mirror_num = mirror_num; + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + printk(KERN_INFO + "New %s%c-block @%llu (%s/%llu/%d)\n", + additional_string, + btrfsic_get_block_type(state, block), + (unsigned long long)block->logical_bytenr, + dev_state->name, + (unsigned long long)block->dev_bytenr, + mirror_num); + list_add(&block->all_blocks_node, &state->all_blocks_list); + btrfsic_block_hashtable_add(block, &state->block_hashtable); + if (NULL != was_created) + *was_created = 1; + } else { + if (NULL != was_created) + *was_created = 0; + } + + return block; +} + +static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state, + u64 bytenr, + struct btrfsic_dev_state *dev_state, + u64 dev_bytenr, char *data) +{ + int num_copies; + int mirror_num; + int ret; + struct btrfsic_block_data_ctx block_ctx; + int match = 0; + + num_copies = btrfs_num_copies(&state->root->fs_info->mapping_tree, + bytenr, PAGE_SIZE); + + for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) { + ret = btrfsic_map_block(state, bytenr, PAGE_SIZE, + &block_ctx, mirror_num); + if (ret) { + printk(KERN_INFO "btrfsic:" + " btrfsic_map_block(logical @%llu," + " mirror %d) failed!\n", + (unsigned long long)bytenr, mirror_num); + continue; + } + + if (dev_state->bdev == block_ctx.dev->bdev && + dev_bytenr == block_ctx.dev_bytenr) { + match++; + btrfsic_release_block_ctx(&block_ctx); + break; + } + btrfsic_release_block_ctx(&block_ctx); + } + + if (!match) { + printk(KERN_INFO "btrfs: attempt to write M-block which contains logical bytenr that doesn't map to dev+physical bytenr of submit_bio," + " buffer->log_bytenr=%llu, submit_bio(bdev=%s," + " phys_bytenr=%llu)!\n", + (unsigned long long)bytenr, dev_state->name, + (unsigned long long)dev_bytenr); + for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) { + ret = btrfsic_map_block(state, bytenr, PAGE_SIZE, + &block_ctx, mirror_num); + if (ret) + continue; + + printk(KERN_INFO "Read logical bytenr @%llu maps to" + " (%s/%llu/%d)\n", + (unsigned long long)bytenr, + block_ctx.dev->name, + (unsigned long long)block_ctx.dev_bytenr, + mirror_num); + } + WARN_ON(1); + } +} + +static struct btrfsic_dev_state *btrfsic_dev_state_lookup( + struct block_device *bdev) +{ + struct btrfsic_dev_state *ds; + + ds = btrfsic_dev_state_hashtable_lookup(bdev, + &btrfsic_dev_state_hashtable); + return ds; +} + +int btrfsic_submit_bh(int rw, struct buffer_head *bh) +{ + struct btrfsic_dev_state *dev_state; + + if (!btrfsic_is_initialized) + return submit_bh(rw, bh); + + mutex_lock(&btrfsic_mutex); + /* since btrfsic_submit_bh() might also be called before + * btrfsic_mount(), this might return NULL */ + dev_state = btrfsic_dev_state_lookup(bh->b_bdev); + + /* Only called to write the superblock (incl. FLUSH/FUA) */ + if (NULL != dev_state && + (rw & WRITE) && bh->b_size > 0) { + u64 dev_bytenr; + + dev_bytenr = 4096 * bh->b_blocknr; + if (dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH) + printk(KERN_INFO + "submit_bh(rw=0x%x, blocknr=%lu (bytenr %llu)," + " size=%lu, data=%p, bdev=%p)\n", + rw, bh->b_blocknr, + (unsigned long long)dev_bytenr, bh->b_size, + bh->b_data, bh->b_bdev); + btrfsic_process_written_block(dev_state, dev_bytenr, + bh->b_data, bh->b_size, NULL, + NULL, bh, rw); + } else if (NULL != dev_state && (rw & REQ_FLUSH)) { + if (dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH) + printk(KERN_INFO + "submit_bh(rw=0x%x) FLUSH, bdev=%p)\n", + rw, bh->b_bdev); + if (!dev_state->dummy_block_for_bio_bh_flush.is_iodone) { + if ((dev_state->state->print_mask & + (BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH | + BTRFSIC_PRINT_MASK_VERBOSE))) + printk(KERN_INFO + "btrfsic_submit_bh(%s) with FLUSH" + " but dummy block already in use" + " (ignored)!\n", + dev_state->name); + } else { + struct btrfsic_block *const block = + &dev_state->dummy_block_for_bio_bh_flush; + + block->is_iodone = 0; + block->never_written = 0; + block->iodone_w_error = 0; + block->flush_gen = dev_state->last_flush_gen + 1; + block->submit_bio_bh_rw = rw; + block->orig_bio_bh_private = bh->b_private; + block->orig_bio_bh_end_io.bh = bh->b_end_io; + block->next_in_same_bio = NULL; + bh->b_private = block; + bh->b_end_io = btrfsic_bh_end_io; + } + } + mutex_unlock(&btrfsic_mutex); + return submit_bh(rw, bh); +} + +void btrfsic_submit_bio(int rw, struct bio *bio) +{ + struct btrfsic_dev_state *dev_state; + + if (!btrfsic_is_initialized) { + submit_bio(rw, bio); + return; + } + + mutex_lock(&btrfsic_mutex); + /* since btrfsic_submit_bio() is also called before + * btrfsic_mount(), this might return NULL */ + dev_state = btrfsic_dev_state_lookup(bio->bi_bdev); + if (NULL != dev_state && + (rw & WRITE) && NULL != bio->bi_io_vec) { + unsigned int i; + u64 dev_bytenr; + int bio_is_patched; + + dev_bytenr = 512 * bio->bi_sector; + bio_is_patched = 0; + if (dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH) + printk(KERN_INFO + "submit_bio(rw=0x%x, bi_vcnt=%u," + " bi_sector=%lu (bytenr %llu), bi_bdev=%p)\n", + rw, bio->bi_vcnt, bio->bi_sector, + (unsigned long long)dev_bytenr, + bio->bi_bdev); + + for (i = 0; i < bio->bi_vcnt; i++) { + u8 *mapped_data; + + mapped_data = kmap(bio->bi_io_vec[i].bv_page); + if ((BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH | + BTRFSIC_PRINT_MASK_VERBOSE) == + (dev_state->state->print_mask & + (BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH | + BTRFSIC_PRINT_MASK_VERBOSE))) + printk(KERN_INFO + "#%u: page=%p, mapped=%p, len=%u," + " offset=%u\n", + i, bio->bi_io_vec[i].bv_page, + mapped_data, + bio->bi_io_vec[i].bv_len, + bio->bi_io_vec[i].bv_offset); + btrfsic_process_written_block(dev_state, dev_bytenr, + mapped_data, + bio->bi_io_vec[i].bv_len, + bio, &bio_is_patched, + NULL, rw); + kunmap(bio->bi_io_vec[i].bv_page); + dev_bytenr += bio->bi_io_vec[i].bv_len; + } + } else if (NULL != dev_state && (rw & REQ_FLUSH)) { + if (dev_state->state->print_mask & + BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH) + printk(KERN_INFO + "submit_bio(rw=0x%x) FLUSH, bdev=%p)\n", + rw, bio->bi_bdev); + if (!dev_state->dummy_block_for_bio_bh_flush.is_iodone) { + if ((dev_state->state->print_mask & + (BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH | + BTRFSIC_PRINT_MASK_VERBOSE))) + printk(KERN_INFO + "btrfsic_submit_bio(%s) with FLUSH" + " but dummy block already in use" + " (ignored)!\n", + dev_state->name); + } else { + struct btrfsic_block *const block = + &dev_state->dummy_block_for_bio_bh_flush; + + block->is_iodone = 0; + block->never_written = 0; + block->iodone_w_error = 0; + block->flush_gen = dev_state->last_flush_gen + 1; + block->submit_bio_bh_rw = rw; + block->orig_bio_bh_private = bio->bi_private; + block->orig_bio_bh_end_io.bio = bio->bi_end_io; + block->next_in_same_bio = NULL; + bio->bi_private = block; + bio->bi_end_io = btrfsic_bio_end_io; + } + } + mutex_unlock(&btrfsic_mutex); + + submit_bio(rw, bio); +} + +int btrfsic_mount(struct btrfs_root *root, + struct btrfs_fs_devices *fs_devices, + int including_extent_data, u32 print_mask) +{ + int ret; + struct btrfsic_state *state; + struct list_head *dev_head = &fs_devices->devices; + struct btrfs_device *device; + + state = kzalloc(sizeof(*state), GFP_NOFS); + if (NULL == state) { + printk(KERN_INFO "btrfs check-integrity: kmalloc() failed!\n"); + return -1; + } + + if (!btrfsic_is_initialized) { + mutex_init(&btrfsic_mutex); + btrfsic_dev_state_hashtable_init(&btrfsic_dev_state_hashtable); + btrfsic_is_initialized = 1; + } + mutex_lock(&btrfsic_mutex); + state->root = root; + state->print_mask = print_mask; + state->include_extent_data = including_extent_data; + state->csum_size = 0; + INIT_LIST_HEAD(&state->all_blocks_list); + btrfsic_block_hashtable_init(&state->block_hashtable); + btrfsic_block_link_hashtable_init(&state->block_link_hashtable); + state->max_superblock_generation = 0; + state->latest_superblock = NULL; + + list_for_each_entry(device, dev_head, dev_list) { + struct btrfsic_dev_state *ds; + char *p; + + if (!device->bdev || !device->name) + continue; + + ds = btrfsic_dev_state_alloc(); + if (NULL == ds) { + printk(KERN_INFO + "btrfs check-integrity: kmalloc() failed!\n"); + mutex_unlock(&btrfsic_mutex); + return -1; + } + ds->bdev = device->bdev; + ds->state = state; + bdevname(ds->bdev, ds->name); + ds->name[BDEVNAME_SIZE - 1] = '\0'; + for (p = ds->name; *p != '\0'; p++); + while (p > ds->name && *p != '/') + p--; + if (*p == '/') + p++; + strlcpy(ds->name, p, sizeof(ds->name)); + btrfsic_dev_state_hashtable_add(ds, + &btrfsic_dev_state_hashtable); + } + + ret = btrfsic_process_superblock(state, fs_devices); + if (0 != ret) { + mutex_unlock(&btrfsic_mutex); + btrfsic_unmount(root, fs_devices); + return ret; + } + + if (state->print_mask & BTRFSIC_PRINT_MASK_INITIAL_DATABASE) + btrfsic_dump_database(state); + if (state->print_mask & BTRFSIC_PRINT_MASK_INITIAL_TREE) + btrfsic_dump_tree(state); + + mutex_unlock(&btrfsic_mutex); + return 0; +} + +void btrfsic_unmount(struct btrfs_root *root, + struct btrfs_fs_devices *fs_devices) +{ + struct list_head *elem_all; + struct list_head *tmp_all; + struct btrfsic_state *state; + struct list_head *dev_head = &fs_devices->devices; + struct btrfs_device *device; + + if (!btrfsic_is_initialized) + return; + + mutex_lock(&btrfsic_mutex); + + state = NULL; + list_for_each_entry(device, dev_head, dev_list) { + struct btrfsic_dev_state *ds; + + if (!device->bdev || !device->name) + continue; + + ds = btrfsic_dev_state_hashtable_lookup( + device->bdev, + &btrfsic_dev_state_hashtable); + if (NULL != ds) { + state = ds->state; + btrfsic_dev_state_hashtable_remove(ds); + btrfsic_dev_state_free(ds); + } + } + + if (NULL == state) { + printk(KERN_INFO + "btrfsic: error, cannot find state information" + " on umount!\n"); + mutex_unlock(&btrfsic_mutex); + return; + } + + /* + * Don't care about keeping the lists' state up to date, + * just free all memory that was allocated dynamically. + * Free the blocks and the block_links. + */ + list_for_each_safe(elem_all, tmp_all, &state->all_blocks_list) { + struct btrfsic_block *const b_all = + list_entry(elem_all, struct btrfsic_block, + all_blocks_node); + struct list_head *elem_ref_to; + struct list_head *tmp_ref_to; + + list_for_each_safe(elem_ref_to, tmp_ref_to, + &b_all->ref_to_list) { + struct btrfsic_block_link *const l = + list_entry(elem_ref_to, + struct btrfsic_block_link, + node_ref_to); + + if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE) + btrfsic_print_rem_link(state, l); + + l->ref_cnt--; + if (0 == l->ref_cnt) + btrfsic_block_link_free(l); + } + + if (b_all->is_iodone) + btrfsic_block_free(b_all); + else + printk(KERN_INFO "btrfs: attempt to free %c-block" + " @%llu (%s/%llu/%d) on umount which is" + " not yet iodone!\n", + btrfsic_get_block_type(state, b_all), + (unsigned long long)b_all->logical_bytenr, + b_all->dev_state->name, + (unsigned long long)b_all->dev_bytenr, + b_all->mirror_num); + } + + mutex_unlock(&btrfsic_mutex); + + kfree(state); +} diff --git a/fs/btrfs/check-integrity.h b/fs/btrfs/check-integrity.h new file mode 100644 index 000000000000..8b59175cc502 --- /dev/null +++ b/fs/btrfs/check-integrity.h @@ -0,0 +1,36 @@ +/* + * Copyright (C) STRATO AG 2011. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ + +#if !defined(__BTRFS_CHECK_INTEGRITY__) +#define __BTRFS_CHECK_INTEGRITY__ + +#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY +int btrfsic_submit_bh(int rw, struct buffer_head *bh); +void btrfsic_submit_bio(int rw, struct bio *bio); +#else +#define btrfsic_submit_bh submit_bh +#define btrfsic_submit_bio submit_bio +#endif + +int btrfsic_mount(struct btrfs_root *root, + struct btrfs_fs_devices *fs_devices, + int including_extent_data, u32 print_mask); +void btrfsic_unmount(struct btrfs_root *root, + struct btrfs_fs_devices *fs_devices); + +#endif From c975dd469d748ce619c510050d4fb407c2398591 Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Tue, 1 Nov 2011 17:06:04 +0100 Subject: [PATCH 166/820] Btrfs: add config option to enable btrfs integrity check Added the BTRFS_FS_CHECK_INTEGRITY option to Kconfig. It depends on BTRFS_FS. Signed-off-by: Stefan Behrens --- fs/btrfs/Kconfig | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/fs/btrfs/Kconfig b/fs/btrfs/Kconfig index ecb9fd3be143..d33f01c08b60 100644 --- a/fs/btrfs/Kconfig +++ b/fs/btrfs/Kconfig @@ -31,3 +31,22 @@ config BTRFS_FS_POSIX_ACL Linux website . If you don't know what Access Control Lists are, say N + +config BTRFS_FS_CHECK_INTEGRITY + bool "Btrfs with integrity check tool compiled in (DANGEROUS)" + depends on BTRFS_FS + help + Adds code that examines all block write requests (including + writes of the super block). The goal is to verify that the + state of the filesystem on disk is always consistent, i.e., + after a power-loss or kernel panic event the filesystem is + in a consistent state. + + If the integrity check tool is included and activated in + the mount options, plenty of kernel memory is used, and + plenty of additional CPU cycles are spent. Enabling this + functionality is not intended for normal use. + + In most cases, unless you are a btrfs developer who needs + to verify the integrity of (super)-block write requests + during the run of a regression test, say N From f11e4d7f533249ddfa110116200c5c3a509f9218 Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Tue, 1 Nov 2011 17:06:39 +0100 Subject: [PATCH 167/820] Btrfs: Makefile changes to optionally include btrfs integrity check If the btrfs integrity check is enabled, the files required to implement the checks are included in the build. Signed-off-by: Stefan Behrens --- fs/btrfs/Makefile | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/btrfs/Makefile b/fs/btrfs/Makefile index c0ddfd29c5e5..bc5b3556cee6 100644 --- a/fs/btrfs/Makefile +++ b/fs/btrfs/Makefile @@ -11,3 +11,4 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \ reada.o backref.o btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o +btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o From 21adbd5cbb5344a3fca6bb7ddb2ab6cb03c44546 Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Wed, 9 Nov 2011 13:44:05 +0100 Subject: [PATCH 168/820] Btrfs: integrate integrity check module into btrfs This is the last part of the patch series. It modifies the btrfs code to use the integrity check module if configured to do so with the define BTRFS_FS_CHECK_INTEGRITY. If this define is not set, the only effective change is that code is added that handles the mount option to activate the integrity check. If the mount option is set and the define BTRFS_FS_CHECK_INTEGRITY is not set, that code complains in the log and the mount fails with EINVAL. Add the mount option to activate the usage of the integrity check code. Add invocation of btrfs integrity check code init and cleanup function on mount and umount, respectively. Add hook to call btrfs integrity check code version of submit_bh/submit_bio. Signed-off-by: Stefan Behrens --- fs/btrfs/ctree.h | 8 +++++++- fs/btrfs/disk-io.c | 26 ++++++++++++++++++++++++-- fs/btrfs/extent_io.c | 5 +++-- fs/btrfs/scrub.c | 5 +++-- fs/btrfs/super.c | 39 ++++++++++++++++++++++++++++++++++++++- fs/btrfs/volumes.c | 7 ++++--- 6 files changed, 79 insertions(+), 11 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 67385033323d..39f6188688e6 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -971,7 +971,7 @@ struct btrfs_fs_info { * is required instead of the faster short fsync log commits */ u64 last_trans_log_full_commit; - unsigned long mount_opt:20; + unsigned long mount_opt:21; unsigned long compress_type:4; u64 max_inline; u64 alloc_start; @@ -1155,6 +1155,10 @@ struct btrfs_fs_info { int scrub_workers_refcnt; struct btrfs_workers scrub_workers; +#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY + u32 check_integrity_print_mask; +#endif + /* filesystem state */ u64 fs_state; @@ -1413,6 +1417,8 @@ struct btrfs_ioctl_defrag_range_args { #define BTRFS_MOUNT_AUTO_DEFRAG (1 << 16) #define BTRFS_MOUNT_INODE_MAP_CACHE (1 << 17) #define BTRFS_MOUNT_RECOVERY (1 << 18) +#define BTRFS_MOUNT_CHECK_INTEGRITY (1 << 19) +#define BTRFS_MOUNT_CHECK_INTEGRITY_INCLUDING_EXTENT_DATA (1 << 20) #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 3f9d5551e582..f363c6d9c3de 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -43,6 +43,7 @@ #include "tree-log.h" #include "free-space-cache.h" #include "inode-map.h" +#include "check-integrity.h" static struct extent_io_ops btree_extent_io_ops; static void end_workqueue_fn(struct btrfs_work *work); @@ -2001,6 +2002,9 @@ struct btrfs_root *open_ctree(struct super_block *sb, init_waitqueue_head(&fs_info->scrub_pause_wait); init_rwsem(&fs_info->scrub_super_lock); fs_info->scrub_workers_refcnt = 0; +#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY + fs_info->check_integrity_print_mask = 0; +#endif sb->s_blocksize = 4096; sb->s_blocksize_bits = blksize_bits(4096); @@ -2356,6 +2360,19 @@ retry_root_backup: btrfs_set_opt(fs_info->mount_opt, SSD); } +#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY + if (btrfs_test_opt(tree_root, CHECK_INTEGRITY)) { + ret = btrfsic_mount(tree_root, fs_devices, + btrfs_test_opt(tree_root, + CHECK_INTEGRITY_INCLUDING_EXTENT_DATA) ? + 1 : 0, + fs_info->check_integrity_print_mask); + if (ret) + printk(KERN_WARNING "btrfs: failed to initialize" + " integrity check module %s\n", sb->s_id); + } +#endif + /* do not make disk changes in broken FS */ if (btrfs_super_log_root(disk_super) != 0 && !(fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR)) { @@ -2634,7 +2651,7 @@ static int write_dev_supers(struct btrfs_device *device, * we fua the first super. The others we allow * to go down lazy. */ - ret = submit_bh(WRITE_FUA, bh); + ret = btrfsic_submit_bh(WRITE_FUA, bh); if (ret) errors++; } @@ -2711,7 +2728,7 @@ static int write_dev_flush(struct btrfs_device *device, int wait) device->flush_bio = bio; bio_get(bio); - submit_bio(WRITE_FLUSH, bio); + btrfsic_submit_bio(WRITE_FLUSH, bio); return 0; } @@ -3057,6 +3074,11 @@ int close_ctree(struct btrfs_root *root) btrfs_stop_workers(&fs_info->caching_workers); btrfs_stop_workers(&fs_info->readahead_workers); +#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY + if (btrfs_test_opt(root, CHECK_INTEGRITY)) + btrfsic_unmount(root, fs_info->fs_devices); +#endif + btrfs_close_devices(fs_info->fs_devices); btrfs_mapping_tree_free(&fs_info->mapping_tree); diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 49f3c9dc09f4..246669296e02 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -18,6 +18,7 @@ #include "ctree.h" #include "btrfs_inode.h" #include "volumes.h" +#include "check-integrity.h" static struct kmem_cache *extent_state_cache; static struct kmem_cache *extent_buffer_cache; @@ -1895,7 +1896,7 @@ int repair_io_failure(struct btrfs_mapping_tree *map_tree, u64 start, } bio->bi_bdev = dev->bdev; bio_add_page(bio, page, length, start-page_offset(page)); - submit_bio(WRITE_SYNC, bio); + btrfsic_submit_bio(WRITE_SYNC, bio); wait_for_completion(&compl); if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) { @@ -2393,7 +2394,7 @@ static int submit_one_bio(int rw, struct bio *bio, int mirror_num, ret = tree->ops->submit_bio_hook(page->mapping->host, rw, bio, mirror_num, bio_flags, start); else - submit_bio(rw, bio); + btrfsic_submit_bio(rw, bio); if (bio_flagged(bio, BIO_EOPNOTSUPP)) ret = -EOPNOTSUPP; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index ddf2c90d3fc0..567e148caca2 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -25,6 +25,7 @@ #include "transaction.h" #include "backref.h" #include "extent_io.h" +#include "check-integrity.h" /* * This is only the first step towards a full-features scrub. It reads all @@ -732,7 +733,7 @@ static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector, bio_add_page(bio, page, PAGE_SIZE, 0); bio->bi_end_io = scrub_fixup_end_io; bio->bi_private = &complete; - submit_bio(rw, bio); + btrfsic_submit_bio(rw, bio); /* this will also unplug the queue */ wait_for_completion(&complete); @@ -958,7 +959,7 @@ static int scrub_submit(struct scrub_dev *sdev) sdev->curr = -1; atomic_inc(&sdev->in_flight); - submit_bio(READ, sbio->bio); + btrfsic_submit_bio(READ, sbio->bio); return 0; } diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 34a8b6112ea4..22a2015f1d7b 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -165,7 +165,10 @@ enum { Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, - Opt_inode_cache, Opt_no_space_cache, Opt_recovery, Opt_err, + Opt_inode_cache, Opt_no_space_cache, Opt_recovery, + Opt_check_integrity, Opt_check_integrity_including_extent_data, + Opt_check_integrity_print_mask, + Opt_err, }; static match_table_t tokens = { @@ -200,6 +203,9 @@ static match_table_t tokens = { {Opt_inode_cache, "inode_cache"}, {Opt_no_space_cache, "nospace_cache"}, {Opt_recovery, "recovery"}, + {Opt_check_integrity, "check_int"}, + {Opt_check_integrity_including_extent_data, "check_int_data"}, + {Opt_check_integrity_print_mask, "check_int_print_mask=%d"}, {Opt_err, NULL}, }; @@ -398,6 +404,37 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) printk(KERN_INFO "btrfs: enabling auto recovery"); btrfs_set_opt(info->mount_opt, RECOVERY); break; +#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY + case Opt_check_integrity_including_extent_data: + printk(KERN_INFO "btrfs: enabling check integrity" + " including extent data\n"); + btrfs_set_opt(info->mount_opt, + CHECK_INTEGRITY_INCLUDING_EXTENT_DATA); + btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY); + break; + case Opt_check_integrity: + printk(KERN_INFO "btrfs: enabling check integrity\n"); + btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY); + break; + case Opt_check_integrity_print_mask: + intarg = 0; + match_int(&args[0], &intarg); + if (intarg) { + info->check_integrity_print_mask = intarg; + printk(KERN_INFO "btrfs:" + " check_integrity_print_mask 0x%x\n", + info->check_integrity_print_mask); + } + break; +#else + case Opt_check_integrity_including_extent_data: + case Opt_check_integrity: + case Opt_check_integrity_print_mask: + printk(KERN_ERR "btrfs: support for check_integrity*" + " not compiled in!\n"); + ret = -EINVAL; + goto out; +#endif case Opt_err: printk(KERN_INFO "btrfs: unrecognized mount option " "'%s'\n", p); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f4b839fd3c9d..821334f6e3a1 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -32,6 +32,7 @@ #include "print-tree.h" #include "volumes.h" #include "async-thread.h" +#include "check-integrity.h" static int init_first_rw_device(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -246,7 +247,7 @@ loop_lock: sync_pending = 0; } - submit_bio(cur->bi_rw, cur); + btrfsic_submit_bio(cur->bi_rw, cur); num_run++; batch_run++; if (need_resched()) @@ -3304,7 +3305,7 @@ static noinline int schedule_bio(struct btrfs_root *root, /* don't bother with additional async steps for reads, right now */ if (!(rw & REQ_WRITE)) { bio_get(bio); - submit_bio(rw, bio); + btrfsic_submit_bio(rw, bio); bio_put(bio); return 0; } @@ -3399,7 +3400,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, if (async_submit) schedule_bio(root, dev, rw, bio); else - submit_bio(rw, bio); + btrfsic_submit_bio(rw, bio); } else { bio->bi_bdev = root->fs_info->fs_devices->latest_bdev; bio->bi_sector = logical >> 9; From da5c81356426c476112f2b59fe64bdb1b37f079d Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 13 Sep 2011 12:29:12 +0200 Subject: [PATCH 169/820] Btrfs: generic data structure to build unique lists ulist is a generic data structures to hold a collection of unique u64 values. The only operations it supports is adding to the list and enumerating it. It is possible to store an auxiliary value along with the key. The implementation is preliminary and can probably be sped up significantly. It is used by btrfs_find_all_roots() quota to translate recursions into iterative loops. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/Makefile | 2 +- fs/btrfs/ulist.c | 220 ++++++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/ulist.h | 68 ++++++++++++++ 3 files changed, 289 insertions(+), 1 deletion(-) create mode 100644 fs/btrfs/ulist.c create mode 100644 fs/btrfs/ulist.h diff --git a/fs/btrfs/Makefile b/fs/btrfs/Makefile index c0ddfd29c5e5..70798407b9a2 100644 --- a/fs/btrfs/Makefile +++ b/fs/btrfs/Makefile @@ -8,6 +8,6 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \ extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.o \ export.o tree-log.o free-space-cache.o zlib.o lzo.o \ compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o \ - reada.o backref.o + reada.o backref.o ulist.o btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o diff --git a/fs/btrfs/ulist.c b/fs/btrfs/ulist.c new file mode 100644 index 000000000000..12f5147bd2b1 --- /dev/null +++ b/fs/btrfs/ulist.c @@ -0,0 +1,220 @@ +/* + * Copyright (C) 2011 STRATO AG + * written by Arne Jansen + * Distributed under the GNU GPL license version 2. + */ + +#include +#include +#include "ulist.h" + +/* + * ulist is a generic data structure to hold a collection of unique u64 + * values. The only operations it supports is adding to the list and + * enumerating it. + * It is possible to store an auxiliary value along with the key. + * + * The implementation is preliminary and can probably be sped up + * significantly. A first step would be to store the values in an rbtree + * as soon as ULIST_SIZE is exceeded. + * + * A sample usage for ulists is the enumeration of directed graphs without + * visiting a node twice. The pseudo-code could look like this: + * + * ulist = ulist_alloc(); + * ulist_add(ulist, root); + * elem = NULL; + * + * while ((elem = ulist_next(ulist, elem)) { + * for (all child nodes n in elem) + * ulist_add(ulist, n); + * do something useful with the node; + * } + * ulist_free(ulist); + * + * This assumes the graph nodes are adressable by u64. This stems from the + * usage for tree enumeration in btrfs, where the logical addresses are + * 64 bit. + * + * It is also useful for tree enumeration which could be done elegantly + * recursively, but is not possible due to kernel stack limitations. The + * loop would be similar to the above. + */ + +/** + * ulist_init - freshly initialize a ulist + * @ulist: the ulist to initialize + * + * Note: don't use this function to init an already used ulist, use + * ulist_reinit instead. + */ +void ulist_init(struct ulist *ulist) +{ + ulist->nnodes = 0; + ulist->nodes = ulist->int_nodes; + ulist->nodes_alloced = ULIST_SIZE; +} +EXPORT_SYMBOL(ulist_init); + +/** + * ulist_fini - free up additionally allocated memory for the ulist + * @ulist: the ulist from which to free the additional memory + * + * This is useful in cases where the base 'struct ulist' has been statically + * allocated. + */ +void ulist_fini(struct ulist *ulist) +{ + /* + * The first ULIST_SIZE elements are stored inline in struct ulist. + * Only if more elements are alocated they need to be freed. + */ + if (ulist->nodes_alloced > ULIST_SIZE) + kfree(ulist->nodes); + ulist->nodes_alloced = 0; /* in case ulist_fini is called twice */ +} +EXPORT_SYMBOL(ulist_fini); + +/** + * ulist_reinit - prepare a ulist for reuse + * @ulist: ulist to be reused + * + * Free up all additional memory allocated for the list elements and reinit + * the ulist. + */ +void ulist_reinit(struct ulist *ulist) +{ + ulist_fini(ulist); + ulist_init(ulist); +} +EXPORT_SYMBOL(ulist_reinit); + +/** + * ulist_alloc - dynamically allocate a ulist + * @gfp_mask: allocation flags to for base allocation + * + * The allocated ulist will be returned in an initialized state. + */ +struct ulist *ulist_alloc(unsigned long gfp_mask) +{ + struct ulist *ulist = kmalloc(sizeof(*ulist), gfp_mask); + + if (!ulist) + return NULL; + + ulist_init(ulist); + + return ulist; +} +EXPORT_SYMBOL(ulist_alloc); + +/** + * ulist_free - free dynamically allocated ulist + * @ulist: ulist to free + * + * It is not necessary to call ulist_fini before. + */ +void ulist_free(struct ulist *ulist) +{ + if (!ulist) + return; + ulist_fini(ulist); + kfree(ulist); +} +EXPORT_SYMBOL(ulist_free); + +/** + * ulist_add - add an element to the ulist + * @ulist: ulist to add the element to + * @val: value to add to ulist + * @aux: auxiliary value to store along with val + * @gfp_mask: flags to use for allocation + * + * Note: locking must be provided by the caller. In case of rwlocks write + * locking is needed + * + * Add an element to a ulist. The @val will only be added if it doesn't + * already exist. If it is added, the auxiliary value @aux is stored along with + * it. In case @val already exists in the ulist, @aux is ignored, even if + * it differs from the already stored value. + * + * ulist_add returns 0 if @val already exists in ulist and 1 if @val has been + * inserted. + * In case of allocation failure -ENOMEM is returned and the ulist stays + * unaltered. + */ +int ulist_add(struct ulist *ulist, u64 val, unsigned long aux, + unsigned long gfp_mask) +{ + int i; + + for (i = 0; i < ulist->nnodes; ++i) { + if (ulist->nodes[i].val == val) + return 0; + } + + if (ulist->nnodes >= ulist->nodes_alloced) { + u64 new_alloced = ulist->nodes_alloced + 128; + struct ulist_node *new_nodes; + void *old = NULL; + + /* + * if nodes_alloced == ULIST_SIZE no memory has been allocated + * yet, so pass NULL to krealloc + */ + if (ulist->nodes_alloced > ULIST_SIZE) + old = ulist->nodes; + + new_nodes = krealloc(old, sizeof(*new_nodes) * new_alloced, + gfp_mask); + if (!new_nodes) + return -ENOMEM; + + if (!old) + memcpy(new_nodes, ulist->int_nodes, + sizeof(ulist->int_nodes)); + + ulist->nodes = new_nodes; + ulist->nodes_alloced = new_alloced; + } + ulist->nodes[ulist->nnodes].val = val; + ulist->nodes[ulist->nnodes].aux = aux; + ++ulist->nnodes; + + return 1; +} +EXPORT_SYMBOL(ulist_add); + +/** + * ulist_next - iterate ulist + * @ulist: ulist to iterate + * @prev: previously returned element or %NULL to start iteration + * + * Note: locking must be provided by the caller. In case of rwlocks only read + * locking is needed + * + * This function is used to iterate an ulist. The iteration is started with + * @prev = %NULL. It returns the next element from the ulist or %NULL when the + * end is reached. No guarantee is made with respect to the order in which + * the elements are returned. They might neither be returned in order of + * addition nor in ascending order. + * It is allowed to call ulist_add during an enumeration. Newly added items + * are guaranteed to show up in the running enumeration. + */ +struct ulist_node *ulist_next(struct ulist *ulist, struct ulist_node *prev) +{ + int next; + + if (ulist->nnodes == 0) + return NULL; + + if (!prev) + return &ulist->nodes[0]; + + next = (prev - ulist->nodes) + 1; + if (next < 0 || next >= ulist->nnodes) + return NULL; + + return &ulist->nodes[next]; +} +EXPORT_SYMBOL(ulist_next); diff --git a/fs/btrfs/ulist.h b/fs/btrfs/ulist.h new file mode 100644 index 000000000000..2e25dec58ec0 --- /dev/null +++ b/fs/btrfs/ulist.h @@ -0,0 +1,68 @@ +/* + * Copyright (C) 2011 STRATO AG + * written by Arne Jansen + * Distributed under the GNU GPL license version 2. + * + */ + +#ifndef __ULIST__ +#define __ULIST__ + +/* + * ulist is a generic data structure to hold a collection of unique u64 + * values. The only operations it supports is adding to the list and + * enumerating it. + * It is possible to store an auxiliary value along with the key. + * + * The implementation is preliminary and can probably be sped up + * significantly. A first step would be to store the values in an rbtree + * as soon as ULIST_SIZE is exceeded. + */ + +/* + * number of elements statically allocated inside struct ulist + */ +#define ULIST_SIZE 16 + +/* + * element of the list + */ +struct ulist_node { + u64 val; /* value to store */ + unsigned long aux; /* auxiliary value saved along with the val */ +}; + +struct ulist { + /* + * number of elements stored in list + */ + unsigned long nnodes; + + /* + * number of nodes we already have room for + */ + unsigned long nodes_alloced; + + /* + * pointer to the array storing the elements. The first ULIST_SIZE + * elements are stored inline. In this case the it points to int_nodes. + * After exceeding ULIST_SIZE, dynamic memory is allocated. + */ + struct ulist_node *nodes; + + /* + * inline storage space for the first ULIST_SIZE entries + */ + struct ulist_node int_nodes[ULIST_SIZE]; +}; + +void ulist_init(struct ulist *ulist); +void ulist_fini(struct ulist *ulist); +void ulist_reinit(struct ulist *ulist); +struct ulist *ulist_alloc(unsigned long gfp_mask); +void ulist_free(struct ulist *ulist); +int ulist_add(struct ulist *ulist, u64 val, unsigned long aux, + unsigned long gfp_mask); +struct ulist_node *ulist_next(struct ulist *ulist, struct ulist_node *prev); + +#endif From c7d22a3c3cdb73d8a0151e2ccc8cf4a48c48310b Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Tue, 22 Nov 2011 15:14:33 +0100 Subject: [PATCH 170/820] Btrfs: added helper btrfs_next_item() btrfs_next_item() makes the btrfs path point to the next item, crossing leaf boundaries if needed. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/ctree.h | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 50634abef9b4..3e4a07b79817 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2482,6 +2482,13 @@ static inline int btrfs_insert_empty_item(struct btrfs_trans_handle *trans, } int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path); +static inline int btrfs_next_item(struct btrfs_root *root, struct btrfs_path *p) +{ + ++p->slots[0]; + if (p->slots[0] >= btrfs_header_nritems(p->nodes[0])) + return btrfs_next_leaf(root, p); + return 0; +} int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path); int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf); void btrfs_drop_snapshot(struct btrfs_root *root, From 66d7e7f09f77456fe68683247d77721032a00ee5 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Mon, 12 Sep 2011 15:26:38 +0200 Subject: [PATCH 171/820] Btrfs: mark delayed refs as for cow Add a for_cow parameter to add_delayed_*_ref and pass the appropriate value from every call site. The for_cow parameter will later on be used to determine if a ref will change anything with respect to qgroups. Delayed refs coming from relocation are always counted as for_cow, as they don't change subvol quota. Also pass in the fs_info for later use. btrfs_find_all_roots() will use this as an optimization, as changes that are for_cow will not change anything with respect to which root points to a certain leaf. Thus, we don't need to add the current sequence number to those delayed refs. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/ctree.c | 42 ++++++++--------- fs/btrfs/ctree.h | 17 +++---- fs/btrfs/delayed-ref.c | 50 ++++++++++++-------- fs/btrfs/delayed-ref.h | 15 ++++-- fs/btrfs/disk-io.c | 3 +- fs/btrfs/extent-tree.c | 101 ++++++++++++++++++++++++----------------- fs/btrfs/file.c | 10 ++-- fs/btrfs/inode.c | 2 +- fs/btrfs/ioctl.c | 5 +- fs/btrfs/relocation.c | 18 ++++---- fs/btrfs/transaction.c | 4 +- fs/btrfs/tree-log.c | 2 +- 12 files changed, 155 insertions(+), 114 deletions(-) diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index dede441bdeee..0639a555e16e 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -240,7 +240,7 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans, cow = btrfs_alloc_free_block(trans, root, buf->len, 0, new_root_objectid, &disk_key, level, - buf->start, 0); + buf->start, 0, 1); if (IS_ERR(cow)) return PTR_ERR(cow); @@ -261,9 +261,9 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans, WARN_ON(btrfs_header_generation(buf) > trans->transid); if (new_root_objectid == BTRFS_TREE_RELOC_OBJECTID) - ret = btrfs_inc_ref(trans, root, cow, 1); + ret = btrfs_inc_ref(trans, root, cow, 1, 1); else - ret = btrfs_inc_ref(trans, root, cow, 0); + ret = btrfs_inc_ref(trans, root, cow, 0, 1); if (ret) return ret; @@ -350,14 +350,14 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, if ((owner == root->root_key.objectid || root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) && !(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)) { - ret = btrfs_inc_ref(trans, root, buf, 1); + ret = btrfs_inc_ref(trans, root, buf, 1, 1); BUG_ON(ret); if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) { - ret = btrfs_dec_ref(trans, root, buf, 0); + ret = btrfs_dec_ref(trans, root, buf, 0, 1); BUG_ON(ret); - ret = btrfs_inc_ref(trans, root, cow, 1); + ret = btrfs_inc_ref(trans, root, cow, 1, 1); BUG_ON(ret); } new_flags |= BTRFS_BLOCK_FLAG_FULL_BACKREF; @@ -365,9 +365,9 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) - ret = btrfs_inc_ref(trans, root, cow, 1); + ret = btrfs_inc_ref(trans, root, cow, 1, 1); else - ret = btrfs_inc_ref(trans, root, cow, 0); + ret = btrfs_inc_ref(trans, root, cow, 0, 1); BUG_ON(ret); } if (new_flags != 0) { @@ -381,11 +381,11 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, if (flags & BTRFS_BLOCK_FLAG_FULL_BACKREF) { if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) - ret = btrfs_inc_ref(trans, root, cow, 1); + ret = btrfs_inc_ref(trans, root, cow, 1, 1); else - ret = btrfs_inc_ref(trans, root, cow, 0); + ret = btrfs_inc_ref(trans, root, cow, 0, 1); BUG_ON(ret); - ret = btrfs_dec_ref(trans, root, buf, 1); + ret = btrfs_dec_ref(trans, root, buf, 1, 1); BUG_ON(ret); } clean_tree_block(trans, root, buf); @@ -446,7 +446,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans, cow = btrfs_alloc_free_block(trans, root, buf->len, parent_start, root->root_key.objectid, &disk_key, - level, search_start, empty_size); + level, search_start, empty_size, 1); if (IS_ERR(cow)) return PTR_ERR(cow); @@ -484,7 +484,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans, rcu_assign_pointer(root->node, cow); btrfs_free_tree_block(trans, root, buf, parent_start, - last_ref); + last_ref, 1); free_extent_buffer(buf); add_root_to_dirty_list(root); } else { @@ -500,7 +500,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans, trans->transid); btrfs_mark_buffer_dirty(parent); btrfs_free_tree_block(trans, root, buf, parent_start, - last_ref); + last_ref, 1); } if (unlock_orig) btrfs_tree_unlock(buf); @@ -957,7 +957,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans, free_extent_buffer(mid); root_sub_used(root, mid->len); - btrfs_free_tree_block(trans, root, mid, 0, 1); + btrfs_free_tree_block(trans, root, mid, 0, 1, 0); /* once for the root ptr */ free_extent_buffer(mid); return 0; @@ -1015,7 +1015,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans, if (wret) ret = wret; root_sub_used(root, right->len); - btrfs_free_tree_block(trans, root, right, 0, 1); + btrfs_free_tree_block(trans, root, right, 0, 1, 0); free_extent_buffer(right); right = NULL; } else { @@ -1055,7 +1055,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans, if (wret) ret = wret; root_sub_used(root, mid->len); - btrfs_free_tree_block(trans, root, mid, 0, 1); + btrfs_free_tree_block(trans, root, mid, 0, 1, 0); free_extent_buffer(mid); mid = NULL; } else { @@ -2089,7 +2089,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans, c = btrfs_alloc_free_block(trans, root, root->nodesize, 0, root->root_key.objectid, &lower_key, - level, root->node->start, 0); + level, root->node->start, 0, 0); if (IS_ERR(c)) return PTR_ERR(c); @@ -2216,7 +2216,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans, split = btrfs_alloc_free_block(trans, root, root->nodesize, 0, root->root_key.objectid, - &disk_key, level, c->start, 0); + &disk_key, level, c->start, 0, 0); if (IS_ERR(split)) return PTR_ERR(split); @@ -2970,7 +2970,7 @@ again: right = btrfs_alloc_free_block(trans, root, root->leafsize, 0, root->root_key.objectid, - &disk_key, 0, l->start, 0); + &disk_key, 0, l->start, 0, 0); if (IS_ERR(right)) return PTR_ERR(right); @@ -3781,7 +3781,7 @@ static noinline int btrfs_del_leaf(struct btrfs_trans_handle *trans, root_sub_used(root, leaf->len); - btrfs_free_tree_block(trans, root, leaf, 0, 1); + btrfs_free_tree_block(trans, root, leaf, 0, 1, 0); return 0; } /* diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3e4a07b79817..543f60bddb39 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2277,11 +2277,11 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, u32 blocksize, u64 parent, u64 root_objectid, struct btrfs_disk_key *key, int level, - u64 hint, u64 empty_size); + u64 hint, u64 empty_size, int for_cow); void btrfs_free_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, - u64 parent, int last_ref); + u64 parent, int last_ref, int for_cow); struct extent_buffer *btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u32 blocksize, @@ -2301,17 +2301,17 @@ int btrfs_reserve_extent(struct btrfs_trans_handle *trans, u64 search_end, struct btrfs_key *ins, u64 data); int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref); + struct extent_buffer *buf, int full_backref, int for_cow); int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref); + struct extent_buffer *buf, int full_backref, int for_cow); int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, u64 flags, int is_data); int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, - u64 bytenr, u64 num_bytes, u64 parent, - u64 root_objectid, u64 owner, u64 offset); + u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid, + u64 owner, u64 offset, int for_cow); int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len); int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, @@ -2323,7 +2323,7 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, u64 parent, - u64 root_objectid, u64 owner, u64 offset); + u64 root_objectid, u64 owner, u64 offset, int for_cow); int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, struct btrfs_root *root); @@ -2492,7 +2492,8 @@ static inline int btrfs_next_item(struct btrfs_root *root, struct btrfs_path *p) int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path); int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf); void btrfs_drop_snapshot(struct btrfs_root *root, - struct btrfs_block_rsv *block_rsv, int update_ref); + struct btrfs_block_rsv *block_rsv, int update_ref, + int for_reloc); int btrfs_drop_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *node, diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index 125cf76fcd08..3a0f0ab804f4 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -390,7 +390,8 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing, * this does all the dirty work in terms of maintaining the correct * overall modification count. */ -static noinline int add_delayed_ref_head(struct btrfs_trans_handle *trans, +static noinline int add_delayed_ref_head(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_node *ref, u64 bytenr, u64 num_bytes, int action, int is_data) @@ -468,10 +469,12 @@ static noinline int add_delayed_ref_head(struct btrfs_trans_handle *trans, /* * helper to insert a delayed tree ref into the rbtree. */ -static noinline int add_delayed_tree_ref(struct btrfs_trans_handle *trans, +static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_node *ref, u64 bytenr, u64 num_bytes, u64 parent, - u64 ref_root, int level, int action) + u64 ref_root, int level, int action, + int for_cow) { struct btrfs_delayed_ref_node *existing; struct btrfs_delayed_tree_ref *full_ref; @@ -522,11 +525,12 @@ static noinline int add_delayed_tree_ref(struct btrfs_trans_handle *trans, /* * helper to insert a delayed data ref into the rbtree. */ -static noinline int add_delayed_data_ref(struct btrfs_trans_handle *trans, +static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_node *ref, u64 bytenr, u64 num_bytes, u64 parent, u64 ref_root, u64 owner, u64 offset, - int action) + int action, int for_cow) { struct btrfs_delayed_ref_node *existing; struct btrfs_delayed_data_ref *full_ref; @@ -554,6 +558,7 @@ static noinline int add_delayed_data_ref(struct btrfs_trans_handle *trans, full_ref->root = ref_root; ref->type = BTRFS_EXTENT_DATA_REF_KEY; } + full_ref->objectid = owner; full_ref->offset = offset; @@ -580,10 +585,12 @@ static noinline int add_delayed_data_ref(struct btrfs_trans_handle *trans, * to make sure the delayed ref is eventually processed before this * transaction commits. */ -int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans, +int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, u64 parent, u64 ref_root, int level, int action, - struct btrfs_delayed_extent_op *extent_op) + struct btrfs_delayed_extent_op *extent_op, + int for_cow) { struct btrfs_delayed_tree_ref *ref; struct btrfs_delayed_ref_head *head_ref; @@ -610,12 +617,13 @@ int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans, * insert both the head node and the new ref without dropping * the spin lock */ - ret = add_delayed_ref_head(trans, &head_ref->node, bytenr, num_bytes, - action, 0); + ret = add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr, + num_bytes, action, 0); BUG_ON(ret); - ret = add_delayed_tree_ref(trans, &ref->node, bytenr, num_bytes, - parent, ref_root, level, action); + ret = add_delayed_tree_ref(fs_info, trans, &ref->node, bytenr, + num_bytes, parent, ref_root, level, action, + for_cow); BUG_ON(ret); spin_unlock(&delayed_refs->lock); return 0; @@ -624,11 +632,13 @@ int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans, /* * add a delayed data ref. it's similar to btrfs_add_delayed_tree_ref. */ -int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans, +int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, u64 parent, u64 ref_root, u64 owner, u64 offset, int action, - struct btrfs_delayed_extent_op *extent_op) + struct btrfs_delayed_extent_op *extent_op, + int for_cow) { struct btrfs_delayed_data_ref *ref; struct btrfs_delayed_ref_head *head_ref; @@ -655,18 +665,20 @@ int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans, * insert both the head node and the new ref without dropping * the spin lock */ - ret = add_delayed_ref_head(trans, &head_ref->node, bytenr, num_bytes, - action, 1); + ret = add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr, + num_bytes, action, 1); BUG_ON(ret); - ret = add_delayed_data_ref(trans, &ref->node, bytenr, num_bytes, - parent, ref_root, owner, offset, action); + ret = add_delayed_data_ref(fs_info, trans, &ref->node, bytenr, + num_bytes, parent, ref_root, owner, offset, + action, for_cow); BUG_ON(ret); spin_unlock(&delayed_refs->lock); return 0; } -int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans, +int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, struct btrfs_delayed_extent_op *extent_op) { @@ -683,7 +695,7 @@ int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans, delayed_refs = &trans->transaction->delayed_refs; spin_lock(&delayed_refs->lock); - ret = add_delayed_ref_head(trans, &head_ref->node, bytenr, + ret = add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr, num_bytes, BTRFS_UPDATE_DELAYED_HEAD, extent_op->is_data); BUG_ON(ret); diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index e287e3b0eab0..8316bff18d30 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -151,16 +151,21 @@ static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref) } } -int btrfs_add_delayed_tree_ref(struct btrfs_trans_handle *trans, +int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, u64 parent, u64 ref_root, int level, int action, - struct btrfs_delayed_extent_op *extent_op); -int btrfs_add_delayed_data_ref(struct btrfs_trans_handle *trans, + struct btrfs_delayed_extent_op *extent_op, + int for_cow); +int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, u64 parent, u64 ref_root, u64 owner, u64 offset, int action, - struct btrfs_delayed_extent_op *extent_op); -int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans, + struct btrfs_delayed_extent_op *extent_op, + int for_cow); +int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, struct btrfs_delayed_extent_op *extent_op); diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 94abc25392f6..6f8cd17c9a9f 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1243,7 +1243,8 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, root->ref_cows = 0; leaf = btrfs_alloc_free_block(trans, root, root->leafsize, 0, - BTRFS_TREE_LOG_OBJECTID, NULL, 0, 0, 0); + BTRFS_TREE_LOG_OBJECTID, NULL, + 0, 0, 0, 0); if (IS_ERR(leaf)) { kfree(root); return ERR_CAST(leaf); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 813c6bb96c9a..dc8b9a834596 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -1872,20 +1872,24 @@ static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr, int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, u64 parent, - u64 root_objectid, u64 owner, u64 offset) + u64 root_objectid, u64 owner, u64 offset, int for_cow) { int ret; + struct btrfs_fs_info *fs_info = root->fs_info; + BUG_ON(owner < BTRFS_FIRST_FREE_OBJECTID && root_objectid == BTRFS_TREE_LOG_OBJECTID); if (owner < BTRFS_FIRST_FREE_OBJECTID) { - ret = btrfs_add_delayed_tree_ref(trans, bytenr, num_bytes, + ret = btrfs_add_delayed_tree_ref(fs_info, trans, bytenr, + num_bytes, parent, root_objectid, (int)owner, - BTRFS_ADD_DELAYED_REF, NULL); + BTRFS_ADD_DELAYED_REF, NULL, for_cow); } else { - ret = btrfs_add_delayed_data_ref(trans, bytenr, num_bytes, + ret = btrfs_add_delayed_data_ref(fs_info, trans, bytenr, + num_bytes, parent, root_objectid, owner, offset, - BTRFS_ADD_DELAYED_REF, NULL); + BTRFS_ADD_DELAYED_REF, NULL, for_cow); } return ret; } @@ -2405,7 +2409,8 @@ int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans, extent_op->update_key = 0; extent_op->is_data = is_data ? 1 : 0; - ret = btrfs_add_delayed_extent_op(trans, bytenr, num_bytes, extent_op); + ret = btrfs_add_delayed_extent_op(root->fs_info, trans, bytenr, + num_bytes, extent_op); if (ret) kfree(extent_op); return ret; @@ -2590,7 +2595,7 @@ out: static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, - int full_backref, int inc) + int full_backref, int inc, int for_cow) { u64 bytenr; u64 num_bytes; @@ -2603,7 +2608,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, int level; int ret = 0; int (*process_func)(struct btrfs_trans_handle *, struct btrfs_root *, - u64, u64, u64, u64, u64, u64); + u64, u64, u64, u64, u64, u64, int); ref_root = btrfs_header_owner(buf); nritems = btrfs_header_nritems(buf); @@ -2640,14 +2645,15 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, key.offset -= btrfs_file_extent_offset(buf, fi); ret = process_func(trans, root, bytenr, num_bytes, parent, ref_root, key.objectid, - key.offset); + key.offset, for_cow); if (ret) goto fail; } else { bytenr = btrfs_node_blockptr(buf, i); num_bytes = btrfs_level_size(root, level - 1); ret = process_func(trans, root, bytenr, num_bytes, - parent, ref_root, level - 1, 0); + parent, ref_root, level - 1, 0, + for_cow); if (ret) goto fail; } @@ -2659,15 +2665,15 @@ fail: } int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref) + struct extent_buffer *buf, int full_backref, int for_cow) { - return __btrfs_mod_ref(trans, root, buf, full_backref, 1); + return __btrfs_mod_ref(trans, root, buf, full_backref, 1, for_cow); } int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, int full_backref) + struct extent_buffer *buf, int full_backref, int for_cow) { - return __btrfs_mod_ref(trans, root, buf, full_backref, 0); + return __btrfs_mod_ref(trans, root, buf, full_backref, 0, for_cow); } static int write_one_cache_group(struct btrfs_trans_handle *trans, @@ -4937,16 +4943,17 @@ out: void btrfs_free_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, - u64 parent, int last_ref) + u64 parent, int last_ref, int for_cow) { struct btrfs_block_group_cache *cache = NULL; int ret; if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) { - ret = btrfs_add_delayed_tree_ref(trans, buf->start, buf->len, - parent, root->root_key.objectid, - btrfs_header_level(buf), - BTRFS_DROP_DELAYED_REF, NULL); + ret = btrfs_add_delayed_tree_ref(root->fs_info, trans, + buf->start, buf->len, + parent, root->root_key.objectid, + btrfs_header_level(buf), + BTRFS_DROP_DELAYED_REF, NULL, for_cow); BUG_ON(ret); } @@ -4981,12 +4988,12 @@ out: btrfs_put_block_group(cache); } -int btrfs_free_extent(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - u64 bytenr, u64 num_bytes, u64 parent, - u64 root_objectid, u64 owner, u64 offset) +int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, + u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid, + u64 owner, u64 offset, int for_cow) { int ret; + struct btrfs_fs_info *fs_info = root->fs_info; /* * tree log blocks never actually go into the extent allocation @@ -4998,14 +5005,17 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, btrfs_pin_extent(root, bytenr, num_bytes, 1); ret = 0; } else if (owner < BTRFS_FIRST_FREE_OBJECTID) { - ret = btrfs_add_delayed_tree_ref(trans, bytenr, num_bytes, + ret = btrfs_add_delayed_tree_ref(fs_info, trans, bytenr, + num_bytes, parent, root_objectid, (int)owner, - BTRFS_DROP_DELAYED_REF, NULL); + BTRFS_DROP_DELAYED_REF, NULL, for_cow); BUG_ON(ret); } else { - ret = btrfs_add_delayed_data_ref(trans, bytenr, num_bytes, - parent, root_objectid, owner, - offset, BTRFS_DROP_DELAYED_REF, NULL); + ret = btrfs_add_delayed_data_ref(fs_info, trans, bytenr, + num_bytes, + parent, root_objectid, owner, + offset, BTRFS_DROP_DELAYED_REF, + NULL, for_cow); BUG_ON(ret); } return ret; @@ -5826,9 +5836,10 @@ int btrfs_alloc_reserved_file_extent(struct btrfs_trans_handle *trans, BUG_ON(root_objectid == BTRFS_TREE_LOG_OBJECTID); - ret = btrfs_add_delayed_data_ref(trans, ins->objectid, ins->offset, - 0, root_objectid, owner, offset, - BTRFS_ADD_DELAYED_EXTENT, NULL); + ret = btrfs_add_delayed_data_ref(root->fs_info, trans, ins->objectid, + ins->offset, 0, + root_objectid, owner, offset, + BTRFS_ADD_DELAYED_EXTENT, NULL, 0); return ret; } @@ -5998,7 +6009,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, u32 blocksize, u64 parent, u64 root_objectid, struct btrfs_disk_key *key, int level, - u64 hint, u64 empty_size) + u64 hint, u64 empty_size, int for_cow) { struct btrfs_key ins; struct btrfs_block_rsv *block_rsv; @@ -6042,10 +6053,11 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, extent_op->update_flags = 1; extent_op->is_data = 0; - ret = btrfs_add_delayed_tree_ref(trans, ins.objectid, + ret = btrfs_add_delayed_tree_ref(root->fs_info, trans, + ins.objectid, ins.offset, parent, root_objectid, level, BTRFS_ADD_DELAYED_EXTENT, - extent_op); + extent_op, for_cow); BUG_ON(ret); } return buf; @@ -6062,6 +6074,7 @@ struct walk_control { int keep_locks; int reada_slot; int reada_count; + int for_reloc; }; #define DROP_REFERENCE 1 @@ -6200,9 +6213,9 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans, /* wc->stage == UPDATE_BACKREF */ if (!(wc->flags[level] & flag)) { BUG_ON(!path->locks[level]); - ret = btrfs_inc_ref(trans, root, eb, 1); + ret = btrfs_inc_ref(trans, root, eb, 1, wc->for_reloc); BUG_ON(ret); - ret = btrfs_dec_ref(trans, root, eb, 0); + ret = btrfs_dec_ref(trans, root, eb, 0, wc->for_reloc); BUG_ON(ret); ret = btrfs_set_disk_extent_flags(trans, root, eb->start, eb->len, flag, 0); @@ -6346,7 +6359,7 @@ skip: } ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent, - root->root_key.objectid, level - 1, 0); + root->root_key.objectid, level - 1, 0, 0); BUG_ON(ret); } btrfs_tree_unlock(next); @@ -6420,9 +6433,11 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans, if (wc->refs[level] == 1) { if (level == 0) { if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) - ret = btrfs_dec_ref(trans, root, eb, 1); + ret = btrfs_dec_ref(trans, root, eb, 1, + wc->for_reloc); else - ret = btrfs_dec_ref(trans, root, eb, 0); + ret = btrfs_dec_ref(trans, root, eb, 0, + wc->for_reloc); BUG_ON(ret); } /* make block locked assertion in clean_tree_block happy */ @@ -6449,7 +6464,7 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans, btrfs_header_owner(path->nodes[level + 1])); } - btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1); + btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1, 0); out: wc->refs[level] = 0; wc->flags[level] = 0; @@ -6533,7 +6548,8 @@ static noinline int walk_up_tree(struct btrfs_trans_handle *trans, * blocks are properly updated. */ void btrfs_drop_snapshot(struct btrfs_root *root, - struct btrfs_block_rsv *block_rsv, int update_ref) + struct btrfs_block_rsv *block_rsv, int update_ref, + int for_reloc) { struct btrfs_path *path; struct btrfs_trans_handle *trans; @@ -6621,6 +6637,7 @@ void btrfs_drop_snapshot(struct btrfs_root *root, wc->stage = DROP_REFERENCE; wc->update_ref = update_ref; wc->keep_locks = 0; + wc->for_reloc = for_reloc; wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(root); while (1) { @@ -6705,6 +6722,7 @@ out: * drop subtree rooted at tree block 'node'. * * NOTE: this function will unlock and release tree block 'node' + * only used by relocation code */ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -6749,6 +6767,7 @@ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, wc->stage = DROP_REFERENCE; wc->update_ref = 0; wc->keep_locks = 1; + wc->for_reloc = 1; wc->reada_count = BTRFS_NODEPTRS_PER_BLOCK(root); while (1) { diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index f2e928289600..d2b60ed6c33f 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -678,7 +678,7 @@ next_slot: disk_bytenr, num_bytes, 0, root->root_key.objectid, new_key.objectid, - start - extent_offset); + start - extent_offset, 0); BUG_ON(ret); *hint_byte = disk_bytenr; } @@ -753,7 +753,7 @@ next_slot: disk_bytenr, num_bytes, 0, root->root_key.objectid, key.objectid, key.offset - - extent_offset); + extent_offset, 0); BUG_ON(ret); inode_sub_bytes(inode, extent_end - key.offset); @@ -962,7 +962,7 @@ again: ret = btrfs_inc_extent_ref(trans, root, bytenr, num_bytes, 0, root->root_key.objectid, - ino, orig_offset); + ino, orig_offset, 0); BUG_ON(ret); if (split == start) { @@ -989,7 +989,7 @@ again: del_nr++; ret = btrfs_free_extent(trans, root, bytenr, num_bytes, 0, root->root_key.objectid, - ino, orig_offset); + ino, orig_offset, 0); BUG_ON(ret); } other_start = 0; @@ -1006,7 +1006,7 @@ again: del_nr++; ret = btrfs_free_extent(trans, root, bytenr, num_bytes, 0, root->root_key.objectid, - ino, orig_offset); + ino, orig_offset, 0); BUG_ON(ret); } if (del_nr == 0) { diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index c5ccec23984c..ea819386b864 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3139,7 +3139,7 @@ delete: ret = btrfs_free_extent(trans, root, extent_start, extent_num_bytes, 0, btrfs_header_owner(leaf), - ino, extent_offset); + ino, extent_offset, 0); BUG_ON(ret); } diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 72d461656f60..c48f2e931ea0 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -358,7 +358,7 @@ static noinline int create_subvol(struct btrfs_root *root, return PTR_ERR(trans); leaf = btrfs_alloc_free_block(trans, root, root->leafsize, - 0, objectid, NULL, 0, 0, 0); + 0, objectid, NULL, 0, 0, 0, 0); if (IS_ERR(leaf)) { ret = PTR_ERR(leaf); goto fail; @@ -2425,7 +2425,8 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, disko, diskl, 0, root->root_key.objectid, btrfs_ino(inode), - new_key.offset - datao); + new_key.offset - datao, + 0); BUG_ON(ret); } } else if (type == BTRFS_FILE_EXTENT_INLINE) { diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index dff29d5e151a..8c1aae2c845d 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1604,12 +1604,12 @@ int replace_file_extents(struct btrfs_trans_handle *trans, ret = btrfs_inc_extent_ref(trans, root, new_bytenr, num_bytes, parent, btrfs_header_owner(leaf), - key.objectid, key.offset); + key.objectid, key.offset, 1); BUG_ON(ret); ret = btrfs_free_extent(trans, root, bytenr, num_bytes, parent, btrfs_header_owner(leaf), - key.objectid, key.offset); + key.objectid, key.offset, 1); BUG_ON(ret); } if (dirty) @@ -1778,21 +1778,23 @@ again: ret = btrfs_inc_extent_ref(trans, src, old_bytenr, blocksize, path->nodes[level]->start, - src->root_key.objectid, level - 1, 0); + src->root_key.objectid, level - 1, 0, + 1); BUG_ON(ret); ret = btrfs_inc_extent_ref(trans, dest, new_bytenr, blocksize, 0, dest->root_key.objectid, level - 1, - 0); + 0, 1); BUG_ON(ret); ret = btrfs_free_extent(trans, src, new_bytenr, blocksize, path->nodes[level]->start, - src->root_key.objectid, level - 1, 0); + src->root_key.objectid, level - 1, 0, + 1); BUG_ON(ret); ret = btrfs_free_extent(trans, dest, old_bytenr, blocksize, 0, dest->root_key.objectid, level - 1, - 0); + 0, 1); BUG_ON(ret); btrfs_unlock_up_safe(path, 0); @@ -2244,7 +2246,7 @@ again: } else { list_del_init(&reloc_root->root_list); } - btrfs_drop_snapshot(reloc_root, rc->block_rsv, 0); + btrfs_drop_snapshot(reloc_root, rc->block_rsv, 0, 1); } if (found) { @@ -2558,7 +2560,7 @@ static int do_relocation(struct btrfs_trans_handle *trans, node->eb->start, blocksize, upper->eb->start, btrfs_header_owner(upper->eb), - node->level, 0); + node->level, 0, 1); BUG_ON(ret); ret = btrfs_drop_subtree(trans, root, eb, upper->eb); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 81376d94cd3c..a2bfedcbcabc 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1393,9 +1393,9 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root) if (btrfs_header_backref_rev(root->node) < BTRFS_MIXED_BACKREF_REV) - btrfs_drop_snapshot(root, NULL, 0); + btrfs_drop_snapshot(root, NULL, 0, 0); else - btrfs_drop_snapshot(root, NULL, 1); + btrfs_drop_snapshot(root, NULL, 1, 0); } return 0; } diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index f4d81c06d48f..fce7b9ec3bfa 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -589,7 +589,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ret = btrfs_inc_extent_ref(trans, root, ins.objectid, ins.offset, 0, root->root_key.objectid, - key->objectid, offset); + key->objectid, offset, 0); BUG_ON(ret); } else { /* From eebe063b7f916087cd5c61de57b20a3a30894a96 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Wed, 14 Sep 2011 14:01:24 +0200 Subject: [PATCH 172/820] Btrfs: always save ref_root in delayed refs For consistent backref walking and (later) qgroup calculation the information to which root a delayed ref belongs is useful even for shared refs. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/delayed-ref.c | 18 ++++++++---------- fs/btrfs/delayed-ref.h | 12 ++++-------- 2 files changed, 12 insertions(+), 18 deletions(-) diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index 3a0f0ab804f4..babd37badb43 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -495,13 +495,12 @@ static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info, ref->in_tree = 1; full_ref = btrfs_delayed_node_to_tree_ref(ref); - if (parent) { - full_ref->parent = parent; + full_ref->parent = parent; + full_ref->root = ref_root; + if (parent) ref->type = BTRFS_SHARED_BLOCK_REF_KEY; - } else { - full_ref->root = ref_root; + else ref->type = BTRFS_TREE_BLOCK_REF_KEY; - } full_ref->level = level; trace_btrfs_delayed_tree_ref(ref, full_ref, action); @@ -551,13 +550,12 @@ static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info, ref->in_tree = 1; full_ref = btrfs_delayed_node_to_data_ref(ref); - if (parent) { - full_ref->parent = parent; + full_ref->parent = parent; + full_ref->root = ref_root; + if (parent) ref->type = BTRFS_SHARED_DATA_REF_KEY; - } else { - full_ref->root = ref_root; + else ref->type = BTRFS_EXTENT_DATA_REF_KEY; - } full_ref->objectid = owner; full_ref->offset = offset; diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index 8316bff18d30..a5fb2bc83732 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -98,19 +98,15 @@ struct btrfs_delayed_ref_head { struct btrfs_delayed_tree_ref { struct btrfs_delayed_ref_node node; - union { - u64 root; - u64 parent; - }; + u64 root; + u64 parent; int level; }; struct btrfs_delayed_data_ref { struct btrfs_delayed_ref_node node; - union { - u64 root; - u64 parent; - }; + u64 root; + u64 parent; u64 objectid; u64 offset; }; From 400312201b0cf4e4deaf75842f5e95212b382e81 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Lothar=20Wa=C3=9Fmann?= Date: Thu, 8 Dec 2011 09:15:41 +0100 Subject: [PATCH 173/820] dma: mxs-dma: fix a typo in comment MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Lothar Waßmann Signed-off-by: Vinod Koul --- drivers/dma/mxs-dma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c index bdf4672b2553..d04662e94467 100644 --- a/drivers/dma/mxs-dma.c +++ b/drivers/dma/mxs-dma.c @@ -268,7 +268,7 @@ static irqreturn_t mxs_dma_int_handler(int irq, void *dev_id) /* * When both completion and error of termination bits set at the * same time, we do not take it as an error. IOW, it only becomes - * an error we need to handler here in case of ether it's (1) an bus + * an error we need to handle here in case of either it's (1) a bus * error or (2) a termination error with no completion. */ stat2 = ((stat2 >> MXS_DMA_CHANNELS) & stat2) | /* (1) */ From feb397de65c3f76e40ef70e264f2cdf688c850c1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Lothar=20Wa=C3=9Fmann?= Date: Thu, 8 Dec 2011 09:15:42 +0100 Subject: [PATCH 174/820] dma: mxs-dma: Always leave mxs_dma_init() with the clock disabled. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is no need to have the clock enabled all the time the driver is loaded. It will be enabled anyway in mxs_dma_alloc_chan_resources() when a channel is actually going to be used. Signed-off-by: Lothar Waßmann Signed-off-by: Vinod Koul --- drivers/dma/mxs-dma.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c index d04662e94467..b0e6ac3e09aa 100644 --- a/drivers/dma/mxs-dma.c +++ b/drivers/dma/mxs-dma.c @@ -580,7 +580,7 @@ static int __init mxs_dma_init(struct mxs_dma_engine *mxs_dma) ret = clk_enable(mxs_dma->clk); if (ret) - goto err_out; + return ret; ret = mxs_reset_block(mxs_dma->base); if (ret) @@ -604,11 +604,8 @@ static int __init mxs_dma_init(struct mxs_dma_engine *mxs_dma) writel(MXS_DMA_CHANNELS_MASK << MXS_DMA_CHANNELS, mxs_dma->base + HW_APBHX_CTRL1 + MXS_SET_ADDR); - clk_disable(mxs_dma->clk); - - return 0; - err_out: + clk_disable(mxs_dma->clk); return ret; } From 6d23ea4b1906f28f5d99ad6aeef7207c48be6bfd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Lothar=20Wa=C3=9Fmann?= Date: Thu, 8 Dec 2011 09:15:43 +0100 Subject: [PATCH 175/820] dma: mxs-dma: make mxs_dma_prep_slave_sg() multi user safe MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Using a static variable for counting the number of CCWs attached to a DMA channel when appending a new descriptor is not multi user safe. Signed-off-by: Lothar Waßmann Signed-off-by: Vinod Koul --- drivers/dma/mxs-dma.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c index b0e6ac3e09aa..1b4c6be3aacb 100644 --- a/drivers/dma/mxs-dma.c +++ b/drivers/dma/mxs-dma.c @@ -111,6 +111,7 @@ struct mxs_dma_chan { int chan_irq; struct mxs_dma_ccw *ccw; dma_addr_t ccw_phys; + int desc_count; dma_cookie_t last_completed; enum dma_status status; unsigned int flags; @@ -386,7 +387,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg( struct scatterlist *sg; int i, j; u32 *pio; - static int idx; + int idx = append ? mxs_chan->desc_count : 0; if (mxs_chan->status == DMA_IN_PROGRESS && !append) return NULL; @@ -462,6 +463,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg( } } } + mxs_chan->desc_count = idx; return &mxs_chan->desc; @@ -523,6 +525,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_dma_cyclic( i++; } + mxs_chan->desc_count = i; return &mxs_chan->desc; From 7ad7a345a4f17c08a1bb9bfdbb62f7793d84aa36 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Lothar=20Wa=C3=9Fmann?= Date: Thu, 8 Dec 2011 09:15:44 +0100 Subject: [PATCH 176/820] dma: mxs-dma: Don't use CLKGATE bits in CTRL0 to disable DMA channels MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This is how the original Freescale code (unintentionally) worked, because the code path which would have asserted the CLKGATE bit was never actually reached in their code. This fixes the nefarious "DMA timout" bug when multiple DMA channels (e.g. GPMI NAND and MMC) are used at the same time. If a better fix for this problem should be found, the clkgate handling could be reinstated. See http://lists.infradead.org/pipermail/linux-arm-kernel/2011-September/065228.html Also reverse the order of mxs_dma_disable_chan() and mxs_dma_reset_chan() in mxs_dma_control() because mxs_dma_reset_chan() can only work when the DMA channel is enabled. Signed-off-by: Lothar Waßmann Signed-off-by: Vinod Koul --- drivers/dma/mxs-dma.c | 29 +---------------------------- 1 file changed, 1 insertion(+), 28 deletions(-) diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c index 1b4c6be3aacb..6548595c26dc 100644 --- a/drivers/dma/mxs-dma.c +++ b/drivers/dma/mxs-dma.c @@ -44,7 +44,6 @@ #define HW_APBHX_CTRL0 0x000 #define BM_APBH_CTRL0_APB_BURST8_EN (1 << 29) #define BM_APBH_CTRL0_APB_BURST_EN (1 << 28) -#define BP_APBH_CTRL0_CLKGATE_CHANNEL 8 #define BP_APBH_CTRL0_RESET_CHANNEL 16 #define HW_APBHX_CTRL1 0x010 #define HW_APBHX_CTRL2 0x020 @@ -131,23 +130,6 @@ struct mxs_dma_engine { struct mxs_dma_chan mxs_chans[MXS_DMA_CHANNELS]; }; -static inline void mxs_dma_clkgate(struct mxs_dma_chan *mxs_chan, int enable) -{ - struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma; - int chan_id = mxs_chan->chan.chan_id; - int set_clr = enable ? MXS_CLR_ADDR : MXS_SET_ADDR; - - /* enable apbh channel clock */ - if (dma_is_apbh()) { - if (apbh_is_old()) - writel(1 << (chan_id + BP_APBH_CTRL0_CLKGATE_CHANNEL), - mxs_dma->base + HW_APBHX_CTRL0 + set_clr); - else - writel(1 << chan_id, - mxs_dma->base + HW_APBHX_CTRL0 + set_clr); - } -} - static void mxs_dma_reset_chan(struct mxs_dma_chan *mxs_chan) { struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma; @@ -166,9 +148,6 @@ static void mxs_dma_enable_chan(struct mxs_dma_chan *mxs_chan) struct mxs_dma_engine *mxs_dma = mxs_chan->mxs_dma; int chan_id = mxs_chan->chan.chan_id; - /* clkgate needs to be enabled before writing other registers */ - mxs_dma_clkgate(mxs_chan, 1); - /* set cmd_addr up */ writel(mxs_chan->ccw_phys, mxs_dma->base + HW_APBHX_CHn_NXTCMDAR(chan_id)); @@ -179,9 +158,6 @@ static void mxs_dma_enable_chan(struct mxs_dma_chan *mxs_chan) static void mxs_dma_disable_chan(struct mxs_dma_chan *mxs_chan) { - /* disable apbh channel clock */ - mxs_dma_clkgate(mxs_chan, 0); - mxs_chan->status = DMA_SUCCESS; } @@ -339,10 +315,7 @@ static int mxs_dma_alloc_chan_resources(struct dma_chan *chan) if (ret) goto err_clk; - /* clkgate needs to be enabled for reset to finish */ - mxs_dma_clkgate(mxs_chan, 1); mxs_dma_reset_chan(mxs_chan); - mxs_dma_clkgate(mxs_chan, 0); dma_async_tx_descriptor_init(&mxs_chan->desc, chan); mxs_chan->desc.tx_submit = mxs_dma_tx_submit; @@ -542,8 +515,8 @@ static int mxs_dma_control(struct dma_chan *chan, enum dma_ctrl_cmd cmd, switch (cmd) { case DMA_TERMINATE_ALL: - mxs_dma_disable_chan(mxs_chan); mxs_dma_reset_chan(mxs_chan); + mxs_dma_disable_chan(mxs_chan); break; case DMA_PAUSE: mxs_dma_pause_chan(mxs_chan); From 62268ce9170c5466332c046ff6ddafcb67751502 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Tue, 13 Dec 2011 23:48:03 +0800 Subject: [PATCH 177/820] dmaengine: add DMA_TRANS_NONE to dma_transfer_direction MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Before dma_transfer_direction was introduced to replace dma_data_direction, some dmaengine device uses DMA_NONE of dma_data_direction for some talk with its client drivers. The mxs-dma and its clients mxs-mmc and gpmi-nand are such case. This patch adds DMA_TRANS_NONE to dma_transfer_direction and migrate the DMA_NONE use in mxs-dma to it. It also fixes the compile warning below. CC drivers/dma/mxs-dma.o drivers/dma/mxs-dma.c: In function ‘mxs_dma_prep_slave_sg’: drivers/dma/mxs-dma.c:420:16: warning: comparison between ‘enum dma_transfer_direction’ and ‘enum dma_data_direction’ Signed-off-by: Shawn Guo Signed-off-by: Vinod Koul --- drivers/dma/mxs-dma.c | 2 +- include/linux/dmaengine.h | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/dma/mxs-dma.c b/drivers/dma/mxs-dma.c index 6548595c26dc..493af2f6e33a 100644 --- a/drivers/dma/mxs-dma.c +++ b/drivers/dma/mxs-dma.c @@ -391,7 +391,7 @@ static struct dma_async_tx_descriptor *mxs_dma_prep_slave_sg( idx = 0; } - if (direction == DMA_NONE) { + if (direction == DMA_TRANS_NONE) { ccw = &mxs_chan->ccw[idx++]; pio = (u32 *) sgl; diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h index 5532bb8b500c..679b349d9b66 100644 --- a/include/linux/dmaengine.h +++ b/include/linux/dmaengine.h @@ -88,6 +88,7 @@ enum dma_transfer_direction { DMA_MEM_TO_DEV, DMA_DEV_TO_MEM, DMA_DEV_TO_DEV, + DMA_TRANS_NONE, }; /** From a4e3e86d66ffc640ac54acffa29450e327a80ae2 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Tue, 13 Dec 2011 23:48:04 +0800 Subject: [PATCH 178/820] mmc: mxs-mmc: fix the dma_transfer_direction migration The commit 05f5799 (mmc-host: move to dma_transfer_direction) left out the DMA_NONE, in turn breaks the driver as below. [ 0.650000] mxs-mmc mxs-mmc.0: initialized [ 0.650000] mxs-mmc mxs-mmc.1: initialized [ 0.690000] mxs-dma mxs-dma-apbh: maximum bytes for sg entry exceeded: -55906 7475 > 65280 [ 0.690000] mxs-mmc mxs-mmc.0: mxs_mmc_ac: failed to prep dma Signed-off-by: Shawn Guo Signed-off-by: Vinod Koul --- drivers/mmc/host/mxs-mmc.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/mmc/host/mxs-mmc.c b/drivers/mmc/host/mxs-mmc.c index 0a878b181341..94d7358187fd 100644 --- a/drivers/mmc/host/mxs-mmc.c +++ b/drivers/mmc/host/mxs-mmc.c @@ -357,6 +357,7 @@ static void mxs_mmc_bc(struct mxs_mmc_host *host) host->ssp_pio_words[1] = cmd0; host->ssp_pio_words[2] = cmd1; host->dma_dir = DMA_NONE; + host->slave_dirn = DMA_TRANS_NONE; desc = mxs_mmc_prep_dma(host, 0); if (!desc) goto out; @@ -396,6 +397,7 @@ static void mxs_mmc_ac(struct mxs_mmc_host *host) host->ssp_pio_words[1] = cmd0; host->ssp_pio_words[2] = cmd1; host->dma_dir = DMA_NONE; + host->slave_dirn = DMA_TRANS_NONE; desc = mxs_mmc_prep_dma(host, 0); if (!desc) goto out; @@ -514,6 +516,7 @@ static void mxs_mmc_adtc(struct mxs_mmc_host *host) host->ssp_pio_words[1] = cmd0; host->ssp_pio_words[2] = cmd1; host->dma_dir = DMA_NONE; + host->slave_dirn = DMA_TRANS_NONE; desc = mxs_mmc_prep_dma(host, 0); if (!desc) goto out; From 4e82786f7039cb707c031dcf0dc84c025e149487 Mon Sep 17 00:00:00 2001 From: Huang Shijie Date: Tue, 13 Dec 2011 23:48:05 +0800 Subject: [PATCH 179/820] mtd: fix compile error for gpmi-nand The driver gpmi-nand should compile at least. This patch adds the missing gpmi-nand.h to fix the compile error below. CC drivers/mtd/nand/gpmi-nand/gpmi-nand.o CC drivers/mtd/nand/gpmi-nand/gpmi-lib.o drivers/mtd/nand/gpmi-nand/gpmi-nand.c:25:33: fatal error: linux/mtd/gpmi-nand.h: No such file or directory drivers/mtd/nand/gpmi-nand/gpmi-lib.c:21:33: fatal error: linux/mtd/gpmi-nand.h: No such file or directory This header is grabbed from patch below, which has not been postponed for merging. [PATCH v8 1/4] ARM: mxs: add GPMI-NAND support for imx23/imx28 http://permalink.gmane.org/gmane.linux.drivers.mtd/37338 Signed-off-by: Huang Shijie Signed-off-by: Shawn Guo Signed-off-by: Vinod Koul --- include/linux/mtd/gpmi-nand.h | 68 +++++++++++++++++++++++++++++++++++ 1 file changed, 68 insertions(+) create mode 100644 include/linux/mtd/gpmi-nand.h diff --git a/include/linux/mtd/gpmi-nand.h b/include/linux/mtd/gpmi-nand.h new file mode 100644 index 000000000000..69b6dbf46b5e --- /dev/null +++ b/include/linux/mtd/gpmi-nand.h @@ -0,0 +1,68 @@ +/* + * Copyright (C) 2011 Freescale Semiconductor, Inc. All Rights Reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + */ + +#ifndef __MACH_MXS_GPMI_NAND_H__ +#define __MACH_MXS_GPMI_NAND_H__ + +/* The size of the resources is fixed. */ +#define GPMI_NAND_RES_SIZE 6 + +/* Resource names for the GPMI NAND driver. */ +#define GPMI_NAND_GPMI_REGS_ADDR_RES_NAME "GPMI NAND GPMI Registers" +#define GPMI_NAND_GPMI_INTERRUPT_RES_NAME "GPMI NAND GPMI Interrupt" +#define GPMI_NAND_BCH_REGS_ADDR_RES_NAME "GPMI NAND BCH Registers" +#define GPMI_NAND_BCH_INTERRUPT_RES_NAME "GPMI NAND BCH Interrupt" +#define GPMI_NAND_DMA_CHANNELS_RES_NAME "GPMI NAND DMA Channels" +#define GPMI_NAND_DMA_INTERRUPT_RES_NAME "GPMI NAND DMA Interrupt" + +/** + * struct gpmi_nand_platform_data - GPMI NAND driver platform data. + * + * This structure communicates platform-specific information to the GPMI NAND + * driver that can't be expressed as resources. + * + * @platform_init: A pointer to a function the driver will call to + * initialize the platform (e.g., set up the pin mux). + * @min_prop_delay_in_ns: Minimum propagation delay of GPMI signals to and + * from the NAND Flash device, in nanoseconds. + * @max_prop_delay_in_ns: Maximum propagation delay of GPMI signals to and + * from the NAND Flash device, in nanoseconds. + * @max_chip_count: The maximum number of chips for which the driver + * should configure the hardware. This value most + * likely reflects the number of pins that are + * connected to a NAND Flash device. If this is + * greater than the SoC hardware can support, the + * driver will print a message and fail to initialize. + * @partitions: An optional pointer to an array of partition + * descriptions. + * @partition_count: The number of elements in the partitions array. + */ +struct gpmi_nand_platform_data { + /* SoC hardware information. */ + int (*platform_init)(void); + + /* NAND Flash information. */ + unsigned int min_prop_delay_in_ns; + unsigned int max_prop_delay_in_ns; + unsigned int max_chip_count; + + /* Medium information. */ + struct mtd_partition *partitions; + unsigned partition_count; +}; +#endif From 0ef7e206d6a5a5de0cd84ed4925a4f688c62e732 Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Tue, 13 Dec 2011 23:48:06 +0800 Subject: [PATCH 180/820] mtd: gpmi-nand: move to dma_transfer_direction This patch fixes usage of dma direction to adopt dma_transfer_direction. Signed-off-by: Shawn Guo Signed-off-by: Vinod Koul --- drivers/mtd/nand/gpmi-nand/gpmi-lib.c | 22 ++++++++++++---------- 1 file changed, 12 insertions(+), 10 deletions(-) diff --git a/drivers/mtd/nand/gpmi-nand/gpmi-lib.c b/drivers/mtd/nand/gpmi-nand/gpmi-lib.c index de4db7604a3f..c4c4d6d89cd5 100644 --- a/drivers/mtd/nand/gpmi-nand/gpmi-lib.c +++ b/drivers/mtd/nand/gpmi-nand/gpmi-lib.c @@ -827,7 +827,7 @@ int gpmi_send_command(struct gpmi_nand_data *this) pio[1] = pio[2] = 0; desc = channel->device->device_prep_slave_sg(channel, (struct scatterlist *)pio, - ARRAY_SIZE(pio), DMA_NONE, 0); + ARRAY_SIZE(pio), DMA_TRANS_NONE, 0); if (!desc) { pr_err("step 1 error\n"); return -1; @@ -839,7 +839,7 @@ int gpmi_send_command(struct gpmi_nand_data *this) sg_init_one(sgl, this->cmd_buffer, this->command_length); dma_map_sg(this->dev, sgl, 1, DMA_TO_DEVICE); desc = channel->device->device_prep_slave_sg(channel, - sgl, 1, DMA_TO_DEVICE, 1); + sgl, 1, DMA_MEM_TO_DEV, 1); if (!desc) { pr_err("step 2 error\n"); return -1; @@ -872,7 +872,7 @@ int gpmi_send_data(struct gpmi_nand_data *this) pio[1] = 0; desc = channel->device->device_prep_slave_sg(channel, (struct scatterlist *)pio, - ARRAY_SIZE(pio), DMA_NONE, 0); + ARRAY_SIZE(pio), DMA_TRANS_NONE, 0); if (!desc) { pr_err("step 1 error\n"); return -1; @@ -881,7 +881,7 @@ int gpmi_send_data(struct gpmi_nand_data *this) /* [2] send DMA request */ prepare_data_dma(this, DMA_TO_DEVICE); desc = channel->device->device_prep_slave_sg(channel, &this->data_sgl, - 1, DMA_TO_DEVICE, 1); + 1, DMA_MEM_TO_DEV, 1); if (!desc) { pr_err("step 2 error\n"); return -1; @@ -908,7 +908,7 @@ int gpmi_read_data(struct gpmi_nand_data *this) pio[1] = 0; desc = channel->device->device_prep_slave_sg(channel, (struct scatterlist *)pio, - ARRAY_SIZE(pio), DMA_NONE, 0); + ARRAY_SIZE(pio), DMA_TRANS_NONE, 0); if (!desc) { pr_err("step 1 error\n"); return -1; @@ -917,7 +917,7 @@ int gpmi_read_data(struct gpmi_nand_data *this) /* [2] : send DMA request */ prepare_data_dma(this, DMA_FROM_DEVICE); desc = channel->device->device_prep_slave_sg(channel, &this->data_sgl, - 1, DMA_FROM_DEVICE, 1); + 1, DMA_DEV_TO_MEM, 1); if (!desc) { pr_err("step 2 error\n"); return -1; @@ -964,7 +964,7 @@ int gpmi_send_page(struct gpmi_nand_data *this, desc = channel->device->device_prep_slave_sg(channel, (struct scatterlist *)pio, - ARRAY_SIZE(pio), DMA_NONE, 0); + ARRAY_SIZE(pio), DMA_TRANS_NONE, 0); if (!desc) { pr_err("step 2 error\n"); return -1; @@ -998,7 +998,8 @@ int gpmi_read_page(struct gpmi_nand_data *this, | BF_GPMI_CTRL0_XFER_COUNT(0); pio[1] = 0; desc = channel->device->device_prep_slave_sg(channel, - (struct scatterlist *)pio, 2, DMA_NONE, 0); + (struct scatterlist *)pio, 2, + DMA_TRANS_NONE, 0); if (!desc) { pr_err("step 1 error\n"); return -1; @@ -1027,7 +1028,7 @@ int gpmi_read_page(struct gpmi_nand_data *this, pio[5] = auxiliary; desc = channel->device->device_prep_slave_sg(channel, (struct scatterlist *)pio, - ARRAY_SIZE(pio), DMA_NONE, 1); + ARRAY_SIZE(pio), DMA_TRANS_NONE, 1); if (!desc) { pr_err("step 2 error\n"); return -1; @@ -1045,7 +1046,8 @@ int gpmi_read_page(struct gpmi_nand_data *this, | BF_GPMI_CTRL0_XFER_COUNT(geo->page_size); pio[1] = 0; desc = channel->device->device_prep_slave_sg(channel, - (struct scatterlist *)pio, 2, DMA_NONE, 1); + (struct scatterlist *)pio, 2, + DMA_TRANS_NONE, 1); if (!desc) { pr_err("step 3 error\n"); return -1; From 1fded07513ea57b5ee128958ff119e05588b7227 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 16 Dec 2011 11:01:38 +0200 Subject: [PATCH 181/820] dmaengine: intel_mid_dma: locking and freeing fixes Two issues are fixed: 1. DMA descriptors are reused so when freeing lli structures that are linked to them, the pointer must be nulled. 2. midc_scan_descriptors() must be called with the channel lock held. Signed-off-by: Adrian Hunter Signed-off-by: Vinod Koul --- drivers/dma/intel_mid_dma.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c index 01929ed6659d..d4b961677e02 100644 --- a/drivers/dma/intel_mid_dma.c +++ b/drivers/dma/intel_mid_dma.c @@ -280,7 +280,8 @@ static void midc_dostart(struct intel_mid_dma_chan *midc, * callbacks but must be called with the lock held. */ static void midc_descriptor_complete(struct intel_mid_dma_chan *midc, - struct intel_mid_dma_desc *desc) + struct intel_mid_dma_desc *desc) + __releases(&midc->lock) __acquires(&midc->lock) { struct dma_async_tx_descriptor *txd = &desc->txd; dma_async_tx_callback callback_txd = NULL; @@ -311,6 +312,7 @@ static void midc_descriptor_complete(struct intel_mid_dma_chan *midc, pci_pool_free(desc->lli_pool, desc->lli, desc->lli_phys); pci_pool_destroy(desc->lli_pool); + desc->lli = NULL; } list_move(&desc->desc_node, &midc->free_list); midc->busy = false; @@ -490,7 +492,9 @@ static enum dma_status intel_mid_dma_tx_status(struct dma_chan *chan, ret = dma_async_is_complete(cookie, last_complete, last_used); if (ret != DMA_SUCCESS) { + spin_lock_bh(&midc->lock); midc_scan_descriptors(to_middma_device(chan->device), midc); + spin_unlock_bh(&midc->lock); last_complete = midc->completed; last_used = chan->cookie; @@ -566,6 +570,7 @@ static int intel_mid_dma_device_control(struct dma_chan *chan, pci_pool_free(desc->lli_pool, desc->lli, desc->lli_phys); pci_pool_destroy(desc->lli_pool); + desc->lli = NULL; } list_move(&desc->desc_node, &midc->free_list); } From 2a0ff7a602714990b10a8749e908dad48401138e Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Fri, 16 Dec 2011 11:01:39 +0200 Subject: [PATCH 182/820] dmaengine: intel_mid_dma: error path fix Add missing iounmap in error path. Signed-off-by: Adrian Hunter Signed-off-by: Vinod Koul --- drivers/dma/intel_mid_dma.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c index d4b961677e02..bbcd470c3236 100644 --- a/drivers/dma/intel_mid_dma.c +++ b/drivers/dma/intel_mid_dma.c @@ -1104,7 +1104,8 @@ static int mid_setup_dma(struct pci_dev *pdev) LNW_PERIPHRAL_MASK_SIZE); if (dma->mask_reg == NULL) { pr_err("ERR_MDMA:Can't map periphral intr space !!\n"); - return -ENOMEM; + err = -ENOMEM; + goto err_ioremap; } } else dma->mask_reg = NULL; @@ -1201,6 +1202,9 @@ static int mid_setup_dma(struct pci_dev *pdev) err_engine: free_irq(pdev->irq, dma); err_irq: + if (dma->mask_reg) + iounmap(dma->mask_reg); +err_ioremap: pci_pool_destroy(dma->dma_pool); err_dma_pool: pr_err("ERR_MDMA:setup_dma failed: %d\n", err); From d6077c92b6c51299eac7c2673631d6394ee2191a Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Wed, 14 Dec 2011 15:47:46 +0800 Subject: [PATCH 183/820] ASoC: mxs: correct 'direction' of device_prep_dma_cyclic The commit 49920bc (dmaengine: add new enum dma_transfer_direction) changes the type of parameter 'direction' of device_prep_dma_cyclic from dma_data_direction to dma_transfer_direction. Signed-off-by: Shawn Guo Signed-off-by: Vinod Koul --- sound/soc/mxs/mxs-pcm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/mxs/mxs-pcm.c b/sound/soc/mxs/mxs-pcm.c index dea5aa4aa647..4b48a169e89c 100644 --- a/sound/soc/mxs/mxs-pcm.c +++ b/sound/soc/mxs/mxs-pcm.c @@ -136,7 +136,7 @@ static int snd_mxs_pcm_hw_params(struct snd_pcm_substream *substream, iprtd->period_bytes * iprtd->periods, iprtd->period_bytes, substream->stream == SNDRV_PCM_STREAM_PLAYBACK ? - DMA_TO_DEVICE : DMA_FROM_DEVICE); + DMA_MEM_TO_DEV : DMA_DEV_TO_MEM); if (!iprtd->desc) { dev_err(&chan->dev->device, "cannot prepare slave dma\n"); return -EINVAL; From 8730790b971d747dd4508d22fbc1a4f9422c2803 Mon Sep 17 00:00:00 2001 From: Kristen Carlson Accardi Date: Fri, 16 Dec 2011 11:01:40 +0200 Subject: [PATCH 184/820] dmaengine: intel_mid_dma: remove legacy pm interface Drivers should not support both legacy pm and new framework. Signed-off-by: Kristen Carlson Accardi Signed-off-by: Vinod Koul --- drivers/dma/intel_mid_dma.c | 12 +++++++----- drivers/dma/intel_mid_dma_regs.h | 2 +- 2 files changed, 8 insertions(+), 6 deletions(-) diff --git a/drivers/dma/intel_mid_dma.c b/drivers/dma/intel_mid_dma.c index bbcd470c3236..74f70aadf9e4 100644 --- a/drivers/dma/intel_mid_dma.c +++ b/drivers/dma/intel_mid_dma.c @@ -873,7 +873,7 @@ static int intel_mid_dma_alloc_chan_resources(struct dma_chan *chan) pm_runtime_get_sync(&mid->pdev->dev); if (mid->state == SUSPENDED) { - if (dma_resume(mid->pdev)) { + if (dma_resume(&mid->pdev->dev)) { pr_err("ERR_MDMA: resume failed"); return -EFAULT; } @@ -1346,8 +1346,9 @@ static void __devexit intel_mid_dma_remove(struct pci_dev *pdev) * * This function is called by OS when a power event occurs */ -int dma_suspend(struct pci_dev *pci, pm_message_t state) +static int dma_suspend(struct device *dev) { + struct pci_dev *pci = to_pci_dev(dev); int i; struct middma_device *device = pci_get_drvdata(pci); pr_debug("MDMA: dma_suspend called\n"); @@ -1371,8 +1372,9 @@ int dma_suspend(struct pci_dev *pci, pm_message_t state) * * This function is called by OS when a power event occurs */ -int dma_resume(struct pci_dev *pci) +int dma_resume(struct device *dev) { + struct pci_dev *pci = to_pci_dev(dev); int ret; struct middma_device *device = pci_get_drvdata(pci); @@ -1438,6 +1440,8 @@ static const struct dev_pm_ops intel_mid_dma_pm = { .runtime_suspend = dma_runtime_suspend, .runtime_resume = dma_runtime_resume, .runtime_idle = dma_runtime_idle, + .suspend = dma_suspend, + .resume = dma_resume, }; static struct pci_driver intel_mid_dma_pci_driver = { @@ -1446,8 +1450,6 @@ static struct pci_driver intel_mid_dma_pci_driver = { .probe = intel_mid_dma_probe, .remove = __devexit_p(intel_mid_dma_remove), #ifdef CONFIG_PM - .suspend = dma_suspend, - .resume = dma_resume, .driver = { .pm = &intel_mid_dma_pm, }, diff --git a/drivers/dma/intel_mid_dma_regs.h b/drivers/dma/intel_mid_dma_regs.h index c6de919a6401..c83d35b97bd8 100644 --- a/drivers/dma/intel_mid_dma_regs.h +++ b/drivers/dma/intel_mid_dma_regs.h @@ -296,6 +296,6 @@ static inline struct intel_mid_dma_slave *to_intel_mid_dma_slave } -int dma_resume(struct pci_dev *pci); +int dma_resume(struct device *dev); #endif /*__INTEL_MID_DMAC_REGS_H__*/ From e9f3a49c9698da6f6b606d5e048ddce308438763 Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 28 Dec 2011 11:27:40 +0530 Subject: [PATCH 185/820] dmaengine/ste_dma40: clear LNK on channel startup Otherwise if a previously physical channel is used as a logical channel, the LNK may have old values which affect the operation of the logical channel since the LNK register has different semantics between physical and logical channels. Signed-off-by: Rabin Vincent Reviewed-by: Jonas ABERG Signed-off-by: Narayanan G Acked-by: Linus Walleij Signed-off-by: Vinod Koul --- drivers/dma/ste_dma40.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c index 972dc35770f9..cc5ecbc067a3 100644 --- a/drivers/dma/ste_dma40.c +++ b/drivers/dma/ste_dma40.c @@ -1104,6 +1104,10 @@ static void d40_config_write(struct d40_chan *d40c) /* Set LIDX for lcla */ writel(lidx, chanbase + D40_CHAN_REG_SSELT); writel(lidx, chanbase + D40_CHAN_REG_SDELT); + + /* Clear LNK which will be used by d40_chan_has_events() */ + writel(0, chanbase + D40_CHAN_REG_SSLNK); + writel(0, chanbase + D40_CHAN_REG_SDLNK); } } From 5b25f70f4200766355cdabda604e131d2fb6010d Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 13 Sep 2011 10:55:48 +0200 Subject: [PATCH 186/820] Btrfs: add nested locking mode for paths This patch adds the possibilty to read-lock an extent even if it is already write-locked from the same thread. btrfs_find_all_roots() needs this capability. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/extent_io.c | 1 + fs/btrfs/extent_io.h | 2 ++ fs/btrfs/locking.c | 53 ++++++++++++++++++++++++++++++++++++++++++-- 3 files changed, 54 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index be1bf627a14b..dd8d140eb27b 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3571,6 +3571,7 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree, atomic_set(&eb->blocking_writers, 0); atomic_set(&eb->spinning_readers, 0); atomic_set(&eb->spinning_writers, 0); + eb->lock_nested = 0; init_waitqueue_head(&eb->write_lock_wq); init_waitqueue_head(&eb->read_lock_wq); diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 7604c3001322..bc6a042cb6fc 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -129,6 +129,7 @@ struct extent_buffer { struct list_head leak_list; struct rcu_head rcu_head; atomic_t refs; + pid_t lock_owner; /* count of read lock holders on the extent buffer */ atomic_t write_locks; @@ -137,6 +138,7 @@ struct extent_buffer { atomic_t blocking_readers; atomic_t spinning_readers; atomic_t spinning_writers; + int lock_nested; /* protects write locks */ rwlock_t lock; diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c index d77b67c4b275..5e178d8f7167 100644 --- a/fs/btrfs/locking.c +++ b/fs/btrfs/locking.c @@ -33,6 +33,14 @@ void btrfs_assert_tree_read_locked(struct extent_buffer *eb); */ void btrfs_set_lock_blocking_rw(struct extent_buffer *eb, int rw) { + if (eb->lock_nested) { + read_lock(&eb->lock); + if (eb->lock_nested && current->pid == eb->lock_owner) { + read_unlock(&eb->lock); + return; + } + read_unlock(&eb->lock); + } if (rw == BTRFS_WRITE_LOCK) { if (atomic_read(&eb->blocking_writers) == 0) { WARN_ON(atomic_read(&eb->spinning_writers) != 1); @@ -57,6 +65,14 @@ void btrfs_set_lock_blocking_rw(struct extent_buffer *eb, int rw) */ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw) { + if (eb->lock_nested) { + read_lock(&eb->lock); + if (&eb->lock_nested && current->pid == eb->lock_owner) { + read_unlock(&eb->lock); + return; + } + read_unlock(&eb->lock); + } if (rw == BTRFS_WRITE_LOCK_BLOCKING) { BUG_ON(atomic_read(&eb->blocking_writers) != 1); write_lock(&eb->lock); @@ -81,12 +97,25 @@ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw) void btrfs_tree_read_lock(struct extent_buffer *eb) { again: + read_lock(&eb->lock); + if (atomic_read(&eb->blocking_writers) && + current->pid == eb->lock_owner) { + /* + * This extent is already write-locked by our thread. We allow + * an additional read lock to be added because it's for the same + * thread. btrfs_find_all_roots() depends on this as it may be + * called on a partly (write-)locked tree. + */ + BUG_ON(eb->lock_nested); + eb->lock_nested = 1; + read_unlock(&eb->lock); + return; + } + read_unlock(&eb->lock); wait_event(eb->write_lock_wq, atomic_read(&eb->blocking_writers) == 0); read_lock(&eb->lock); if (atomic_read(&eb->blocking_writers)) { read_unlock(&eb->lock); - wait_event(eb->write_lock_wq, - atomic_read(&eb->blocking_writers) == 0); goto again; } atomic_inc(&eb->read_locks); @@ -129,6 +158,7 @@ int btrfs_try_tree_write_lock(struct extent_buffer *eb) } atomic_inc(&eb->write_locks); atomic_inc(&eb->spinning_writers); + eb->lock_owner = current->pid; return 1; } @@ -137,6 +167,15 @@ int btrfs_try_tree_write_lock(struct extent_buffer *eb) */ void btrfs_tree_read_unlock(struct extent_buffer *eb) { + if (eb->lock_nested) { + read_lock(&eb->lock); + if (eb->lock_nested && current->pid == eb->lock_owner) { + eb->lock_nested = 0; + read_unlock(&eb->lock); + return; + } + read_unlock(&eb->lock); + } btrfs_assert_tree_read_locked(eb); WARN_ON(atomic_read(&eb->spinning_readers) == 0); atomic_dec(&eb->spinning_readers); @@ -149,6 +188,15 @@ void btrfs_tree_read_unlock(struct extent_buffer *eb) */ void btrfs_tree_read_unlock_blocking(struct extent_buffer *eb) { + if (eb->lock_nested) { + read_lock(&eb->lock); + if (eb->lock_nested && current->pid == eb->lock_owner) { + eb->lock_nested = 0; + read_unlock(&eb->lock); + return; + } + read_unlock(&eb->lock); + } btrfs_assert_tree_read_locked(eb); WARN_ON(atomic_read(&eb->blocking_readers) == 0); if (atomic_dec_and_test(&eb->blocking_readers)) @@ -181,6 +229,7 @@ again: WARN_ON(atomic_read(&eb->spinning_writers)); atomic_inc(&eb->spinning_writers); atomic_inc(&eb->write_locks); + eb->lock_owner = current->pid; return 0; } From 00f04b88791ff49dc64ada18819d40a5b0671709 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Wed, 14 Sep 2011 12:37:00 +0200 Subject: [PATCH 187/820] Btrfs: add sequence numbers to delayed refs Sequence numbers are needed to reconstruct the backrefs of a given extent to a certain point in time. The total set of backrefs consist of the set of backrefs recorded on disk plus the enqueued delayed refs for it that existed at that moment. This patch also adds a list that records all delayed refs which are currently in the process of being added. When walking all refs of an extent in btrfs_find_all_roots(), we freeze the current state of delayed refs, honor anythinh up to this point and prevent processing newer delayed refs to assert consistency. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/delayed-ref.c | 34 ++++++++++++++++++++ fs/btrfs/delayed-ref.h | 70 ++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/transaction.c | 4 +++ 3 files changed, 108 insertions(+) diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index babd37badb43..a405db0320e8 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -101,6 +101,11 @@ static int comp_entry(struct btrfs_delayed_ref_node *ref2, return -1; if (ref1->type > ref2->type) return 1; + /* merging of sequenced refs is not allowed */ + if (ref1->seq < ref2->seq) + return -1; + if (ref1->seq > ref2->seq) + return 1; if (ref1->type == BTRFS_TREE_BLOCK_REF_KEY || ref1->type == BTRFS_SHARED_BLOCK_REF_KEY) { return comp_tree_refs(btrfs_delayed_node_to_tree_ref(ref2), @@ -209,6 +214,24 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, return 0; } +int btrfs_check_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, + u64 seq) +{ + struct seq_list *elem; + + assert_spin_locked(&delayed_refs->lock); + if (list_empty(&delayed_refs->seq_head)) + return 0; + + elem = list_first_entry(&delayed_refs->seq_head, struct seq_list, list); + if (seq >= elem->seq) { + pr_debug("holding back delayed_ref %llu, lowest is %llu (%p)\n", + seq, elem->seq, delayed_refs); + return 1; + } + return 0; +} + int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, struct list_head *cluster, u64 start) { @@ -438,6 +461,7 @@ static noinline int add_delayed_ref_head(struct btrfs_fs_info *fs_info, ref->action = 0; ref->is_head = 1; ref->in_tree = 1; + ref->seq = 0; head_ref = btrfs_delayed_node_to_head(ref); head_ref->must_insert_reserved = must_insert_reserved; @@ -479,6 +503,7 @@ static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_node *existing; struct btrfs_delayed_tree_ref *full_ref; struct btrfs_delayed_ref_root *delayed_refs; + u64 seq = 0; if (action == BTRFS_ADD_DELAYED_EXTENT) action = BTRFS_ADD_DELAYED_REF; @@ -494,6 +519,10 @@ static noinline int add_delayed_tree_ref(struct btrfs_fs_info *fs_info, ref->is_head = 0; ref->in_tree = 1; + if (need_ref_seq(for_cow, ref_root)) + seq = inc_delayed_seq(delayed_refs); + ref->seq = seq; + full_ref = btrfs_delayed_node_to_tree_ref(ref); full_ref->parent = parent; full_ref->root = ref_root; @@ -534,6 +563,7 @@ static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_node *existing; struct btrfs_delayed_data_ref *full_ref; struct btrfs_delayed_ref_root *delayed_refs; + u64 seq = 0; if (action == BTRFS_ADD_DELAYED_EXTENT) action = BTRFS_ADD_DELAYED_REF; @@ -549,6 +579,10 @@ static noinline int add_delayed_data_ref(struct btrfs_fs_info *fs_info, ref->is_head = 0; ref->in_tree = 1; + if (need_ref_seq(for_cow, ref_root)) + seq = inc_delayed_seq(delayed_refs); + ref->seq = seq; + full_ref = btrfs_delayed_node_to_data_ref(ref); full_ref->parent = parent; full_ref->root = ref_root; diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index a5fb2bc83732..174416f7882b 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -33,6 +33,9 @@ struct btrfs_delayed_ref_node { /* the size of the extent */ u64 num_bytes; + /* seq number to keep track of insertion order */ + u64 seq; + /* ref count on this data structure */ atomic_t refs; @@ -136,6 +139,20 @@ struct btrfs_delayed_ref_root { int flushing; u64 run_delayed_start; + + /* + * seq number of delayed refs. We need to know if a backref was being + * added before the currently processed ref or afterwards. + */ + u64 seq; + + /* + * seq_list holds a list of all seq numbers that are currently being + * added to the list. While walking backrefs (btrfs_find_all_roots, + * qgroups), which might take some time, no newer ref must be processed, + * as it might influence the outcome of the walk. + */ + struct list_head seq_head; }; static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref) @@ -171,6 +188,59 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *head); int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, struct list_head *cluster, u64 search_start); + +struct seq_list { + struct list_head list; + u64 seq; +}; + +static inline u64 inc_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs) +{ + assert_spin_locked(&delayed_refs->lock); + ++delayed_refs->seq; + return delayed_refs->seq; +} + +static inline void +btrfs_get_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, + struct seq_list *elem) +{ + assert_spin_locked(&delayed_refs->lock); + elem->seq = delayed_refs->seq; + list_add_tail(&elem->list, &delayed_refs->seq_head); +} + +static inline void +btrfs_put_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, + struct seq_list *elem) +{ + spin_lock(&delayed_refs->lock); + list_del(&elem->list); + spin_unlock(&delayed_refs->lock); +} + +int btrfs_check_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, + u64 seq); + +/* + * delayed refs with a ref_seq > 0 must be held back during backref walking. + * this only applies to items in one of the fs-trees. for_cow items never need + * to be held back, so they won't get a ref_seq number. + */ +static inline int need_ref_seq(int for_cow, u64 rootid) +{ + if (for_cow) + return 0; + + if (rootid == BTRFS_FS_TREE_OBJECTID) + return 1; + + if ((s64)rootid >= (s64)BTRFS_FIRST_FREE_OBJECTID) + return 1; + + return 0; +} + /* * a node might live in a head or a regular ref, this lets you * test for the proper type to use. diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index a2bfedcbcabc..31a7393af64e 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -36,6 +36,8 @@ static noinline void put_transaction(struct btrfs_transaction *transaction) WARN_ON(atomic_read(&transaction->use_count) == 0); if (atomic_dec_and_test(&transaction->use_count)) { BUG_ON(!list_empty(&transaction->list)); + WARN_ON(transaction->delayed_refs.root.rb_node); + WARN_ON(!list_empty(&transaction->delayed_refs.seq_head)); memset(transaction, 0, sizeof(*transaction)); kmem_cache_free(btrfs_transaction_cachep, transaction); } @@ -108,8 +110,10 @@ loop: cur_trans->delayed_refs.num_heads = 0; cur_trans->delayed_refs.flushing = 0; cur_trans->delayed_refs.run_delayed_start = 0; + cur_trans->delayed_refs.seq = 1; spin_lock_init(&cur_trans->commit_lock); spin_lock_init(&cur_trans->delayed_refs.lock); + INIT_LIST_HEAD(&cur_trans->delayed_refs.seq_head); INIT_LIST_HEAD(&cur_trans->pending_snapshots); list_add_tail(&cur_trans->list, &root->fs_info->trans_list); From d1270cd91f308c9d22b2804720c36ccd32dbc35e Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 13 Sep 2011 15:16:43 +0200 Subject: [PATCH 188/820] Btrfs: put back delayed refs that are too new When processing a delayed ref, first check if there are still old refs in the process of being added. If so, put this ref back to the tree. To avoid looping on this ref, choose a newer one in the next loop. btrfs_find_ref_cluster has to take care of that. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/delayed-ref.c | 43 ++++++++++++++++++++++++------------------ fs/btrfs/extent-tree.c | 27 +++++++++++++++++++++----- 2 files changed, 47 insertions(+), 23 deletions(-) diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index a405db0320e8..ee181989d444 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -155,16 +155,22 @@ static struct btrfs_delayed_ref_node *tree_insert(struct rb_root *root, /* * find an head entry based on bytenr. This returns the delayed ref - * head if it was able to find one, or NULL if nothing was in that spot + * head if it was able to find one, or NULL if nothing was in that spot. + * If return_bigger is given, the next bigger entry is returned if no exact + * match is found. */ static struct btrfs_delayed_ref_node *find_ref_head(struct rb_root *root, u64 bytenr, - struct btrfs_delayed_ref_node **last) + struct btrfs_delayed_ref_node **last, + int return_bigger) { - struct rb_node *n = root->rb_node; + struct rb_node *n; struct btrfs_delayed_ref_node *entry; - int cmp; + int cmp = 0; +again: + n = root->rb_node; + entry = NULL; while (n) { entry = rb_entry(n, struct btrfs_delayed_ref_node, rb_node); WARN_ON(!entry->in_tree); @@ -187,6 +193,19 @@ static struct btrfs_delayed_ref_node *find_ref_head(struct rb_root *root, else return entry; } + if (entry && return_bigger) { + if (cmp > 0) { + n = rb_next(&entry->rb_node); + if (!n) + n = rb_first(root); + entry = rb_entry(n, struct btrfs_delayed_ref_node, + rb_node); + bytenr = entry->bytenr; + return_bigger = 0; + goto again; + } + return entry; + } return NULL; } @@ -246,20 +265,8 @@ int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, node = rb_first(&delayed_refs->root); } else { ref = NULL; - find_ref_head(&delayed_refs->root, start, &ref); + find_ref_head(&delayed_refs->root, start + 1, &ref, 1); if (ref) { - struct btrfs_delayed_ref_node *tmp; - - node = rb_prev(&ref->rb_node); - while (node) { - tmp = rb_entry(node, - struct btrfs_delayed_ref_node, - rb_node); - if (tmp->bytenr < start) - break; - ref = tmp; - node = rb_prev(&ref->rb_node); - } node = &ref->rb_node; } else node = rb_first(&delayed_refs->root); @@ -748,7 +755,7 @@ btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr) struct btrfs_delayed_ref_root *delayed_refs; delayed_refs = &trans->transaction->delayed_refs; - ref = find_ref_head(&delayed_refs->root, bytenr, NULL); + ref = find_ref_head(&delayed_refs->root, bytenr, NULL, 0); if (ref) return btrfs_delayed_node_to_head(ref); return NULL; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index dc8b9a834596..bbcca12fbbba 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2236,6 +2236,28 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, } } + /* + * locked_ref is the head node, so we have to go one + * node back for any delayed ref updates + */ + ref = select_delayed_ref(locked_ref); + + if (ref && ref->seq && + btrfs_check_delayed_seq(delayed_refs, ref->seq)) { + /* + * there are still refs with lower seq numbers in the + * process of being added. Don't run this ref yet. + */ + list_del_init(&locked_ref->cluster); + mutex_unlock(&locked_ref->mutex); + locked_ref = NULL; + delayed_refs->num_heads_ready++; + spin_unlock(&delayed_refs->lock); + cond_resched(); + spin_lock(&delayed_refs->lock); + continue; + } + /* * record the must insert reserved flag before we * drop the spin lock. @@ -2246,11 +2268,6 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, extent_op = locked_ref->extent_op; locked_ref->extent_op = NULL; - /* - * locked_ref is the head node, so we have to go one - * node back for any delayed ref updates - */ - ref = select_delayed_ref(locked_ref); if (!ref) { /* All delayed refs have been processed, Go ahead * and send the head node to run_one_delayed_ref, From a168650c08300434e1456abe7b6451f1448230d3 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Mon, 12 Dec 2011 16:10:07 +0100 Subject: [PATCH 189/820] Btrfs: add waitqueue instead of doing busy waiting for more delayed refs Now that we may be holding back delayed refs for a limited period, we might end up having no runnable delayed refs. Without this commit, we'd do busy waiting in that thread until another (runnable) ref arives. Instead, we're detecting this situation and use a waitqueue, such that we only try to run more refs after a) another runnable ref was added or b) delayed refs are no longer held back Signed-off-by: Jan Schmidt --- fs/btrfs/delayed-ref.c | 8 ++++++ fs/btrfs/delayed-ref.h | 7 +++++ fs/btrfs/extent-tree.c | 59 +++++++++++++++++++++++++++++++++++++++++- fs/btrfs/transaction.c | 1 + 4 files changed, 74 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index ee181989d444..66e4f29505a3 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -664,6 +664,9 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, num_bytes, parent, ref_root, level, action, for_cow); BUG_ON(ret); + if (!need_ref_seq(for_cow, ref_root) && + waitqueue_active(&delayed_refs->seq_wait)) + wake_up(&delayed_refs->seq_wait); spin_unlock(&delayed_refs->lock); return 0; } @@ -712,6 +715,9 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, num_bytes, parent, ref_root, owner, offset, action, for_cow); BUG_ON(ret); + if (!need_ref_seq(for_cow, ref_root) && + waitqueue_active(&delayed_refs->seq_wait)) + wake_up(&delayed_refs->seq_wait); spin_unlock(&delayed_refs->lock); return 0; } @@ -739,6 +745,8 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info, extent_op->is_data); BUG_ON(ret); + if (waitqueue_active(&delayed_refs->seq_wait)) + wake_up(&delayed_refs->seq_wait); spin_unlock(&delayed_refs->lock); return 0; } diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index 174416f7882b..d8f244d94925 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -153,6 +153,12 @@ struct btrfs_delayed_ref_root { * as it might influence the outcome of the walk. */ struct list_head seq_head; + + /* + * when the only refs we have in the list must not be processed, we want + * to wait for more refs to show up or for the end of backref walking. + */ + wait_queue_head_t seq_wait; }; static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref) @@ -216,6 +222,7 @@ btrfs_put_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs, { spin_lock(&delayed_refs->lock); list_del(&elem->list); + wake_up(&delayed_refs->seq_wait); spin_unlock(&delayed_refs->lock); } diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index bbcca12fbbba..0a435e2e143e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2300,7 +2300,12 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, ref->in_tree = 0; rb_erase(&ref->rb_node, &delayed_refs->root); delayed_refs->num_entries--; - + /* + * we modified num_entries, but as we're currently running + * delayed refs, skip + * wake_up(&delayed_refs->seq_wait); + * here. + */ spin_unlock(&delayed_refs->lock); ret = run_one_delayed_ref(trans, root, ref, extent_op, @@ -2317,6 +2322,23 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, return count; } + +static void wait_for_more_refs(struct btrfs_delayed_ref_root *delayed_refs, + unsigned long num_refs) +{ + struct list_head *first_seq = delayed_refs->seq_head.next; + + spin_unlock(&delayed_refs->lock); + pr_debug("waiting for more refs (num %ld, first %p)\n", + num_refs, first_seq); + wait_event(delayed_refs->seq_wait, + num_refs != delayed_refs->num_entries || + delayed_refs->seq_head.next != first_seq); + pr_debug("done waiting for more refs (num %ld, first %p)\n", + delayed_refs->num_entries, delayed_refs->seq_head.next); + spin_lock(&delayed_refs->lock); +} + /* * this starts processing the delayed reference count updates and * extent insertions we have queued up so far. count can be @@ -2332,8 +2354,11 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_node *ref; struct list_head cluster; int ret; + u64 delayed_start; int run_all = count == (unsigned long)-1; int run_most = 0; + unsigned long num_refs = 0; + int consider_waiting; if (root == root->fs_info->extent_root) root = root->fs_info->tree_root; @@ -2341,6 +2366,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, delayed_refs = &trans->transaction->delayed_refs; INIT_LIST_HEAD(&cluster); again: + consider_waiting = 0; spin_lock(&delayed_refs->lock); if (count == 0) { count = delayed_refs->num_entries * 2; @@ -2357,11 +2383,35 @@ again: * of refs to process starting at the first one we are able to * lock */ + delayed_start = delayed_refs->run_delayed_start; ret = btrfs_find_ref_cluster(trans, &cluster, delayed_refs->run_delayed_start); if (ret) break; + if (delayed_start >= delayed_refs->run_delayed_start) { + if (consider_waiting == 0) { + /* + * btrfs_find_ref_cluster looped. let's do one + * more cycle. if we don't run any delayed ref + * during that cycle (because we can't because + * all of them are blocked) and if the number of + * refs doesn't change, we avoid busy waiting. + */ + consider_waiting = 1; + num_refs = delayed_refs->num_entries; + } else { + wait_for_more_refs(delayed_refs, num_refs); + /* + * after waiting, things have changed. we + * dropped the lock and someone else might have + * run some refs, built new clusters and so on. + * therefore, we restart staleness detection. + */ + consider_waiting = 0; + } + } + ret = run_clustered_refs(trans, root, &cluster); BUG_ON(ret < 0); @@ -2369,6 +2419,11 @@ again: if (count == 0) break; + + if (ret || delayed_refs->run_delayed_start == 0) { + /* refs were run, let's reset staleness detection */ + consider_waiting = 0; + } } if (run_all) { @@ -4933,6 +4988,8 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans, rb_erase(&head->node.rb_node, &delayed_refs->root); delayed_refs->num_entries--; + if (waitqueue_active(&delayed_refs->seq_wait)) + wake_up(&delayed_refs->seq_wait); /* * we don't take a ref on the node because we're removing it from the diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 31a7393af64e..04c5c7c2c32f 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -111,6 +111,7 @@ loop: cur_trans->delayed_refs.flushing = 0; cur_trans->delayed_refs.run_delayed_start = 0; cur_trans->delayed_refs.seq = 1; + init_waitqueue_head(&cur_trans->delayed_refs.seq_wait); spin_lock_init(&cur_trans->commit_lock); spin_lock_init(&cur_trans->delayed_refs.lock); INIT_LIST_HEAD(&cur_trans->delayed_refs.seq_head); From 8da6d5815c592b713ecaf4f4f8b631f8359c96c4 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Wed, 23 Nov 2011 18:55:04 +0100 Subject: [PATCH 190/820] Btrfs: added btrfs_find_all_roots() This function gets a byte number (a data extent), collects all the leafs pointing to it and walks up the trees to find all fs roots pointing to those leafs. It also returns the list of all leafs pointing to that extent. It does proper locking for the involved trees, can be used on busy file systems and honors delayed refs. Signed-off-by: Arne Jansen Signed-off-by: Jan Schmidt --- fs/btrfs/backref.c | 783 +++++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/backref.h | 5 + 2 files changed, 788 insertions(+) diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 22c64fff1bd5..03c30a1836f4 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -19,6 +19,9 @@ #include "ctree.h" #include "disk-io.h" #include "backref.h" +#include "ulist.h" +#include "transaction.h" +#include "delayed-ref.h" struct __data_ref { struct list_head list; @@ -32,6 +35,786 @@ struct __shared_ref { u64 disk_byte; }; +/* + * this structure records all encountered refs on the way up to the root + */ +struct __prelim_ref { + struct list_head list; + u64 root_id; + struct btrfs_key key; + int level; + int count; + u64 parent; + u64 wanted_disk_byte; +}; + +static int __add_prelim_ref(struct list_head *head, u64 root_id, + struct btrfs_key *key, int level, u64 parent, + u64 wanted_disk_byte, int count) +{ + struct __prelim_ref *ref; + + /* in case we're adding delayed refs, we're holding the refs spinlock */ + ref = kmalloc(sizeof(*ref), GFP_ATOMIC); + if (!ref) + return -ENOMEM; + + ref->root_id = root_id; + if (key) + ref->key = *key; + else + memset(&ref->key, 0, sizeof(ref->key)); + + ref->level = level; + ref->count = count; + ref->parent = parent; + ref->wanted_disk_byte = wanted_disk_byte; + list_add_tail(&ref->list, head); + + return 0; +} + +static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path, + struct ulist *parents, + struct extent_buffer *eb, int level, + u64 wanted_objectid, u64 wanted_disk_byte) +{ + int ret; + int slot; + struct btrfs_file_extent_item *fi; + struct btrfs_key key; + u64 disk_byte; + +add_parent: + ret = ulist_add(parents, eb->start, 0, GFP_NOFS); + if (ret < 0) + return ret; + + if (level != 0) + return 0; + + /* + * if the current leaf is full with EXTENT_DATA items, we must + * check the next one if that holds a reference as well. + * ref->count cannot be used to skip this check. + * repeat this until we don't find any additional EXTENT_DATA items. + */ + while (1) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + return ret; + if (ret) + return 0; + + eb = path->nodes[0]; + for (slot = 0; slot < btrfs_header_nritems(eb); ++slot) { + btrfs_item_key_to_cpu(eb, &key, slot); + if (key.objectid != wanted_objectid || + key.type != BTRFS_EXTENT_DATA_KEY) + return 0; + fi = btrfs_item_ptr(eb, slot, + struct btrfs_file_extent_item); + disk_byte = btrfs_file_extent_disk_bytenr(eb, fi); + if (disk_byte == wanted_disk_byte) + goto add_parent; + } + } + + return 0; +} + +/* + * resolve an indirect backref in the form (root_id, key, level) + * to a logical address + */ +static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info, + struct __prelim_ref *ref, + struct ulist *parents) +{ + struct btrfs_path *path; + struct btrfs_root *root; + struct btrfs_key root_key; + struct btrfs_key key = {0}; + struct extent_buffer *eb; + int ret = 0; + int root_level; + int level = ref->level; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + root_key.objectid = ref->root_id; + root_key.type = BTRFS_ROOT_ITEM_KEY; + root_key.offset = (u64)-1; + root = btrfs_read_fs_root_no_name(fs_info, &root_key); + if (IS_ERR(root)) { + ret = PTR_ERR(root); + goto out; + } + + rcu_read_lock(); + root_level = btrfs_header_level(root->node); + rcu_read_unlock(); + + if (root_level + 1 == level) + goto out; + + path->lowest_level = level; + ret = btrfs_search_slot(NULL, root, &ref->key, path, 0, 0); + pr_debug("search slot in root %llu (level %d, ref count %d) returned " + "%d for key (%llu %u %llu)\n", + (unsigned long long)ref->root_id, level, ref->count, ret, + (unsigned long long)ref->key.objectid, ref->key.type, + (unsigned long long)ref->key.offset); + if (ret < 0) + goto out; + + eb = path->nodes[level]; + if (!eb) { + WARN_ON(1); + ret = 1; + goto out; + } + + if (level == 0) { + if (ret == 1 && path->slots[0] >= btrfs_header_nritems(eb)) { + ret = btrfs_next_leaf(root, path); + if (ret) + goto out; + eb = path->nodes[0]; + } + + btrfs_item_key_to_cpu(eb, &key, path->slots[0]); + } + + /* the last two parameters will only be used for level == 0 */ + ret = add_all_parents(root, path, parents, eb, level, key.objectid, + ref->wanted_disk_byte); +out: + btrfs_free_path(path); + return ret; +} + +/* + * resolve all indirect backrefs from the list + */ +static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info, + struct list_head *head) +{ + int err; + int ret = 0; + struct __prelim_ref *ref; + struct __prelim_ref *ref_safe; + struct __prelim_ref *new_ref; + struct ulist *parents; + struct ulist_node *node; + + parents = ulist_alloc(GFP_NOFS); + if (!parents) + return -ENOMEM; + + /* + * _safe allows us to insert directly after the current item without + * iterating over the newly inserted items. + * we're also allowed to re-assign ref during iteration. + */ + list_for_each_entry_safe(ref, ref_safe, head, list) { + if (ref->parent) /* already direct */ + continue; + if (ref->count == 0) + continue; + err = __resolve_indirect_ref(fs_info, ref, parents); + if (err) { + if (ret == 0) + ret = err; + continue; + } + + /* we put the first parent into the ref at hand */ + node = ulist_next(parents, NULL); + ref->parent = node ? node->val : 0; + + /* additional parents require new refs being added here */ + while ((node = ulist_next(parents, node))) { + new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS); + if (!new_ref) { + ret = -ENOMEM; + break; + } + memcpy(new_ref, ref, sizeof(*ref)); + new_ref->parent = node->val; + list_add(&new_ref->list, &ref->list); + } + ulist_reinit(parents); + } + + ulist_free(parents); + return ret; +} + +/* + * merge two lists of backrefs and adjust counts accordingly + * + * mode = 1: merge identical keys, if key is set + * mode = 2: merge identical parents + */ +static int __merge_refs(struct list_head *head, int mode) +{ + struct list_head *pos1; + + list_for_each(pos1, head) { + struct list_head *n2; + struct list_head *pos2; + struct __prelim_ref *ref1; + + ref1 = list_entry(pos1, struct __prelim_ref, list); + + if (mode == 1 && ref1->key.type == 0) + continue; + for (pos2 = pos1->next, n2 = pos2->next; pos2 != head; + pos2 = n2, n2 = pos2->next) { + struct __prelim_ref *ref2; + + ref2 = list_entry(pos2, struct __prelim_ref, list); + + if (mode == 1) { + if (memcmp(&ref1->key, &ref2->key, + sizeof(ref1->key)) || + ref1->level != ref2->level || + ref1->root_id != ref2->root_id) + continue; + ref1->count += ref2->count; + } else { + if (ref1->parent != ref2->parent) + continue; + ref1->count += ref2->count; + } + list_del(&ref2->list); + kfree(ref2); + } + + } + return 0; +} + +/* + * add all currently queued delayed refs from this head whose seq nr is + * smaller or equal that seq to the list + */ +static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq, + struct btrfs_key *info_key, + struct list_head *prefs) +{ + struct btrfs_delayed_extent_op *extent_op = head->extent_op; + struct rb_node *n = &head->node.rb_node; + int sgn; + int ret; + + if (extent_op && extent_op->update_key) + btrfs_disk_key_to_cpu(info_key, &extent_op->key); + + while ((n = rb_prev(n))) { + struct btrfs_delayed_ref_node *node; + node = rb_entry(n, struct btrfs_delayed_ref_node, + rb_node); + if (node->bytenr != head->node.bytenr) + break; + WARN_ON(node->is_head); + + if (node->seq > seq) + continue; + + switch (node->action) { + case BTRFS_ADD_DELAYED_EXTENT: + case BTRFS_UPDATE_DELAYED_HEAD: + WARN_ON(1); + continue; + case BTRFS_ADD_DELAYED_REF: + sgn = 1; + break; + case BTRFS_DROP_DELAYED_REF: + sgn = -1; + break; + default: + BUG_ON(1); + } + switch (node->type) { + case BTRFS_TREE_BLOCK_REF_KEY: { + struct btrfs_delayed_tree_ref *ref; + + ref = btrfs_delayed_node_to_tree_ref(node); + ret = __add_prelim_ref(prefs, ref->root, info_key, + ref->level + 1, 0, node->bytenr, + node->ref_mod * sgn); + break; + } + case BTRFS_SHARED_BLOCK_REF_KEY: { + struct btrfs_delayed_tree_ref *ref; + + ref = btrfs_delayed_node_to_tree_ref(node); + ret = __add_prelim_ref(prefs, ref->root, info_key, + ref->level + 1, ref->parent, + node->bytenr, + node->ref_mod * sgn); + break; + } + case BTRFS_EXTENT_DATA_REF_KEY: { + struct btrfs_delayed_data_ref *ref; + struct btrfs_key key; + + ref = btrfs_delayed_node_to_data_ref(node); + + key.objectid = ref->objectid; + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = ref->offset; + ret = __add_prelim_ref(prefs, ref->root, &key, 0, 0, + node->bytenr, + node->ref_mod * sgn); + break; + } + case BTRFS_SHARED_DATA_REF_KEY: { + struct btrfs_delayed_data_ref *ref; + struct btrfs_key key; + + ref = btrfs_delayed_node_to_data_ref(node); + + key.objectid = ref->objectid; + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = ref->offset; + ret = __add_prelim_ref(prefs, ref->root, &key, 0, + ref->parent, node->bytenr, + node->ref_mod * sgn); + break; + } + default: + WARN_ON(1); + } + BUG_ON(ret); + } + + return 0; +} + +/* + * add all inline backrefs for bytenr to the list + */ +static int __add_inline_refs(struct btrfs_fs_info *fs_info, + struct btrfs_path *path, u64 bytenr, + struct btrfs_key *info_key, int *info_level, + struct list_head *prefs) +{ + int ret; + int slot; + struct extent_buffer *leaf; + struct btrfs_key key; + unsigned long ptr; + unsigned long end; + struct btrfs_extent_item *ei; + u64 flags; + u64 item_size; + + /* + * enumerate all inline refs + */ + leaf = path->nodes[0]; + slot = path->slots[0] - 1; + + item_size = btrfs_item_size_nr(leaf, slot); + BUG_ON(item_size < sizeof(*ei)); + + ei = btrfs_item_ptr(leaf, slot, struct btrfs_extent_item); + flags = btrfs_extent_flags(leaf, ei); + + ptr = (unsigned long)(ei + 1); + end = (unsigned long)ei + item_size; + + if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { + struct btrfs_tree_block_info *info; + struct btrfs_disk_key disk_key; + + info = (struct btrfs_tree_block_info *)ptr; + *info_level = btrfs_tree_block_level(leaf, info); + btrfs_tree_block_key(leaf, info, &disk_key); + btrfs_disk_key_to_cpu(info_key, &disk_key); + ptr += sizeof(struct btrfs_tree_block_info); + BUG_ON(ptr > end); + } else { + BUG_ON(!(flags & BTRFS_EXTENT_FLAG_DATA)); + } + + while (ptr < end) { + struct btrfs_extent_inline_ref *iref; + u64 offset; + int type; + + iref = (struct btrfs_extent_inline_ref *)ptr; + type = btrfs_extent_inline_ref_type(leaf, iref); + offset = btrfs_extent_inline_ref_offset(leaf, iref); + + switch (type) { + case BTRFS_SHARED_BLOCK_REF_KEY: + ret = __add_prelim_ref(prefs, 0, info_key, + *info_level + 1, offset, + bytenr, 1); + break; + case BTRFS_SHARED_DATA_REF_KEY: { + struct btrfs_shared_data_ref *sdref; + int count; + + sdref = (struct btrfs_shared_data_ref *)(iref + 1); + count = btrfs_shared_data_ref_count(leaf, sdref); + ret = __add_prelim_ref(prefs, 0, NULL, 0, offset, + bytenr, count); + break; + } + case BTRFS_TREE_BLOCK_REF_KEY: + ret = __add_prelim_ref(prefs, offset, info_key, + *info_level + 1, 0, bytenr, 1); + break; + case BTRFS_EXTENT_DATA_REF_KEY: { + struct btrfs_extent_data_ref *dref; + int count; + u64 root; + + dref = (struct btrfs_extent_data_ref *)(&iref->offset); + count = btrfs_extent_data_ref_count(leaf, dref); + key.objectid = btrfs_extent_data_ref_objectid(leaf, + dref); + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = btrfs_extent_data_ref_offset(leaf, dref); + root = btrfs_extent_data_ref_root(leaf, dref); + ret = __add_prelim_ref(prefs, root, &key, 0, 0, bytenr, + count); + break; + } + default: + WARN_ON(1); + } + BUG_ON(ret); + ptr += btrfs_extent_inline_ref_size(type); + } + + return 0; +} + +/* + * add all non-inline backrefs for bytenr to the list + */ +static int __add_keyed_refs(struct btrfs_fs_info *fs_info, + struct btrfs_path *path, u64 bytenr, + struct btrfs_key *info_key, int info_level, + struct list_head *prefs) +{ + struct btrfs_root *extent_root = fs_info->extent_root; + int ret; + int slot; + struct extent_buffer *leaf; + struct btrfs_key key; + + while (1) { + ret = btrfs_next_item(extent_root, path); + if (ret < 0) + break; + if (ret) { + ret = 0; + break; + } + + slot = path->slots[0]; + leaf = path->nodes[0]; + btrfs_item_key_to_cpu(leaf, &key, slot); + + if (key.objectid != bytenr) + break; + if (key.type < BTRFS_TREE_BLOCK_REF_KEY) + continue; + if (key.type > BTRFS_SHARED_DATA_REF_KEY) + break; + + switch (key.type) { + case BTRFS_SHARED_BLOCK_REF_KEY: + ret = __add_prelim_ref(prefs, 0, info_key, + info_level + 1, key.offset, + bytenr, 1); + break; + case BTRFS_SHARED_DATA_REF_KEY: { + struct btrfs_shared_data_ref *sdref; + int count; + + sdref = btrfs_item_ptr(leaf, slot, + struct btrfs_shared_data_ref); + count = btrfs_shared_data_ref_count(leaf, sdref); + ret = __add_prelim_ref(prefs, 0, NULL, 0, key.offset, + bytenr, count); + break; + } + case BTRFS_TREE_BLOCK_REF_KEY: + ret = __add_prelim_ref(prefs, key.offset, info_key, + info_level + 1, 0, bytenr, 1); + break; + case BTRFS_EXTENT_DATA_REF_KEY: { + struct btrfs_extent_data_ref *dref; + int count; + u64 root; + + dref = btrfs_item_ptr(leaf, slot, + struct btrfs_extent_data_ref); + count = btrfs_extent_data_ref_count(leaf, dref); + key.objectid = btrfs_extent_data_ref_objectid(leaf, + dref); + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = btrfs_extent_data_ref_offset(leaf, dref); + root = btrfs_extent_data_ref_root(leaf, dref); + ret = __add_prelim_ref(prefs, root, &key, 0, 0, + bytenr, count); + break; + } + default: + WARN_ON(1); + } + BUG_ON(ret); + } + + return ret; +} + +/* + * this adds all existing backrefs (inline backrefs, backrefs and delayed + * refs) for the given bytenr to the refs list, merges duplicates and resolves + * indirect refs to their parent bytenr. + * When roots are found, they're added to the roots list + * + * FIXME some caching might speed things up + */ +static int find_parent_nodes(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, u64 bytenr, + u64 seq, struct ulist *refs, struct ulist *roots) +{ + struct btrfs_key key; + struct btrfs_path *path; + struct btrfs_key info_key = { 0 }; + struct btrfs_delayed_ref_root *delayed_refs = NULL; + struct btrfs_delayed_ref_head *head = NULL; + int info_level = 0; + int ret; + struct list_head prefs_delayed; + struct list_head prefs; + struct __prelim_ref *ref; + + INIT_LIST_HEAD(&prefs); + INIT_LIST_HEAD(&prefs_delayed); + + key.objectid = bytenr; + key.type = BTRFS_EXTENT_ITEM_KEY; + key.offset = (u64)-1; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + /* + * grab both a lock on the path and a lock on the delayed ref head. + * We need both to get a consistent picture of how the refs look + * at a specified point in time + */ +again: + ret = btrfs_search_slot(trans, fs_info->extent_root, &key, path, 0, 0); + if (ret < 0) + goto out; + BUG_ON(ret == 0); + + /* + * look if there are updates for this ref queued and lock the head + */ + delayed_refs = &trans->transaction->delayed_refs; + spin_lock(&delayed_refs->lock); + head = btrfs_find_delayed_ref_head(trans, bytenr); + if (head) { + if (!mutex_trylock(&head->mutex)) { + atomic_inc(&head->node.refs); + spin_unlock(&delayed_refs->lock); + + btrfs_release_path(path); + + /* + * Mutex was contended, block until it's + * released and try again + */ + mutex_lock(&head->mutex); + mutex_unlock(&head->mutex); + btrfs_put_delayed_ref(&head->node); + goto again; + } + ret = __add_delayed_refs(head, seq, &info_key, &prefs_delayed); + if (ret) + goto out; + } + spin_unlock(&delayed_refs->lock); + + if (path->slots[0]) { + struct extent_buffer *leaf; + int slot; + + leaf = path->nodes[0]; + slot = path->slots[0] - 1; + btrfs_item_key_to_cpu(leaf, &key, slot); + if (key.objectid == bytenr && + key.type == BTRFS_EXTENT_ITEM_KEY) { + ret = __add_inline_refs(fs_info, path, bytenr, + &info_key, &info_level, &prefs); + if (ret) + goto out; + ret = __add_keyed_refs(fs_info, path, bytenr, &info_key, + info_level, &prefs); + if (ret) + goto out; + } + } + btrfs_release_path(path); + + /* + * when adding the delayed refs above, the info_key might not have + * been known yet. Go over the list and replace the missing keys + */ + list_for_each_entry(ref, &prefs_delayed, list) { + if ((ref->key.offset | ref->key.type | ref->key.objectid) == 0) + memcpy(&ref->key, &info_key, sizeof(ref->key)); + } + list_splice_init(&prefs_delayed, &prefs); + + ret = __merge_refs(&prefs, 1); + if (ret) + goto out; + + ret = __resolve_indirect_refs(fs_info, &prefs); + if (ret) + goto out; + + ret = __merge_refs(&prefs, 2); + if (ret) + goto out; + + while (!list_empty(&prefs)) { + ref = list_first_entry(&prefs, struct __prelim_ref, list); + list_del(&ref->list); + if (ref->count < 0) + WARN_ON(1); + if (ref->count && ref->root_id && ref->parent == 0) { + /* no parent == root of tree */ + ret = ulist_add(roots, ref->root_id, 0, GFP_NOFS); + BUG_ON(ret < 0); + } + if (ref->count && ref->parent) { + ret = ulist_add(refs, ref->parent, 0, GFP_NOFS); + BUG_ON(ret < 0); + } + kfree(ref); + } + +out: + if (head) + mutex_unlock(&head->mutex); + btrfs_free_path(path); + while (!list_empty(&prefs)) { + ref = list_first_entry(&prefs, struct __prelim_ref, list); + list_del(&ref->list); + kfree(ref); + } + while (!list_empty(&prefs_delayed)) { + ref = list_first_entry(&prefs_delayed, struct __prelim_ref, + list); + list_del(&ref->list); + kfree(ref); + } + + return ret; +} + +/* + * Finds all leafs with a reference to the specified combination of bytenr and + * offset. key_list_head will point to a list of corresponding keys (caller must + * free each list element). The leafs will be stored in the leafs ulist, which + * must be freed with ulist_free. + * + * returns 0 on success, <0 on error + */ +static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, u64 bytenr, + u64 num_bytes, u64 seq, struct ulist **leafs) +{ + struct ulist *tmp; + int ret; + + tmp = ulist_alloc(GFP_NOFS); + if (!tmp) + return -ENOMEM; + *leafs = ulist_alloc(GFP_NOFS); + if (!*leafs) { + ulist_free(tmp); + return -ENOMEM; + } + + ret = find_parent_nodes(trans, fs_info, bytenr, seq, *leafs, tmp); + ulist_free(tmp); + + if (ret < 0 && ret != -ENOENT) { + ulist_free(*leafs); + return ret; + } + + return 0; +} + +/* + * walk all backrefs for a given extent to find all roots that reference this + * extent. Walking a backref means finding all extents that reference this + * extent and in turn walk the backrefs of those, too. Naturally this is a + * recursive process, but here it is implemented in an iterative fashion: We + * find all referencing extents for the extent in question and put them on a + * list. In turn, we find all referencing extents for those, further appending + * to the list. The way we iterate the list allows adding more elements after + * the current while iterating. The process stops when we reach the end of the + * list. Found roots are added to the roots list. + * + * returns 0 on success, < 0 on error. + */ +int btrfs_find_all_roots(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, u64 bytenr, + u64 num_bytes, u64 seq, struct ulist **roots) +{ + struct ulist *tmp; + struct ulist_node *node = NULL; + int ret; + + tmp = ulist_alloc(GFP_NOFS); + if (!tmp) + return -ENOMEM; + *roots = ulist_alloc(GFP_NOFS); + if (!*roots) { + ulist_free(tmp); + return -ENOMEM; + } + + while (1) { + ret = find_parent_nodes(trans, fs_info, bytenr, seq, + tmp, *roots); + if (ret < 0 && ret != -ENOENT) { + ulist_free(tmp); + ulist_free(*roots); + return ret; + } + node = ulist_next(tmp, node); + if (!node) + break; + bytenr = node->val; + } + + ulist_free(tmp); + return 0; +} + + static int __inode_info(u64 inum, u64 ioff, u8 key_type, struct btrfs_root *fs_root, struct btrfs_path *path, struct btrfs_key *found_key) diff --git a/fs/btrfs/backref.h b/fs/btrfs/backref.h index 92618837cb8f..d00dfa9ca934 100644 --- a/fs/btrfs/backref.h +++ b/fs/btrfs/backref.h @@ -20,6 +20,7 @@ #define __BTRFS_BACKREF__ #include "ioctl.h" +#include "ulist.h" struct inode_fs_paths { struct btrfs_path *btrfs_path; @@ -54,6 +55,10 @@ int iterate_inodes_from_logical(u64 logical, struct btrfs_fs_info *fs_info, int paths_from_inode(u64 inum, struct inode_fs_paths *ipath); +int btrfs_find_all_roots(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, u64 bytenr, + u64 num_bytes, u64 seq, struct ulist **roots); + struct btrfs_data_container *init_data_container(u32 total_bytes); struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root, struct btrfs_path *path); From 4692cf58aa7b81f721c1653d48db99ea41421d58 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Fri, 2 Dec 2011 14:56:41 +0100 Subject: [PATCH 191/820] Btrfs: new backref walking code The old backref iteration code could only safely be used on commit roots. Besides this limitation, it had bugs in finding the roots for these references. This commit replaces large parts of it by btrfs_find_all_roots() which a) really finds all roots and the correct roots, b) works correctly under heavy file system load, c) considers delayed refs. Signed-off-by: Jan Schmidt --- fs/btrfs/backref.c | 358 +++++++++++++-------------------------------- fs/btrfs/ioctl.c | 8 +- fs/btrfs/scrub.c | 7 +- 3 files changed, 109 insertions(+), 264 deletions(-) diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 03c30a1836f4..b9a843226de8 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -23,18 +23,6 @@ #include "transaction.h" #include "delayed-ref.h" -struct __data_ref { - struct list_head list; - u64 inum; - u64 root; - u64 extent_data_item_offset; -}; - -struct __shared_ref { - struct list_head list; - u64 disk_byte; -}; - /* * this structure records all encountered refs on the way up to the root */ @@ -964,8 +952,11 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical, btrfs_item_key_to_cpu(path->nodes[0], found_key, path->slots[0]); if (found_key->type != BTRFS_EXTENT_ITEM_KEY || found_key->objectid > logical || - found_key->objectid + found_key->offset <= logical) + found_key->objectid + found_key->offset <= logical) { + pr_debug("logical %llu is not within any extent\n", + (unsigned long long)logical); return -ENOENT; + } eb = path->nodes[0]; item_size = btrfs_item_size_nr(eb, path->slots[0]); @@ -974,6 +965,13 @@ int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical, ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); flags = btrfs_extent_flags(eb, ei); + pr_debug("logical %llu is at position %llu within the extent (%llu " + "EXTENT_ITEM %llu) flags %#llx size %u\n", + (unsigned long long)logical, + (unsigned long long)(logical - found_key->objectid), + (unsigned long long)found_key->objectid, + (unsigned long long)found_key->offset, + (unsigned long long)flags, item_size); if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) return BTRFS_EXTENT_FLAG_TREE_BLOCK; if (flags & BTRFS_EXTENT_FLAG_DATA) @@ -1070,128 +1068,11 @@ int tree_backref_for_extent(unsigned long *ptr, struct extent_buffer *eb, return 0; } -static int __data_list_add(struct list_head *head, u64 inum, - u64 extent_data_item_offset, u64 root) -{ - struct __data_ref *ref; - - ref = kmalloc(sizeof(*ref), GFP_NOFS); - if (!ref) - return -ENOMEM; - - ref->inum = inum; - ref->extent_data_item_offset = extent_data_item_offset; - ref->root = root; - list_add_tail(&ref->list, head); - - return 0; -} - -static int __data_list_add_eb(struct list_head *head, struct extent_buffer *eb, - struct btrfs_extent_data_ref *dref) -{ - return __data_list_add(head, btrfs_extent_data_ref_objectid(eb, dref), - btrfs_extent_data_ref_offset(eb, dref), - btrfs_extent_data_ref_root(eb, dref)); -} - -static int __shared_list_add(struct list_head *head, u64 disk_byte) -{ - struct __shared_ref *ref; - - ref = kmalloc(sizeof(*ref), GFP_NOFS); - if (!ref) - return -ENOMEM; - - ref->disk_byte = disk_byte; - list_add_tail(&ref->list, head); - - return 0; -} - -static int __iter_shared_inline_ref_inodes(struct btrfs_fs_info *fs_info, - u64 logical, u64 inum, - u64 extent_data_item_offset, - u64 extent_offset, - struct btrfs_path *path, - struct list_head *data_refs, - iterate_extent_inodes_t *iterate, - void *ctx) -{ - u64 ref_root; - u32 item_size; - struct btrfs_key key; - struct extent_buffer *eb; - struct btrfs_extent_item *ei; - struct btrfs_extent_inline_ref *eiref; - struct __data_ref *ref; - int ret; - int type; - int last; - unsigned long ptr = 0; - - WARN_ON(!list_empty(data_refs)); - ret = extent_from_logical(fs_info, logical, path, &key); - if (ret & BTRFS_EXTENT_FLAG_DATA) - ret = -EIO; - if (ret < 0) - goto out; - - eb = path->nodes[0]; - ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); - item_size = btrfs_item_size_nr(eb, path->slots[0]); - - ret = 0; - ref_root = 0; - /* - * as done in iterate_extent_inodes, we first build a list of refs to - * iterate, then free the path and then iterate them to avoid deadlocks. - */ - do { - last = __get_extent_inline_ref(&ptr, eb, ei, item_size, - &eiref, &type); - if (last < 0) { - ret = last; - goto out; - } - if (type == BTRFS_TREE_BLOCK_REF_KEY || - type == BTRFS_SHARED_BLOCK_REF_KEY) { - ref_root = btrfs_extent_inline_ref_offset(eb, eiref); - ret = __data_list_add(data_refs, inum, - extent_data_item_offset, - ref_root); - } - } while (!ret && !last); - - btrfs_release_path(path); - - if (ref_root == 0) { - printk(KERN_ERR "btrfs: failed to find tree block ref " - "for shared data backref %llu\n", logical); - WARN_ON(1); - ret = -EIO; - } - -out: - while (!list_empty(data_refs)) { - ref = list_first_entry(data_refs, struct __data_ref, list); - list_del(&ref->list); - if (!ret) - ret = iterate(ref->inum, extent_offset + - ref->extent_data_item_offset, - ref->root, ctx); - kfree(ref); - } - - return ret; -} - -static int __iter_shared_inline_ref(struct btrfs_fs_info *fs_info, - u64 logical, u64 orig_extent_item_objectid, - u64 extent_offset, struct btrfs_path *path, - struct list_head *data_refs, - iterate_extent_inodes_t *iterate, - void *ctx) +static int iterate_leaf_refs(struct btrfs_fs_info *fs_info, + struct btrfs_path *path, u64 logical, + u64 orig_extent_item_objectid, + u64 extent_item_pos, u64 root, + iterate_extent_inodes_t *iterate, void *ctx) { u64 disk_byte; struct btrfs_key key; @@ -1199,8 +1080,10 @@ static int __iter_shared_inline_ref(struct btrfs_fs_info *fs_info, struct extent_buffer *eb; int slot; int nritems; - int ret; - int found = 0; + int ret = 0; + int extent_type; + u64 data_offset; + u64 data_len; eb = read_tree_block(fs_info->tree_root, logical, fs_info->tree_root->leafsize, 0); @@ -1218,149 +1101,99 @@ static int __iter_shared_inline_ref(struct btrfs_fs_info *fs_info, if (key.type != BTRFS_EXTENT_DATA_KEY) continue; fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item); - if (!fi) { - free_extent_buffer(eb); - return -EIO; - } + extent_type = btrfs_file_extent_type(eb, fi); + if (extent_type == BTRFS_FILE_EXTENT_INLINE) + continue; + /* don't skip BTRFS_FILE_EXTENT_PREALLOC, we can handle that */ disk_byte = btrfs_file_extent_disk_bytenr(eb, fi); - if (disk_byte != orig_extent_item_objectid) { - if (found) - break; - else - continue; - } - ++found; - ret = __iter_shared_inline_ref_inodes(fs_info, logical, - key.objectid, - key.offset, - extent_offset, path, - data_refs, - iterate, ctx); - if (ret) - break; - } + if (disk_byte != orig_extent_item_objectid) + continue; - if (!found) { - printk(KERN_ERR "btrfs: failed to follow shared data backref " - "to parent %llu\n", logical); - WARN_ON(1); - ret = -EIO; + data_offset = btrfs_file_extent_offset(eb, fi); + data_len = btrfs_file_extent_num_bytes(eb, fi); + + if (extent_item_pos < data_offset || + extent_item_pos >= data_offset + data_len) + continue; + + pr_debug("ref for %llu resolved, key (%llu EXTEND_DATA %llu), " + "root %llu\n", orig_extent_item_objectid, + key.objectid, key.offset, root); + ret = iterate(key.objectid, + key.offset + (extent_item_pos - data_offset), + root, ctx); + if (ret) { + pr_debug("stopping iteration because ret=%d\n", ret); + break; + } } free_extent_buffer(eb); + return ret; } /* * calls iterate() for every inode that references the extent identified by - * the given parameters. will use the path given as a parameter and return it - * released. + * the given parameters. * when the iterator function returns a non-zero value, iteration stops. + * path is guaranteed to be in released state when iterate() is called. */ int iterate_extent_inodes(struct btrfs_fs_info *fs_info, struct btrfs_path *path, - u64 extent_item_objectid, - u64 extent_offset, + u64 extent_item_objectid, u64 extent_item_pos, iterate_extent_inodes_t *iterate, void *ctx) { - unsigned long ptr = 0; - int last; int ret; - int type; - u64 logical; - u32 item_size; - struct btrfs_extent_inline_ref *eiref; - struct btrfs_extent_data_ref *dref; - struct extent_buffer *eb; - struct btrfs_extent_item *ei; - struct btrfs_key key; struct list_head data_refs = LIST_HEAD_INIT(data_refs); struct list_head shared_refs = LIST_HEAD_INIT(shared_refs); - struct __data_ref *ref_d; - struct __shared_ref *ref_s; + struct btrfs_trans_handle *trans; + struct ulist *refs; + struct ulist *roots; + struct ulist_node *ref_node = NULL; + struct ulist_node *root_node = NULL; + struct seq_list seq_elem; + struct btrfs_delayed_ref_root *delayed_refs; - eb = path->nodes[0]; - ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); - item_size = btrfs_item_size_nr(eb, path->slots[0]); + trans = btrfs_join_transaction(fs_info->extent_root); + if (IS_ERR(trans)) + return PTR_ERR(trans); - /* first we iterate the inline refs, ... */ - do { - last = __get_extent_inline_ref(&ptr, eb, ei, item_size, - &eiref, &type); - if (last == -ENOENT) { - ret = 0; + pr_debug("resolving all inodes for extent %llu\n", + extent_item_objectid); + + delayed_refs = &trans->transaction->delayed_refs; + spin_lock(&delayed_refs->lock); + btrfs_get_delayed_seq(delayed_refs, &seq_elem); + spin_unlock(&delayed_refs->lock); + + ret = btrfs_find_all_leafs(trans, fs_info, extent_item_objectid, + extent_item_pos, seq_elem.seq, + &refs); + + if (ret) + goto out; + + while (!ret && (ref_node = ulist_next(refs, ref_node))) { + ret = btrfs_find_all_roots(trans, fs_info, ref_node->val, -1, + seq_elem.seq, &roots); + if (ret) break; - } - if (last < 0) { - ret = last; - break; - } - - if (type == BTRFS_EXTENT_DATA_REF_KEY) { - dref = (struct btrfs_extent_data_ref *)(&eiref->offset); - ret = __data_list_add_eb(&data_refs, eb, dref); - } else if (type == BTRFS_SHARED_DATA_REF_KEY) { - logical = btrfs_extent_inline_ref_offset(eb, eiref); - ret = __shared_list_add(&shared_refs, logical); - } - } while (!ret && !last); - - /* ... then we proceed to in-tree references and ... */ - while (!ret) { - ++path->slots[0]; - if (path->slots[0] > btrfs_header_nritems(eb)) { - ret = btrfs_next_leaf(fs_info->extent_root, path); - if (ret) { - if (ret == 1) - ret = 0; /* we're done */ - break; - } - eb = path->nodes[0]; - } - btrfs_item_key_to_cpu(eb, &key, path->slots[0]); - if (key.objectid != extent_item_objectid) - break; - if (key.type == BTRFS_EXTENT_DATA_REF_KEY) { - dref = btrfs_item_ptr(eb, path->slots[0], - struct btrfs_extent_data_ref); - ret = __data_list_add_eb(&data_refs, eb, dref); - } else if (key.type == BTRFS_SHARED_DATA_REF_KEY) { - ret = __shared_list_add(&shared_refs, key.offset); + while (!ret && (root_node = ulist_next(roots, root_node))) { + pr_debug("root %llu references leaf %llu\n", + root_node->val, ref_node->val); + ret = iterate_leaf_refs(fs_info, path, ref_node->val, + extent_item_objectid, + extent_item_pos, root_node->val, + iterate, ctx); } } - btrfs_release_path(path); - - /* - * ... only at the very end we can process the refs we found. this is - * because the iterator function we call is allowed to make tree lookups - * and we have to avoid deadlocks. additionally, we need more tree - * lookups ourselves for shared data refs. - */ - while (!list_empty(&data_refs)) { - ref_d = list_first_entry(&data_refs, struct __data_ref, list); - list_del(&ref_d->list); - if (!ret) - ret = iterate(ref_d->inum, extent_offset + - ref_d->extent_data_item_offset, - ref_d->root, ctx); - kfree(ref_d); - } - - while (!list_empty(&shared_refs)) { - ref_s = list_first_entry(&shared_refs, struct __shared_ref, - list); - list_del(&ref_s->list); - if (!ret) - ret = __iter_shared_inline_ref(fs_info, - ref_s->disk_byte, - extent_item_objectid, - extent_offset, path, - &data_refs, - iterate, ctx); - kfree(ref_s); - } - + ulist_free(refs); + ulist_free(roots); +out: + btrfs_put_delayed_seq(delayed_refs, &seq_elem); + btrfs_end_transaction(trans, fs_info->extent_root); return ret; } @@ -1369,19 +1202,20 @@ int iterate_inodes_from_logical(u64 logical, struct btrfs_fs_info *fs_info, iterate_extent_inodes_t *iterate, void *ctx) { int ret; - u64 offset; + u64 extent_item_pos; struct btrfs_key found_key; ret = extent_from_logical(fs_info, logical, path, &found_key); + btrfs_release_path(path); if (ret & BTRFS_EXTENT_FLAG_TREE_BLOCK) ret = -EINVAL; if (ret < 0) return ret; - offset = logical - found_key.objectid; + extent_item_pos = logical - found_key.objectid; ret = iterate_extent_inodes(fs_info, path, found_key.objectid, - offset, iterate, ctx); + extent_item_pos, iterate, ctx); return ret; } @@ -1426,6 +1260,10 @@ static int iterate_irefs(u64 inum, struct btrfs_root *fs_root, for (cur = 0; cur < btrfs_item_size(eb, item); cur += len) { name_len = btrfs_inode_ref_name_len(eb, iref); /* path must be released before calling iterate()! */ + pr_debug("following ref at offset %u for inode %llu in " + "tree %llu\n", cur, + (unsigned long long)found_key.objectid, + (unsigned long long)fs_root->objectid); ret = iterate(parent, iref, eb, ctx); if (ret) { free_extent_buffer(eb); @@ -1466,10 +1304,14 @@ static int inode_to_path(u64 inum, struct btrfs_inode_ref *iref, return PTR_ERR(fspath); if (fspath > fspath_min) { + pr_debug("path resolved: %s\n", fspath); ipath->fspath->val[i] = (u64)(unsigned long)fspath; ++ipath->fspath->elem_cnt; ipath->fspath->bytes_left = fspath - fspath_min; } else { + pr_debug("missed path, not enough space. missing bytes: %lu, " + "constructed so far: %s\n", + (unsigned long)(fspath_min - fspath), fspath_min); ++ipath->fspath->elem_missed; ipath->fspath->bytes_missing += fspath_min - fspath; ipath->fspath->bytes_left = 0; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index c48f2e931ea0..9b0526872b7b 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2976,7 +2976,7 @@ static long btrfs_ioctl_logical_to_ino(struct btrfs_root *root, { int ret = 0; int size; - u64 extent_offset; + u64 extent_item_pos; struct btrfs_ioctl_logical_ino_args *loi; struct btrfs_data_container *inodes = NULL; struct btrfs_path *path = NULL; @@ -3007,15 +3007,17 @@ static long btrfs_ioctl_logical_to_ino(struct btrfs_root *root, } ret = extent_from_logical(root->fs_info, loi->logical, path, &key); + btrfs_release_path(path); if (ret & BTRFS_EXTENT_FLAG_TREE_BLOCK) ret = -ENOENT; if (ret < 0) goto out; - extent_offset = loi->logical - key.objectid; + extent_item_pos = loi->logical - key.objectid; ret = iterate_extent_inodes(root->fs_info, path, key.objectid, - extent_offset, build_ino_list, inodes); + extent_item_pos, build_ino_list, + inodes); if (ret < 0) goto out; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index c27bcb67f330..b5edff25a53f 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -309,7 +309,7 @@ static void scrub_print_warning(const char *errstr, struct scrub_bio *sbio, u8 ref_level; unsigned long ptr = 0; const int bufsize = 4096; - u64 extent_offset; + u64 extent_item_pos; path = btrfs_alloc_path(); @@ -329,12 +329,13 @@ static void scrub_print_warning(const char *errstr, struct scrub_bio *sbio, if (ret < 0) goto out; - extent_offset = swarn.logical - found_key.objectid; + extent_item_pos = swarn.logical - found_key.objectid; swarn.extent_item_size = found_key.offset; eb = path->nodes[0]; ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); item_size = btrfs_item_size_nr(eb, path->slots[0]); + btrfs_release_path(path); if (ret & BTRFS_EXTENT_FLAG_TREE_BLOCK) { do { @@ -351,7 +352,7 @@ static void scrub_print_warning(const char *errstr, struct scrub_bio *sbio, } else { swarn.path = path; iterate_extent_inodes(fs_info, path, found_key.objectid, - extent_offset, + extent_item_pos, scrub_print_warning_inode, &swarn); } From 6bf7e080d5bcb0d399ee38ce3dabbfad64448192 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Thu, 1 Dec 2011 14:35:19 +0100 Subject: [PATCH 192/820] Btrfs: make sure we're not using obsolete code in btrfs_get_extent There's code in btrfs_get_extent that should never be used. This patch turns a WARN_ON(1) into a BUG(), hoping we can remove the transaction code from btrfs_get_extent soon. Signed-off-by: Jan Schmidt --- fs/btrfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index ea819386b864..603d740f0f1c 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5022,7 +5022,7 @@ again: } flush_dcache_page(page); } else if (create && PageUptodate(page)) { - WARN_ON(1); + BUG(); if (!trans) { kunmap(page); free_extent_map(em); From f69f2e264f6388df6d3cae45dd67ddfd52aaa14b Mon Sep 17 00:00:00 2001 From: Haitao Zhang Date: Sun, 1 Jan 2012 11:30:06 +0800 Subject: [PATCH 193/820] dma/imx-sdma: save irq flags when use spin_lock in sdma_tx_submit mx53_loco: fix deadlock report from sdma_tx_submit() during boot BugLink: http://bugs.launchpad.net/bugs/878701 Adjust to use spin_lock_irqsave()/spin_unlock_irqresotre(), so to make it safe when called from interrupt context. Signed-off-by: Haitao Zhang Signed-off-by: Eric Miao Signed-off-by: Richard Zhao Acked-by: Sascha Hauer Signed-off-by: Vinod Koul --- drivers/dma/imx-sdma.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c index f59fd8fffa88..a8af379680c1 100644 --- a/drivers/dma/imx-sdma.c +++ b/drivers/dma/imx-sdma.c @@ -836,17 +836,18 @@ static struct sdma_channel *to_sdma_chan(struct dma_chan *chan) static dma_cookie_t sdma_tx_submit(struct dma_async_tx_descriptor *tx) { + unsigned long flags; struct sdma_channel *sdmac = to_sdma_chan(tx->chan); struct sdma_engine *sdma = sdmac->sdma; dma_cookie_t cookie; - spin_lock_irq(&sdmac->lock); + spin_lock_irqsave(&sdmac->lock, flags); cookie = sdma_assign_cookie(sdmac); sdma_enable_channel(sdma, sdmac->channel); - spin_unlock_irq(&sdmac->lock); + spin_unlock_irqrestore(&sdmac->lock, flags); return cookie; } From c11b46c32c8a9bf05fdb76d70d8dc74fcbfd02d1 Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Wed, 4 Jan 2012 15:34:17 +0100 Subject: [PATCH 194/820] dma: shdma: fix runtime PM: clear channel buffers on reset On platforms, supporting power domains, if the domain, containing a DMAC instance is powered down, the driver fails to resume correctly. On those platforms DMAC channels have an additional CHCLR register for clearing channel buffers. Using this register during runtime resume fixes the problem. Signed-off-by: Guennadi Liakhovetski Signed-off-by: Vinod Koul --- drivers/dma/shdma.c | 47 ++++++++++++++++++++++++++++-------------- include/linux/sh_dma.h | 2 ++ 2 files changed, 33 insertions(+), 16 deletions(-) diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c index 592304fb41a6..54043cd831c8 100644 --- a/drivers/dma/shdma.c +++ b/drivers/dma/shdma.c @@ -56,6 +56,15 @@ static LIST_HEAD(sh_dmae_devices); static unsigned long sh_dmae_slave_used[BITS_TO_LONGS(SH_DMA_SLAVE_NUMBER)]; static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan, bool all); +static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan); + +static void chclr_write(struct sh_dmae_chan *sh_dc, u32 data) +{ + struct sh_dmae_device *shdev = to_sh_dev(sh_dc); + + __raw_writel(data, shdev->chan_reg + + shdev->pdata->channel[sh_dc->id].chclr_offset); +} static void sh_dmae_writel(struct sh_dmae_chan *sh_dc, u32 data, u32 reg) { @@ -128,6 +137,15 @@ static int sh_dmae_rst(struct sh_dmae_device *shdev) dmaor = dmaor_read(shdev) & ~(DMAOR_NMIF | DMAOR_AE | DMAOR_DME); + if (shdev->pdata->chclr_present) { + int i; + for (i = 0; i < shdev->pdata->channel_num; i++) { + struct sh_dmae_chan *sh_chan = shdev->chan[i]; + if (sh_chan) + chclr_write(sh_chan, 0); + } + } + dmaor_write(shdev, dmaor | shdev->pdata->dmaor_init); dmaor = dmaor_read(shdev); @@ -138,6 +156,10 @@ static int sh_dmae_rst(struct sh_dmae_device *shdev) dev_warn(shdev->common.dev, "Can't initialize DMAOR.\n"); return -EIO; } + if (shdev->pdata->dmaor_init & ~dmaor) + dev_warn(shdev->common.dev, + "DMAOR=0x%x hasn't latched the initial value 0x%x.\n", + dmaor, shdev->pdata->dmaor_init); return 0; } @@ -258,8 +280,6 @@ static int dmae_set_dmars(struct sh_dmae_chan *sh_chan, u16 val) return 0; } -static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan); - static dma_cookie_t sh_dmae_tx_submit(struct dma_async_tx_descriptor *tx) { struct sh_desc *desc = tx_to_sh_desc(tx), *chunk, *last = desc, *c; @@ -339,6 +359,8 @@ static dma_cookie_t sh_dmae_tx_submit(struct dma_async_tx_descriptor *tx) sh_chan_xfer_ld_queue(sh_chan); sh_chan->pm_state = DMAE_PM_ESTABLISHED; } + } else { + sh_chan->pm_state = DMAE_PM_PENDING; } spin_unlock_irq(&sh_chan->desc_lock); @@ -1224,6 +1246,8 @@ static int __init sh_dmae_probe(struct platform_device *pdev) platform_set_drvdata(pdev, shdev); + shdev->common.dev = &pdev->dev; + pm_runtime_enable(&pdev->dev); pm_runtime_get_sync(&pdev->dev); @@ -1253,7 +1277,6 @@ static int __init sh_dmae_probe(struct platform_device *pdev) shdev->common.device_prep_slave_sg = sh_dmae_prep_slave_sg; shdev->common.device_control = sh_dmae_control; - shdev->common.dev = &pdev->dev; /* Default transfer size of 32 bytes requires 32-byte alignment */ shdev->common.copy_align = LOG2_DEFAULT_XFER_SIZE; @@ -1434,22 +1457,17 @@ static int sh_dmae_runtime_resume(struct device *dev) #ifdef CONFIG_PM static int sh_dmae_suspend(struct device *dev) { - struct sh_dmae_device *shdev = dev_get_drvdata(dev); - int i; - - for (i = 0; i < shdev->pdata->channel_num; i++) { - struct sh_dmae_chan *sh_chan = shdev->chan[i]; - if (sh_chan->descs_allocated) - sh_chan->pm_error = pm_runtime_put_sync(dev); - } - return 0; } static int sh_dmae_resume(struct device *dev) { struct sh_dmae_device *shdev = dev_get_drvdata(dev); - int i; + int i, ret; + + ret = sh_dmae_rst(shdev); + if (ret < 0) + dev_err(dev, "Failed to reset!\n"); for (i = 0; i < shdev->pdata->channel_num; i++) { struct sh_dmae_chan *sh_chan = shdev->chan[i]; @@ -1458,9 +1476,6 @@ static int sh_dmae_resume(struct device *dev) if (!sh_chan->descs_allocated) continue; - if (!sh_chan->pm_error) - pm_runtime_get_sync(dev); - if (param) { const struct sh_dmae_slave_config *cfg = param->config; dmae_set_dmars(sh_chan, cfg->mid_rid); diff --git a/include/linux/sh_dma.h b/include/linux/sh_dma.h index 62ef6938da10..8cd7fe59cf1a 100644 --- a/include/linux/sh_dma.h +++ b/include/linux/sh_dma.h @@ -48,6 +48,7 @@ struct sh_dmae_channel { unsigned int offset; unsigned int dmars; unsigned int dmars_bit; + unsigned int chclr_offset; }; struct sh_dmae_pdata { @@ -68,6 +69,7 @@ struct sh_dmae_pdata { unsigned int dmaor_is_32bit:1; unsigned int needs_tend_set:1; unsigned int no_dmars:1; + unsigned int chclr_present:1; }; /* DMA register */ From e08b881a69d638175bfa99b5af4d72b731633ea7 Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Wed, 4 Jan 2012 15:34:21 +0100 Subject: [PATCH 195/820] ARM: mach-shmobile: specify CHCLR registers on SH7372 Signed-off-by: Guennadi Liakhovetski Signed-off-by: Vinod Koul --- arch/arm/mach-shmobile/setup-sh7372.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/arch/arm/mach-shmobile/setup-sh7372.c b/arch/arm/mach-shmobile/setup-sh7372.c index 2380389e6ac5..83dc940a21b7 100644 --- a/arch/arm/mach-shmobile/setup-sh7372.c +++ b/arch/arm/mach-shmobile/setup-sh7372.c @@ -445,31 +445,39 @@ static const struct sh_dmae_slave_config sh7372_dmae_slaves[] = { }, }; +#define SH7372_CHCLR 0x220 + static const struct sh_dmae_channel sh7372_dmae_channels[] = { { .offset = 0, .dmars = 0, .dmars_bit = 0, + .chclr_offset = SH7372_CHCLR + 0, }, { .offset = 0x10, .dmars = 0, .dmars_bit = 8, + .chclr_offset = SH7372_CHCLR + 0x10, }, { .offset = 0x20, .dmars = 4, .dmars_bit = 0, + .chclr_offset = SH7372_CHCLR + 0x20, }, { .offset = 0x30, .dmars = 4, .dmars_bit = 8, + .chclr_offset = SH7372_CHCLR + 0x30, }, { .offset = 0x50, .dmars = 8, .dmars_bit = 0, + .chclr_offset = SH7372_CHCLR + 0x50, }, { .offset = 0x60, .dmars = 8, .dmars_bit = 8, + .chclr_offset = SH7372_CHCLR + 0x60, } }; @@ -487,6 +495,7 @@ static struct sh_dmae_pdata dma_platform_data = { .ts_shift = ts_shift, .ts_shift_num = ARRAY_SIZE(ts_shift), .dmaor_init = DMAOR_DME, + .chclr_present = 1, }; /* Resource order important! */ @@ -494,7 +503,7 @@ static struct resource sh7372_dmae0_resources[] = { { /* Channel registers and DMAOR */ .start = 0xfe008020, - .end = 0xfe00808f, + .end = 0xfe00828f, .flags = IORESOURCE_MEM, }, { @@ -522,7 +531,7 @@ static struct resource sh7372_dmae1_resources[] = { { /* Channel registers and DMAOR */ .start = 0xfe018020, - .end = 0xfe01808f, + .end = 0xfe01828f, .flags = IORESOURCE_MEM, }, { @@ -550,7 +559,7 @@ static struct resource sh7372_dmae2_resources[] = { { /* Channel registers and DMAOR */ .start = 0xfe028020, - .end = 0xfe02808f, + .end = 0xfe02828f, .flags = IORESOURCE_MEM, }, { From 5e1a0f9f8d604c21fd72a09b632e3cd9ae66d3a6 Mon Sep 17 00:00:00 2001 From: David Brown Date: Thu, 5 Jan 2012 15:55:46 -0800 Subject: [PATCH 196/820] ARM: msm: vreg is a module and so needs module.h The MSM7201 vreg appears to have gotten missed in the module.h cleanup. Signed-off-by: David Brown --- arch/arm/mach-msm/vreg.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm/mach-msm/vreg.c b/arch/arm/mach-msm/vreg.c index a9103bc6615f..bd66ed04d6dc 100644 --- a/arch/arm/mach-msm/vreg.c +++ b/arch/arm/mach-msm/vreg.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include From 203bf287cb01a5dc26c20bd3737cecf3aeba1d48 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 6 Jan 2012 15:23:57 -0500 Subject: [PATCH 197/820] Btrfs: run chunk allocations while we do delayed refs Btrfs tries to batch extent allocation tree changes to improve performance and reduce metadata trashing. But it doesn't allocate new metadata chunks while it is doing allocations for the extent allocation tree. This commit changes the delayed refence code to do chunk allocations if we're getting low on room. It prevents crashes and improves performance. Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 14 ++++++++++---- fs/btrfs/transaction.c | 9 +-------- 2 files changed, 11 insertions(+), 12 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index f5fbe576d2ba..71549d11a09e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2267,9 +2267,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, BUG_ON(ret); kfree(extent_op); - cond_resched(); - spin_lock(&delayed_refs->lock); - continue; + goto next; } list_del_init(&locked_ref->cluster); @@ -2289,7 +2287,11 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, btrfs_put_delayed_ref(ref); kfree(extent_op); count++; - +next: + do_chunk_alloc(trans, root->fs_info->extent_root, + 2 * 1024 * 1024, + btrfs_get_alloc_profile(root, 0), + CHUNK_ALLOC_NO_FORCE); cond_resched(); spin_lock(&delayed_refs->lock); } @@ -2317,6 +2319,10 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, if (root == root->fs_info->extent_root) root = root->fs_info->tree_root; + do_chunk_alloc(trans, root->fs_info->extent_root, + 2 * 1024 * 1024, btrfs_get_alloc_profile(root, 0), + CHUNK_ALLOC_NO_FORCE); + delayed_refs = &trans->transaction->delayed_refs; INIT_LIST_HEAD(&cluster); again: diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 81376d94cd3c..360c2dfd1ee6 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -467,19 +467,12 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, btrfs_trans_release_metadata(trans, root); trans->block_rsv = NULL; - while (count < 4) { + while (count < 2) { unsigned long cur = trans->delayed_ref_updates; trans->delayed_ref_updates = 0; if (cur && trans->transaction->delayed_refs.num_heads_ready > 64) { trans->delayed_ref_updates = 0; - - /* - * do a full flush if the transaction is trying - * to close - */ - if (trans->transaction->delayed_refs.flushing) - cur = 0; btrfs_run_delayed_refs(trans, root, cur); } else { break; From cf1d72c9ceec391d34c48724da57282e97f01122 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 6 Jan 2012 15:41:34 -0500 Subject: [PATCH 198/820] Btrfs: lower the bar for chunk allocation The chunk allocation code has tried to keep a pretty tight lid on creating new metadata chunks. This is partially because in the past the reservation code didn't give us an accurate idea of how much space was being used. The new code is much more accurate, so we're able to get rid of some of these checks. Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 21 +++------------------ 1 file changed, 3 insertions(+), 18 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 71549d11a09e..247d2c94f8ec 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3263,27 +3263,12 @@ static int should_alloc_chunk(struct btrfs_root *root, if (num_bytes - num_allocated < thresh) return 1; } - - /* - * we have two similar checks here, one based on percentage - * and once based on a hard number of 256MB. The idea - * is that if we have a good amount of free - * room, don't allocate a chunk. A good mount is - * less than 80% utilized of the chunks we have allocated, - * or more than 256MB free - */ - if (num_allocated + alloc_bytes + 256 * 1024 * 1024 < num_bytes) - return 0; - - if (num_allocated + alloc_bytes < div_factor(num_bytes, 8)) - return 0; - thresh = btrfs_super_total_bytes(root->fs_info->super_copy); - /* 256MB or 5% of the FS */ - thresh = max_t(u64, 256 * 1024 * 1024, div_factor_fine(thresh, 5)); + /* 256MB or 2% of the FS */ + thresh = max_t(u64, 256 * 1024 * 1024, div_factor_fine(thresh, 2)); - if (num_bytes > thresh && sinfo->bytes_used < div_factor(num_bytes, 3)) + if (num_bytes > thresh && sinfo->bytes_used < div_factor(num_bytes, 8)) return 0; return 1; } From 1100373f8aa69e377386499350496e3d8565605f Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 6 Jan 2012 15:47:38 -0500 Subject: [PATCH 199/820] Btrfs: use bigger metadata chunks on bigger filesystems The 256MB chunk is a little small on a huge FS. This scales up the chunk size. Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f4b839fd3c9d..ac00e3aa80a1 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2441,7 +2441,11 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, max_stripe_size = 1024 * 1024 * 1024; max_chunk_size = 10 * max_stripe_size; } else if (type & BTRFS_BLOCK_GROUP_METADATA) { - max_stripe_size = 256 * 1024 * 1024; + /* for larger filesystems, use larger metadata chunks */ + if (fs_devices->total_rw_bytes > 50ULL * 1024 * 1024 * 1024) + max_stripe_size = 1024 * 1024 * 1024; + else + max_stripe_size = 256 * 1024 * 1024; max_chunk_size = max_stripe_size; } else if (type & BTRFS_BLOCK_GROUP_SYSTEM) { max_stripe_size = 8 * 1024 * 1024; From a5f6f719a5cd7caeee8ed8137cf3f94c3bbebc65 Mon Sep 17 00:00:00 2001 From: Alexandre Oliva Date: Mon, 12 Dec 2011 04:48:19 -0200 Subject: [PATCH 200/820] Btrfs: test free space only for unclustered allocation Since the clustered allocation may be taking extents from a different block group, there's no point in spin-locking and testing the current block group free space before attempting to allocate space from a cluster, even more so when we might refrain from even trying the cluster in the current block group because, after the cluster was set up, not enough free space remained. Furthermore, cluster creation attempts fail fast when the block group doesn't have enough free space, so the test was completely superfluous. I've move the free space test past the cluster allocation attempt, where it is more useful, and arranged for a cluster in the current block group to be released before trying an unclustered allocation, when we reach the LOOP_NO_EMPTY_SIZE stage, so that the free space in the cluster stands a chance of being combined with additional free space in the block group so as to succeed in the allocation attempt. Signed-off-by: Alexandre Oliva Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 34 +++++++++++++++++++++++----------- 1 file changed, 23 insertions(+), 11 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 247d2c94f8ec..5ea3acc53241 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5286,15 +5286,6 @@ alloc: if (unlikely(block_group->ro)) goto loop; - spin_lock(&block_group->free_space_ctl->tree_lock); - if (cached && - block_group->free_space_ctl->free_space < - num_bytes + empty_cluster + empty_size) { - spin_unlock(&block_group->free_space_ctl->tree_lock); - goto loop; - } - spin_unlock(&block_group->free_space_ctl->tree_lock); - /* * Ok we want to try and use the cluster allocator, so * lets look there @@ -5340,8 +5331,15 @@ refill_cluster: * plenty of times and not have found * anything, so we are likely way too * fragmented for the clustering stuff to find - * anything. */ - if (loop >= LOOP_NO_EMPTY_SIZE) { + * anything. + * + * However, if the cluster is taken from the + * current block group, release the cluster + * first, so that we stand a better chance of + * succeeding in the unclustered + * allocation. */ + if (loop >= LOOP_NO_EMPTY_SIZE && + last_ptr->block_group != block_group) { spin_unlock(&last_ptr->refill_lock); goto unclustered_alloc; } @@ -5352,6 +5350,11 @@ refill_cluster: */ btrfs_return_cluster_to_free_space(NULL, last_ptr); + if (loop >= LOOP_NO_EMPTY_SIZE) { + spin_unlock(&last_ptr->refill_lock); + goto unclustered_alloc; + } + /* allocate a cluster in this block group */ ret = btrfs_find_space_cluster(trans, root, block_group, last_ptr, @@ -5392,6 +5395,15 @@ refill_cluster: } unclustered_alloc: + spin_lock(&block_group->free_space_ctl->tree_lock); + if (cached && + block_group->free_space_ctl->free_space < + num_bytes + empty_cluster + empty_size) { + spin_unlock(&block_group->free_space_ctl->tree_lock); + goto loop; + } + spin_unlock(&block_group->free_space_ctl->tree_lock); + offset = btrfs_find_space_for_alloc(block_group, search_start, num_bytes, empty_size); /* From fc7c1077ceb99c35e5f9d0ce03dc7740565bb2bf Mon Sep 17 00:00:00 2001 From: Alexandre Oliva Date: Mon, 28 Nov 2011 12:36:17 -0200 Subject: [PATCH 201/820] Btrfs: don't set up allocation result twice We store the allocation start and length twice in ins, once right after the other, but with intervening calls that may prevent the duplicate from being optimized out by the compiler. Remove one of the assignments. Signed-off-by: Alexandre Oliva Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 5ea3acc53241..37594e4bf660 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5441,9 +5441,6 @@ checks: goto loop; } - ins->objectid = search_start; - ins->offset = num_bytes; - if (offset < search_start) btrfs_add_free_space(used_block_group, offset, search_start - offset); From 1bb91902dc90e25449893e693ad45605cb08fbe5 Mon Sep 17 00:00:00 2001 From: Alexandre Oliva Date: Fri, 14 Oct 2011 12:10:36 -0300 Subject: [PATCH 202/820] Btrfs: revamp clustered allocation logic Parameterize clusters on minimum total size, minimum chunk size and minimum contiguous size for at least one chunk, without limits on cluster, window or gap sizes. Don't tolerate any fragmentation for SSD_SPREAD; accept it for metadata, but try to keep data dense. Signed-off-by: Alexandre Oliva Signed-off-by: Chris Mason --- fs/btrfs/free-space-cache.c | 112 ++++++++++++++++-------------------- 1 file changed, 49 insertions(+), 63 deletions(-) diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index ec23d43d0c35..ce40db59c706 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -2283,23 +2283,23 @@ out: static int btrfs_bitmap_cluster(struct btrfs_block_group_cache *block_group, struct btrfs_free_space *entry, struct btrfs_free_cluster *cluster, - u64 offset, u64 bytes, u64 min_bytes) + u64 offset, u64 bytes, + u64 cont1_bytes, u64 min_bytes) { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; unsigned long next_zero; unsigned long i; - unsigned long search_bits; - unsigned long total_bits; + unsigned long want_bits; + unsigned long min_bits; unsigned long found_bits; unsigned long start = 0; unsigned long total_found = 0; int ret; - bool found = false; i = offset_to_bit(entry->offset, block_group->sectorsize, max_t(u64, offset, entry->offset)); - search_bits = bytes_to_bits(bytes, block_group->sectorsize); - total_bits = bytes_to_bits(min_bytes, block_group->sectorsize); + want_bits = bytes_to_bits(bytes, block_group->sectorsize); + min_bits = bytes_to_bits(min_bytes, block_group->sectorsize); again: found_bits = 0; @@ -2308,7 +2308,7 @@ again: i = find_next_bit(entry->bitmap, BITS_PER_BITMAP, i + 1)) { next_zero = find_next_zero_bit(entry->bitmap, BITS_PER_BITMAP, i); - if (next_zero - i >= search_bits) { + if (next_zero - i >= min_bits) { found_bits = next_zero - i; break; } @@ -2318,10 +2318,9 @@ again: if (!found_bits) return -ENOSPC; - if (!found) { + if (!total_found) { start = i; cluster->max_size = 0; - found = true; } total_found += found_bits; @@ -2329,13 +2328,8 @@ again: if (cluster->max_size < found_bits * block_group->sectorsize) cluster->max_size = found_bits * block_group->sectorsize; - if (total_found < total_bits) { - i = find_next_bit(entry->bitmap, BITS_PER_BITMAP, next_zero); - if (i - start > total_bits * 2) { - total_found = 0; - cluster->max_size = 0; - found = false; - } + if (total_found < want_bits || cluster->max_size < cont1_bytes) { + i = next_zero + 1; goto again; } @@ -2351,23 +2345,23 @@ again: /* * This searches the block group for just extents to fill the cluster with. + * Try to find a cluster with at least bytes total bytes, at least one + * extent of cont1_bytes, and other clusters of at least min_bytes. */ static noinline int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster, struct list_head *bitmaps, u64 offset, u64 bytes, - u64 min_bytes) + u64 cont1_bytes, u64 min_bytes) { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *first = NULL; struct btrfs_free_space *entry = NULL; - struct btrfs_free_space *prev = NULL; struct btrfs_free_space *last; struct rb_node *node; u64 window_start; u64 window_free; u64 max_extent; - u64 max_gap = 128 * 1024; entry = tree_search_offset(ctl, offset, 0, 1); if (!entry) @@ -2377,8 +2371,8 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, * We don't want bitmaps, so just move along until we find a normal * extent entry. */ - while (entry->bitmap) { - if (list_empty(&entry->list)) + while (entry->bitmap || entry->bytes < min_bytes) { + if (entry->bitmap && list_empty(&entry->list)) list_add_tail(&entry->list, bitmaps); node = rb_next(&entry->offset_index); if (!node) @@ -2391,12 +2385,9 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, max_extent = entry->bytes; first = entry; last = entry; - prev = entry; - while (window_free <= min_bytes) { - node = rb_next(&entry->offset_index); - if (!node) - return -ENOSPC; + for (node = rb_next(&entry->offset_index); node; + node = rb_next(&entry->offset_index)) { entry = rb_entry(node, struct btrfs_free_space, offset_index); if (entry->bitmap) { @@ -2405,26 +2396,18 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, continue; } - /* - * we haven't filled the empty size and the window is - * very large. reset and try again - */ - if (entry->offset - (prev->offset + prev->bytes) > max_gap || - entry->offset - window_start > (min_bytes * 2)) { - first = entry; - window_start = entry->offset; - window_free = entry->bytes; - last = entry; + if (entry->bytes < min_bytes) + continue; + + last = entry; + window_free += entry->bytes; + if (entry->bytes > max_extent) max_extent = entry->bytes; - } else { - last = entry; - window_free += entry->bytes; - if (entry->bytes > max_extent) - max_extent = entry->bytes; - } - prev = entry; } + if (window_free < bytes || max_extent < cont1_bytes) + return -ENOSPC; + cluster->window_start = first->offset; node = &first->offset_index; @@ -2438,7 +2421,7 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, entry = rb_entry(node, struct btrfs_free_space, offset_index); node = rb_next(&entry->offset_index); - if (entry->bitmap) + if (entry->bitmap || entry->bytes < min_bytes) continue; rb_erase(&entry->offset_index, &ctl->free_space_offset); @@ -2460,7 +2443,7 @@ static noinline int setup_cluster_bitmap(struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster, struct list_head *bitmaps, u64 offset, u64 bytes, - u64 min_bytes) + u64 cont1_bytes, u64 min_bytes) { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry; @@ -2485,7 +2468,7 @@ setup_cluster_bitmap(struct btrfs_block_group_cache *block_group, if (entry->bytes < min_bytes) continue; ret = btrfs_bitmap_cluster(block_group, entry, cluster, offset, - bytes, min_bytes); + bytes, cont1_bytes, min_bytes); if (!ret) return 0; } @@ -2499,7 +2482,7 @@ setup_cluster_bitmap(struct btrfs_block_group_cache *block_group, /* * here we try to find a cluster of blocks in a block group. The goal - * is to find at least bytes free and up to empty_size + bytes free. + * is to find at least bytes+empty_size. * We might not find them all in one contiguous area. * * returns zero and sets up cluster if things worked out, otherwise @@ -2515,23 +2498,24 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, struct btrfs_free_space *entry, *tmp; LIST_HEAD(bitmaps); u64 min_bytes; + u64 cont1_bytes; int ret; - /* for metadata, allow allocates with more holes */ + /* + * Choose the minimum extent size we'll require for this + * cluster. For SSD_SPREAD, don't allow any fragmentation. + * For metadata, allow allocates with smaller extents. For + * data, keep it dense. + */ if (btrfs_test_opt(root, SSD_SPREAD)) { - min_bytes = bytes + empty_size; + cont1_bytes = min_bytes = bytes + empty_size; } else if (block_group->flags & BTRFS_BLOCK_GROUP_METADATA) { - /* - * we want to do larger allocations when we are - * flushing out the delayed refs, it helps prevent - * making more work as we go along. - */ - if (trans->transaction->delayed_refs.flushing) - min_bytes = max(bytes, (bytes + empty_size) >> 1); - else - min_bytes = max(bytes, (bytes + empty_size) >> 4); - } else - min_bytes = max(bytes, (bytes + empty_size) >> 2); + cont1_bytes = bytes; + min_bytes = block_group->sectorsize; + } else { + cont1_bytes = max(bytes, (bytes + empty_size) >> 2); + min_bytes = block_group->sectorsize; + } spin_lock(&ctl->tree_lock); @@ -2539,7 +2523,7 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, * If we know we don't have enough space to make a cluster don't even * bother doing all the work to try and find one. */ - if (ctl->free_space < min_bytes) { + if (ctl->free_space < bytes) { spin_unlock(&ctl->tree_lock); return -ENOSPC; } @@ -2553,10 +2537,12 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, } ret = setup_cluster_no_bitmap(block_group, cluster, &bitmaps, offset, - bytes, min_bytes); + bytes + empty_size, + cont1_bytes, min_bytes); if (ret) ret = setup_cluster_bitmap(block_group, cluster, &bitmaps, - offset, bytes, min_bytes); + offset, bytes + empty_size, + cont1_bytes, min_bytes); /* Clear our temporary list */ list_for_each_entry_safe(entry, tmp, &bitmaps, list) From 98c7089c769048f941bd5c5285287f8fc301f8b1 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:00:31 -0500 Subject: [PATCH 203/820] btrfs: preparation to fixing mount/umount race We need fs_info and root to live until the moment when the victim superblock leaves the list, so we need to postpone free_fs_info() until after ->put_super(). The call is buried in close_ctree(), though, so we need to lift it into the callers (including btrfs_put_super()) first. Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 3 +-- fs/btrfs/super.c | 6 +++++- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index f99a099a7747..dcb5d949b543 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2424,6 +2424,7 @@ retry_root_backup: up_read(&fs_info->cleanup_work_sem); if (err) { close_ctree(tree_root); + free_fs_info(fs_info); return ERR_PTR(err); } } @@ -3059,8 +3060,6 @@ int close_ctree(struct btrfs_root *root) bdi_destroy(&fs_info->bdi); cleanup_srcu_struct(&fs_info->subvol_srcu); - free_fs_info(fs_info); - return 0; } diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index ae488aa1966a..a3f435e58987 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -151,6 +151,7 @@ static void btrfs_put_super(struct super_block *sb) int ret; ret = close_ctree(root); + free_fs_info(root->fs_info); sb->s_fs_info = NULL; (void)ret; /* FIXME: need to fix VFS to return error? */ @@ -589,6 +590,7 @@ static int btrfs_fill_super(struct super_block *sb, struct inode *inode; struct dentry *root_dentry; struct btrfs_root *tree_root; + struct btrfs_fs_info *fs_info; struct btrfs_key key; int err; @@ -609,12 +611,13 @@ static int btrfs_fill_super(struct super_block *sb, printk("btrfs: open_ctree failed\n"); return PTR_ERR(tree_root); } + fs_info = tree_root->fs_info; sb->s_fs_info = tree_root; key.objectid = BTRFS_FIRST_FREE_OBJECTID; key.type = BTRFS_INODE_ITEM_KEY; key.offset = 0; - inode = btrfs_iget(sb, &key, tree_root->fs_info->fs_root, NULL); + inode = btrfs_iget(sb, &key, fs_info->fs_root, NULL); if (IS_ERR(inode)) { err = PTR_ERR(inode); goto fail_close; @@ -635,6 +638,7 @@ static int btrfs_fill_super(struct super_block *sb, fail_close: close_ctree(tree_root); + free_fs_info(fs_info); return err; } From aea52e19dd2085617dd7d247afb76a90cf2ea40c Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:22:46 -0500 Subject: [PATCH 204/820] btrfs: get ->kill_sb() of its own ... and move free_fs_info() to that, out of ->put_super(). Do NOT set ->s_fs_info to NULL in the latter; we need it for sget() to be able to see and wait for fs in the middle of umount if we get a mount/umount race. Signed-off-by: Al Viro --- fs/btrfs/super.c | 27 ++++++++++++++++++--------- 1 file changed, 18 insertions(+), 9 deletions(-) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index a3f435e58987..eca48624a4f0 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -147,14 +147,13 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, static void btrfs_put_super(struct super_block *sb) { - struct btrfs_root *root = btrfs_sb(sb); - int ret; - - ret = close_ctree(root); - free_fs_info(root->fs_info); - sb->s_fs_info = NULL; - - (void)ret; /* FIXME: need to fix VFS to return error? */ + (void)close_ctree(btrfs_sb(sb)); + /* FIXME: need to fix VFS to return error? */ + /* AV: return it _where_? ->put_super() can be triggered by any number + * of async events, up to and including delivery of SIGKILL to the + * last process that kept it busy. Or segfault in the aforementioned + * process... Whom would you report that to? + */ } enum { @@ -1223,11 +1222,21 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) return 0; } +static void btrfs_kill_super(struct super_block *sb) +{ + struct btrfs_fs_info *fs_info = NULL; + if (sb->s_root) + fs_info = btrfs_sb(sb)->fs_info; + kill_anon_super(sb); + if (fs_info) + free_fs_info(fs_info); +} + static struct file_system_type btrfs_fs_type = { .owner = THIS_MODULE, .name = "btrfs", .mount = btrfs_mount, - .kill_sb = kill_anon_super, + .kill_sb = btrfs_kill_super, .fs_flags = FS_REQUIRES_DEV, }; From 6de1d09d9677dce7a04ef485d426821159ab7de9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:29:09 -0500 Subject: [PATCH 205/820] btrfs: fix mount/umount race Do *NOT* skip doomed superblocks in btrfs_test_super(); we want sget() to wait for their shutdown to complete. Since we don't mutilate ->s_fs_info in ->put_super() anymore (or free what it used to point to until the superblock is past being findable by sget()), we can just DTRT there and report a match. Signed-off-by: Al Viro --- fs/btrfs/super.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index eca48624a4f0..b9fd62a0fca2 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -733,20 +733,15 @@ static int btrfs_test_super(struct super_block *s, void *data) struct btrfs_root *test_root = data; struct btrfs_root *root = btrfs_sb(s); - /* - * If this super block is going away, return false as it - * can't match as an existing super block. - */ - if (!atomic_read(&s->s_active)) - return 0; return root->fs_info->fs_devices == test_root->fs_info->fs_devices; } static int btrfs_set_super(struct super_block *s, void *data) { - s->s_fs_info = data; - - return set_anon_super(s, data); + int err = set_anon_super(s, data); + if (!err) + s->s_fs_info = data; + return err; } /* From 10f6327b5d26660e06120ba17cff993cb616b1d2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 15:05:22 -0500 Subject: [PATCH 206/820] btrfs: fix a deadlock in btrfs_scan_one_device() pathname resolution under a global mutex, taken on some paths in ->mount() is a Bad Idea(tm) - think what happens if said pathname resolution triggers automount of some btrfs instance and walks into attempt to grab the same mutex. Deadlock - we are waiting for daemon to finish walking the path, daemon is waiting for us to release the mutex... Signed-off-by: Al Viro --- fs/btrfs/volumes.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f4b839fd3c9d..96bb3c0a79b0 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -706,8 +706,6 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, u64 devid; u64 transid; - mutex_lock(&uuid_mutex); - flags |= FMODE_EXCL; bdev = blkdev_get_by_path(path, flags, holder); @@ -716,6 +714,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, goto error; } + mutex_lock(&uuid_mutex); ret = set_blocksize(bdev, 4096); if (ret) goto error_close; @@ -737,9 +736,9 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, brelse(bh); error_close: + mutex_unlock(&uuid_mutex); blkdev_put(bdev, flags); error: - mutex_unlock(&uuid_mutex); return ret; } From 2eb34cd368b3e5cef960ae9d9971a037cd6fa9d3 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 00:32:06 -0500 Subject: [PATCH 207/820] btrfs: sanitizing ->fs_info, part 1 take assignment of ->fs_info to callers of __setup_root() Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index dcb5d949b543..04eba10af825 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1142,7 +1142,6 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, root->orphan_item_inserted = 0; root->orphan_cleanup_state = 0; - root->fs_info = fs_info; root->objectid = objectid; root->last_trans = 0; root->highest_objectid = 0; @@ -1193,6 +1192,7 @@ static int find_and_setup_root(struct btrfs_root *tree_root, u32 blocksize; u64 generation; + root->fs_info = fs_info; __setup_root(tree_root->nodesize, tree_root->leafsize, tree_root->sectorsize, tree_root->stripesize, root, fs_info, objectid); @@ -1227,6 +1227,7 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, if (!root) return ERR_PTR(-ENOMEM); + root->fs_info = fs_info; __setup_root(tree_root->nodesize, tree_root->leafsize, tree_root->sectorsize, tree_root->stripesize, root, fs_info, BTRFS_TREE_LOG_OBJECTID); @@ -1330,6 +1331,7 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, goto out; } + root->fs_info = fs_info; __setup_root(tree_root->nodesize, tree_root->leafsize, tree_root->sectorsize, tree_root->stripesize, root, fs_info, location->objectid); @@ -2055,6 +2057,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, init_waitqueue_head(&fs_info->transaction_blocked_wait); init_waitqueue_head(&fs_info->async_submit_wait); + tree_root->fs_info = fs_info; __setup_root(4096, 4096, 4096, 4096, tree_root, fs_info, BTRFS_ROOT_TREE_OBJECTID); @@ -2247,6 +2250,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_super_chunk_root_level(disk_super)); generation = btrfs_super_chunk_root_generation(disk_super); + chunk_root->fs_info = fs_info; __setup_root(nodesize, leafsize, sectorsize, stripesize, chunk_root, fs_info, BTRFS_CHUNK_TREE_OBJECTID); @@ -2373,6 +2377,7 @@ retry_root_backup: goto fail_trans_kthread; } + log_tree_root->fs_info = fs_info; __setup_root(nodesize, leafsize, sectorsize, stripesize, log_tree_root, fs_info, BTRFS_TREE_LOG_OBJECTID); From 1233f546ec29ac32424327baf6ae5df81e3d9eae Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 00:34:00 -0500 Subject: [PATCH 208/820] btrfs: sanitizing ->fs_info, part 2 lift assignment to callers of find_and_setup_root() Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 04eba10af825..b8dae517f80c 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1192,7 +1192,6 @@ static int find_and_setup_root(struct btrfs_root *tree_root, u32 blocksize; u64 generation; - root->fs_info = fs_info; __setup_root(tree_root->nodesize, tree_root->leafsize, tree_root->sectorsize, tree_root->stripesize, root, fs_info, objectid); @@ -1322,6 +1321,7 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, if (!root) return ERR_PTR(-ENOMEM); if (location->offset == (u64)-1) { + root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, location->objectid, root); if (ret) { @@ -2300,18 +2300,21 @@ retry_root_backup: btrfs_set_root_node(&tree_root->root_item, tree_root->node); tree_root->commit_root = btrfs_root_node(tree_root); + extent_root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, BTRFS_EXTENT_TREE_OBJECTID, extent_root); if (ret) goto recovery_tree_root; extent_root->track_dirty = 1; + dev_root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, BTRFS_DEV_TREE_OBJECTID, dev_root); if (ret) goto recovery_tree_root; dev_root->track_dirty = 1; + csum_root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, BTRFS_CSUM_TREE_OBJECTID, csum_root); if (ret) From 38a77db49ad8f78369dcdfb693b8e5a818a60104 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 00:38:54 -0500 Subject: [PATCH 209/820] btrfs: sanitizing ->fs_info, part 3 move assignments to ->fs_info in open_ctree() up, to the place just after the original allocations. Assignment for tree_root becomes a no-op - we'd obtained fs_info from tree_root->fs_info in the first place. Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index b8dae517f80c..06480fcf0fb3 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1320,8 +1320,8 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, root = kzalloc(sizeof(*root), GFP_NOFS); if (!root) return ERR_PTR(-ENOMEM); + root->fs_info = fs_info; if (location->offset == (u64)-1) { - root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, location->objectid, root); if (ret) { @@ -1331,7 +1331,6 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, goto out; } - root->fs_info = fs_info; __setup_root(tree_root->nodesize, tree_root->leafsize, tree_root->sectorsize, tree_root->stripesize, root, fs_info, location->objectid); @@ -1914,6 +1913,10 @@ struct btrfs_root *open_ctree(struct super_block *sb, err = -ENOMEM; goto fail; } + chunk_root->fs_info = fs_info; + extent_root->fs_info = fs_info; + dev_root->fs_info = fs_info; + csum_root->fs_info = fs_info; ret = init_srcu_struct(&fs_info->subvol_srcu); if (ret) { @@ -2057,7 +2060,6 @@ struct btrfs_root *open_ctree(struct super_block *sb, init_waitqueue_head(&fs_info->transaction_blocked_wait); init_waitqueue_head(&fs_info->async_submit_wait); - tree_root->fs_info = fs_info; __setup_root(4096, 4096, 4096, 4096, tree_root, fs_info, BTRFS_ROOT_TREE_OBJECTID); @@ -2250,7 +2252,6 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_super_chunk_root_level(disk_super)); generation = btrfs_super_chunk_root_generation(disk_super); - chunk_root->fs_info = fs_info; __setup_root(nodesize, leafsize, sectorsize, stripesize, chunk_root, fs_info, BTRFS_CHUNK_TREE_OBJECTID); @@ -2300,21 +2301,18 @@ retry_root_backup: btrfs_set_root_node(&tree_root->root_item, tree_root->node); tree_root->commit_root = btrfs_root_node(tree_root); - extent_root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, BTRFS_EXTENT_TREE_OBJECTID, extent_root); if (ret) goto recovery_tree_root; extent_root->track_dirty = 1; - dev_root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, BTRFS_DEV_TREE_OBJECTID, dev_root); if (ret) goto recovery_tree_root; dev_root->track_dirty = 1; - csum_root->fs_info = fs_info; ret = find_and_setup_root(tree_root, fs_info, BTRFS_CSUM_TREE_OBJECTID, csum_root); if (ret) From 6f07e42ee6fcc252a210781d7262f4051e9fd8f6 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 00:46:16 -0500 Subject: [PATCH 210/820] btrfs: sanitizing ->fs_info, part 4 A new helper: btrfs_alloc_root(fs_info); allocates btrfs_root and sets ->fs_info. All places allocating the suckers converted to it. At that point we *never* reassign ->fs_info of btrfs_root; it's set before anyone sees the address of newly allocated struct btrfs_root and never assigned anywhere else. Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 33 +++++++++++++++------------------ fs/btrfs/disk-io.h | 2 ++ fs/btrfs/super.c | 3 +-- 3 files changed, 18 insertions(+), 20 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 06480fcf0fb3..ee846ce58846 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1215,6 +1215,14 @@ static int find_and_setup_root(struct btrfs_root *tree_root, return 0; } +struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info) +{ + struct btrfs_root *root = kzalloc(sizeof(*root), GFP_NOFS); + if (root) + root->fs_info = fs_info; + return root; +} + static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info) { @@ -1222,11 +1230,10 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, struct btrfs_root *tree_root = fs_info->tree_root; struct extent_buffer *leaf; - root = kzalloc(sizeof(*root), GFP_NOFS); + root = btrfs_alloc_root(fs_info); if (!root) return ERR_PTR(-ENOMEM); - root->fs_info = fs_info; __setup_root(tree_root->nodesize, tree_root->leafsize, tree_root->sectorsize, tree_root->stripesize, root, fs_info, BTRFS_TREE_LOG_OBJECTID); @@ -1317,10 +1324,9 @@ struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, u32 blocksize; int ret = 0; - root = kzalloc(sizeof(*root), GFP_NOFS); + root = btrfs_alloc_root(fs_info); if (!root) return ERR_PTR(-ENOMEM); - root->fs_info = fs_info; if (location->offset == (u64)-1) { ret = find_and_setup_root(tree_root, fs_info, location->objectid, root); @@ -1900,23 +1906,15 @@ struct btrfs_root *open_ctree(struct super_block *sb, int num_backups_tried = 0; int backup_index = 0; - extent_root = fs_info->extent_root = - kzalloc(sizeof(struct btrfs_root), GFP_NOFS); - csum_root = fs_info->csum_root = - kzalloc(sizeof(struct btrfs_root), GFP_NOFS); - chunk_root = fs_info->chunk_root = - kzalloc(sizeof(struct btrfs_root), GFP_NOFS); - dev_root = fs_info->dev_root = - kzalloc(sizeof(struct btrfs_root), GFP_NOFS); + extent_root = fs_info->extent_root = btrfs_alloc_root(fs_info); + csum_root = fs_info->csum_root = btrfs_alloc_root(fs_info); + chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info); + dev_root = fs_info->dev_root = btrfs_alloc_root(fs_info); if (!extent_root || !csum_root || !chunk_root || !dev_root) { err = -ENOMEM; goto fail; } - chunk_root->fs_info = fs_info; - extent_root->fs_info = fs_info; - dev_root->fs_info = fs_info; - csum_root->fs_info = fs_info; ret = init_srcu_struct(&fs_info->subvol_srcu); if (ret) { @@ -2372,13 +2370,12 @@ retry_root_backup: btrfs_level_size(tree_root, btrfs_super_log_root_level(disk_super)); - log_tree_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS); + log_tree_root = btrfs_alloc_root(fs_info); if (!log_tree_root) { err = -ENOMEM; goto fail_trans_kthread; } - log_tree_root->fs_info = fs_info; __setup_root(nodesize, leafsize, sectorsize, stripesize, log_tree_root, fs_info, BTRFS_TREE_LOG_OBJECTID); diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index c99d0a8f13fa..2bb5f59ddf95 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -86,6 +86,8 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans, int btrfs_add_log_tree(struct btrfs_trans_handle *trans, struct btrfs_root *root); +struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info); + #ifdef CONFIG_DEBUG_LOCK_ALLOC void btrfs_init_lockdep(void); void btrfs_set_buffer_lockdep_class(u64 objectid, diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index b9fd62a0fca2..e9f876a1655b 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -901,12 +901,11 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, if (!fs_info) return ERR_PTR(-ENOMEM); - fs_info->tree_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS); + fs_info->tree_root = btrfs_alloc_root(fs_info); if (!fs_info->tree_root) { error = -ENOMEM; goto error_fs_info; } - fs_info->tree_root->fs_info = fs_info; fs_info->fs_devices = fs_devices; fs_info->super_copy = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS); From e3029d9fd426c8f582210ba35551ae5506218345 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 00:56:18 -0500 Subject: [PATCH 211/820] btrfs: sanitizing ->fs_info, part 5 close_ctree() uses a weird mix of accesses to root->fs_info and its value at the beginning of function stored in local variable. Since ->fs_info *never* changes, let's just use the local variable to avoid confusion. Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index ee846ce58846..9308c7f13c17 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2982,7 +2982,7 @@ int close_ctree(struct btrfs_root *root) (atomic_read(&fs_info->defrag_running) == 0)); /* clear out the rbtree of defraggable inodes */ - btrfs_run_defrag_inodes(root->fs_info); + btrfs_run_defrag_inodes(fs_info); /* * Here come 2 situations when btrfs is broken to flip readonly: @@ -3011,8 +3011,8 @@ int close_ctree(struct btrfs_root *root) btrfs_put_block_group_cache(fs_info); - kthread_stop(root->fs_info->transaction_kthread); - kthread_stop(root->fs_info->cleaner_kthread); + kthread_stop(fs_info->transaction_kthread); + kthread_stop(fs_info->cleaner_kthread); fs_info->closing = 2; smp_mb(); @@ -3030,14 +3030,14 @@ int close_ctree(struct btrfs_root *root) free_extent_buffer(fs_info->extent_root->commit_root); free_extent_buffer(fs_info->tree_root->node); free_extent_buffer(fs_info->tree_root->commit_root); - free_extent_buffer(root->fs_info->chunk_root->node); - free_extent_buffer(root->fs_info->chunk_root->commit_root); - free_extent_buffer(root->fs_info->dev_root->node); - free_extent_buffer(root->fs_info->dev_root->commit_root); - free_extent_buffer(root->fs_info->csum_root->node); - free_extent_buffer(root->fs_info->csum_root->commit_root); + free_extent_buffer(fs_info->chunk_root->node); + free_extent_buffer(fs_info->chunk_root->commit_root); + free_extent_buffer(fs_info->dev_root->node); + free_extent_buffer(fs_info->dev_root->commit_root); + free_extent_buffer(fs_info->csum_root->node); + free_extent_buffer(fs_info->csum_root->commit_root); - btrfs_free_block_groups(root->fs_info); + btrfs_free_block_groups(fs_info); del_fs_roots(fs_info); From ad2b2c802be2d3e8ed8364fef5ffaddabe448219 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:10:02 -0500 Subject: [PATCH 212/820] btrfs: make open_ctree() return int It returns either ERR_PTR(-ve) or sb->s_fs_info. The latter can be found by caller just as well, TYVM, no need to return it. Just return -ve or 0... Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 12 ++++++------ fs/btrfs/disk-io.h | 6 +++--- fs/btrfs/super.c | 8 ++++---- 3 files changed, 13 insertions(+), 13 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 9308c7f13c17..78247522c693 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1880,9 +1880,9 @@ static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root) } -struct btrfs_root *open_ctree(struct super_block *sb, - struct btrfs_fs_devices *fs_devices, - char *options) +int open_ctree(struct super_block *sb, + struct btrfs_fs_devices *fs_devices, + char *options) { u32 sectorsize; u32 nodesize; @@ -2428,11 +2428,11 @@ retry_root_backup: if (err) { close_ctree(tree_root); free_fs_info(fs_info); - return ERR_PTR(err); + return err; } } - return tree_root; + return 0; fail_trans_kthread: kthread_stop(fs_info->transaction_kthread); @@ -2479,7 +2479,7 @@ fail_srcu: fail: btrfs_close_devices(fs_info->fs_devices); free_fs_info(fs_info); - return ERR_PTR(err); + return err; recovery_tree_root: if (!btrfs_test_opt(tree_root, RECOVERY)) diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 2bb5f59ddf95..6f5f48778b00 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -46,9 +46,9 @@ struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize); int clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf); -struct btrfs_root *open_ctree(struct super_block *sb, - struct btrfs_fs_devices *fs_devices, - char *options); +int open_ctree(struct super_block *sb, + struct btrfs_fs_devices *fs_devices, + char *options); int close_ctree(struct btrfs_root *root); int write_ctree_super(struct btrfs_trans_handle *trans, struct btrfs_root *root, int max_mirrors); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index e9f876a1655b..56e007fd6702 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -604,12 +604,12 @@ static int btrfs_fill_super(struct super_block *sb, sb->s_flags |= MS_POSIXACL; #endif - tree_root = open_ctree(sb, fs_devices, (char *)data); - - if (IS_ERR(tree_root)) { + err = open_ctree(sb, fs_devices, (char *)data); + if (err) { printk("btrfs: open_ctree failed\n"); - return PTR_ERR(tree_root); + return err; } + tree_root = sb->s_fs_info; fs_info = tree_root->fs_info; sb->s_fs_info = tree_root; From 29db78aa0ac82319b764b87a1c5030d74523e296 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:12:38 -0500 Subject: [PATCH 213/820] btrfs: kill pointless reassignment of ->s_fs_info in btrfs_fill_super() We do not (fortunately) modify ->s_fs_info of superblock on the fly in btrfs_fill_super(); apparent assignment is a no-op. Signed-off-by: Al Viro --- fs/btrfs/super.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 56e007fd6702..a381f9f9b0c2 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -588,8 +588,8 @@ static int btrfs_fill_super(struct super_block *sb, { struct inode *inode; struct dentry *root_dentry; - struct btrfs_root *tree_root; - struct btrfs_fs_info *fs_info; + struct btrfs_root *tree_root = sb->s_fs_info; + struct btrfs_fs_info *fs_info = tree_root->fs_info; struct btrfs_key key; int err; @@ -609,9 +609,6 @@ static int btrfs_fill_super(struct super_block *sb, printk("btrfs: open_ctree failed\n"); return err; } - tree_root = sb->s_fs_info; - fs_info = tree_root->fs_info; - sb->s_fs_info = tree_root; key.objectid = BTRFS_FIRST_FREE_OBJECTID; key.type = BTRFS_INODE_ITEM_KEY; From be7e0950def403e90b5295ff2192c39967bf2aec Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:40:05 -0500 Subject: [PATCH 214/820] btrfs: merge free_fs_info() calls on fill_super failures ... all the way up into btrfs_mount(). Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 2 -- fs/btrfs/super.c | 2 +- 2 files changed, 1 insertion(+), 3 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 78247522c693..b6d11eb4e430 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2427,7 +2427,6 @@ retry_root_backup: up_read(&fs_info->cleanup_work_sem); if (err) { close_ctree(tree_root); - free_fs_info(fs_info); return err; } } @@ -2478,7 +2477,6 @@ fail_srcu: cleanup_srcu_struct(&fs_info->subvol_srcu); fail: btrfs_close_devices(fs_info->fs_devices); - free_fs_info(fs_info); return err; recovery_tree_root: diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index a381f9f9b0c2..8901b6c85260 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -634,7 +634,6 @@ static int btrfs_fill_super(struct super_block *sb, fail_close: close_ctree(tree_root); - free_fs_info(fs_info); return err; } @@ -947,6 +946,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, error = btrfs_fill_super(s, fs_devices, data, flags & MS_SILENT ? 1 : 0); if (error) { + free_fs_info(fs_info); deactivate_locked_super(s); return ERR_PTR(error); } From d22ca7de770e2a683eac000ba4db5a95e2f4c969 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:46:50 -0500 Subject: [PATCH 215/820] btrfs: make free_fs_info() call ->kill_sb() unconditional ... and don't bother with it after btrfs_fill_super() failure - ->kill_sb() (unlike ->put_super()) will be called even if we have not got non-NULL ->s_root. Signed-off-by: Al Viro --- fs/btrfs/super.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 8901b6c85260..f628a6a79ca6 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -946,7 +946,6 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, error = btrfs_fill_super(s, fs_devices, data, flags & MS_SILENT ? 1 : 0); if (error) { - free_fs_info(fs_info); deactivate_locked_super(s); return ERR_PTR(error); } @@ -1215,12 +1214,9 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) static void btrfs_kill_super(struct super_block *sb) { - struct btrfs_fs_info *fs_info = NULL; - if (sb->s_root) - fs_info = btrfs_sb(sb)->fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(sb)->fs_info; kill_anon_super(sb); - if (fs_info) - free_fs_info(fs_info); + free_fs_info(fs_info); } static struct file_system_type btrfs_fs_type = { From 59553edf110e5576d91be9dd5bd53d110e0d0290 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 01:56:28 -0500 Subject: [PATCH 216/820] btrfs: consolidate failure exits in btrfs_mount() a bit Signed-off-by: Al Viro --- fs/btrfs/super.c | 21 +++++---------------- 1 file changed, 5 insertions(+), 16 deletions(-) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index f628a6a79ca6..15c8ae571e40 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -630,6 +630,7 @@ static int btrfs_fill_super(struct super_block *sb, save_mount_options(sb, data); cleancache_init_fs(sb); + sb->s_flags |= MS_ACTIVE; return 0; fail_close: @@ -929,14 +930,10 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, } if (s->s_root) { - if ((flags ^ s->s_flags) & MS_RDONLY) { - deactivate_locked_super(s); - error = -EBUSY; - goto error_close_devices; - } - btrfs_close_devices(fs_devices); free_fs_info(fs_info); + if ((flags ^ s->s_flags) & MS_RDONLY) + error = -EBUSY; } else { char b[BDEVNAME_SIZE]; @@ -945,19 +942,11 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, btrfs_sb(s)->fs_info->bdev_holder = fs_type; error = btrfs_fill_super(s, fs_devices, data, flags & MS_SILENT ? 1 : 0); - if (error) { - deactivate_locked_super(s); - return ERR_PTR(error); - } - - s->s_flags |= MS_ACTIVE; } - root = get_default_root(s, subvol_objectid); - if (IS_ERR(root)) { + root = !error ? get_default_root(s, subvol_objectid) : ERR_PTR(error); + if (IS_ERR(root)) deactivate_locked_super(s); - return root; - } return root; From 815745cf3e46681241ad8025602ffbf2a452d514 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 15:40:49 -0500 Subject: [PATCH 217/820] btrfs: let ->s_fs_info point to fs_info, not root... the latter can be obtained from the former (by looking as ->tree_root) just as cheaply as we currently are doing the other way round. Signed-off-by: Al Viro --- fs/btrfs/ctree.h | 2 +- fs/btrfs/disk-io.c | 4 +-- fs/btrfs/export.c | 2 +- fs/btrfs/ioctl.c | 7 ++--- fs/btrfs/super.c | 75 +++++++++++++++++++++++----------------------- 5 files changed, 45 insertions(+), 45 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 67385033323d..dbb999a38272 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2196,7 +2196,7 @@ static inline u32 btrfs_file_extent_inline_item_len(struct extent_buffer *eb, return btrfs_item_size(eb, e) - offset; } -static inline struct btrfs_root *btrfs_sb(struct super_block *sb) +static inline struct btrfs_fs_info *btrfs_sb(struct super_block *sb) { return sb->s_fs_info; } diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index b6d11eb4e430..f7d8b9ba519b 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1894,8 +1894,8 @@ int open_ctree(struct super_block *sb, struct btrfs_key location; struct buffer_head *bh; struct btrfs_super_block *disk_super; - struct btrfs_root *tree_root = btrfs_sb(sb); - struct btrfs_fs_info *fs_info = tree_root->fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + struct btrfs_root *tree_root = fs_info->tree_root; struct btrfs_root *extent_root; struct btrfs_root *csum_root; struct btrfs_root *chunk_root; diff --git a/fs/btrfs/export.c b/fs/btrfs/export.c index 1b8dc33778f9..5f77166fd01c 100644 --- a/fs/btrfs/export.c +++ b/fs/btrfs/export.c @@ -67,7 +67,7 @@ static struct dentry *btrfs_get_dentry(struct super_block *sb, u64 objectid, u64 root_objectid, u32 generation, int check_generation) { - struct btrfs_fs_info *fs_info = btrfs_sb(sb)->fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(sb); struct btrfs_root *root; struct inode *inode; struct btrfs_key key; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 5441ff1480fd..0c55f8f38091 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -276,14 +276,13 @@ static int btrfs_ioctl_getversion(struct file *file, int __user *arg) static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) { - struct btrfs_root *root = fdentry(file)->d_sb->s_fs_info; - struct btrfs_fs_info *fs_info = root->fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(fdentry(file)->d_sb); struct btrfs_device *device; struct request_queue *q; struct fstrim_range range; u64 minlen = ULLONG_MAX; u64 num_devices = 0; - u64 total_bytes = btrfs_super_total_bytes(root->fs_info->super_copy); + u64 total_bytes = btrfs_super_total_bytes(fs_info->super_copy); int ret; if (!capable(CAP_SYS_ADMIN)) @@ -312,7 +311,7 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) range.len = min(range.len, total_bytes - range.start); range.minlen = max(range.minlen, minlen); - ret = btrfs_trim_fs(root, &range); + ret = btrfs_trim_fs(fs_info->tree_root, &range); if (ret < 0) return ret; diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 15c8ae571e40..305adf6dc09a 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -147,7 +147,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, static void btrfs_put_super(struct super_block *sb) { - (void)close_ctree(btrfs_sb(sb)); + (void)close_ctree(btrfs_sb(sb)->tree_root); /* FIXME: need to fix VFS to return error? */ /* AV: return it _where_? ->put_super() can be triggered by any number * of async events, up to and including delivery of SIGKILL to the @@ -500,7 +500,8 @@ out: static struct dentry *get_default_root(struct super_block *sb, u64 subvol_objectid) { - struct btrfs_root *root = sb->s_fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + struct btrfs_root *root = fs_info->tree_root; struct btrfs_root *new_root; struct btrfs_dir_item *di; struct btrfs_path *path; @@ -530,7 +531,7 @@ static struct dentry *get_default_root(struct super_block *sb, * will mount by default if we haven't been given a specific subvolume * to mount. */ - dir_id = btrfs_super_root_dir(root->fs_info->super_copy); + dir_id = btrfs_super_root_dir(fs_info->super_copy); di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0); if (IS_ERR(di)) { btrfs_free_path(path); @@ -544,7 +545,7 @@ static struct dentry *get_default_root(struct super_block *sb, */ btrfs_free_path(path); dir_id = BTRFS_FIRST_FREE_OBJECTID; - new_root = root->fs_info->fs_root; + new_root = fs_info->fs_root; goto setup_root; } @@ -552,7 +553,7 @@ static struct dentry *get_default_root(struct super_block *sb, btrfs_free_path(path); find_root: - new_root = btrfs_read_fs_root_no_name(root->fs_info, &location); + new_root = btrfs_read_fs_root_no_name(fs_info, &location); if (IS_ERR(new_root)) return ERR_CAST(new_root); @@ -588,8 +589,7 @@ static int btrfs_fill_super(struct super_block *sb, { struct inode *inode; struct dentry *root_dentry; - struct btrfs_root *tree_root = sb->s_fs_info; - struct btrfs_fs_info *fs_info = tree_root->fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(sb); struct btrfs_key key; int err; @@ -634,20 +634,21 @@ static int btrfs_fill_super(struct super_block *sb, return 0; fail_close: - close_ctree(tree_root); + close_ctree(fs_info->tree_root); return err; } int btrfs_sync_fs(struct super_block *sb, int wait) { struct btrfs_trans_handle *trans; - struct btrfs_root *root = btrfs_sb(sb); + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + struct btrfs_root *root = fs_info->tree_root; int ret; trace_btrfs_sync_fs(wait); if (!wait) { - filemap_flush(root->fs_info->btree_inode->i_mapping); + filemap_flush(fs_info->btree_inode->i_mapping); return 0; } @@ -663,8 +664,8 @@ int btrfs_sync_fs(struct super_block *sb, int wait) static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry) { - struct btrfs_root *root = btrfs_sb(dentry->d_sb); - struct btrfs_fs_info *info = root->fs_info; + struct btrfs_fs_info *info = btrfs_sb(dentry->d_sb); + struct btrfs_root *root = info->tree_root; char *compress_type; if (btrfs_test_opt(root, DEGRADED)) @@ -727,10 +728,10 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry) static int btrfs_test_super(struct super_block *s, void *data) { - struct btrfs_root *test_root = data; - struct btrfs_root *root = btrfs_sb(s); + struct btrfs_fs_info *p = data; + struct btrfs_fs_info *fs_info = btrfs_sb(s); - return root->fs_info->fs_devices == test_root->fs_info->fs_devices; + return fs_info->fs_devices == p->fs_devices; } static int btrfs_set_super(struct super_block *s, void *data) @@ -922,8 +923,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, } bdev = fs_devices->latest_bdev; - s = sget(fs_type, btrfs_test_super, btrfs_set_super, - fs_info->tree_root); + s = sget(fs_type, btrfs_test_super, btrfs_set_super, fs_info); if (IS_ERR(s)) { error = PTR_ERR(s); goto error_close_devices; @@ -939,7 +939,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, s->s_flags = flags | MS_NOSEC; strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id)); - btrfs_sb(s)->fs_info->bdev_holder = fs_type; + btrfs_sb(s)->bdev_holder = fs_type; error = btrfs_fill_super(s, fs_devices, data, flags & MS_SILENT ? 1 : 0); } @@ -959,7 +959,8 @@ error_fs_info: static int btrfs_remount(struct super_block *sb, int *flags, char *data) { - struct btrfs_root *root = btrfs_sb(sb); + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + struct btrfs_root *root = fs_info->tree_root; int ret; ret = btrfs_parse_options(root, data); @@ -975,13 +976,13 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) ret = btrfs_commit_super(root); WARN_ON(ret); } else { - if (root->fs_info->fs_devices->rw_devices == 0) + if (fs_info->fs_devices->rw_devices == 0) return -EACCES; - if (btrfs_super_log_root(root->fs_info->super_copy) != 0) + if (btrfs_super_log_root(fs_info->super_copy) != 0) return -EINVAL; - ret = btrfs_cleanup_fs_roots(root->fs_info); + ret = btrfs_cleanup_fs_roots(fs_info); WARN_ON(ret); /* recover relocation */ @@ -1150,18 +1151,18 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes) static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) { - struct btrfs_root *root = btrfs_sb(dentry->d_sb); - struct btrfs_super_block *disk_super = root->fs_info->super_copy; - struct list_head *head = &root->fs_info->space_info; + struct btrfs_fs_info *fs_info = btrfs_sb(dentry->d_sb); + struct btrfs_super_block *disk_super = fs_info->super_copy; + struct list_head *head = &fs_info->space_info; struct btrfs_space_info *found; u64 total_used = 0; u64 total_free_data = 0; int bits = dentry->d_sb->s_blocksize_bits; - __be32 *fsid = (__be32 *)root->fs_info->fsid; + __be32 *fsid = (__be32 *)fs_info->fsid; int ret; /* holding chunk_muext to avoid allocating new chunks */ - mutex_lock(&root->fs_info->chunk_mutex); + mutex_lock(&fs_info->chunk_mutex); rcu_read_lock(); list_for_each_entry_rcu(found, head, list) { if (found->flags & BTRFS_BLOCK_GROUP_DATA) { @@ -1180,14 +1181,14 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_bsize = dentry->d_sb->s_blocksize; buf->f_type = BTRFS_SUPER_MAGIC; buf->f_bavail = total_free_data; - ret = btrfs_calc_avail_data_space(root, &total_free_data); + ret = btrfs_calc_avail_data_space(fs_info->tree_root, &total_free_data); if (ret) { - mutex_unlock(&root->fs_info->chunk_mutex); + mutex_unlock(&fs_info->chunk_mutex); return ret; } buf->f_bavail += total_free_data; buf->f_bavail = buf->f_bavail >> bits; - mutex_unlock(&root->fs_info->chunk_mutex); + mutex_unlock(&fs_info->chunk_mutex); /* We treat it as constant endianness (it doesn't matter _which_) because we want the fsid to come out the same whether mounted @@ -1203,7 +1204,7 @@ static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf) static void btrfs_kill_super(struct super_block *sb) { - struct btrfs_fs_info *fs_info = btrfs_sb(sb)->fs_info; + struct btrfs_fs_info *fs_info = btrfs_sb(sb); kill_anon_super(sb); free_fs_info(fs_info); } @@ -1246,17 +1247,17 @@ static long btrfs_control_ioctl(struct file *file, unsigned int cmd, static int btrfs_freeze(struct super_block *sb) { - struct btrfs_root *root = btrfs_sb(sb); - mutex_lock(&root->fs_info->transaction_kthread_mutex); - mutex_lock(&root->fs_info->cleaner_mutex); + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + mutex_lock(&fs_info->transaction_kthread_mutex); + mutex_lock(&fs_info->cleaner_mutex); return 0; } static int btrfs_unfreeze(struct super_block *sb) { - struct btrfs_root *root = btrfs_sb(sb); - mutex_unlock(&root->fs_info->cleaner_mutex); - mutex_unlock(&root->fs_info->transaction_kthread_mutex); + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + mutex_unlock(&fs_info->cleaner_mutex); + mutex_unlock(&fs_info->transaction_kthread_mutex); return 0; } From f84a8bd60e3ee49eacc9ba824babf149ba3dad7e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 17 Nov 2011 15:57:57 -0500 Subject: [PATCH 218/820] btrfs: take allocation of ->tree_root into open_ctree() now that we don't need it for sget() anymore... Signed-off-by: Al Viro --- fs/btrfs/disk-io.c | 8 +++++--- fs/btrfs/disk-io.h | 2 -- fs/btrfs/super.c | 5 ----- 3 files changed, 5 insertions(+), 10 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index f7d8b9ba519b..dd71be875939 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1215,7 +1215,7 @@ static int find_and_setup_root(struct btrfs_root *tree_root, return 0; } -struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info) +static struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info) { struct btrfs_root *root = kzalloc(sizeof(*root), GFP_NOFS); if (root) @@ -1895,7 +1895,7 @@ int open_ctree(struct super_block *sb, struct buffer_head *bh; struct btrfs_super_block *disk_super; struct btrfs_fs_info *fs_info = btrfs_sb(sb); - struct btrfs_root *tree_root = fs_info->tree_root; + struct btrfs_root *tree_root; struct btrfs_root *extent_root; struct btrfs_root *csum_root; struct btrfs_root *chunk_root; @@ -1906,12 +1906,14 @@ int open_ctree(struct super_block *sb, int num_backups_tried = 0; int backup_index = 0; + tree_root = fs_info->tree_root = btrfs_alloc_root(fs_info); extent_root = fs_info->extent_root = btrfs_alloc_root(fs_info); csum_root = fs_info->csum_root = btrfs_alloc_root(fs_info); chunk_root = fs_info->chunk_root = btrfs_alloc_root(fs_info); dev_root = fs_info->dev_root = btrfs_alloc_root(fs_info); - if (!extent_root || !csum_root || !chunk_root || !dev_root) { + if (!tree_root || !extent_root || !csum_root || + !chunk_root || !dev_root) { err = -ENOMEM; goto fail; } diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 6f5f48778b00..e4bc4741319b 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -86,8 +86,6 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans, int btrfs_add_log_tree(struct btrfs_trans_handle *trans, struct btrfs_root *root); -struct btrfs_root *btrfs_alloc_root(struct btrfs_fs_info *fs_info); - #ifdef CONFIG_DEBUG_LOCK_ALLOC void btrfs_init_lockdep(void); void btrfs_set_buffer_lockdep_class(u64 objectid, diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 305adf6dc09a..044652ee7ef5 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -899,11 +899,6 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, if (!fs_info) return ERR_PTR(-ENOMEM); - fs_info->tree_root = btrfs_alloc_root(fs_info); - if (!fs_info->tree_root) { - error = -ENOMEM; - goto error_fs_info; - } fs_info->fs_devices = fs_devices; fs_info->super_copy = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS); From 6b2a05584c5939eba24d22037dcb4cc71e5345e5 Mon Sep 17 00:00:00 2001 From: Jamie Iles Date: Mon, 9 Jan 2012 18:10:56 +0000 Subject: [PATCH 219/820] ARM: picoxcell: fix sched_clock() cleanup fallout Commit 2f0778afa (ARM: 7205/2: sched_clock: allow sched_clock to be selected at runtime) replaced the picoxcell specific sched_clock() with a generic runtime selectable version but replaced "unsigned long long notrace sched_clock(void)" with "unsigned u32 notrace picoxcell_read_sched_clock(void)" fix this up to return a u32 as expected. Cc: Marc Zyngier Signed-off-by: Jamie Iles --- arch/arm/mach-picoxcell/time.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mach-picoxcell/time.c b/arch/arm/mach-picoxcell/time.c index 6c89cf8ab22e..2ecba6743b8e 100644 --- a/arch/arm/mach-picoxcell/time.c +++ b/arch/arm/mach-picoxcell/time.c @@ -67,7 +67,7 @@ static void picoxcell_add_clocksource(struct device_node *source_timer) static void __iomem *sched_io_base; -unsigned u32 notrace picoxcell_read_sched_clock(void) +static u32 picoxcell_read_sched_clock(void) { return __raw_readl(sched_io_base); } From c2f5b65020869215814df03c3941dac9436f99fb Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Sat, 15 Oct 2011 07:33:46 -0400 Subject: [PATCH 220/820] NVMe: Simplify completion handling Instead of encoding the handler type in the bottom two bits of the per-completion context pointer, store the handler function as well as the context pointer. This gives us more flexibility and the code is clearer. It comes at the cost of an extra 8k of memory per queue, but this feels like a reasonable price to pay. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 167 +++++++++++++++++++++---------------------- 1 file changed, 81 insertions(+), 86 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index a17f80fa3881..4724655a6ebf 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -135,8 +135,12 @@ static inline void _nvme_check_size(void) BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64); } +typedef void (*nvme_completion_fn)(struct nvme_queue *, void *, + struct nvme_completion *); + struct nvme_cmd_info { - unsigned long ctx; + nvme_completion_fn fn; + void *ctx; unsigned long timeout; }; @@ -149,7 +153,7 @@ static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) * alloc_cmdid() - Allocate a Command ID * @nvmeq: The queue that will be used for this command * @ctx: A pointer that will be passed to the handler - * @handler: The ID of the handler to call + * @handler: The function to call on completion * * Allocate a Command ID for a queue. The data passed in will * be passed to the completion handler. This is implemented by using @@ -160,28 +164,27 @@ static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) * May be called with local interrupts disabled and the q_lock held, * or with interrupts enabled and no locks held. */ -static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, int handler, - unsigned timeout) +static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, + nvme_completion_fn handler, unsigned timeout) { int depth = nvmeq->q_depth - 1; struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); int cmdid; - BUG_ON((unsigned long)ctx & 3); - do { cmdid = find_first_zero_bit(nvmeq->cmdid_data, depth); if (cmdid >= depth) return -EBUSY; } while (test_and_set_bit(cmdid, nvmeq->cmdid_data)); - info[cmdid].ctx = (unsigned long)ctx | handler; + info[cmdid].fn = handler; + info[cmdid].ctx = ctx; info[cmdid].timeout = jiffies + timeout; return cmdid; } static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, - int handler, unsigned timeout) + nvme_completion_fn handler, unsigned timeout) { int cmdid; wait_event_killable(nvmeq->sq_full, @@ -189,47 +192,69 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, return (cmdid < 0) ? -EINTR : cmdid; } -/* - * If you need more than four handlers, you'll need to change how - * alloc_cmdid and nvme_process_cq work. Consider using a special - * CMD_CTX value instead, if that works for your situation. - */ -enum { - sync_completion_id = 0, - bio_completion_id, -}; - -/* Special values must be a multiple of 4, and less than 0x1000 */ -#define CMD_CTX_BASE (POISON_POINTER_DELTA + sync_completion_id) +/* Special values must be less than 0x1000 */ +#define CMD_CTX_BASE ((void *)POISON_POINTER_DELTA) #define CMD_CTX_CANCELLED (0x30C + CMD_CTX_BASE) #define CMD_CTX_COMPLETED (0x310 + CMD_CTX_BASE) #define CMD_CTX_INVALID (0x314 + CMD_CTX_BASE) #define CMD_CTX_FLUSH (0x318 + CMD_CTX_BASE) +static void special_completion(struct nvme_queue *nvmeq, void *ctx, + struct nvme_completion *cqe) +{ + if (ctx == CMD_CTX_CANCELLED) + return; + if (ctx == CMD_CTX_FLUSH) + return; + if (ctx == CMD_CTX_COMPLETED) { + dev_warn(nvmeq->q_dmadev, + "completed id %d twice on queue %d\n", + cqe->command_id, le16_to_cpup(&cqe->sq_id)); + return; + } + if (ctx == CMD_CTX_INVALID) { + dev_warn(nvmeq->q_dmadev, + "invalid id %d completed on queue %d\n", + cqe->command_id, le16_to_cpup(&cqe->sq_id)); + return; + } + + dev_warn(nvmeq->q_dmadev, "Unknown special completion %p\n", ctx); +} + /* * Called with local interrupts disabled and the q_lock held. May not sleep. */ -static unsigned long free_cmdid(struct nvme_queue *nvmeq, int cmdid) +static void *free_cmdid(struct nvme_queue *nvmeq, int cmdid, + nvme_completion_fn *fn) { - unsigned long data; + void *ctx; struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); - if (cmdid >= nvmeq->q_depth) + if (cmdid >= nvmeq->q_depth) { + *fn = special_completion; return CMD_CTX_INVALID; - data = info[cmdid].ctx; + } + *fn = info[cmdid].fn; + ctx = info[cmdid].ctx; + info[cmdid].fn = special_completion; info[cmdid].ctx = CMD_CTX_COMPLETED; clear_bit(cmdid, nvmeq->cmdid_data); wake_up(&nvmeq->sq_full); - return data; + return ctx; } -static unsigned long cancel_cmdid(struct nvme_queue *nvmeq, int cmdid) +static void *cancel_cmdid(struct nvme_queue *nvmeq, int cmdid, + nvme_completion_fn *fn) { - unsigned long data; + void *ctx; struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); - data = info[cmdid].ctx; + if (fn) + *fn = info[cmdid].fn; + ctx = info[cmdid].ctx; + info[cmdid].fn = special_completion; info[cmdid].ctx = CMD_CTX_CANCELLED; - return data; + return ctx; } static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) @@ -485,7 +510,7 @@ static int nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns, static int nvme_submit_flush_data(struct nvme_queue *nvmeq, struct nvme_ns *ns) { int cmdid = alloc_cmdid(nvmeq, (void *)CMD_CTX_FLUSH, - sync_completion_id, IO_TIMEOUT); + special_completion, IO_TIMEOUT); if (unlikely(cmdid < 0)) return cmdid; @@ -518,7 +543,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, nbio->bio = bio; result = -EBUSY; - cmdid = alloc_cmdid(nvmeq, nbio, bio_completion_id, IO_TIMEOUT); + cmdid = alloc_cmdid(nvmeq, nbio, bio_completion, IO_TIMEOUT); if (unlikely(cmdid < 0)) goto free_nbio; @@ -599,45 +624,6 @@ static int nvme_make_request(struct request_queue *q, struct bio *bio) return 0; } -struct sync_cmd_info { - struct task_struct *task; - u32 result; - int status; -}; - -static void sync_completion(struct nvme_queue *nvmeq, void *ctx, - struct nvme_completion *cqe) -{ - struct sync_cmd_info *cmdinfo = ctx; - if (unlikely((unsigned long)cmdinfo == CMD_CTX_CANCELLED)) - return; - if ((unsigned long)cmdinfo == CMD_CTX_FLUSH) - return; - if (unlikely((unsigned long)cmdinfo == CMD_CTX_COMPLETED)) { - dev_warn(nvmeq->q_dmadev, - "completed id %d twice on queue %d\n", - cqe->command_id, le16_to_cpup(&cqe->sq_id)); - return; - } - if (unlikely((unsigned long)cmdinfo == CMD_CTX_INVALID)) { - dev_warn(nvmeq->q_dmadev, - "invalid id %d completed on queue %d\n", - cqe->command_id, le16_to_cpup(&cqe->sq_id)); - return; - } - cmdinfo->result = le32_to_cpup(&cqe->result); - cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; - wake_up_process(cmdinfo->task); -} - -typedef void (*completion_fn)(struct nvme_queue *, void *, - struct nvme_completion *); - -static const completion_fn nvme_completions[4] = { - [sync_completion_id] = sync_completion, - [bio_completion_id] = bio_completion, -}; - static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) { u16 head, phase; @@ -646,9 +632,8 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) phase = nvmeq->cq_phase; for (;;) { - unsigned long data; - void *ptr; - unsigned char handler; + void *ctx; + nvme_completion_fn fn; struct nvme_completion cqe = nvmeq->cqes[head]; if ((le16_to_cpu(cqe.status) & 1) != phase) break; @@ -658,10 +643,8 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) phase = !phase; } - data = free_cmdid(nvmeq, cqe.command_id); - handler = data & 3; - ptr = (void *)(data & ~3UL); - nvme_completions[handler](nvmeq, ptr, &cqe); + ctx = free_cmdid(nvmeq, cqe.command_id, &fn); + fn(nvmeq, ctx, &cqe); } /* If the controller ignores the cq head doorbell and continuously @@ -702,10 +685,25 @@ static irqreturn_t nvme_irq_check(int irq, void *data) static void nvme_abort_command(struct nvme_queue *nvmeq, int cmdid) { spin_lock_irq(&nvmeq->q_lock); - cancel_cmdid(nvmeq, cmdid); + cancel_cmdid(nvmeq, cmdid, NULL); spin_unlock_irq(&nvmeq->q_lock); } +struct sync_cmd_info { + struct task_struct *task; + u32 result; + int status; +}; + +static void sync_completion(struct nvme_queue *nvmeq, void *ctx, + struct nvme_completion *cqe) +{ + struct sync_cmd_info *cmdinfo = ctx; + cmdinfo->result = le32_to_cpup(&cqe->result); + cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; + wake_up_process(cmdinfo->task); +} + /* * Returns 0 on success. If the result is negative, it's a Linux error code; * if the result is positive, it's an NVM Express status code @@ -719,7 +717,7 @@ static int nvme_submit_sync_cmd(struct nvme_queue *nvmeq, cmdinfo.task = current; cmdinfo.status = -EINTR; - cmdid = alloc_cmdid_killable(nvmeq, &cmdinfo, sync_completion_id, + cmdid = alloc_cmdid_killable(nvmeq, &cmdinfo, sync_completion, timeout); if (cmdid < 0) return cmdid; @@ -1201,18 +1199,15 @@ static void nvme_timeout_ios(struct nvme_queue *nvmeq) int cmdid; for_each_set_bit(cmdid, nvmeq->cmdid_data, depth) { - unsigned long data; - void *ptr; - unsigned char handler; + void *ctx; + nvme_completion_fn fn; static struct nvme_completion cqe = { .status = cpu_to_le16(NVME_SC_ABORT_REQ) << 1, }; if (!time_after(now, info[cmdid].timeout)) continue; dev_warn(nvmeq->q_dmadev, "Timing out I/O %d\n", cmdid); - data = cancel_cmdid(nvmeq, cmdid); - handler = data & 3; - ptr = (void *)(data & ~3UL); - nvme_completions[handler](nvmeq, ptr, &cqe); + ctx = cancel_cmdid(nvmeq, cmdid, &fn); + fn(nvmeq, ctx, &cqe); } } From 040a93b52a9eee8177ebaf2ba0ee0f9f518d1bf8 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 20 Dec 2011 11:04:12 -0500 Subject: [PATCH 221/820] NVMe: Change get_nvmeq to take a dev instead of a namespace Upcoming patches require calling get_nvmeq when we don't have a namespace. Some callers already have the device in a local variable anyway. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 4724655a6ebf..aa2fd66aabd6 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -257,9 +257,9 @@ static void *cancel_cmdid(struct nvme_queue *nvmeq, int cmdid, return ctx; } -static struct nvme_queue *get_nvmeq(struct nvme_ns *ns) +static struct nvme_queue *get_nvmeq(struct nvme_dev *dev) { - return ns->dev->queues[get_cpu() + 1]; + return dev->queues[get_cpu() + 1]; } static void put_nvmeq(struct nvme_queue *nvmeq) @@ -606,7 +606,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, static int nvme_make_request(struct request_queue *q, struct bio *bio) { struct nvme_ns *ns = q->queuedata; - struct nvme_queue *nvmeq = get_nvmeq(ns); + struct nvme_queue *nvmeq = get_nvmeq(ns->dev); int result = -EBUSY; spin_lock_irq(&nvmeq->q_lock); @@ -1103,7 +1103,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) /* XXX: metadata */ prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); - nvmeq = get_nvmeq(ns); + nvmeq = get_nvmeq(dev); /* * Since nvme_submit_sync_cmd sleeps, we can't keep preemption * disabled. We may be preempted at any point, and be rescheduled From 5c1281a3bf5655ec1b90db495da3a2b77826ba88 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 20 Dec 2011 11:54:53 -0500 Subject: [PATCH 222/820] NVMe: Change nvme_completion_fn to take a dev The queue is only needed for some rare occasions, and it's more consistent to pass the device around. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 43 +++++++++++++++++++++++++------------------ 1 file changed, 25 insertions(+), 18 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index aa2fd66aabd6..b0e8a6dd33b1 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -135,7 +135,7 @@ static inline void _nvme_check_size(void) BUILD_BUG_ON(sizeof(struct nvme_lba_range_type) != 64); } -typedef void (*nvme_completion_fn)(struct nvme_queue *, void *, +typedef void (*nvme_completion_fn)(struct nvme_dev *, void *, struct nvme_completion *); struct nvme_cmd_info { @@ -199,7 +199,7 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, #define CMD_CTX_INVALID (0x314 + CMD_CTX_BASE) #define CMD_CTX_FLUSH (0x318 + CMD_CTX_BASE) -static void special_completion(struct nvme_queue *nvmeq, void *ctx, +static void special_completion(struct nvme_dev *dev, void *ctx, struct nvme_completion *cqe) { if (ctx == CMD_CTX_CANCELLED) @@ -207,19 +207,19 @@ static void special_completion(struct nvme_queue *nvmeq, void *ctx, if (ctx == CMD_CTX_FLUSH) return; if (ctx == CMD_CTX_COMPLETED) { - dev_warn(nvmeq->q_dmadev, + dev_warn(&dev->pci_dev->dev, "completed id %d twice on queue %d\n", cqe->command_id, le16_to_cpup(&cqe->sq_id)); return; } if (ctx == CMD_CTX_INVALID) { - dev_warn(nvmeq->q_dmadev, + dev_warn(&dev->pci_dev->dev, "invalid id %d completed on queue %d\n", cqe->command_id, le16_to_cpup(&cqe->sq_id)); return; } - dev_warn(nvmeq->q_dmadev, "Unknown special completion %p\n", ctx); + dev_warn(&dev->pci_dev->dev, "Unknown special completion %p\n", ctx); } /* @@ -332,29 +332,36 @@ static struct nvme_bio *alloc_nbio(unsigned nseg, gfp_t gfp) sizeof(struct scatterlist) * nseg, gfp); } -static void free_nbio(struct nvme_queue *nvmeq, struct nvme_bio *nbio) +static void free_nbio(struct nvme_dev *dev, struct nvme_bio *nbio) { - nvme_free_prps(nvmeq->dev, nbio->prps); + nvme_free_prps(dev, nbio->prps); kfree(nbio); } -static void bio_completion(struct nvme_queue *nvmeq, void *ctx, +static void requeue_bio(struct nvme_dev *dev, struct bio *bio) +{ + struct nvme_queue *nvmeq = get_nvmeq(dev); + if (bio_list_empty(&nvmeq->sq_cong)) + add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait); + bio_list_add(&nvmeq->sq_cong, bio); + put_nvmeq(nvmeq); + wake_up_process(nvme_thread); +} + +static void bio_completion(struct nvme_dev *dev, void *ctx, struct nvme_completion *cqe) { struct nvme_bio *nbio = ctx; struct bio *bio = nbio->bio; u16 status = le16_to_cpup(&cqe->status) >> 1; - dma_unmap_sg(nvmeq->q_dmadev, nbio->sg, nbio->nents, + dma_unmap_sg(&dev->pci_dev->dev, nbio->sg, nbio->nents, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); - free_nbio(nvmeq, nbio); + free_nbio(dev, nbio); if (status) { bio_endio(bio, -EIO); } else if (bio->bi_vcnt > bio->bi_idx) { - if (bio_list_empty(&nvmeq->sq_cong)) - add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait); - bio_list_add(&nvmeq->sq_cong, bio); - wake_up_process(nvme_thread); + requeue_bio(dev, bio); } else { bio_endio(bio, 0); } @@ -594,7 +601,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return 0; free_nbio: - free_nbio(nvmeq, nbio); + free_nbio(nvmeq->dev, nbio); nomem: return result; } @@ -644,7 +651,7 @@ static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) } ctx = free_cmdid(nvmeq, cqe.command_id, &fn); - fn(nvmeq, ctx, &cqe); + fn(nvmeq->dev, ctx, &cqe); } /* If the controller ignores the cq head doorbell and continuously @@ -695,7 +702,7 @@ struct sync_cmd_info { int status; }; -static void sync_completion(struct nvme_queue *nvmeq, void *ctx, +static void sync_completion(struct nvme_dev *dev, void *ctx, struct nvme_completion *cqe) { struct sync_cmd_info *cmdinfo = ctx; @@ -1207,7 +1214,7 @@ static void nvme_timeout_ios(struct nvme_queue *nvmeq) continue; dev_warn(nvmeq->q_dmadev, "Timing out I/O %d\n", cmdid); ctx = cancel_cmdid(nvmeq, cmdid, &fn); - fn(nvmeq, ctx, &cqe); + fn(nvmeq->dev, ctx, &cqe); } } From eca18b2394a9387feeaf14cd884ddddd7a809d19 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 20 Dec 2011 13:34:52 -0500 Subject: [PATCH 223/820] NVMe: Merge the nvme_bio and nvme_prp data structures The new merged data structure is called nvme_iod. This improves performance for mid-sized I/Os (in the 16k range) since we save a memory allocation. It is also a slightly simpler interface to use. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 241 ++++++++++++++++++++++--------------------- 1 file changed, 125 insertions(+), 116 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b0e8a6dd33b1..4517608c068f 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -290,52 +290,70 @@ static int nvme_submit_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd) return 0; } -struct nvme_prps { - int npages; /* 0 means small pool in use */ +/* + * The nvme_iod describes the data in an I/O, including the list of PRP + * entries. You can't see it in this data structure because C doesn't let + * me express that. Use nvme_alloc_iod to ensure there's enough space + * allocated to store the PRP list. + */ +struct nvme_iod { + void *private; /* For the use of the submitter of the I/O */ + int npages; /* In the PRP list. 0 means small pool in use */ + int offset; /* Of PRP list */ + int nents; /* Used in scatterlist */ + int length; /* Of data, in bytes */ dma_addr_t first_dma; - __le64 *list[0]; + struct scatterlist sg[0]; }; -static void nvme_free_prps(struct nvme_dev *dev, struct nvme_prps *prps) +static __le64 **iod_list(struct nvme_iod *iod) +{ + return ((void *)iod) + iod->offset; +} + +/* + * Will slightly overestimate the number of pages needed. This is OK + * as it only leads to a small amount of wasted memory for the lifetime of + * the I/O. + */ +static int nvme_npages(unsigned size) +{ + unsigned nprps = DIV_ROUND_UP(size + PAGE_SIZE, PAGE_SIZE); + return DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8); +} + +static struct nvme_iod * +nvme_alloc_iod(unsigned nseg, unsigned nbytes, gfp_t gfp) +{ + struct nvme_iod *iod = kmalloc(sizeof(struct nvme_iod) + + sizeof(__le64 *) * nvme_npages(nbytes) + + sizeof(struct scatterlist) * nseg, gfp); + + if (iod) { + iod->offset = offsetof(struct nvme_iod, sg[nseg]); + iod->npages = -1; + iod->length = nbytes; + } + + return iod; +} + +static void nvme_free_iod(struct nvme_dev *dev, struct nvme_iod *iod) { const int last_prp = PAGE_SIZE / 8 - 1; int i; - dma_addr_t prp_dma; + __le64 **list = iod_list(iod); + dma_addr_t prp_dma = iod->first_dma; - if (!prps) - return; - - prp_dma = prps->first_dma; - - if (prps->npages == 0) - dma_pool_free(dev->prp_small_pool, prps->list[0], prp_dma); - for (i = 0; i < prps->npages; i++) { - __le64 *prp_list = prps->list[i]; + if (iod->npages == 0) + dma_pool_free(dev->prp_small_pool, list[0], prp_dma); + for (i = 0; i < iod->npages; i++) { + __le64 *prp_list = list[i]; dma_addr_t next_prp_dma = le64_to_cpu(prp_list[last_prp]); dma_pool_free(dev->prp_page_pool, prp_list, prp_dma); prp_dma = next_prp_dma; } - kfree(prps); -} - -struct nvme_bio { - struct bio *bio; - int nents; - struct nvme_prps *prps; - struct scatterlist sg[0]; -}; - -/* XXX: use a mempool */ -static struct nvme_bio *alloc_nbio(unsigned nseg, gfp_t gfp) -{ - return kzalloc(sizeof(struct nvme_bio) + - sizeof(struct scatterlist) * nseg, gfp); -} - -static void free_nbio(struct nvme_dev *dev, struct nvme_bio *nbio) -{ - nvme_free_prps(dev, nbio->prps); - kfree(nbio); + kfree(iod); } static void requeue_bio(struct nvme_dev *dev, struct bio *bio) @@ -351,13 +369,13 @@ static void requeue_bio(struct nvme_dev *dev, struct bio *bio) static void bio_completion(struct nvme_dev *dev, void *ctx, struct nvme_completion *cqe) { - struct nvme_bio *nbio = ctx; - struct bio *bio = nbio->bio; + struct nvme_iod *iod = ctx; + struct bio *bio = iod->private; u16 status = le16_to_cpup(&cqe->status) >> 1; - dma_unmap_sg(&dev->pci_dev->dev, nbio->sg, nbio->nents, + dma_unmap_sg(&dev->pci_dev->dev, iod->sg, iod->nents, bio_data_dir(bio) ? DMA_TO_DEVICE : DMA_FROM_DEVICE); - free_nbio(dev, nbio); + nvme_free_iod(dev, iod); if (status) { bio_endio(bio, -EIO); } else if (bio->bi_vcnt > bio->bi_idx) { @@ -368,25 +386,25 @@ static void bio_completion(struct nvme_dev *dev, void *ctx, } /* length is in bytes. gfp flags indicates whether we may sleep. */ -static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, - struct nvme_common_command *cmd, - struct scatterlist *sg, int *len, - gfp_t gfp) +static int nvme_setup_prps(struct nvme_dev *dev, + struct nvme_common_command *cmd, struct nvme_iod *iod, + int total_len, gfp_t gfp) { struct dma_pool *pool; - int length = *len; + int length = total_len; + struct scatterlist *sg = iod->sg; int dma_len = sg_dma_len(sg); u64 dma_addr = sg_dma_address(sg); int offset = offset_in_page(dma_addr); __le64 *prp_list; + __le64 **list = iod_list(iod); dma_addr_t prp_dma; - int nprps, npages, i; - struct nvme_prps *prps = NULL; + int nprps, i; cmd->prp1 = cpu_to_le64(dma_addr); length -= (PAGE_SIZE - offset); if (length <= 0) - return prps; + return total_len; dma_len -= (PAGE_SIZE - offset); if (dma_len) { @@ -399,46 +417,35 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, if (length <= PAGE_SIZE) { cmd->prp2 = cpu_to_le64(dma_addr); - return prps; + return total_len; } nprps = DIV_ROUND_UP(length, PAGE_SIZE); - npages = DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8); - prps = kmalloc(sizeof(*prps) + sizeof(__le64 *) * npages, gfp); - if (!prps) { - cmd->prp2 = cpu_to_le64(dma_addr); - *len = (*len - length) + PAGE_SIZE; - return prps; - } - if (nprps <= (256 / 8)) { pool = dev->prp_small_pool; - prps->npages = 0; + iod->npages = 0; } else { pool = dev->prp_page_pool; - prps->npages = 1; + iod->npages = 1; } prp_list = dma_pool_alloc(pool, gfp, &prp_dma); if (!prp_list) { cmd->prp2 = cpu_to_le64(dma_addr); - *len = (*len - length) + PAGE_SIZE; - kfree(prps); - return NULL; + iod->npages = -1; + return (total_len - length) + PAGE_SIZE; } - prps->list[0] = prp_list; - prps->first_dma = prp_dma; + list[0] = prp_list; + iod->first_dma = prp_dma; cmd->prp2 = cpu_to_le64(prp_dma); i = 0; for (;;) { if (i == PAGE_SIZE / 8) { __le64 *old_prp_list = prp_list; prp_list = dma_pool_alloc(pool, gfp, &prp_dma); - if (!prp_list) { - *len = (*len - length); - return prps; - } - prps->list[prps->npages++] = prp_list; + if (!prp_list) + return total_len - length; + list[iod->npages++] = prp_list; prp_list[0] = old_prp_list[i - 1]; old_prp_list[i - 1] = cpu_to_le64(prp_dma); i = 1; @@ -457,21 +464,21 @@ static struct nvme_prps *nvme_setup_prps(struct nvme_dev *dev, dma_len = sg_dma_len(sg); } - return prps; + return total_len; } /* NVMe scatterlists require no holes in the virtual address */ #define BIOVEC_NOT_VIRT_MERGEABLE(vec1, vec2) ((vec2)->bv_offset || \ (((vec1)->bv_offset + (vec1)->bv_len) % PAGE_SIZE)) -static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, +static int nvme_map_bio(struct device *dev, struct nvme_iod *iod, struct bio *bio, enum dma_data_direction dma_dir, int psegs) { struct bio_vec *bvec, *bvprv = NULL; struct scatterlist *sg = NULL; int i, old_idx, length = 0, nsegs = 0; - sg_init_table(nbio->sg, psegs); + sg_init_table(iod->sg, psegs); old_idx = bio->bi_idx; bio_for_each_segment(bvec, bio, i) { if (bvprv && BIOVEC_PHYS_MERGEABLE(bvprv, bvec)) { @@ -479,7 +486,7 @@ static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, } else { if (bvprv && BIOVEC_NOT_VIRT_MERGEABLE(bvprv, bvec)) break; - sg = sg ? sg + 1 : nbio->sg; + sg = sg ? sg + 1 : iod->sg; sg_set_page(sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); nsegs++; @@ -488,9 +495,9 @@ static int nvme_map_bio(struct device *dev, struct nvme_bio *nbio, bvprv = bvec; } bio->bi_idx = i; - nbio->nents = nsegs; + iod->nents = nsegs; sg_mark_end(sg); - if (dma_map_sg(dev, nbio->sg, nbio->nents, dma_dir) == 0) { + if (dma_map_sg(dev, iod->sg, iod->nents, dma_dir) == 0) { bio->bi_idx = old_idx; return -ENOMEM; } @@ -531,7 +538,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, struct bio *bio) { struct nvme_command *cmnd; - struct nvme_bio *nbio; + struct nvme_iod *iod; enum dma_data_direction dma_dir; int cmdid, length, result = -ENOMEM; u16 control; @@ -544,15 +551,15 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return result; } - nbio = alloc_nbio(psegs, GFP_ATOMIC); - if (!nbio) + iod = nvme_alloc_iod(psegs, bio->bi_size, GFP_ATOMIC); + if (!iod) goto nomem; - nbio->bio = bio; + iod->private = bio; result = -EBUSY; - cmdid = alloc_cmdid(nvmeq, nbio, bio_completion, IO_TIMEOUT); + cmdid = alloc_cmdid(nvmeq, iod, bio_completion, IO_TIMEOUT); if (unlikely(cmdid < 0)) - goto free_nbio; + goto free_iod; if ((bio->bi_rw & REQ_FLUSH) && !psegs) return nvme_submit_flush(nvmeq, ns, cmdid); @@ -578,15 +585,15 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, dma_dir = DMA_FROM_DEVICE; } - result = nvme_map_bio(nvmeq->q_dmadev, nbio, bio, dma_dir, psegs); + result = nvme_map_bio(nvmeq->q_dmadev, iod, bio, dma_dir, psegs); if (result < 0) - goto free_nbio; + goto free_iod; length = result; cmnd->rw.command_id = cmdid; cmnd->rw.nsid = cpu_to_le32(ns->ns_id); - nbio->prps = nvme_setup_prps(nvmeq->dev, &cmnd->common, nbio->sg, - &length, GFP_ATOMIC); + length = nvme_setup_prps(nvmeq->dev, &cmnd->common, iod, length, + GFP_ATOMIC); cmnd->rw.slba = cpu_to_le64(bio->bi_sector >> (ns->lba_shift - 9)); cmnd->rw.length = cpu_to_le16((length >> ns->lba_shift) - 1); cmnd->rw.control = cpu_to_le16(control); @@ -600,8 +607,8 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return 0; - free_nbio: - free_nbio(nvmeq->dev, nbio); + free_iod: + nvme_free_iod(nvmeq->dev, iod); nomem: return result; } @@ -1005,18 +1012,18 @@ static int __devinit nvme_configure_admin_queue(struct nvme_dev *dev) return result; } -static int nvme_map_user_pages(struct nvme_dev *dev, int write, - unsigned long addr, unsigned length, - struct scatterlist **sgp) +static struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write, + unsigned long addr, unsigned length) { int i, err, count, nents, offset; struct scatterlist *sg; struct page **pages; + struct nvme_iod *iod; if (addr & 3) - return -EINVAL; + return ERR_PTR(-EINVAL); if (!length) - return -EINVAL; + return ERR_PTR(-EINVAL); offset = offset_in_page(addr); count = DIV_ROUND_UP(offset + length, PAGE_SIZE); @@ -1029,7 +1036,8 @@ static int nvme_map_user_pages(struct nvme_dev *dev, int write, goto put_pages; } - sg = kcalloc(count, sizeof(*sg), GFP_KERNEL); + iod = nvme_alloc_iod(count, length, GFP_KERNEL); + sg = iod->sg; sg_init_table(sg, count); for (i = 0; i < count; i++) { sg_set_page(&sg[i], pages[i], @@ -1042,22 +1050,24 @@ static int nvme_map_user_pages(struct nvme_dev *dev, int write, nents = dma_map_sg(&dev->pci_dev->dev, sg, count, write ? DMA_TO_DEVICE : DMA_FROM_DEVICE); if (!nents) - goto put_pages; + goto free_iod; kfree(pages); - *sgp = sg; - return nents; + return iod; + free_iod: + kfree(iod); put_pages: for (i = 0; i < count; i++) put_page(pages[i]); kfree(pages); - return err; + return ERR_PTR(err); } static void nvme_unmap_user_pages(struct nvme_dev *dev, int write, - unsigned long addr, int length, struct scatterlist *sg) + unsigned long addr, int length, struct nvme_iod *iod) { + struct scatterlist *sg = iod->sg; int i, count; count = DIV_ROUND_UP(offset_in_page(addr) + length, PAGE_SIZE); @@ -1074,9 +1084,8 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) struct nvme_user_io io; struct nvme_command c; unsigned length; - int nents, status; - struct scatterlist *sg; - struct nvme_prps *prps; + int status; + struct nvme_iod *iod; if (copy_from_user(&io, uio, sizeof(io))) return -EFAULT; @@ -1086,15 +1095,14 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) case nvme_cmd_write: case nvme_cmd_read: case nvme_cmd_compare: - nents = nvme_map_user_pages(dev, io.opcode & 1, io.addr, - length, &sg); + iod = nvme_map_user_pages(dev, io.opcode & 1, io.addr, length); break; default: return -EINVAL; } - if (nents < 0) - return nents; + if (IS_ERR(iod)) + return PTR_ERR(iod); memset(&c, 0, sizeof(c)); c.rw.opcode = io.opcode; @@ -1108,7 +1116,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) c.rw.apptag = io.apptag; c.rw.appmask = io.appmask; /* XXX: metadata */ - prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); + length = nvme_setup_prps(dev, &c.common, iod, length, GFP_KERNEL); nvmeq = get_nvmeq(dev); /* @@ -1123,8 +1131,8 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) else status = nvme_submit_sync_cmd(nvmeq, &c, NULL, IO_TIMEOUT); - nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, sg); - nvme_free_prps(dev, prps); + nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, iod); + nvme_free_iod(dev, iod); return status; } @@ -1134,9 +1142,8 @@ static int nvme_user_admin_cmd(struct nvme_ns *ns, struct nvme_dev *dev = ns->dev; struct nvme_admin_cmd cmd; struct nvme_command c; - int status, length, nents = 0; - struct scatterlist *sg; - struct nvme_prps *prps = NULL; + int status, length; + struct nvme_iod *iod; if (!capable(CAP_SYS_ADMIN)) return -EACCES; @@ -1158,19 +1165,21 @@ static int nvme_user_admin_cmd(struct nvme_ns *ns, length = cmd.data_len; if (cmd.data_len) { - nents = nvme_map_user_pages(dev, 1, cmd.addr, length, &sg); - if (nents < 0) - return nents; - prps = nvme_setup_prps(dev, &c.common, sg, &length, GFP_KERNEL); + iod = nvme_map_user_pages(dev, 1, cmd.addr, length); + if (IS_ERR(iod)) + return PTR_ERR(iod); + length = nvme_setup_prps(dev, &c.common, iod, length, + GFP_KERNEL); } if (length != cmd.data_len) status = -ENOMEM; else status = nvme_submit_admin_cmd(dev, &c, NULL); + if (cmd.data_len) { - nvme_unmap_user_pages(dev, 0, cmd.addr, cmd.data_len, sg); - nvme_free_prps(dev, prps); + nvme_unmap_user_pages(dev, 0, cmd.addr, cmd.data_len, iod); + nvme_free_iod(dev, iod); } return status; } From ff976d724a74e4522e9ca2de1fb37ac4520f454f Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 20 Dec 2011 13:53:01 -0500 Subject: [PATCH 224/820] NVMe: Rename IO_TIMEOUT to NVME_IO_TIMEOUT IO_TIMEOUT is a little too generic and might be used by other parts of the kernel in the future. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 4517608c068f..1cc01872f6dc 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -45,7 +45,7 @@ #define SQ_SIZE(depth) (depth * sizeof(struct nvme_command)) #define CQ_SIZE(depth) (depth * sizeof(struct nvme_completion)) #define NVME_MINORS 64 -#define IO_TIMEOUT (5 * HZ) +#define NVME_IO_TIMEOUT (5 * HZ) #define ADMIN_TIMEOUT (60 * HZ) static int nvme_major; @@ -524,7 +524,7 @@ static int nvme_submit_flush(struct nvme_queue *nvmeq, struct nvme_ns *ns, static int nvme_submit_flush_data(struct nvme_queue *nvmeq, struct nvme_ns *ns) { int cmdid = alloc_cmdid(nvmeq, (void *)CMD_CTX_FLUSH, - special_completion, IO_TIMEOUT); + special_completion, NVME_IO_TIMEOUT); if (unlikely(cmdid < 0)) return cmdid; @@ -557,7 +557,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, iod->private = bio; result = -EBUSY; - cmdid = alloc_cmdid(nvmeq, iod, bio_completion, IO_TIMEOUT); + cmdid = alloc_cmdid(nvmeq, iod, bio_completion, NVME_IO_TIMEOUT); if (unlikely(cmdid < 0)) goto free_iod; @@ -1129,7 +1129,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) if (length != (io.nblocks + 1) << ns->lba_shift) status = -ENOMEM; else - status = nvme_submit_sync_cmd(nvmeq, &c, NULL, IO_TIMEOUT); + status = nvme_submit_sync_cmd(nvmeq, &c, NULL, NVME_IO_TIMEOUT); nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, iod); nvme_free_iod(dev, iod); From 497421880acecd0281d3182d534f3d28c927caec Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 6 Jan 2012 13:42:45 -0700 Subject: [PATCH 225/820] NVMe: Fix DMA mapping for admin commands We were always mapping as DMA_FROM_DEVICE then unmapping with DMA_TO_DEVICE which was clearly not correct. Follow the same pattern as nvme_submit_io() and key off the bottom bit of the opcode to determine whether this is a read or a write. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 1cc01872f6dc..3f8cae9dc960 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1165,7 +1165,8 @@ static int nvme_user_admin_cmd(struct nvme_ns *ns, length = cmd.data_len; if (cmd.data_len) { - iod = nvme_map_user_pages(dev, 1, cmd.addr, length); + iod = nvme_map_user_pages(dev, cmd.opcode & 1, cmd.addr, + length); if (IS_ERR(iod)) return PTR_ERR(iod); length = nvme_setup_prps(dev, &c.common, iod, length, @@ -1178,7 +1179,8 @@ static int nvme_user_admin_cmd(struct nvme_ns *ns, status = nvme_submit_admin_cmd(dev, &c, NULL); if (cmd.data_len) { - nvme_unmap_user_pages(dev, 0, cmd.addr, cmd.data_len, iod); + nvme_unmap_user_pages(dev, cmd.opcode & 1, cmd.addr, + cmd.data_len, iod); nvme_free_iod(dev, iod); } return status; From fe304c43c6d63e29ed4fc46a874d7a74313788c5 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 6 Jan 2012 13:49:25 -0700 Subject: [PATCH 226/820] NVMe: Mark the end of the sg list For user I/O and admin commands, we were forgetting to mark the end of the SG list. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 3f8cae9dc960..71fc9030b4df 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1045,6 +1045,7 @@ static struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write, length -= (PAGE_SIZE - offset); offset = 0; } + sg_mark_end(&sg[i - 1]); err = -ENOMEM; nents = dma_map_sg(&dev->pci_dev->dev, sg, count, From 1c2ad9faaf662b4a525348775deca3ac8e6c35a0 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 6 Jan 2012 13:52:56 -0700 Subject: [PATCH 227/820] NVMe: Simplify nvme_unmap_user_pages By using the iod->nents field (the same way other I/O paths do), we can avoid recalculating the number of sg entries at unmap time, and make nvme_unmap_user_pages() easier to call. Also, use the 'write' parameter instead of assuming DMA_FROM_DEVICE. Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 71fc9030b4df..3cf82c27a544 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1046,6 +1046,7 @@ static struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write, offset = 0; } sg_mark_end(&sg[i - 1]); + iod->nents = count; err = -ENOMEM; nents = dma_map_sg(&dev->pci_dev->dev, sg, count, @@ -1066,16 +1067,15 @@ static struct nvme_iod *nvme_map_user_pages(struct nvme_dev *dev, int write, } static void nvme_unmap_user_pages(struct nvme_dev *dev, int write, - unsigned long addr, int length, struct nvme_iod *iod) + struct nvme_iod *iod) { - struct scatterlist *sg = iod->sg; - int i, count; + int i; - count = DIV_ROUND_UP(offset_in_page(addr) + length, PAGE_SIZE); - dma_unmap_sg(&dev->pci_dev->dev, sg, count, DMA_FROM_DEVICE); + dma_unmap_sg(&dev->pci_dev->dev, iod->sg, iod->nents, + write ? DMA_TO_DEVICE : DMA_FROM_DEVICE); - for (i = 0; i < count; i++) - put_page(sg_page(&sg[i])); + for (i = 0; i < iod->nents; i++) + put_page(sg_page(&iod->sg[i])); } static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) @@ -1132,7 +1132,7 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) else status = nvme_submit_sync_cmd(nvmeq, &c, NULL, NVME_IO_TIMEOUT); - nvme_unmap_user_pages(dev, io.opcode & 1, io.addr, length, iod); + nvme_unmap_user_pages(dev, io.opcode & 1, iod); nvme_free_iod(dev, iod); return status; } @@ -1180,8 +1180,7 @@ static int nvme_user_admin_cmd(struct nvme_ns *ns, status = nvme_submit_admin_cmd(dev, &c, NULL); if (cmd.data_len) { - nvme_unmap_user_pages(dev, cmd.opcode & 1, cmd.addr, - cmd.data_len, iod); + nvme_unmap_user_pages(dev, cmd.opcode & 1, iod); nvme_free_iod(dev, iod); } return status; From 4eeb9215a0d5c9494ca8b20158cc8ee82618840c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 10 Jan 2012 14:35:08 -0700 Subject: [PATCH 228/820] NVMe: Set queue flags correctly QUEUE_FLAG_* are flags (other than QUEUE_FLAG_DEFAULT), so they cannot be ORed together. Set the queue flags using queue_flag_set_unlocked(). Reported-by: Donald Wood Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 3cf82c27a544..b583603fae5b 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1315,8 +1315,10 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, int nsid, ns->queue = blk_alloc_queue(GFP_KERNEL); if (!ns->queue) goto out_free_ns; - ns->queue->queue_flags = QUEUE_FLAG_DEFAULT | QUEUE_FLAG_NOMERGES | - QUEUE_FLAG_NONROT | QUEUE_FLAG_DISCARD; + ns->queue->queue_flags = QUEUE_FLAG_DEFAULT; + queue_flag_set_unlocked(QUEUE_FLAG_NOMERGES, ns->queue); + queue_flag_set_unlocked(QUEUE_FLAG_NONROT, ns->queue); +/* queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, ns->queue); */ blk_queue_make_request(ns->queue, nvme_make_request); ns->dev = dev; ns->queue->queuedata = ns; From 366e8217e5ec6ce9f73aec19c46d983110fb4a98 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 10 Jan 2012 16:30:15 -0500 Subject: [PATCH 229/820] NVMe: Version 0.8 Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index b583603fae5b..28c84b18712d 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -1726,6 +1726,6 @@ static void __exit nvme_exit(void) MODULE_AUTHOR("Matthew Wilcox "); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.7"); +MODULE_VERSION("0.8"); module_init(nvme_init); module_exit(nvme_exit); From 9fc397fc0878c9540af20cbffc4d546541fe8b23 Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Fri, 18 Nov 2011 01:44:27 +0100 Subject: [PATCH 230/820] [SCSI] scsi_dh_rdac: Fix error path If create_singlethread_workqueue() failes, rdac_init should fail too. Signed-off-by: Richard Weinberger Acked-by: "Moger, Babu" Signed-off-by: James Bottomley --- drivers/scsi/device_handler/scsi_dh_rdac.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/scsi/device_handler/scsi_dh_rdac.c b/drivers/scsi/device_handler/scsi_dh_rdac.c index 841ebf4a6788..53a31c753cb1 100644 --- a/drivers/scsi/device_handler/scsi_dh_rdac.c +++ b/drivers/scsi/device_handler/scsi_dh_rdac.c @@ -953,6 +953,8 @@ static int __init rdac_init(void) if (!kmpath_rdacd) { scsi_unregister_device_handler(&rdac_dh); printk(KERN_ERR "kmpath_rdacd creation failed.\n"); + + r = -EINVAL; } done: return r; From cced5041ed5a2d1352186510944b0ddfbdbe4c0b Mon Sep 17 00:00:00 2001 From: Stratos Psomadakis Date: Sun, 4 Dec 2011 02:23:54 +0200 Subject: [PATCH 231/820] [SCSI] sym53c8xx: Fix NULL pointer dereference in slave_destroy sym53c8xx_slave_destroy unconditionally assumes that sym53c8xx_slave_alloc has succesesfully allocated a sym_lcb. This can lead to a NULL pointer dereference (exposed by commit 4e6c82b). Signed-off-by: Stratos Psomadakis Cc: stable@vger.kernel.org Signed-off-by: James Bottomley --- drivers/scsi/sym53c8xx_2/sym_glue.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/scsi/sym53c8xx_2/sym_glue.c b/drivers/scsi/sym53c8xx_2/sym_glue.c index b4543f575f46..36d1ed7817eb 100644 --- a/drivers/scsi/sym53c8xx_2/sym_glue.c +++ b/drivers/scsi/sym53c8xx_2/sym_glue.c @@ -839,6 +839,10 @@ static void sym53c8xx_slave_destroy(struct scsi_device *sdev) struct sym_lcb *lp = sym_lp(tp, sdev->lun); unsigned long flags; + /* if slave_alloc returned before allocating a sym_lcb, return */ + if (!lp) + return; + spin_lock_irqsave(np->s.host->host_lock, flags); if (lp->busy_itlq || lp->busy_itl) { From 00c4a09bb0840457f5f8f5753a562e5e19a91baf Mon Sep 17 00:00:00 2001 From: Thadeu Lima de Souza Cascardo Date: Wed, 14 Dec 2011 13:46:23 -0200 Subject: [PATCH 232/820] [SCSI] libcxgbi: do not print a message when memory allocation fails In alloc_pdu, libcxgbi tries to allocate a skb with GFP_ATOMIC, which may potentially fail. When it happens, the current code prints a warning message. When the system is under IO stress, this failure may happen lots of times and it usually scares users. Instead of printing the warning message, the code now increases the tx_dropped statistics for the ethernet interface wich is doing the iscsi task. Signed-off-by: Thadeu Lima de Souza Cascardo Acked-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/cxgbi/libcxgbi.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/cxgbi/libcxgbi.c b/drivers/scsi/cxgbi/libcxgbi.c index c5360ffb4bed..d3ff9cd40234 100644 --- a/drivers/scsi/cxgbi/libcxgbi.c +++ b/drivers/scsi/cxgbi/libcxgbi.c @@ -1868,8 +1868,9 @@ int cxgbi_conn_alloc_pdu(struct iscsi_task *task, u8 opcode) tdata->skb = alloc_skb(cdev->skb_tx_rsvd + headroom, GFP_ATOMIC); if (!tdata->skb) { - pr_warn("alloc skb %u+%u, opcode 0x%x failed.\n", - cdev->skb_tx_rsvd, headroom, opcode); + struct cxgbi_sock *csk = cconn->cep->csk; + struct net_device *ndev = cdev->ports[csk->port_id]; + ndev->stats.tx_dropped++; return -ENOMEM; } From 4a4bc2e90c1c689bf929914256310361d4012df1 Mon Sep 17 00:00:00 2001 From: Lalit Chandivade Date: Fri, 16 Dec 2011 01:58:55 -0800 Subject: [PATCH 233/820] [SCSI] qla4xxx: cleanup, make qla4xxx_build_ddb_list short Make qla4xxx_build_ddb_list shorter by adding more helper functions. Signed-off-by: Lalit Chandivade Signed-off-by: Vikas Chaudhary Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_os.c | 419 ++++++++++++++++++++-------------- 1 file changed, 247 insertions(+), 172 deletions(-) diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index ec393a00c038..87e3699d7ea2 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -2078,7 +2078,7 @@ void qla4_8xxx_watchdog(struct scsi_qla_host *ha) } } -void qla4xxx_check_relogin_flash_ddb(struct iscsi_cls_session *cls_sess) +static void qla4xxx_check_relogin_flash_ddb(struct iscsi_cls_session *cls_sess) { struct iscsi_session *sess; struct ddb_entry *ddb_entry; @@ -3826,16 +3826,14 @@ exit_check: return ret; } -static void qla4xxx_free_nt_list(struct list_head *list_nt) +static void qla4xxx_free_ddb_list(struct list_head *list_ddb) { - struct qla_ddb_index *nt_ddb_idx, *nt_ddb_idx_tmp; + struct qla_ddb_index *ddb_idx, *ddb_idx_tmp; - /* Free up the normaltargets list */ - list_for_each_entry_safe(nt_ddb_idx, nt_ddb_idx_tmp, list_nt, list) { - list_del_init(&nt_ddb_idx->list); - vfree(nt_ddb_idx); + list_for_each_entry_safe(ddb_idx, ddb_idx_tmp, list_ddb, list) { + list_del_init(&ddb_idx->list); + vfree(ddb_idx); } - } static struct iscsi_endpoint *qla4xxx_get_ep_fwdb(struct scsi_qla_host *ha, @@ -3934,7 +3932,6 @@ static void qla4xxx_wait_for_ip_configuration(struct scsi_qla_host *ha) ip_state == IP_ADDRSTATE_DEPRICATED || ip_state == IP_ADDRSTATE_DISABLING) ip_idx[idx] = -1; - } /* Break if all IP states checked */ @@ -3947,52 +3944,34 @@ static void qla4xxx_wait_for_ip_configuration(struct scsi_qla_host *ha) } while (time_after(wtime, jiffies)); } -void qla4xxx_build_ddb_list(struct scsi_qla_host *ha, int is_reset) +static void qla4xxx_build_st_list(struct scsi_qla_host *ha, + struct list_head *list_st) { + struct qla_ddb_index *st_ddb_idx; int max_ddbs; + int fw_idx_size; + struct dev_db_entry *fw_ddb_entry; + dma_addr_t fw_ddb_dma; int ret; uint32_t idx = 0, next_idx = 0; uint32_t state = 0, conn_err = 0; - uint16_t conn_id; - struct dev_db_entry *fw_ddb_entry; - struct ddb_entry *ddb_entry = NULL; - dma_addr_t fw_ddb_dma; - struct iscsi_cls_session *cls_sess; - struct iscsi_session *sess; - struct iscsi_cls_conn *cls_conn; - struct iscsi_endpoint *ep; - uint16_t cmds_max = 32, tmo = 0; - uint32_t initial_cmdsn = 0; - struct list_head list_st, list_nt; /* List of sendtargets */ - struct qla_ddb_index *st_ddb_idx, *st_ddb_idx_tmp; - int fw_idx_size; - unsigned long wtime; - struct qla_ddb_index *nt_ddb_idx; - - if (!test_bit(AF_LINK_UP, &ha->flags)) { - set_bit(AF_BUILD_DDB_LIST, &ha->flags); - ha->is_reset = is_reset; - return; - } - max_ddbs = is_qla40XX(ha) ? MAX_DEV_DB_ENTRIES_40XX : - MAX_DEV_DB_ENTRIES; + uint16_t conn_id = 0; fw_ddb_entry = dma_pool_alloc(ha->fw_ddb_dma_pool, GFP_KERNEL, &fw_ddb_dma); if (fw_ddb_entry == NULL) { DEBUG2(ql4_printk(KERN_ERR, ha, "Out of memory\n")); - goto exit_ddb_list; + goto exit_st_list; } - INIT_LIST_HEAD(&list_st); - INIT_LIST_HEAD(&list_nt); + max_ddbs = is_qla40XX(ha) ? MAX_DEV_DB_ENTRIES_40XX : + MAX_DEV_DB_ENTRIES; fw_idx_size = sizeof(struct qla_ddb_index); for (idx = 0; idx < max_ddbs; idx = next_idx) { - ret = qla4xxx_get_fwddb_entry(ha, idx, fw_ddb_entry, - fw_ddb_dma, NULL, - &next_idx, &state, &conn_err, - NULL, &conn_id); + ret = qla4xxx_get_fwddb_entry(ha, idx, fw_ddb_entry, fw_ddb_dma, + NULL, &next_idx, &state, + &conn_err, NULL, &conn_id); if (ret == QLA_ERROR) break; @@ -4009,12 +3988,235 @@ void qla4xxx_build_ddb_list(struct scsi_qla_host *ha, int is_reset) st_ddb_idx->fw_ddb_idx = idx; - list_add_tail(&st_ddb_idx->list, &list_st); + list_add_tail(&st_ddb_idx->list, list_st); continue_next_st: if (next_idx == 0) break; } +exit_st_list: + if (fw_ddb_entry) + dma_pool_free(ha->fw_ddb_dma_pool, fw_ddb_entry, fw_ddb_dma); +} + +/** + * qla4xxx_remove_failed_ddb - Remove inactive or failed ddb from list + * @ha: pointer to adapter structure + * @list_ddb: List from which failed ddb to be removed + * + * Iterate over the list of DDBs and find and remove DDBs that are either in + * no connection active state or failed state + **/ +static void qla4xxx_remove_failed_ddb(struct scsi_qla_host *ha, + struct list_head *list_ddb) +{ + struct qla_ddb_index *ddb_idx, *ddb_idx_tmp; + uint32_t next_idx = 0; + uint32_t state = 0, conn_err = 0; + int ret; + + list_for_each_entry_safe(ddb_idx, ddb_idx_tmp, list_ddb, list) { + ret = qla4xxx_get_fwddb_entry(ha, ddb_idx->fw_ddb_idx, + NULL, 0, NULL, &next_idx, &state, + &conn_err, NULL, NULL); + if (ret == QLA_ERROR) + continue; + + if (state == DDB_DS_NO_CONNECTION_ACTIVE || + state == DDB_DS_SESSION_FAILED) { + list_del_init(&ddb_idx->list); + vfree(ddb_idx); + } + } +} + +static int qla4xxx_sess_conn_setup(struct scsi_qla_host *ha, + struct dev_db_entry *fw_ddb_entry, + int is_reset) +{ + struct iscsi_cls_session *cls_sess; + struct iscsi_session *sess; + struct iscsi_cls_conn *cls_conn; + struct iscsi_endpoint *ep; + uint16_t cmds_max = 32; + uint16_t conn_id = 0; + uint32_t initial_cmdsn = 0; + int ret = QLA_SUCCESS; + + struct ddb_entry *ddb_entry = NULL; + + /* Create session object, with INVALID_ENTRY, + * the targer_id would get set when we issue the login + */ + cls_sess = iscsi_session_setup(&qla4xxx_iscsi_transport, ha->host, + cmds_max, sizeof(struct ddb_entry), + sizeof(struct ql4_task_data), + initial_cmdsn, INVALID_ENTRY); + if (!cls_sess) { + ret = QLA_ERROR; + goto exit_setup; + } + + /* + * so calling module_put function to decrement the + * reference count. + **/ + module_put(qla4xxx_iscsi_transport.owner); + sess = cls_sess->dd_data; + ddb_entry = sess->dd_data; + ddb_entry->sess = cls_sess; + + cls_sess->recovery_tmo = ql4xsess_recovery_tmo; + memcpy(&ddb_entry->fw_ddb_entry, fw_ddb_entry, + sizeof(struct dev_db_entry)); + + qla4xxx_setup_flash_ddb_entry(ha, ddb_entry); + + cls_conn = iscsi_conn_setup(cls_sess, sizeof(struct qla_conn), conn_id); + + if (!cls_conn) { + ret = QLA_ERROR; + goto exit_setup; + } + + ddb_entry->conn = cls_conn; + + /* Setup ep, for displaying attributes in sysfs */ + ep = qla4xxx_get_ep_fwdb(ha, fw_ddb_entry); + if (ep) { + ep->conn = cls_conn; + cls_conn->ep = ep; + } else { + DEBUG2(ql4_printk(KERN_ERR, ha, "Unable to get ep\n")); + ret = QLA_ERROR; + goto exit_setup; + } + + /* Update sess/conn params */ + qla4xxx_copy_fwddb_param(ha, fw_ddb_entry, cls_sess, cls_conn); + + if (is_reset == RESET_ADAPTER) { + iscsi_block_session(cls_sess); + /* Use the relogin path to discover new devices + * by short-circuting the logic of setting + * timer to relogin - instead set the flags + * to initiate login right away. + */ + set_bit(DPC_RELOGIN_DEVICE, &ha->dpc_flags); + set_bit(DF_RELOGIN, &ddb_entry->flags); + } + +exit_setup: + return ret; +} + +static void qla4xxx_build_nt_list(struct scsi_qla_host *ha, + struct list_head *list_nt, int is_reset) +{ + struct dev_db_entry *fw_ddb_entry; + dma_addr_t fw_ddb_dma; + int max_ddbs; + int fw_idx_size; + int ret; + uint32_t idx = 0, next_idx = 0; + uint32_t state = 0, conn_err = 0; + uint16_t conn_id = 0; + struct qla_ddb_index *nt_ddb_idx; + + fw_ddb_entry = dma_pool_alloc(ha->fw_ddb_dma_pool, GFP_KERNEL, + &fw_ddb_dma); + if (fw_ddb_entry == NULL) { + DEBUG2(ql4_printk(KERN_ERR, ha, "Out of memory\n")); + goto exit_nt_list; + } + max_ddbs = is_qla40XX(ha) ? MAX_DEV_DB_ENTRIES_40XX : + MAX_DEV_DB_ENTRIES; + fw_idx_size = sizeof(struct qla_ddb_index); + + for (idx = 0; idx < max_ddbs; idx = next_idx) { + ret = qla4xxx_get_fwddb_entry(ha, idx, fw_ddb_entry, fw_ddb_dma, + NULL, &next_idx, &state, + &conn_err, NULL, &conn_id); + if (ret == QLA_ERROR) + break; + + if (qla4xxx_verify_boot_idx(ha, idx) != QLA_SUCCESS) + goto continue_next_nt; + + /* Check if NT, then add to list it */ + if (strlen((char *) fw_ddb_entry->iscsi_name) == 0) + goto continue_next_nt; + + if (!(state == DDB_DS_NO_CONNECTION_ACTIVE || + state == DDB_DS_SESSION_FAILED)) + goto continue_next_nt; + + DEBUG2(ql4_printk(KERN_INFO, ha, + "Adding DDB to session = 0x%x\n", idx)); + if (is_reset == INIT_ADAPTER) { + nt_ddb_idx = vmalloc(fw_idx_size); + if (!nt_ddb_idx) + break; + + nt_ddb_idx->fw_ddb_idx = idx; + + memcpy(&nt_ddb_idx->fw_ddb, fw_ddb_entry, + sizeof(struct dev_db_entry)); + + if (qla4xxx_is_flash_ddb_exists(ha, list_nt, + fw_ddb_entry) == QLA_SUCCESS) { + vfree(nt_ddb_idx); + goto continue_next_nt; + } + list_add_tail(&nt_ddb_idx->list, list_nt); + } else if (is_reset == RESET_ADAPTER) { + if (qla4xxx_is_session_exists(ha, fw_ddb_entry) == + QLA_SUCCESS) + goto continue_next_nt; + } + + ret = qla4xxx_sess_conn_setup(ha, fw_ddb_entry, is_reset); + if (ret == QLA_ERROR) + goto exit_nt_list; + +continue_next_nt: + if (next_idx == 0) + break; + } + +exit_nt_list: + if (fw_ddb_entry) + dma_pool_free(ha->fw_ddb_dma_pool, fw_ddb_entry, fw_ddb_dma); +} + +/** + * qla4xxx_build_ddb_list - Build ddb list and setup sessions + * @ha: pointer to adapter structure + * @is_reset: Is this init path or reset path + * + * Create a list of sendtargets (st) from firmware DDBs, issue send targets + * using connection open, then create the list of normal targets (nt) + * from firmware DDBs. Based on the list of nt setup session and connection + * objects. + **/ +void qla4xxx_build_ddb_list(struct scsi_qla_host *ha, int is_reset) +{ + uint16_t tmo = 0; + struct list_head list_st, list_nt; + struct qla_ddb_index *st_ddb_idx, *st_ddb_idx_tmp; + unsigned long wtime; + + if (!test_bit(AF_LINK_UP, &ha->flags)) { + set_bit(AF_BUILD_DDB_LIST, &ha->flags); + ha->is_reset = is_reset; + return; + } + + INIT_LIST_HEAD(&list_st); + INIT_LIST_HEAD(&list_nt); + + qla4xxx_build_st_list(ha, &list_st); + /* Before issuing conn open mbox, ensure all IPs states are configured * Note, conn open fails if IPs are not configured */ @@ -4032,147 +4234,20 @@ continue_next_st: wtime = jiffies + (HZ * tmo); do { - list_for_each_entry_safe(st_ddb_idx, st_ddb_idx_tmp, &list_st, - list) { - ret = qla4xxx_get_fwddb_entry(ha, - st_ddb_idx->fw_ddb_idx, - NULL, 0, NULL, &next_idx, - &state, &conn_err, NULL, - NULL); - if (ret == QLA_ERROR) - continue; - - if (state == DDB_DS_NO_CONNECTION_ACTIVE || - state == DDB_DS_SESSION_FAILED) { - list_del_init(&st_ddb_idx->list); - vfree(st_ddb_idx); - } - } + qla4xxx_remove_failed_ddb(ha, &list_st); schedule_timeout_uninterruptible(HZ / 10); } while (time_after(wtime, jiffies)); /* Free up the sendtargets list */ - list_for_each_entry_safe(st_ddb_idx, st_ddb_idx_tmp, &list_st, list) { - list_del_init(&st_ddb_idx->list); - vfree(st_ddb_idx); - } + qla4xxx_free_ddb_list(&list_st); - for (idx = 0; idx < max_ddbs; idx = next_idx) { - ret = qla4xxx_get_fwddb_entry(ha, idx, fw_ddb_entry, - fw_ddb_dma, NULL, - &next_idx, &state, &conn_err, - NULL, &conn_id); - if (ret == QLA_ERROR) - break; + qla4xxx_build_nt_list(ha, &list_nt, is_reset); - if (qla4xxx_verify_boot_idx(ha, idx) != QLA_SUCCESS) - goto continue_next_nt; - - /* Check if NT, then add to list it */ - if (strlen((char *) fw_ddb_entry->iscsi_name) == 0) - goto continue_next_nt; - - if (state == DDB_DS_NO_CONNECTION_ACTIVE || - state == DDB_DS_SESSION_FAILED) { - DEBUG2(ql4_printk(KERN_INFO, ha, - "Adding DDB to session = 0x%x\n", - idx)); - if (is_reset == INIT_ADAPTER) { - nt_ddb_idx = vmalloc(fw_idx_size); - if (!nt_ddb_idx) - break; - - nt_ddb_idx->fw_ddb_idx = idx; - - memcpy(&nt_ddb_idx->fw_ddb, fw_ddb_entry, - sizeof(struct dev_db_entry)); - - if (qla4xxx_is_flash_ddb_exists(ha, &list_nt, - fw_ddb_entry) == QLA_SUCCESS) { - vfree(nt_ddb_idx); - goto continue_next_nt; - } - list_add_tail(&nt_ddb_idx->list, &list_nt); - } else if (is_reset == RESET_ADAPTER) { - if (qla4xxx_is_session_exists(ha, - fw_ddb_entry) == QLA_SUCCESS) - goto continue_next_nt; - } - - /* Create session object, with INVALID_ENTRY, - * the targer_id would get set when we issue the login - */ - cls_sess = iscsi_session_setup(&qla4xxx_iscsi_transport, - ha->host, cmds_max, - sizeof(struct ddb_entry), - sizeof(struct ql4_task_data), - initial_cmdsn, INVALID_ENTRY); - if (!cls_sess) - goto exit_ddb_list; - - /* - * iscsi_session_setup increments the driver reference - * count which wouldn't let the driver to be unloaded. - * so calling module_put function to decrement the - * reference count. - **/ - module_put(qla4xxx_iscsi_transport.owner); - sess = cls_sess->dd_data; - ddb_entry = sess->dd_data; - ddb_entry->sess = cls_sess; - - cls_sess->recovery_tmo = ql4xsess_recovery_tmo; - memcpy(&ddb_entry->fw_ddb_entry, fw_ddb_entry, - sizeof(struct dev_db_entry)); - - qla4xxx_setup_flash_ddb_entry(ha, ddb_entry); - - cls_conn = iscsi_conn_setup(cls_sess, - sizeof(struct qla_conn), - conn_id); - if (!cls_conn) - goto exit_ddb_list; - - ddb_entry->conn = cls_conn; - - /* Setup ep, for displaying attributes in sysfs */ - ep = qla4xxx_get_ep_fwdb(ha, fw_ddb_entry); - if (ep) { - ep->conn = cls_conn; - cls_conn->ep = ep; - } else { - DEBUG2(ql4_printk(KERN_ERR, ha, - "Unable to get ep\n")); - } - - /* Update sess/conn params */ - qla4xxx_copy_fwddb_param(ha, fw_ddb_entry, cls_sess, - cls_conn); - - if (is_reset == RESET_ADAPTER) { - iscsi_block_session(cls_sess); - /* Use the relogin path to discover new devices - * by short-circuting the logic of setting - * timer to relogin - instead set the flags - * to initiate login right away. - */ - set_bit(DPC_RELOGIN_DEVICE, &ha->dpc_flags); - set_bit(DF_RELOGIN, &ddb_entry->flags); - } - } -continue_next_nt: - if (next_idx == 0) - break; - } -exit_ddb_list: - qla4xxx_free_nt_list(&list_nt); - if (fw_ddb_entry) - dma_pool_free(ha->fw_ddb_dma_pool, fw_ddb_entry, fw_ddb_dma); + qla4xxx_free_ddb_list(&list_nt); qla4xxx_free_ddb_index(ha); } - /** * qla4xxx_probe_adapter - callback function to probe HBA * @pdev: pointer to pci_dev structure From c28eaaca7acea214937fb9b3ae45e001d49947bf Mon Sep 17 00:00:00 2001 From: Nilesh Javali Date: Sun, 18 Dec 2011 21:40:44 -0800 Subject: [PATCH 234/820] [SCSI] qla4xxx: Limit the ACB Default Timeout value to 12s A wrong default timeout value programmed in the adapter causes driver to wait for that much time while waiting for target discoveries to complete. This could add huge delays during the driver load time. To avoid this, limit the default timeout value to 12 seconds if the default timeout value set in adapter is less than 12 seconds and greater than 120 seconds. Signed-off-by: Nilesh Javali Signed-off-by: Vikas Chaudhary Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_os.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index 87e3699d7ea2..09b28f8513a0 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -1630,7 +1630,9 @@ void qla4xxx_update_session_conn_param(struct scsi_qla_host *ha, /* Update timers after login */ ddb_entry->default_relogin_timeout = - le16_to_cpu(fw_ddb_entry->def_timeout); + (le16_to_cpu(fw_ddb_entry->def_timeout) > LOGIN_TOV) && + (le16_to_cpu(fw_ddb_entry->def_timeout) < LOGIN_TOV * 10) ? + le16_to_cpu(fw_ddb_entry->def_timeout) : LOGIN_TOV; ddb_entry->default_time2wait = le16_to_cpu(fw_ddb_entry->iscsi_def_time2wait); @@ -3882,6 +3884,8 @@ static int qla4xxx_verify_boot_idx(struct scsi_qla_host *ha, uint16_t idx) static void qla4xxx_setup_flash_ddb_entry(struct scsi_qla_host *ha, struct ddb_entry *ddb_entry) { + uint16_t def_timeout; + ddb_entry->ddb_type = FLASH_DDB; ddb_entry->fw_ddb_index = INVALID_ENTRY; ddb_entry->fw_ddb_device_state = DDB_DS_NO_CONNECTION_ACTIVE; @@ -3892,9 +3896,10 @@ static void qla4xxx_setup_flash_ddb_entry(struct scsi_qla_host *ha, atomic_set(&ddb_entry->retry_relogin_timer, INVALID_ENTRY); atomic_set(&ddb_entry->relogin_timer, 0); atomic_set(&ddb_entry->relogin_retry_count, 0); - + def_timeout = le16_to_cpu(ddb_entry->fw_ddb_entry.def_timeout); ddb_entry->default_relogin_timeout = - le16_to_cpu(ddb_entry->fw_ddb_entry.def_timeout); + (def_timeout > LOGIN_TOV) && (def_timeout < LOGIN_TOV * 10) ? + def_timeout : LOGIN_TOV; ddb_entry->default_time2wait = le16_to_cpu(ddb_entry->fw_ddb_entry.iscsi_def_time2wait); } @@ -4228,7 +4233,10 @@ void qla4xxx_build_ddb_list(struct scsi_qla_host *ha, int is_reset) } /* Wait to ensure all sendtargets are done for min 12 sec wait */ - tmo = ((ha->def_timeout < LOGIN_TOV) ? LOGIN_TOV : ha->def_timeout); + tmo = ((ha->def_timeout > LOGIN_TOV) && + (ha->def_timeout < LOGIN_TOV * 10) ? + ha->def_timeout : LOGIN_TOV); + DEBUG2(ql4_printk(KERN_INFO, ha, "Default time to wait for build ddb %d\n", tmo)); From f1f2e60e1a2320d0ce11102b72b15d5dedb29fe2 Mon Sep 17 00:00:00 2001 From: Nilesh Javali Date: Fri, 16 Dec 2011 01:58:57 -0800 Subject: [PATCH 235/820] [SCSI] qla4xxx: Break the loop if the sendtargets list was empty Signed-off-by: Nilesh Javali Signed-off-by: Vikas Chaudhary Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_os.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index 09b28f8513a0..ae5d512bf43b 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -4242,6 +4242,9 @@ void qla4xxx_build_ddb_list(struct scsi_qla_host *ha, int is_reset) wtime = jiffies + (HZ * tmo); do { + if (list_empty(&list_st)) + break; + qla4xxx_remove_failed_ddb(ha, &list_st); schedule_timeout_uninterruptible(HZ / 10); } while (time_after(wtime, jiffies)); From 368d45cc61c062a1291d0b29257a0697d6aeabd9 Mon Sep 17 00:00:00 2001 From: Manish Rangankar Date: Fri, 16 Dec 2011 01:58:58 -0800 Subject: [PATCH 236/820] [SCSI] qla4xxx: Fixed BFS with sendtargets as boot index. If ql4xdisablesysfsboot = 0 and sendtargets entry as boot index then driver does export sendtarget entries in sysfs but iscsistart does not do discovery. So in this case let driver do the discovery and login to the targets. Signed-off-by: Manish Rangankar Signed-off-by: Vikas Chaudhary Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_os.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index ae5d512bf43b..0048a3facd7a 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -3980,9 +3980,6 @@ static void qla4xxx_build_st_list(struct scsi_qla_host *ha, if (ret == QLA_ERROR) break; - if (qla4xxx_verify_boot_idx(ha, idx) != QLA_SUCCESS) - goto continue_next_st; - /* Check if ST, add to the list_st */ if (strlen((char *) fw_ddb_entry->iscsi_name) != 0) goto continue_next_st; From 3a19cbf530648a63fb3f73630046c561f70c6b71 Mon Sep 17 00:00:00 2001 From: Vikas Chaudhary Date: Fri, 16 Dec 2011 01:58:59 -0800 Subject: [PATCH 237/820] [SCSI] qla4xxx: Update driver version to 5.02.00-k11 Signed-off-by: Vikas Chaudhary Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_version.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/qla4xxx/ql4_version.h b/drivers/scsi/qla4xxx/ql4_version.h index 26a3fa34a33c..7d04eb05c45e 100644 --- a/drivers/scsi/qla4xxx/ql4_version.h +++ b/drivers/scsi/qla4xxx/ql4_version.h @@ -5,4 +5,4 @@ * See LICENSE.qla4xxx for copyright and licensing details. */ -#define QLA4XXX_DRIVER_VERSION "5.02.00-k10" +#define QLA4XXX_DRIVER_VERSION "5.02.00-k11" From f1633011e4fcb65bf65aebeaf4024efdc4c49d6e Mon Sep 17 00:00:00 2001 From: Robert Love Date: Fri, 16 Dec 2011 14:24:49 -0800 Subject: [PATCH 238/820] [SCSI] fcoe: remove double check if skb is nonlinear skb_linearize already has a check for skb_is_nonlinear, there is no need to duplicate the check in fcoe.c. This patch simply removes the unnecessary check and calls skb_linearize unconditionally. Reported-by: patrick kelle Signed-off-by: Robert Love Acked-by: patrick kelle Signed-off-by: James Bottomley --- drivers/scsi/fcoe/fcoe.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c index 8d67467dd9ce..e535f95e4772 100644 --- a/drivers/scsi/fcoe/fcoe.c +++ b/drivers/scsi/fcoe/fcoe.c @@ -1671,8 +1671,7 @@ static void fcoe_recv_frame(struct sk_buff *skb) skb->dev ? skb->dev->name : ""); port = lport_priv(lport); - if (skb_is_nonlinear(skb)) - skb_linearize(skb); /* not ideal */ + skb_linearize(skb); /* check for skb_is_nonlinear is within skb_linearize */ /* * Frame length checks and setting up the header pointers From 6bc6204e3b7f4ac235e98e362714213369d6a012 Mon Sep 17 00:00:00 2001 From: Krishna Gudipati Date: Tue, 20 Dec 2011 18:54:30 -0800 Subject: [PATCH 239/820] [SCSI] bfa: Fix possible NULL pointer dereference in lunmask add/delete. Patch fixes the possible NULL pointer dereference when we try to add or delete a rpwwn to the lunmask config which is not zoned to this port. Check if the FCS rport is not NULL before de-referencing it. Signed-off-by: Krishna Gudipati Signed-off-by: James Bottomley --- drivers/scsi/bfa/bfa_fcpim.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/bfa/bfa_fcpim.c b/drivers/scsi/bfa/bfa_fcpim.c index e07bd4745d8b..846f15b6e1d8 100644 --- a/drivers/scsi/bfa/bfa_fcpim.c +++ b/drivers/scsi/bfa/bfa_fcpim.c @@ -2647,7 +2647,8 @@ bfa_fcpim_lunmask_add(struct bfa_s *bfa, u16 vf_id, wwn_t *pwwn, if (port) { *pwwn = port->port_cfg.pwwn; rp_fcs = bfa_fcs_lport_get_rport_by_pwwn(port, rpwwn); - rp = rp_fcs->bfa_rport; + if (rp_fcs) + rp = rp_fcs->bfa_rport; } lunm_list = bfa_get_lun_mask_list(bfa); @@ -2715,7 +2716,8 @@ bfa_fcpim_lunmask_delete(struct bfa_s *bfa, u16 vf_id, wwn_t *pwwn, if (port) { *pwwn = port->port_cfg.pwwn; rp_fcs = bfa_fcs_lport_get_rport_by_pwwn(port, rpwwn); - rp = rp_fcs->bfa_rport; + if (rp_fcs) + rp = rp_fcs->bfa_rport; } } From 8ca2dd87e7f29df99389e9dc41f7a45a35223672 Mon Sep 17 00:00:00 2001 From: Krishna Gudipati Date: Tue, 20 Dec 2011 18:55:07 -0800 Subject: [PATCH 240/820] [SCSI] bfa: Revert back the current LUN Masking Implementation. This patch reverts the current LUN Masking Implementation. We re-implemented this feature using the SCSI Slave Callout's as per the review comments. Signed-off-by: Krishna Gudipati Signed-off-by: James Bottomley --- drivers/scsi/bfa/bfa_defs_svc.h | 7 +- drivers/scsi/bfa/bfa_fc.h | 155 ------------ drivers/scsi/bfa/bfa_fcpim.c | 410 +------------------------------- drivers/scsi/bfa/bfa_fcpim.h | 7 - drivers/scsi/bfa/bfa_svc.h | 5 - 5 files changed, 2 insertions(+), 582 deletions(-) diff --git a/drivers/scsi/bfa/bfa_defs_svc.h b/drivers/scsi/bfa/bfa_defs_svc.h index 78963be2c4fb..cb07c628b2f1 100644 --- a/drivers/scsi/bfa/bfa_defs_svc.h +++ b/drivers/scsi/bfa/bfa_defs_svc.h @@ -673,12 +673,7 @@ struct bfa_itnim_iostats_s { u32 tm_iocdowns; /* TM cleaned-up due to IOC down */ u32 tm_cleanups; /* TM cleanup requests */ u32 tm_cleanup_comps; /* TM cleanup completions */ - u32 lm_lun_across_sg; /* LM lun is across sg data buf */ - u32 lm_lun_not_sup; /* LM lun not supported */ - u32 lm_rpl_data_changed; /* LM report-lun data changed */ - u32 lm_wire_residue_changed; /* LM report-lun rsp residue changed */ - u32 lm_small_buf_addresidue; /* LM buf smaller than reported cnt */ - u32 lm_lun_not_rdy; /* LM lun not ready */ + u32 rsvd[6]; }; /* Modify char* port_stt[] in bfal_port.c if a new state was added */ diff --git a/drivers/scsi/bfa/bfa_fc.h b/drivers/scsi/bfa/bfa_fc.h index 50b6a1c86195..8d0b88f67a38 100644 --- a/drivers/scsi/bfa/bfa_fc.h +++ b/drivers/scsi/bfa/bfa_fc.h @@ -56,161 +56,6 @@ struct scsi_cdb_s { #define SCSI_MAX_ALLOC_LEN 0xFF /* maximum allocarion length */ -#define SCSI_SENSE_CUR_ERR 0x70 -#define SCSI_SENSE_DEF_ERR 0x71 - -/* - * SCSI additional sense codes - */ -#define SCSI_ASC_LUN_NOT_READY 0x04 -#define SCSI_ASC_LUN_NOT_SUPPORTED 0x25 -#define SCSI_ASC_TOCC 0x3F - -/* - * SCSI additional sense code qualifiers - */ -#define SCSI_ASCQ_MAN_INTR_REQ 0x03 /* manual intervention req */ -#define SCSI_ASCQ_RL_DATA_CHANGED 0x0E /* report luns data changed */ - -/* - * Methods of reporting informational exceptions - */ -#define SCSI_MP_IEC_UNIT_ATTN 0x2 /* generate unit attention */ - -struct scsi_report_luns_data_s { - u32 lun_list_length; /* length of LUN list length */ - u32 reserved; - struct scsi_lun lun[1]; /* first LUN in lun list */ -}; - -struct scsi_inquiry_vendor_s { - u8 vendor_id[8]; -}; - -struct scsi_inquiry_prodid_s { - u8 product_id[16]; -}; - -struct scsi_inquiry_prodrev_s { - u8 product_rev[4]; -}; - -struct scsi_inquiry_data_s { -#ifdef __BIG_ENDIAN - u8 peripheral_qual:3; /* peripheral qualifier */ - u8 device_type:5; /* peripheral device type */ - u8 rmb:1; /* removable medium bit */ - u8 device_type_mod:7; /* device type modifier */ - u8 version; - u8 aenc:1; /* async evt notification capability */ - u8 trm_iop:1; /* terminate I/O process */ - u8 norm_aca:1; /* normal ACA supported */ - u8 hi_support:1; /* SCSI-3: supports REPORT LUNS */ - u8 rsp_data_format:4; - u8 additional_len; - u8 sccs:1; - u8 reserved1:7; - u8 reserved2:1; - u8 enc_serv:1; /* enclosure service component */ - u8 reserved3:1; - u8 multi_port:1; /* multi-port device */ - u8 m_chngr:1; /* device in medium transport element */ - u8 ack_req_q:1; /* SIP specific bit */ - u8 addr32:1; /* SIP specific bit */ - u8 addr16:1; /* SIP specific bit */ - u8 rel_adr:1; /* relative address */ - u8 w_bus32:1; - u8 w_bus16:1; - u8 synchronous:1; - u8 linked_commands:1; - u8 trans_dis:1; - u8 cmd_queue:1; /* command queueing supported */ - u8 soft_reset:1; /* soft reset alternative (VS) */ -#else - u8 device_type:5; /* peripheral device type */ - u8 peripheral_qual:3; /* peripheral qualifier */ - u8 device_type_mod:7; /* device type modifier */ - u8 rmb:1; /* removable medium bit */ - u8 version; - u8 rsp_data_format:4; - u8 hi_support:1; /* SCSI-3: supports REPORT LUNS */ - u8 norm_aca:1; /* normal ACA supported */ - u8 terminate_iop:1;/* terminate I/O process */ - u8 aenc:1; /* async evt notification capability */ - u8 additional_len; - u8 reserved1:7; - u8 sccs:1; - u8 addr16:1; /* SIP specific bit */ - u8 addr32:1; /* SIP specific bit */ - u8 ack_req_q:1; /* SIP specific bit */ - u8 m_chngr:1; /* device in medium transport element */ - u8 multi_port:1; /* multi-port device */ - u8 reserved3:1; /* TBD - Vendor Specific */ - u8 enc_serv:1; /* enclosure service component */ - u8 reserved2:1; - u8 soft_seset:1; /* soft reset alternative (VS) */ - u8 cmd_queue:1; /* command queueing supported */ - u8 trans_dis:1; - u8 linked_commands:1; - u8 synchronous:1; - u8 w_bus16:1; - u8 w_bus32:1; - u8 rel_adr:1; /* relative address */ -#endif - struct scsi_inquiry_vendor_s vendor_id; - struct scsi_inquiry_prodid_s product_id; - struct scsi_inquiry_prodrev_s product_rev; - u8 vendor_specific[20]; - u8 reserved4[40]; -}; - -/* - * SCSI sense data format - */ -struct scsi_sense_s { -#ifdef __BIG_ENDIAN - u8 valid:1; - u8 rsp_code:7; -#else - u8 rsp_code:7; - u8 valid:1; -#endif - u8 seg_num; -#ifdef __BIG_ENDIAN - u8 file_mark:1; - u8 eom:1; /* end of media */ - u8 ili:1; /* incorrect length indicator */ - u8 reserved:1; - u8 sense_key:4; -#else - u8 sense_key:4; - u8 reserved:1; - u8 ili:1; /* incorrect length indicator */ - u8 eom:1; /* end of media */ - u8 file_mark:1; -#endif - u8 information[4]; /* device-type or cmd specific info */ - u8 add_sense_length; /* additional sense length */ - u8 command_info[4];/* command specific information */ - u8 asc; /* additional sense code */ - u8 ascq; /* additional sense code qualifier */ - u8 fru_code; /* field replaceable unit code */ -#ifdef __BIG_ENDIAN - u8 sksv:1; /* sense key specific valid */ - u8 c_d:1; /* command/data bit */ - u8 res1:2; - u8 bpv:1; /* bit pointer valid */ - u8 bpointer:3; /* bit pointer */ -#else - u8 bpointer:3; /* bit pointer */ - u8 bpv:1; /* bit pointer valid */ - u8 res1:2; - u8 c_d:1; /* command/data bit */ - u8 sksv:1; /* sense key specific valid */ -#endif - u8 fpointer[2]; /* field pointer */ -}; - /* * Fibre Channel Header Structure (FCHS) definition */ diff --git a/drivers/scsi/bfa/bfa_fcpim.c b/drivers/scsi/bfa/bfa_fcpim.c index 846f15b6e1d8..f0f80e282e39 100644 --- a/drivers/scsi/bfa/bfa_fcpim.c +++ b/drivers/scsi/bfa/bfa_fcpim.c @@ -24,8 +24,6 @@ BFA_TRC_FILE(HAL, FCPIM); * BFA ITNIM Related definitions */ static void bfa_itnim_update_del_itn_stats(struct bfa_itnim_s *itnim); -static bfa_boolean_t bfa_ioim_lm_proc_rpl_data(struct bfa_ioim_s *ioim); -static bfa_boolean_t bfa_ioim_lm_proc_inq_data(struct bfa_ioim_s *ioim); static void bfa_ioim_lm_init(struct bfa_s *bfa); #define BFA_ITNIM_FROM_TAG(_fcpim, _tag) \ @@ -60,14 +58,6 @@ static void bfa_ioim_lm_init(struct bfa_s *bfa); } \ } while (0) -#define bfa_ioim_rp_wwn(__ioim) \ - (((struct bfa_fcs_rport_s *) \ - (__ioim)->itnim->rport->rport_drv)->pwwn) - -#define bfa_ioim_lp_wwn(__ioim) \ - ((BFA_LPS_FROM_TAG(BFA_LPS_MOD((__ioim)->bfa), \ - (__ioim)->itnim->rport->rport_info.lp_tag))->pwwn) \ - #define bfa_itnim_sler_cb(__itnim) do { \ if ((__itnim)->bfa->fcs) \ bfa_cb_itnim_sler((__itnim)->ditn); \ @@ -77,13 +67,6 @@ static void bfa_ioim_lm_init(struct bfa_s *bfa); } \ } while (0) -enum bfa_ioim_lm_status { - BFA_IOIM_LM_PRESENT = 1, - BFA_IOIM_LM_LUN_NOT_SUP = 2, - BFA_IOIM_LM_RPL_DATA_CHANGED = 3, - BFA_IOIM_LM_LUN_NOT_RDY = 4, -}; - enum bfa_ioim_lm_ua_status { BFA_IOIM_LM_UA_RESET = 0, BFA_IOIM_LM_UA_SET = 1, @@ -145,9 +128,6 @@ enum bfa_ioim_event { BFA_IOIM_SM_TMDONE = 16, /* IO cleanup from tskim */ BFA_IOIM_SM_HWFAIL = 17, /* IOC h/w failure event */ BFA_IOIM_SM_IOTOV = 18, /* ITN offline TOV */ - BFA_IOIM_SM_LM_LUN_NOT_SUP = 19,/* lunmask lun not supported */ - BFA_IOIM_SM_LM_RPL_DC = 20, /* lunmask report-lun data changed */ - BFA_IOIM_SM_LM_LUN_NOT_RDY = 21,/* lunmask lun not ready */ }; @@ -245,9 +225,6 @@ static void __bfa_cb_ioim_abort(void *cbarg, bfa_boolean_t complete); static void __bfa_cb_ioim_failed(void *cbarg, bfa_boolean_t complete); static void __bfa_cb_ioim_pathtov(void *cbarg, bfa_boolean_t complete); static bfa_boolean_t bfa_ioim_is_abortable(struct bfa_ioim_s *ioim); -static void __bfa_cb_ioim_lm_lun_not_sup(void *cbarg, bfa_boolean_t complete); -static void __bfa_cb_ioim_lm_rpl_dc(void *cbarg, bfa_boolean_t complete); -static void __bfa_cb_ioim_lm_lun_not_rdy(void *cbarg, bfa_boolean_t complete); /* * forward declaration of BFA IO state machine @@ -445,12 +422,6 @@ bfa_fcpim_add_stats(struct bfa_itnim_iostats_s *lstats, bfa_fcpim_add_iostats(lstats, rstats, output_reqs); bfa_fcpim_add_iostats(lstats, rstats, rd_throughput); bfa_fcpim_add_iostats(lstats, rstats, wr_throughput); - bfa_fcpim_add_iostats(lstats, rstats, lm_lun_across_sg); - bfa_fcpim_add_iostats(lstats, rstats, lm_lun_not_sup); - bfa_fcpim_add_iostats(lstats, rstats, lm_rpl_data_changed); - bfa_fcpim_add_iostats(lstats, rstats, lm_wire_residue_changed); - bfa_fcpim_add_iostats(lstats, rstats, lm_small_buf_addresidue); - bfa_fcpim_add_iostats(lstats, rstats, lm_lun_not_rdy); } bfa_status_t @@ -1580,27 +1551,6 @@ bfa_ioim_sm_uninit(struct bfa_ioim_s *ioim, enum bfa_ioim_event event) __bfa_cb_ioim_abort, ioim); break; - case BFA_IOIM_SM_LM_LUN_NOT_SUP: - bfa_sm_set_state(ioim, bfa_ioim_sm_hcb); - bfa_ioim_move_to_comp_q(ioim); - bfa_cb_queue(ioim->bfa, &ioim->hcb_qe, - __bfa_cb_ioim_lm_lun_not_sup, ioim); - break; - - case BFA_IOIM_SM_LM_RPL_DC: - bfa_sm_set_state(ioim, bfa_ioim_sm_hcb); - bfa_ioim_move_to_comp_q(ioim); - bfa_cb_queue(ioim->bfa, &ioim->hcb_qe, - __bfa_cb_ioim_lm_rpl_dc, ioim); - break; - - case BFA_IOIM_SM_LM_LUN_NOT_RDY: - bfa_sm_set_state(ioim, bfa_ioim_sm_hcb); - bfa_ioim_move_to_comp_q(ioim); - bfa_cb_queue(ioim->bfa, &ioim->hcb_qe, - __bfa_cb_ioim_lm_lun_not_rdy, ioim); - break; - default: bfa_sm_fault(ioim->bfa, event); } @@ -2160,243 +2110,6 @@ bfa_ioim_lm_init(struct bfa_s *bfa) } } -/* - * Validate LUN for LUN masking - */ -static enum bfa_ioim_lm_status -bfa_ioim_lm_check(struct bfa_ioim_s *ioim, struct bfa_lps_s *lps, - struct bfa_rport_s *rp, struct scsi_lun lun) -{ - u8 i; - struct bfa_lun_mask_s *lun_list = bfa_get_lun_mask_list(ioim->bfa); - struct scsi_cmnd *cmnd = (struct scsi_cmnd *)ioim->dio; - struct scsi_cdb_s *cdb = (struct scsi_cdb_s *)cmnd->cmnd; - - if ((cdb->scsi_cdb[0] == REPORT_LUNS) && - (scsilun_to_int((struct scsi_lun *)&lun) == 0)) { - ioim->proc_rsp_data = bfa_ioim_lm_proc_rpl_data; - return BFA_IOIM_LM_PRESENT; - } - - for (i = 0; i < MAX_LUN_MASK_CFG; i++) { - - if (lun_list[i].state != BFA_IOIM_LUN_MASK_ACTIVE) - continue; - - if ((scsilun_to_int((struct scsi_lun *)&lun_list[i].lun) == - scsilun_to_int((struct scsi_lun *)&lun)) - && (rp->rport_tag == lun_list[i].rp_tag) - && ((u8)ioim->itnim->rport->rport_info.lp_tag == - lun_list[i].lp_tag)) { - bfa_trc(ioim->bfa, lun_list[i].rp_tag); - bfa_trc(ioim->bfa, lun_list[i].lp_tag); - bfa_trc(ioim->bfa, scsilun_to_int( - (struct scsi_lun *)&lun_list[i].lun)); - - if ((lun_list[i].ua == BFA_IOIM_LM_UA_SET) && - ((cdb->scsi_cdb[0] != INQUIRY) || - (cdb->scsi_cdb[0] != REPORT_LUNS))) { - lun_list[i].ua = BFA_IOIM_LM_UA_RESET; - return BFA_IOIM_LM_RPL_DATA_CHANGED; - } - - if (cdb->scsi_cdb[0] == REPORT_LUNS) - ioim->proc_rsp_data = bfa_ioim_lm_proc_rpl_data; - - return BFA_IOIM_LM_PRESENT; - } - } - - if ((cdb->scsi_cdb[0] == INQUIRY) && - (scsilun_to_int((struct scsi_lun *)&lun) == 0)) { - ioim->proc_rsp_data = bfa_ioim_lm_proc_inq_data; - return BFA_IOIM_LM_PRESENT; - } - - if (cdb->scsi_cdb[0] == TEST_UNIT_READY) - return BFA_IOIM_LM_LUN_NOT_RDY; - - return BFA_IOIM_LM_LUN_NOT_SUP; -} - -static bfa_boolean_t -bfa_ioim_lm_proc_rsp_data_dummy(struct bfa_ioim_s *ioim) -{ - return BFA_TRUE; -} - -static void -bfa_ioim_lm_fetch_lun(struct bfa_ioim_s *ioim, u8 *rl_data, int offset, - int buf_lun_cnt) -{ - struct bfa_lun_mask_s *lun_list = bfa_get_lun_mask_list(ioim->bfa); - struct scsi_lun *lun_data = (struct scsi_lun *)(rl_data + offset); - struct scsi_lun lun; - int i, j; - - bfa_trc(ioim->bfa, buf_lun_cnt); - for (j = 0; j < buf_lun_cnt; j++) { - lun = *((struct scsi_lun *)(lun_data + j)); - for (i = 0; i < MAX_LUN_MASK_CFG; i++) { - if (lun_list[i].state != BFA_IOIM_LUN_MASK_ACTIVE) - continue; - if ((lun_list[i].rp_wwn == bfa_ioim_rp_wwn(ioim)) && - (lun_list[i].lp_wwn == bfa_ioim_lp_wwn(ioim)) && - (scsilun_to_int((struct scsi_lun *)&lun_list[i].lun) - == scsilun_to_int((struct scsi_lun *)&lun))) { - lun_list[i].state = BFA_IOIM_LUN_MASK_FETCHED; - break; - } - } /* next lun in mask DB */ - } /* next lun in buf */ -} - -static int -bfa_ioim_lm_update_lun_sg(struct bfa_ioim_s *ioim, u32 *pgdlen, - struct scsi_report_luns_data_s *rl) -{ - struct scsi_cmnd *cmnd = (struct scsi_cmnd *)ioim->dio; - struct scatterlist *sg = scsi_sglist(cmnd); - struct bfa_lun_mask_s *lun_list = bfa_get_lun_mask_list(ioim->bfa); - struct scsi_lun *prev_rl_data = NULL, *base_rl_data; - int i, j, sgeid, lun_fetched_cnt = 0, prev_sg_len = 0, base_count; - int lun_across_sg_bytes, bytes_from_next_buf; - u64 last_lun, temp_last_lun; - - /* fetch luns from the first sg element */ - bfa_ioim_lm_fetch_lun(ioim, (u8 *)(rl->lun), 0, - (sg_dma_len(sg) / sizeof(struct scsi_lun)) - 1); - - /* fetch luns from multiple sg elements */ - scsi_for_each_sg(cmnd, sg, scsi_sg_count(cmnd), sgeid) { - if (sgeid == 0) { - prev_sg_len = sg_dma_len(sg); - prev_rl_data = (struct scsi_lun *) - phys_to_virt(sg_dma_address(sg)); - continue; - } - - /* if the buf is having more data */ - lun_across_sg_bytes = prev_sg_len % sizeof(struct scsi_lun); - if (lun_across_sg_bytes) { - bfa_trc(ioim->bfa, lun_across_sg_bytes); - bfa_stats(ioim->itnim, lm_lun_across_sg); - bytes_from_next_buf = sizeof(struct scsi_lun) - - lun_across_sg_bytes; - - /* from next buf take higher bytes */ - temp_last_lun = *((u64 *) - phys_to_virt(sg_dma_address(sg))); - last_lun |= temp_last_lun >> - (lun_across_sg_bytes * BITS_PER_BYTE); - - /* from prev buf take higher bytes */ - temp_last_lun = *((u64 *)(prev_rl_data + - (prev_sg_len - lun_across_sg_bytes))); - temp_last_lun >>= bytes_from_next_buf * BITS_PER_BYTE; - last_lun = last_lun | (temp_last_lun << - (bytes_from_next_buf * BITS_PER_BYTE)); - - bfa_ioim_lm_fetch_lun(ioim, (u8 *)&last_lun, 0, 1); - } else - bytes_from_next_buf = 0; - - *pgdlen += sg_dma_len(sg); - prev_sg_len = sg_dma_len(sg); - prev_rl_data = (struct scsi_lun *) - phys_to_virt(sg_dma_address(sg)); - bfa_ioim_lm_fetch_lun(ioim, (u8 *)prev_rl_data, - bytes_from_next_buf, - sg_dma_len(sg) / sizeof(struct scsi_lun)); - } - - /* update the report luns data - based on fetched luns */ - sg = scsi_sglist(cmnd); - base_rl_data = (struct scsi_lun *)rl->lun; - base_count = (sg_dma_len(sg) / sizeof(struct scsi_lun)) - 1; - for (i = 0, j = 0; i < MAX_LUN_MASK_CFG; i++) { - if (lun_list[i].state == BFA_IOIM_LUN_MASK_FETCHED) { - base_rl_data[j] = lun_list[i].lun; - lun_list[i].state = BFA_IOIM_LUN_MASK_ACTIVE; - j++; - lun_fetched_cnt++; - } - - if (j > base_count) { - j = 0; - sg = sg_next(sg); - base_rl_data = (struct scsi_lun *) - phys_to_virt(sg_dma_address(sg)); - base_count = sg_dma_len(sg) / sizeof(struct scsi_lun); - } - } - - bfa_trc(ioim->bfa, lun_fetched_cnt); - return lun_fetched_cnt; -} - -static bfa_boolean_t -bfa_ioim_lm_proc_inq_data(struct bfa_ioim_s *ioim) -{ - struct scsi_inquiry_data_s *inq; - struct scatterlist *sg = scsi_sglist((struct scsi_cmnd *)ioim->dio); - - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; - inq = (struct scsi_inquiry_data_s *)phys_to_virt(sg_dma_address(sg)); - - bfa_trc(ioim->bfa, inq->device_type); - inq->peripheral_qual = SCSI_INQ_PQ_NOT_CON; - return 0; -} - -static bfa_boolean_t -bfa_ioim_lm_proc_rpl_data(struct bfa_ioim_s *ioim) -{ - struct scsi_cmnd *cmnd = (struct scsi_cmnd *)ioim->dio; - struct scatterlist *sg = scsi_sglist(cmnd); - struct bfi_ioim_rsp_s *m; - struct scsi_report_luns_data_s *rl = NULL; - int lun_count = 0, lun_fetched_cnt = 0; - u32 residue, pgdlen = 0; - - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; - if (bfa_get_lun_mask_status(ioim->bfa) != BFA_LUNMASK_ENABLED) - return BFA_TRUE; - - m = (struct bfi_ioim_rsp_s *) &ioim->iosp->comp_rspmsg; - if (m->scsi_status == SCSI_STATUS_CHECK_CONDITION) - return BFA_TRUE; - - pgdlen = sg_dma_len(sg); - bfa_trc(ioim->bfa, pgdlen); - rl = (struct scsi_report_luns_data_s *)phys_to_virt(sg_dma_address(sg)); - lun_count = cpu_to_be32(rl->lun_list_length) / sizeof(struct scsi_lun); - lun_fetched_cnt = bfa_ioim_lm_update_lun_sg(ioim, &pgdlen, rl); - - if (lun_count == lun_fetched_cnt) - return BFA_TRUE; - - bfa_trc(ioim->bfa, lun_count); - bfa_trc(ioim->bfa, lun_fetched_cnt); - bfa_trc(ioim->bfa, be32_to_cpu(rl->lun_list_length)); - - if (be32_to_cpu(rl->lun_list_length) <= pgdlen) - rl->lun_list_length = be32_to_cpu(lun_fetched_cnt) * - sizeof(struct scsi_lun); - else - bfa_stats(ioim->itnim, lm_small_buf_addresidue); - - bfa_trc(ioim->bfa, be32_to_cpu(rl->lun_list_length)); - bfa_trc(ioim->bfa, be32_to_cpu(m->residue)); - - residue = be32_to_cpu(m->residue); - residue += (lun_count - lun_fetched_cnt) * sizeof(struct scsi_lun); - bfa_stats(ioim->itnim, lm_wire_residue_changed); - m->residue = be32_to_cpu(residue); - bfa_trc(ioim->bfa, ioim->nsges); - return BFA_FALSE; -} - static void __bfa_cb_ioim_good_comp(void *cbarg, bfa_boolean_t complete) { @@ -2454,83 +2167,6 @@ __bfa_cb_ioim_comp(void *cbarg, bfa_boolean_t complete) m->scsi_status, sns_len, snsinfo, residue); } -static void -__bfa_cb_ioim_lm_lun_not_sup(void *cbarg, bfa_boolean_t complete) -{ - struct bfa_ioim_s *ioim = cbarg; - int sns_len = 0xD; - u32 residue = scsi_bufflen((struct scsi_cmnd *)ioim->dio); - struct scsi_sense_s *snsinfo; - - if (!complete) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_HCB); - return; - } - - snsinfo = (struct scsi_sense_s *)BFA_SNSINFO_FROM_TAG( - ioim->fcpim->fcp, ioim->iotag); - snsinfo->rsp_code = SCSI_SENSE_CUR_ERR; - snsinfo->add_sense_length = 0xa; - snsinfo->asc = SCSI_ASC_LUN_NOT_SUPPORTED; - snsinfo->sense_key = ILLEGAL_REQUEST; - bfa_trc(ioim->bfa, residue); - bfa_cb_ioim_done(ioim->bfa->bfad, ioim->dio, BFI_IOIM_STS_OK, - SCSI_STATUS_CHECK_CONDITION, sns_len, - (u8 *)snsinfo, residue); -} - -static void -__bfa_cb_ioim_lm_rpl_dc(void *cbarg, bfa_boolean_t complete) -{ - struct bfa_ioim_s *ioim = cbarg; - int sns_len = 0xD; - u32 residue = scsi_bufflen((struct scsi_cmnd *)ioim->dio); - struct scsi_sense_s *snsinfo; - - if (!complete) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_HCB); - return; - } - - snsinfo = (struct scsi_sense_s *)BFA_SNSINFO_FROM_TAG(ioim->fcpim->fcp, - ioim->iotag); - snsinfo->rsp_code = SCSI_SENSE_CUR_ERR; - snsinfo->sense_key = SCSI_MP_IEC_UNIT_ATTN; - snsinfo->asc = SCSI_ASC_TOCC; - snsinfo->add_sense_length = 0x6; - snsinfo->ascq = SCSI_ASCQ_RL_DATA_CHANGED; - bfa_trc(ioim->bfa, residue); - bfa_cb_ioim_done(ioim->bfa->bfad, ioim->dio, BFI_IOIM_STS_OK, - SCSI_STATUS_CHECK_CONDITION, sns_len, - (u8 *)snsinfo, residue); -} - -static void -__bfa_cb_ioim_lm_lun_not_rdy(void *cbarg, bfa_boolean_t complete) -{ - struct bfa_ioim_s *ioim = cbarg; - int sns_len = 0xD; - u32 residue = scsi_bufflen((struct scsi_cmnd *)ioim->dio); - struct scsi_sense_s *snsinfo; - - if (!complete) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_HCB); - return; - } - - snsinfo = (struct scsi_sense_s *)BFA_SNSINFO_FROM_TAG( - ioim->fcpim->fcp, ioim->iotag); - snsinfo->rsp_code = SCSI_SENSE_CUR_ERR; - snsinfo->add_sense_length = 0xa; - snsinfo->sense_key = NOT_READY; - snsinfo->asc = SCSI_ASC_LUN_NOT_READY; - snsinfo->ascq = SCSI_ASCQ_MAN_INTR_REQ; - bfa_trc(ioim->bfa, residue); - bfa_cb_ioim_done(ioim->bfa->bfad, ioim->dio, BFI_IOIM_STS_OK, - SCSI_STATUS_CHECK_CONDITION, sns_len, - (u8 *)snsinfo, residue); -} - void bfa_fcpim_lunmask_rp_update(struct bfa_s *bfa, wwn_t lp_wwn, wwn_t rp_wwn, u16 rp_tag, u8 lp_tag) @@ -2759,7 +2395,6 @@ __bfa_cb_ioim_failed(void *cbarg, bfa_boolean_t complete) return; } - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; bfa_cb_ioim_done(ioim->bfa->bfad, ioim->dio, BFI_IOIM_STS_ABORTED, 0, 0, NULL, 0); } @@ -2775,7 +2410,6 @@ __bfa_cb_ioim_pathtov(void *cbarg, bfa_boolean_t complete) return; } - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; bfa_cb_ioim_done(ioim->bfa->bfad, ioim->dio, BFI_IOIM_STS_PATHTOV, 0, 0, NULL, 0); } @@ -2790,7 +2424,6 @@ __bfa_cb_ioim_abort(void *cbarg, bfa_boolean_t complete) return; } - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; bfa_cb_ioim_abort(ioim->bfa->bfad, ioim->dio); } @@ -3134,7 +2767,6 @@ bfa_ioim_attach(struct bfa_fcpim_s *fcpim) ioim->bfa = fcpim->bfa; ioim->fcpim = fcpim; ioim->iosp = iosp; - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; INIT_LIST_HEAD(&ioim->sgpg_q); bfa_reqq_winit(&ioim->iosp->reqq_wait, bfa_ioim_qresume, ioim); @@ -3172,7 +2804,6 @@ bfa_ioim_isr(struct bfa_s *bfa, struct bfi_msg_s *m) evt = BFA_IOIM_SM_DONE; else evt = BFA_IOIM_SM_COMP; - ioim->proc_rsp_data(ioim); break; case BFI_IOIM_STS_TIMEDOUT: @@ -3208,7 +2839,6 @@ bfa_ioim_isr(struct bfa_s *bfa, struct bfi_msg_s *m) if (rsp->abort_tag != ioim->abort_tag) { bfa_trc(ioim->bfa, rsp->abort_tag); bfa_trc(ioim->bfa, ioim->abort_tag); - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; return; } @@ -3227,7 +2857,6 @@ bfa_ioim_isr(struct bfa_s *bfa, struct bfi_msg_s *m) WARN_ON(1); } - ioim->proc_rsp_data = bfa_ioim_lm_proc_rsp_data_dummy; bfa_sm_send_event(ioim, evt); } @@ -3246,15 +2875,7 @@ bfa_ioim_good_comp_isr(struct bfa_s *bfa, struct bfi_msg_s *m) bfa_ioim_cb_profile_comp(fcpim, ioim); - if (bfa_get_lun_mask_status(bfa) != BFA_LUNMASK_ENABLED) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_COMP_GOOD); - return; - } - - if (ioim->proc_rsp_data(ioim) == BFA_TRUE) - bfa_sm_send_event(ioim, BFA_IOIM_SM_COMP_GOOD); - else - bfa_sm_send_event(ioim, BFA_IOIM_SM_COMP); + bfa_sm_send_event(ioim, BFA_IOIM_SM_COMP_GOOD); } /* @@ -3366,35 +2987,6 @@ bfa_ioim_free(struct bfa_ioim_s *ioim) void bfa_ioim_start(struct bfa_ioim_s *ioim) { - struct scsi_cmnd *cmnd = (struct scsi_cmnd *)ioim->dio; - struct bfa_lps_s *lps; - enum bfa_ioim_lm_status status; - struct scsi_lun scsilun; - - if (bfa_get_lun_mask_status(ioim->bfa) == BFA_LUNMASK_ENABLED) { - lps = BFA_IOIM_TO_LPS(ioim); - int_to_scsilun(cmnd->device->lun, &scsilun); - status = bfa_ioim_lm_check(ioim, lps, - ioim->itnim->rport, scsilun); - if (status == BFA_IOIM_LM_LUN_NOT_RDY) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_LM_LUN_NOT_RDY); - bfa_stats(ioim->itnim, lm_lun_not_rdy); - return; - } - - if (status == BFA_IOIM_LM_LUN_NOT_SUP) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_LM_LUN_NOT_SUP); - bfa_stats(ioim->itnim, lm_lun_not_sup); - return; - } - - if (status == BFA_IOIM_LM_RPL_DATA_CHANGED) { - bfa_sm_send_event(ioim, BFA_IOIM_SM_LM_RPL_DC); - bfa_stats(ioim->itnim, lm_rpl_data_changed); - return; - } - } - bfa_ioim_cb_profile_start(ioim->fcpim, ioim); /* diff --git a/drivers/scsi/bfa/bfa_fcpim.h b/drivers/scsi/bfa/bfa_fcpim.h index 1080bcb81cb7..36f26da80f76 100644 --- a/drivers/scsi/bfa/bfa_fcpim.h +++ b/drivers/scsi/bfa/bfa_fcpim.h @@ -110,7 +110,6 @@ struct bfad_ioim_s; struct bfad_tskim_s; typedef void (*bfa_fcpim_profile_t) (struct bfa_ioim_s *ioim); -typedef bfa_boolean_t (*bfa_ioim_lm_proc_rsp_data_t) (struct bfa_ioim_s *ioim); struct bfa_fcpim_s { struct bfa_s *bfa; @@ -124,7 +123,6 @@ struct bfa_fcpim_s { u32 path_tov; u16 q_depth; u8 reqq; /* Request queue to be used */ - u8 lun_masking_pending; struct list_head itnim_q; /* queue of active itnim */ struct list_head ioim_resfree_q; /* IOs waiting for f/w */ struct list_head ioim_comp_q; /* IO global comp Q */ @@ -181,7 +179,6 @@ struct bfa_ioim_s { u8 reqq; /* Request queue for I/O */ u8 mode; /* IO is passthrough or not */ u64 start_time; /* IO's Profile start val */ - bfa_ioim_lm_proc_rsp_data_t proc_rsp_data; /* RSP data adjust */ }; struct bfa_ioim_sp_s { @@ -261,10 +258,6 @@ struct bfa_itnim_s { (__ioim)->iotag |= k << BFA_IOIM_RETRY_TAG_OFFSET; \ } while (0) -#define BFA_IOIM_TO_LPS(__ioim) \ - BFA_LPS_FROM_TAG(BFA_LPS_MOD(__ioim->bfa), \ - __ioim->itnim->rport->rport_info.lp_tag) - static inline bfa_boolean_t bfa_ioim_maxretry_reached(struct bfa_ioim_s *ioim) { diff --git a/drivers/scsi/bfa/bfa_svc.h b/drivers/scsi/bfa/bfa_svc.h index 95adb86d3769..b52cbb6bcd5a 100644 --- a/drivers/scsi/bfa/bfa_svc.h +++ b/drivers/scsi/bfa/bfa_svc.h @@ -582,11 +582,6 @@ void bfa_cb_rport_qos_scn_prio(void *rport, #define BFA_LP_TAG_INVALID 0xff void bfa_rport_set_lunmask(struct bfa_s *bfa, struct bfa_rport_s *rp); void bfa_rport_unset_lunmask(struct bfa_s *bfa, struct bfa_rport_s *rp); -bfa_boolean_t bfa_rport_lunmask_active(struct bfa_rport_s *rp); -wwn_t bfa_rport_get_pwwn(struct bfa_s *bfa, struct bfa_rport_s *rp); -struct bfa_rport_s *bfa_rport_get_by_wwn(struct bfa_s *bfa, u16 vf_id, - wwn_t *lpwwn, wwn_t rpwwn); -void *bfa_cb_get_rp_by_wwn(void *arg, u16 vf_id, wwn_t *lpwwn, wwn_t rpwwn); /* * bfa fcxp API functions From 5b7db7af522d9f281ff8bf540d2b5cbea2206b27 Mon Sep 17 00:00:00 2001 From: Krishna Gudipati Date: Tue, 20 Dec 2011 18:58:32 -0800 Subject: [PATCH 241/820] [SCSI] bfa: Implement LUN Masking feature using the SCSI Slave Callouts. This patch re-implements LUN Masking feature using SCSI Slave Callouts. With the new design in the slave_alloc entry point; for each new LUN discovered we check with our internal LUN Masking config whether to expose or to mask this particular LUN. We return -ENXIO (No such device or address) from slave_alloc for the LUNs we don't want to be exposed. We also notify the SCSI mid-layer to do a sequential LUN scan rather than REPORT_LUNS based scan if LUN masking is enabled on our HBA port, since a -ENXIO from any LUN in REPORT_LUNS based scan translates to a scan abort. This patch also handles the dynamic lun masking config change from enable to disable or vice-versa by resetting sdev_bflags of LUN 0 appropriately. Signed-off-by: Krishna Gudipati Signed-off-by: James Bottomley --- drivers/scsi/bfa/bfad.c | 2 ++ drivers/scsi/bfa/bfad_attr.c | 2 +- drivers/scsi/bfa/bfad_bsg.c | 27 +++++++++++++++-- drivers/scsi/bfa/bfad_drv.h | 2 ++ drivers/scsi/bfa/bfad_im.c | 56 ++++++++++++++++++++++++++++++++++++ drivers/scsi/bfa/bfad_im.h | 27 +++++++++++++++++ 6 files changed, 112 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/bfa/bfad.c b/drivers/scsi/bfa/bfad.c index 66fb72531b34..404fd10ddb21 100644 --- a/drivers/scsi/bfa/bfad.c +++ b/drivers/scsi/bfa/bfad.c @@ -674,6 +674,7 @@ bfad_vport_create(struct bfad_s *bfad, u16 vf_id, spin_lock_irqsave(&bfad->bfad_lock, flags); bfa_fcs_vport_start(&vport->fcs_vport); + list_add_tail(&vport->list_entry, &bfad->vport_list); spin_unlock_irqrestore(&bfad->bfad_lock, flags); return BFA_STATUS_OK; @@ -1404,6 +1405,7 @@ bfad_pci_probe(struct pci_dev *pdev, const struct pci_device_id *pid) bfad->ref_count = 0; bfad->pport.bfad = bfad; INIT_LIST_HEAD(&bfad->pbc_vport_list); + INIT_LIST_HEAD(&bfad->vport_list); /* Setup the debugfs node for this bfad */ if (bfa_debugfs_enable) diff --git a/drivers/scsi/bfa/bfad_attr.c b/drivers/scsi/bfa/bfad_attr.c index 9d95844ab463..1938fe0473e9 100644 --- a/drivers/scsi/bfa/bfad_attr.c +++ b/drivers/scsi/bfa/bfad_attr.c @@ -491,7 +491,7 @@ bfad_im_vport_delete(struct fc_vport *fc_vport) free_scsi_host: bfad_scsi_host_free(bfad, im_port); - + list_del(&vport->list_entry); kfree(vport); return 0; diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c index 06fc00caeb41..530de2b1200a 100644 --- a/drivers/scsi/bfa/bfad_bsg.c +++ b/drivers/scsi/bfa/bfad_bsg.c @@ -2394,6 +2394,21 @@ out: return 0; } +/* Function to reset the LUN SCAN mode */ +static void +bfad_iocmd_lunmask_reset_lunscan_mode(struct bfad_s *bfad, int lunmask_cfg) +{ + struct bfad_im_port_s *pport_im = bfad->pport.im_port; + struct bfad_vport_s *vport = NULL; + + /* Set the scsi device LUN SCAN flags for base port */ + bfad_reset_sdev_bflags(pport_im, lunmask_cfg); + + /* Set the scsi device LUN SCAN flags for the vports */ + list_for_each_entry(vport, &bfad->vport_list, list_entry) + bfad_reset_sdev_bflags(vport->drv_port.im_port, lunmask_cfg); +} + int bfad_iocmd_lunmask(struct bfad_s *bfad, void *pcmd, unsigned int v_cmd) { @@ -2401,11 +2416,17 @@ bfad_iocmd_lunmask(struct bfad_s *bfad, void *pcmd, unsigned int v_cmd) unsigned long flags; spin_lock_irqsave(&bfad->bfad_lock, flags); - if (v_cmd == IOCMD_FCPIM_LUNMASK_ENABLE) + if (v_cmd == IOCMD_FCPIM_LUNMASK_ENABLE) { iocmd->status = bfa_fcpim_lunmask_update(&bfad->bfa, BFA_TRUE); - else if (v_cmd == IOCMD_FCPIM_LUNMASK_DISABLE) + /* Set the LUN Scanning mode to be Sequential scan */ + if (iocmd->status == BFA_STATUS_OK) + bfad_iocmd_lunmask_reset_lunscan_mode(bfad, BFA_TRUE); + } else if (v_cmd == IOCMD_FCPIM_LUNMASK_DISABLE) { iocmd->status = bfa_fcpim_lunmask_update(&bfad->bfa, BFA_FALSE); - else if (v_cmd == IOCMD_FCPIM_LUNMASK_CLEAR) + /* Set the LUN Scanning mode to default REPORT_LUNS scan */ + if (iocmd->status == BFA_STATUS_OK) + bfad_iocmd_lunmask_reset_lunscan_mode(bfad, BFA_FALSE); + } else if (v_cmd == IOCMD_FCPIM_LUNMASK_CLEAR) iocmd->status = bfa_fcpim_lunmask_clear(&bfad->bfa); spin_unlock_irqrestore(&bfad->bfad_lock, flags); return 0; diff --git a/drivers/scsi/bfa/bfad_drv.h b/drivers/scsi/bfa/bfad_drv.h index 5e19a5f820ec..dc5b9d99c450 100644 --- a/drivers/scsi/bfa/bfad_drv.h +++ b/drivers/scsi/bfa/bfad_drv.h @@ -43,6 +43,7 @@ #include #include #include +#include #include "bfa_modules.h" #include "bfa_fcs.h" @@ -227,6 +228,7 @@ struct bfad_s { struct list_head active_aen_q; struct bfa_aen_entry_s aen_list[BFA_AEN_MAX_ENTRY]; spinlock_t bfad_aen_spinlock; + struct list_head vport_list; }; /* BFAD state machine events */ diff --git a/drivers/scsi/bfa/bfad_im.c b/drivers/scsi/bfa/bfad_im.c index e5db649e8eb7..3153923f5b60 100644 --- a/drivers/scsi/bfa/bfad_im.c +++ b/drivers/scsi/bfa/bfad_im.c @@ -917,6 +917,37 @@ bfad_get_itnim(struct bfad_im_port_s *im_port, int id) return NULL; } +/* + * Function is invoked from the SCSI Host Template slave_alloc() entry point. + * Has the logic to query the LUN Mask database to check if this LUN needs to + * be made visible to the SCSI mid-layer or not. + * + * Returns BFA_STATUS_OK if this LUN needs to be added to the OS stack. + * Returns -ENXIO to notify SCSI mid-layer to not add this LUN to the OS stack. + */ +static int +bfad_im_check_if_make_lun_visible(struct scsi_device *sdev, + struct fc_rport *rport) +{ + struct bfad_itnim_data_s *itnim_data = + (struct bfad_itnim_data_s *) rport->dd_data; + struct bfa_s *bfa = itnim_data->itnim->bfa_itnim->bfa; + struct bfa_rport_s *bfa_rport = itnim_data->itnim->bfa_itnim->rport; + struct bfa_lun_mask_s *lun_list = bfa_get_lun_mask_list(bfa); + int i = 0, ret = -ENXIO; + + for (i = 0; i < MAX_LUN_MASK_CFG; i++) { + if (lun_list[i].state == BFA_IOIM_LUN_MASK_ACTIVE && + scsilun_to_int(&lun_list[i].lun) == sdev->lun && + lun_list[i].rp_tag == bfa_rport->rport_tag && + lun_list[i].lp_tag == (u8)bfa_rport->rport_info.lp_tag) { + ret = BFA_STATUS_OK; + break; + } + } + return ret; +} + /* * Scsi_Host template entry slave_alloc */ @@ -924,10 +955,33 @@ static int bfad_im_slave_alloc(struct scsi_device *sdev) { struct fc_rport *rport = starget_to_rport(scsi_target(sdev)); + struct bfad_itnim_data_s *itnim_data = + (struct bfad_itnim_data_s *) rport->dd_data; + struct bfa_s *bfa = itnim_data->itnim->bfa_itnim->bfa; if (!rport || fc_remote_port_chkready(rport)) return -ENXIO; + if (bfa_get_lun_mask_status(bfa) == BFA_LUNMASK_ENABLED) { + /* + * We should not mask LUN 0 - since this will translate + * to no LUN / TARGET for SCSI ml resulting no scan. + */ + if (sdev->lun == 0) { + sdev->sdev_bflags |= BLIST_NOREPORTLUN | + BLIST_SPARSELUN; + goto done; + } + + /* + * Query LUN Mask configuration - to expose this LUN + * to the SCSI mid-layer or to mask it. + */ + if (bfad_im_check_if_make_lun_visible(sdev, rport) != + BFA_STATUS_OK) + return -ENXIO; + } +done: sdev->hostdata = rport->dd_data; return 0; @@ -1037,6 +1091,8 @@ bfad_im_fc_rport_add(struct bfad_im_port_s *im_port, struct bfad_itnim_s *itnim) && (fc_rport->scsi_target_id < MAX_FCP_TARGET)) itnim->scsi_tgt_id = fc_rport->scsi_target_id; + itnim->channel = fc_rport->channel; + return; } diff --git a/drivers/scsi/bfa/bfad_im.h b/drivers/scsi/bfa/bfad_im.h index 004b6cf848d9..0814367ef101 100644 --- a/drivers/scsi/bfa/bfad_im.h +++ b/drivers/scsi/bfa/bfad_im.h @@ -91,6 +91,7 @@ struct bfad_itnim_s { struct fc_rport *fc_rport; struct bfa_itnim_s *bfa_itnim; u16 scsi_tgt_id; + u16 channel; u16 queue_work; unsigned long last_ramp_up_time; unsigned long last_queue_full_time; @@ -166,4 +167,30 @@ irqreturn_t bfad_intx(int irq, void *dev_id); int bfad_im_bsg_request(struct fc_bsg_job *job); int bfad_im_bsg_timeout(struct fc_bsg_job *job); +/* + * Macro to set the SCSI device sdev_bflags - sdev_bflags are used by the + * SCSI mid-layer to choose LUN Scanning mode REPORT_LUNS vs. Sequential Scan + * + * Internally iterate's over all the ITNIM's part of the im_port & set's the + * sdev_bflags for the scsi_device associated with LUN #0. + */ +#define bfad_reset_sdev_bflags(__im_port, __lunmask_cfg) do { \ + struct scsi_device *__sdev = NULL; \ + struct bfad_itnim_s *__itnim = NULL; \ + u32 scan_flags = BLIST_NOREPORTLUN | BLIST_SPARSELUN; \ + list_for_each_entry(__itnim, &((__im_port)->itnim_mapped_list), \ + list_entry) { \ + __sdev = scsi_device_lookup((__im_port)->shost, \ + __itnim->channel, \ + __itnim->scsi_tgt_id, 0); \ + if (__sdev) { \ + if ((__lunmask_cfg) == BFA_TRUE) \ + __sdev->sdev_bflags |= scan_flags; \ + else \ + __sdev->sdev_bflags &= ~scan_flags; \ + scsi_device_put(__sdev); \ + } \ + } \ +} while (0) + #endif From 410f02d813212eef1dedfcfd43460dd11a0ff707 Mon Sep 17 00:00:00 2001 From: "Moger, Babu" Date: Wed, 21 Dec 2011 18:01:37 -0500 Subject: [PATCH 242/820] [SCSI] scsi_dh_alua: Retry the check-condition in case Mode Parameters Changed This patch adds a check-condition in scsi_dh_alua handler for a retry. Sometimes, I have seen attach failing due to this check-condition with following error messages on NetApp E series storage. Dec 7 15:31:01 nilgiris kernel: [102979.696673] scsi 3:0:2:9: alua: port group 00 rel port 01 Dec 7 15:31:01 nilgiris kernel: [102979.697082] scsi 3:0:2:9: alua: rtpg failed with 8000002 Dec 7 15:31:01 nilgiris kernel: [102979.697086] scsi 3:0:2:9: alua: rtpg sense code 06/2a/01 Dec 7 15:31:01 nilgiris kernel: [102979.697088] scsi 3:0:2:9: alua: not attached Signed-off-by: Babu Moger Signed-off-by: James Bottomley --- drivers/scsi/device_handler/scsi_dh_alua.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/scsi/device_handler/scsi_dh_alua.c b/drivers/scsi/device_handler/scsi_dh_alua.c index 4ef021291a4d..04c5cea47a22 100644 --- a/drivers/scsi/device_handler/scsi_dh_alua.c +++ b/drivers/scsi/device_handler/scsi_dh_alua.c @@ -466,6 +466,11 @@ static int alua_check_sense(struct scsi_device *sdev, * Power On, Reset, or Bus Device Reset, just retry. */ return ADD_TO_MLQUEUE; + if (sense_hdr->asc == 0x2a && sense_hdr->ascq == 0x01) + /* + * Mode Parameters Changed + */ + return ADD_TO_MLQUEUE; if (sense_hdr->asc == 0x2a && sense_hdr->ascq == 0x06) /* * ALUA state changed From e51e07e0ac7e3ff847d640f41b7527db04d4a4e7 Mon Sep 17 00:00:00 2001 From: Tkhai Kirill Date: Tue, 10 Jan 2012 13:17:03 +0000 Subject: [PATCH 243/820] sparc32: forced setting of mode of sun4m per-cpu timers SUN4M per-cpu timers have two modes of work. These are timer mode and counter mode. Kernel doesn't write anything to the register, which is connected with mode choice. So, the mode is chosen by bootloader. This patch forces to use timer mode from the kernel and to be independent of bootloader. I had this problem with OpenBIOS. Timers don't tick and kernel fails on QEMU, when it's compiled with SMP support. The patch fixes problem. Signed-off-by: Tkhai Kirill Acked-by: Sam Ravnborg Signed-off-by: David S. Miller --- arch/sparc/kernel/sun4m_irq.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/sparc/kernel/sun4m_irq.c b/arch/sparc/kernel/sun4m_irq.c index 422c16dad1f6..e61165161dd3 100644 --- a/arch/sparc/kernel/sun4m_irq.c +++ b/arch/sparc/kernel/sun4m_irq.c @@ -399,6 +399,9 @@ static void __init sun4m_init_timers(irq_handler_t counter_fn) timers_global = (void __iomem *) (unsigned long) addr[num_cpu_timers]; + /* Every per-cpu timer works in timer mode */ + sbus_writel(0x00000000, &timers_global->timer_config); + sbus_writel((((1000000/HZ) + 1) << 10), &timers_global->l10_limit); master_l10_counter = &timers_global->l10_count; From db804f23a72bada58f083dfad6a65d019ddb3bd4 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Tue, 10 Jan 2012 16:41:01 +0800 Subject: [PATCH 244/820] Btrfs: add pinned extents to on-disk free space cache correctly I got this while running xfstests: [24256.836098] block group 317849600 has an wrong amount of free space [24256.836100] btrfs: failed to load free space cache for block group 317849600 We should clamp the extent returned by find_first_extent_bit(), so the start of the extent won't smaller than the start of the block group. Signed-off-by: Li Zefan --- fs/btrfs/free-space-cache.c | 41 ++++++++++++++++++------------------- 1 file changed, 20 insertions(+), 21 deletions(-) diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index ec23d43d0c35..01840ef95a32 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -838,7 +838,7 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, struct io_ctl io_ctl; struct list_head bitmap_list; struct btrfs_key key; - u64 start, end, len; + u64 start, extent_start, extent_end, len; int entries = 0; int bitmaps = 0; int ret; @@ -857,25 +857,12 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, struct btrfs_free_cluster, block_group_list); - /* - * We shouldn't have switched the pinned extents yet so this is the - * right one - */ - unpin = root->fs_info->pinned_extents; - /* Lock all pages first so we can lock the extent safely. */ io_ctl_prepare_pages(&io_ctl, inode, 0); lock_extent_bits(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, 0, &cached_state, GFP_NOFS); - /* - * When searching for pinned extents, we need to start at our start - * offset. - */ - if (block_group) - start = block_group->key.objectid; - node = rb_first(&ctl->free_space_offset); if (!node && cluster) { node = rb_first(&cluster->root); @@ -918,9 +905,20 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, * We want to add any pinned extents to our free space cache * so we don't leak the space */ + + /* + * We shouldn't have switched the pinned extents yet so this is the + * right one + */ + unpin = root->fs_info->pinned_extents; + + if (block_group) + start = block_group->key.objectid; + while (block_group && (start < block_group->key.objectid + block_group->key.offset)) { - ret = find_first_extent_bit(unpin, start, &start, &end, + ret = find_first_extent_bit(unpin, start, + &extent_start, &extent_end, EXTENT_DIRTY); if (ret) { ret = 0; @@ -928,20 +926,21 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, } /* This pinned extent is out of our range */ - if (start >= block_group->key.objectid + + if (extent_start >= block_group->key.objectid + block_group->key.offset) break; - len = block_group->key.objectid + - block_group->key.offset - start; - len = min(len, end + 1 - start); + extent_start = max(extent_start, start); + extent_end = min(block_group->key.objectid + + block_group->key.offset, extent_end + 1); + len = extent_end - extent_start; entries++; - ret = io_ctl_add_entry(&io_ctl, start, len, NULL); + ret = io_ctl_add_entry(&io_ctl, extent_start, len, NULL); if (ret) goto out_nospc; - start = end + 1; + start = extent_end; } /* Write out the bitmaps */ From a1ee5a45818acc7f9c13e560827cf3e8735ac919 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 9 Jan 2012 14:27:42 +0800 Subject: [PATCH 245/820] Btrfs: avoid possible NULL deref in io_ctl_drop_pages() If we run into some failure path in io_ctl_prepare_pages(), io_ctl->pages[] array may have some NULL pointers. Signed-off-by: Li Zefan --- fs/btrfs/free-space-cache.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 01840ef95a32..4e55af333e19 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -319,9 +319,11 @@ static void io_ctl_drop_pages(struct io_ctl *io_ctl) io_ctl_unmap_page(io_ctl); for (i = 0; i < io_ctl->num_pages; i++) { - ClearPageChecked(io_ctl->pages[i]); - unlock_page(io_ctl->pages[i]); - page_cache_release(io_ctl->pages[i]); + if (io_ctl->pages[i]) { + ClearPageChecked(io_ctl->pages[i]); + unlock_page(io_ctl->pages[i]); + page_cache_release(io_ctl->pages[i]); + } } } From 706efc6630c2722602541a6a2fc5900a4e38456a Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 9 Jan 2012 14:36:28 +0800 Subject: [PATCH 246/820] Btrfs: check the return value of io_ctl_init() It can return -ENOMEM. Signed-off-by: Li Zefan --- fs/btrfs/free-space-cache.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 4e55af333e19..e4eb222147cc 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -637,7 +637,10 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, if (!num_entries) return 0; - io_ctl_init(&io_ctl, inode, root); + ret = io_ctl_init(&io_ctl, inode, root); + if (ret) + return ret; + ret = readahead_cache(inode); if (ret) goto out; @@ -851,7 +854,9 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, if (!i_size_read(inode)) return -1; - io_ctl_init(&io_ctl, inode, root); + ret = io_ctl_init(&io_ctl, inode, root); + if (ret) + return -1; /* Get the cluster for this block_group if it exists */ if (block_group && !list_empty(&block_group->cluster_list)) From f062abf089ff705e09bbaa6fa1e2fd7688a0f2ea Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 29 Dec 2011 13:36:45 +0800 Subject: [PATCH 247/820] Btrfs: remove BUG_ON()s in btrfs_ioctl_setflags() We can recover from errors and return -errno to user space. Signed-off-by: Li Zefan --- fs/btrfs/ioctl.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index c04f02c7d5bb..9619fb03a5d6 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -176,6 +176,8 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) struct btrfs_trans_handle *trans; unsigned int flags, oldflags; int ret; + u64 ip_oldflags; + unsigned int i_oldflags; if (btrfs_root_readonly(root)) return -EROFS; @@ -192,6 +194,9 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) mutex_lock(&inode->i_mutex); + ip_oldflags = ip->flags; + i_oldflags = inode->i_flags; + flags = btrfs_mask_flags(inode->i_mode, flags); oldflags = btrfs_flags_to_ioctl(ip->flags); if ((flags ^ oldflags) & (FS_APPEND_FL | FS_IMMUTABLE_FL)) { @@ -250,18 +255,23 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) } trans = btrfs_join_transaction(root); - BUG_ON(IS_ERR(trans)); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out_drop; + } btrfs_update_iflags(inode); inode->i_ctime = CURRENT_TIME; ret = btrfs_update_inode(trans, root, inode); - BUG_ON(ret); btrfs_end_transaction(trans, root); + out_drop: + if (ret) { + ip->flags = ip_oldflags; + inode->i_flags = i_oldflags; + } mnt_drop_write(file->f_path.mnt); - - ret = 0; out_unlock: mutex_unlock(&inode->i_mutex); return ret; From 4da6f1a332f6c16b6594c7892f13c31459b9b1c8 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 29 Dec 2011 13:39:50 +0800 Subject: [PATCH 248/820] Btrfs: reserve metadata space in btrfs_ioctl_setflags() Check and reserve space for btrfs_update_inode(). Signed-off-by: Li Zefan --- fs/btrfs/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 9619fb03a5d6..fe8a60c865eb 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -254,7 +254,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS); } - trans = btrfs_join_transaction(root); + trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto out_drop; From 125ccb0ae6806dbec31abf4a85448971df3b4e39 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 8 Dec 2011 15:07:24 +0800 Subject: [PATCH 249/820] Btrfs: don't pass a trans handle unnecessarily in volumes.c Some functions never use the transaction handle passed to them. Signed-off-by: Li Zefan --- fs/btrfs/extent-tree.c | 2 +- fs/btrfs/volumes.c | 18 +++++++----------- fs/btrfs/volumes.h | 3 +-- 3 files changed, 9 insertions(+), 14 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 8603ee4e3dfd..5b53479ce07b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -7084,7 +7084,7 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr) * space to fit our block group in. */ if (device->total_bytes > device->bytes_used + min_free) { - ret = find_free_dev_extent(NULL, device, min_free, + ret = find_free_dev_extent(device, min_free, &dev_offset, NULL); if (!ret) dev_nr++; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f4b839fd3c9d..73f673c8d8d8 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -829,7 +829,6 @@ out: /* * find_free_dev_extent - find free space in the specified device - * @trans: transaction handler * @device: the device which we search the free space in * @num_bytes: the size of the free space that we need * @start: store the start of the free space. @@ -848,8 +847,7 @@ out: * But if we don't find suitable free space, it is used to store the size of * the max free space. */ -int find_free_dev_extent(struct btrfs_trans_handle *trans, - struct btrfs_device *device, u64 num_bytes, +int find_free_dev_extent(struct btrfs_device *device, u64 num_bytes, u64 *start, u64 *len) { struct btrfs_key key; @@ -893,7 +891,7 @@ int find_free_dev_extent(struct btrfs_trans_handle *trans, key.offset = search_start; key.type = BTRFS_DEV_EXTENT_KEY; - ret = btrfs_search_slot(trans, root, &key, path, 0, 0); + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto out; if (ret > 0) { @@ -1469,8 +1467,7 @@ error_undo: /* * does all the dirty work required for changing file system's UUID. */ -static int btrfs_prepare_sprout(struct btrfs_trans_handle *trans, - struct btrfs_root *root) +static int btrfs_prepare_sprout(struct btrfs_root *root) { struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices; struct btrfs_fs_devices *old_devices; @@ -1695,7 +1692,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) if (seeding_dev) { sb->s_flags &= ~MS_RDONLY; - ret = btrfs_prepare_sprout(trans, root); + ret = btrfs_prepare_sprout(root); BUG_ON(ret); } @@ -2323,8 +2320,7 @@ done: return ret; } -static int btrfs_add_system_chunk(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +static int btrfs_add_system_chunk(struct btrfs_root *root, struct btrfs_key *key, struct btrfs_chunk *chunk, int item_size) { @@ -2496,7 +2492,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, if (total_avail == 0) continue; - ret = find_free_dev_extent(trans, device, + ret = find_free_dev_extent(device, max_stripe_size * dev_stripes, &dev_offset, &max_avail); if (ret && ret != -ENOSPC) @@ -2687,7 +2683,7 @@ static int __finish_chunk_alloc(struct btrfs_trans_handle *trans, BUG_ON(ret); if (map->type & BTRFS_BLOCK_GROUP_SYSTEM) { - ret = btrfs_add_system_chunk(trans, chunk_root, &key, chunk, + ret = btrfs_add_system_chunk(chunk_root, &key, chunk, item_size); BUG_ON(ret); } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 78f2d4d4f37f..c1701ec9d49f 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -230,7 +230,6 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_root *dev_root); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); -int find_free_dev_extent(struct btrfs_trans_handle *trans, - struct btrfs_device *device, u64 num_bytes, +int find_free_dev_extent(struct btrfs_device *device, u64 num_bytes, u64 *start, u64 *max_avail); #endif From de11cc12df17337979e0929d2831887432f236ca Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 1 Dec 2011 12:55:47 +0800 Subject: [PATCH 250/820] Btrfs: don't pre-allocate btrfs bio We pre-allocate a btrfs bio with fixed size, and then may re-allocate memory if we find stripes are bigger than the fixed size. But this pre-allocation is not necessary. Also we don't have to calcuate the stripe number twice. Signed-off-by: Li Zefan --- fs/btrfs/volumes.c | 67 +++++++++++++++------------------------------- 1 file changed, 21 insertions(+), 46 deletions(-) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 73f673c8d8d8..540fdd25fb5e 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2897,26 +2897,13 @@ static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, u64 stripe_nr; u64 stripe_nr_orig; u64 stripe_nr_end; - int stripes_allocated = 8; - int stripes_required = 1; int stripe_index; int i; + int ret = 0; int num_stripes; int max_errors = 0; struct btrfs_bio *bbio = NULL; - if (bbio_ret && !(rw & (REQ_WRITE | REQ_DISCARD))) - stripes_allocated = 1; -again: - if (bbio_ret) { - bbio = kzalloc(btrfs_bio_size(stripes_allocated), - GFP_NOFS); - if (!bbio) - return -ENOMEM; - - atomic_set(&bbio->error, 0); - } - read_lock(&em_tree->lock); em = lookup_extent_mapping(em_tree, logical, *length); read_unlock(&em_tree->lock); @@ -2935,32 +2922,6 @@ again: if (mirror_num > map->num_stripes) mirror_num = 0; - /* if our btrfs_bio struct is too small, back off and try again */ - if (rw & REQ_WRITE) { - if (map->type & (BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_DUP)) { - stripes_required = map->num_stripes; - max_errors = 1; - } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) { - stripes_required = map->sub_stripes; - max_errors = 1; - } - } - if (rw & REQ_DISCARD) { - if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | - BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_DUP | - BTRFS_BLOCK_GROUP_RAID10)) { - stripes_required = map->num_stripes; - } - } - if (bbio_ret && (rw & (REQ_WRITE | REQ_DISCARD)) && - stripes_allocated < stripes_required) { - stripes_allocated = map->num_stripes; - free_extent_map(em); - kfree(bbio); - goto again; - } stripe_nr = offset; /* * stripe_nr counts the total number of stripes we have to stride @@ -3055,6 +3016,13 @@ again: } BUG_ON(stripe_index >= map->num_stripes); + bbio = kzalloc(btrfs_bio_size(num_stripes), GFP_NOFS); + if (!bbio) { + ret = -ENOMEM; + goto out; + } + atomic_set(&bbio->error, 0); + if (rw & REQ_DISCARD) { for (i = 0; i < num_stripes; i++) { bbio->stripes[i].physical = @@ -3151,15 +3119,22 @@ again: stripe_index++; } } - if (bbio_ret) { - *bbio_ret = bbio; - bbio->num_stripes = num_stripes; - bbio->max_errors = max_errors; - bbio->mirror_num = mirror_num; + + if (rw & REQ_WRITE) { + if (map->type & (BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID10 | + BTRFS_BLOCK_GROUP_DUP)) { + max_errors = 1; + } } + + *bbio_ret = bbio; + bbio->num_stripes = num_stripes; + bbio->max_errors = max_errors; + bbio->mirror_num = mirror_num; out: free_extent_map(em); - return 0; + return ret; } int btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, From ec9ef7a13be4dcce964c8503e8999087945e5b9e Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 1 Dec 2011 14:06:42 +0800 Subject: [PATCH 251/820] Btrfs: simplfy calculation of stripe length for discard operation For btrfs raid, while discarding a range of space, we'll need to know the start offset and length to discard for each device, and it's done in btrfs_map_block(). However the calculation is a bit complex for raid0 and raid10, so I reimplement it based on a fact that: dev1 dev2 dev3 (raid0) ----------------------------------- s0 s3 s6 s1 s4 s7 s2 s5 Each device has (total_stripes / nr_dev) stripes, or plus one. Signed-off-by: Li Zefan --- fs/btrfs/volumes.c | 97 +++++++++++++++------------------------------- 1 file changed, 32 insertions(+), 65 deletions(-) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 540fdd25fb5e..563ef650850e 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3024,80 +3024,47 @@ static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, atomic_set(&bbio->error, 0); if (rw & REQ_DISCARD) { + int factor = 0; + int sub_stripes = 0; + u64 stripes_per_dev = 0; + u32 remaining_stripes = 0; + + if (map->type & + (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID10)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID0) + sub_stripes = 1; + else + sub_stripes = map->sub_stripes; + + factor = map->num_stripes / sub_stripes; + stripes_per_dev = div_u64_rem(stripe_nr_end - + stripe_nr_orig, + factor, + &remaining_stripes); + } + for (i = 0; i < num_stripes; i++) { bbio->stripes[i].physical = map->stripes[stripe_index].physical + stripe_offset + stripe_nr * map->stripe_len; bbio->stripes[i].dev = map->stripes[stripe_index].dev; - if (map->type & BTRFS_BLOCK_GROUP_RAID0) { - u64 stripes; - u32 last_stripe = 0; - int j; - - div_u64_rem(stripe_nr_end - 1, - map->num_stripes, - &last_stripe); - - for (j = 0; j < map->num_stripes; j++) { - u32 test; - - div_u64_rem(stripe_nr_end - 1 - j, - map->num_stripes, &test); - if (test == stripe_index) - break; - } - stripes = stripe_nr_end - 1 - j; - do_div(stripes, map->num_stripes); - bbio->stripes[i].length = map->stripe_len * - (stripes - stripe_nr + 1); - - if (i == 0) { + if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID10)) { + bbio->stripes[i].length = stripes_per_dev * + map->stripe_len; + if (i / sub_stripes < remaining_stripes) + bbio->stripes[i].length += + map->stripe_len; + if (i < sub_stripes) bbio->stripes[i].length -= stripe_offset; + if ((i / sub_stripes + 1) % + sub_stripes == remaining_stripes) + bbio->stripes[i].length -= + stripe_end_offset; + if (i == sub_stripes - 1) stripe_offset = 0; - } - if (stripe_index == last_stripe) - bbio->stripes[i].length -= - stripe_end_offset; - } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) { - u64 stripes; - int j; - int factor = map->num_stripes / - map->sub_stripes; - u32 last_stripe = 0; - - div_u64_rem(stripe_nr_end - 1, - factor, &last_stripe); - last_stripe *= map->sub_stripes; - - for (j = 0; j < factor; j++) { - u32 test; - - div_u64_rem(stripe_nr_end - 1 - j, - factor, &test); - - if (test == - stripe_index / map->sub_stripes) - break; - } - stripes = stripe_nr_end - 1 - j; - do_div(stripes, factor); - bbio->stripes[i].length = map->stripe_len * - (stripes - stripe_nr + 1); - - if (i < map->sub_stripes) { - bbio->stripes[i].length -= - stripe_offset; - if (i == map->sub_stripes - 1) - stripe_offset = 0; - } - if (stripe_index >= last_stripe && - stripe_index <= (last_stripe + - map->sub_stripes - 1)) { - bbio->stripes[i].length -= - stripe_end_offset; - } } else bbio->stripes[i].length = *length; From 7fe1e641502616220437079258506196bc4d8cbf Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 29 Dec 2011 14:47:27 +0800 Subject: [PATCH 252/820] Btrfs: rewrite btrfs_trim_block_group() There are various bugs in block group trimming: - It may trim from offset smaller than user-specified offset. - It may trim beyond user-specified range. - It may leak free space for extents smaller than specified minlen. - It may truncate the last trimmed extent thus leak free space. - With mixed extents+bitmaps, some extents may not be trimmed. - With mixed extents+bitmaps, some bitmaps may not be trimmed (even none will be trimmed). Even for those trimmed, not all the free space in the bitmaps will be trimmed. I rewrite btrfs_trim_block_group() and break it into two functions. One is to trim extents only, and the other is to trim bitmaps only. Before patching: # fstrim -v /mnt/ /mnt/: 1496465408 bytes were trimmed After patching: # fstrim -v /mnt/ /mnt/: 2193768448 bytes were trimmed And this matches the total free space: # btrfs fi df /mnt Data: total=3.58GB, used=1.79GB System, DUP: total=8.00MB, used=4.00KB System: total=4.00MB, used=0.00 Metadata, DUP: total=205.12MB, used=97.14MB Metadata: total=8.00MB, used=0.00 Signed-off-by: Li Zefan --- fs/btrfs/free-space-cache.c | 237 +++++++++++++++++++++++++----------- 1 file changed, 165 insertions(+), 72 deletions(-) diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index e4eb222147cc..b3cbb8939fa3 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -2594,17 +2594,57 @@ void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster) cluster->block_group = NULL; } -int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, - u64 *trimmed, u64 start, u64 end, u64 minlen) +static int do_trimming(struct btrfs_block_group_cache *block_group, + u64 *total_trimmed, u64 start, u64 bytes, + u64 reserved_start, u64 reserved_bytes) +{ + struct btrfs_space_info *space_info = block_group->space_info; + struct btrfs_fs_info *fs_info = block_group->fs_info; + int ret; + int update = 0; + u64 trimmed = 0; + + spin_lock(&space_info->lock); + spin_lock(&block_group->lock); + if (!block_group->ro) { + block_group->reserved += reserved_bytes; + space_info->bytes_reserved += reserved_bytes; + update = 1; + } + spin_unlock(&block_group->lock); + spin_unlock(&space_info->lock); + + ret = btrfs_error_discard_extent(fs_info->extent_root, + start, bytes, &trimmed); + if (!ret) + *total_trimmed += trimmed; + + btrfs_add_free_space(block_group, reserved_start, reserved_bytes); + + if (update) { + spin_lock(&space_info->lock); + spin_lock(&block_group->lock); + if (block_group->ro) + space_info->bytes_readonly += reserved_bytes; + block_group->reserved -= reserved_bytes; + space_info->bytes_reserved -= reserved_bytes; + spin_unlock(&space_info->lock); + spin_unlock(&block_group->lock); + } + + return ret; +} + +static int trim_no_bitmap(struct btrfs_block_group_cache *block_group, + u64 *total_trimmed, u64 start, u64 end, u64 minlen) { struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; - struct btrfs_free_space *entry = NULL; - struct btrfs_fs_info *fs_info = block_group->fs_info; - u64 bytes = 0; - u64 actually_trimmed; + struct btrfs_free_space *entry; + struct rb_node *node; int ret = 0; - - *trimmed = 0; + u64 extent_start; + u64 extent_bytes; + u64 bytes; while (start < end) { spin_lock(&ctl->tree_lock); @@ -2615,81 +2655,47 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, } entry = tree_search_offset(ctl, start, 0, 1); - if (!entry) - entry = tree_search_offset(ctl, - offset_to_bitmap(ctl, start), - 1, 1); - - if (!entry || entry->offset >= end) { + if (!entry) { spin_unlock(&ctl->tree_lock); break; } - if (entry->bitmap) { - ret = search_bitmap(ctl, entry, &start, &bytes); - if (!ret) { - if (start >= end) { - spin_unlock(&ctl->tree_lock); - break; - } - bytes = min(bytes, end - start); - bitmap_clear_bits(ctl, entry, start, bytes); - if (entry->bytes == 0) - free_bitmap(ctl, entry); - } else { - start = entry->offset + BITS_PER_BITMAP * - block_group->sectorsize; + /* skip bitmaps */ + while (entry->bitmap) { + node = rb_next(&entry->offset_index); + if (!node) { spin_unlock(&ctl->tree_lock); - ret = 0; - continue; + goto out; } - } else { - start = entry->offset; - bytes = min(entry->bytes, end - start); - unlink_free_space(ctl, entry); - kmem_cache_free(btrfs_free_space_cachep, entry); + entry = rb_entry(node, struct btrfs_free_space, + offset_index); } + if (entry->offset >= end) { + spin_unlock(&ctl->tree_lock); + break; + } + + extent_start = entry->offset; + extent_bytes = entry->bytes; + start = max(start, extent_start); + bytes = min(extent_start + extent_bytes, end) - start; + if (bytes < minlen) { + spin_unlock(&ctl->tree_lock); + goto next; + } + + unlink_free_space(ctl, entry); + kmem_cache_free(btrfs_free_space_cachep, entry); + spin_unlock(&ctl->tree_lock); - if (bytes >= minlen) { - struct btrfs_space_info *space_info; - int update = 0; - - space_info = block_group->space_info; - spin_lock(&space_info->lock); - spin_lock(&block_group->lock); - if (!block_group->ro) { - block_group->reserved += bytes; - space_info->bytes_reserved += bytes; - update = 1; - } - spin_unlock(&block_group->lock); - spin_unlock(&space_info->lock); - - ret = btrfs_error_discard_extent(fs_info->extent_root, - start, - bytes, - &actually_trimmed); - - btrfs_add_free_space(block_group, start, bytes); - if (update) { - spin_lock(&space_info->lock); - spin_lock(&block_group->lock); - if (block_group->ro) - space_info->bytes_readonly += bytes; - block_group->reserved -= bytes; - space_info->bytes_reserved -= bytes; - spin_unlock(&space_info->lock); - spin_unlock(&block_group->lock); - } - - if (ret) - break; - *trimmed += actually_trimmed; - } + ret = do_trimming(block_group, total_trimmed, start, bytes, + extent_start, extent_bytes); + if (ret) + break; +next: start += bytes; - bytes = 0; if (fatal_signal_pending(current)) { ret = -ERESTARTSYS; @@ -2698,6 +2704,93 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, cond_resched(); } +out: + return ret; +} + +static int trim_bitmaps(struct btrfs_block_group_cache *block_group, + u64 *total_trimmed, u64 start, u64 end, u64 minlen) +{ + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; + struct btrfs_free_space *entry; + int ret = 0; + int ret2; + u64 bytes; + u64 offset = offset_to_bitmap(ctl, start); + + while (offset < end) { + bool next_bitmap = false; + + spin_lock(&ctl->tree_lock); + + if (ctl->free_space < minlen) { + spin_unlock(&ctl->tree_lock); + break; + } + + entry = tree_search_offset(ctl, offset, 1, 0); + if (!entry) { + spin_unlock(&ctl->tree_lock); + next_bitmap = true; + goto next; + } + + bytes = minlen; + ret2 = search_bitmap(ctl, entry, &start, &bytes); + if (ret2 || start >= end) { + spin_unlock(&ctl->tree_lock); + next_bitmap = true; + goto next; + } + + bytes = min(bytes, end - start); + if (bytes < minlen) { + spin_unlock(&ctl->tree_lock); + goto next; + } + + bitmap_clear_bits(ctl, entry, start, bytes); + if (entry->bytes == 0) + free_bitmap(ctl, entry); + + spin_unlock(&ctl->tree_lock); + + ret = do_trimming(block_group, total_trimmed, start, bytes, + start, bytes); + if (ret) + break; +next: + if (next_bitmap) { + offset += BITS_PER_BITMAP * ctl->unit; + } else { + start += bytes; + if (start >= offset + BITS_PER_BITMAP * ctl->unit) + offset += BITS_PER_BITMAP * ctl->unit; + } + + if (fatal_signal_pending(current)) { + ret = -ERESTARTSYS; + break; + } + + cond_resched(); + } + + return ret; +} + +int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, + u64 *trimmed, u64 start, u64 end, u64 minlen) +{ + int ret; + + *trimmed = 0; + + ret = trim_no_bitmap(block_group, trimmed, start, end, minlen); + if (ret) + return ret; + + ret = trim_bitmaps(block_group, trimmed, start, end, minlen); return ret; } From c7c144db531fda414e532adac56e965ce332e2a5 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 7 Dec 2011 10:39:22 +0800 Subject: [PATCH 253/820] Btrfs: update global block_rsv when creating a new block group A bug was triggered while using seed device: # mkfs.btrfs /dev/loop1 # btrfstune -S 1 /dev/loop1 # mount -o /dev/loop1 /mnt # btrfs dev add /dev/loop2 /mnt btrfs: block rsv returned -28 ------------[ cut here ]------------ WARNING: at fs/btrfs/extent-tree.c:5969 btrfs_alloc_free_block+0x166/0x396 [btrfs]() ... Call Trace: ... [] btrfs_cow_block+0x101/0x147 [btrfs] [] btrfs_search_slot+0x1b8/0x55f [btrfs] [] btrfs_insert_empty_items+0x42/0x7f [btrfs] [] btrfs_insert_item+0x40/0x7e [btrfs] [] btrfs_make_block_group+0x243/0x2aa [btrfs] [] __btrfs_alloc_chunk+0x672/0x70e [btrfs] [] init_first_rw_device+0x77/0x13c [btrfs] [] btrfs_init_new_device+0x664/0x9fd [btrfs] [] btrfs_ioctl+0x694/0xdbe [btrfs] [] do_vfs_ioctl+0x496/0x4cc [] sys_ioctl+0x33/0x4f [] sysenter_do_call+0x12/0x38 ---[ end trace 906adac595facc7d ]--- Since seed device is readonly, there's no usable space in the filesystem. Afterwards we add a sprout device to it, and the kernel creates a METADATA block group and a SYSTEM block group where comes free space we can reserve, but we still get revervation failure because the global block_rsv hasn't been updated accordingly. Signed-off-by: Li Zefan --- fs/btrfs/extent-tree.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 5b53479ce07b..bf30f670cda9 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -7446,6 +7446,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, ret = update_space_info(root->fs_info, cache->flags, size, bytes_used, &cache->space_info); BUG_ON(ret); + update_global_block_rsv(root->fs_info); spin_lock(&cache->space_info->lock); cache->space_info->bytes_readonly += cache->bytes_super; From b367e47fb3a70f5d24ebd6faf7d42436d485fb2d Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 7 Dec 2011 11:38:24 +0800 Subject: [PATCH 254/820] Btrfs: fix possible deadlock when opening a seed device The correct lock order is uuid_mutex -> volume_mutex -> chunk_mutex, but when we mount a filesystem which has backing seed devices, we have this lock chain: open_ctree() lock(chunk_mutex); read_chunk_tree(); read_one_dev(); open_seed_devices(); lock(uuid_mutex); and then we hit a lockdep splat. Signed-off-by: Li Zefan --- fs/btrfs/disk-io.c | 2 -- fs/btrfs/volumes.c | 9 +++++++-- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 3f9d5551e582..858ab347413e 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2270,9 +2270,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, (unsigned long)btrfs_header_chunk_tree_uuid(chunk_root->node), BTRFS_UUID_SIZE); - mutex_lock(&fs_info->chunk_mutex); ret = btrfs_read_chunk_tree(chunk_root); - mutex_unlock(&fs_info->chunk_mutex); if (ret) { printk(KERN_WARNING "btrfs: failed to read chunk tree on %s\n", sb->s_id); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 563ef650850e..fbb493b28d5a 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3506,7 +3506,7 @@ static int open_seed_devices(struct btrfs_root *root, u8 *fsid) struct btrfs_fs_devices *fs_devices; int ret; - mutex_lock(&uuid_mutex); + BUG_ON(!mutex_is_locked(&uuid_mutex)); fs_devices = root->fs_info->fs_devices->seed; while (fs_devices) { @@ -3544,7 +3544,6 @@ static int open_seed_devices(struct btrfs_root *root, u8 *fsid) fs_devices->seed = root->fs_info->fs_devices->seed; root->fs_info->fs_devices->seed = fs_devices; out: - mutex_unlock(&uuid_mutex); return ret; } @@ -3687,6 +3686,9 @@ int btrfs_read_chunk_tree(struct btrfs_root *root) if (!path) return -ENOMEM; + mutex_lock(&uuid_mutex); + lock_chunks(root); + /* first we search for all of the device items, and then we * read in all of the chunk items. This way we can create chunk * mappings that reference all of the devices that are afound @@ -3737,6 +3739,9 @@ again: } ret = 0; error: + unlock_chunks(root); + mutex_unlock(&uuid_mutex); + btrfs_free_path(path); return ret; } From d761f0c521868e59cd0bc59159cbdb4686fe210d Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Tue, 10 Jan 2012 12:15:58 +0100 Subject: [PATCH 255/820] microblaze: Add topology init Init topology subsystem by cpu registration. Microblaze Linux kernel is fauling by "Oops: kernel access of bad area, sig: 11" because cpu is not initialized. Signed-off-by: Michal Simek --- arch/microblaze/kernel/setup.c | 21 ++++++++++++++++++++- 1 file changed, 20 insertions(+), 1 deletion(-) diff --git a/arch/microblaze/kernel/setup.c b/arch/microblaze/kernel/setup.c index 604cd9dd1333..d4fc1a971779 100644 --- a/arch/microblaze/kernel/setup.c +++ b/arch/microblaze/kernel/setup.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include @@ -226,5 +227,23 @@ static int __init setup_bus_notifier(void) return 0; } - arch_initcall(setup_bus_notifier); + +static DEFINE_PER_CPU(struct cpu, cpu_devices); + +static int __init topology_init(void) +{ + int i, ret; + + for_each_present_cpu(i) { + struct cpu *c = &per_cpu(cpu_devices, i); + + ret = register_cpu(c, i); + if (ret) + printk(KERN_WARNING "topology_init: register_cpu %d " + "failed (%d)\n", i, ret); + } + + return 0; +} +subsys_initcall(topology_init); From 353b67d8ced4dc53281c88150ad295e24bc4b4c5 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sat, 26 Nov 2011 00:35:39 +0100 Subject: [PATCH 256/820] jbd: Issue cache flush after checkpointing When we reach cleanup_journal_tail(), there is no guarantee that checkpointed buffers are on a stable storage - especially if buffers were written out by log_do_checkpoint(), they are likely to be only in disk's caches. Thus when we update journal superblock, effectively removing old transaction from journal, this write of superblock can get to stable storage before those checkpointed buffers which can result in filesystem corruption after a crash. A similar problem can happen if we replay the journal and wipe it before flushing disk's caches. Thus we must unconditionally issue a cache flush before we update journal superblock in these cases. The fix is slightly complicated by the fact that we have to get log tail before we issue cache flush but we can store it in the journal superblock only after the cache flush. Otherwise we risk races where new tail is written before appropriate cache flush is finished. I managed to reproduce the corruption using somewhat tweaked Chris Mason's barrier-test scheduler. Also this should fix occasional reports of 'Bit already freed' filesystem errors which are totally unreproducible but inspection of several fs images I've gathered over time points to a problem like this. CC: stable@kernel.org Signed-off-by: Jan Kara --- fs/jbd/checkpoint.c | 27 ++++++++++++++++++++++----- fs/jbd/recovery.c | 4 ++++ 2 files changed, 26 insertions(+), 5 deletions(-) diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c index 5d1a00a5041b..05f0754f2b46 100644 --- a/fs/jbd/checkpoint.c +++ b/fs/jbd/checkpoint.c @@ -453,8 +453,6 @@ out: * * Return <0 on error, 0 on success, 1 if there was nothing to clean up. * - * Called with the journal lock held. - * * This is the only part of the journaling code which really needs to be * aware of transaction aborts. Checkpointing involves writing to the * main filesystem area rather than to the journal, so it can proceed @@ -472,13 +470,14 @@ int cleanup_journal_tail(journal_t *journal) if (is_journal_aborted(journal)) return 1; - /* OK, work out the oldest transaction remaining in the log, and + /* + * OK, work out the oldest transaction remaining in the log, and * the log block it starts at. * * If the log is now empty, we need to work out which is the * next transaction ID we will write, and where it will - * start. */ - + * start. + */ spin_lock(&journal->j_state_lock); spin_lock(&journal->j_list_lock); transaction = journal->j_checkpoint_transactions; @@ -504,7 +503,25 @@ int cleanup_journal_tail(journal_t *journal) spin_unlock(&journal->j_state_lock); return 1; } + spin_unlock(&journal->j_state_lock); + /* + * We need to make sure that any blocks that were recently written out + * --- perhaps by log_do_checkpoint() --- are flushed out before we + * drop the transactions from the journal. It's unlikely this will be + * necessary, especially with an appropriately sized journal, but we + * need this to guarantee correctness. Fortunately + * cleanup_journal_tail() doesn't get called all that often. + */ + if (journal->j_flags & JFS_BARRIER) + blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL); + + spin_lock(&journal->j_state_lock); + if (!tid_gt(first_tid, journal->j_tail_sequence)) { + spin_unlock(&journal->j_state_lock); + /* Someone else cleaned up journal so return 0 */ + return 0; + } /* OK, update the superblock to recover the freed space. * Physical blocks come first: have we wrapped beyond the end of * the log? */ diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c index 5b43e96788e6..008bf062fd26 100644 --- a/fs/jbd/recovery.c +++ b/fs/jbd/recovery.c @@ -20,6 +20,7 @@ #include #include #include +#include #endif /* @@ -263,6 +264,9 @@ int journal_recover(journal_t *journal) err2 = sync_blockdev(journal->j_fs_dev); if (!err) err = err2; + /* Flush disk caches to get replayed data on the permanent storage */ + if (journal->j_flags & JFS_BARRIER) + blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL); return err; } From 34b07840565004cfa444e165e88bf77a5cbcdb25 Mon Sep 17 00:00:00 2001 From: Djalal Harouni Date: Mon, 9 Jan 2012 15:58:37 +0100 Subject: [PATCH 257/820] ext2: protect inode changes in the SETVERSION and SETFLAGS ioctls Unlock mutex after i_flags and i_ctime updates in the EXT2_IOC_SETFLAGS ioctl. Use i_mutex in the EXT2_IOC_SETVERSION ioctl to protect i_ctime and i_generation updates and make the ioctl consistent since i_mutex is also used in other places to protect timestamps and inode changes. Cc: Andreas Dilger Cc: Jan Kara Signed-off-by: Djalal Harouni Signed-off-by: Jan Kara --- fs/ext2/ioctl.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/fs/ext2/ioctl.c b/fs/ext2/ioctl.c index 1089f760c847..2de655f5d625 100644 --- a/fs/ext2/ioctl.c +++ b/fs/ext2/ioctl.c @@ -77,10 +77,11 @@ long ext2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) flags = flags & EXT2_FL_USER_MODIFIABLE; flags |= oldflags & ~EXT2_FL_USER_MODIFIABLE; ei->i_flags = flags; - mutex_unlock(&inode->i_mutex); ext2_set_inode_flags(inode); inode->i_ctime = CURRENT_TIME_SEC; + mutex_unlock(&inode->i_mutex); + mark_inode_dirty(inode); setflags_out: mnt_drop_write_file(filp); @@ -88,20 +89,29 @@ setflags_out: } case EXT2_IOC_GETVERSION: return put_user(inode->i_generation, (int __user *) arg); - case EXT2_IOC_SETVERSION: + case EXT2_IOC_SETVERSION: { + __u32 generation; + if (!inode_owner_or_capable(inode)) return -EPERM; ret = mnt_want_write_file(filp); if (ret) return ret; - if (get_user(inode->i_generation, (int __user *) arg)) { + if (get_user(generation, (int __user *) arg)) { ret = -EFAULT; - } else { - inode->i_ctime = CURRENT_TIME_SEC; - mark_inode_dirty(inode); + goto setversion_out; } + + mutex_lock(&inode->i_mutex); + inode->i_ctime = CURRENT_TIME_SEC; + inode->i_generation = generation; + mutex_unlock(&inode->i_mutex); + + mark_inode_dirty(inode); +setversion_out: mnt_drop_write_file(filp); return ret; + } case EXT2_IOC_GETRSVSZ: if (test_opt(inode->i_sb, RESERVATION) && S_ISREG(inode->i_mode) From df3481399042200792822b6243e36a95a557b57e Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Wed, 11 Jan 2012 07:29:56 -0700 Subject: [PATCH 258/820] NVMe: Set number of queues correctly The number of submission & completion queues should be set by calling Set Features, not Get Features. Reported-by: Kwok Kong Signed-off-by: Matthew Wilcox --- drivers/block/nvme.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index 28c84b18712d..f4996b0e4b1a 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -840,7 +840,7 @@ static int nvme_identify(struct nvme_dev *dev, unsigned nsid, unsigned cns, } static int nvme_get_features(struct nvme_dev *dev, unsigned fid, - unsigned dword11, dma_addr_t dma_addr, u32 *result) + unsigned dword11, dma_addr_t dma_addr) { struct nvme_command c; @@ -850,6 +850,20 @@ static int nvme_get_features(struct nvme_dev *dev, unsigned fid, c.features.fid = cpu_to_le32(fid); c.features.dword11 = cpu_to_le32(dword11); + return nvme_submit_admin_cmd(dev, &c, NULL); +} + +static int nvme_set_features(struct nvme_dev *dev, unsigned fid, + unsigned dword11, dma_addr_t dma_addr, u32 *result) +{ + struct nvme_command c; + + memset(&c, 0, sizeof(c)); + c.features.opcode = nvme_admin_set_features; + c.features.prp1 = cpu_to_le64(dma_addr); + c.features.fid = cpu_to_le32(fid); + c.features.dword11 = cpu_to_le32(dword11); + return nvme_submit_admin_cmd(dev, &c, result); } @@ -1365,7 +1379,7 @@ static int set_queue_count(struct nvme_dev *dev, int count) u32 result; u32 q_count = (count - 1) | ((count - 1) << 16); - status = nvme_get_features(dev, NVME_FEAT_NUM_QUEUES, q_count, 0, + status = nvme_set_features(dev, NVME_FEAT_NUM_QUEUES, q_count, 0, &result); if (status) return -EIO; @@ -1482,7 +1496,7 @@ static int __devinit nvme_dev_add(struct nvme_dev *dev) continue; res = nvme_get_features(dev, NVME_FEAT_LBA_RANGE, i, - dma_addr + 4096, NULL); + dma_addr + 4096); if (res) continue; From 8ca0686aabbf145014fa22bc7fd06189b564f119 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 26 Dec 2011 20:28:23 +0900 Subject: [PATCH 259/820] ARM: SAMSUNG: Guard against multiple inclusion of plat/dma.h Otherwise it'll generate errors if included twice. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/plat-samsung/include/plat/dma.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/arch/arm/plat-samsung/include/plat/dma.h b/arch/arm/plat-samsung/include/plat/dma.h index b9061128abde..7b02143ccd9a 100644 --- a/arch/arm/plat-samsung/include/plat/dma.h +++ b/arch/arm/plat-samsung/include/plat/dma.h @@ -10,6 +10,9 @@ * published by the Free Software Foundation. */ +#ifndef __PLAT_DMA_H +#define __PLAT_DMA_H + #include enum s3c2410_dma_buffresult { @@ -122,5 +125,6 @@ extern int s3c2410_dma_getposition(enum dma_ch channel, extern int s3c2410_dma_set_opfn(enum dma_ch, s3c2410_dma_opfn_t rtn); extern int s3c2410_dma_set_buffdone_fn(enum dma_ch, s3c2410_dma_cbfn_t rtn); - #include + +#endif From 6fd9dd38a1124db66d2438bf9b225aef09597478 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 26 Dec 2011 20:28:57 +0900 Subject: [PATCH 260/820] ARM: SAMSUNG: dma-ops.h needs mach/dma.h dma-ops.h uses samsung_dma_is_dmadev so it needs to have mach/dma.h included. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/plat-samsung/include/plat/dma-ops.h | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm/plat-samsung/include/plat/dma-ops.h b/arch/arm/plat-samsung/include/plat/dma-ops.h index 22eafc310bd7..7c6c2a8dd552 100644 --- a/arch/arm/plat-samsung/include/plat/dma-ops.h +++ b/arch/arm/plat-samsung/include/plat/dma-ops.h @@ -14,6 +14,7 @@ #define __SAMSUNG_DMA_OPS_H_ __FILE__ #include +#include struct samsung_dma_prep_info { enum dma_transaction_type cap; From 5b0b34ea755810462346b24e24643eb93ebc919c Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 29 Dec 2011 18:01:08 +0900 Subject: [PATCH 261/820] ARM: SAMSUNG: Declare struct platform_device in plat/s3c64xx-spi.h The function declarations in plat/s3c64xx-spi.h rely on struct platform_device but it's not declared by the header causing compiler warnings if it is included prior to another header which does that. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/plat-samsung/include/plat/s3c64xx-spi.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h b/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h index aea68b60ef98..fa95e9a00972 100644 --- a/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h +++ b/arch/arm/plat-samsung/include/plat/s3c64xx-spi.h @@ -11,6 +11,8 @@ #ifndef __S3C64XX_PLAT_SPI_H #define __S3C64XX_PLAT_SPI_H +struct platform_device; + /** * struct s3c64xx_spi_csinfo - ChipSelect description * @fb_delay: Slave specific feedback delay. From d5b9cdbcf74c192907e8290b85c3dcb3117984fb Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Fri, 30 Dec 2011 09:38:13 +0900 Subject: [PATCH 262/820] ARM: S3C64XX: Remove unconditional power domain disables Now that we have software runtime power management support this code is redundant. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/mach-s3c64xx/pm.c | 13 ------------- 1 file changed, 13 deletions(-) diff --git a/arch/arm/mach-s3c64xx/pm.c b/arch/arm/mach-s3c64xx/pm.c index 055dac90e0e2..7d3e81b9dd06 100644 --- a/arch/arm/mach-s3c64xx/pm.c +++ b/arch/arm/mach-s3c64xx/pm.c @@ -346,23 +346,10 @@ int __init s3c64xx_pm_init(void) static __init int s3c64xx_pm_initcall(void) { - u32 val; - pm_cpu_prep = s3c64xx_pm_prepare; pm_cpu_sleep = s3c64xx_cpu_suspend; pm_uart_udivslot = 1; - /* - * Unconditionally disable power domains that contain only - * blocks which have no mainline driver support. - */ - val = __raw_readl(S3C64XX_NORMAL_CFG); - val &= ~(S3C64XX_NORMALCFG_DOMAIN_G_ON | - S3C64XX_NORMALCFG_DOMAIN_V_ON | - S3C64XX_NORMALCFG_DOMAIN_I_ON | - S3C64XX_NORMALCFG_DOMAIN_P_ON); - __raw_writel(val, S3C64XX_NORMAL_CFG); - #ifdef CONFIG_S3C_PM_DEBUG_LED_SMDK gpio_request(S3C64XX_GPN(12), "DEBUG_LED0"); gpio_request(S3C64XX_GPN(13), "DEBUG_LED1"); From d7d3077a2843faa6877280a4fac8cc732f5e7dce Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 29 Dec 2011 17:09:23 +0900 Subject: [PATCH 263/820] ARM: S3C64XX: Remove hsmmc1 from Cragganmore The second MMC slot was only present on revision 1 of Cragganmore so remove it. Only MMC0 and MMC2 are there on production systems. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/mach-s3c64xx/mach-crag6410.c | 8 -------- 1 file changed, 8 deletions(-) diff --git a/arch/arm/mach-s3c64xx/mach-crag6410.c b/arch/arm/mach-s3c64xx/mach-crag6410.c index 680fd758ff2d..2e34fe861944 100644 --- a/arch/arm/mach-s3c64xx/mach-crag6410.c +++ b/arch/arm/mach-s3c64xx/mach-crag6410.c @@ -328,7 +328,6 @@ static struct platform_device wallvdd_device = { static struct platform_device *crag6410_devices[] __initdata = { &s3c_device_hsmmc0, - &s3c_device_hsmmc1, &s3c_device_hsmmc2, &s3c_device_i2c0, &s3c_device_i2c1, @@ -683,12 +682,6 @@ static struct s3c_sdhci_platdata crag6410_hsmmc2_pdata = { .cd_type = S3C_SDHCI_CD_PERMANENT, }; -static struct s3c_sdhci_platdata crag6410_hsmmc1_pdata = { - .max_width = 4, - .cd_type = S3C_SDHCI_CD_GPIO, - .ext_cd_gpio = S3C64XX_GPF(11), -}; - static void crag6410_cfg_sdhci0(struct platform_device *dev, int width) { /* Set all the necessary GPG pins to special-function 2 */ @@ -723,7 +716,6 @@ static void __init crag6410_machine_init(void) gpio_direction_output(S3C64XX_GPF(10), 1); s3c_sdhci0_set_platdata(&crag6410_hsmmc0_pdata); - s3c_sdhci1_set_platdata(&crag6410_hsmmc1_pdata); s3c_sdhci2_set_platdata(&crag6410_hsmmc2_pdata); s3c_i2c0_set_platdata(&i2c0_pdata); From 91b60b1d5a56e5971237f9c88de2faf168a5ceaf Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 29 Dec 2011 18:02:39 +0900 Subject: [PATCH 264/820] ARM: S3C64XX: Fix the memory mapped GPIOs on Cragganmore Rather than letting them get allocated dynamically where we don't know where they are, and also name the data line resource as gpio-generic requires that. Without these changes the GPIOs are useless. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/mach-s3c64xx/include/mach/crag6410.h | 1 + arch/arm/mach-s3c64xx/mach-crag6410.c | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/arm/mach-s3c64xx/include/mach/crag6410.h b/arch/arm/mach-s3c64xx/include/mach/crag6410.h index 5d55ab018b6b..4cb2f951f1e9 100644 --- a/arch/arm/mach-s3c64xx/include/mach/crag6410.h +++ b/arch/arm/mach-s3c64xx/include/mach/crag6410.h @@ -21,5 +21,6 @@ #define CODEC_GPIO_BASE (GPIO_BOARD_START + 8) #define GLENFARCLAS_PMIC_GPIO_BASE (GPIO_BOARD_START + 32) #define BANFF_PMIC_GPIO_BASE (GPIO_BOARD_START + 64) +#define MMGPIO_GPIO_BASE (GPIO_BOARD_START + 96) #endif diff --git a/arch/arm/mach-s3c64xx/mach-crag6410.c b/arch/arm/mach-s3c64xx/mach-crag6410.c index 2e34fe861944..6998b5b54141 100644 --- a/arch/arm/mach-s3c64xx/mach-crag6410.c +++ b/arch/arm/mach-s3c64xx/mach-crag6410.c @@ -260,6 +260,7 @@ static struct platform_device crag6410_dm9k_device = { static struct resource crag6410_mmgpio_resource[] = { [0] = { + .name = "dat", .start = S3C64XX_PA_XM0CSN4 + 1, .end = S3C64XX_PA_XM0CSN4 + 1, .flags = IORESOURCE_MEM, @@ -272,7 +273,7 @@ static struct platform_device crag6410_mmgpio = { .resource = crag6410_mmgpio_resource, .num_resources = ARRAY_SIZE(crag6410_mmgpio_resource), .dev.platform_data = &(struct bgpio_pdata) { - .base = -1, + .base = MMGPIO_GPIO_BASE, }, }; From 6e11e0bd82ecacd13b839558a49f65ebc958e1bd Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Fri, 30 Dec 2011 09:46:40 +0900 Subject: [PATCH 265/820] ARM: S3C64XX: Fix interrupt configuration for PCA935x on Cragganmore 0 is a valid IRQ but the interrupt line for the pca935x certainly isn't connected there which causes it to fail to probe. Signed-off-by: Mark Brown Signed-off-by: Kukjin Kim --- arch/arm/mach-s3c64xx/mach-crag6410.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mach-s3c64xx/mach-crag6410.c b/arch/arm/mach-s3c64xx/mach-crag6410.c index 6998b5b54141..3d8b20bcd21a 100644 --- a/arch/arm/mach-s3c64xx/mach-crag6410.c +++ b/arch/arm/mach-s3c64xx/mach-crag6410.c @@ -355,7 +355,7 @@ static struct platform_device *crag6410_devices[] __initdata = { static struct pca953x_platform_data crag6410_pca_data = { .gpio_base = PCA935X_GPIO_BASE, - .irq_base = 0, + .irq_base = -1, }; /* VDDARM is controlled by DVS1 connected to GPK(0) */ From 815ed6fc0a4d82bb39ed43d230c4e516214987e7 Mon Sep 17 00:00:00 2001 From: Tushar Behera Date: Thu, 29 Dec 2011 16:48:08 +0900 Subject: [PATCH 266/820] ARM: EXYNOS: Invert VCLK polarity for framebuffer on ORIGEN Framebuffer driver needs to fetch the video data during the rising edge of the VCLK. Otherwise, there are some glitches in the LCD display. Signed-off-by: Tushar Behera Signed-off-by: Kukjin Kim --- arch/arm/mach-exynos/mach-origen.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm/mach-exynos/mach-origen.c b/arch/arm/mach-exynos/mach-origen.c index 2b11e046d391..0679b8ad2d1e 100644 --- a/arch/arm/mach-exynos/mach-origen.c +++ b/arch/arm/mach-exynos/mach-origen.c @@ -597,7 +597,8 @@ static struct s3c_fb_pd_win origen_fb_win0 = { static struct s3c_fb_platdata origen_lcd_pdata __initdata = { .win[0] = &origen_fb_win0, .vidcon0 = VIDCON0_VIDOUT_RGB | VIDCON0_PNRMODE_RGB, - .vidcon1 = VIDCON1_INV_HSYNC | VIDCON1_INV_VSYNC, + .vidcon1 = VIDCON1_INV_HSYNC | VIDCON1_INV_VSYNC | + VIDCON1_INV_VCLK, .setup_gpio = exynos4_fimd0_gpio_setup_24bpp, }; From 46fe44ce8777f087aa8ad4a2605fdcfb9c2d63af Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 16 Nov 2011 15:03:59 +0100 Subject: [PATCH 267/820] quota: Pass information that quota is stored in system file to userspace Quota tools need to know whether quota is stored in a system file or in classical aquota.{user|group} files. So pass this information as a flag in GETINFO quotactl. Signed-off-by: Jan Kara --- fs/quota/dquot.c | 8 +++++--- include/linux/quota.h | 6 +++++- 2 files changed, 10 insertions(+), 4 deletions(-) diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 5ec59b20cf76..46741970371b 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2125,6 +2125,8 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id, mutex_unlock(&dqopt->dqio_mutex); goto out_file_init; } + if (dqopt->flags & DQUOT_QUOTA_SYS_FILE) + dqopt->info[type].dqi_flags |= DQF_SYS_FILE; mutex_unlock(&dqopt->dqio_mutex); spin_lock(&dq_state_lock); dqopt->flags |= dquot_state_flag(flags, type); @@ -2464,7 +2466,7 @@ int dquot_get_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii) spin_lock(&dq_data_lock); ii->dqi_bgrace = mi->dqi_bgrace; ii->dqi_igrace = mi->dqi_igrace; - ii->dqi_flags = mi->dqi_flags & DQF_MASK; + ii->dqi_flags = mi->dqi_flags & DQF_GETINFO_MASK; ii->dqi_valid = IIF_ALL; spin_unlock(&dq_data_lock); mutex_unlock(&sb_dqopt(sb)->dqonoff_mutex); @@ -2490,8 +2492,8 @@ int dquot_set_dqinfo(struct super_block *sb, int type, struct if_dqinfo *ii) if (ii->dqi_valid & IIF_IGRACE) mi->dqi_igrace = ii->dqi_igrace; if (ii->dqi_valid & IIF_FLAGS) - mi->dqi_flags = (mi->dqi_flags & ~DQF_MASK) | - (ii->dqi_flags & DQF_MASK); + mi->dqi_flags = (mi->dqi_flags & ~DQF_SETINFO_MASK) | + (ii->dqi_flags & DQF_SETINFO_MASK); spin_unlock(&dq_data_lock); mark_info_dirty(sb, type); /* Force write to disk */ diff --git a/include/linux/quota.h b/include/linux/quota.h index cb7855699037..c09fa042b5ea 100644 --- a/include/linux/quota.h +++ b/include/linux/quota.h @@ -230,7 +230,11 @@ struct mem_dqinfo { struct super_block; #define DQF_MASK 0xffff /* Mask for format specific flags */ -#define DQF_INFO_DIRTY_B 16 +#define DQF_GETINFO_MASK 0x1ffff /* Mask for flags passed to userspace */ +#define DQF_SETINFO_MASK 0xffff /* Mask for flags modifiable from userspace */ +#define DQF_SYS_FILE_B 16 +#define DQF_SYS_FILE (1 << DQF_SYS_FILE_B) /* Quota file stored as system file */ +#define DQF_INFO_DIRTY_B 31 #define DQF_INFO_DIRTY (1 << DQF_INFO_DIRTY_B) /* Is info dirty? */ extern void mark_info_dirty(struct super_block *sb, int type); From 8ea11f7f11c811d1f502c2dabc5259d447c2f2a0 Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Thu, 12 Jan 2012 11:35:50 -0500 Subject: [PATCH 268/820] xen/balloon: Move the registration from device to subsystem. With git commit 070680218379e15c1901f4bf21b98e3cbf12b527 "xen-balloon: convert sysdev_class to a regular subsystem" we would end up with the attributes being put in: /sys/devices/xen_memory0/target_kb instead of /sys/devices/system/xen_memory/xen_memory0/target_kb Making the tools inable to deflate the kernel to make more space for launching another guest and printing: Error: Failed to query current memory allocation of dom0 Reported-by: Sander Eikelenboom Suggested-by: Kay Sievers Signed-off-by: Konrad Rzeszutek Wilk --- drivers/xen/xen-balloon.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/xen/xen-balloon.c b/drivers/xen/xen-balloon.c index 3832e303c33a..596e6a7b17d6 100644 --- a/drivers/xen/xen-balloon.c +++ b/drivers/xen/xen-balloon.c @@ -221,7 +221,7 @@ static int register_balloon(struct device *dev) { int i, error; - error = bus_register(&balloon_subsys); + error = subsys_system_register(&balloon_subsys, NULL); if (error) return error; From 7250170c9ed00f3b74b11b98afefab45020672dd Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 11 Jan 2012 10:46:27 +0300 Subject: [PATCH 269/820] cifs: integer overflow in parse_dacl() On 32 bit systems num_aces * sizeof(struct cifs_ace *) could overflow leading to a smaller ppace buffer than we expected. Signed-off-by: Dan Carpenter Acked-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsacl.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index 72ddf23ef6f7..c1b254487388 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -909,6 +909,8 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, umode_t group_mask = S_IRWXG; umode_t other_mask = S_IRWXU | S_IRWXG | S_IRWXO; + if (num_aces > ULONG_MAX / sizeof(struct cifs_ace *)) + return; ppace = kmalloc(num_aces * sizeof(struct cifs_ace *), GFP_KERNEL); if (!ppace) { From 220d9f259536e22afef30347b7b42148640282a7 Mon Sep 17 00:00:00 2001 From: Javier Martin Date: Thu, 12 Jan 2012 12:30:38 +0100 Subject: [PATCH 270/820] ASoC: Fix DMA channel leak in imx-pcm-dma-mx2 driver. "snd_imx_pcm_hw_params" callback can be called several times by the user (i.e. OSS emulation) leading to a DMA channel leak. Signed-off-by: Javier Martin Signed-off-by: Mark Brown --- sound/soc/imx/imx-pcm-dma-mx2.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/imx/imx-pcm-dma-mx2.c b/sound/soc/imx/imx-pcm-dma-mx2.c index 1cf2fe889f6a..7d7fbb5d2a91 100644 --- a/sound/soc/imx/imx-pcm-dma-mx2.c +++ b/sound/soc/imx/imx-pcm-dma-mx2.c @@ -88,11 +88,13 @@ static int imx_ssi_dma_alloc(struct snd_pcm_substream *substream, iprtd->dma_data.dma_request = dma_params->dma; /* Try to grab a DMA channel */ - dma_cap_zero(mask); - dma_cap_set(DMA_SLAVE, mask); - iprtd->dma_chan = dma_request_channel(mask, filter, iprtd); - if (!iprtd->dma_chan) - return -EINVAL; + if (!iprtd->dma_chan) { + dma_cap_zero(mask); + dma_cap_set(DMA_SLAVE, mask); + iprtd->dma_chan = dma_request_channel(mask, filter, iprtd); + if (!iprtd->dma_chan) + return -EINVAL; + } switch (params_format(params)) { case SNDRV_PCM_FORMAT_S16_LE: From b01de4fb40137fbda7530550ff0cd37171dafb0c Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Thu, 12 Jan 2012 16:31:14 +0100 Subject: [PATCH 271/820] ALSA: HDA: Use LPIB position fix for Macbook Pro 7,1 Several users have reported "choppy" audio under the 3.2 kernel, and that changing position_fix to 1 has resolved their problem. The chip is an nVidia Corporation MCP89 High Definition Audio, [10de:0d94] (rev a2). Cc: stable@kernel.org (v3.2+) BugLink: https://bugs.launchpad.net/bugs/909419 Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_intel.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index 0852e204a4c8..fb35474c1203 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -2498,6 +2498,7 @@ static struct snd_pci_quirk position_fix_list[] __devinitdata = { SND_PCI_QUIRK(0x1043, 0x81b3, "ASUS", POS_FIX_LPIB), SND_PCI_QUIRK(0x1043, 0x81e7, "ASUS M2V", POS_FIX_LPIB), SND_PCI_QUIRK(0x104d, 0x9069, "Sony VPCS11V9E", POS_FIX_LPIB), + SND_PCI_QUIRK(0x10de, 0xcb89, "Macbook Pro 7,1", POS_FIX_LPIB), SND_PCI_QUIRK(0x1297, 0x3166, "Shuttle", POS_FIX_LPIB), SND_PCI_QUIRK(0x1458, 0xa022, "ga-ma770-ud3", POS_FIX_LPIB), SND_PCI_QUIRK(0x1462, 0x1002, "MSI Wind U115", POS_FIX_LPIB), From 5c845beb4220ba7cddb099bd47c72be7e6919712 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 13 Jan 2012 09:53:53 +0100 Subject: [PATCH 272/820] ALSA: Don't prompt for CONFIG_SND_COMPRESS_OFFLOAD CONFIG_SND_COMPRESS_OFFLOAD is an item to be selected by the dirver just like CONFIG_SND_PCM, and no need to prompt for explicit selection. Signed-off-by: Takashi Iwai --- sound/core/Kconfig | 13 +++---------- 1 file changed, 3 insertions(+), 10 deletions(-) diff --git a/sound/core/Kconfig b/sound/core/Kconfig index ca6cdf6f9968..349756978ed3 100644 --- a/sound/core/Kconfig +++ b/sound/core/Kconfig @@ -13,6 +13,9 @@ config SND_HWDEP config SND_RAWMIDI tristate +config SND_COMPRESS_OFFLOAD + tristate + # To be effective this also requires INPUT - users should say: # select SND_JACK if INPUT=y || INPUT=SND # to avoid having to force INPUT on. @@ -155,16 +158,6 @@ config SND_DYNAMIC_MINORS If you are unsure about this, say N here. -config SND_COMPRESS_OFFLOAD - tristate "ALSA Compressed audio offload support" - default n - help - If you want support for offloading compressed audio and have such - a hardware, then you should say Y here and also to the DSP driver - of your platform. - - If you are unsure about this, say N here. - config SND_SUPPORT_OLD_API bool "Support old ALSA API" default y From 252c3d84ed398b090ac2dace46fc6faa6cfaea99 Mon Sep 17 00:00:00 2001 From: "RongQing.Li" Date: Thu, 12 Jan 2012 22:33:46 +0000 Subject: [PATCH 273/820] ipv6: release idev when ip6_neigh_lookup failed in icmp6_dst_alloc release idev when ip6_neigh_lookup failed in icmp6_dst_alloc Signed-off-by: RongQing.Li Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/route.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 07361dfa8085..8c2e3ab58f2a 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1091,6 +1091,7 @@ struct dst_entry *icmp6_dst_alloc(struct net_device *dev, else { neigh = ip6_neigh_lookup(&rt->dst, &fl6->daddr); if (IS_ERR(neigh)) { + in6_dev_put(idev); dst_free(&rt->dst); return ERR_CAST(neigh); } From 673e8e597c06eb81954bf21a10f5cce74a1de8f1 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:04 +0000 Subject: [PATCH 274/820] xfs: remove xfs_itruncate_data This wrapper isn't overly useful, not to say rather confusing. Around the call to xfs_itruncate_extents it does: - add tracing - add a few asserts in debug builds - conditionally update the inode size in two places - log the inode Both the tracing and the inode logging can be moved to xfs_itruncate_extents as they are useful for the attribute fork as well - in fact the attr code already does an equivalent xfs_trans_log_inode call just after calling xfs_itruncate_extents. The conditional size updates are a mess, and there was no reason to do them in two places anyway, as the first one was conditional on the inode having extents - but without extents we xfs_itruncate_extents would be a no-op and the placement wouldn't matter anyway. Instead move the size assignments and the asserts that make sense to the callers that want it. As a side effect of this clean up xfs_setattr_size by introducing variables for the old and new inode size, and moving the size updates into a common place. Reviewed-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Ben Myers --- fs/xfs/xfs_attr.c | 4 -- fs/xfs/xfs_inode.c | 124 ++++----------------------------------- fs/xfs/xfs_inode.h | 2 - fs/xfs/xfs_iops.c | 47 +++++++++------ fs/xfs/xfs_qm_syscalls.c | 9 ++- fs/xfs/xfs_trace.h | 4 +- fs/xfs/xfs_vnodeops.c | 17 +++++- 7 files changed, 65 insertions(+), 142 deletions(-) diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index 1e5d97f86ea8..08b9ac644c31 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -827,10 +827,6 @@ xfs_attr_inactive(xfs_inode_t *dp) if (error) goto out; - /* - * Commit the last in the sequence of transactions. - */ - xfs_trans_log_inode(trans, dp, XFS_ILOG_CORE); error = xfs_trans_commit(trans, XFS_TRANS_RELEASE_LOG_RES); xfs_iunlock(dp, XFS_ILOCK_EXCL); diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 9dda7cc32848..ccd619a993f6 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -1165,52 +1165,6 @@ xfs_ialloc( return 0; } -/* - * Check to make sure that there are no blocks allocated to the - * file beyond the size of the file. We don't check this for - * files with fixed size extents or real time extents, but we - * at least do it for regular files. - */ -#ifdef DEBUG -STATIC void -xfs_isize_check( - struct xfs_inode *ip, - xfs_fsize_t isize) -{ - struct xfs_mount *mp = ip->i_mount; - xfs_fileoff_t map_first; - int nimaps; - xfs_bmbt_irec_t imaps[2]; - int error; - - if (!S_ISREG(ip->i_d.di_mode)) - return; - - if (XFS_IS_REALTIME_INODE(ip)) - return; - - if (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) - return; - - nimaps = 2; - map_first = XFS_B_TO_FSB(mp, (xfs_ufsize_t)isize); - /* - * The filesystem could be shutting down, so bmapi may return - * an error. - */ - error = xfs_bmapi_read(ip, map_first, - (XFS_B_TO_FSB(mp, - (xfs_ufsize_t)XFS_MAXIOFFSET(mp)) - map_first), - imaps, &nimaps, XFS_BMAPI_ENTIRE); - if (error) - return; - ASSERT(nimaps == 1); - ASSERT(imaps[0].br_startblock == HOLESTARTBLOCK); -} -#else /* DEBUG */ -#define xfs_isize_check(ip, isize) -#endif /* DEBUG */ - /* * Free up the underlying blocks past new_size. The new size must be smaller * than the current size. This routine can be used both for the attribute and @@ -1258,6 +1212,8 @@ xfs_itruncate_extents( ASSERT(ip->i_itemp->ili_lock_flags == 0); ASSERT(!XFS_NOT_DQATTACHED(mp, ip)); + trace_xfs_itruncate_extents_start(ip, new_size); + /* * Since it is possible for space to become allocated beyond * the end of the file (in a crash where the space is allocated @@ -1325,6 +1281,14 @@ xfs_itruncate_extents( goto out; } + /* + * Always re-log the inode so that our permanent transaction can keep + * on rolling it forward in the log. + */ + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + + trace_xfs_itruncate_extents_end(ip, new_size); + out: *tpp = tp; return error; @@ -1338,74 +1302,6 @@ out_bmap_cancel: goto out; } -int -xfs_itruncate_data( - struct xfs_trans **tpp, - struct xfs_inode *ip, - xfs_fsize_t new_size) -{ - int error; - - trace_xfs_itruncate_data_start(ip, new_size); - - /* - * The first thing we do is set the size to new_size permanently on - * disk. This way we don't have to worry about anyone ever being able - * to look at the data being freed even in the face of a crash. - * What we're getting around here is the case where we free a block, it - * is allocated to another file, it is written to, and then we crash. - * If the new data gets written to the file but the log buffers - * containing the free and reallocation don't, then we'd end up with - * garbage in the blocks being freed. As long as we make the new_size - * permanent before actually freeing any blocks it doesn't matter if - * they get written to. - */ - if (ip->i_d.di_nextents > 0) { - /* - * If we are not changing the file size then do not update - * the on-disk file size - we may be called from - * xfs_inactive_free_eofblocks(). If we update the on-disk - * file size and then the system crashes before the contents - * of the file are flushed to disk then the files may be - * full of holes (ie NULL files bug). - */ - if (ip->i_size != new_size) { - ip->i_d.di_size = new_size; - ip->i_size = new_size; - xfs_trans_log_inode(*tpp, ip, XFS_ILOG_CORE); - } - } - - error = xfs_itruncate_extents(tpp, ip, XFS_DATA_FORK, new_size); - if (error) - return error; - - /* - * If we are not changing the file size then do not update the on-disk - * file size - we may be called from xfs_inactive_free_eofblocks(). - * If we update the on-disk file size and then the system crashes - * before the contents of the file are flushed to disk then the files - * may be full of holes (ie NULL files bug). - */ - xfs_isize_check(ip, new_size); - if (ip->i_size != new_size) { - ip->i_d.di_size = new_size; - ip->i_size = new_size; - } - - ASSERT(new_size != 0 || ip->i_delayed_blks == 0); - ASSERT(new_size != 0 || ip->i_d.di_nextents == 0); - - /* - * Always re-log the inode so that our permanent transaction can keep - * on rolling it forward in the log. - */ - xfs_trans_log_inode(*tpp, ip, XFS_ILOG_CORE); - - trace_xfs_itruncate_data_end(ip, new_size); - return 0; -} - /* * This is called when the inode's link count goes to 0. * We place the on-disk inode on a list in the AGI. It diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index f0e6b151ba37..440f2acebfa8 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -491,8 +491,6 @@ int xfs_ifree(struct xfs_trans *, xfs_inode_t *, struct xfs_bmap_free *); int xfs_itruncate_extents(struct xfs_trans **, struct xfs_inode *, int, xfs_fsize_t); -int xfs_itruncate_data(struct xfs_trans **, struct xfs_inode *, - xfs_fsize_t); int xfs_iunlink(struct xfs_trans *, xfs_inode_t *); void xfs_iext_realloc(xfs_inode_t *, int, int); diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index f9babd179223..f02eaa298d3c 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -750,6 +750,7 @@ xfs_setattr_size( struct xfs_mount *mp = ip->i_mount; struct inode *inode = VFS_I(ip); int mask = iattr->ia_valid; + xfs_off_t oldsize, newsize; struct xfs_trans *tp; int error; uint lock_flags; @@ -777,11 +778,13 @@ xfs_setattr_size( lock_flags |= XFS_IOLOCK_EXCL; xfs_ilock(ip, lock_flags); + oldsize = ip->i_size; + newsize = iattr->ia_size; + /* * Short circuit the truncate case for zero length files. */ - if (iattr->ia_size == 0 && - ip->i_size == 0 && ip->i_d.di_nextents == 0) { + if (newsize == 0 && oldsize == 0 && ip->i_d.di_nextents == 0) { if (!(mask & (ATTR_CTIME|ATTR_MTIME))) goto out_unlock; @@ -807,14 +810,14 @@ xfs_setattr_size( * the inode to the transaction, because the inode cannot be unlocked * once it is a part of the transaction. */ - if (iattr->ia_size > ip->i_size) { + if (newsize > oldsize) { /* * Do the first part of growing a file: zero any data in the * last block that is beyond the old EOF. We need to do this * before the inode is joined to the transaction to modify * i_size. */ - error = xfs_zero_eof(ip, iattr->ia_size, ip->i_size); + error = xfs_zero_eof(ip, newsize, oldsize); if (error) goto out_unlock; } @@ -833,8 +836,8 @@ xfs_setattr_size( * here and prevents waiting for other data not within the range we * care about here. */ - if (ip->i_size != ip->i_d.di_size && iattr->ia_size > ip->i_d.di_size) { - error = xfs_flush_pages(ip, ip->i_d.di_size, iattr->ia_size, 0, + if (oldsize != ip->i_d.di_size && newsize > ip->i_d.di_size) { + error = xfs_flush_pages(ip, ip->i_d.di_size, newsize, 0, FI_NONE); if (error) goto out_unlock; @@ -845,8 +848,7 @@ xfs_setattr_size( */ inode_dio_wait(inode); - error = -block_truncate_page(inode->i_mapping, iattr->ia_size, - xfs_get_blocks); + error = -block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks); if (error) goto out_unlock; @@ -857,7 +859,7 @@ xfs_setattr_size( if (error) goto out_trans_cancel; - truncate_setsize(inode, iattr->ia_size); + truncate_setsize(inode, newsize); commit_flags = XFS_TRANS_RELEASE_LOG_RES; lock_flags |= XFS_ILOCK_EXCL; @@ -876,19 +878,30 @@ xfs_setattr_size( * these flags set. For all other operations the VFS set these flags * explicitly if it wants a timestamp update. */ - if (iattr->ia_size != ip->i_size && - (!(mask & (ATTR_CTIME | ATTR_MTIME)))) { + if (newsize != oldsize && (!(mask & (ATTR_CTIME | ATTR_MTIME)))) { iattr->ia_ctime = iattr->ia_mtime = current_fs_time(inode->i_sb); mask |= ATTR_CTIME | ATTR_MTIME; } - if (iattr->ia_size > ip->i_size) { - ip->i_d.di_size = iattr->ia_size; - ip->i_size = iattr->ia_size; - } else if (iattr->ia_size <= ip->i_size || - (iattr->ia_size == 0 && ip->i_d.di_nextents)) { - error = xfs_itruncate_data(&tp, ip, iattr->ia_size); + /* + * The first thing we do is set the size to new_size permanently on + * disk. This way we don't have to worry about anyone ever being able + * to look at the data being freed even in the face of a crash. + * What we're getting around here is the case where we free a block, it + * is allocated to another file, it is written to, and then we crash. + * If the new data gets written to the file but the log buffers + * containing the free and reallocation don't, then we'd end up with + * garbage in the blocks being freed. As long as we make the new size + * permanent before actually freeing any blocks it doesn't matter if + * they get written to. + */ + ip->i_d.di_size = newsize; + ip->i_size = newsize; + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + + if (newsize <= oldsize) { + error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, newsize); if (error) goto out_trans_abort; diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 5cc3dde1bc90..27378650b5cb 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -31,6 +31,7 @@ #include "xfs_mount.h" #include "xfs_bmap_btree.h" #include "xfs_inode.h" +#include "xfs_inode_item.h" #include "xfs_itable.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" @@ -263,13 +264,19 @@ xfs_qm_scall_trunc_qfile( xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, ip, 0); - error = xfs_itruncate_data(&tp, ip, 0); + ip->i_d.di_size = 0; + ip->i_size = 0; + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + + error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, 0); if (error) { xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); goto out_unlock; } + ASSERT(ip->i_d.di_nextents == 0); + xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index a9d5b1e06efe..297f9fa6fb64 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -1090,8 +1090,8 @@ DECLARE_EVENT_CLASS(xfs_itrunc_class, DEFINE_EVENT(xfs_itrunc_class, name, \ TP_PROTO(struct xfs_inode *ip, xfs_fsize_t new_size), \ TP_ARGS(ip, new_size)) -DEFINE_ITRUNC_EVENT(xfs_itruncate_data_start); -DEFINE_ITRUNC_EVENT(xfs_itruncate_data_end); +DEFINE_ITRUNC_EVENT(xfs_itruncate_extents_start); +DEFINE_ITRUNC_EVENT(xfs_itruncate_extents_end); TRACE_EVENT(xfs_pagecache_inval, TP_PROTO(struct xfs_inode *ip, xfs_off_t start, xfs_off_t finish), diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index f2fea868d4db..96ff03421753 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -226,7 +226,14 @@ xfs_free_eofblocks( xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, ip, 0); - error = xfs_itruncate_data(&tp, ip, ip->i_size); + /* + * Do not update the on-disk file size. If we update the + * on-disk file size and then the system crashes before the + * contents of the file are flushed to disk then the files + * may be full of holes (ie NULL files bug). + */ + error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, + ip->i_size); if (error) { /* * If we get an error at this point we simply don't @@ -670,13 +677,19 @@ xfs_inactive( xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, ip, 0); - error = xfs_itruncate_data(&tp, ip, 0); + ip->i_d.di_size = 0; + ip->i_size = 0; + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + + error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, 0); if (error) { xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); xfs_iunlock(ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL); return VN_INACTIVE_CACHE; } + + ASSERT(ip->i_d.di_nextents == 0); } else if (S_ISLNK(ip->i_d.di_mode)) { /* From bf322d983e540f66517db85b6870017613bb1e8d Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:05 +0000 Subject: [PATCH 275/820] xfs: cleanup xfs_iomap_eof_align_last_fsb Replace the nasty if, else if, elseif condition with more natural C flow that expressed the logic we want here better. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_iomap.c | 36 ++++++++++++++++++------------------ 1 file changed, 18 insertions(+), 18 deletions(-) diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 9afa282aa937..a27a44659da6 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -57,26 +57,26 @@ xfs_iomap_eof_align_last_fsb( xfs_fileoff_t *last_fsb) { xfs_fileoff_t new_last_fsb = 0; - xfs_extlen_t align; + xfs_extlen_t align = 0; int eof, error; - if (XFS_IS_REALTIME_INODE(ip)) - ; - /* - * If mounted with the "-o swalloc" option, roundup the allocation - * request to a stripe width boundary if the file size is >= - * stripe width and we are allocating past the allocation eof. - */ - else if (mp->m_swidth && (mp->m_flags & XFS_MOUNT_SWALLOC) && - (ip->i_size >= XFS_FSB_TO_B(mp, mp->m_swidth))) - new_last_fsb = roundup_64(*last_fsb, mp->m_swidth); - /* - * Roundup the allocation request to a stripe unit (m_dalign) boundary - * if the file size is >= stripe unit size, and we are allocating past - * the allocation eof. - */ - else if (mp->m_dalign && (ip->i_size >= XFS_FSB_TO_B(mp, mp->m_dalign))) - new_last_fsb = roundup_64(*last_fsb, mp->m_dalign); + if (!XFS_IS_REALTIME_INODE(ip)) { + /* + * Round up the allocation request to a stripe unit + * (m_dalign) boundary if the file size is >= stripe unit + * size, and we are allocating past the allocation eof. + * + * If mounted with the "-o swalloc" option the alignment is + * increased from the strip unit size to the stripe width. + */ + if (mp->m_swidth && (mp->m_flags & XFS_MOUNT_SWALLOC)) + align = mp->m_swidth; + else if (mp->m_dalign) + align = mp->m_dalign; + + if (align && ip->i_size >= XFS_FSB_TO_B(mp, align)) + new_last_fsb = roundup_64(*last_fsb, align); + } /* * Always round up the allocation request to an extent boundary From 3d2b3129c2c48cf0153e0f2058cf87e4b45ca3ac Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:06 +0000 Subject: [PATCH 276/820] xfs: remove the unused dm_attrs structure .. and the just as dead bhv_desc forward declaration while we're at it. Signed-off-by: Christoph Hellwig Reviewed-by: Alex Elder Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_inode.h | 7 ------- 1 file changed, 7 deletions(-) diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 440f2acebfa8..c1574721a191 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -211,7 +211,6 @@ typedef struct xfs_icdinode { #ifdef __KERNEL__ -struct bhv_desc; struct xfs_buf; struct xfs_bmap_free; struct xfs_bmbt_irec; @@ -220,12 +219,6 @@ struct xfs_mount; struct xfs_trans; struct xfs_dquot; -typedef struct dm_attrs_s { - __uint32_t da_dmevmask; /* DMIG event mask */ - __uint16_t da_dmstate; /* DMIG state info */ - __uint16_t da_pad; /* DMIG extra padding */ -} dm_attrs_t; - typedef struct xfs_inode { /* Inode linking and identification information. */ struct xfs_mount *i_mount; /* fs mount struct ptr */ From 7c7c7f01cc5e3e423120a4848a73dd5e4586f2f9 Mon Sep 17 00:00:00 2001 From: stephen hemminger Date: Wed, 11 Jan 2012 19:30:38 +0000 Subject: [PATCH 277/820] vhost-net: add module alias (v2.1) By adding some module aliases, programs (or users) won't have to explicitly call modprobe. Vhost-net will always be available if built into the kernel. It does require assigning a permanent minor number for depmod to work. Also: - use C99 style initialization. - add missing entry in documentation for loop-control Signed-off-by: Stephen Hemminger Acked-By: Kay Sievers Signed-off-by: David S. Miller --- Documentation/devices.txt | 3 +++ drivers/vhost/net.c | 8 +++++--- include/linux/miscdevice.h | 1 + 3 files changed, 9 insertions(+), 3 deletions(-) diff --git a/Documentation/devices.txt b/Documentation/devices.txt index cec8864ce4e8..00383186d8fb 100644 --- a/Documentation/devices.txt +++ b/Documentation/devices.txt @@ -447,6 +447,9 @@ Your cooperation is appreciated. 234 = /dev/btrfs-control Btrfs control device 235 = /dev/autofs Autofs control device 236 = /dev/mapper/control Device-Mapper control device + 237 = /dev/loop-control Loopback control device + 238 = /dev/vhost-net Host kernel accelerator for virtio net + 240-254 Reserved for local use 255 Reserved for MISC_DYNAMIC_MINOR diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c index 882a51fe7b3c..9dab1f51dd43 100644 --- a/drivers/vhost/net.c +++ b/drivers/vhost/net.c @@ -856,9 +856,9 @@ static const struct file_operations vhost_net_fops = { }; static struct miscdevice vhost_net_misc = { - MISC_DYNAMIC_MINOR, - "vhost-net", - &vhost_net_fops, + .minor = VHOST_NET_MINOR, + .name = "vhost-net", + .fops = &vhost_net_fops, }; static int vhost_net_init(void) @@ -879,3 +879,5 @@ MODULE_VERSION("0.0.1"); MODULE_LICENSE("GPL v2"); MODULE_AUTHOR("Michael S. Tsirkin"); MODULE_DESCRIPTION("Host kernel accelerator for virtio net"); +MODULE_ALIAS_MISCDEV(VHOST_NET_MINOR); +MODULE_ALIAS("devname:vhost-net"); diff --git a/include/linux/miscdevice.h b/include/linux/miscdevice.h index 32085249e9cb..0549d2115507 100644 --- a/include/linux/miscdevice.h +++ b/include/linux/miscdevice.h @@ -42,6 +42,7 @@ #define AUTOFS_MINOR 235 #define MAPPER_CTRL_MINOR 236 #define LOOP_CTRL_MINOR 237 +#define VHOST_NET_MINOR 238 #define MISC_DYNAMIC_MINOR 255 struct device; From ccde8a45cbd3ea9a6e785e4393d3a1f6b4e6b495 Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Fri, 6 Jan 2012 12:58:16 +0100 Subject: [PATCH 278/820] wireless: b43: fix Oops on card eject during transfer An Oops has once been observed, when the SDIO card had been ejected during IO. The PC value shows, that the dev pointer in b43_op_stop() was NULL. (I moved the NULL check before the lock, based upon a suggestion from Julian Calaby . -- JWL) Signed-off-by: Guennadi Liakhovetski Signed-off-by: John W. Linville --- drivers/net/wireless/b43/main.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index 1c6f19393efa..b91f28ef1032 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -4852,6 +4852,9 @@ static void b43_op_stop(struct ieee80211_hw *hw) cancel_work_sync(&(wl->beacon_update_trigger)); + if (!dev) + goto out; + mutex_lock(&wl->mutex); if (b43_status(dev) >= B43_STAT_STARTED) { dev = b43_wireless_core_stop(dev); @@ -4863,7 +4866,7 @@ static void b43_op_stop(struct ieee80211_hw *hw) out_unlock: mutex_unlock(&wl->mutex); - +out: cancel_work_sync(&(wl->txpower_adjust_work)); } From 92c1ff1fa772ee76a18ae1edd2e4baecf381f17e Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 12 Jan 2012 09:34:50 +0300 Subject: [PATCH 279/820] ipw2x00: signedness bug handling frame length This is basically just a cleanup. Large positive numbers get counted as negative but then get implicitly cast to positive again for the checks that matter. This does make a small difference in ipw_handle_promiscuous_rx() when we test "if (unlikely((len + IPW_RX_FRAME_SIZE) > skb_tailroom(rxb->skb)))" It should return there, but we don't return until a couple lines later when we test "if (len > IPW_RX_BUF_SIZE - sizeof(struct ipw_rt_hdr)) {". The difference is that in the second test the sizeof() means that there is an implied cast to unsigned. Signed-off-by: Dan Carpenter Reviewed-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/ipw2x00/ipw2200.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ipw2x00/ipw2200.c b/drivers/net/wireless/ipw2x00/ipw2200.c index 018a8deb88a8..4fcdac63a300 100644 --- a/drivers/net/wireless/ipw2x00/ipw2200.c +++ b/drivers/net/wireless/ipw2x00/ipw2200.c @@ -7848,7 +7848,7 @@ static void ipw_handle_data_packet_monitor(struct ipw_priv *priv, * more efficiently than we can parse it. ORDER MATTERS HERE */ struct ipw_rt_hdr *ipw_rt; - short len = le16_to_cpu(pkt->u.frame.length); + unsigned short len = le16_to_cpu(pkt->u.frame.length); /* We received data from the HW, so stop the watchdog */ dev->trans_start = jiffies; @@ -8023,7 +8023,7 @@ static void ipw_handle_promiscuous_rx(struct ipw_priv *priv, s8 signal = frame->rssi_dbm - IPW_RSSI_TO_DBM; s8 noise = (s8) le16_to_cpu(frame->noise); u8 rate = frame->rate; - short len = le16_to_cpu(pkt->u.frame.length); + unsigned short len = le16_to_cpu(pkt->u.frame.length); struct sk_buff *skb; int hdr_only = 0; u16 filter = priv->prom_priv->filter; From dfd00c4c8f3dfa1fd7cec45f83d98b2a49743dcd Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Fri, 13 Jan 2012 12:59:32 +0100 Subject: [PATCH 280/820] rt2800pci: fix spurious interrupts generation Same devices can generate interrupt without properly setting bit in INT_SOURCE_CSR register (spurious interrupt), what will cause IRQ line will be disabled by interrupts controller driver. We discovered that clearing INT_MASK_CSR stops such behaviour. We previously first read that register, and then clear all know interrupt sources bits and do not touch reserved bits. After this patch, we write to all register content (I believe writing to reserved bits on that register will not cause any problems, I tested that on my rt2800pci device). This fix very bad performance problem, practically making device unusable (since worked without interrupts), reported in: https://bugzilla.redhat.com/show_bug.cgi?id=658451 We previously tried to workaround that issue in commit 4ba7d9997869d25bd223dea7536fc1ce9fab3b3b "rt2800pci: handle spurious interrupts", but it was reverted in commit 82e5fc2a34fa9ffea38f00c4066b7e600a0ca5e6 as thing, that will prevent to detect real spurious interrupts. Reported-and-tested-by: Amir Hedayaty Cc: stable@vger.kernel.org Signed-off-by: Stanislaw Gruszka Acked-by: Gertjan van Wingerde Signed-off-by: John W. Linville --- drivers/net/wireless/rt2x00/rt2800pci.c | 28 +++++++------------------ 1 file changed, 8 insertions(+), 20 deletions(-) diff --git a/drivers/net/wireless/rt2x00/rt2800pci.c b/drivers/net/wireless/rt2x00/rt2800pci.c index 4941a1a23219..dc88baefa72e 100644 --- a/drivers/net/wireless/rt2x00/rt2800pci.c +++ b/drivers/net/wireless/rt2x00/rt2800pci.c @@ -422,7 +422,6 @@ static int rt2800pci_init_queues(struct rt2x00_dev *rt2x00dev) static void rt2800pci_toggle_irq(struct rt2x00_dev *rt2x00dev, enum dev_state state) { - int mask = (state == STATE_RADIO_IRQ_ON); u32 reg; unsigned long flags; @@ -436,25 +435,14 @@ static void rt2800pci_toggle_irq(struct rt2x00_dev *rt2x00dev, } spin_lock_irqsave(&rt2x00dev->irqmask_lock, flags); - rt2x00pci_register_read(rt2x00dev, INT_MASK_CSR, ®); - rt2x00_set_field32(®, INT_MASK_CSR_RXDELAYINT, 0); - rt2x00_set_field32(®, INT_MASK_CSR_TXDELAYINT, 0); - rt2x00_set_field32(®, INT_MASK_CSR_RX_DONE, mask); - rt2x00_set_field32(®, INT_MASK_CSR_AC0_DMA_DONE, 0); - rt2x00_set_field32(®, INT_MASK_CSR_AC1_DMA_DONE, 0); - rt2x00_set_field32(®, INT_MASK_CSR_AC2_DMA_DONE, 0); - rt2x00_set_field32(®, INT_MASK_CSR_AC3_DMA_DONE, 0); - rt2x00_set_field32(®, INT_MASK_CSR_HCCA_DMA_DONE, 0); - rt2x00_set_field32(®, INT_MASK_CSR_MGMT_DMA_DONE, 0); - rt2x00_set_field32(®, INT_MASK_CSR_MCU_COMMAND, 0); - rt2x00_set_field32(®, INT_MASK_CSR_RXTX_COHERENT, 0); - rt2x00_set_field32(®, INT_MASK_CSR_TBTT, mask); - rt2x00_set_field32(®, INT_MASK_CSR_PRE_TBTT, mask); - rt2x00_set_field32(®, INT_MASK_CSR_TX_FIFO_STATUS, mask); - rt2x00_set_field32(®, INT_MASK_CSR_AUTO_WAKEUP, mask); - rt2x00_set_field32(®, INT_MASK_CSR_GPTIMER, 0); - rt2x00_set_field32(®, INT_MASK_CSR_RX_COHERENT, 0); - rt2x00_set_field32(®, INT_MASK_CSR_TX_COHERENT, 0); + reg = 0; + if (state == STATE_RADIO_IRQ_ON) { + rt2x00_set_field32(®, INT_MASK_CSR_RX_DONE, 1); + rt2x00_set_field32(®, INT_MASK_CSR_TBTT, 1); + rt2x00_set_field32(®, INT_MASK_CSR_PRE_TBTT, 1); + rt2x00_set_field32(®, INT_MASK_CSR_TX_FIFO_STATUS, 1); + rt2x00_set_field32(®, INT_MASK_CSR_AUTO_WAKEUP, 1); + } rt2x00pci_register_write(rt2x00dev, INT_MASK_CSR, reg); spin_unlock_irqrestore(&rt2x00dev->irqmask_lock, flags); From 543d1b92d5500841e70f15e79d197bbffed36561 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 13 Jan 2012 14:17:59 +0100 Subject: [PATCH 281/820] mac80211: fix no-op authorized transitions When userspace attempts to authorize a station that is already authorized, nothing happens as you'd expect. Similarly, when it unauthorizes a station that is associated, nothing happens. However, when it unauthorizes a station that isn't even associated yet, we erroneously try to move the station to associated. This seems to happen occasionally as a result of a race when wpa_supplicant attempts to unauthorize the port in managed mode. Particularly with my new patches to keep stations, it can then move a station into ASSOCIATED state before we have really associated, which is really confusing. I introduced this bug in "mac80211: refactor station state transitions" Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/cfg.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index e60df48fa4d4..296620d6ca0c 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -791,7 +791,7 @@ static int sta_apply_parameters(struct ieee80211_local *local, if (set & BIT(NL80211_STA_FLAG_AUTHORIZED)) ret = sta_info_move_state_checked(sta, IEEE80211_STA_AUTHORIZED); - else + else if (test_sta_flag(sta, WLAN_STA_AUTHORIZED)) ret = sta_info_move_state_checked(sta, IEEE80211_STA_ASSOC); if (ret) From 84e31fdb7c797a7303e0cc295cb9bc8b73fb872d Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sat, 14 Jan 2012 21:27:37 +0300 Subject: [PATCH 282/820] crypto: sha512 - make it work, undo percpu message schedule commit f9e2bca6c22d75a289a349f869701214d63b5060 aka "crypto: sha512 - Move message schedule W[80] to static percpu area" created global message schedule area. If sha512_update will ever be entered twice, hash will be silently calculated incorrectly. Probably the easiest way to notice incorrect hashes being calculated is to run 2 ping floods over AH with hmac(sha512): #!/usr/sbin/setkey -f flush; spdflush; add IP1 IP2 ah 25 -A hmac-sha512 0x00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000025; add IP2 IP1 ah 52 -A hmac-sha512 0x00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000052; spdadd IP1 IP2 any -P out ipsec ah/transport//require; spdadd IP2 IP1 any -P in ipsec ah/transport//require; XfrmInStateProtoError will start ticking with -EBADMSG being returned from ah_input(). This never happens with, say, hmac(sha1). With patch applied (on BOTH sides), XfrmInStateProtoError does not tick with multiple bidirectional ping flood streams like it doesn't tick with SHA-1. After this patch sha512_transform() will start using ~750 bytes of stack on x86_64. This is OK for simple loads, for something more heavy, stack reduction will be done separatedly. Signed-off-by: Alexey Dobriyan Cc: stable@vger.kernel.org Signed-off-by: Herbert Xu --- crypto/sha512_generic.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/crypto/sha512_generic.c b/crypto/sha512_generic.c index 9ed9f60316e5..8b9035b0189c 100644 --- a/crypto/sha512_generic.c +++ b/crypto/sha512_generic.c @@ -21,8 +21,6 @@ #include #include -static DEFINE_PER_CPU(u64[80], msg_schedule); - static inline u64 Ch(u64 x, u64 y, u64 z) { return z ^ (x & (y ^ z)); @@ -89,7 +87,7 @@ sha512_transform(u64 *state, const u8 *input) u64 a, b, c, d, e, f, g, h, t1, t2; int i; - u64 *W = get_cpu_var(msg_schedule); + u64 W[80]; /* load the input */ for (i = 0; i < 16; i++) @@ -128,8 +126,6 @@ sha512_transform(u64 *state, const u8 *input) /* erase our data */ a = b = c = d = e = f = g = h = t1 = t2 = 0; - memset(W, 0, sizeof(__get_cpu_var(msg_schedule))); - put_cpu_var(msg_schedule); } static int From 51fc6dc8f948047364f7d42a4ed89b416c6cc0a3 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sat, 14 Jan 2012 21:40:57 +0300 Subject: [PATCH 283/820] crypto: sha512 - reduce stack usage to safe number For rounds 16--79, W[i] only depends on W[i - 2], W[i - 7], W[i - 15] and W[i - 16]. Consequently, keeping all W[80] array on stack is unnecessary, only 16 values are really needed. Using W[16] instead of W[80] greatly reduces stack usage (~750 bytes to ~340 bytes on x86_64). Line by line explanation: * BLEND_OP array is "circular" now, all indexes have to be modulo 16. Round number is positive, so remainder operation should be without surprises. * initial full message scheduling is trimmed to first 16 values which come from data block, the rest is calculated before it's needed. * original loop body is unrolled version of new SHA512_0_15 and SHA512_16_79 macros, unrolling was done to not do explicit variable renaming. Otherwise it's the very same code after preprocessing. See sha1_transform() code which does the same trick. Patch survives in-tree crypto test and original bugreport test (ping flood with hmac(sha512). See FIPS 180-2 for SHA-512 definition http://csrc.nist.gov/publications/fips/fips180-2/fips180-2withchangenotice.pdf Signed-off-by: Alexey Dobriyan Cc: stable@vger.kernel.org Signed-off-by: Herbert Xu --- crypto/sha512_generic.c | 58 ++++++++++++++++++++++++----------------- 1 file changed, 34 insertions(+), 24 deletions(-) diff --git a/crypto/sha512_generic.c b/crypto/sha512_generic.c index 8b9035b0189c..88f160b77b1f 100644 --- a/crypto/sha512_generic.c +++ b/crypto/sha512_generic.c @@ -78,7 +78,7 @@ static inline void LOAD_OP(int I, u64 *W, const u8 *input) static inline void BLEND_OP(int I, u64 *W) { - W[I] = s1(W[I-2]) + W[I-7] + s0(W[I-15]) + W[I-16]; + W[I % 16] += s1(W[(I-2) % 16]) + W[(I-7) % 16] + s0(W[(I-15) % 16]); } static void @@ -87,38 +87,48 @@ sha512_transform(u64 *state, const u8 *input) u64 a, b, c, d, e, f, g, h, t1, t2; int i; - u64 W[80]; + u64 W[16]; /* load the input */ for (i = 0; i < 16; i++) LOAD_OP(i, W, input); - for (i = 16; i < 80; i++) { - BLEND_OP(i, W); - } - /* load the state into our registers */ a=state[0]; b=state[1]; c=state[2]; d=state[3]; e=state[4]; f=state[5]; g=state[6]; h=state[7]; - /* now iterate */ - for (i=0; i<80; i+=8) { - t1 = h + e1(e) + Ch(e,f,g) + sha512_K[i ] + W[i ]; - t2 = e0(a) + Maj(a,b,c); d+=t1; h=t1+t2; - t1 = g + e1(d) + Ch(d,e,f) + sha512_K[i+1] + W[i+1]; - t2 = e0(h) + Maj(h,a,b); c+=t1; g=t1+t2; - t1 = f + e1(c) + Ch(c,d,e) + sha512_K[i+2] + W[i+2]; - t2 = e0(g) + Maj(g,h,a); b+=t1; f=t1+t2; - t1 = e + e1(b) + Ch(b,c,d) + sha512_K[i+3] + W[i+3]; - t2 = e0(f) + Maj(f,g,h); a+=t1; e=t1+t2; - t1 = d + e1(a) + Ch(a,b,c) + sha512_K[i+4] + W[i+4]; - t2 = e0(e) + Maj(e,f,g); h+=t1; d=t1+t2; - t1 = c + e1(h) + Ch(h,a,b) + sha512_K[i+5] + W[i+5]; - t2 = e0(d) + Maj(d,e,f); g+=t1; c=t1+t2; - t1 = b + e1(g) + Ch(g,h,a) + sha512_K[i+6] + W[i+6]; - t2 = e0(c) + Maj(c,d,e); f+=t1; b=t1+t2; - t1 = a + e1(f) + Ch(f,g,h) + sha512_K[i+7] + W[i+7]; - t2 = e0(b) + Maj(b,c,d); e+=t1; a=t1+t2; +#define SHA512_0_15(i, a, b, c, d, e, f, g, h) \ + t1 = h + e1(e) + Ch(e, f, g) + sha512_K[i] + W[i]; \ + t2 = e0(a) + Maj(a, b, c); \ + d += t1; \ + h = t1 + t2 + +#define SHA512_16_79(i, a, b, c, d, e, f, g, h) \ + BLEND_OP(i, W); \ + t1 = h + e1(e) + Ch(e, f, g) + sha512_K[i] + W[(i)%16]; \ + t2 = e0(a) + Maj(a, b, c); \ + d += t1; \ + h = t1 + t2 + + for (i = 0; i < 16; i += 8) { + SHA512_0_15(i, a, b, c, d, e, f, g, h); + SHA512_0_15(i + 1, h, a, b, c, d, e, f, g); + SHA512_0_15(i + 2, g, h, a, b, c, d, e, f); + SHA512_0_15(i + 3, f, g, h, a, b, c, d, e); + SHA512_0_15(i + 4, e, f, g, h, a, b, c, d); + SHA512_0_15(i + 5, d, e, f, g, h, a, b, c); + SHA512_0_15(i + 6, c, d, e, f, g, h, a, b); + SHA512_0_15(i + 7, b, c, d, e, f, g, h, a); + } + for (i = 16; i < 80; i += 8) { + SHA512_16_79(i, a, b, c, d, e, f, g, h); + SHA512_16_79(i + 1, h, a, b, c, d, e, f, g); + SHA512_16_79(i + 2, g, h, a, b, c, d, e, f); + SHA512_16_79(i + 3, f, g, h, a, b, c, d, e); + SHA512_16_79(i + 4, e, f, g, h, a, b, c, d); + SHA512_16_79(i + 5, d, e, f, g, h, a, b, c); + SHA512_16_79(i + 6, c, d, e, f, g, h, a, b); + SHA512_16_79(i + 7, b, c, d, e, f, g, h, a); } state[0] += a; state[1] += b; state[2] += c; state[3] += d; From f4b1952a453f4b4b4c4cd3436b9cd632bf4202aa Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Fri, 13 Jan 2012 11:09:23 -0200 Subject: [PATCH 284/820] [media] dvb-core: fix a regression with MythTV As reported by Lawrence[1], MythTV 0.24.1 does the wrong thing with a DVBv5 call: it fills the delivery system with SYS_UNDEFINED, expecting that the DVB core would work with that. This used to work by accident, as the DVB core were missing the check for the supported delivery systems. Yet, fixing it is easy, so let's add a logic to handle this case, to provide backward compatibility. [1] http://patchwork.linuxtv.org/patch/8314/ Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-core/dvb_frontend.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/media/dvb/dvb-core/dvb_frontend.c b/drivers/media/dvb/dvb-core/dvb_frontend.c index b15db4fe347b..2ad7faf4529d 100644 --- a/drivers/media/dvb/dvb-core/dvb_frontend.c +++ b/drivers/media/dvb/dvb-core/dvb_frontend.c @@ -1413,6 +1413,15 @@ static int set_delivery_system(struct dvb_frontend *fe, u32 desired_system) struct dtv_frontend_properties *c = &fe->dtv_property_cache; enum dvbv3_emulation_type type; + /* + * It was reported that some old DVBv5 applications were + * filling delivery_system with SYS_UNDEFINED. If this happens, + * assume that the application wants to use the first supported + * delivery system. + */ + if (c->delivery_system == SYS_UNDEFINED) + c->delivery_system = fe->ops.delsys[0]; + if (desired_system == SYS_UNDEFINED) { /* * A DVBv3 call doesn't know what's the desired system. From 149709f5b8a4a8678401facb5c670119751f6087 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Fri, 13 Jan 2012 11:46:36 -0200 Subject: [PATCH 285/820] [media] dvb-core: preserve the delivery system at cache clear The changeset 240ab508aa is incomplete, as the first thing that happens at cache clear is to do a memset with 0 to the cache. So, the delivery system needs to be explicitly preserved there. Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-core/dvb_frontend.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/media/dvb/dvb-core/dvb_frontend.c b/drivers/media/dvb/dvb-core/dvb_frontend.c index 2ad7faf4529d..f5fa7aa89618 100644 --- a/drivers/media/dvb/dvb-core/dvb_frontend.c +++ b/drivers/media/dvb/dvb-core/dvb_frontend.c @@ -904,8 +904,11 @@ static int dvb_frontend_clear_cache(struct dvb_frontend *fe) { struct dtv_frontend_properties *c = &fe->dtv_property_cache; int i; + u32 delsys; + delsys = c->delivery_system; memset(c, 0, sizeof(struct dtv_frontend_properties)); + c->delivery_system = delsys; c->state = DTV_CLEAR; From 0eecee413d3157bac4477362bfc4b907b4b2988b Mon Sep 17 00:00:00 2001 From: James Smart Date: Thu, 3 Nov 2011 11:10:40 -0400 Subject: [PATCH 286/820] [SCSI] scsi_transport_fc: Clear Devloss Callback Done flag in fc_remote_port_rolechg This patch fixes a bug where devloss is not called on fc_host teardown. The issue is seen if the LLDD uses rport_rolechg to add the target role to an rport. When an rport goes away, the LLDD will call fc_remote_port_delete, which will start the devloss timer. If the timer expires, the transport will call the devloss callback and set the FC_RPORT_DEVLOSS_CALLBK_DONE flag. However, the rport structure is not deleted, it is retained to store the SCSI id mappings for the rport in case it comes back. In the scenario where it does come back, and the driver calls fc_remote_port_add, but does not indicate the "target" role for the rport - the create will clear the structure, but forgets to clear FC_RPORT_DEVLOSS_CALLBK_DONE flag (which is cleared if it's added with the target role). The secondary call, of fc_remote_port_rolechg to add the target role also does not clear the flag. Thus, the next time the rport goes away, the resulting devloss timer expiration will not call the driver callback as the flag is still set. This patch adds the FC_RPORT_DEVLOSS_CALLBK_DONE flags to the list of those that are cleared upon reuse of the rport structure. Signed-off-by: Alex Iannicelli Signed-off-by: James Smart Signed-off-by: James Bottomley --- drivers/scsi/scsi_transport_fc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c index 1b214910b714..f59d4a05ecd7 100644 --- a/drivers/scsi/scsi_transport_fc.c +++ b/drivers/scsi/scsi_transport_fc.c @@ -3048,7 +3048,8 @@ fc_remote_port_rolechg(struct fc_rport *rport, u32 roles) spin_lock_irqsave(shost->host_lock, flags); rport->flags &= ~(FC_RPORT_FAST_FAIL_TIMEDOUT | - FC_RPORT_DEVLOSS_PENDING); + FC_RPORT_DEVLOSS_PENDING | + FC_RPORT_DEVLOSS_CALLBK_DONE); spin_unlock_irqrestore(shost->host_lock, flags); /* ensure any stgt delete functions are done */ From f2d7e40e8069f0d6b4958dbda4383f412f53c7d1 Mon Sep 17 00:00:00 2001 From: Alok Kataria Date: Thu, 10 Nov 2011 20:04:03 -0800 Subject: [PATCH 287/820] [SCSI] MAINTAINERS: pvscsi maintainers update Welcome to Arvind Kumar, our new pvscsi maintainer. Signed-off-by: Alok N Kataria Signed-off-by: James Bottomley --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index a2883a283698..9a2b8f0f94ff 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7171,7 +7171,7 @@ S: Maintained F: drivers/net/vmxnet3/ VMware PVSCSI driver -M: Alok Kataria +M: Arvind Kumar M: VMware PV-Drivers L: linux-scsi@vger.kernel.org S: Maintained From 775bf2773858c50d2acfcdf71889984be94e7037 Mon Sep 17 00:00:00 2001 From: "Stephen M. Cameron" Date: Thu, 5 Jan 2012 10:15:31 -0600 Subject: [PATCH 288/820] [SCSI] hpsa: do not sleep in atomic context in rmmod path. Don't call kthread_stop with a spin lock held and interrupts disabled because kthread_stop will sleep waiting for the thread to stop. Signed-off-by: Stephen M. Cameron Signed-off-by: James Bottomley --- drivers/scsi/hpsa.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c index 5140f5d0fd6b..b96962c39449 100644 --- a/drivers/scsi/hpsa.c +++ b/drivers/scsi/hpsa.c @@ -4271,7 +4271,9 @@ static void stop_controller_lockup_detector(struct ctlr_info *h) remove_ctlr_from_lockup_detector_list(h); /* If the list of ctlr's to monitor is empty, stop the thread */ if (list_empty(&hpsa_ctlr_list)) { + spin_unlock_irqrestore(&lockup_detector_lock, flags); kthread_stop(hpsa_lockup_detector); + spin_lock_irqsave(&lockup_detector_lock, flags); hpsa_lockup_detector = NULL; } spin_unlock_irqrestore(&lockup_detector_lock, flags); From f7e6746ebae984ea67b0a1a1e23c7e6698240631 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 14 Jan 2012 00:01:01 +0000 Subject: [PATCH 289/820] sched/accounting, proc: Fix /proc/stat interrupts sum Commit 3292beb340c7688 ("sched/accounting: Change cpustat fields to an array") deleted the code which provides us with the sum of all interrupts in the system, causing vmstat to report zero interrupts occuring in the system. Fix this by restoring the code. Signed-off-by: Russell King Tested-by: Russell King # [on ARM] Tested-by: Tony Luck Tested-by: Steven Rostedt Cc: Glauber Costa Cc: KAMEZAWA Hiroyuki Cc: Linus Torvalds Cc: Andrew Morton Cc: Paul Tuner Cc: Peter Zijlstra Signed-off-by: Ingo Molnar --- fs/proc/stat.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/proc/stat.c b/fs/proc/stat.c index d76ca6ae2b1b..121f77cfef76 100644 --- a/fs/proc/stat.c +++ b/fs/proc/stat.c @@ -77,6 +77,8 @@ static int show_stat(struct seq_file *p, void *v) steal += kcpustat_cpu(i).cpustat[CPUTIME_STEAL]; guest += kcpustat_cpu(i).cpustat[CPUTIME_GUEST]; guest_nice += kcpustat_cpu(i).cpustat[CPUTIME_GUEST_NICE]; + sum += kstat_cpu_irqs_sum(i); + sum += arch_irq_stat_cpu(i); for (j = 0; j < NR_SOFTIRQS; j++) { unsigned int softirq_stat = kstat_softirqs_cpu(j, i); From 8d973b624ece3b85cfae9474935795d034f72faf Mon Sep 17 00:00:00 2001 From: Ulrich Drepper Date: Sun, 15 Jan 2012 19:40:24 -0500 Subject: [PATCH 290/820] x86/kprobes: Fix typo transferred from Intel manual The arch/x86/lib/x86-opcode-map.txt file [used by the kprobes instruction decoder] contains the line: af: SCAS/W/D/Q rAX,Xv This is what the Intel manuals show, but it's not correct. The 'X' stands for: Memory addressed by the DS:rSI register pair (for example, MOVS, CMPS, OUTS, or LODS). On the other hand 'Y' means (also see the ae byte entry for SCASB): Memory addressed by the ES:rDI register pair (for example, MOVS, CMPS, INS, STOS, or SCAS). Signed-off-by: Ulrich Drepper Acked-by: Masami Hiramatsu Cc: yrl.pp-manager.tt@hitachi.com Link: http://lkml.kernel.org/r/CAOPLpQfytPyDEBF1Hbkpo7ovUerEsstVGxBr%3DEpDL-BKEMaqLA@mail.gmail.com Signed-off-by: Ingo Molnar --- arch/x86/lib/x86-opcode-map.txt | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/x86/lib/x86-opcode-map.txt b/arch/x86/lib/x86-opcode-map.txt index a793da5e560e..8641bbb8e006 100644 --- a/arch/x86/lib/x86-opcode-map.txt +++ b/arch/x86/lib/x86-opcode-map.txt @@ -210,7 +210,9 @@ ab: STOS/W/D/Q Yv,rAX ac: LODS/B AL,Xb ad: LODS/W/D/Q rAX,Xv ae: SCAS/B AL,Yb -af: SCAS/W/D/Q rAX,Xv +# Note: The May 2011 Intel manual shows Xv for the second parameter of the +# next instruction but Yv is correct +af: SCAS/W/D/Q rAX,Yv # 0xb0 - 0xbf b0: MOV AL/R8L,Ib b1: MOV CL/R9L,Ib From a1c611745c8c4e8996c1877d4e5d0fc95f227c38 Mon Sep 17 00:00:00 2001 From: "xiyou.wangcong@gmail.com" Date: Sun, 15 Jan 2012 20:02:17 +0800 Subject: [PATCH 291/820] x86/kprobes: Add arch/x86/tools/insn_sanity to .gitignore After compiling the kernel, I got: % git status # On branch master # Untracked files: # (use "git add ..." to include in what will be committed) # # arch/x86/tools/insn_sanity nothing added to commit but untracked files present (use "git add" to track) it should be added to .gitignore. Signed-off-by: WANG Cong Acked-by: Masami Hiramatsu Link: http://lkml.kernel.org/r/1326628937-27609-1-git-send-email-xiyou.wangcong@gmail.com Signed-off-by: Ingo Molnar --- arch/x86/.gitignore | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/.gitignore b/arch/x86/.gitignore index 028079065af6..7cab8c08e6d1 100644 --- a/arch/x86/.gitignore +++ b/arch/x86/.gitignore @@ -1,3 +1,4 @@ boot/compressed/vmlinux tools/test_get_len +tools/insn_sanity From 7f788ec07d169123bb1bd14e6221a2b228079ce8 Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Thu, 12 Jan 2012 09:18:03 +0100 Subject: [PATCH 292/820] USB: EHCI: Don't use NO_IRQ in xilinx ehci driver Drivers shouldn't use NO_IRQ. This driver is used by Microblaze and PPC. PPC defines NO_IRQ as 0 and Microblaze has removed it. Signed-off-by: Michal Simek Acked-by: Alan Stern CC: Greg Kroah-Hartman CC: Grant Likely CC: linux-usb@vger.kernel.org CC: devicetree-discuss@lists.ozlabs.org --- drivers/usb/host/ehci-xilinx-of.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/host/ehci-xilinx-of.c b/drivers/usb/host/ehci-xilinx-of.c index 32793ce3d9e9..9c2cc4633894 100644 --- a/drivers/usb/host/ehci-xilinx-of.c +++ b/drivers/usb/host/ehci-xilinx-of.c @@ -183,7 +183,7 @@ static int __devinit ehci_hcd_xilinx_of_probe(struct platform_device *op) } irq = irq_of_parse_and_map(dn, 0); - if (irq == NO_IRQ) { + if (!irq) { printk(KERN_ERR "%s: irq_of_parse_and_map failed\n", __FILE__); rv = -EBUSY; goto err_irq; From 0d4c24cc1bf62e75fdb31e73c68daa50e0877564 Mon Sep 17 00:00:00 2001 From: adam radford Date: Fri, 6 Jan 2012 17:02:26 -0800 Subject: [PATCH 293/820] [SCSI] megaraid_sas: Fix reglockFlags for degraded raid5/6 The following patch for megaraid_sas fixes the reglockFlags field for degraded raid5/6 for MR9360/9380, which will result in a performance improvement. Signed-off-by: Adam Radford Signed-off-by: James Bottomley --- drivers/scsi/megaraid/megaraid_sas_fp.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/megaraid/megaraid_sas_fp.c b/drivers/scsi/megaraid/megaraid_sas_fp.c index 5255dd688aca..294abb0defa6 100644 --- a/drivers/scsi/megaraid/megaraid_sas_fp.c +++ b/drivers/scsi/megaraid/megaraid_sas_fp.c @@ -282,7 +282,9 @@ u8 MR_GetPhyParams(struct megasas_instance *instance, u32 ld, u64 stripRow, else { *pDevHandle = MR_PD_INVALID; /* set dev handle as invalid. */ if ((raid->level >= 5) && - (instance->pdev->device != PCI_DEVICE_ID_LSI_INVADER)) + ((instance->pdev->device != PCI_DEVICE_ID_LSI_INVADER) || + (instance->pdev->device == PCI_DEVICE_ID_LSI_INVADER && + raid->regTypeReqOnRead != REGION_TYPE_UNUSED))) pRAID_Context->regLockFlags = REGION_TYPE_EXCLUSIVE; else if (raid->level == 1) { /* Get alternate Pd. */ From 882be7c3be72f6d72ae7a81e707154287ea5d289 Mon Sep 17 00:00:00 2001 From: adam radford Date: Fri, 6 Jan 2012 17:02:34 -0800 Subject: [PATCH 294/820] [SCSI] megaraid_sas: mask off flags in ioctl path Mask off flags in the ioctl path to prevent memory scribble with older MegaCLI versions. Signed-off-by: Adam Radford Signed-off-by: James Bottomley --- drivers/scsi/megaraid/megaraid_sas_base.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c b/drivers/scsi/megaraid/megaraid_sas_base.c index 29a994f9c4f1..2bdc70754b14 100644 --- a/drivers/scsi/megaraid/megaraid_sas_base.c +++ b/drivers/scsi/megaraid/megaraid_sas_base.c @@ -4773,6 +4773,8 @@ megasas_mgmt_fw_ioctl(struct megasas_instance *instance, memcpy(cmd->frame, ioc->frame.raw, 2 * MEGAMFI_FRAME_SIZE); cmd->frame->hdr.context = cmd->index; cmd->frame->hdr.pad_0 = 0; + cmd->frame->hdr.flags &= ~(MFI_FRAME_IEEE | MFI_FRAME_SGL64 | + MFI_FRAME_SENSE64); /* * The management interface between applications and the fw uses From 7895f9c957ff5dbce8e1b96ed11919bb2e20641f Mon Sep 17 00:00:00 2001 From: adam radford Date: Fri, 6 Jan 2012 17:02:40 -0800 Subject: [PATCH 295/820] [SCSI] megaraid_sas: remove poll_mode_io code This code has never worked correctly, doesn't disable interrupts when set as a module parameter, doesn't disable interrupts when set after driver load time in sysfs node, etc. Signed-off-by: Adam Radford Signed-off-by: James Bottomley --- drivers/scsi/megaraid/megaraid_sas.h | 2 - drivers/scsi/megaraid/megaraid_sas_base.c | 141 ---------------------- 2 files changed, 143 deletions(-) diff --git a/drivers/scsi/megaraid/megaraid_sas.h b/drivers/scsi/megaraid/megaraid_sas.h index dd94c7d574fb..7a03a9a6bde0 100644 --- a/drivers/scsi/megaraid/megaraid_sas.h +++ b/drivers/scsi/megaraid/megaraid_sas.h @@ -773,7 +773,6 @@ struct megasas_ctrl_info { #define MFI_OB_INTR_STATUS_MASK 0x00000002 #define MFI_POLL_TIMEOUT_SECS 60 -#define MEGASAS_COMPLETION_TIMER_INTERVAL (HZ/10) #define MFI_REPLY_1078_MESSAGE_INTERRUPT 0x80000000 #define MFI_REPLY_GEN2_MESSAGE_INTERRUPT 0x00000001 @@ -1353,7 +1352,6 @@ struct megasas_instance { u32 mfiStatus; u32 last_seq_num; - struct timer_list io_completion_timer; struct list_head internal_reset_pending_q; /* Ptr to hba specific information */ diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c b/drivers/scsi/megaraid/megaraid_sas_base.c index 2bdc70754b14..f42bd88a30e0 100644 --- a/drivers/scsi/megaraid/megaraid_sas_base.c +++ b/drivers/scsi/megaraid/megaraid_sas_base.c @@ -58,14 +58,6 @@ #include "megaraid_sas_fusion.h" #include "megaraid_sas.h" -/* - * poll_mode_io:1- schedule complete completion from q cmd - */ -static unsigned int poll_mode_io; -module_param_named(poll_mode_io, poll_mode_io, int, 0); -MODULE_PARM_DESC(poll_mode_io, - "Complete cmds from IO path, (default=0)"); - /* * Number of sectors per IO command * Will be set in megasas_init_mfi if user does not provide @@ -1439,11 +1431,6 @@ megasas_build_and_issue_cmd(struct megasas_instance *instance, instance->instancet->fire_cmd(instance, cmd->frame_phys_addr, cmd->frame_count-1, instance->reg_set); - /* - * Check if we have pend cmds to be completed - */ - if (poll_mode_io && atomic_read(&instance->fw_outstanding)) - tasklet_schedule(&instance->isr_tasklet); return 0; out_return_cmd: @@ -3370,47 +3357,6 @@ fail_fw_init: return -EINVAL; } -/** - * megasas_start_timer - Initializes a timer object - * @instance: Adapter soft state - * @timer: timer object to be initialized - * @fn: timer function - * @interval: time interval between timer function call - */ -static inline void -megasas_start_timer(struct megasas_instance *instance, - struct timer_list *timer, - void *fn, unsigned long interval) -{ - init_timer(timer); - timer->expires = jiffies + interval; - timer->data = (unsigned long)instance; - timer->function = fn; - add_timer(timer); -} - -/** - * megasas_io_completion_timer - Timer fn - * @instance_addr: Address of adapter soft state - * - * Schedules tasklet for cmd completion - * if poll_mode_io is set - */ -static void -megasas_io_completion_timer(unsigned long instance_addr) -{ - struct megasas_instance *instance = - (struct megasas_instance *)instance_addr; - - if (atomic_read(&instance->fw_outstanding)) - tasklet_schedule(&instance->isr_tasklet); - - /* Restart timer */ - if (poll_mode_io) - mod_timer(&instance->io_completion_timer, - jiffies + MEGASAS_COMPLETION_TIMER_INTERVAL); -} - static u32 megasas_init_adapter_mfi(struct megasas_instance *instance) { @@ -3638,11 +3584,6 @@ static int megasas_init_fw(struct megasas_instance *instance) tasklet_init(&instance->isr_tasklet, instance->instancet->tasklet, (unsigned long)instance); - /* Initialize the cmd completion timer */ - if (poll_mode_io) - megasas_start_timer(instance, &instance->io_completion_timer, - megasas_io_completion_timer, - MEGASAS_COMPLETION_TIMER_INTERVAL); return 0; fail_init_adapter: @@ -4369,9 +4310,6 @@ megasas_suspend(struct pci_dev *pdev, pm_message_t state) host = instance->host; instance->unload = 1; - if (poll_mode_io) - del_timer_sync(&instance->io_completion_timer); - megasas_flush_cache(instance); megasas_shutdown_controller(instance, MR_DCMD_HIBERNATE_SHUTDOWN); @@ -4511,12 +4449,6 @@ megasas_resume(struct pci_dev *pdev) } instance->instancet->enable_intr(instance->reg_set); - - /* Initialize the cmd completion timer */ - if (poll_mode_io) - megasas_start_timer(instance, &instance->io_completion_timer, - megasas_io_completion_timer, - MEGASAS_COMPLETION_TIMER_INTERVAL); instance->unload = 0; /* @@ -4570,9 +4502,6 @@ static void __devexit megasas_detach_one(struct pci_dev *pdev) host = instance->host; fusion = instance->ctrl_context; - if (poll_mode_io) - del_timer_sync(&instance->io_completion_timer); - scsi_remove_host(instance->host); megasas_flush_cache(instance); megasas_shutdown_controller(instance, MR_DCMD_CTRL_SHUTDOWN); @@ -5221,60 +5150,6 @@ megasas_sysfs_set_dbg_lvl(struct device_driver *dd, const char *buf, size_t coun static DRIVER_ATTR(dbg_lvl, S_IRUGO|S_IWUSR, megasas_sysfs_show_dbg_lvl, megasas_sysfs_set_dbg_lvl); -static ssize_t -megasas_sysfs_show_poll_mode_io(struct device_driver *dd, char *buf) -{ - return sprintf(buf, "%u\n", poll_mode_io); -} - -static ssize_t -megasas_sysfs_set_poll_mode_io(struct device_driver *dd, - const char *buf, size_t count) -{ - int retval = count; - int tmp = poll_mode_io; - int i; - struct megasas_instance *instance; - - if (sscanf(buf, "%u", &poll_mode_io) < 1) { - printk(KERN_ERR "megasas: could not set poll_mode_io\n"); - retval = -EINVAL; - } - - /* - * Check if poll_mode_io is already set or is same as previous value - */ - if ((tmp && poll_mode_io) || (tmp == poll_mode_io)) - goto out; - - if (poll_mode_io) { - /* - * Start timers for all adapters - */ - for (i = 0; i < megasas_mgmt_info.max_index; i++) { - instance = megasas_mgmt_info.instance[i]; - if (instance) { - megasas_start_timer(instance, - &instance->io_completion_timer, - megasas_io_completion_timer, - MEGASAS_COMPLETION_TIMER_INTERVAL); - } - } - } else { - /* - * Delete timers for all adapters - */ - for (i = 0; i < megasas_mgmt_info.max_index; i++) { - instance = megasas_mgmt_info.instance[i]; - if (instance) - del_timer_sync(&instance->io_completion_timer); - } - } - -out: - return retval; -} - static void megasas_aen_polling(struct work_struct *work) { @@ -5504,11 +5379,6 @@ megasas_aen_polling(struct work_struct *work) kfree(ev); } - -static DRIVER_ATTR(poll_mode_io, S_IRUGO|S_IWUSR, - megasas_sysfs_show_poll_mode_io, - megasas_sysfs_set_poll_mode_io); - /** * megasas_init - Driver load entry point */ @@ -5567,11 +5437,6 @@ static int __init megasas_init(void) &driver_attr_dbg_lvl); if (rval) goto err_dcf_dbg_lvl; - rval = driver_create_file(&megasas_pci_driver.driver, - &driver_attr_poll_mode_io); - if (rval) - goto err_dcf_poll_mode_io; - rval = driver_create_file(&megasas_pci_driver.driver, &driver_attr_support_device_change); if (rval) @@ -5580,10 +5445,6 @@ static int __init megasas_init(void) return rval; err_dcf_support_device_change: - driver_remove_file(&megasas_pci_driver.driver, - &driver_attr_poll_mode_io); - -err_dcf_poll_mode_io: driver_remove_file(&megasas_pci_driver.driver, &driver_attr_dbg_lvl); err_dcf_dbg_lvl: @@ -5608,8 +5469,6 @@ err_pcidrv: */ static void __exit megasas_exit(void) { - driver_remove_file(&megasas_pci_driver.driver, - &driver_attr_poll_mode_io); driver_remove_file(&megasas_pci_driver.driver, &driver_attr_dbg_lvl); driver_remove_file(&megasas_pci_driver.driver, From 798edaadba97612224c250d50f880405d75a3af1 Mon Sep 17 00:00:00 2001 From: adam radford Date: Fri, 6 Jan 2012 17:02:45 -0800 Subject: [PATCH 296/820] [SCSI] megaraid_sas: Version and Changelog update Signed-off-by: Adam Radford Signed-off-by: James Bottomley --- Documentation/scsi/ChangeLog.megaraid_sas | 10 ++++++++++ drivers/scsi/megaraid/megaraid_sas.h | 6 +++--- drivers/scsi/megaraid/megaraid_sas_base.c | 2 +- 3 files changed, 14 insertions(+), 4 deletions(-) diff --git a/Documentation/scsi/ChangeLog.megaraid_sas b/Documentation/scsi/ChangeLog.megaraid_sas index 64adb98b181c..57566bacb4c5 100644 --- a/Documentation/scsi/ChangeLog.megaraid_sas +++ b/Documentation/scsi/ChangeLog.megaraid_sas @@ -1,3 +1,13 @@ +Release Date : Fri. Jan 6, 2012 17:00:00 PST 2010 - + (emaild-id:megaraidlinux@lsi.com) + Adam Radford +Current Version : 00.00.06.14-rc1 +Old Version : 00.00.06.12-rc1 + 1. Fix reglockFlags for degraded raid5/6 for MR 9360/9380. + 2. Mask off flags in ioctl path to prevent memory scribble with older + MegaCLI versions. + 3. Remove poll_mode_io module paramater, sysfs node, and associated code. +------------------------------------------------------------------------------- Release Date : Wed. Oct 5, 2011 17:00:00 PST 2010 - (emaild-id:megaraidlinux@lsi.com) Adam Radford diff --git a/drivers/scsi/megaraid/megaraid_sas.h b/drivers/scsi/megaraid/megaraid_sas.h index 7a03a9a6bde0..e5f416f8042d 100644 --- a/drivers/scsi/megaraid/megaraid_sas.h +++ b/drivers/scsi/megaraid/megaraid_sas.h @@ -33,9 +33,9 @@ /* * MegaRAID SAS Driver meta data */ -#define MEGASAS_VERSION "00.00.06.12-rc1" -#define MEGASAS_RELDATE "Oct. 5, 2011" -#define MEGASAS_EXT_VERSION "Wed. Oct. 5 17:00:00 PDT 2011" +#define MEGASAS_VERSION "00.00.06.14-rc1" +#define MEGASAS_RELDATE "Jan. 6, 2012" +#define MEGASAS_EXT_VERSION "Fri. Jan. 6 17:00:00 PDT 2012" /* * Device IDs diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c b/drivers/scsi/megaraid/megaraid_sas_base.c index f42bd88a30e0..8b300be44284 100644 --- a/drivers/scsi/megaraid/megaraid_sas_base.c +++ b/drivers/scsi/megaraid/megaraid_sas_base.c @@ -18,7 +18,7 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * * FILE: megaraid_sas_base.c - * Version : v00.00.06.12-rc1 + * Version : v00.00.06.14-rc1 * * Authors: LSI Corporation * Sreenivas Bagalkote From 7d99b3abaf8412932269a2a2e561138326d61f8b Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Wed, 4 Jan 2012 01:32:28 -0800 Subject: [PATCH 297/820] [SCSI] isci, firmware: Remove isci fallback parameter blob and generator This parameter blob and generator program have been moved to the linux-firmware.git repository. Signed-off-by: Ben Hutchings Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/firmware/Makefile | 19 ----- drivers/scsi/isci/firmware/README | 36 ---------- drivers/scsi/isci/firmware/create_fw.c | 99 -------------------------- drivers/scsi/isci/firmware/create_fw.h | 77 -------------------- firmware/Makefile | 1 - firmware/isci/isci_firmware.bin.ihex | 16 ----- 6 files changed, 248 deletions(-) delete mode 100644 drivers/scsi/isci/firmware/Makefile delete mode 100644 drivers/scsi/isci/firmware/README delete mode 100644 drivers/scsi/isci/firmware/create_fw.c delete mode 100644 drivers/scsi/isci/firmware/create_fw.h delete mode 100644 firmware/isci/isci_firmware.bin.ihex diff --git a/drivers/scsi/isci/firmware/Makefile b/drivers/scsi/isci/firmware/Makefile deleted file mode 100644 index 5f54461cabc5..000000000000 --- a/drivers/scsi/isci/firmware/Makefile +++ /dev/null @@ -1,19 +0,0 @@ -# Makefile for create_fw -# -CC=gcc -CFLAGS=-c -Wall -O2 -g -LDFLAGS= -SOURCES=create_fw.c -OBJECTS=$(SOURCES:.cpp=.o) -EXECUTABLE=create_fw - -all: $(SOURCES) $(EXECUTABLE) - -$(EXECUTABLE): $(OBJECTS) - $(CC) $(LDFLAGS) $(OBJECTS) -o $@ - -.c.o: - $(CC) $(CFLAGS) $< -O $@ - -clean: - rm -f *.o $(EXECUTABLE) diff --git a/drivers/scsi/isci/firmware/README b/drivers/scsi/isci/firmware/README deleted file mode 100644 index 8056d2bd233b..000000000000 --- a/drivers/scsi/isci/firmware/README +++ /dev/null @@ -1,36 +0,0 @@ -This defines the temporary binary blow we are to pass to the SCU -driver to emulate the binary firmware that we will eventually be -able to access via NVRAM on the SCU controller. - -The current size of the binary blob is expected to be 149 bytes or larger - -Header Types: -0x1: Phy Masks -0x2: Phy Gens -0x3: SAS Addrs -0xff: End of Data - -ID string - u8[12]: "#SCU MAGIC#\0" -Version - u8: 1 -SubVersion - u8: 0 - -Header Type - u8: 0x1 -Size - u8: 8 -Phy Mask - u32[8] - -Header Type - u8: 0x2 -Size - u8: 8 -Phy Gen - u32[8] - -Header Type - u8: 0x3 -Size - u8: 8 -Sas Addr - u64[8] - -Header Type - u8: 0xf - - -============================================================================== - -Place isci_firmware.bin in /lib/firmware -Be sure to recreate the initramfs image to include the firmware. - diff --git a/drivers/scsi/isci/firmware/create_fw.c b/drivers/scsi/isci/firmware/create_fw.c deleted file mode 100644 index c7a2887a7e95..000000000000 --- a/drivers/scsi/isci/firmware/create_fw.c +++ /dev/null @@ -1,99 +0,0 @@ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include "create_fw.h" -#include "../probe_roms.h" - -int write_blob(struct isci_orom *isci_orom) -{ - FILE *fd; - int err; - size_t count; - - fd = fopen(blob_name, "w+"); - if (!fd) { - perror("Open file for write failed"); - fclose(fd); - return -EIO; - } - - count = fwrite(isci_orom, sizeof(struct isci_orom), 1, fd); - if (count != 1) { - perror("Write data failed"); - fclose(fd); - return -EIO; - } - - fclose(fd); - - return 0; -} - -void set_binary_values(struct isci_orom *isci_orom) -{ - int ctrl_idx, phy_idx, port_idx; - - /* setting OROM signature */ - strncpy(isci_orom->hdr.signature, sig, strlen(sig)); - isci_orom->hdr.version = version; - isci_orom->hdr.total_block_length = sizeof(struct isci_orom); - isci_orom->hdr.hdr_length = sizeof(struct sci_bios_oem_param_block_hdr); - isci_orom->hdr.num_elements = num_elements; - - for (ctrl_idx = 0; ctrl_idx < 2; ctrl_idx++) { - isci_orom->ctrl[ctrl_idx].controller.mode_type = mode_type; - isci_orom->ctrl[ctrl_idx].controller.max_concurrent_dev_spin_up = - max_num_concurrent_dev_spin_up; - isci_orom->ctrl[ctrl_idx].controller.do_enable_ssc = - enable_ssc; - - for (port_idx = 0; port_idx < 4; port_idx++) - isci_orom->ctrl[ctrl_idx].ports[port_idx].phy_mask = - phy_mask[ctrl_idx][port_idx]; - - for (phy_idx = 0; phy_idx < 4; phy_idx++) { - isci_orom->ctrl[ctrl_idx].phys[phy_idx].sas_address.high = - (__u32)(sas_addr[ctrl_idx][phy_idx] >> 32); - isci_orom->ctrl[ctrl_idx].phys[phy_idx].sas_address.low = - (__u32)(sas_addr[ctrl_idx][phy_idx]); - - isci_orom->ctrl[ctrl_idx].phys[phy_idx].afe_tx_amp_control0 = - afe_tx_amp_control0; - isci_orom->ctrl[ctrl_idx].phys[phy_idx].afe_tx_amp_control1 = - afe_tx_amp_control1; - isci_orom->ctrl[ctrl_idx].phys[phy_idx].afe_tx_amp_control2 = - afe_tx_amp_control2; - isci_orom->ctrl[ctrl_idx].phys[phy_idx].afe_tx_amp_control3 = - afe_tx_amp_control3; - } - } -} - -int main(void) -{ - int err; - struct isci_orom *isci_orom; - - isci_orom = malloc(sizeof(struct isci_orom)); - memset(isci_orom, 0, sizeof(struct isci_orom)); - - set_binary_values(isci_orom); - - err = write_blob(isci_orom); - if (err < 0) { - free(isci_orom); - return err; - } - - free(isci_orom); - return 0; -} diff --git a/drivers/scsi/isci/firmware/create_fw.h b/drivers/scsi/isci/firmware/create_fw.h deleted file mode 100644 index 5f298828d22e..000000000000 --- a/drivers/scsi/isci/firmware/create_fw.h +++ /dev/null @@ -1,77 +0,0 @@ -#ifndef _CREATE_FW_H_ -#define _CREATE_FW_H_ -#include "../probe_roms.h" - - -/* we are configuring for 2 SCUs */ -static const int num_elements = 2; - -/* - * For all defined arrays: - * elements 0-3 are for SCU0, ports 0-3 - * elements 4-7 are for SCU1, ports 0-3 - * - * valid configurations for one SCU are: - * P0 P1 P2 P3 - * ---------------- - * 0xF,0x0,0x0,0x0 # 1 x4 port - * 0x3,0x0,0x4,0x8 # Phys 0 and 1 are a x2 port, phy 2 and phy 3 are each x1 - * # ports - * 0x1,0x2,0xC,0x0 # Phys 0 and 1 are each x1 ports, phy 2 and phy 3 are a x2 - * # port - * 0x3,0x0,0xC,0x0 # Phys 0 and 1 are a x2 port, phy 2 and phy 3 are a x2 port - * 0x1,0x2,0x4,0x8 # Each phy is a x1 port (this is the default configuration) - * - * if there is a port/phy on which you do not wish to override the default - * values, use the value assigned to UNINIT_PARAM (255). - */ - -/* discovery mode type (port auto config mode by default ) */ - -/* - * if there is a port/phy on which you do not wish to override the default - * values, use the value "0000000000000000". SAS address of zero's is - * considered invalid and will not be used. - */ -#ifdef MPC -static const int mode_type = SCIC_PORT_MANUAL_CONFIGURATION_MODE; -static const __u8 phy_mask[2][4] = { {1, 2, 4, 8}, - {1, 2, 4, 8} }; -static const unsigned long long sas_addr[2][4] = { { 0x5FCFFFFFF0000001ULL, - 0x5FCFFFFFF0000002ULL, - 0x5FCFFFFFF0000003ULL, - 0x5FCFFFFFF0000004ULL }, - { 0x5FCFFFFFF0000005ULL, - 0x5FCFFFFFF0000006ULL, - 0x5FCFFFFFF0000007ULL, - 0x5FCFFFFFF0000008ULL } }; -#else /* APC (default) */ -static const int mode_type = SCIC_PORT_AUTOMATIC_CONFIGURATION_MODE; -static const __u8 phy_mask[2][4]; -static const unsigned long long sas_addr[2][4] = { { 0x5FCFFFFF00000001ULL, - 0x5FCFFFFF00000001ULL, - 0x5FCFFFFF00000001ULL, - 0x5FCFFFFF00000001ULL }, - { 0x5FCFFFFF00000002ULL, - 0x5FCFFFFF00000002ULL, - 0x5FCFFFFF00000002ULL, - 0x5FCFFFFF00000002ULL } }; -#endif - -/* Maximum number of concurrent device spin up */ -static const int max_num_concurrent_dev_spin_up = 1; - -/* enable of ssc operation */ -static const int enable_ssc; - -/* AFE_TX_AMP_CONTROL */ -static const unsigned int afe_tx_amp_control0 = 0x000bdd08; -static const unsigned int afe_tx_amp_control1 = 0x000ffc00; -static const unsigned int afe_tx_amp_control2 = 0x000b7c09; -static const unsigned int afe_tx_amp_control3 = 0x000afc6e; - -static const char blob_name[] = "isci_firmware.bin"; -static const char sig[] = "ISCUOEMB"; -static const unsigned char version = 0x10; - -#endif diff --git a/firmware/Makefile b/firmware/Makefile index 5f43bfba3c7a..0d15a3d113a2 100644 --- a/firmware/Makefile +++ b/firmware/Makefile @@ -82,7 +82,6 @@ fw-shipped-$(CONFIG_SERIAL_8250_CS) += cis/MT5634ZLX.cis cis/RS-COM-2P.cis \ fw-shipped-$(CONFIG_PCMCIA_SMC91C92) += ositech/Xilinx7OD.bin fw-shipped-$(CONFIG_SCSI_ADVANSYS) += advansys/mcode.bin advansys/38C1600.bin \ advansys/3550.bin advansys/38C0800.bin -fw-shipped-$(CONFIG_SCSI_ISCI) += isci/isci_firmware.bin fw-shipped-$(CONFIG_SCSI_QLOGIC_1280) += qlogic/1040.bin qlogic/1280.bin \ qlogic/12160.bin fw-shipped-$(CONFIG_SCSI_QLOGICPTI) += qlogic/isp1000.bin diff --git a/firmware/isci/isci_firmware.bin.ihex b/firmware/isci/isci_firmware.bin.ihex deleted file mode 100644 index 2e6619570072..000000000000 --- a/firmware/isci/isci_firmware.bin.ihex +++ /dev/null @@ -1,16 +0,0 @@ -:10000000495343554F454D42E80018100002000087 -:1000100000000000000000000101000000000000DE -:10002000FFFFCF5F0100000008DD0B0000FC0F00A8 -:10003000097C0B006EFC0A00FFFFCF5F010000008F -:1000400008DD0B0000FC0F00097C0B006EFC0A00B1 -:10005000FFFFCF5F0100000008DD0B0000FC0F0078 -:10006000097C0B006EFC0A00FFFFCF5F010000005F -:1000700008DD0B0000FC0F00097C0B006EFC0A0081 -:100080000101000000000000FFFFCF5F0200000040 -:1000900008DD0B0000FC0F00097C0B006EFC0A0061 -:1000A000FFFFCF5F0200000008DD0B0000FC0F0027 -:1000B000097C0B006EFC0A00FFFFCF5F020000000E -:1000C00008DD0B0000FC0F00097C0B006EFC0A0031 -:1000D000FFFFCF5F0200000008DD0B0000FC0F00F7 -:0800E000097C0B006EFC0A0014 -:00000001FF From 5d53cb27d849c899136c048ec84c940ac449494b Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 13 Jan 2012 10:14:12 -0800 Subject: [PATCH 298/820] memblock: Fix alloc failure due to dumb underflow protection in memblock_find_in_range_node() 7bd0b0f0da ("memblock: Reimplement memblock allocation using reverse free area iterator") implemented a simple top-down allocator using a reverse memblock iterator. To avoid underflow in the allocator loop, it simply raised the lower boundary to the requested size under the assumption that requested size would be far smaller than available memblocks. This causes early page table allocation failure under certain configurations in Xen. Fix it by checking for underflow directly instead of bumping up lower bound. Signed-off-by: Tejun Heo Reported-by: Konrad Rzeszutek Wilk Cc: rjw@sisk.pl Cc: xen-devel@lists.xensource.com Cc: Benjamin Herrenschmidt Cc: Andrew Morton Cc: Linus Torvalds Link: http://lkml.kernel.org/r/20120113181412.GA11112@google.com Signed-off-by: Ingo Molnar --- mm/memblock.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/mm/memblock.c b/mm/memblock.c index 2f55f19b7c86..77b5f227e1d8 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -106,14 +106,17 @@ phys_addr_t __init_memblock memblock_find_in_range_node(phys_addr_t start, if (end == MEMBLOCK_ALLOC_ACCESSIBLE) end = memblock.current_limit; - /* adjust @start to avoid underflow and allocating the first page */ - start = max3(start, size, (phys_addr_t)PAGE_SIZE); + /* avoid allocating the first page */ + start = max_t(phys_addr_t, start, PAGE_SIZE); end = max(start, end); for_each_free_mem_range_reverse(i, nid, &this_start, &this_end, NULL) { this_start = clamp(this_start, start, end); this_end = clamp(this_end, start, end); + if (this_end < size) + continue; + cand = round_down(this_end - size, align); if (cand >= this_start) return cand; From 2e5da889d44a3a9629f895de3488306e7f5ddf16 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Wed, 4 Jan 2012 01:32:34 -0800 Subject: [PATCH 299/820] [SCSI] isci: cleanup oem parameter and recipe handling Before updating the code to support the latest platform updates and silicon revision cleanup some of the long deref chains. Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 107 +++++++++++++++++---------------------- drivers/scsi/isci/phy.c | 81 ++++++++++++++--------------- 2 files changed, 86 insertions(+), 102 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index e7fe9c4c85b8..8e7de192cf6d 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -1910,160 +1910,147 @@ void sci_controller_power_control_queue_remove(struct isci_host *ihost, #define AFE_REGISTER_WRITE_DELAY 10 -/* Initialize the AFE for this phy index. We need to read the AFE setup from - * the OEM parameters - */ static void sci_controller_afe_initialization(struct isci_host *ihost) { + struct scu_afe_registers __iomem *afe = &ihost->scu_registers->afe; const struct sci_oem_params *oem = &ihost->oem_parameters; struct pci_dev *pdev = ihost->pdev; u32 afe_status; u32 phy_id; /* Clear DFX Status registers */ - writel(0x0081000f, &ihost->scu_registers->afe.afe_dfx_master_control0); + writel(0x0081000f, &afe->afe_dfx_master_control0); udelay(AFE_REGISTER_WRITE_DELAY); if (is_b0(pdev)) { /* PM Rx Equalization Save, PM SPhy Rx Acknowledgement - * Timer, PM Stagger Timer */ - writel(0x0007BFFF, &ihost->scu_registers->afe.afe_pmsn_master_control2); + * Timer, PM Stagger Timer + */ + writel(0x0007BFFF, &afe->afe_pmsn_master_control2); udelay(AFE_REGISTER_WRITE_DELAY); } /* Configure bias currents to normal */ if (is_a2(pdev)) - writel(0x00005A00, &ihost->scu_registers->afe.afe_bias_control); + writel(0x00005A00, &afe->afe_bias_control); else if (is_b0(pdev) || is_c0(pdev)) - writel(0x00005F00, &ihost->scu_registers->afe.afe_bias_control); + writel(0x00005F00, &afe->afe_bias_control); udelay(AFE_REGISTER_WRITE_DELAY); /* Enable PLL */ if (is_b0(pdev) || is_c0(pdev)) - writel(0x80040A08, &ihost->scu_registers->afe.afe_pll_control0); + writel(0x80040A08, &afe->afe_pll_control0); else - writel(0x80040908, &ihost->scu_registers->afe.afe_pll_control0); + writel(0x80040908, &afe->afe_pll_control0); udelay(AFE_REGISTER_WRITE_DELAY); /* Wait for the PLL to lock */ do { - afe_status = readl(&ihost->scu_registers->afe.afe_common_block_status); + afe_status = readl(&afe->afe_common_block_status); udelay(AFE_REGISTER_WRITE_DELAY); } while ((afe_status & 0x00001000) == 0); if (is_a2(pdev)) { - /* Shorten SAS SNW lock time (RxLock timer value from 76 us to 50 us) */ - writel(0x7bcc96ad, &ihost->scu_registers->afe.afe_pmsn_master_control0); + /* Shorten SAS SNW lock time (RxLock timer value from 76 + * us to 50 us) + */ + writel(0x7bcc96ad, &afe->afe_pmsn_master_control0); udelay(AFE_REGISTER_WRITE_DELAY); } for (phy_id = 0; phy_id < SCI_MAX_PHYS; phy_id++) { + struct scu_afe_transceiver *xcvr = &afe->scu_afe_xcvr[phy_id]; const struct sci_phy_oem_params *oem_phy = &oem->phys[phy_id]; if (is_b0(pdev)) { /* Configure transmitter SSC parameters */ - writel(0x00030000, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_ssc_control); + writel(0x00030000, &xcvr->afe_tx_ssc_control); udelay(AFE_REGISTER_WRITE_DELAY); } else if (is_c0(pdev)) { /* Configure transmitter SSC parameters */ - writel(0x0003000, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_ssc_control); + writel(0x0003000, &xcvr->afe_tx_ssc_control); udelay(AFE_REGISTER_WRITE_DELAY); - /* - * All defaults, except the Receive Word Alignament/Comma Detect - * Enable....(0xe800) */ - writel(0x00004500, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_xcvr_control0); + /* All defaults, except the Receive Word + * Alignament/Comma Detect Enable....(0xe800) + */ + writel(0x00004500, &xcvr->afe_xcvr_control0); udelay(AFE_REGISTER_WRITE_DELAY); } else { - /* - * All defaults, except the Receive Word Alignament/Comma Detect - * Enable....(0xe800) */ - writel(0x00004512, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_xcvr_control0); + /* All defaults, except the Receive Word + * Alignament/Comma Detect Enable....(0xe800) + */ + writel(0x00004512, &xcvr->afe_xcvr_control0); udelay(AFE_REGISTER_WRITE_DELAY); - writel(0x0050100F, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_xcvr_control1); + writel(0x0050100F, &xcvr->afe_xcvr_control1); udelay(AFE_REGISTER_WRITE_DELAY); } - /* - * Power up TX and RX out from power down (PWRDNTX and PWRDNRX) - * & increase TX int & ext bias 20%....(0xe85c) */ + /* Power up TX and RX out from power down (PWRDNTX and PWRDNRX) + * & increase TX int & ext bias 20%....(0xe85c) + */ if (is_a2(pdev)) - writel(0x000003F0, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_channel_control); + writel(0x000003F0, &xcvr->afe_channel_control); else if (is_b0(pdev)) { /* Power down TX and RX (PWRDNTX and PWRDNRX) */ - writel(0x000003D7, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_channel_control); + writel(0x000003D7, &xcvr->afe_channel_control); udelay(AFE_REGISTER_WRITE_DELAY); - - /* - * Power up TX and RX out from power down (PWRDNTX and PWRDNRX) - * & increase TX int & ext bias 20%....(0xe85c) */ - writel(0x000003D4, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_channel_control); + writel(0x000003D4, &xcvr->afe_channel_control); } else { - writel(0x000001E7, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_channel_control); + writel(0x000001E7, &xcvr->afe_channel_control); udelay(AFE_REGISTER_WRITE_DELAY); - - /* - * Power up TX and RX out from power down (PWRDNTX and PWRDNRX) - * & increase TX int & ext bias 20%....(0xe85c) */ - writel(0x000001E4, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_channel_control); + writel(0x000001E4, &xcvr->afe_channel_control); } udelay(AFE_REGISTER_WRITE_DELAY); if (is_a2(pdev)) { /* Enable TX equalization (0xe824) */ - writel(0x00040000, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_control); + writel(0x00040000, &xcvr->afe_tx_control); udelay(AFE_REGISTER_WRITE_DELAY); } - /* - * RDPI=0x0(RX Power On), RXOOBDETPDNC=0x0, TPD=0x0(TX Power On), - * RDD=0x0(RX Detect Enabled) ....(0xe800) */ - writel(0x00004100, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_xcvr_control0); + writel(0x00004100, &xcvr->afe_xcvr_control0); udelay(AFE_REGISTER_WRITE_DELAY); /* Leave DFE/FFE on */ if (is_a2(pdev)) - writel(0x3F11103F, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_rx_ssc_control0); + writel(0x3F11103F, &xcvr->afe_rx_ssc_control0); else if (is_b0(pdev)) { - writel(0x3F11103F, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_rx_ssc_control0); + writel(0x3F11103F, &xcvr->afe_rx_ssc_control0); udelay(AFE_REGISTER_WRITE_DELAY); /* Enable TX equalization (0xe824) */ - writel(0x00040000, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_control); + writel(0x00040000, &xcvr->afe_tx_control); } else { - writel(0x0140DF0F, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_rx_ssc_control1); + writel(0x0140DF0F, &xcvr->afe_rx_ssc_control1); udelay(AFE_REGISTER_WRITE_DELAY); - writel(0x3F6F103F, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_rx_ssc_control0); + writel(0x3F6F103F, &xcvr->afe_rx_ssc_control0); udelay(AFE_REGISTER_WRITE_DELAY); /* Enable TX equalization (0xe824) */ - writel(0x00040000, &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_control); + writel(0x00040000, &xcvr->afe_tx_control); } udelay(AFE_REGISTER_WRITE_DELAY); - writel(oem_phy->afe_tx_amp_control0, - &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_amp_control0); + writel(oem_phy->afe_tx_amp_control0, &xcvr->afe_tx_amp_control0); udelay(AFE_REGISTER_WRITE_DELAY); - writel(oem_phy->afe_tx_amp_control1, - &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_amp_control1); + writel(oem_phy->afe_tx_amp_control1, &xcvr->afe_tx_amp_control1); udelay(AFE_REGISTER_WRITE_DELAY); - writel(oem_phy->afe_tx_amp_control2, - &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_amp_control2); + writel(oem_phy->afe_tx_amp_control2, &xcvr->afe_tx_amp_control2); udelay(AFE_REGISTER_WRITE_DELAY); - writel(oem_phy->afe_tx_amp_control3, - &ihost->scu_registers->afe.scu_afe_xcvr[phy_id].afe_tx_amp_control3); + writel(oem_phy->afe_tx_amp_control3, &xcvr->afe_tx_amp_control3); udelay(AFE_REGISTER_WRITE_DELAY); } /* Transfer control to the PEs */ - writel(0x00010f00, &ihost->scu_registers->afe.afe_dfx_master_control0); + writel(0x00010f00, &afe->afe_dfx_master_control0); udelay(AFE_REGISTER_WRITE_DELAY); } diff --git a/drivers/scsi/isci/phy.c b/drivers/scsi/isci/phy.c index 35f50c2183e1..0ae0990b39dd 100644 --- a/drivers/scsi/isci/phy.c +++ b/drivers/scsi/isci/phy.c @@ -91,22 +91,23 @@ sci_phy_transport_layer_initialization(struct isci_phy *iphy, static enum sci_status sci_phy_link_layer_initialization(struct isci_phy *iphy, - struct scu_link_layer_registers __iomem *reg) + struct scu_link_layer_registers __iomem *llr) { struct isci_host *ihost = iphy->owning_port->owning_controller; + struct sci_phy_user_params *phy_user; + struct sci_phy_oem_params *phy_oem; int phy_idx = iphy->phy_index; - struct sci_phy_user_params *phy_user = &ihost->user_parameters.phys[phy_idx]; - struct sci_phy_oem_params *phy_oem = - &ihost->oem_parameters.phys[phy_idx]; - u32 phy_configuration; struct sci_phy_cap phy_cap; + u32 phy_configuration; u32 parity_check = 0; u32 parity_count = 0; u32 llctl, link_rate; u32 clksm_value = 0; u32 sp_timeouts = 0; - iphy->link_layer_registers = reg; + phy_user = &ihost->user_parameters.phys[phy_idx]; + phy_oem = &ihost->oem_parameters.phys[phy_idx]; + iphy->link_layer_registers = llr; /* Set our IDENTIFY frame data */ #define SCI_END_DEVICE 0x01 @@ -116,32 +117,26 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, SCU_SAS_TIID_GEN_BIT(STP_INITIATOR) | SCU_SAS_TIID_GEN_BIT(DA_SATA_HOST) | SCU_SAS_TIID_GEN_VAL(DEVICE_TYPE, SCI_END_DEVICE), - &iphy->link_layer_registers->transmit_identification); + &llr->transmit_identification); /* Write the device SAS Address */ - writel(0xFEDCBA98, - &iphy->link_layer_registers->sas_device_name_high); - writel(phy_idx, &iphy->link_layer_registers->sas_device_name_low); + writel(0xFEDCBA98, &llr->sas_device_name_high); + writel(phy_idx, &llr->sas_device_name_low); /* Write the source SAS Address */ - writel(phy_oem->sas_address.high, - &iphy->link_layer_registers->source_sas_address_high); - writel(phy_oem->sas_address.low, - &iphy->link_layer_registers->source_sas_address_low); + writel(phy_oem->sas_address.high, &llr->source_sas_address_high); + writel(phy_oem->sas_address.low, &llr->source_sas_address_low); /* Clear and Set the PHY Identifier */ - writel(0, &iphy->link_layer_registers->identify_frame_phy_id); - writel(SCU_SAS_TIPID_GEN_VALUE(ID, phy_idx), - &iphy->link_layer_registers->identify_frame_phy_id); + writel(0, &llr->identify_frame_phy_id); + writel(SCU_SAS_TIPID_GEN_VALUE(ID, phy_idx), &llr->identify_frame_phy_id); /* Change the initial state of the phy configuration register */ - phy_configuration = - readl(&iphy->link_layer_registers->phy_configuration); + phy_configuration = readl(&llr->phy_configuration); /* Hold OOB state machine in reset */ phy_configuration |= SCU_SAS_PCFG_GEN_BIT(OOB_RESET); - writel(phy_configuration, - &iphy->link_layer_registers->phy_configuration); + writel(phy_configuration, &llr->phy_configuration); /* Configure the SNW capabilities */ phy_cap.all = 0; @@ -155,9 +150,9 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, phy_cap.gen1_ssc = 1; } - /* - * The SAS specification indicates that the phy_capabilities that - * are transmitted shall have an even parity. Calculate the parity. */ + /* The SAS specification indicates that the phy_capabilities that + * are transmitted shall have an even parity. Calculate the parity. + */ parity_check = phy_cap.all; while (parity_check != 0) { if (parity_check & 0x1) @@ -165,20 +160,20 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, parity_check >>= 1; } - /* - * If parity indicates there are an odd number of bits set, then - * set the parity bit to 1 in the phy capabilities. */ + /* If parity indicates there are an odd number of bits set, then + * set the parity bit to 1 in the phy capabilities. + */ if ((parity_count % 2) != 0) phy_cap.parity = 1; - writel(phy_cap.all, &iphy->link_layer_registers->phy_capabilities); + writel(phy_cap.all, &llr->phy_capabilities); /* Set the enable spinup period but disable the ability to send * notify enable spinup */ writel(SCU_ENSPINUP_GEN_VAL(COUNT, phy_user->notify_enable_spin_up_insertion_frequency), - &iphy->link_layer_registers->notify_enable_spinup_control); + &llr->notify_enable_spinup_control); /* Write the ALIGN Insertion Ferequency for connected phy and * inpendent of connected state @@ -189,11 +184,10 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, clksm_value |= SCU_ALIGN_INSERTION_FREQUENCY_GEN_VAL(GENERAL, phy_user->align_insertion_frequency); - writel(clksm_value, &iphy->link_layer_registers->clock_skew_management); + writel(clksm_value, &llr->clock_skew_management); /* @todo Provide a way to write this register correctly */ - writel(0x02108421, - &iphy->link_layer_registers->afe_lookup_table_control); + writel(0x02108421, &llr->afe_lookup_table_control); llctl = SCU_SAS_LLCTL_GEN_VAL(NO_OUTBOUND_TASK_TIMEOUT, (u8)ihost->user_parameters.no_outbound_task_timeout); @@ -210,9 +204,9 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, break; } llctl |= SCU_SAS_LLCTL_GEN_VAL(MAX_LINK_RATE, link_rate); - writel(llctl, &iphy->link_layer_registers->link_layer_control); + writel(llctl, &llr->link_layer_control); - sp_timeouts = readl(&iphy->link_layer_registers->sas_phy_timeouts); + sp_timeouts = readl(&llr->sas_phy_timeouts); /* Clear the default 0x36 (54us) RATE_CHANGE timeout value. */ sp_timeouts &= ~SCU_SAS_PHYTOV_GEN_VAL(RATE_CHANGE, 0xFF); @@ -222,20 +216,23 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, */ sp_timeouts |= SCU_SAS_PHYTOV_GEN_VAL(RATE_CHANGE, 0x3B); - writel(sp_timeouts, &iphy->link_layer_registers->sas_phy_timeouts); + writel(sp_timeouts, &llr->sas_phy_timeouts); if (is_a2(ihost->pdev)) { - /* Program the max ARB time for the PHY to 700us so we inter-operate with - * the PMC expander which shuts down PHYs if the expander PHY generates too - * many breaks. This time value will guarantee that the initiator PHY will - * generate the break. + /* Program the max ARB time for the PHY to 700us so we + * inter-operate with the PMC expander which shuts down + * PHYs if the expander PHY generates too many breaks. + * This time value will guarantee that the initiator PHY + * will generate the break. */ writel(SCIC_SDS_PHY_MAX_ARBITRATION_WAIT_TIME, - &iphy->link_layer_registers->maximum_arbitration_wait_timer_timeout); + &llr->maximum_arbitration_wait_timer_timeout); } - /* Disable link layer hang detection, rely on the OS timeout for I/O timeouts. */ - writel(0, &iphy->link_layer_registers->link_layer_hang_detection_timeout); + /* Disable link layer hang detection, rely on the OS timeout for + * I/O timeouts. + */ + writel(0, &llr->link_layer_hang_detection_timeout); /* We can exit the initial state to the stopped state */ sci_change_state(&iphy->sm, SCI_PHY_STOPPED); From afd13a1f2b05157c7621d87dfe89ea6ea9061bd8 Mon Sep 17 00:00:00 2001 From: Jeff Skirvin Date: Wed, 4 Jan 2012 01:32:39 -0800 Subject: [PATCH 300/820] [SCSI] isci: update afe (analog-front-end) recipe for C1 C1 silicon requires updates to the phy tuning recipe and also support for user provided cable selects (per-phy) for short, medium, and long cables. Default to 'short' awaiting support for selecting the cable via oem parameters. Reviewed-by: Jiangbi Liu Signed-off-by: Marcin Tomczak Signed-off-by: Jeff Skirvin Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 121 +++++++++++++++++++++++++-------- drivers/scsi/isci/host.h | 9 ++- drivers/scsi/isci/phy.c | 7 +- drivers/scsi/isci/probe_roms.c | 2 +- 4 files changed, 108 insertions(+), 31 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index 8e7de192cf6d..383bb6913087 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -1908,12 +1908,23 @@ void sci_controller_power_control_queue_remove(struct isci_host *ihost, ihost->power_control.requesters[iphy->phy_index] = NULL; } +static int is_long_cable(int phy, unsigned char selection_byte) +{ + return 0; +} + +static int is_medium_cable(int phy, unsigned char selection_byte) +{ + return 0; +} + #define AFE_REGISTER_WRITE_DELAY 10 static void sci_controller_afe_initialization(struct isci_host *ihost) { struct scu_afe_registers __iomem *afe = &ihost->scu_registers->afe; const struct sci_oem_params *oem = &ihost->oem_parameters; + unsigned char cable_selection_mask = 0; struct pci_dev *pdev = ihost->pdev; u32 afe_status; u32 phy_id; @@ -1922,11 +1933,11 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) writel(0x0081000f, &afe->afe_dfx_master_control0); udelay(AFE_REGISTER_WRITE_DELAY); - if (is_b0(pdev)) { + if (is_b0(pdev) || is_c0(pdev) || is_c1(pdev)) { /* PM Rx Equalization Save, PM SPhy Rx Acknowledgement * Timer, PM Stagger Timer */ - writel(0x0007BFFF, &afe->afe_pmsn_master_control2); + writel(0x0007FFFF, &afe->afe_pmsn_master_control2); udelay(AFE_REGISTER_WRITE_DELAY); } @@ -1935,14 +1946,23 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) writel(0x00005A00, &afe->afe_bias_control); else if (is_b0(pdev) || is_c0(pdev)) writel(0x00005F00, &afe->afe_bias_control); + else if (is_c1(pdev)) + writel(0x00005500, &afe->afe_bias_control); udelay(AFE_REGISTER_WRITE_DELAY); /* Enable PLL */ - if (is_b0(pdev) || is_c0(pdev)) - writel(0x80040A08, &afe->afe_pll_control0); - else + if (is_a2(pdev)) writel(0x80040908, &afe->afe_pll_control0); + else if (is_b0(pdev) || is_c0(pdev)) + writel(0x80040A08, &afe->afe_pll_control0); + else if (is_c1(pdev)) { + writel(0x80000B08, &afe->afe_pll_control0); + udelay(AFE_REGISTER_WRITE_DELAY); + writel(0x00000B08, &afe->afe_pll_control0); + udelay(AFE_REGISTER_WRITE_DELAY); + writel(0x80000B08, &afe->afe_pll_control0); + } udelay(AFE_REGISTER_WRITE_DELAY); @@ -1963,22 +1983,12 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) for (phy_id = 0; phy_id < SCI_MAX_PHYS; phy_id++) { struct scu_afe_transceiver *xcvr = &afe->scu_afe_xcvr[phy_id]; const struct sci_phy_oem_params *oem_phy = &oem->phys[phy_id]; + int cable_length_long = + is_long_cable(phy_id, cable_selection_mask); + int cable_length_medium = + is_medium_cable(phy_id, cable_selection_mask); - if (is_b0(pdev)) { - /* Configure transmitter SSC parameters */ - writel(0x00030000, &xcvr->afe_tx_ssc_control); - udelay(AFE_REGISTER_WRITE_DELAY); - } else if (is_c0(pdev)) { - /* Configure transmitter SSC parameters */ - writel(0x0003000, &xcvr->afe_tx_ssc_control); - udelay(AFE_REGISTER_WRITE_DELAY); - - /* All defaults, except the Receive Word - * Alignament/Comma Detect Enable....(0xe800) - */ - writel(0x00004500, &xcvr->afe_xcvr_control0); - udelay(AFE_REGISTER_WRITE_DELAY); - } else { + if (is_a2(pdev)) { /* All defaults, except the Receive Word * Alignament/Comma Detect Enable....(0xe800) */ @@ -1987,22 +1997,54 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) writel(0x0050100F, &xcvr->afe_xcvr_control1); udelay(AFE_REGISTER_WRITE_DELAY); + } else if (is_b0(pdev)) { + /* Configure transmitter SSC parameters */ + writel(0x00030000, &xcvr->afe_tx_ssc_control); + udelay(AFE_REGISTER_WRITE_DELAY); + } else if (is_c0(pdev)) { + /* Configure transmitter SSC parameters */ + writel(0x00010202, &xcvr->afe_tx_ssc_control); + udelay(AFE_REGISTER_WRITE_DELAY); + + /* All defaults, except the Receive Word + * Alignament/Comma Detect Enable....(0xe800) + */ + writel(0x00014500, &xcvr->afe_xcvr_control0); + udelay(AFE_REGISTER_WRITE_DELAY); + } else if (is_c1(pdev)) { + /* Configure transmitter SSC parameters */ + writel(0x00010202, &xcvr->afe_tx_ssc_control); + udelay(AFE_REGISTER_WRITE_DELAY); + + /* All defaults, except the Receive Word + * Alignament/Comma Detect Enable....(0xe800) + */ + writel(0x0001C500, &xcvr->afe_xcvr_control0); + udelay(AFE_REGISTER_WRITE_DELAY); } - /* Power up TX and RX out from power down (PWRDNTX and PWRDNRX) - * & increase TX int & ext bias 20%....(0xe85c) + /* Power up TX and RX out from power down (PWRDNTX and + * PWRDNRX) & increase TX int & ext bias 20%....(0xe85c) */ if (is_a2(pdev)) writel(0x000003F0, &xcvr->afe_channel_control); else if (is_b0(pdev)) { - /* Power down TX and RX (PWRDNTX and PWRDNRX) */ writel(0x000003D7, &xcvr->afe_channel_control); udelay(AFE_REGISTER_WRITE_DELAY); + writel(0x000003D4, &xcvr->afe_channel_control); - } else { + } else if (is_c0(pdev)) { writel(0x000001E7, &xcvr->afe_channel_control); udelay(AFE_REGISTER_WRITE_DELAY); + writel(0x000001E4, &xcvr->afe_channel_control); + } else if (is_c1(pdev)) { + writel(cable_length_long ? 0x000002F7 : 0x000001F7, + &xcvr->afe_channel_control); + udelay(AFE_REGISTER_WRITE_DELAY); + + writel(cable_length_long ? 0x000002F4 : 0x000001F4, + &xcvr->afe_channel_control); } udelay(AFE_REGISTER_WRITE_DELAY); @@ -2012,7 +2054,16 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) udelay(AFE_REGISTER_WRITE_DELAY); } - writel(0x00004100, &xcvr->afe_xcvr_control0); + if (is_a2(pdev) || is_b0(pdev)) + /* RDPI=0x0(RX Power On), RXOOBDETPDNC=0x0, + * TPD=0x0(TX Power On), RDD=0x0(RX Detect + * Enabled) ....(0xe800) + */ + writel(0x00004100, &xcvr->afe_xcvr_control0); + else if (is_c0(pdev)) + writel(0x00014100, &xcvr->afe_xcvr_control0); + else if (is_c1(pdev)) + writel(0x0001C100, &xcvr->afe_xcvr_control0); udelay(AFE_REGISTER_WRITE_DELAY); /* Leave DFE/FFE on */ @@ -2023,13 +2074,29 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) udelay(AFE_REGISTER_WRITE_DELAY); /* Enable TX equalization (0xe824) */ writel(0x00040000, &xcvr->afe_tx_control); - } else { - writel(0x0140DF0F, &xcvr->afe_rx_ssc_control1); + } else if (is_c0(pdev)) { + writel(0x01400C0F, &xcvr->afe_rx_ssc_control1); udelay(AFE_REGISTER_WRITE_DELAY); writel(0x3F6F103F, &xcvr->afe_rx_ssc_control0); udelay(AFE_REGISTER_WRITE_DELAY); + /* Enable TX equalization (0xe824) */ + writel(0x00040000, &xcvr->afe_tx_control); + } else if (is_c1(pdev)) { + writel(cable_length_long ? 0x01500C0C : + cable_length_medium ? 0x01400C0D : 0x02400C0D, + &xcvr->afe_xcvr_control1); + udelay(AFE_REGISTER_WRITE_DELAY); + + writel(0x000003E0, &xcvr->afe_dfx_rx_control1); + udelay(AFE_REGISTER_WRITE_DELAY); + + writel(cable_length_long ? 0x33091C1F : + cable_length_medium ? 0x3315181F : 0x2B17161F, + &xcvr->afe_rx_ssc_control0); + udelay(AFE_REGISTER_WRITE_DELAY); + /* Enable TX equalization (0xe824) */ writel(0x00040000, &xcvr->afe_tx_control); } diff --git a/drivers/scsi/isci/host.h b/drivers/scsi/isci/host.h index 646051afd3cb..4573075a6b97 100644 --- a/drivers/scsi/isci/host.h +++ b/drivers/scsi/isci/host.h @@ -435,7 +435,14 @@ static inline bool is_b0(struct pci_dev *pdev) static inline bool is_c0(struct pci_dev *pdev) { - if (pdev->revision >= 5) + if (pdev->revision == 5) + return true; + return false; +} + +static inline bool is_c1(struct pci_dev *pdev) +{ + if (pdev->revision >= 6) return true; return false; } diff --git a/drivers/scsi/isci/phy.c b/drivers/scsi/isci/phy.c index 0ae0990b39dd..c650d3003c22 100644 --- a/drivers/scsi/isci/phy.c +++ b/drivers/scsi/isci/phy.c @@ -186,8 +186,11 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, writel(clksm_value, &llr->clock_skew_management); - /* @todo Provide a way to write this register correctly */ - writel(0x02108421, &llr->afe_lookup_table_control); + if (is_c0(ihost->pdev) || is_c1(ihost->pdev)) { + writel(0x04210400, &llr->afe_lookup_table_control); + writel(0x020A7C05, &llr->sas_primitive_timeout); + } else + writel(0x02108421, &llr->afe_lookup_table_control); llctl = SCU_SAS_LLCTL_GEN_VAL(NO_OUTBOUND_TASK_TIMEOUT, (u8)ihost->user_parameters.no_outbound_task_timeout); diff --git a/drivers/scsi/isci/probe_roms.c b/drivers/scsi/isci/probe_roms.c index b5f4341de243..9b8117b9d756 100644 --- a/drivers/scsi/isci/probe_roms.c +++ b/drivers/scsi/isci/probe_roms.c @@ -147,7 +147,7 @@ struct isci_orom *isci_request_firmware(struct pci_dev *pdev, const struct firmw memcpy(orom, fw->data, fw->size); - if (is_c0(pdev)) + if (is_c0(pdev) || is_c1(pdev)) goto out; /* From 594e566ae5985e0cc3185ac21509a86e90aad577 Mon Sep 17 00:00:00 2001 From: Dave Jiang Date: Wed, 4 Jan 2012 01:32:44 -0800 Subject: [PATCH 301/820] [SCSI] isci: oem parameter format v1.1 (ssc select) v1.1 allows finer grained tuning of the SSC (spread-spectrum-clocking) settings for SAS and SATA. See notes in probe_roms.h Signed-off-by: Dave Jiang Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 47 ++++++++++++++++++++++++++++-- drivers/scsi/isci/init.c | 3 +- drivers/scsi/isci/phy.c | 51 +++++++++++++++++++++++++++++++- drivers/scsi/isci/probe_roms.h | 53 ++++++++++++++++++++++++++++++++-- 4 files changed, 148 insertions(+), 6 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index 383bb6913087..ed1441c89577 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -1759,7 +1759,7 @@ static enum sci_status sci_controller_construct(struct isci_host *ihost, return sci_controller_reset(ihost); } -int sci_oem_parameters_validate(struct sci_oem_params *oem) +int sci_oem_parameters_validate(struct sci_oem_params *oem, u8 version) { int i; @@ -1791,18 +1791,61 @@ int sci_oem_parameters_validate(struct sci_oem_params *oem) oem->controller.max_concurr_spin_up < 1) return -EINVAL; + if (oem->controller.do_enable_ssc) { + if (version < ISCI_ROM_VER_1_1 && oem->controller.do_enable_ssc != 1) + return -EINVAL; + + if (version >= ISCI_ROM_VER_1_1) { + u8 test = oem->controller.ssc_sata_tx_spread_level; + + switch (test) { + case 0: + case 2: + case 3: + case 6: + case 7: + break; + default: + return -EINVAL; + } + + test = oem->controller.ssc_sas_tx_spread_level; + if (oem->controller.ssc_sas_tx_type == 0) { + switch (test) { + case 0: + case 2: + case 3: + break; + default: + return -EINVAL; + } + } else if (oem->controller.ssc_sas_tx_type == 1) { + switch (test) { + case 0: + case 3: + case 6: + break; + default: + return -EINVAL; + } + } + } + } + return 0; } static enum sci_status sci_oem_parameters_set(struct isci_host *ihost) { u32 state = ihost->sm.current_state_id; + struct isci_pci_info *pci_info = to_pci_info(ihost->pdev); if (state == SCIC_RESET || state == SCIC_INITIALIZING || state == SCIC_INITIALIZED) { - if (sci_oem_parameters_validate(&ihost->oem_parameters)) + if (sci_oem_parameters_validate(&ihost->oem_parameters, + pci_info->orom->hdr.version)) return SCI_FAILURE_INVALID_PARAMETER_VALUE; return SCI_SUCCESS; diff --git a/drivers/scsi/isci/init.c b/drivers/scsi/isci/init.c index a97edabcb85a..8a34fd92e42e 100644 --- a/drivers/scsi/isci/init.c +++ b/drivers/scsi/isci/init.c @@ -466,7 +466,8 @@ static int __devinit isci_pci_probe(struct pci_dev *pdev, const struct pci_devic orom = isci_request_oprom(pdev); for (i = 0; orom && i < ARRAY_SIZE(orom->ctrl); i++) { - if (sci_oem_parameters_validate(&orom->ctrl[i])) { + if (sci_oem_parameters_validate(&orom->ctrl[i], + orom->hdr.version)) { dev_warn(&pdev->dev, "[%d]: invalid oem parameters detected, falling back to firmware\n", i); devm_kfree(&pdev->dev, orom); diff --git a/drivers/scsi/isci/phy.c b/drivers/scsi/isci/phy.c index c650d3003c22..61000cde84c7 100644 --- a/drivers/scsi/isci/phy.c +++ b/drivers/scsi/isci/phy.c @@ -144,10 +144,59 @@ sci_phy_link_layer_initialization(struct isci_phy *iphy, phy_cap.gen3_no_ssc = 1; phy_cap.gen2_no_ssc = 1; phy_cap.gen1_no_ssc = 1; - if (ihost->oem_parameters.controller.do_enable_ssc == true) { + if (ihost->oem_parameters.controller.do_enable_ssc) { + struct scu_afe_registers __iomem *afe = &ihost->scu_registers->afe; + struct scu_afe_transceiver *xcvr = &afe->scu_afe_xcvr[phy_idx]; + struct isci_pci_info *pci_info = to_pci_info(ihost->pdev); + bool en_sas = false; + bool en_sata = false; + u32 sas_type = 0; + u32 sata_spread = 0x2; + u32 sas_spread = 0x2; + phy_cap.gen3_ssc = 1; phy_cap.gen2_ssc = 1; phy_cap.gen1_ssc = 1; + + if (pci_info->orom->hdr.version < ISCI_ROM_VER_1_1) + en_sas = en_sata = true; + else { + sata_spread = ihost->oem_parameters.controller.ssc_sata_tx_spread_level; + sas_spread = ihost->oem_parameters.controller.ssc_sas_tx_spread_level; + + if (sata_spread) + en_sata = true; + + if (sas_spread) { + en_sas = true; + sas_type = ihost->oem_parameters.controller.ssc_sas_tx_type; + } + + } + + if (en_sas) { + u32 reg; + + reg = readl(&xcvr->afe_xcvr_control0); + reg |= (0x00100000 | (sas_type << 19)); + writel(reg, &xcvr->afe_xcvr_control0); + + reg = readl(&xcvr->afe_tx_ssc_control); + reg |= sas_spread << 8; + writel(reg, &xcvr->afe_tx_ssc_control); + } + + if (en_sata) { + u32 reg; + + reg = readl(&xcvr->afe_tx_ssc_control); + reg |= sata_spread; + writel(reg, &xcvr->afe_tx_ssc_control); + + reg = readl(&llr->stp_control); + reg |= 1 << 12; + writel(reg, &llr->stp_control); + } } /* The SAS specification indicates that the phy_capabilities that diff --git a/drivers/scsi/isci/probe_roms.h b/drivers/scsi/isci/probe_roms.h index 2c75248ca326..42dd05414f3b 100644 --- a/drivers/scsi/isci/probe_roms.h +++ b/drivers/scsi/isci/probe_roms.h @@ -152,7 +152,7 @@ struct sci_user_parameters { #define MAX_CONCURRENT_DEVICE_SPIN_UP_COUNT 4 struct sci_oem_params; -int sci_oem_parameters_validate(struct sci_oem_params *oem); +int sci_oem_parameters_validate(struct sci_oem_params *oem, u8 version); struct isci_orom; struct isci_orom *isci_request_oprom(struct pci_dev *pdev); @@ -191,6 +191,10 @@ struct isci_oem_hdr { 0x1a, 0x04, 0xc6) #define ISCI_EFI_VAR_NAME "RstScuO" +#define ISCI_ROM_VER_1_0 0x10 +#define ISCI_ROM_VER_1_1 0x11 +#define ISCI_ROM_VER_LATEST ISCI_ROM_VER_1_1 + /* Allowed PORT configuration modes APC Automatic PORT configuration mode is * defined by the OEM configuration parameters providing no PHY_MASK parameters * for any PORT. i.e. There are no phys assigned to any of the ports at start. @@ -220,7 +224,52 @@ struct sci_oem_params { struct { uint8_t mode_type; uint8_t max_concurr_spin_up; - uint8_t do_enable_ssc; + /* + * This bitfield indicates the OEM's desired default Tx + * Spread Spectrum Clocking (SSC) settings for SATA and SAS. + * NOTE: Default SSC Modulation Frequency is 31.5KHz. + */ + union { + struct { + /* + * NOTE: Max spread for SATA is +0 / -5000 PPM. + * Down-spreading SSC (only method allowed for SATA): + * SATA SSC Tx Disabled = 0x0 + * SATA SSC Tx at +0 / -1419 PPM Spread = 0x2 + * SATA SSC Tx at +0 / -2129 PPM Spread = 0x3 + * SATA SSC Tx at +0 / -4257 PPM Spread = 0x6 + * SATA SSC Tx at +0 / -4967 PPM Spread = 0x7 + */ + uint8_t ssc_sata_tx_spread_level:4; + /* + * SAS SSC Tx Disabled = 0x0 + * + * NOTE: Max spread for SAS down-spreading +0 / + * -2300 PPM + * Down-spreading SSC: + * SAS SSC Tx at +0 / -1419 PPM Spread = 0x2 + * SAS SSC Tx at +0 / -2129 PPM Spread = 0x3 + * + * NOTE: Max spread for SAS center-spreading +2300 / + * -2300 PPM + * Center-spreading SSC: + * SAS SSC Tx at +1064 / -1064 PPM Spread = 0x3 + * SAS SSC Tx at +2129 / -2129 PPM Spread = 0x6 + */ + uint8_t ssc_sas_tx_spread_level:3; + /* + * NOTE: Refer to the SSC section of the SAS 2.x + * Specification for proper setting of this field. + * For standard SAS Initiator SAS PHY operation it + * should be 0 for Down-spreading. + * SAS SSC Tx spread type: + * Down-spreading SSC = 0 + * Center-spreading SSC = 1 + */ + uint8_t ssc_sas_tx_type:1; + }; + uint8_t do_enable_ssc; + }; uint8_t reserved; } controller; From 9fee607f0b29adabd72265a847b8e421dff10d66 Mon Sep 17 00:00:00 2001 From: Jeff Skirvin Date: Wed, 4 Jan 2012 01:32:49 -0800 Subject: [PATCH 302/820] [SCSI] isci: oem parameter format v1.3 (cable select) v1.3 allows the attenuation of the attached cables to be specified to the driver in terms of 'short', 'medium', and 'long' (see probe_roms.h). These settings (per phy) are retrieved from the platform oem-parameters (BIOS rom) or via a module parameter override. Reviewed-by: Jiangbi Liu Signed-off-by: Jeff Skirvin Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 43 +++++++++++++++++++++++++++++++--- drivers/scsi/isci/host.h | 18 ++++++++++++++ drivers/scsi/isci/init.c | 16 +++++++++++++ drivers/scsi/isci/isci.h | 1 + drivers/scsi/isci/probe_roms.h | 38 ++++++++++++++++++++++++++++-- 5 files changed, 111 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index ed1441c89577..9a52b984620f 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -1666,6 +1666,9 @@ static void sci_controller_set_default_config_parameters(struct isci_host *ihost /* Default to no SSC operation. */ ihost->oem_parameters.controller.do_enable_ssc = false; + /* Default to short cables on all phys. */ + ihost->oem_parameters.controller.cable_selection_mask = 0; + /* Initialize all of the port parameter information to narrow ports. */ for (index = 0; index < SCI_MAX_PORTS; index++) { ihost->oem_parameters.ports[index].phy_mask = 0; @@ -1953,12 +1956,46 @@ void sci_controller_power_control_queue_remove(struct isci_host *ihost, static int is_long_cable(int phy, unsigned char selection_byte) { - return 0; + return !!(selection_byte & (1 << phy)); } static int is_medium_cable(int phy, unsigned char selection_byte) { - return 0; + return !!(selection_byte & (1 << (phy + 4))); +} + +static enum cable_selections decode_selection_byte( + int phy, + unsigned char selection_byte) +{ + return ((selection_byte & (1 << phy)) ? 1 : 0) + + (selection_byte & (1 << (phy + 4)) ? 2 : 0); +} + +static unsigned char *to_cable_select(struct isci_host *ihost) +{ + if (is_cable_select_overridden()) + return ((unsigned char *)&cable_selection_override) + + ihost->id; + else + return &ihost->oem_parameters.controller.cable_selection_mask; +} + +enum cable_selections decode_cable_selection(struct isci_host *ihost, int phy) +{ + return decode_selection_byte(phy, *to_cable_select(ihost)); +} + +char *lookup_cable_names(enum cable_selections selection) +{ + static char *cable_names[] = { + [short_cable] = "short", + [long_cable] = "long", + [medium_cable] = "medium", + [undefined_cable] = "" /* bit 0==1 */ + }; + return (selection <= undefined_cable) ? cable_names[selection] + : cable_names[undefined_cable]; } #define AFE_REGISTER_WRITE_DELAY 10 @@ -1967,10 +2004,10 @@ static void sci_controller_afe_initialization(struct isci_host *ihost) { struct scu_afe_registers __iomem *afe = &ihost->scu_registers->afe; const struct sci_oem_params *oem = &ihost->oem_parameters; - unsigned char cable_selection_mask = 0; struct pci_dev *pdev = ihost->pdev; u32 afe_status; u32 phy_id; + unsigned char cable_selection_mask = *to_cable_select(ihost); /* Clear DFX Status registers */ writel(0x0081000f, &afe->afe_dfx_master_control0); diff --git a/drivers/scsi/isci/host.h b/drivers/scsi/isci/host.h index 4573075a6b97..5477f0fa8233 100644 --- a/drivers/scsi/isci/host.h +++ b/drivers/scsi/isci/host.h @@ -447,6 +447,24 @@ static inline bool is_c1(struct pci_dev *pdev) return false; } +enum cable_selections { + short_cable = 0, + long_cable = 1, + medium_cable = 2, + undefined_cable = 3 +}; + +#define CABLE_OVERRIDE_DISABLED (0x10000) + +static inline int is_cable_select_overridden(void) +{ + return cable_selection_override < CABLE_OVERRIDE_DISABLED; +} + +enum cable_selections decode_cable_selection(struct isci_host *ihost, int phy); +void validate_cable_selections(struct isci_host *ihost); +char *lookup_cable_names(enum cable_selections); + /* set hw control for 'activity', even though active enclosures seem to drive * the activity led on their own. Skip setting FSENG control on 'status' due * to unexpected operation and 'error' due to not being a supported automatic diff --git a/drivers/scsi/isci/init.c b/drivers/scsi/isci/init.c index 8a34fd92e42e..1047108ba4de 100644 --- a/drivers/scsi/isci/init.c +++ b/drivers/scsi/isci/init.c @@ -122,6 +122,14 @@ unsigned char max_concurr_spinup; module_param(max_concurr_spinup, byte, 0); MODULE_PARM_DESC(max_concurr_spinup, "Max concurrent device spinup"); +uint cable_selection_override = CABLE_OVERRIDE_DISABLED; +module_param(cable_selection_override, uint, 0); + +MODULE_PARM_DESC(cable_selection_override, + "This field indicates length of the SAS/SATA cable between " + "host and device. If any bits > 15 are set (default) " + "indicates \"use platform defaults\""); + static ssize_t isci_show_id(struct device *dev, struct device_attribute *attr, char *buf) { struct Scsi_Host *shost = container_of(dev, typeof(*shost), shost_dev); @@ -412,6 +420,14 @@ static struct isci_host *isci_host_alloc(struct pci_dev *pdev, int id) return NULL; isci_host->shost = shost; + dev_info(&pdev->dev, "%sSCU controller %d: phy 3-0 cables: " + "{%s, %s, %s, %s}\n", + (is_cable_select_overridden() ? "* " : ""), isci_host->id, + lookup_cable_names(decode_cable_selection(isci_host, 3)), + lookup_cable_names(decode_cable_selection(isci_host, 2)), + lookup_cable_names(decode_cable_selection(isci_host, 1)), + lookup_cable_names(decode_cable_selection(isci_host, 0))); + err = isci_host_init(isci_host); if (err) goto err_shost; diff --git a/drivers/scsi/isci/isci.h b/drivers/scsi/isci/isci.h index 8efeb6b08321..234ab46fce33 100644 --- a/drivers/scsi/isci/isci.h +++ b/drivers/scsi/isci/isci.h @@ -480,6 +480,7 @@ extern u16 ssp_inactive_to; extern u16 stp_inactive_to; extern unsigned char phy_gen; extern unsigned char max_concurr_spinup; +extern uint cable_selection_override; irqreturn_t isci_msix_isr(int vec, void *data); irqreturn_t isci_intx_isr(int vec, void *data); diff --git a/drivers/scsi/isci/probe_roms.h b/drivers/scsi/isci/probe_roms.h index 42dd05414f3b..bb0e9d4d97c9 100644 --- a/drivers/scsi/isci/probe_roms.h +++ b/drivers/scsi/isci/probe_roms.h @@ -193,7 +193,8 @@ struct isci_oem_hdr { #define ISCI_ROM_VER_1_0 0x10 #define ISCI_ROM_VER_1_1 0x11 -#define ISCI_ROM_VER_LATEST ISCI_ROM_VER_1_1 +#define ISCI_ROM_VER_1_3 0x13 +#define ISCI_ROM_VER_LATEST ISCI_ROM_VER_1_3 /* Allowed PORT configuration modes APC Automatic PORT configuration mode is * defined by the OEM configuration parameters providing no PHY_MASK parameters @@ -270,7 +271,40 @@ struct sci_oem_params { }; uint8_t do_enable_ssc; }; - uint8_t reserved; + /* + * This field indicates length of the SAS/SATA cable between + * host and device. + * This field is used make relationship between analog + * parameters of the phy in the silicon and length of the cable. + * Supported cable attenuation levels: + * "short"- up to 3m, "medium"-3m to 6m, and "long"- more than + * 6m. + * + * This is bit mask field: + * + * BIT: (MSB) 7 6 5 4 + * ASSIGNMENT: - Medium cable + * length assignment + * BIT: 3 2 1 0 (LSB) + * ASSIGNMENT: - Long cable length + * assignment + * + * BITS 7-4 are set when the cable length is assigned to medium + * BITS 3-0 are set when the cable length is assigned to long + * + * The BIT positions are clear when the cable length is + * assigned to short. + * + * Setting the bits for both long and medium cable length is + * undefined. + * + * A value of 0x84 would assign + * phy3 - medium + * phy2 - long + * phy1 - short + * phy0 - short + */ + uint8_t cable_selection_mask; } controller; struct { From 6024d38b86ecc7943e398ef01b0dc2765870d444 Mon Sep 17 00:00:00 2001 From: Marcin Tomczak Date: Wed, 4 Jan 2012 01:32:54 -0800 Subject: [PATCH 303/820] [SCSI] isci: performance-fix, shorten default "no outbound task" timeout "No task timeout timer reduced from 20 to 2 This timer controls how long the SCU hardware will hold open the TX side of the connection before sending a DONE. The timer allows the hardware to attempt to optimize the DONE/CLOSE behavior to allow for new COMMAND IU to be posted. In practice closing the connection quicker is better." Signed-off-by: Marcin Tomczak Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 2 +- drivers/scsi/isci/init.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index 9a52b984620f..670ecb456d48 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -1697,7 +1697,7 @@ static void sci_controller_set_default_config_parameters(struct isci_host *ihost ihost->user_parameters.ssp_inactivity_timeout = 5; ihost->user_parameters.stp_max_occupancy_timeout = 5; ihost->user_parameters.ssp_max_occupancy_timeout = 20; - ihost->user_parameters.no_outbound_task_timeout = 20; + ihost->user_parameters.no_outbound_task_timeout = 2; } static void controller_timeout(unsigned long data) diff --git a/drivers/scsi/isci/init.c b/drivers/scsi/isci/init.c index 1047108ba4de..3c4ddad97053 100644 --- a/drivers/scsi/isci/init.c +++ b/drivers/scsi/isci/init.c @@ -94,7 +94,7 @@ MODULE_DEVICE_TABLE(pci, isci_id_table); /* linux isci specific settings */ -unsigned char no_outbound_task_to = 20; +unsigned char no_outbound_task_to = 2; module_param(no_outbound_task_to, byte, 0); MODULE_PARM_DESC(no_outbound_task_to, "No Outbound Task Timeout (1us incr)"); From be168a3b89b43b8bb4ef71066e91439d400e20a6 Mon Sep 17 00:00:00 2001 From: Jeff Skirvin Date: Wed, 4 Jan 2012 01:33:00 -0800 Subject: [PATCH 304/820] [SCSI] isci: link speeds default to gen 2 Gen-3 operation is marginal, default to gen-2 for now. Signed-off-by: Jeff Skirvin Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 5 +++-- drivers/scsi/isci/init.c | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index 670ecb456d48..7e4d709dc201 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -1676,8 +1676,9 @@ static void sci_controller_set_default_config_parameters(struct isci_host *ihost /* Initialize all of the phy parameter information. */ for (index = 0; index < SCI_MAX_PHYS; index++) { - /* Default to 6G (i.e. Gen 3) for now. */ - ihost->user_parameters.phys[index].max_speed_generation = 3; + /* Default to 3G (i.e. Gen 2). */ + ihost->user_parameters.phys[index].max_speed_generation = + SCIC_SDS_PARM_GEN2_SPEED; /* the frequencies cannot be 0 */ ihost->user_parameters.phys[index].align_insertion_frequency = 0x7f; diff --git a/drivers/scsi/isci/init.c b/drivers/scsi/isci/init.c index 3c4ddad97053..2bcfb40ae9ad 100644 --- a/drivers/scsi/isci/init.c +++ b/drivers/scsi/isci/init.c @@ -114,7 +114,7 @@ u16 stp_inactive_to = 5; module_param(stp_inactive_to, ushort, 0); MODULE_PARM_DESC(stp_inactive_to, "STP inactivity timeout (100us incr)"); -unsigned char phy_gen = 3; +unsigned char phy_gen = SCIC_SDS_PARM_GEN2_SPEED; module_param(phy_gen, byte, 0); MODULE_PARM_DESC(phy_gen, "PHY generation (1: 1.5Gbps 2: 3.0Gbps 3: 6.0Gbps)"); From 27234ab44f97d85bab062a9d18aaff99addd267d Mon Sep 17 00:00:00 2001 From: Maciej Trela Date: Wed, 4 Jan 2012 01:33:05 -0800 Subject: [PATCH 305/820] [SCSI] isci: remove unused 'isci_tmf->device' field As the field was never set, isci_print_tmf() using 'isci_tmf->device' sometimes causes a kernel crash if the dev_dbg() statement is enabled. Remove the unused field both from isci_tmf struct definition and from isci_print_tmf() Signed-off-by: Maciej Trela Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/task.c | 2 +- drivers/scsi/isci/task.h | 7 +++---- 2 files changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/isci/task.c b/drivers/scsi/isci/task.c index 66ad3dc89498..f5a3f7d2bdab 100644 --- a/drivers/scsi/isci/task.c +++ b/drivers/scsi/isci/task.c @@ -496,7 +496,7 @@ static int isci_task_execute_tmf(struct isci_host *ihost, } } - isci_print_tmf(tmf); + isci_print_tmf(ihost, tmf); if (tmf->status == SCI_SUCCESS) ret = TMF_RESP_FUNC_COMPLETE; diff --git a/drivers/scsi/isci/task.h b/drivers/scsi/isci/task.h index bc78c0a41d5c..1b27b3797c6c 100644 --- a/drivers/scsi/isci/task.h +++ b/drivers/scsi/isci/task.h @@ -106,7 +106,6 @@ struct isci_tmf { } resp; unsigned char lun[8]; u16 io_tag; - struct isci_remote_device *device; enum isci_tmf_function_codes tmf_code; int status; @@ -120,10 +119,10 @@ struct isci_tmf { }; -static inline void isci_print_tmf(struct isci_tmf *tmf) +static inline void isci_print_tmf(struct isci_host *ihost, struct isci_tmf *tmf) { if (SAS_PROTOCOL_SATA == tmf->proto) - dev_dbg(&tmf->device->isci_port->isci_host->pdev->dev, + dev_dbg(&ihost->pdev->dev, "%s: status = %x\n" "tmf->resp.d2h_fis.status = %x\n" "tmf->resp.d2h_fis.error = %x\n", @@ -132,7 +131,7 @@ static inline void isci_print_tmf(struct isci_tmf *tmf) tmf->resp.d2h_fis.status, tmf->resp.d2h_fis.error); else - dev_dbg(&tmf->device->isci_port->isci_host->pdev->dev, + dev_dbg(&ihost->pdev->dev, "%s: status = %x\n" "tmf->resp.resp_iu.data_present = %x\n" "tmf->resp.resp_iu.status = %x\n" From d3fd2e2bc29b1f0979352994f9035359003461c3 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Wed, 4 Jan 2012 01:33:10 -0800 Subject: [PATCH 306/820] [SCSI] isci: update version to 1.1 Bump the version now that the driver has atapi support and the initial round of hotplug fixes. The EXPERIMENTAL tag should have been removed a while back. While we're here also kill the "select SCSI_SAS_HOST_SMP" as the build error was separately fixed by commit d962480e "[SCSI] libsas: fix try_test_sas_gpio_gp_bit() build error". Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/Kconfig | 5 ----- drivers/scsi/isci/init.c | 2 +- 2 files changed, 1 insertion(+), 6 deletions(-) diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 06ea3bcfdd2a..16570aa84aac 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -830,16 +830,11 @@ config SCSI_ISCI tristate "Intel(R) C600 Series Chipset SAS Controller" depends on PCI && SCSI depends on X86 - # (temporary): known alpha quality driver - depends on EXPERIMENTAL select SCSI_SAS_LIBSAS - select SCSI_SAS_HOST_SMP ---help--- This driver supports the 6Gb/s SAS capabilities of the storage control unit found in the Intel(R) C600 series chipset. - The experimental tag will be removed after the driver exits alpha - config SCSI_GENERIC_NCR5380 tristate "Generic NCR5380/53c400 SCSI PIO support" depends on ISA && SCSI diff --git a/drivers/scsi/isci/init.c b/drivers/scsi/isci/init.c index 2bcfb40ae9ad..17c4c2c89c2e 100644 --- a/drivers/scsi/isci/init.c +++ b/drivers/scsi/isci/init.c @@ -65,7 +65,7 @@ #include "probe_roms.h" #define MAJ 1 -#define MIN 0 +#define MIN 1 #define BUILD 0 #define DRV_VERSION __stringify(MAJ) "." __stringify(MIN) "." \ __stringify(BUILD) From d4ec1cf61fb081a9dde0c0e0b2d0201f4005f937 Mon Sep 17 00:00:00 2001 From: Marcin Tomczak Date: Wed, 4 Jan 2012 01:33:15 -0800 Subject: [PATCH 307/820] [SCSI] isci: Fix IO fails when pull cable from phy in x4 wideport in MPC mode. Failure seen pulling a cable from a x4 port configured in manual port configuration (MPC) mode (MPC mode is set by the the OEM paramaters provided by the platform or isci_firmware.bin). While IO running to devices behind and expander, plugging out the cable from phy is causing IO failures and IO drops on disks and never recover. It happens because during link up/down the phy were being taken out of the port. Fix: during link down the phy is kept in the same logical port. Signed-off-by: Marcin Tomczak Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/port.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/scsi/isci/port.c b/drivers/scsi/isci/port.c index ac7f27749f97..4cb071d825bb 100644 --- a/drivers/scsi/isci/port.c +++ b/drivers/scsi/isci/port.c @@ -674,9 +674,13 @@ void sci_port_deactivate_phy(struct isci_port *iport, struct isci_phy *iphy, iphy->max_negotiated_speed = SAS_LINK_RATE_UNKNOWN; - /* Re-assign the phy back to the LP as if it were a narrow port */ - writel(iphy->phy_index, - &iport->port_pe_configuration_register[iphy->phy_index]); + /* Re-assign the phy back to the LP as if it were a narrow port for APC + * mode. For MPC mode, the phy will remain in the port. + */ + if (iport->owning_controller->oem_parameters.controller.mode_type == + SCIC_PORT_AUTOMATIC_CONFIGURATION_MODE) + writel(iphy->phy_index, + &iport->port_pe_configuration_register[iphy->phy_index]); if (do_notify_user == true) isci_port_link_down(ihost, iphy, iport); From 7e629841b8f8f7bc84d74de994d748b0a3282225 Mon Sep 17 00:00:00 2001 From: Bartek Nowakowski Date: Wed, 4 Jan 2012 01:33:20 -0800 Subject: [PATCH 308/820] [SCSI] isci: enable wide port targets Arrange for task_contexts prepared for the wide targets to account for all the attached phys in the port. Signed-off-by: Bartek Nowakowski Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/port.c | 2 +- drivers/scsi/isci/port.h | 4 ++++ drivers/scsi/isci/remote_device.c | 10 +++++++--- 3 files changed, 12 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/isci/port.c b/drivers/scsi/isci/port.c index 4cb071d825bb..49e8a72d1c5b 100644 --- a/drivers/scsi/isci/port.c +++ b/drivers/scsi/isci/port.c @@ -114,7 +114,7 @@ static u32 sci_port_get_phys(struct isci_port *iport) * value is returned if the specified port is not valid. When this value is * returned, no data is copied to the properties output parameter. */ -static enum sci_status sci_port_get_properties(struct isci_port *iport, +enum sci_status sci_port_get_properties(struct isci_port *iport, struct sci_port_properties *prop) { if (!iport || iport->logical_port_index == SCIC_SDS_DUMMY_PORT) diff --git a/drivers/scsi/isci/port.h b/drivers/scsi/isci/port.h index cb5ffbc38603..b3a7f12803f5 100644 --- a/drivers/scsi/isci/port.h +++ b/drivers/scsi/isci/port.h @@ -250,6 +250,10 @@ bool sci_port_link_detected( struct isci_port *iport, struct isci_phy *iphy); +enum sci_status sci_port_get_properties( + struct isci_port *iport, + struct sci_port_properties *prop); + enum sci_status sci_port_link_up(struct isci_port *iport, struct isci_phy *iphy); enum sci_status sci_port_link_down(struct isci_port *iport, diff --git a/drivers/scsi/isci/remote_device.c b/drivers/scsi/isci/remote_device.c index b207cd3b15a0..dd74b6ceeb82 100644 --- a/drivers/scsi/isci/remote_device.c +++ b/drivers/scsi/isci/remote_device.c @@ -53,6 +53,7 @@ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ #include +#include #include "isci.h" #include "port.h" #include "remote_device.h" @@ -1101,6 +1102,7 @@ static enum sci_status sci_remote_device_da_construct(struct isci_port *iport, struct isci_remote_device *idev) { enum sci_status status; + struct sci_port_properties properties; struct domain_device *dev = idev->domain_dev; sci_remote_device_construct(iport, idev); @@ -1110,6 +1112,11 @@ static enum sci_status sci_remote_device_da_construct(struct isci_port *iport, * entries will be needed to store the remote node. */ idev->is_direct_attached = true; + + sci_port_get_properties(iport, &properties); + /* Get accurate port width from port's phy mask for a DA device. */ + idev->device_port_width = hweight32(properties.phy_mask); + status = sci_controller_allocate_remote_node_context(iport->owning_controller, idev, &idev->rnc.remote_node_index); @@ -1125,9 +1132,6 @@ static enum sci_status sci_remote_device_da_construct(struct isci_port *iport, idev->connection_rate = sci_port_get_max_allowed_speed(iport); - /* / @todo Should I assign the port width by reading all of the phys on the port? */ - idev->device_port_width = 1; - return SCI_SUCCESS; } From 472d4d2cfbc169f3868a5f63ce727a482a2fd487 Mon Sep 17 00:00:00 2001 From: Marcin Tomczak Date: Wed, 4 Jan 2012 01:33:26 -0800 Subject: [PATCH 309/820] [SCSI] isci: allow more time for wide port targets When hot insert the wide port device through the mini-sas port, the first IOs (Report Luns or Inquiry) may fail due to the device trying to open to a SCU phy that is still in suspended state. This IO failure causes the wide port device stuck in UPDATING_PORT_WIDTH state. Signed-off-by: Marcin Tomczak Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/port_config.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/isci/port_config.c b/drivers/scsi/isci/port_config.c index 38a99d281141..37d15d3165da 100644 --- a/drivers/scsi/isci/port_config.c +++ b/drivers/scsi/isci/port_config.c @@ -57,7 +57,7 @@ #define SCIC_SDS_MPC_RECONFIGURATION_TIMEOUT (10) #define SCIC_SDS_APC_RECONFIGURATION_TIMEOUT (10) -#define SCIC_SDS_APC_WAIT_LINK_UP_NOTIFICATION (100) +#define SCIC_SDS_APC_WAIT_LINK_UP_NOTIFICATION (250) enum SCIC_SDS_APC_ACTIVITY { SCIC_SDS_APC_SKIP_PHY, From be778341812dc75b1c515fab6ebd39c0daf1e2bc Mon Sep 17 00:00:00 2001 From: Marcin Tomczak Date: Wed, 4 Jan 2012 01:33:31 -0800 Subject: [PATCH 310/820] [SCSI] isci: fix io failures while wide port links are coming up When the first phy of a wide port comes up, don't report the port ready yet, always wait for 250 miliseconds then config the port with all phys added to the port. So that we can avoid reporting wide port device too early to kernel, which caused the first IOs (report luns, inquirys) failed due to not all the phys are configured into its port. Changes also made that the phys in a wide port don't need to go through half second wait time for consuming power. Signed-off-by: Marcin Tomczak Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/host.c | 59 ++++++++++++++++++++++++++++++--- drivers/scsi/isci/port_config.c | 33 +++++++++++------- 2 files changed, 76 insertions(+), 16 deletions(-) diff --git a/drivers/scsi/isci/host.c b/drivers/scsi/isci/host.c index 7e4d709dc201..1a65d6514237 100644 --- a/drivers/scsi/isci/host.c +++ b/drivers/scsi/isci/host.c @@ -899,7 +899,8 @@ static enum sci_status sci_controller_start_next_phy(struct isci_host *ihost) */ if ((iphy->is_in_link_training == false && state == SCI_PHY_INITIAL) || (iphy->is_in_link_training == false && state == SCI_PHY_STOPPED) || - (iphy->is_in_link_training == true && is_phy_starting(iphy))) { + (iphy->is_in_link_training == true && is_phy_starting(iphy)) || + (ihost->port_agent.phy_ready_mask != ihost->port_agent.phy_configured_mask)) { is_controller_start_complete = false; break; } @@ -1904,6 +1905,31 @@ static void power_control_timeout(unsigned long data) ihost->power_control.phys_waiting--; ihost->power_control.phys_granted_power++; sci_phy_consume_power_handler(iphy); + + if (iphy->protocol == SCIC_SDS_PHY_PROTOCOL_SAS) { + u8 j; + + for (j = 0; j < SCI_MAX_PHYS; j++) { + struct isci_phy *requester = ihost->power_control.requesters[j]; + + /* + * Search the power_control queue to see if there are other phys + * attached to the same remote device. If found, take all of + * them out of await_sas_power state. + */ + if (requester != NULL && requester != iphy) { + u8 other = memcmp(requester->frame_rcvd.iaf.sas_addr, + iphy->frame_rcvd.iaf.sas_addr, + sizeof(requester->frame_rcvd.iaf.sas_addr)); + + if (other == 0) { + ihost->power_control.requesters[j] = NULL; + ihost->power_control.phys_waiting--; + sci_phy_consume_power_handler(requester); + } + } + } + } } /* @@ -1938,9 +1964,34 @@ void sci_controller_power_control_queue_insert(struct isci_host *ihost, ihost->power_control.timer_started = true; } else { - /* Add the phy in the waiting list */ - ihost->power_control.requesters[iphy->phy_index] = iphy; - ihost->power_control.phys_waiting++; + /* + * There are phys, attached to the same sas address as this phy, are + * already in READY state, this phy don't need wait. + */ + u8 i; + struct isci_phy *current_phy; + + for (i = 0; i < SCI_MAX_PHYS; i++) { + u8 other; + current_phy = &ihost->phys[i]; + + other = memcmp(current_phy->frame_rcvd.iaf.sas_addr, + iphy->frame_rcvd.iaf.sas_addr, + sizeof(current_phy->frame_rcvd.iaf.sas_addr)); + + if (current_phy->sm.current_state_id == SCI_PHY_READY && + current_phy->protocol == SCIC_SDS_PHY_PROTOCOL_SAS && + other == 0) { + sci_phy_consume_power_handler(iphy); + break; + } + } + + if (i == SCI_MAX_PHYS) { + /* Add the phy in the waiting list */ + ihost->power_control.requesters[iphy->phy_index] = iphy; + ihost->power_control.phys_waiting++; + } } } diff --git a/drivers/scsi/isci/port_config.c b/drivers/scsi/isci/port_config.c index 37d15d3165da..6d1e9544cbe5 100644 --- a/drivers/scsi/isci/port_config.c +++ b/drivers/scsi/isci/port_config.c @@ -466,6 +466,23 @@ sci_apc_agent_validate_phy_configuration(struct isci_host *ihost, return sci_port_configuration_agent_validate_ports(ihost, port_agent); } +/* + * This routine will restart the automatic port configuration timeout + * timer for the next time period. This could be caused by either a link + * down event or a link up event where we can not yet tell to which a phy + * belongs. + */ +static void sci_apc_agent_start_timer( + struct sci_port_configuration_agent *port_agent, + u32 timeout) +{ + if (port_agent->timer_pending) + sci_del_timer(&port_agent->timer); + + port_agent->timer_pending = true; + sci_mod_timer(&port_agent->timer, timeout); +} + static void sci_apc_agent_configure_ports(struct isci_host *ihost, struct sci_port_configuration_agent *port_agent, struct isci_phy *iphy, @@ -565,17 +582,8 @@ static void sci_apc_agent_configure_ports(struct isci_host *ihost, break; case SCIC_SDS_APC_START_TIMER: - /* - * This can occur for either a link down event, or a link - * up event where we cannot yet tell the port to which a - * phy belongs. - */ - if (port_agent->timer_pending) - sci_del_timer(&port_agent->timer); - - port_agent->timer_pending = true; - sci_mod_timer(&port_agent->timer, - SCIC_SDS_APC_WAIT_LINK_UP_NOTIFICATION); + sci_apc_agent_start_timer(port_agent, + SCIC_SDS_APC_WAIT_LINK_UP_NOTIFICATION); break; case SCIC_SDS_APC_SKIP_PHY: @@ -607,7 +615,8 @@ static void sci_apc_agent_link_up(struct isci_host *ihost, if (!iport) { /* the phy is not the part of this port */ port_agent->phy_ready_mask |= 1 << phy_index; - sci_apc_agent_configure_ports(ihost, port_agent, iphy, true); + sci_apc_agent_start_timer(port_agent, + SCIC_SDS_APC_WAIT_LINK_UP_NOTIFICATION); } else { /* the phy is already the part of the port */ u32 port_state = iport->sm.current_state_id; From 0953dbea1d9a84c8443b5f5bb45229a6c9d7f9f3 Mon Sep 17 00:00:00 2001 From: Marcin Tomczak Date: Wed, 4 Jan 2012 01:33:36 -0800 Subject: [PATCH 311/820] [SCSI] isci: fix start OOB Split scu_link_layer_start_oob function to reset and enable and add flush after all steps. Signed-off-by: Marcin Tomczak Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/phy.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/drivers/scsi/isci/phy.c b/drivers/scsi/isci/phy.c index 61000cde84c7..fe18acfd6eb3 100644 --- a/drivers/scsi/isci/phy.c +++ b/drivers/scsi/isci/phy.c @@ -1098,24 +1098,25 @@ static void scu_link_layer_stop_protocol_engine( writel(enable_spinup_value, &iphy->link_layer_registers->notify_enable_spinup_control); } -/** - * - * - * This method will start the OOB/SN state machine for this struct isci_phy object. - */ -static void scu_link_layer_start_oob( - struct isci_phy *iphy) +static void scu_link_layer_start_oob(struct isci_phy *iphy) { - u32 scu_sas_pcfg_value; + struct scu_link_layer_registers __iomem *ll = iphy->link_layer_registers; + u32 val; - scu_sas_pcfg_value = - readl(&iphy->link_layer_registers->phy_configuration); - scu_sas_pcfg_value |= SCU_SAS_PCFG_GEN_BIT(OOB_ENABLE); - scu_sas_pcfg_value &= - ~(SCU_SAS_PCFG_GEN_BIT(OOB_RESET) | - SCU_SAS_PCFG_GEN_BIT(HARD_RESET)); - writel(scu_sas_pcfg_value, - &iphy->link_layer_registers->phy_configuration); + /** Reset OOB sequence - start */ + val = readl(&ll->phy_configuration); + val &= ~(SCU_SAS_PCFG_GEN_BIT(OOB_RESET) | + SCU_SAS_PCFG_GEN_BIT(HARD_RESET)); + writel(val, &ll->phy_configuration); + readl(&ll->phy_configuration); /* flush */ + /** Reset OOB sequence - end */ + + /** Start OOB sequence - start */ + val = readl(&ll->phy_configuration); + val |= SCU_SAS_PCFG_GEN_BIT(OOB_ENABLE); + writel(val, &ll->phy_configuration); + readl(&ll->phy_configuration); /* flush */ + /** Start OOB sequence - end */ } /** From 05b080fc933bb068b32fa119db00e8efcc10e3bd Mon Sep 17 00:00:00 2001 From: Marcin Tomczak Date: Wed, 4 Jan 2012 01:33:41 -0800 Subject: [PATCH 312/820] [SCSI] isci: fix, prevent port from getting stuck in the 'configuring' state When expander connected in x2 or x4 mode and with IO runnning, if a cable from wideport is plugged out from the phy, IO's start failing on all the targets. Observed that when cable is pulled with IO running, cominit is happening on all the links and IO's start dropping to 0 and eventually the whole IO fails. Second observation, target is trying to open and SCU is responding with "Open reject no destination". A cause of the problem is when the port went from the "ready configuring substate" back to "ready configuring substate" as a result of phy being pulled off, scic suspended the port task scheduler register. As a result no IO was allowed and in the "substate configuring enter" routine the IO never goes back to 0. As a result the port never comes out of "ready substate configuring". The patch adds a mechanism of activate and deactivate phy when a port link up, which fixes the problem. Signed-off-by: Bartek Nowakowski Signed-off-by: Maciej Trela Signed-off-by: Marcin Tomczak Signed-off-by: Dan Williams Signed-off-by: James Bottomley --- drivers/scsi/isci/port.c | 92 ++++++++++++++++++++++------------------ drivers/scsi/isci/port.h | 6 +++ 2 files changed, 56 insertions(+), 42 deletions(-) diff --git a/drivers/scsi/isci/port.c b/drivers/scsi/isci/port.c index 49e8a72d1c5b..7c6ac58a5c4c 100644 --- a/drivers/scsi/isci/port.c +++ b/drivers/scsi/isci/port.c @@ -647,19 +647,26 @@ void sci_port_setup_transports(struct isci_port *iport, u32 device_id) } } -static void sci_port_activate_phy(struct isci_port *iport, struct isci_phy *iphy, - bool do_notify_user) +static void sci_port_resume_phy(struct isci_port *iport, struct isci_phy *iphy) +{ + sci_phy_resume(iphy); + iport->enabled_phy_mask |= 1 << iphy->phy_index; +} + +static void sci_port_activate_phy(struct isci_port *iport, + struct isci_phy *iphy, + u8 flags) { struct isci_host *ihost = iport->owning_controller; - if (iphy->protocol != SCIC_SDS_PHY_PROTOCOL_SATA) + if (iphy->protocol != SCIC_SDS_PHY_PROTOCOL_SATA && (flags & PF_RESUME)) sci_phy_resume(iphy); iport->active_phy_mask |= 1 << iphy->phy_index; sci_controller_clear_invalid_phy(ihost, iphy); - if (do_notify_user == true) + if (flags & PF_NOTIFY) isci_port_link_up(ihost, iport, iphy); } @@ -669,6 +676,7 @@ void sci_port_deactivate_phy(struct isci_port *iport, struct isci_phy *iphy, struct isci_host *ihost = iport->owning_controller; iport->active_phy_mask &= ~(1 << iphy->phy_index); + iport->enabled_phy_mask &= ~(1 << iphy->phy_index); if (!iport->active_phy_mask) iport->last_active_phy = iphy->phy_index; @@ -705,18 +713,16 @@ static void sci_port_invalid_link_up(struct isci_port *iport, struct isci_phy *i * sci_port_general_link_up_handler - phy can be assigned to port? * @sci_port: sci_port object for which has a phy that has gone link up. * @sci_phy: This is the struct isci_phy object that has gone link up. - * @do_notify_user: This parameter specifies whether to inform the user (via - * sci_port_link_up()) as to the fact that a new phy as become ready. + * @flags: PF_RESUME, PF_NOTIFY to sci_port_activate_phy * - * Determine if this phy can be assigned to this - * port . If the phy is not a valid PHY for - * this port then the function will notify the user. A PHY can only be - * part of a port if it's attached SAS ADDRESS is the same as all other PHYs in - * the same port. none + * Determine if this phy can be assigned to this port . If the phy is + * not a valid PHY for this port then the function will notify the user. + * A PHY can only be part of a port if it's attached SAS ADDRESS is the + * same as all other PHYs in the same port. */ static void sci_port_general_link_up_handler(struct isci_port *iport, - struct isci_phy *iphy, - bool do_notify_user) + struct isci_phy *iphy, + u8 flags) { struct sci_sas_address port_sas_address; struct sci_sas_address phy_sas_address; @@ -734,7 +740,7 @@ static void sci_port_general_link_up_handler(struct isci_port *iport, iport->active_phy_mask == 0) { struct sci_base_state_machine *sm = &iport->sm; - sci_port_activate_phy(iport, iphy, do_notify_user); + sci_port_activate_phy(iport, iphy, flags); if (sm->current_state_id == SCI_PORT_RESETTING) port_state_machine_change(iport, SCI_PORT_READY); } else @@ -785,11 +791,16 @@ bool sci_port_link_detected( struct isci_phy *iphy) { if ((iport->logical_port_index != SCIC_SDS_DUMMY_PORT) && - (iphy->protocol == SCIC_SDS_PHY_PROTOCOL_SATA) && - sci_port_is_wide(iport)) { - sci_port_invalid_link_up(iport, iphy); - - return false; + (iphy->protocol == SCIC_SDS_PHY_PROTOCOL_SATA)) { + if (sci_port_is_wide(iport)) { + sci_port_invalid_link_up(iport, iphy); + return false; + } else { + struct isci_host *ihost = iport->owning_controller; + struct isci_port *dst_port = &(ihost->ports[iphy->phy_index]); + writel(iphy->phy_index, + &dst_port->port_pe_configuration_register[iphy->phy_index]); + } } return true; @@ -979,6 +990,13 @@ static void sci_port_ready_substate_waiting_enter(struct sci_base_state_machine } } +static void scic_sds_port_ready_substate_waiting_exit( + struct sci_base_state_machine *sm) +{ + struct isci_port *iport = container_of(sm, typeof(*iport), sm); + sci_port_resume_port_task_scheduler(iport); +} + static void sci_port_ready_substate_operational_enter(struct sci_base_state_machine *sm) { u32 index; @@ -992,13 +1010,13 @@ static void sci_port_ready_substate_operational_enter(struct sci_base_state_mach writel(iport->physical_port_index, &iport->port_pe_configuration_register[ iport->phy_table[index]->phy_index]); + if (((iport->active_phy_mask^iport->enabled_phy_mask) & (1 << index)) != 0) + sci_port_resume_phy(iport, iport->phy_table[index]); } } sci_port_update_viit_entry(iport); - sci_port_resume_port_task_scheduler(iport); - /* * Post the dummy task for the port so the hardware can schedule * io correctly @@ -1065,20 +1083,9 @@ static void sci_port_ready_substate_configuring_enter(struct sci_base_state_mach if (iport->active_phy_mask == 0) { isci_port_not_ready(ihost, iport); - port_state_machine_change(iport, - SCI_PORT_SUB_WAITING); - } else if (iport->started_request_count == 0) - port_state_machine_change(iport, - SCI_PORT_SUB_OPERATIONAL); -} - -static void sci_port_ready_substate_configuring_exit(struct sci_base_state_machine *sm) -{ - struct isci_port *iport = container_of(sm, typeof(*iport), sm); - - sci_port_suspend_port_task_scheduler(iport); - if (iport->ready_exit) - sci_port_invalidate_dummy_remote_node(iport); + port_state_machine_change(iport, SCI_PORT_SUB_WAITING); + } else + port_state_machine_change(iport, SCI_PORT_SUB_OPERATIONAL); } enum sci_status sci_port_start(struct isci_port *iport) @@ -1256,7 +1263,7 @@ enum sci_status sci_port_add_phy(struct isci_port *iport, if (status != SCI_SUCCESS) return status; - sci_port_general_link_up_handler(iport, iphy, true); + sci_port_general_link_up_handler(iport, iphy, PF_NOTIFY|PF_RESUME); iport->not_ready_reason = SCIC_PORT_NOT_READY_RECONFIGURING; port_state_machine_change(iport, SCI_PORT_SUB_CONFIGURING); @@ -1266,7 +1273,7 @@ enum sci_status sci_port_add_phy(struct isci_port *iport, if (status != SCI_SUCCESS) return status; - sci_port_general_link_up_handler(iport, iphy, true); + sci_port_general_link_up_handler(iport, iphy, PF_NOTIFY); /* Re-enter the configuring state since this may be the last phy in * the port. @@ -1342,13 +1349,13 @@ enum sci_status sci_port_link_up(struct isci_port *iport, /* Since this is the first phy going link up for the port we * can just enable it and continue */ - sci_port_activate_phy(iport, iphy, true); + sci_port_activate_phy(iport, iphy, PF_NOTIFY|PF_RESUME); port_state_machine_change(iport, SCI_PORT_SUB_OPERATIONAL); return SCI_SUCCESS; case SCI_PORT_SUB_OPERATIONAL: - sci_port_general_link_up_handler(iport, iphy, true); + sci_port_general_link_up_handler(iport, iphy, PF_NOTIFY|PF_RESUME); return SCI_SUCCESS; case SCI_PORT_RESETTING: /* TODO We should make sure that the phy that has gone @@ -1365,7 +1372,7 @@ enum sci_status sci_port_link_up(struct isci_port *iport, /* In the resetting state we don't notify the user regarding * link up and link down notifications. */ - sci_port_general_link_up_handler(iport, iphy, false); + sci_port_general_link_up_handler(iport, iphy, PF_RESUME); return SCI_SUCCESS; default: dev_warn(sciport_to_dev(iport), @@ -1588,14 +1595,14 @@ static const struct sci_base_state sci_port_state_table[] = { }, [SCI_PORT_SUB_WAITING] = { .enter_state = sci_port_ready_substate_waiting_enter, + .exit_state = scic_sds_port_ready_substate_waiting_exit, }, [SCI_PORT_SUB_OPERATIONAL] = { .enter_state = sci_port_ready_substate_operational_enter, .exit_state = sci_port_ready_substate_operational_exit }, [SCI_PORT_SUB_CONFIGURING] = { - .enter_state = sci_port_ready_substate_configuring_enter, - .exit_state = sci_port_ready_substate_configuring_exit + .enter_state = sci_port_ready_substate_configuring_enter }, [SCI_PORT_RESETTING] = { .exit_state = sci_port_resetting_state_exit @@ -1613,6 +1620,7 @@ void sci_port_construct(struct isci_port *iport, u8 index, iport->logical_port_index = SCIC_SDS_DUMMY_PORT; iport->physical_port_index = index; iport->active_phy_mask = 0; + iport->enabled_phy_mask = 0; iport->last_active_phy = 0; iport->ready_exit = false; diff --git a/drivers/scsi/isci/port.h b/drivers/scsi/isci/port.h index b3a7f12803f5..08116090eb70 100644 --- a/drivers/scsi/isci/port.h +++ b/drivers/scsi/isci/port.h @@ -63,6 +63,9 @@ #define SCIC_SDS_DUMMY_PORT 0xFF +#define PF_NOTIFY (1 << 0) +#define PF_RESUME (1 << 1) + struct isci_phy; struct isci_host; @@ -83,6 +86,8 @@ enum isci_status { * @logical_port_index: software port index * @physical_port_index: hardware port index * @active_phy_mask: identifies phy members + * @enabled_phy_mask: phy mask for the port + * that are already part of the port * @reserved_tag: * @reserved_rni: reserver for port task scheduler workaround * @started_request_count: reference count for outstanding commands @@ -104,6 +109,7 @@ struct isci_port { u8 logical_port_index; u8 physical_port_index; u8 active_phy_mask; + u8 enabled_phy_mask; u8 last_active_phy; u16 reserved_rni; u16 reserved_tag; From 466c08c71a7dc19528e9b336c5bfa5ec41730c7c Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Tue, 10 Jan 2012 11:27:01 +0800 Subject: [PATCH 313/820] [SCSI] don't change sdev starvation list order without request dispatched The sdev is deleted from starved list and then try to dispatch from this device. It's quite possible the sdev can't eventually dispatch a request, then the sdev will be in starved list tail. This isn't fair. There are two cases here: 1. unplug path. scsi_request_fn() calls to scsi_target_queue_ready(), then the dev is removed from starved list, but quite possible host queue isn't ready, the dev is moved to starved list without dispatching any request. 2. scsi_run_queue path. It deletes the dev from starved list first (both global and local starved lists), then handles the dev. Then we could have the same process like case 1. This patch fixes the first case. Case 2 isn't fixed, because there is a rare case scsi_run_queue finds host isn't busy but scsi_request_fn finds host is busy (other CPU is faster to get host queue depth). Not deleting the dev from starved list in scsi_run_queue will keep scsi_run_queue looping (though this is very rare case, because host will become busy). Fortunately fixing case 1 already gives big improvement for starvation in my test. In a 12 disk JBOD setup, running file creation under EXT4, this gives 12% more throughput. Signed-off-by: Shaohua Li Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index f85cfa6c47b5..b2c95dbe9d65 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1316,15 +1316,10 @@ static inline int scsi_target_queue_ready(struct Scsi_Host *shost, } if (scsi_target_is_busy(starget)) { - if (list_empty(&sdev->starved_entry)) - list_add_tail(&sdev->starved_entry, - &shost->starved_list); + list_move_tail(&sdev->starved_entry, &shost->starved_list); return 0; } - /* We're OK to process the command, so we can't be starved */ - if (!list_empty(&sdev->starved_entry)) - list_del_init(&sdev->starved_entry); return 1; } From 7e95fffe080d4dbe826dfe864eb084916cdc6468 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Tue, 10 Jan 2012 15:42:34 -0800 Subject: [PATCH 314/820] [SCSI] sg: convert to kstrtoul_from_user() Instead of open coding this function use kstrtoul_from_user() directly. Signed-off-by: Stephen Boyd Acked-by: Douglas Gilbert Signed-off-by: Andrew Morton Signed-off-by: James Bottomley --- drivers/scsi/sg.c | 25 +++++++++++-------------- 1 file changed, 11 insertions(+), 14 deletions(-) diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 02d99982a74d..eacd46bb36b9 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -2368,16 +2368,15 @@ static ssize_t sg_proc_write_adio(struct file *filp, const char __user *buffer, size_t count, loff_t *off) { - int num; - char buff[11]; + int err; + unsigned long num; if (!capable(CAP_SYS_ADMIN) || !capable(CAP_SYS_RAWIO)) return -EACCES; - num = (count < 10) ? count : 10; - if (copy_from_user(buff, buffer, num)) - return -EFAULT; - buff[num] = '\0'; - sg_allow_dio = simple_strtoul(buff, NULL, 10) ? 1 : 0; + err = kstrtoul_from_user(buffer, count, 0, &num); + if (err) + return err; + sg_allow_dio = num ? 1 : 0; return count; } @@ -2390,17 +2389,15 @@ static ssize_t sg_proc_write_dressz(struct file *filp, const char __user *buffer, size_t count, loff_t *off) { - int num; + int err; unsigned long k = ULONG_MAX; - char buff[11]; if (!capable(CAP_SYS_ADMIN) || !capable(CAP_SYS_RAWIO)) return -EACCES; - num = (count < 10) ? count : 10; - if (copy_from_user(buff, buffer, num)) - return -EFAULT; - buff[num] = '\0'; - k = simple_strtoul(buff, NULL, 10); + + err = kstrtoul_from_user(buffer, count, 0, &k); + if (err) + return err; if (k <= 1048576) { /* limit "big buff" to 1 MB */ sg_big_buff = k; return count; From 124dd90f6525ee785b47b59aebadd4a35f797dc1 Mon Sep 17 00:00:00 2001 From: adam radford Date: Tue, 10 Jan 2012 18:07:56 -0800 Subject: [PATCH 315/820] [SCSI] megaraid: fix sparse warnings There's a zero day mistake in the megaraid driver in that the code that obtains the version number does a >> 8 on a char quantity. This >>8 causes a sparse warning because it always produces zero. Al Viro suggested these shifts should be >> 4 thus treating the firmware version as a BCD quantity. However, in the interests of safety we've elected to replace the >> 8 quantities with an explicit zero, thus quieting the sparse warning while preserving the same (albeit incorrect) version number as had previously been seen. Signed-off-by: Adam Radford Signed-off-by: James Bottomley --- drivers/scsi/megaraid.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/megaraid.c b/drivers/scsi/megaraid.c index 5c1776406c96..15eefa1d61fd 100644 --- a/drivers/scsi/megaraid.c +++ b/drivers/scsi/megaraid.c @@ -306,19 +306,22 @@ mega_query_adapter(adapter_t *adapter) adapter->host->sg_tablesize = adapter->sglen; - /* use HP firmware and bios version encoding */ + /* use HP firmware and bios version encoding + Note: fw_version[0|1] and bios_version[0|1] were originally shifted + right 8 bits making them zero. This 0 value was hardcoded to fix + sparse warnings. */ if (adapter->product_info.subsysvid == HP_SUBSYS_VID) { sprintf (adapter->fw_version, "%c%d%d.%d%d", adapter->product_info.fw_version[2], - adapter->product_info.fw_version[1] >> 8, + 0, adapter->product_info.fw_version[1] & 0x0f, - adapter->product_info.fw_version[0] >> 8, + 0, adapter->product_info.fw_version[0] & 0x0f); sprintf (adapter->bios_version, "%c%d%d.%d%d", adapter->product_info.bios_version[2], - adapter->product_info.bios_version[1] >> 8, + 0, adapter->product_info.bios_version[1] & 0x0f, - adapter->product_info.bios_version[0] >> 8, + 0, adapter->product_info.bios_version[0] & 0x0f); } else { memcpy(adapter->fw_version, From 4f77083ed0325ceb9cd5701c0f335583df3dded2 Mon Sep 17 00:00:00 2001 From: Mike Hernandez Date: Wed, 11 Jan 2012 02:44:15 -0800 Subject: [PATCH 316/820] [SCSI] qla4xxx: Temperature monitoring for ISP82XX core. During watchdog, need to monitor temperature of ISP82XX core and set device state to FAILED when temperature reaches "Panic" level. Signed-off-by: Mike Hernandez Signed-off-by: Vikas Chaudhary Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_def.h | 1 + drivers/scsi/qla4xxx/ql4_nx.h | 19 ++++++++++++--- drivers/scsi/qla4xxx/ql4_os.c | 42 +++++++++++++++++++++++++++++++++- 3 files changed, 58 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/qla4xxx/ql4_def.h b/drivers/scsi/qla4xxx/ql4_def.h index 22a3ff02e48a..ec48dc30b9a2 100644 --- a/drivers/scsi/qla4xxx/ql4_def.h +++ b/drivers/scsi/qla4xxx/ql4_def.h @@ -671,6 +671,7 @@ struct scsi_qla_host { uint16_t pri_ddb_idx; uint16_t sec_ddb_idx; int is_reset; + uint16_t temperature; }; struct ql4_task_data { diff --git a/drivers/scsi/qla4xxx/ql4_nx.h b/drivers/scsi/qla4xxx/ql4_nx.h index 35376a1c3f1b..cfb2f2edac3a 100644 --- a/drivers/scsi/qla4xxx/ql4_nx.h +++ b/drivers/scsi/qla4xxx/ql4_nx.h @@ -19,12 +19,25 @@ #define PHAN_PEG_RCV_INITIALIZED 0xff01 /*CRB_RELATED*/ -#define QLA82XX_CRB_BASE QLA82XX_CAM_RAM(0x200) -#define QLA82XX_REG(X) (QLA82XX_CRB_BASE+(X)) - +#define QLA82XX_CRB_BASE (QLA82XX_CAM_RAM(0x200)) +#define QLA82XX_REG(X) (QLA82XX_CRB_BASE+(X)) #define CRB_CMDPEG_STATE QLA82XX_REG(0x50) #define CRB_RCVPEG_STATE QLA82XX_REG(0x13c) #define CRB_DMA_SHIFT QLA82XX_REG(0xcc) +#define CRB_TEMP_STATE QLA82XX_REG(0x1b4) + +#define qla82xx_get_temp_val(x) ((x) >> 16) +#define qla82xx_get_temp_state(x) ((x) & 0xffff) +#define qla82xx_encode_temp(val, state) (((val) << 16) | (state)) + +/* + * Temperature control. + */ +enum { + QLA82XX_TEMP_NORMAL = 0x1, /* Normal operating range */ + QLA82XX_TEMP_WARN, /* Sound alert, temperature getting high */ + QLA82XX_TEMP_PANIC /* Fatal error, hardware has shut down. */ +}; #define QLA82XX_HW_H0_CH_HUB_ADR 0x05 #define QLA82XX_HW_H1_CH_HUB_ADR 0x0E diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index 0048a3facd7a..9d3eab513935 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -1971,6 +1971,42 @@ mem_alloc_error_exit: return QLA_ERROR; } +/** + * qla4_8xxx_check_temp - Check the ISP82XX temperature. + * @ha: adapter block pointer. + * + * Note: The caller should not hold the idc lock. + **/ +static int qla4_8xxx_check_temp(struct scsi_qla_host *ha) +{ + uint32_t temp, temp_state, temp_val; + int status = QLA_SUCCESS; + + temp = qla4_8xxx_rd_32(ha, CRB_TEMP_STATE); + + temp_state = qla82xx_get_temp_state(temp); + temp_val = qla82xx_get_temp_val(temp); + + if (temp_state == QLA82XX_TEMP_PANIC) { + ql4_printk(KERN_WARNING, ha, "Device temperature %d degrees C" + " exceeds maximum allowed. Hardware has been shut" + " down.\n", temp_val); + status = QLA_ERROR; + } else if (temp_state == QLA82XX_TEMP_WARN) { + if (ha->temperature == QLA82XX_TEMP_NORMAL) + ql4_printk(KERN_WARNING, ha, "Device temperature %d" + " degrees C exceeds operating range." + " Immediate action needed.\n", temp_val); + } else { + if (ha->temperature == QLA82XX_TEMP_WARN) + ql4_printk(KERN_INFO, ha, "Device temperature is" + " now %d degrees C in normal range.\n", + temp_val); + } + ha->temperature = temp_state; + return status; +} + /** * qla4_8xxx_check_fw_alive - Check firmware health * @ha: Pointer to host adapter structure. @@ -2042,7 +2078,11 @@ void qla4_8xxx_watchdog(struct scsi_qla_host *ha) test_bit(DPC_RESET_HA, &ha->dpc_flags) || test_bit(DPC_RETRY_RESET_HA, &ha->dpc_flags))) { dev_state = qla4_8xxx_rd_32(ha, QLA82XX_CRB_DEV_STATE); - if (dev_state == QLA82XX_DEV_NEED_RESET && + + if (qla4_8xxx_check_temp(ha)) { + set_bit(DPC_HA_UNRECOVERABLE, &ha->dpc_flags); + qla4xxx_wake_dpc(ha); + } else if (dev_state == QLA82XX_DEV_NEED_RESET && !test_bit(DPC_RESET_HA, &ha->dpc_flags)) { if (!ql4xdontresethba) { ql4_printk(KERN_INFO, ha, "%s: HW State: " From e6bd0ebd4a8ba9118e970c0214e912e35895c92b Mon Sep 17 00:00:00 2001 From: Giridhar Malavali Date: Wed, 11 Jan 2012 02:44:16 -0800 Subject: [PATCH 317/820] [SCSI] qla4xxx: Disable generating pause frames in case of FW hung In case of FW hung ISP82xx generates continuous pause frames which causes switch to disable port. Added fix to disable generating pause frames in case of FW hung Signed-off-by: Giridhar Malavali Signed-off-by: Vikas Chaudhary Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_mbx.c | 7 +++++++ drivers/scsi/qla4xxx/ql4_nx.h | 3 +++ drivers/scsi/qla4xxx/ql4_os.c | 10 ++++++++++ 3 files changed, 20 insertions(+) diff --git a/drivers/scsi/qla4xxx/ql4_mbx.c b/drivers/scsi/qla4xxx/ql4_mbx.c index c2593782fbbe..e1e66a45e4d0 100644 --- a/drivers/scsi/qla4xxx/ql4_mbx.c +++ b/drivers/scsi/qla4xxx/ql4_mbx.c @@ -219,6 +219,13 @@ int qla4xxx_mailbox_command(struct scsi_qla_host *ha, uint8_t inCount, ha->mailbox_timeout_count++; mbx_sts[0] = (-1); set_bit(DPC_RESET_HA, &ha->dpc_flags); + if (is_qla8022(ha)) { + ql4_printk(KERN_INFO, ha, + "disabling pause transmit on port 0 & 1.\n"); + qla4_8xxx_wr_32(ha, QLA82XX_CRB_NIU + 0x98, + CRB_NIU_XG_PAUSE_CTL_P0 | + CRB_NIU_XG_PAUSE_CTL_P1); + } goto mbox_exit; } diff --git a/drivers/scsi/qla4xxx/ql4_nx.h b/drivers/scsi/qla4xxx/ql4_nx.h index cfb2f2edac3a..dc45ac923691 100644 --- a/drivers/scsi/qla4xxx/ql4_nx.h +++ b/drivers/scsi/qla4xxx/ql4_nx.h @@ -39,6 +39,9 @@ enum { QLA82XX_TEMP_PANIC /* Fatal error, hardware has shut down. */ }; +#define CRB_NIU_XG_PAUSE_CTL_P0 0x1 +#define CRB_NIU_XG_PAUSE_CTL_P1 0x8 + #define QLA82XX_HW_H0_CH_HUB_ADR 0x05 #define QLA82XX_HW_H1_CH_HUB_ADR 0x0E #define QLA82XX_HW_H2_CH_HUB_ADR 0x03 diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index 9d3eab513935..e2ef7762dc24 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -2080,6 +2080,11 @@ void qla4_8xxx_watchdog(struct scsi_qla_host *ha) dev_state = qla4_8xxx_rd_32(ha, QLA82XX_CRB_DEV_STATE); if (qla4_8xxx_check_temp(ha)) { + ql4_printk(KERN_INFO, ha, "disabling pause" + " transmit on port 0 & 1.\n"); + qla4_8xxx_wr_32(ha, QLA82XX_CRB_NIU + 0x98, + CRB_NIU_XG_PAUSE_CTL_P0 | + CRB_NIU_XG_PAUSE_CTL_P1); set_bit(DPC_HA_UNRECOVERABLE, &ha->dpc_flags); qla4xxx_wake_dpc(ha); } else if (dev_state == QLA82XX_DEV_NEED_RESET && @@ -2099,6 +2104,11 @@ void qla4_8xxx_watchdog(struct scsi_qla_host *ha) } else { /* Check firmware health */ if (qla4_8xxx_check_fw_alive(ha)) { + ql4_printk(KERN_INFO, ha, "disabling pause" + " transmit on port 0 & 1.\n"); + qla4_8xxx_wr_32(ha, QLA82XX_CRB_NIU + 0x98, + CRB_NIU_XG_PAUSE_CTL_P0 | + CRB_NIU_XG_PAUSE_CTL_P1); halt_status = qla4_8xxx_rd_32(ha, QLA82XX_PEG_HALT_STATUS1); From 527c8b2e962d21baa38a96b22e1bf50a47fdf4fb Mon Sep 17 00:00:00 2001 From: Nilesh Javali Date: Wed, 11 Jan 2012 02:44:17 -0800 Subject: [PATCH 318/820] [SCSI] qla4xxx: Added error logging for firmware abort Added debug print with error code in case of firmware error. Signed-off-by: Nilesh Javali Signed-off-by: Vikas Chaudhary Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_def.h | 2 ++ drivers/scsi/qla4xxx/ql4_os.c | 7 +++++++ 2 files changed, 9 insertions(+) diff --git a/drivers/scsi/qla4xxx/ql4_def.h b/drivers/scsi/qla4xxx/ql4_def.h index ec48dc30b9a2..bfe68545203f 100644 --- a/drivers/scsi/qla4xxx/ql4_def.h +++ b/drivers/scsi/qla4xxx/ql4_def.h @@ -150,6 +150,8 @@ #define QL4_SESS_RECOVERY_TMO 120 /* iSCSI session */ /* recovery timeout */ +#define MSB(x) ((uint8_t)((uint16_t)(x) >> 8)) +#define LSW(x) ((uint16_t)(x)) #define LSDW(x) ((u32)((u64)(x))) #define MSDW(x) ((u32)((((u64)(x)) >> 16) >> 16)) diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index e2ef7762dc24..b75590af8ed3 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -2112,6 +2112,13 @@ void qla4_8xxx_watchdog(struct scsi_qla_host *ha) halt_status = qla4_8xxx_rd_32(ha, QLA82XX_PEG_HALT_STATUS1); + if (LSW(MSB(halt_status)) == 0x67) + ql4_printk(KERN_ERR, ha, "%s:" + " Firmware aborted with" + " error code 0x00006700." + " Device is being reset\n", + __func__); + /* Since we cannot change dev_state in interrupt * context, set appropriate DPC flag then wakeup * DPC */ From 787649993230eb1d6fc326f13be285a840092538 Mon Sep 17 00:00:00 2001 From: Sarang Radke Date: Wed, 11 Jan 2012 02:44:18 -0800 Subject: [PATCH 319/820] [SCSI] qla4xxx: Clear the RISC interrupt bit during FW init This patch fix kernel panic during kdump. Signed-off-by: Sarang Radke Signed-off-by: Vikas Chaudhary Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_init.c | 3 +++ drivers/scsi/qla4xxx/ql4_nx.c | 5 +++++ 2 files changed, 8 insertions(+) diff --git a/drivers/scsi/qla4xxx/ql4_init.c b/drivers/scsi/qla4xxx/ql4_init.c index 1bdfa8120ac8..90614f38b55d 100644 --- a/drivers/scsi/qla4xxx/ql4_init.c +++ b/drivers/scsi/qla4xxx/ql4_init.c @@ -697,6 +697,9 @@ int qla4xxx_start_firmware(struct scsi_qla_host *ha) writel(set_rmask(CSR_SCSI_PROCESSOR_INTR), &ha->reg->ctrl_status); readl(&ha->reg->ctrl_status); + writel(set_rmask(CSR_SCSI_COMPLETION_INTR), + &ha->reg->ctrl_status); + readl(&ha->reg->ctrl_status); spin_unlock_irqrestore(&ha->hardware_lock, flags); if (qla4xxx_get_firmware_state(ha) == QLA_SUCCESS) { DEBUG2(printk("scsi%ld: %s: Get firmware " diff --git a/drivers/scsi/qla4xxx/ql4_nx.c b/drivers/scsi/qla4xxx/ql4_nx.c index 8d6bc1b2ff17..78f1111158d7 100644 --- a/drivers/scsi/qla4xxx/ql4_nx.c +++ b/drivers/scsi/qla4xxx/ql4_nx.c @@ -1875,6 +1875,11 @@ exit: int qla4_8xxx_load_risc(struct scsi_qla_host *ha) { int retval; + + /* clear the interrupt */ + writel(0, &ha->qla4_8xxx_reg->host_int); + readl(&ha->qla4_8xxx_reg->host_int); + retval = qla4_8xxx_device_state_handler(ha); if (retval == QLA_SUCCESS && !test_bit(AF_INIT_DONE, &ha->flags)) From 991039f9f40d0809108e8d453c68c2ae965b3ac4 Mon Sep 17 00:00:00 2001 From: Vikas Chaudhary Date: Wed, 11 Jan 2012 02:44:19 -0800 Subject: [PATCH 320/820] [SCSI] qla4xxx: Update license Signed-off-by: Vikas Chaudhary Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- Documentation/scsi/LICENSE.qla4xxx | 23 +---------------------- 1 file changed, 1 insertion(+), 22 deletions(-) diff --git a/Documentation/scsi/LICENSE.qla4xxx b/Documentation/scsi/LICENSE.qla4xxx index 494980e40491..ab899591ecb7 100644 --- a/Documentation/scsi/LICENSE.qla4xxx +++ b/Documentation/scsi/LICENSE.qla4xxx @@ -1,32 +1,11 @@ Copyright (c) 2003-2011 QLogic Corporation -QLogic Linux iSCSI HBA Driver +QLogic Linux iSCSI Driver This program includes a device driver for Linux 3.x. You may modify and redistribute the device driver code under the GNU General Public License (a copy of which is attached hereto as Exhibit A) published by the Free Software Foundation (version 2). -REGARDLESS OF WHAT LICENSING MECHANISM IS USED OR APPLICABLE, -THIS PROGRAM IS PROVIDED BY QLOGIC CORPORATION "AS IS'' AND ANY -EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A -PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR -BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, -EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED -TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, -DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON -ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, -OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY -OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE -POSSIBILITY OF SUCH DAMAGE. - -USER ACKNOWLEDGES AND AGREES THAT USE OF THIS PROGRAM WILL NOT -CREATE OR GIVE GROUNDS FOR A LICENSE BY IMPLICATION, ESTOPPEL, OR -OTHERWISE IN ANY INTELLECTUAL PROPERTY RIGHTS (PATENT, COPYRIGHT, -TRADE SECRET, MASK WORK, OR OTHER PROPRIETARY RIGHT) EMBODIED IN -ANY OTHER QLOGIC HARDWARE OR SOFTWARE EITHER SOLELY OR IN -COMBINATION WITH THIS PROGRAM. - EXHIBIT A From a4e8a715a39e5d29f7d35a7289b914ef205a866b Mon Sep 17 00:00:00 2001 From: Karen Higgins Date: Wed, 11 Jan 2012 02:44:20 -0800 Subject: [PATCH 321/820] [SCSI] qla4xxx: Cleanup modinfo display Beautify modinfo display. Display correct info for ql4xextended_error_logging [jejb: fixup checkpatch warning] Signed-off-by: Karen Higgins Signed-off-by: Vikas Chaudhary Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_os.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c index b75590af8ed3..ce6d3b7f0c61 100644 --- a/drivers/scsi/qla4xxx/ql4_os.c +++ b/drivers/scsi/qla4xxx/ql4_os.c @@ -35,43 +35,44 @@ static struct kmem_cache *srb_cachep; int ql4xdisablesysfsboot = 1; module_param(ql4xdisablesysfsboot, int, S_IRUGO | S_IWUSR); MODULE_PARM_DESC(ql4xdisablesysfsboot, - "Set to disable exporting boot targets to sysfs\n" - " 0 - Export boot targets\n" - " 1 - Do not export boot targets (Default)"); + " Set to disable exporting boot targets to sysfs.\n" + "\t\t 0 - Export boot targets\n" + "\t\t 1 - Do not export boot targets (Default)"); int ql4xdontresethba = 0; module_param(ql4xdontresethba, int, S_IRUGO | S_IWUSR); MODULE_PARM_DESC(ql4xdontresethba, - "Don't reset the HBA for driver recovery \n" - " 0 - It will reset HBA (Default)\n" - " 1 - It will NOT reset HBA"); + " Don't reset the HBA for driver recovery.\n" + "\t\t 0 - It will reset HBA (Default)\n" + "\t\t 1 - It will NOT reset HBA"); -int ql4xextended_error_logging = 0; /* 0 = off, 1 = log errors */ +int ql4xextended_error_logging; module_param(ql4xextended_error_logging, int, S_IRUGO | S_IWUSR); MODULE_PARM_DESC(ql4xextended_error_logging, - "Option to enable extended error logging, " - "Default is 0 - no logging, 1 - debug logging"); + " Option to enable extended error logging.\n" + "\t\t 0 - no logging (Default)\n" + "\t\t 2 - debug logging"); int ql4xenablemsix = 1; module_param(ql4xenablemsix, int, S_IRUGO|S_IWUSR); MODULE_PARM_DESC(ql4xenablemsix, - "Set to enable MSI or MSI-X interrupt mechanism.\n" - " 0 = enable INTx interrupt mechanism.\n" - " 1 = enable MSI-X interrupt mechanism (Default).\n" - " 2 = enable MSI interrupt mechanism."); + " Set to enable MSI or MSI-X interrupt mechanism.\n" + "\t\t 0 = enable INTx interrupt mechanism.\n" + "\t\t 1 = enable MSI-X interrupt mechanism (Default).\n" + "\t\t 2 = enable MSI interrupt mechanism."); #define QL4_DEF_QDEPTH 32 static int ql4xmaxqdepth = QL4_DEF_QDEPTH; module_param(ql4xmaxqdepth, int, S_IRUGO | S_IWUSR); MODULE_PARM_DESC(ql4xmaxqdepth, - "Maximum queue depth to report for target devices.\n" - " Default: 32."); + " Maximum queue depth to report for target devices.\n" + "\t\t Default: 32."); static int ql4xsess_recovery_tmo = QL4_SESS_RECOVERY_TMO; module_param(ql4xsess_recovery_tmo, int, S_IRUGO); MODULE_PARM_DESC(ql4xsess_recovery_tmo, "Target Session Recovery Timeout.\n" - " Default: 120 sec."); + "\t\t Default: 120 sec."); static int qla4xxx_wait_for_hba_online(struct scsi_qla_host *ha); /* From 19b628dfc60eda4459b65e4b6b7a748932d28929 Mon Sep 17 00:00:00 2001 From: Vikas Chaudhary Date: Wed, 11 Jan 2012 02:44:21 -0800 Subject: [PATCH 322/820] [SCSI] qla4xxx: Update driver version to 5.02.00-k12 Signed-off-by: Vikas Chaudhary Signed-off-by: James Bottomley --- drivers/scsi/qla4xxx/ql4_version.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/qla4xxx/ql4_version.h b/drivers/scsi/qla4xxx/ql4_version.h index 7d04eb05c45e..133989b3a9f4 100644 --- a/drivers/scsi/qla4xxx/ql4_version.h +++ b/drivers/scsi/qla4xxx/ql4_version.h @@ -5,4 +5,4 @@ * See LICENSE.qla4xxx for copyright and licensing details. */ -#define QLA4XXX_DRIVER_VERSION "5.02.00-k11" +#define QLA4XXX_DRIVER_VERSION "5.02.00-k12" From f10448689d95b9516c656ccd4078839e656656e7 Mon Sep 17 00:00:00 2001 From: Anton Vorontsov Date: Wed, 11 Jan 2012 05:11:46 +0400 Subject: [PATCH 323/820] x86: Get rid of dubious one-bit signed bitfield This very noisy sparse warning appears on almost every file in the kernel: CHECK init/main.c arch/x86/include/asm/thread_info.h:43:55: error: dubious one-bit signed bitfield arch/x86/include/asm/thread_info.h:44:46: error: dubious one-bit signed bitfield Sparse is right and this patch changes sig_on_uaccess_error and uaccess_err flags to unsigned type and thus fixes the warning. Signed-off-by: Anton Vorontsov Acked-by: Andy Lutomirski Cc: Linus Torvalds Cc: H. Peter Anvin Cc: Dan Carpenter Link: http://lkml.kernel.org/r/20120111011146.GA30428@oksana.dev.rtsoft.ru Signed-off-by: Ingo Molnar --- arch/x86/include/asm/thread_info.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index 185b719ec61a..56a63ff7665e 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -40,8 +40,8 @@ struct thread_info { */ __u8 supervisor_stack[0]; #endif - int sig_on_uaccess_error:1; - int uaccess_err:1; /* uaccess failed */ + unsigned int sig_on_uaccess_error:1; + unsigned int uaccess_err:1; /* uaccess failed */ }; #define INIT_THREAD_INFO(tsk) \ From a762dce41cb5742a143f6aa2d80ee1aac7e1f5eb Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Fri, 13 Jan 2012 17:26:15 -0800 Subject: [PATCH 324/820] [SCSI] fcoe: fix regression on offload em matching function for initiator/target This is a regression introduced by commit 1ff9918b625457ce20d450d00f9ed0a12ba191b7 The else statement here is breaking the initiator logic of allocating xid from the offloaded em xid pool for READ I/O only to use DDP, as shown by the snippet of trace below, where the WRITE is using xid 0x5 from the offloaded em xid pool: Protocol VID Len S_ID D_ID OX_ID RX_ID Summary .. *FCP 228 96 0b.08.01 -> 01.0f.00 0x0005 0xffff SCSI: Write(10) LUN: 0x00 FCP 228 76 01.0f.00 -> 0b.08.01 0x0005 0x828d XFER_RDY ... The bug is in the else statement, for both initiator and target, the new command will have FC frame header bit 23 (FC_FC_EX_CTX) cleared as it was originated from the initiator. Also, this is assuming the frame header is already filled up, which is only true for target since for initiator, this is a new frame and oem_match gets called when em tries get xid for this i/o before it is filled up and sent out. The fix is to check if there is a fc_fcp_pkt associated w/ this frame from fr_fsp(fp), since fr_fsp(fp) is NULL for tcm_fc target and non-I/O frame in initiator. This should also return true for target only if it is an FC_RCTL_DD_UNSOL_CMD and rx_id is not allocated. Signed-off-by: Yi Zou Tested-by: Ross Brattain Signed-off-by: Robert Love Signed-off-by: James Bottomley --- drivers/scsi/fcoe/fcoe.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c index e535f95e4772..507504b77880 100644 --- a/drivers/scsi/fcoe/fcoe.c +++ b/drivers/scsi/fcoe/fcoe.c @@ -756,11 +756,12 @@ bool fcoe_oem_match(struct fc_frame *fp) if (fc_fcp_is_read(fr_fsp(fp)) && (fr_fsp(fp)->data_len > fcoe_ddp_min)) return true; - else if (!(ntoh24(fh->fh_f_ctl) & FC_FC_EX_CTX)) { + else if ((fr_fsp(fp) == NULL) && + (fh->fh_r_ctl == FC_RCTL_DD_UNSOL_CMD) && + (ntohs(fh->fh_rx_id) == FC_XID_UNKNOWN)) { fcp = fc_frame_payload_get(fp, sizeof(*fcp)); - if (ntohs(fh->fh_rx_id) == FC_XID_UNKNOWN && - fcp && (ntohl(fcp->fc_dl) > fcoe_ddp_min) && - (fcp->fc_flags & FCP_CFL_WRDATA)) + if ((fcp->fc_flags & FCP_CFL_WRDATA) && + (ntohl(fcp->fc_dl) > fcoe_ddp_min)) return true; } return false; From c6b21c93c1794113c68f3d43f321968191d87b1b Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Fri, 13 Jan 2012 17:26:20 -0800 Subject: [PATCH 325/820] [SCSI] libfc: Declare local functions static Avoid that sparse complains about missing declarations for local functions by declaring these static or by adding an #include directive. Add the __percpu annotation where it is missing. Signed-off-by: Bart Van Assche Reviewed-by: Yi Zou Signed-off-by: Robert Love Signed-off-by: James Bottomley --- drivers/scsi/libfc/fc_disc.c | 6 +++--- drivers/scsi/libfc/fc_elsct.c | 1 + drivers/scsi/libfc/fc_exch.c | 2 +- drivers/scsi/libfc/fc_lport.c | 5 +++-- drivers/scsi/libfc/fc_rport.c | 10 +++++----- include/scsi/libfc.h | 2 +- 6 files changed, 14 insertions(+), 12 deletions(-) diff --git a/drivers/scsi/libfc/fc_disc.c b/drivers/scsi/libfc/fc_disc.c index 7269e928824a..1d1b0c9da29b 100644 --- a/drivers/scsi/libfc/fc_disc.c +++ b/drivers/scsi/libfc/fc_disc.c @@ -61,7 +61,7 @@ static void fc_disc_restart(struct fc_disc *); * Locking Note: This function expects that the lport mutex is locked before * calling it. */ -void fc_disc_stop_rports(struct fc_disc *disc) +static void fc_disc_stop_rports(struct fc_disc *disc) { struct fc_lport *lport; struct fc_rport_priv *rdata; @@ -682,7 +682,7 @@ static int fc_disc_single(struct fc_lport *lport, struct fc_disc_port *dp) * fc_disc_stop() - Stop discovery for a given lport * @lport: The local port that discovery should stop on */ -void fc_disc_stop(struct fc_lport *lport) +static void fc_disc_stop(struct fc_lport *lport) { struct fc_disc *disc = &lport->disc; @@ -698,7 +698,7 @@ void fc_disc_stop(struct fc_lport *lport) * This function will block until discovery has been * completely stopped and all rports have been deleted. */ -void fc_disc_stop_final(struct fc_lport *lport) +static void fc_disc_stop_final(struct fc_lport *lport) { fc_disc_stop(lport); lport->tt.rport_flush_queue(); diff --git a/drivers/scsi/libfc/fc_elsct.c b/drivers/scsi/libfc/fc_elsct.c index fb9161dc4ca6..e17a28d324d0 100644 --- a/drivers/scsi/libfc/fc_elsct.c +++ b/drivers/scsi/libfc/fc_elsct.c @@ -28,6 +28,7 @@ #include #include #include +#include "fc_libfc.h" /** * fc_elsct_send() - Send an ELS or CT frame diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c index 9de9db27e874..4d70d96fa5dc 100644 --- a/drivers/scsi/libfc/fc_exch.c +++ b/drivers/scsi/libfc/fc_exch.c @@ -91,7 +91,7 @@ struct fc_exch_pool { * It manages the allocation of exchange IDs. */ struct fc_exch_mgr { - struct fc_exch_pool *pool; + struct fc_exch_pool __percpu *pool; mempool_t *ep_pool; enum fc_class class; struct kref kref; diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c index e77094a587ed..83750ebb527f 100644 --- a/drivers/scsi/libfc/fc_lport.c +++ b/drivers/scsi/libfc/fc_lport.c @@ -677,7 +677,8 @@ EXPORT_SYMBOL(fc_set_mfs); * @lport: The local port receiving the event * @event: The discovery event */ -void fc_lport_disc_callback(struct fc_lport *lport, enum fc_disc_event event) +static void fc_lport_disc_callback(struct fc_lport *lport, + enum fc_disc_event event) { switch (event) { case DISC_EV_SUCCESS: @@ -1568,7 +1569,7 @@ EXPORT_SYMBOL(fc_lport_flogi_resp); * Locking Note: The lport lock is expected to be held before calling * this routine. */ -void fc_lport_enter_flogi(struct fc_lport *lport) +static void fc_lport_enter_flogi(struct fc_lport *lport) { struct fc_frame *fp; diff --git a/drivers/scsi/libfc/fc_rport.c b/drivers/scsi/libfc/fc_rport.c index b9e434844a69..83aa1efec875 100644 --- a/drivers/scsi/libfc/fc_rport.c +++ b/drivers/scsi/libfc/fc_rport.c @@ -391,7 +391,7 @@ static void fc_rport_work(struct work_struct *work) * If it appears we are already logged in, ADISC is used to verify * the setup. */ -int fc_rport_login(struct fc_rport_priv *rdata) +static int fc_rport_login(struct fc_rport_priv *rdata) { mutex_lock(&rdata->rp_mutex); @@ -451,7 +451,7 @@ static void fc_rport_enter_delete(struct fc_rport_priv *rdata, * function will hold the rport lock, call an _enter_* * function and then unlock the rport. */ -int fc_rport_logoff(struct fc_rport_priv *rdata) +static int fc_rport_logoff(struct fc_rport_priv *rdata) { mutex_lock(&rdata->rp_mutex); @@ -653,8 +653,8 @@ static int fc_rport_login_complete(struct fc_rport_priv *rdata, * @fp: The FLOGI response frame * @rp_arg: The remote port that received the FLOGI response */ -void fc_rport_flogi_resp(struct fc_seq *sp, struct fc_frame *fp, - void *rp_arg) +static void fc_rport_flogi_resp(struct fc_seq *sp, struct fc_frame *fp, + void *rp_arg) { struct fc_rport_priv *rdata = rp_arg; struct fc_lport *lport = rdata->local_port; @@ -1520,7 +1520,7 @@ reject: * * Locking Note: Called with the lport lock held. */ -void fc_rport_recv_req(struct fc_lport *lport, struct fc_frame *fp) +static void fc_rport_recv_req(struct fc_lport *lport, struct fc_frame *fp) { struct fc_seq_els_data els_data; diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h index 5d1a758e0595..6a3922fe0be0 100644 --- a/include/scsi/libfc.h +++ b/include/scsi/libfc.h @@ -857,7 +857,7 @@ struct fc_lport { enum fc_lport_state state; unsigned long boot_time; struct fc_host_statistics host_stats; - struct fcoe_dev_stats *dev_stats; + struct fcoe_dev_stats __percpu *dev_stats; u8 retry_count; /* Fabric information */ From 7c9c684160bc2c6668abbd2701b440e18bb9ef35 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Fri, 13 Jan 2012 17:26:25 -0800 Subject: [PATCH 326/820] [SCSI] fcoe: Move fcoe_debug_logging from fcoe.h to fcoe.c Move the definition of the global variable fcoe_debug_logging from fcoe.h to fcoe.c. Avoid that sparse complains about missing declarations for local functions or variables by declaring these static. Signed-off-by: Bart Van Assche Reviewed-by: Yi Zou Signed-off-by: Robert Love Signed-off-by: James Bottomley --- drivers/scsi/fcoe/fcoe.c | 34 +++++++++++++++++++--------------- drivers/scsi/fcoe/fcoe.h | 4 +--- 2 files changed, 20 insertions(+), 18 deletions(-) diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c index 507504b77880..e9599600aa23 100644 --- a/drivers/scsi/fcoe/fcoe.c +++ b/drivers/scsi/fcoe/fcoe.c @@ -58,7 +58,11 @@ module_param_named(ddp_min, fcoe_ddp_min, uint, S_IRUGO | S_IWUSR); MODULE_PARM_DESC(ddp_min, "Minimum I/O size in bytes for " \ "Direct Data Placement (DDP)."); -DEFINE_MUTEX(fcoe_config_mutex); +unsigned int fcoe_debug_logging; +module_param_named(debug_logging, fcoe_debug_logging, int, S_IRUGO|S_IWUSR); +MODULE_PARM_DESC(debug_logging, "a bit mask of logging levels"); + +static DEFINE_MUTEX(fcoe_config_mutex); static struct workqueue_struct *fcoe_wq; @@ -67,8 +71,8 @@ static DECLARE_COMPLETION(fcoe_flush_completion); /* fcoe host list */ /* must only by accessed under the RTNL mutex */ -LIST_HEAD(fcoe_hostlist); -DEFINE_PER_CPU(struct fcoe_percpu_s, fcoe_percpu); +static LIST_HEAD(fcoe_hostlist); +static DEFINE_PER_CPU(struct fcoe_percpu_s, fcoe_percpu); /* Function Prototypes */ static int fcoe_reset(struct Scsi_Host *); @@ -157,7 +161,7 @@ static struct libfc_function_template fcoe_libfc_fcn_templ = { .lport_set_port_id = fcoe_set_port_id, }; -struct fc_function_template fcoe_nport_fc_functions = { +static struct fc_function_template fcoe_nport_fc_functions = { .show_host_node_name = 1, .show_host_port_name = 1, .show_host_supported_classes = 1, @@ -197,7 +201,7 @@ struct fc_function_template fcoe_nport_fc_functions = { .bsg_request = fc_lport_bsg_request, }; -struct fc_function_template fcoe_vport_fc_functions = { +static struct fc_function_template fcoe_vport_fc_functions = { .show_host_node_name = 1, .show_host_port_name = 1, .show_host_supported_classes = 1, @@ -433,7 +437,7 @@ static inline void fcoe_interface_put(struct fcoe_interface *fcoe) * * Caller must be holding the RTNL mutex */ -void fcoe_interface_cleanup(struct fcoe_interface *fcoe) +static void fcoe_interface_cleanup(struct fcoe_interface *fcoe) { struct net_device *netdev = fcoe->netdev; struct fcoe_ctlr *fip = &fcoe->ctlr; @@ -748,7 +752,7 @@ static int fcoe_shost_config(struct fc_lport *lport, struct device *dev) * * Returns: True for read types I/O, otherwise returns false. */ -bool fcoe_oem_match(struct fc_frame *fp) +static bool fcoe_oem_match(struct fc_frame *fp) { struct fc_frame_header *fh = fc_frame_header_get(fp); struct fcp_cmnd *fcp; @@ -1107,7 +1111,7 @@ static int __init fcoe_if_init(void) * * Returns: 0 on success */ -int __exit fcoe_if_exit(void) +static int __exit fcoe_if_exit(void) { fc_release_transport(fcoe_nport_scsi_transport); fc_release_transport(fcoe_vport_scsi_transport); @@ -1296,7 +1300,7 @@ static inline unsigned int fcoe_select_cpu(void) * * Returns: 0 for success */ -int fcoe_rcv(struct sk_buff *skb, struct net_device *netdev, +static int fcoe_rcv(struct sk_buff *skb, struct net_device *netdev, struct packet_type *ptype, struct net_device *olddev) { struct fc_lport *lport; @@ -1452,7 +1456,7 @@ static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen) * * Return: 0 for success */ -int fcoe_xmit(struct fc_lport *lport, struct fc_frame *fp) +static int fcoe_xmit(struct fc_lport *lport, struct fc_frame *fp) { int wlen; u32 crc; @@ -1728,7 +1732,7 @@ drop: * * Return: 0 for success */ -int fcoe_percpu_receive_thread(void *arg) +static int fcoe_percpu_receive_thread(void *arg) { struct fcoe_percpu_s *p = arg; struct sk_buff *skb; @@ -2146,7 +2150,7 @@ out_nortnl: * Returns: 0 if the ethtool query was successful * -1 if the ethtool query failed */ -int fcoe_link_speed_update(struct fc_lport *lport) +static int fcoe_link_speed_update(struct fc_lport *lport) { struct net_device *netdev = fcoe_netdev(lport); struct ethtool_cmd ecmd; @@ -2180,7 +2184,7 @@ int fcoe_link_speed_update(struct fc_lport *lport) * Returns: 0 if link is UP and OK, -1 if not * */ -int fcoe_link_ok(struct fc_lport *lport) +static int fcoe_link_ok(struct fc_lport *lport) { struct net_device *netdev = fcoe_netdev(lport); @@ -2200,7 +2204,7 @@ int fcoe_link_ok(struct fc_lport *lport) * there no packets that will be handled by the lport, but also that any * threads already handling packet have returned. */ -void fcoe_percpu_clean(struct fc_lport *lport) +static void fcoe_percpu_clean(struct fc_lport *lport) { struct fcoe_percpu_s *pp; struct fcoe_rcv_info *fr; @@ -2251,7 +2255,7 @@ void fcoe_percpu_clean(struct fc_lport *lport) * * Returns: Always 0 (return value required by FC transport template) */ -int fcoe_reset(struct Scsi_Host *shost) +static int fcoe_reset(struct Scsi_Host *shost) { struct fc_lport *lport = shost_priv(shost); struct fcoe_port *port = lport_priv(lport); diff --git a/drivers/scsi/fcoe/fcoe.h b/drivers/scsi/fcoe/fcoe.h index 6c6884bcf840..bcc89e639495 100644 --- a/drivers/scsi/fcoe/fcoe.h +++ b/drivers/scsi/fcoe/fcoe.h @@ -40,9 +40,7 @@ #define FCOE_MIN_XID 0x0000 /* the min xid supported by fcoe_sw */ #define FCOE_MAX_XID 0x0FFF /* the max xid supported by fcoe_sw */ -unsigned int fcoe_debug_logging; -module_param_named(debug_logging, fcoe_debug_logging, int, S_IRUGO|S_IWUSR); -MODULE_PARM_DESC(debug_logging, "a bit mask of logging levels"); +extern unsigned int fcoe_debug_logging; #define FCOE_LOGGING 0x01 /* General logging, not categorized */ #define FCOE_NETDEV_LOGGING 0x02 /* Netdevice logging */ From 76ffe8a3f766358a0ade543153625b3e4e66159d Mon Sep 17 00:00:00 2001 From: Yi Zou Date: Fri, 13 Jan 2012 17:26:30 -0800 Subject: [PATCH 327/820] [SCSI] libfc: remove redundant timer init for fcp The fcp timer is already initialized when it gets allocated. Signed-off-by: Yi Zou Tested-by: Ross Brattain Signed-off-by: Robert Love Signed-off-by: James Bottomley --- drivers/scsi/libfc/fc_fcp.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c index 221875ec3d7c..f607314810ac 100644 --- a/drivers/scsi/libfc/fc_fcp.c +++ b/drivers/scsi/libfc/fc_fcp.c @@ -155,6 +155,7 @@ static struct fc_fcp_pkt *fc_fcp_pkt_alloc(struct fc_lport *lport, gfp_t gfp) fsp->xfer_ddp = FC_XID_UNKNOWN; atomic_set(&fsp->ref_cnt, 1); init_timer(&fsp->timer); + fsp->timer.data = (unsigned long)fsp; INIT_LIST_HEAD(&fsp->list); spin_lock_init(&fsp->scsi_pkt_lock); } @@ -1850,9 +1851,6 @@ int fc_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc_cmd) } put_cpu(); - init_timer(&fsp->timer); - fsp->timer.data = (unsigned long)fsp; - /* * send it to the lower layer * if we get -1 return then put the request in the pending From ffe535edb9a9c5b4d5fe03dfa3d89a1495580f1b Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Mon, 16 Jan 2012 10:52:20 +0100 Subject: [PATCH 328/820] ALSA: HDA: Fix internal microphone on Dell Studio 16 XPS 1645 More than one user reports that changing the model from "both" to "dmic" makes their Internal Mic work. Cc: stable@kernel.org Tested-by: Martin Ling BugLink: https://bugs.launchpad.net/bugs/795823 Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_sigmatel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 87e684fa830f..3556408d6ece 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -1596,7 +1596,7 @@ static const struct snd_pci_quirk stac92hd73xx_cfg_tbl[] = { SND_PCI_QUIRK(PCI_VENDOR_ID_DELL, 0x02bd, "Dell Studio 1557", STAC_DELL_M6_DMIC), SND_PCI_QUIRK(PCI_VENDOR_ID_DELL, 0x02fe, - "Dell Studio XPS 1645", STAC_DELL_M6_BOTH), + "Dell Studio XPS 1645", STAC_DELL_M6_DMIC), SND_PCI_QUIRK(PCI_VENDOR_ID_DELL, 0x0413, "Dell Studio 1558", STAC_DELL_M6_DMIC), {} /* terminator */ From 7b32486ca449a20d64fa381262cc5e56893f0ac1 Mon Sep 17 00:00:00 2001 From: Raymond Yau Date: Fri, 13 Jan 2012 10:30:08 +0800 Subject: [PATCH 329/820] ALSA: Au88x0 - Support 4 channels playback when AC97 codecs has SDAC bit - Check SDAC bit of AC97 codec for supporting 4 channels playback. Signed-off-by: Raymond Yau Signed-off-by: Takashi Iwai --- sound/pci/au88x0/au88x0.c | 10 +++++----- sound/pci/au88x0/au88x0_pcm.c | 1 + 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/sound/pci/au88x0/au88x0.c b/sound/pci/au88x0/au88x0.c index 762bb108c51c..b8bc11556ed0 100644 --- a/sound/pci/au88x0/au88x0.c +++ b/sound/pci/au88x0/au88x0.c @@ -268,6 +268,11 @@ snd_vortex_probe(struct pci_dev *pci, const struct pci_device_id *pci_id) card->shortname, chip->io, chip->irq); // (4) Alloc components. + err = snd_vortex_mixer(chip); + if (err < 0) { + snd_card_free(card); + return err; + } // ADB pcm. if ((err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_ADB)) < 0) { snd_card_free(card); @@ -299,11 +304,6 @@ snd_vortex_probe(struct pci_dev *pci, const struct pci_device_id *pci_id) return err; } #endif - // snd_ac97_mixer and Vortex mixer. - if ((err = snd_vortex_mixer(chip)) < 0) { - snd_card_free(card); - return err; - } if ((err = snd_vortex_midi(chip)) < 0) { snd_card_free(card); return err; diff --git a/sound/pci/au88x0/au88x0_pcm.c b/sound/pci/au88x0/au88x0_pcm.c index 0488633ea874..0ef2f9712208 100644 --- a/sound/pci/au88x0/au88x0_pcm.c +++ b/sound/pci/au88x0/au88x0_pcm.c @@ -168,6 +168,7 @@ static int snd_vortex_pcm_open(struct snd_pcm_substream *substream) runtime->hw = snd_vortex_playback_hw_adb; #ifdef CHIP_AU8830 if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK && + VORTEX_IS_QUAD(vortex) && VORTEX_PCM_TYPE(substream->pcm) == VORTEX_PCM_ADB) { runtime->hw.channels_max = 4; snd_pcm_hw_constraint_list(runtime, 0, From 49b9c40e9b5e47ef64f8601b84981fa45c3c4f86 Mon Sep 17 00:00:00 2001 From: Raymond Yau Date: Fri, 13 Jan 2012 10:35:01 +0800 Subject: [PATCH 330/820] ALSA: Au88x0 - Reduce the number of playback subdevices of au8830 from 32 to 16 - The maximum number of playback streams depend on the number of sample rate conveters (16) and the number of DMA channels (32). Signed-off-by: Raymond Yau Signed-off-by: Takashi Iwai --- sound/pci/au88x0/au88x0.c | 3 ++- sound/pci/au88x0/au88x0.h | 1 + 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/sound/pci/au88x0/au88x0.c b/sound/pci/au88x0/au88x0.c index b8bc11556ed0..f13ad536b2d5 100644 --- a/sound/pci/au88x0/au88x0.c +++ b/sound/pci/au88x0/au88x0.c @@ -274,7 +274,8 @@ snd_vortex_probe(struct pci_dev *pci, const struct pci_device_id *pci_id) return err; } // ADB pcm. - if ((err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_ADB)) < 0) { + err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_PCM); + if (err < 0) { snd_card_free(card); return err; } diff --git a/sound/pci/au88x0/au88x0.h b/sound/pci/au88x0/au88x0.h index 02f6e08f7592..bb938153a964 100644 --- a/sound/pci/au88x0/au88x0.h +++ b/sound/pci/au88x0/au88x0.h @@ -105,6 +105,7 @@ #define MIX_SPDIF(x) (vortex->mixspdif[x]) #define NR_WTPB 0x20 /* WT channels per each bank. */ +#define NR_PCM 0x10 /* Structs */ typedef struct { From f0e48b6bd4e407459715240cd241ddb6b89bdf81 Mon Sep 17 00:00:00 2001 From: Clemens Ladisch Date: Sat, 14 Jan 2012 16:42:24 +0100 Subject: [PATCH 331/820] ALSA: virtuoso: Xonar DS: fix polarity of front output The two DACs for the front output and the surround/center/LFE/back outputs are wired up out of phase, so when channels are duplicated, their sound can cancel out each other and result in a weaker bass response. To fix this, reverse the polarity of the neutron flow to the front output. Reported-any-tested-by: Daniel Hill Signed-off-by: Clemens Ladisch Cc: 2.6.34+ Signed-off-by: Takashi Iwai --- sound/pci/oxygen/xonar_wm87x6.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/oxygen/xonar_wm87x6.c b/sound/pci/oxygen/xonar_wm87x6.c index 478303e6c2b0..63cff90706bf 100644 --- a/sound/pci/oxygen/xonar_wm87x6.c +++ b/sound/pci/oxygen/xonar_wm87x6.c @@ -177,6 +177,7 @@ static void wm8776_registers_init(struct oxygen *chip) struct xonar_wm87x6 *data = chip->model_data; wm8776_write(chip, WM8776_RESET, 0); + wm8776_write(chip, WM8776_PHASESWAP, WM8776_PH_MASK); wm8776_write(chip, WM8776_DACCTRL1, WM8776_DZCEN | WM8776_PL_LEFT_LEFT | WM8776_PL_RIGHT_RIGHT); wm8776_write(chip, WM8776_DACMUTE, chip->dac_mute ? WM8776_DMUTE : 0); From 0cb64f02b8c8f8181d6eaa6d1931fe0bc4195a25 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 10 Jan 2012 05:04:39 -0300 Subject: [PATCH 332/820] [media] cx23885: handle errors from videobuf_dvb_get_frontend() The error handling in the original code wasn't complete so static checkers complained about a potential NULL deference. Signed-off-by: Dan Carpenter Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/cx23885/cx23885-video.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/media/video/cx23885/cx23885-video.c b/drivers/media/video/cx23885/cx23885-video.c index 4bbf9bb97bde..c654bdc7ccb2 100644 --- a/drivers/media/video/cx23885/cx23885-video.c +++ b/drivers/media/video/cx23885/cx23885-video.c @@ -1550,7 +1550,6 @@ static int cx23885_set_freq_via_ops(struct cx23885_dev *dev, struct v4l2_control ctrl; struct videobuf_dvb_frontend *vfe; struct dvb_frontend *fe; - int err = 0; struct analog_parameters params = { .mode = V4L2_TUNER_ANALOG_TV, @@ -1572,8 +1571,10 @@ static int cx23885_set_freq_via_ops(struct cx23885_dev *dev, params.frequency, f->tuner, params.std); vfe = videobuf_dvb_get_frontend(&dev->ts2.frontends, 1); - if (!vfe) - err = -EINVAL; + if (!vfe) { + mutex_unlock(&dev->lock); + return -EINVAL; + } fe = vfe->dvb.frontend; From 05834fcf58fe77b0519a8dc40e3c4a5a28dfc88d Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 22 Nov 2011 08:59:34 -0300 Subject: [PATCH 333/820] [media] v4l2 spec: clarify usage of V4L2_FBUF_FLAG_OVERLAY Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- .../DocBook/media/v4l/vidioc-g-fbuf.xml | 23 +++++++++++-------- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/Documentation/DocBook/media/v4l/vidioc-g-fbuf.xml b/Documentation/DocBook/media/v4l/vidioc-g-fbuf.xml index 93817f337033..7c63815e7afd 100644 --- a/Documentation/DocBook/media/v4l/vidioc-g-fbuf.xml +++ b/Documentation/DocBook/media/v4l/vidioc-g-fbuf.xml @@ -364,15 +364,20 @@ capability and it is cleared otherwise. V4L2_FBUF_FLAG_OVERLAY 0x0002 - The frame buffer is an overlay surface the same -size as the capture. [?] - - - The purpose of -V4L2_FBUF_FLAG_OVERLAY was never quite clear. -Most drivers seem to ignore this flag. For compatibility with the -bttv driver applications should set the -V4L2_FBUF_FLAG_OVERLAY flag. + If this flag is set for a video capture device, then the +driver will set the initial overlay size to cover the full framebuffer size, +otherwise the existing overlay size (as set by &VIDIOC-S-FMT;) will be used. + +Only one video capture driver (bttv) supports this flag. The use of this flag +for capture devices is deprecated. There is no way to detect which drivers +support this flag, so the only reliable method of setting the overlay size is +through &VIDIOC-S-FMT;. + +If this flag is set for a video output device, then the video output overlay +window is relative to the top-left corner of the framebuffer and restricted +to the size of the framebuffer. If it is cleared, then the video output +overlay window is relative to the video output display. + V4L2_FBUF_FLAG_CHROMAKEY From 20c9fe895a2356924a31a1238f75cf0d05704593 Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 22 Nov 2011 08:48:47 -0300 Subject: [PATCH 334/820] [media] zoran: do not set V4L2_FBUF_FLAG_OVERLAY The zoran driver does not support this flag, so don't set it. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/zoran/zoran_driver.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/media/video/zoran/zoran_driver.c b/drivers/media/video/zoran/zoran_driver.c index d4d05d2ace65..f7d236a2402a 100644 --- a/drivers/media/video/zoran/zoran_driver.c +++ b/drivers/media/video/zoran/zoran_driver.c @@ -1958,7 +1958,6 @@ static int zoran_g_fbuf(struct file *file, void *__fh, mutex_unlock(&zr->resource_lock); fb->fmt.colorspace = V4L2_COLORSPACE_SRGB; fb->fmt.field = V4L2_FIELD_INTERLACED; - fb->flags = V4L2_FBUF_FLAG_OVERLAY; fb->capability = V4L2_FBUF_CAP_LIST_CLIPPING; return 0; From 047a01fdc807ffa461db87e2d30d6d28f95c091a Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 22 Nov 2011 08:23:19 -0300 Subject: [PATCH 335/820] [media] omap_vout: add missing OVERLAY_OUTPUT cap and set V4L2_FBUF_FLAG_OVERLAY The omap_vout driver has an output overlay, but never advertised that capability. The driver should also set the V4L2_FBUF_FLAG_OVERLAY flag. Signed-off-by: Hans Verkuil CC: Archit Taneja CC: Vaibhav Hiremath Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/omap/omap_vout.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/media/video/omap/omap_vout.c b/drivers/media/video/omap/omap_vout.c index 00d6d192b88c..1e7c0d6205e6 100644 --- a/drivers/media/video/omap/omap_vout.c +++ b/drivers/media/video/omap/omap_vout.c @@ -1046,7 +1046,8 @@ static int vidioc_querycap(struct file *file, void *fh, strlcpy(cap->driver, VOUT_NAME, sizeof(cap->driver)); strlcpy(cap->card, vout->vfd->name, sizeof(cap->card)); cap->bus_info[0] = '\0'; - cap->capabilities = V4L2_CAP_STREAMING | V4L2_CAP_VIDEO_OUTPUT; + cap->capabilities = V4L2_CAP_STREAMING | V4L2_CAP_VIDEO_OUTPUT | + V4L2_CAP_VIDEO_OUTPUT_OVERLAY; return 0; } @@ -1828,7 +1829,9 @@ static int vidioc_g_fbuf(struct file *file, void *fh, ovid = &vout->vid_info; ovl = ovid->overlays[0]; - a->flags = 0x0; + /* The video overlay must stay within the framebuffer and can't be + positioned independently. */ + a->flags = V4L2_FBUF_FLAG_OVERLAY; a->capability = V4L2_FBUF_CAP_LOCAL_ALPHA | V4L2_FBUF_CAP_CHROMAKEY | V4L2_FBUF_CAP_SRC_CHROMAKEY; From 6055b12ae0f433314cba98f423fb8c5fc211ca8d Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Wed, 11 Jan 2012 07:30:55 -0300 Subject: [PATCH 336/820] [media] V4L2 Spec: fix extended control documentation Update the spec to the behavior implemented by the control framework. This should have been documented long ago but for some reason it was never done. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- .../DocBook/media/v4l/vidioc-g-ext-ctrls.xml | 18 +++++++++++----- Documentation/video4linux/v4l2-controls.txt | 21 ------------------- 2 files changed, 13 insertions(+), 26 deletions(-) diff --git a/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml b/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml index 6f1f9a629dc3..b17a7aac6997 100644 --- a/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml +++ b/Documentation/DocBook/media/v4l/vidioc-g-ext-ctrls.xml @@ -183,7 +183,12 @@ applications must set the array to zero. __u32 ctrl_class The control class to which all controls belong, see -. +. Drivers that use a kernel framework for handling +controls will also accept a value of 0 here, meaning that the controls can +belong to any control class. Whether drivers support this can be tested by setting +ctrl_class to 0 and calling VIDIOC_TRY_EXT_CTRLS +with a count of 0. If that succeeds, then the driver +supports this feature. __u32 @@ -194,10 +199,13 @@ also be zero. __u32 error_idx - Set by the driver in case of an error. It is the -index of the control causing the error or equal to 'count' when the -error is not associated with a particular control. Undefined when the -ioctl returns 0 (success). + Set by the driver in case of an error. If it is equal +to count, then no actual changes were made to +controls. In other words, the error was not associated with setting a particular +control. If it is another value, then only the controls up to error_idx-1 +were modified and control error_idx is the one that +caused the error. The error_idx value is undefined +if the ioctl returned 0 (success). __u32 diff --git a/Documentation/video4linux/v4l2-controls.txt b/Documentation/video4linux/v4l2-controls.txt index 26aa0573933e..e2492a9d1027 100644 --- a/Documentation/video4linux/v4l2-controls.txt +++ b/Documentation/video4linux/v4l2-controls.txt @@ -666,27 +666,6 @@ a control of this type whenever the first control belonging to a new control class is added. -Differences from the Spec -========================= - -There are a few places where the framework acts slightly differently from the -V4L2 Specification. Those differences are described in this section. We will -have to see whether we need to adjust the spec or not. - -1) It is no longer required to have all controls contained in a -v4l2_ext_control array be from the same control class. The framework will be -able to handle any type of control in the array. You need to set ctrl_class -to 0 in order to enable this. If ctrl_class is non-zero, then it will still -check that all controls belong to that control class. - -If you set ctrl_class to 0 and count to 0, then it will only return an error -if there are no controls at all. - -2) Clarified the way error_idx works. For get and set it will be equal to -count if nothing was done yet. If it is less than count then only the controls -up to error_idx-1 were successfully applied. - - Proposals for Extensions ======================== From dacd4fa37dbafe209a90d6d5b729d2257df4b734 Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Wed, 11 Jan 2012 07:37:54 -0300 Subject: [PATCH 337/820] [media] V4L2 Spec: improve the G/S_INPUT/OUTPUT documentation As Rupert pointed out, the phrase "It is good practice" should be replaced with "You must". Signed-off-by: Hans Verkuil Cc: Rupert Eibauer Signed-off-by: Mauro Carvalho Chehab --- Documentation/DocBook/media/v4l/vidioc-g-input.xml | 4 ++-- Documentation/DocBook/media/v4l/vidioc-g-output.xml | 5 +++-- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/Documentation/DocBook/media/v4l/vidioc-g-input.xml b/Documentation/DocBook/media/v4l/vidioc-g-input.xml index 08ae82f131f2..1d43065090dd 100644 --- a/Documentation/DocBook/media/v4l/vidioc-g-input.xml +++ b/Documentation/DocBook/media/v4l/vidioc-g-input.xml @@ -61,8 +61,8 @@ desired input in an integer and call the VIDIOC_S_INPUT ioctl with a pointer to this integer. Side effects are possible. For example inputs may support different video standards, so the driver may implicitly switch the -current standard. It is good practice to select an input before -querying or negotiating any other parameters. +current standard. Because of these possible side effects applications +must select an input before querying or negotiating any other parameters. Information about video inputs is available using the &VIDIOC-ENUMINPUT; ioctl. diff --git a/Documentation/DocBook/media/v4l/vidioc-g-output.xml b/Documentation/DocBook/media/v4l/vidioc-g-output.xml index fd45f1c13ccf..4533068ecb8a 100644 --- a/Documentation/DocBook/media/v4l/vidioc-g-output.xml +++ b/Documentation/DocBook/media/v4l/vidioc-g-output.xml @@ -61,8 +61,9 @@ desired output in an integer and call the VIDIOC_S_OUTPUT ioctl with a pointer to this integer. Side effects are possible. For example outputs may support different video standards, so the driver may implicitly switch the current -standard. It is good practice to select an output before querying or -negotiating any other parameters. +standard. +standard. Because of these possible side effects applications +must select an output before querying or negotiating any other parameters. Information about video outputs is available using the &VIDIOC-ENUMOUTPUT; ioctl. From aa07eec5320cf1793062c03b8409e7541ca37c31 Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 11 Oct 2011 05:12:52 -0300 Subject: [PATCH 338/820] [media] v4l2-ioctl: make tuner 'type' check more strict for S_FREQUENCY As per the feature removal document, make the tuner type check more strict so that it is no longer possible to set the radio frequency through a video node or the TV frequency through a radio node. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- .../DocBook/media/v4l/vidioc-g-frequency.xml | 7 +++++-- Documentation/feature-removal-schedule.txt | 11 ----------- drivers/media/video/v4l2-ioctl.c | 8 +++++++- 3 files changed, 12 insertions(+), 14 deletions(-) diff --git a/Documentation/DocBook/media/v4l/vidioc-g-frequency.xml b/Documentation/DocBook/media/v4l/vidioc-g-frequency.xml index 16431813bebd..66e9a5257861 100644 --- a/Documentation/DocBook/media/v4l/vidioc-g-frequency.xml +++ b/Documentation/DocBook/media/v4l/vidioc-g-frequency.xml @@ -98,8 +98,11 @@ the &v4l2-output; modulator field and the &v4l2-tuner-type; type The tuner type. This is the same value as in the -&v4l2-tuner; type field. The field is not -applicable to modulators, &ie; ignored by drivers. +&v4l2-tuner; type field. The type must be set +to V4L2_TUNER_RADIO for /dev/radioX +device nodes, and to V4L2_TUNER_ANALOG_TV +for all others. The field is not applicable to modulators, &ie; ignored +by drivers. __u32 diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index c4362ef3a793..fed5a34a727d 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -460,17 +460,6 @@ Who: Jean Delvare ---------------------------- -What: For VIDIOC_S_FREQUENCY the type field must match the device node's type. - If not, return -EINVAL. -When: 3.2 -Why: It makes no sense to switch the tuner to radio mode by calling - VIDIOC_S_FREQUENCY on a video node, or to switch the tuner to tv mode by - calling VIDIOC_S_FREQUENCY on a radio node. This is the first step of a - move to more consistent handling of tv and radio tuners. -Who: Hans Verkuil - ----------------------------- - What: Opening a radio device node will no longer automatically switch the tuner mode from tv to radio. When: 3.3 diff --git a/drivers/media/video/v4l2-ioctl.c b/drivers/media/video/v4l2-ioctl.c index 77feeb67e2db..3f623859a337 100644 --- a/drivers/media/video/v4l2-ioctl.c +++ b/drivers/media/video/v4l2-ioctl.c @@ -1871,6 +1871,7 @@ static long __video_do_ioctl(struct file *file, case VIDIOC_S_FREQUENCY: { struct v4l2_frequency *p = arg; + enum v4l2_tuner_type type; if (!ops->vidioc_s_frequency) break; @@ -1878,9 +1879,14 @@ static long __video_do_ioctl(struct file *file, ret = ret_prio; break; } + type = (vfd->vfl_type == VFL_TYPE_RADIO) ? + V4L2_TUNER_RADIO : V4L2_TUNER_ANALOG_TV; dbgarg(cmd, "tuner=%d, type=%d, frequency=%d\n", p->tuner, p->type, p->frequency); - ret = ops->vidioc_s_frequency(file, fh, p); + if (p->type != type) + ret = -EINVAL; + else + ret = ops->vidioc_s_frequency(file, fh, p); break; } case VIDIOC_G_SLICED_VBI_CAP: From 3f3edd7c97d0a5ae355aace36000576acbc2f2cc Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 11 Oct 2011 05:18:08 -0300 Subject: [PATCH 339/820] [media] ivtv: remove exclusive radio open For some reason the /dev/radio device was implemented as an exclusive open: you could open it only once and not a second time. Remove this limitation. Signed-off-by: Hans Verkuil Acked-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/ivtv/ivtv-fileops.c | 57 +++++++++++-------------- 1 file changed, 24 insertions(+), 33 deletions(-) diff --git a/drivers/media/video/ivtv/ivtv-fileops.c b/drivers/media/video/ivtv/ivtv-fileops.c index 38f052257f46..69a3de9bae62 100644 --- a/drivers/media/video/ivtv/ivtv-fileops.c +++ b/drivers/media/video/ivtv/ivtv-fileops.c @@ -861,20 +861,10 @@ int ivtv_v4l2_close(struct file *filp) IVTV_DEBUG_FILE("close %s\n", s->name); - v4l2_fh_del(fh); - v4l2_fh_exit(fh); - - /* Easy case first: this stream was never claimed by us */ - if (s->id != id->open_id) { - kfree(id); - return 0; - } - - /* 'Unclaim' this stream */ - - /* Stop radio */ mutex_lock(&itv->serialize_lock); - if (id->type == IVTV_ENC_STREAM_TYPE_RAD) { + /* Stop radio */ + if (id->type == IVTV_ENC_STREAM_TYPE_RAD && + v4l2_fh_is_singular_file(filp)) { /* Closing radio device, return to TV mode */ ivtv_mute(itv); /* Mark that the radio is no longer in use */ @@ -890,13 +880,26 @@ int ivtv_v4l2_close(struct file *filp) if (atomic_read(&itv->capturing) > 0) { /* Undo video mute */ ivtv_vapi(itv, CX2341X_ENC_MUTE_VIDEO, 1, - v4l2_ctrl_g_ctrl(itv->cxhdl.video_mute) | - (v4l2_ctrl_g_ctrl(itv->cxhdl.video_mute_yuv) << 8)); + v4l2_ctrl_g_ctrl(itv->cxhdl.video_mute) | + (v4l2_ctrl_g_ctrl(itv->cxhdl.video_mute_yuv) << 8)); } /* Done! Unmute and continue. */ ivtv_unmute(itv); - ivtv_release_stream(s); - } else if (s->type >= IVTV_DEC_STREAM_TYPE_MPG) { + } + + v4l2_fh_del(fh); + v4l2_fh_exit(fh); + + /* Easy case first: this stream was never claimed by us */ + if (s->id != id->open_id) { + kfree(id); + mutex_unlock(&itv->serialize_lock); + return 0; + } + + /* 'Unclaim' this stream */ + + if (s->type >= IVTV_DEC_STREAM_TYPE_MPG) { struct ivtv_stream *s_vout = &itv->streams[IVTV_DEC_STREAM_TYPE_VOUT]; ivtv_stop_decoding(id, VIDEO_CMD_STOP_TO_BLACK | VIDEO_CMD_STOP_IMMEDIATELY, 0); @@ -966,31 +969,20 @@ static int ivtv_serialized_open(struct ivtv_stream *s, struct file *filp) return -ENOMEM; } v4l2_fh_init(&item->fh, s->vdev); - if (res < 0) { - v4l2_fh_exit(&item->fh); - kfree(item); - return res; - } item->itv = itv; item->type = s->type; item->open_id = itv->open_id++; filp->private_data = &item->fh; + v4l2_fh_add(&item->fh); - if (item->type == IVTV_ENC_STREAM_TYPE_RAD) { - /* Try to claim this stream */ - if (ivtv_claim_stream(item, item->type)) { - /* No, it's already in use */ - v4l2_fh_exit(&item->fh); - kfree(item); - return -EBUSY; - } - + if (item->type == IVTV_ENC_STREAM_TYPE_RAD && + v4l2_fh_is_singular_file(filp)) { if (!test_bit(IVTV_F_I_RADIO_USER, &itv->i_flags)) { if (atomic_read(&itv->capturing) > 0) { /* switching to radio while capture is in progress is not polite */ - ivtv_release_stream(s); + v4l2_fh_del(&item->fh); v4l2_fh_exit(&item->fh); kfree(item); return -EBUSY; @@ -1022,7 +1014,6 @@ static int ivtv_serialized_open(struct ivtv_stream *s, struct file *filp) 1080 * ((itv->yuv_info.v4l2_src_h + 31) & ~31); itv->yuv_info.stream_size = 0; } - v4l2_fh_add(&item->fh); return 0; } From 4d68e700d6a192a5a8b394b26ac056a1c0fa6ebc Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 11 Oct 2011 05:23:36 -0300 Subject: [PATCH 340/820] [media] cx18: remove exclusive open of radio device For some reason the cx18 driver could open the radio device only once. Remove this limitation. Signed-off-by: Hans Verkuil Acked-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/cx18/cx18-fileops.c | 41 +++++++++---------------- 1 file changed, 14 insertions(+), 27 deletions(-) diff --git a/drivers/media/video/cx18/cx18-fileops.c b/drivers/media/video/cx18/cx18-fileops.c index 14cb961c22bd..4bfd865a4106 100644 --- a/drivers/media/video/cx18/cx18-fileops.c +++ b/drivers/media/video/cx18/cx18-fileops.c @@ -751,20 +751,10 @@ int cx18_v4l2_close(struct file *filp) CX18_DEBUG_IOCTL("close() of %s\n", s->name); - v4l2_fh_del(fh); - v4l2_fh_exit(fh); - - /* Easy case first: this stream was never claimed by us */ - if (s->id != id->open_id) { - kfree(id); - return 0; - } - - /* 'Unclaim' this stream */ - - /* Stop radio */ mutex_lock(&cx->serialize_lock); - if (id->type == CX18_ENC_STREAM_TYPE_RAD) { + /* Stop radio */ + if (id->type == CX18_ENC_STREAM_TYPE_RAD && + v4l2_fh_is_singular_file(filp)) { /* Closing radio device, return to TV mode */ cx18_mute(cx); /* Mark that the radio is no longer in use */ @@ -781,10 +771,14 @@ int cx18_v4l2_close(struct file *filp) } /* Done! Unmute and continue. */ cx18_unmute(cx); - cx18_release_stream(s); - } else { - cx18_stop_capture(id, 0); } + + v4l2_fh_del(fh); + v4l2_fh_exit(fh); + + /* 'Unclaim' this stream */ + if (s->id == id->open_id) + cx18_stop_capture(id, 0); kfree(id); mutex_unlock(&cx->serialize_lock); return 0; @@ -810,21 +804,15 @@ static int cx18_serialized_open(struct cx18_stream *s, struct file *filp) item->open_id = cx->open_id++; filp->private_data = &item->fh; + v4l2_fh_add(&item->fh); - if (item->type == CX18_ENC_STREAM_TYPE_RAD) { - /* Try to claim this stream */ - if (cx18_claim_stream(item, item->type)) { - /* No, it's already in use */ - v4l2_fh_exit(&item->fh); - kfree(item); - return -EBUSY; - } - + if (item->type == CX18_ENC_STREAM_TYPE_RAD && + v4l2_fh_is_singular_file(filp)) { if (!test_bit(CX18_F_I_RADIO_USER, &cx->i_flags)) { if (atomic_read(&cx->ana_capturing) > 0) { /* switching to radio while capture is in progress is not polite */ - cx18_release_stream(s); + v4l2_fh_del(&item->fh); v4l2_fh_exit(&item->fh); kfree(item); return -EBUSY; @@ -842,7 +830,6 @@ static int cx18_serialized_open(struct cx18_stream *s, struct file *filp) /* Done! Unmute and continue. */ cx18_unmute(cx); } - v4l2_fh_add(&item->fh); return 0; } From cdc037817cc15caf931cd3476970860d62f1985c Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Tue, 11 Oct 2011 06:06:58 -0300 Subject: [PATCH 341/820] [media] ivtv: switch to the v4l core lock Signed-off-by: Hans Verkuil Acked-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/ivtv/ivtv-fileops.c | 48 ++++++++----------------- drivers/media/video/ivtv/ivtv-ioctl.c | 22 +++--------- drivers/media/video/ivtv/ivtv-streams.c | 1 + drivers/media/video/ivtv/ivtv-yuv.c | 22 ++++++++++-- 4 files changed, 39 insertions(+), 54 deletions(-) diff --git a/drivers/media/video/ivtv/ivtv-fileops.c b/drivers/media/video/ivtv/ivtv-fileops.c index 69a3de9bae62..116ece4d6f27 100644 --- a/drivers/media/video/ivtv/ivtv-fileops.c +++ b/drivers/media/video/ivtv/ivtv-fileops.c @@ -268,11 +268,13 @@ static struct ivtv_buffer *ivtv_get_buffer(struct ivtv_stream *s, int non_block, } /* wait for more data to arrive */ + mutex_unlock(&itv->serialize_lock); prepare_to_wait(&s->waitq, &wait, TASK_INTERRUPTIBLE); /* New buffers might have become available before we were added to the waitqueue */ if (!s->q_full.buffers) schedule(); finish_wait(&s->waitq, &wait); + mutex_lock(&itv->serialize_lock); if (signal_pending(current)) { /* return if a signal was received */ IVTV_DEBUG_INFO("User stopped %s\n", s->name); @@ -507,9 +509,7 @@ ssize_t ivtv_v4l2_read(struct file * filp, char __user *buf, size_t count, loff_ IVTV_DEBUG_HI_FILE("read %zd bytes from %s\n", count, s->name); - mutex_lock(&itv->serialize_lock); rc = ivtv_start_capture(id); - mutex_unlock(&itv->serialize_lock); if (rc) return rc; return ivtv_read_pos(s, buf, count, pos, filp->f_flags & O_NONBLOCK); @@ -584,9 +584,7 @@ ssize_t ivtv_v4l2_write(struct file *filp, const char __user *user_buf, size_t c set_bit(IVTV_F_S_APPL_IO, &s->s_flags); /* Start decoder (returns 0 if already started) */ - mutex_lock(&itv->serialize_lock); rc = ivtv_start_decoding(id, itv->speed); - mutex_unlock(&itv->serialize_lock); if (rc) { IVTV_DEBUG_WARN("Failed start decode stream %s\n", s->name); @@ -627,11 +625,13 @@ retry: break; if (filp->f_flags & O_NONBLOCK) return -EAGAIN; + mutex_unlock(&itv->serialize_lock); prepare_to_wait(&s->waitq, &wait, TASK_INTERRUPTIBLE); /* New buffers might have become free before we were added to the waitqueue */ if (!s->q_free.buffers) schedule(); finish_wait(&s->waitq, &wait); + mutex_lock(&itv->serialize_lock); if (signal_pending(current)) { IVTV_DEBUG_INFO("User stopped %s\n", s->name); return -EINTR; @@ -686,12 +686,14 @@ retry: if (mode == OUT_YUV) ivtv_yuv_setup_stream_frame(itv); + mutex_unlock(&itv->serialize_lock); prepare_to_wait(&itv->dma_waitq, &wait, TASK_INTERRUPTIBLE); while (!(got_sig = signal_pending(current)) && test_bit(IVTV_F_S_DMA_PENDING, &s->s_flags)) { schedule(); } finish_wait(&itv->dma_waitq, &wait); + mutex_lock(&itv->serialize_lock); if (got_sig) { IVTV_DEBUG_INFO("User interrupted %s\n", s->name); return -EINTR; @@ -756,9 +758,7 @@ unsigned int ivtv_v4l2_enc_poll(struct file *filp, poll_table * wait) if (!eof && !test_bit(IVTV_F_S_STREAMING, &s->s_flags)) { int rc; - mutex_lock(&itv->serialize_lock); rc = ivtv_start_capture(id); - mutex_unlock(&itv->serialize_lock); if (rc) { IVTV_DEBUG_INFO("Could not start capture for %s (%d)\n", s->name, rc); @@ -861,7 +861,6 @@ int ivtv_v4l2_close(struct file *filp) IVTV_DEBUG_FILE("close %s\n", s->name); - mutex_lock(&itv->serialize_lock); /* Stop radio */ if (id->type == IVTV_ENC_STREAM_TYPE_RAD && v4l2_fh_is_singular_file(filp)) { @@ -893,7 +892,6 @@ int ivtv_v4l2_close(struct file *filp) /* Easy case first: this stream was never claimed by us */ if (s->id != id->open_id) { kfree(id); - mutex_unlock(&itv->serialize_lock); return 0; } @@ -914,21 +912,25 @@ int ivtv_v4l2_close(struct file *filp) ivtv_stop_capture(id, 0); } kfree(id); - mutex_unlock(&itv->serialize_lock); return 0; } -static int ivtv_serialized_open(struct ivtv_stream *s, struct file *filp) +int ivtv_v4l2_open(struct file *filp) { -#ifdef CONFIG_VIDEO_ADV_DEBUG struct video_device *vdev = video_devdata(filp); -#endif + struct ivtv_stream *s = video_get_drvdata(vdev); struct ivtv *itv = s->itv; struct ivtv_open_id *item; int res = 0; IVTV_DEBUG_FILE("open %s\n", s->name); + if (ivtv_init_on_first_open(itv)) { + IVTV_ERR("Failed to initialize on device %s\n", + video_device_node_name(vdev)); + return -ENXIO; + } + #ifdef CONFIG_VIDEO_ADV_DEBUG /* Unless ivtv_fw_debug is set, error out if firmware dead. */ if (ivtv_fw_debug) { @@ -1017,28 +1019,6 @@ static int ivtv_serialized_open(struct ivtv_stream *s, struct file *filp) return 0; } -int ivtv_v4l2_open(struct file *filp) -{ - int res; - struct ivtv *itv = NULL; - struct ivtv_stream *s = NULL; - struct video_device *vdev = video_devdata(filp); - - s = video_get_drvdata(vdev); - itv = s->itv; - - mutex_lock(&itv->serialize_lock); - if (ivtv_init_on_first_open(itv)) { - IVTV_ERR("Failed to initialize on device %s\n", - video_device_node_name(vdev)); - mutex_unlock(&itv->serialize_lock); - return -ENXIO; - } - res = ivtv_serialized_open(s, filp); - mutex_unlock(&itv->serialize_lock); - return res; -} - void ivtv_mute(struct ivtv *itv) { if (atomic_read(&itv->capturing)) diff --git a/drivers/media/video/ivtv/ivtv-ioctl.c b/drivers/media/video/ivtv/ivtv-ioctl.c index ecafa697326e..c4bc48143098 100644 --- a/drivers/media/video/ivtv/ivtv-ioctl.c +++ b/drivers/media/video/ivtv/ivtv-ioctl.c @@ -179,6 +179,7 @@ int ivtv_set_speed(struct ivtv *itv, int speed) ivtv_vapi(itv, CX2341X_DEC_PAUSE_PLAYBACK, 1, 0); /* Wait for any DMA to finish */ + mutex_unlock(&itv->serialize_lock); prepare_to_wait(&itv->dma_waitq, &wait, TASK_INTERRUPTIBLE); while (test_bit(IVTV_F_I_DMA, &itv->i_flags)) { got_sig = signal_pending(current); @@ -188,6 +189,7 @@ int ivtv_set_speed(struct ivtv *itv, int speed) schedule(); } finish_wait(&itv->dma_waitq, &wait); + mutex_lock(&itv->serialize_lock); if (got_sig) return -EINTR; @@ -1107,6 +1109,7 @@ void ivtv_s_std_dec(struct ivtv *itv, v4l2_std_id *std) * happens within the first 100 lines of the top field. * Make 4 attempts to sync to the decoder before giving up. */ + mutex_unlock(&itv->serialize_lock); for (f = 0; f < 4; f++) { prepare_to_wait(&itv->vsync_waitq, &wait, TASK_UNINTERRUPTIBLE); @@ -1115,6 +1118,7 @@ void ivtv_s_std_dec(struct ivtv *itv, v4l2_std_id *std) schedule_timeout(msecs_to_jiffies(25)); } finish_wait(&itv->vsync_waitq, &wait); + mutex_lock(&itv->serialize_lock); if (f == 4) IVTV_WARN("Mode change failed to sync to decoder\n"); @@ -1842,8 +1846,7 @@ static long ivtv_default(struct file *file, void *fh, bool valid_prio, return 0; } -static long ivtv_serialized_ioctl(struct ivtv *itv, struct file *filp, - unsigned int cmd, unsigned long arg) +long ivtv_v4l2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { struct video_device *vfd = video_devdata(filp); long ret; @@ -1855,21 +1858,6 @@ static long ivtv_serialized_ioctl(struct ivtv *itv, struct file *filp, return ret; } -long ivtv_v4l2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) -{ - struct ivtv_open_id *id = fh2id(filp->private_data); - struct ivtv *itv = id->itv; - long res; - - /* DQEVENT can block, so this should not run with the serialize lock */ - if (cmd == VIDIOC_DQEVENT) - return ivtv_serialized_ioctl(itv, filp, cmd, arg); - mutex_lock(&itv->serialize_lock); - res = ivtv_serialized_ioctl(itv, filp, cmd, arg); - mutex_unlock(&itv->serialize_lock); - return res; -} - static const struct v4l2_ioctl_ops ivtv_ioctl_ops = { .vidioc_querycap = ivtv_querycap, .vidioc_s_audio = ivtv_s_audio, diff --git a/drivers/media/video/ivtv/ivtv-streams.c b/drivers/media/video/ivtv/ivtv-streams.c index e7794dc1330e..d598df046605 100644 --- a/drivers/media/video/ivtv/ivtv-streams.c +++ b/drivers/media/video/ivtv/ivtv-streams.c @@ -214,6 +214,7 @@ static int ivtv_prep_dev(struct ivtv *itv, int type) s->vdev->fops = ivtv_stream_info[type].fops; s->vdev->release = video_device_release; s->vdev->tvnorms = V4L2_STD_ALL; + s->vdev->lock = &itv->serialize_lock; set_bit(V4L2_FL_USE_FH_PRIO, &s->vdev->flags); ivtv_set_funcs(s->vdev); return 0; diff --git a/drivers/media/video/ivtv/ivtv-yuv.c b/drivers/media/video/ivtv/ivtv-yuv.c index dcbab6ad4c26..2ad65eb29832 100644 --- a/drivers/media/video/ivtv/ivtv-yuv.c +++ b/drivers/media/video/ivtv/ivtv-yuv.c @@ -1149,23 +1149,37 @@ int ivtv_yuv_udma_stream_frame(struct ivtv *itv, void __user *src) { struct yuv_playback_info *yi = &itv->yuv_info; struct ivtv_dma_frame dma_args; + int res; ivtv_yuv_setup_stream_frame(itv); /* We only need to supply source addresses for this */ dma_args.y_source = src; dma_args.uv_source = src + 720 * ((yi->v4l2_src_h + 31) & ~31); - return ivtv_yuv_udma_frame(itv, &dma_args); + /* Wait for frame DMA. Note that serialize_lock is locked, + so to allow other processes to access the driver while + we are waiting unlock first and later lock again. */ + mutex_unlock(&itv->serialize_lock); + res = ivtv_yuv_udma_frame(itv, &dma_args); + mutex_lock(&itv->serialize_lock); + return res; } /* IVTV_IOC_DMA_FRAME ioctl handler */ int ivtv_yuv_prep_frame(struct ivtv *itv, struct ivtv_dma_frame *args) { -/* IVTV_DEBUG_INFO("yuv_prep_frame\n"); */ + int res; +/* IVTV_DEBUG_INFO("yuv_prep_frame\n"); */ ivtv_yuv_next_free(itv); ivtv_yuv_setup_frame(itv, args); - return ivtv_yuv_udma_frame(itv, args); + /* Wait for frame DMA. Note that serialize_lock is locked, + so to allow other processes to access the driver while + we are waiting unlock first and later lock again. */ + mutex_unlock(&itv->serialize_lock); + res = ivtv_yuv_udma_frame(itv, args); + mutex_lock(&itv->serialize_lock); + return res; } void ivtv_yuv_close(struct ivtv *itv) @@ -1174,7 +1188,9 @@ void ivtv_yuv_close(struct ivtv *itv) int h_filter, v_filter_1, v_filter_2; IVTV_DEBUG_YUV("ivtv_yuv_close\n"); + mutex_unlock(&itv->serialize_lock); ivtv_waitq(&itv->vsync_waitq); + mutex_lock(&itv->serialize_lock); yi->running = 0; atomic_set(&yi->next_dma_frame, -1); From 61bb725ef5a646d3fc8a64e41b020a65542cdae1 Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Wed, 12 Oct 2011 05:36:04 -0300 Subject: [PATCH 342/820] [media] ivtv: remove open_id/id from the filehandle code Instead of messing around with id's it's much easier to just compare against a filehandle pointer. Signed-off-by: Hans Verkuil Acked-by: Andy Walls Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/ivtv/ivtv-driver.c | 3 --- drivers/media/video/ivtv/ivtv-driver.h | 3 +-- drivers/media/video/ivtv/ivtv-fileops.c | 19 +++++++++---------- drivers/media/video/ivtv/ivtv-irq.c | 4 ++-- drivers/media/video/ivtv/ivtv-streams.c | 1 - 5 files changed, 12 insertions(+), 18 deletions(-) diff --git a/drivers/media/video/ivtv/ivtv-driver.c b/drivers/media/video/ivtv/ivtv-driver.c index 41108a9a195e..461ae449cb43 100644 --- a/drivers/media/video/ivtv/ivtv-driver.c +++ b/drivers/media/video/ivtv/ivtv-driver.c @@ -731,9 +731,6 @@ static int __devinit ivtv_init_struct1(struct ivtv *itv) init_kthread_work(&itv->irq_work, ivtv_irq_work_handler); - /* start counting open_id at 1 */ - itv->open_id = 1; - /* Initial settings */ itv->cxhdl.port = CX2341X_PORT_MEMORY; itv->cxhdl.capabilities = CX2341X_CAP_HAS_SLICED_VBI; diff --git a/drivers/media/video/ivtv/ivtv-driver.h b/drivers/media/video/ivtv/ivtv-driver.h index 8f9cc17b518e..06f3d78389bf 100644 --- a/drivers/media/video/ivtv/ivtv-driver.h +++ b/drivers/media/video/ivtv/ivtv-driver.h @@ -332,7 +332,7 @@ struct ivtv_stream { const char *name; /* name of the stream */ int type; /* stream type */ - u32 id; + struct v4l2_fh *fh; /* pointer to the streaming filehandle */ spinlock_t qlock; /* locks access to the queues */ unsigned long s_flags; /* status flags, see above */ int dma; /* can be PCI_DMA_TODEVICE, PCI_DMA_FROMDEVICE or PCI_DMA_NONE */ @@ -379,7 +379,6 @@ struct ivtv_stream { struct ivtv_open_id { struct v4l2_fh fh; - u32 open_id; /* unique ID for this file descriptor */ int type; /* stream type */ int yuv_frames; /* 1: started OUT_UDMA_YUV output mode */ struct ivtv *itv; diff --git a/drivers/media/video/ivtv/ivtv-fileops.c b/drivers/media/video/ivtv/ivtv-fileops.c index 116ece4d6f27..2cd6c89b7d91 100644 --- a/drivers/media/video/ivtv/ivtv-fileops.c +++ b/drivers/media/video/ivtv/ivtv-fileops.c @@ -50,16 +50,16 @@ static int ivtv_claim_stream(struct ivtv_open_id *id, int type) if (test_and_set_bit(IVTV_F_S_CLAIMED, &s->s_flags)) { /* someone already claimed this stream */ - if (s->id == id->open_id) { + if (s->fh == &id->fh) { /* yes, this file descriptor did. So that's OK. */ return 0; } - if (s->id == -1 && (type == IVTV_DEC_STREAM_TYPE_VBI || + if (s->fh == NULL && (type == IVTV_DEC_STREAM_TYPE_VBI || type == IVTV_ENC_STREAM_TYPE_VBI)) { /* VBI is handled already internally, now also assign the file descriptor to this stream for external reading of the stream. */ - s->id = id->open_id; + s->fh = &id->fh; IVTV_DEBUG_INFO("Start Read VBI\n"); return 0; } @@ -67,7 +67,7 @@ static int ivtv_claim_stream(struct ivtv_open_id *id, int type) IVTV_DEBUG_INFO("Stream %d is busy\n", type); return -EBUSY; } - s->id = id->open_id; + s->fh = &id->fh; if (type == IVTV_DEC_STREAM_TYPE_VBI) { /* Enable reinsertion interrupt */ ivtv_clear_irq_mask(itv, IVTV_IRQ_DEC_VBI_RE_INSERT); @@ -104,7 +104,7 @@ void ivtv_release_stream(struct ivtv_stream *s) struct ivtv *itv = s->itv; struct ivtv_stream *s_vbi; - s->id = -1; + s->fh = NULL; if ((s->type == IVTV_DEC_STREAM_TYPE_VBI || s->type == IVTV_ENC_STREAM_TYPE_VBI) && test_bit(IVTV_F_S_INTERNAL_USE, &s->s_flags)) { /* this stream is still in use internally */ @@ -136,7 +136,7 @@ void ivtv_release_stream(struct ivtv_stream *s) /* was already cleared */ return; } - if (s_vbi->id != -1) { + if (s_vbi->fh) { /* VBI stream still claimed by a file descriptor */ return; } @@ -359,7 +359,7 @@ static ssize_t ivtv_read(struct ivtv_stream *s, char __user *ubuf, size_t tot_co size_t tot_written = 0; int single_frame = 0; - if (atomic_read(&itv->capturing) == 0 && s->id == -1) { + if (atomic_read(&itv->capturing) == 0 && s->fh == NULL) { /* shouldn't happen */ IVTV_DEBUG_WARN("Stream %s not initialized before read\n", s->name); return -EIO; @@ -808,7 +808,7 @@ void ivtv_stop_capture(struct ivtv_open_id *id, int gop_end) id->type == IVTV_ENC_STREAM_TYPE_VBI) && test_bit(IVTV_F_S_INTERNAL_USE, &s->s_flags)) { /* Also used internally, don't stop capturing */ - s->id = -1; + s->fh = NULL; } else { ivtv_stop_v4l2_encode_stream(s, gop_end); @@ -890,7 +890,7 @@ int ivtv_v4l2_close(struct file *filp) v4l2_fh_exit(fh); /* Easy case first: this stream was never claimed by us */ - if (s->id != id->open_id) { + if (s->fh != &id->fh) { kfree(id); return 0; } @@ -974,7 +974,6 @@ int ivtv_v4l2_open(struct file *filp) item->itv = itv; item->type = s->type; - item->open_id = itv->open_id++; filp->private_data = &item->fh; v4l2_fh_add(&item->fh); diff --git a/drivers/media/video/ivtv/ivtv-irq.c b/drivers/media/video/ivtv/ivtv-irq.c index 9c29e964d400..1b3b9578bf47 100644 --- a/drivers/media/video/ivtv/ivtv-irq.c +++ b/drivers/media/video/ivtv/ivtv-irq.c @@ -288,13 +288,13 @@ static void dma_post(struct ivtv_stream *s) ivtv_process_vbi_data(itv, buf, 0, s->type); s->q_dma.bytesused += buf->bytesused; } - if (s->id == -1) { + if (s->fh == NULL) { ivtv_queue_move(s, &s->q_dma, NULL, &s->q_free, 0); return; } } ivtv_queue_move(s, &s->q_dma, NULL, &s->q_full, s->q_dma.bytesused); - if (s->id != -1) + if (s->fh) wake_up(&s->waitq); } diff --git a/drivers/media/video/ivtv/ivtv-streams.c b/drivers/media/video/ivtv/ivtv-streams.c index d598df046605..c6e28b4ebbed 100644 --- a/drivers/media/video/ivtv/ivtv-streams.c +++ b/drivers/media/video/ivtv/ivtv-streams.c @@ -159,7 +159,6 @@ static void ivtv_stream_init(struct ivtv *itv, int type) s->buffers = (itv->options.kilobytes[type] * 1024 + s->buf_size - 1) / s->buf_size; spin_lock_init(&s->qlock); init_waitqueue_head(&s->waitq); - s->id = -1; s->sg_handle = IVTV_DMA_UNMAPPED; ivtv_queue_init(&s->q_free); ivtv_queue_init(&s->q_full); From 412662d204eca981458156fd64d9d5f3b533d7b6 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 11 Jan 2012 11:52:17 +1100 Subject: [PATCH 343/820] netfilter: xt_hashlimit: fix unused variable warning if IPv6 disabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fixes this warning when CONFIG_IP6_NF_IPTABLES is not enabled: net/netfilter/xt_hashlimit.c: In function ‘hashlimit_init_dst’: net/netfilter/xt_hashlimit.c:448:9: warning: unused variable ‘frag_off’ [-Wunused-variable] Signed-off-by: Stephen Rothwell Signed-off-by: Pablo Neira Ayuso --- net/netfilter/xt_hashlimit.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/netfilter/xt_hashlimit.c b/net/netfilter/xt_hashlimit.c index 8e4992101875..d95f9c963cde 100644 --- a/net/netfilter/xt_hashlimit.c +++ b/net/netfilter/xt_hashlimit.c @@ -445,7 +445,6 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo, { __be16 _ports[2], *ports; u8 nexthdr; - __be16 frag_off; int poff; memset(dst, 0, sizeof(*dst)); @@ -466,6 +465,9 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo, break; #if IS_ENABLED(CONFIG_IP6_NF_IPTABLES) case NFPROTO_IPV6: + { + __be16 frag_off; + if (hinfo->cfg.mode & XT_HASHLIMIT_HASH_DIP) { memcpy(&dst->ip6.dst, &ipv6_hdr(skb)->daddr, sizeof(dst->ip6.dst)); @@ -485,6 +487,7 @@ hashlimit_init_dst(const struct xt_hashlimit_htable *hinfo, if ((int)protoff < 0) return -1; break; + } #endif default: BUG(); From 9bf04646b0b41c5438ed8a27c5f8dbe0ff40d756 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Sun, 15 Jan 2012 16:57:12 +0100 Subject: [PATCH 344/820] netfilter: revert user-space expectation helper support This patch partially reverts: 3d058d7 netfilter: rework user-space expectation helper support that was applied during the 3.2 development cycle. After this patch, the tree remains just like before patch bc01bef, that initially added the preliminary infrastructure. I decided to partially revert this patch because the approach that I proposed to resolve this problem is broken in NAT setups. Moreover, a new infrastructure will be submitted for the 3.3.x development cycle that resolve the existing issues while providing a neat solution. Since nobody has been seriously using this infrastructure in user-space, the removal of this feature should affect any know FOSS project (to my knowledge). Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/nf_conntrack_common.h | 4 ---- include/linux/netfilter/xt_CT.h | 3 +-- net/netfilter/nf_conntrack_helper.c | 12 ------------ net/netfilter/nf_conntrack_netlink.c | 4 ---- net/netfilter/xt_CT.c | 8 +++----- 5 files changed, 4 insertions(+), 27 deletions(-) diff --git a/include/linux/netfilter/nf_conntrack_common.h b/include/linux/netfilter/nf_conntrack_common.h index 9e3a2838291b..0d3dd66322ec 100644 --- a/include/linux/netfilter/nf_conntrack_common.h +++ b/include/linux/netfilter/nf_conntrack_common.h @@ -83,10 +83,6 @@ enum ip_conntrack_status { /* Conntrack is a fake untracked entry */ IPS_UNTRACKED_BIT = 12, IPS_UNTRACKED = (1 << IPS_UNTRACKED_BIT), - - /* Conntrack has a userspace helper. */ - IPS_USERSPACE_HELPER_BIT = 13, - IPS_USERSPACE_HELPER = (1 << IPS_USERSPACE_HELPER_BIT), }; /* Connection tracking event types */ diff --git a/include/linux/netfilter/xt_CT.h b/include/linux/netfilter/xt_CT.h index 6390f0992f36..b56e76811c04 100644 --- a/include/linux/netfilter/xt_CT.h +++ b/include/linux/netfilter/xt_CT.h @@ -3,8 +3,7 @@ #include -#define XT_CT_NOTRACK 0x1 -#define XT_CT_USERSPACE_HELPER 0x2 +#define XT_CT_NOTRACK 0x1 struct xt_ct_target_info { __u16 flags; diff --git a/net/netfilter/nf_conntrack_helper.c b/net/netfilter/nf_conntrack_helper.c index 299fec91f741..bbe23baa19b6 100644 --- a/net/netfilter/nf_conntrack_helper.c +++ b/net/netfilter/nf_conntrack_helper.c @@ -121,18 +121,6 @@ int __nf_ct_try_assign_helper(struct nf_conn *ct, struct nf_conn *tmpl, int ret = 0; if (tmpl != NULL) { - /* we've got a userspace helper. */ - if (tmpl->status & IPS_USERSPACE_HELPER) { - help = nf_ct_helper_ext_add(ct, flags); - if (help == NULL) { - ret = -ENOMEM; - goto out; - } - rcu_assign_pointer(help->helper, NULL); - __set_bit(IPS_USERSPACE_HELPER_BIT, &ct->status); - ret = 0; - goto out; - } help = nfct_help(tmpl); if (help != NULL) helper = help->helper; diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 2a4834b83332..9307b033c0c9 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -2042,10 +2042,6 @@ ctnetlink_create_expect(struct net *net, u16 zone, } help = nfct_help(ct); if (!help) { - err = -EOPNOTSUPP; - goto out; - } - if (test_bit(IPS_USERSPACE_HELPER_BIT, &ct->status)) { if (!cda[CTA_EXPECT_TIMEOUT]) { err = -EINVAL; goto out; diff --git a/net/netfilter/xt_CT.c b/net/netfilter/xt_CT.c index 8e87123f1373..0221d10de75a 100644 --- a/net/netfilter/xt_CT.c +++ b/net/netfilter/xt_CT.c @@ -62,8 +62,8 @@ static int xt_ct_tg_check(const struct xt_tgchk_param *par) int ret = 0; u8 proto; - if (info->flags & ~(XT_CT_NOTRACK | XT_CT_USERSPACE_HELPER)) - return -EOPNOTSUPP; + if (info->flags & ~XT_CT_NOTRACK) + return -EINVAL; if (info->flags & XT_CT_NOTRACK) { ct = nf_ct_untracked_get(); @@ -92,9 +92,7 @@ static int xt_ct_tg_check(const struct xt_tgchk_param *par) GFP_KERNEL)) goto err3; - if (info->flags & XT_CT_USERSPACE_HELPER) { - __set_bit(IPS_USERSPACE_HELPER_BIT, &ct->status); - } else if (info->helper[0]) { + if (info->helper[0]) { ret = -ENOENT; proto = xt_ct_find_proto(par); if (!proto) { From 54d3fb3b11a7c38b112585e54b7af7cb3faa5c91 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Sun, 8 Jan 2012 07:19:29 -0300 Subject: [PATCH 345/820] [media] pwc: Make fps runtime configurable through s_parm, drop fps module param Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-ctrl.c | 32 ++++++++++------ drivers/media/video/pwc/pwc-if.c | 17 +-------- drivers/media/video/pwc/pwc-v4l.c | 61 +++++++++++++++++++++++++++++- 3 files changed, 82 insertions(+), 28 deletions(-) diff --git a/drivers/media/video/pwc/pwc-ctrl.c b/drivers/media/video/pwc/pwc-ctrl.c index 905d41d90c6a..c6dea6bc3225 100644 --- a/drivers/media/video/pwc/pwc-ctrl.c +++ b/drivers/media/video/pwc/pwc-ctrl.c @@ -194,8 +194,12 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int frames, 7 /* 30 */ }; - if (size < 0 || size > PSZ_CIF || frames < 4 || frames > 25) + if (size < 0 || size > PSZ_CIF) return -EINVAL; + if (frames < 4) + frames = 4; + else if (frames > 25) + frames = 25; frames = frames2frames[frames]; fps = frames2table[frames]; pEntry = &Nala_table[size][fps]; @@ -250,11 +254,14 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int frames, const struct Timon_table_entry *pChoose; int ret, fps; - if (size >= PSZ_MAX || frames < 5 || frames > 30 || - *compression < 0 || *compression > 3) - return -EINVAL; - if (size == PSZ_VGA && frames > 15) + if (size >= PSZ_MAX || *compression < 0 || *compression > 3) return -EINVAL; + if (frames < 5) + frames = 5; + else if (size == PSZ_VGA && frames > 15) + frames = 15; + else if (frames > 30) + frames = 30; fps = (frames / 5) - 1; /* Find a supported framerate with progressively higher compression */ @@ -283,7 +290,7 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int frames, memcpy(pdev->cmd_buf, buf, 13); /* Set various parameters */ - pdev->vframes = frames; + pdev->vframes = (fps + 1) * 5; pdev->valternate = pChoose->alternate; pdev->width = pwc_image_sizes[size][0]; pdev->height = pwc_image_sizes[size][1]; @@ -303,11 +310,14 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int frames, int fps, ret; unsigned char buf[12]; - if (size >= PSZ_MAX || frames < 5 || frames > 30 || - *compression < 0 || *compression > 3) - return -EINVAL; - if (size == PSZ_VGA && frames > 15) + if (size >= PSZ_MAX || *compression < 0 || *compression > 3) return -EINVAL; + if (frames < 5) + frames = 5; + else if (size == PSZ_VGA && frames > 15) + frames = 15; + else if (frames > 30) + frames = 30; fps = (frames / 5) - 1; /* Find a supported framerate with progressively higher compression */ @@ -339,7 +349,7 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int frames, pdev->cmd_len = 12; memcpy(pdev->cmd_buf, buf, 12); /* All set and go */ - pdev->vframes = frames; + pdev->vframes = (fps + 1) * 5; pdev->valternate = pChoose->alternate; pdev->width = pwc_image_sizes[size][0]; pdev->height = pwc_image_sizes[size][1]; diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index 943d37ad0d33..5592068b2f92 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -128,7 +128,6 @@ static struct usb_driver pwc_driver = { #define MAX_DEV_HINTS 20 #define MAX_ISOC_ERRORS 20 -static int default_fps = 10; #ifdef CONFIG_USB_PWC_DEBUG int pwc_trace = PWC_DEBUG_LEVEL; #endif @@ -1076,7 +1075,6 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id return -ENOMEM; } pdev->type = type_id; - pdev->vframes = default_fps; pdev->features = features; pwc_construct(pdev); /* set min/max sizes correct */ @@ -1138,8 +1136,7 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id pwc_set_leds(pdev, 0, 0); /* Setup intial videomode */ - rc = pwc_set_video_mode(pdev, MAX_WIDTH, MAX_HEIGHT, pdev->vframes, - &compression); + rc = pwc_set_video_mode(pdev, MAX_WIDTH, MAX_HEIGHT, 30, &compression); if (rc) goto err_free_mem; @@ -1243,13 +1240,11 @@ static void usb_pwc_disconnect(struct usb_interface *intf) * Initialization code & module stuff */ -static int fps; static int leds[2] = { -1, -1 }; static unsigned int leds_nargs; static char *dev_hint[MAX_DEV_HINTS]; static unsigned int dev_hint_nargs; -module_param(fps, int, 0444); #ifdef CONFIG_USB_PWC_DEBUG module_param_named(trace, pwc_trace, int, 0644); #endif @@ -1257,7 +1252,6 @@ module_param(power_save, int, 0644); module_param_array(leds, int, &leds_nargs, 0444); module_param_array(dev_hint, charp, &dev_hint_nargs, 0444); -MODULE_PARM_DESC(fps, "Initial frames per second. Varies with model, useful range 5-30"); #ifdef CONFIG_USB_PWC_DEBUG MODULE_PARM_DESC(trace, "For debugging purposes"); #endif @@ -1286,15 +1280,6 @@ static int __init usb_pwc_init(void) } #endif - if (fps) { - if (fps < 4 || fps > 30) { - PWC_ERROR("Framerate out of bounds (4-30).\n"); - return -EINVAL; - } - default_fps = fps; - PWC_DEBUG_MODULE("Default framerate set to %d.\n", default_fps); - } - if (leds[0] >= 0) led_on = leds[0]; if (leds[1] >= 0) diff --git a/drivers/media/video/pwc/pwc-v4l.c b/drivers/media/video/pwc/pwc-v4l.c index 80e25842e84a..ef422d0ac599 100644 --- a/drivers/media/video/pwc/pwc-v4l.c +++ b/drivers/media/video/pwc/pwc-v4l.c @@ -493,7 +493,7 @@ static int pwc_s_fmt_vid_cap(struct file *file, void *fh, struct v4l2_format *f) (pixelformat>>24)&255); ret = pwc_set_video_mode(pdev, f->fmt.pix.width, f->fmt.pix.height, - pdev->vframes, &compression); + 30, &compression); PWC_DEBUG_IOCTL("pwc_set_video_mode(), return=%d\n", ret); @@ -1094,6 +1094,63 @@ static int pwc_enum_frameintervals(struct file *file, void *fh, return 0; } +static int pwc_g_parm(struct file *file, void *fh, + struct v4l2_streamparm *parm) +{ + struct pwc_device *pdev = video_drvdata(file); + + if (parm->type != V4L2_BUF_TYPE_VIDEO_CAPTURE) + return -EINVAL; + + memset(parm, 0, sizeof(*parm)); + + parm->type = V4L2_BUF_TYPE_VIDEO_CAPTURE; + parm->parm.capture.readbuffers = MIN_FRAMES; + parm->parm.capture.capability |= V4L2_CAP_TIMEPERFRAME; + parm->parm.capture.timeperframe.denominator = pdev->vframes; + parm->parm.capture.timeperframe.numerator = 1; + + return 0; +} + +static int pwc_s_parm(struct file *file, void *fh, + struct v4l2_streamparm *parm) +{ + struct pwc_device *pdev = video_drvdata(file); + int compression = 0; + int ret, fps; + + if (parm->type != V4L2_BUF_TYPE_VIDEO_CAPTURE || + parm->parm.capture.timeperframe.numerator == 0) + return -EINVAL; + + if (pwc_test_n_set_capt_file(pdev, file)) + return -EBUSY; + + fps = parm->parm.capture.timeperframe.denominator / + parm->parm.capture.timeperframe.numerator; + + mutex_lock(&pdev->udevlock); + if (!pdev->udev) { + ret = -ENODEV; + goto leave; + } + + if (pdev->iso_init) { + ret = -EBUSY; + goto leave; + } + + ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, fps, + &compression); + + pwc_g_parm(file, fh, parm); + +leave: + mutex_unlock(&pdev->udevlock); + return ret; +} + static int pwc_log_status(struct file *file, void *priv) { struct pwc_device *pdev = video_drvdata(file); @@ -1120,4 +1177,6 @@ const struct v4l2_ioctl_ops pwc_ioctl_ops = { .vidioc_log_status = pwc_log_status, .vidioc_enum_framesizes = pwc_enum_framesizes, .vidioc_enum_frameintervals = pwc_enum_frameintervals, + .vidioc_g_parm = pwc_g_parm, + .vidioc_s_parm = pwc_s_parm, }; From 56ae24aad8f9e25dfef995c3e898e5f394cf0beb Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Sun, 8 Jan 2012 11:29:19 -0300 Subject: [PATCH 346/820] [media] pwc: Make decoder data part of the main pwc struct Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-ctrl.c | 15 ++++-------- drivers/media/video/pwc/pwc-dec1.c | 16 ++++--------- drivers/media/video/pwc/pwc-dec1.h | 6 +++-- drivers/media/video/pwc/pwc-dec23.c | 37 ++++++++++------------------- drivers/media/video/pwc/pwc-dec23.h | 6 ++--- drivers/media/video/pwc/pwc-if.c | 8 ------- drivers/media/video/pwc/pwc.h | 7 +++++- 7 files changed, 35 insertions(+), 60 deletions(-) diff --git a/drivers/media/video/pwc/pwc-ctrl.c b/drivers/media/video/pwc/pwc-ctrl.c index c6dea6bc3225..34a01b209251 100644 --- a/drivers/media/video/pwc/pwc-ctrl.c +++ b/drivers/media/video/pwc/pwc-ctrl.c @@ -213,9 +213,7 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int frames, return ret; } if (pEntry->compressed && pdev->pixfmt == V4L2_PIX_FMT_YUV420) { - ret = pwc_dec1_init(pdev, pdev->type, pdev->release, buf); - if (ret < 0) - return ret; + pwc_dec1_init(pdev, buf); } pdev->cmd_len = 3; @@ -281,9 +279,7 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int frames, return ret; if (pChoose->bandlength > 0 && pdev->pixfmt == V4L2_PIX_FMT_YUV420) { - ret = pwc_dec23_init(pdev, pdev->type, buf); - if (ret < 0) - return ret; + pwc_dec23_init(pdev, buf); } pdev->cmd_len = 13; @@ -341,9 +337,7 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int frames, return ret; if (pChoose->bandlength > 0 && pdev->pixfmt == V4L2_PIX_FMT_YUV420) { - ret = pwc_dec23_init(pdev, pdev->type, buf); - if (ret < 0) - return ret; + pwc_dec23_init(pdev, buf); } pdev->cmd_len = 12; @@ -368,7 +362,8 @@ int pwc_set_video_mode(struct pwc_device *pdev, int width, int height, { int ret, size; - PWC_DEBUG_FLOW("set_video_mode(%dx%d @ %d, pixfmt %08x).\n", width, height, frames, pdev->pixfmt); + PWC_DEBUG_FLOW("set_video_mode(%dx%d @ %d, pixfmt %08x).\n", + width, height, frames, pdev->pixfmt); size = pwc_get_size(pdev, width, height); PWC_TRACE("decode_size = %d.\n", size); diff --git a/drivers/media/video/pwc/pwc-dec1.c b/drivers/media/video/pwc/pwc-dec1.c index be0e02cb487f..bac0d83fe119 100644 --- a/drivers/media/video/pwc/pwc-dec1.c +++ b/drivers/media/video/pwc/pwc-dec1.c @@ -22,19 +22,11 @@ along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include "pwc-dec1.h" +#include "pwc.h" -int pwc_dec1_init(struct pwc_device *pwc, int type, int release, void *buffer) +void pwc_dec1_init(struct pwc_device *pdev, void *buffer) { - struct pwc_dec1_private *pdec; + struct pwc_dec1_private *pdec = &pdev->dec1; - if (pwc->decompress_data == NULL) { - pdec = kmalloc(sizeof(struct pwc_dec1_private), GFP_KERNEL); - if (pdec == NULL) - return -ENOMEM; - pwc->decompress_data = pdec; - } - pdec = pwc->decompress_data; - - return 0; + pdec->version = pdev->release; } diff --git a/drivers/media/video/pwc/pwc-dec1.h b/drivers/media/video/pwc/pwc-dec1.h index a57d8601080b..6e8f3c561c9b 100644 --- a/drivers/media/video/pwc/pwc-dec1.h +++ b/drivers/media/video/pwc/pwc-dec1.h @@ -25,13 +25,15 @@ #ifndef PWC_DEC1_H #define PWC_DEC1_H -#include "pwc.h" +#include + +struct pwc_device; struct pwc_dec1_private { int version; }; -int pwc_dec1_init(struct pwc_device *pwc, int type, int release, void *buffer); +void pwc_dec1_init(struct pwc_device *pdev, void *buffer); #endif diff --git a/drivers/media/video/pwc/pwc-dec23.c b/drivers/media/video/pwc/pwc-dec23.c index 2c6709112b2f..6d48c1f90683 100644 --- a/drivers/media/video/pwc/pwc-dec23.c +++ b/drivers/media/video/pwc/pwc-dec23.c @@ -294,22 +294,14 @@ static unsigned char pwc_crop_table[256 + 2*MAX_OUTER_CROP_VALUE]; /* If the type or the command change, we rebuild the lookup table */ -int pwc_dec23_init(struct pwc_device *pwc, int type, unsigned char *cmd) +void pwc_dec23_init(struct pwc_device *pdev, unsigned char *cmd) { int flags, version, shift, i; - struct pwc_dec23_private *pdec; - - if (pwc->decompress_data == NULL) { - pdec = kmalloc(sizeof(struct pwc_dec23_private), GFP_KERNEL); - if (pdec == NULL) - return -ENOMEM; - pwc->decompress_data = pdec; - } - pdec = pwc->decompress_data; + struct pwc_dec23_private *pdec = &pdev->dec23; mutex_init(&pdec->lock); - if (DEVICE_USE_CODEC3(type)) { + if (DEVICE_USE_CODEC3(pdev->type)) { flags = cmd[2] & 0x18; if (flags == 8) pdec->nbits = 7; /* More bits, mean more bits to encode the stream, but better quality */ @@ -355,8 +347,6 @@ int pwc_dec23_init(struct pwc_device *pwc, int type, unsigned char *cmd) for (i=0; idecompress_data; + struct pwc_dec23_private *pdec = &pdev->dec23; /* YUV420P image format */ unsigned char *pout_planar_y; @@ -674,23 +664,22 @@ void pwc_dec23_decompress(const struct pwc_device *pwc, mutex_lock(&pdec->lock); - bandlines_left = pwc->height / 4; - bytes_per_block = pwc->width * 4; - plane_size = pwc->height * pwc->width; + bandlines_left = pdev->height / 4; + bytes_per_block = pdev->width * 4; + plane_size = pdev->height * pdev->width; pout_planar_y = dst; pout_planar_u = dst + plane_size; pout_planar_v = dst + plane_size + plane_size / 4; while (bandlines_left--) { - DecompressBand23(pwc->decompress_data, - src, + DecompressBand23(pdec, src, pout_planar_y, pout_planar_u, pout_planar_v, - pwc->width, pwc->width); - src += pwc->vbandlength; + pdev->width, pdev->width); + src += pdev->vbandlength; pout_planar_y += bytes_per_block; - pout_planar_u += pwc->width; - pout_planar_v += pwc->width; + pout_planar_u += pdev->width; + pout_planar_v += pdev->width; } mutex_unlock(&pdec->lock); } diff --git a/drivers/media/video/pwc/pwc-dec23.h b/drivers/media/video/pwc/pwc-dec23.h index d64a3c281af6..a29068ee428b 100644 --- a/drivers/media/video/pwc/pwc-dec23.h +++ b/drivers/media/video/pwc/pwc-dec23.h @@ -25,7 +25,7 @@ #ifndef PWC_DEC23_H #define PWC_DEC23_H -#include "pwc.h" +struct pwc_device; struct pwc_dec23_private { @@ -51,8 +51,8 @@ struct pwc_dec23_private }; -int pwc_dec23_init(struct pwc_device *pwc, int type, unsigned char *cmd); -void pwc_dec23_decompress(const struct pwc_device *pwc, +void pwc_dec23_init(struct pwc_device *pdev, unsigned char *cmd); +void pwc_dec23_decompress(struct pwc_device *pdev, const void *src, void *dst); #endif diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index 5592068b2f92..250ad4c3e9d8 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -603,14 +603,6 @@ static void pwc_video_release(struct v4l2_device *v) if (device_hint[hint].pdev == pdev) device_hint[hint].pdev = NULL; - /* Free intermediate decompression buffer & tables */ - if (pdev->decompress_data != NULL) { - PWC_DEBUG_MEMORY("Freeing decompression buffer at %p.\n", - pdev->decompress_data); - kfree(pdev->decompress_data); - pdev->decompress_data = NULL; - } - v4l2_ctrl_handler_free(&pdev->ctrl_handler); kfree(pdev); diff --git a/drivers/media/video/pwc/pwc.h b/drivers/media/video/pwc/pwc.h index 47c518fef179..0f3bc1b3a971 100644 --- a/drivers/media/video/pwc/pwc.h +++ b/drivers/media/video/pwc/pwc.h @@ -44,6 +44,8 @@ #ifdef CONFIG_USB_PWC_INPUT_EVDEV #include #endif +#include "pwc-dec1.h" +#include "pwc-dec23.h" /* Version block */ #define PWC_VERSION "10.0.15" @@ -272,7 +274,10 @@ struct pwc_device int frame_total_size; /* including header & trailer */ int drop_frames; - void *decompress_data; /* private data for decompression engine */ + union { /* private data for decompression engine */ + struct pwc_dec1_private dec1; + struct pwc_dec23_private dec23; + }; /* * We have an 'image' and a 'view', where 'image' is the fixed-size img From d167a85c5fb45b1ecdacdb9b7733833a9af78da8 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Tue, 10 Jan 2012 13:01:41 -0300 Subject: [PATCH 347/820] [media] pwc: Fix pixfmt handling Before this patch various code in the mode setting patch checked pdev->pixfmt, but that was not set until the mode setting succeeded, so it was looking at the old pixfmt! This patch fixes this by making the pixfmt a parameter to set_video_mode, and setting it from set_video_mode on success. Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-ctrl.c | 39 +++++++++++++++--------------- drivers/media/video/pwc/pwc-if.c | 5 ++-- drivers/media/video/pwc/pwc-misc.c | 1 - drivers/media/video/pwc/pwc-v4l.c | 13 +++------- drivers/media/video/pwc/pwc.h | 2 +- 5 files changed, 28 insertions(+), 32 deletions(-) diff --git a/drivers/media/video/pwc/pwc-ctrl.c b/drivers/media/video/pwc/pwc-ctrl.c index 34a01b209251..51ab4c570d0b 100644 --- a/drivers/media/video/pwc/pwc-ctrl.c +++ b/drivers/media/video/pwc/pwc-ctrl.c @@ -168,8 +168,8 @@ int send_control_msg(struct pwc_device *pdev, request, value, pdev->vcinterface, buf, buflen); } -static int set_video_mode_Nala(struct pwc_device *pdev, int size, int frames, - int *compression) +static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, + int frames, int *compression) { unsigned char buf[3]; int ret, fps; @@ -212,14 +212,14 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int frames, PWC_DEBUG_MODULE("Failed to send video command... %d\n", ret); return ret; } - if (pEntry->compressed && pdev->pixfmt == V4L2_PIX_FMT_YUV420) { + if (pEntry->compressed && pixfmt == V4L2_PIX_FMT_YUV420) pwc_dec1_init(pdev, buf); - } pdev->cmd_len = 3; memcpy(pdev->cmd_buf, buf, 3); /* Set various parameters */ + pdev->pixfmt = pixfmt; pdev->vframes = frames; pdev->valternate = pEntry->alternate; pdev->width = pwc_image_sizes[size][0]; @@ -245,8 +245,8 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int frames, } -static int set_video_mode_Timon(struct pwc_device *pdev, int size, int frames, - int *compression) +static int set_video_mode_Timon(struct pwc_device *pdev, int size, int pixfmt, + int frames, int *compression) { unsigned char buf[13]; const struct Timon_table_entry *pChoose; @@ -278,14 +278,14 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int frames, if (ret < 0) return ret; - if (pChoose->bandlength > 0 && pdev->pixfmt == V4L2_PIX_FMT_YUV420) { + if (pChoose->bandlength > 0 && pixfmt == V4L2_PIX_FMT_YUV420) pwc_dec23_init(pdev, buf); - } pdev->cmd_len = 13; memcpy(pdev->cmd_buf, buf, 13); /* Set various parameters */ + pdev->pixfmt = pixfmt; pdev->vframes = (fps + 1) * 5; pdev->valternate = pChoose->alternate; pdev->width = pwc_image_sizes[size][0]; @@ -299,8 +299,8 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int frames, } -static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int frames, - int *compression) +static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int pixfmt, + int frames, int *compression) { const struct Kiara_table_entry *pChoose = NULL; int fps, ret; @@ -336,13 +336,13 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int frames, if (ret < 0) return ret; - if (pChoose->bandlength > 0 && pdev->pixfmt == V4L2_PIX_FMT_YUV420) { + if (pChoose->bandlength > 0 && pixfmt == V4L2_PIX_FMT_YUV420) pwc_dec23_init(pdev, buf); - } pdev->cmd_len = 12; memcpy(pdev->cmd_buf, buf, 12); /* All set and go */ + pdev->pixfmt = pixfmt; pdev->vframes = (fps + 1) * 5; pdev->valternate = pChoose->alternate; pdev->width = pwc_image_sizes[size][0]; @@ -358,23 +358,24 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int frames, } int pwc_set_video_mode(struct pwc_device *pdev, int width, int height, - int frames, int *compression) + int pixfmt, int frames, int *compression) { int ret, size; PWC_DEBUG_FLOW("set_video_mode(%dx%d @ %d, pixfmt %08x).\n", - width, height, frames, pdev->pixfmt); + width, height, frames, pixfmt); size = pwc_get_size(pdev, width, height); PWC_TRACE("decode_size = %d.\n", size); if (DEVICE_USE_CODEC1(pdev->type)) { - ret = set_video_mode_Nala(pdev, size, frames, compression); - + ret = set_video_mode_Nala(pdev, size, pixfmt, frames, + compression); } else if (DEVICE_USE_CODEC3(pdev->type)) { - ret = set_video_mode_Kiara(pdev, size, frames, compression); - + ret = set_video_mode_Kiara(pdev, size, pixfmt, frames, + compression); } else { - ret = set_video_mode_Timon(pdev, size, frames, compression); + ret = set_video_mode_Timon(pdev, size, pixfmt, frames, + compression); } if (ret < 0) { PWC_ERROR("Failed to set video mode %s@%d fps; return code = %d\n", size2name[size], frames, ret); diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index 250ad4c3e9d8..1f3386da05fa 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -386,7 +386,7 @@ retry: /* We first try with low compression and then retry with a higher compression setting if there is not enough bandwidth. */ ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, - pdev->vframes, &compression); + pdev->pixfmt, pdev->vframes, &compression); /* Get the current alternate interface, adjust packet size */ intf = usb_ifnum_to_if(udev, 0); @@ -1128,7 +1128,8 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id pwc_set_leds(pdev, 0, 0); /* Setup intial videomode */ - rc = pwc_set_video_mode(pdev, MAX_WIDTH, MAX_HEIGHT, 30, &compression); + rc = pwc_set_video_mode(pdev, MAX_WIDTH, MAX_HEIGHT, + V4L2_PIX_FMT_YUV420, 30, &compression); if (rc) goto err_free_mem; diff --git a/drivers/media/video/pwc/pwc-misc.c b/drivers/media/video/pwc/pwc-misc.c index 23a55b5814fc..9be5adffa874 100644 --- a/drivers/media/video/pwc/pwc-misc.c +++ b/drivers/media/video/pwc/pwc-misc.c @@ -90,5 +90,4 @@ void pwc_construct(struct pwc_device *pdev) pdev->frame_header_size = 0; pdev->frame_trailer_size = 0; } - pdev->pixfmt = V4L2_PIX_FMT_YUV420; /* default */ } diff --git a/drivers/media/video/pwc/pwc-v4l.c b/drivers/media/video/pwc/pwc-v4l.c index ef422d0ac599..b275fad2927b 100644 --- a/drivers/media/video/pwc/pwc-v4l.c +++ b/drivers/media/video/pwc/pwc-v4l.c @@ -493,16 +493,11 @@ static int pwc_s_fmt_vid_cap(struct file *file, void *fh, struct v4l2_format *f) (pixelformat>>24)&255); ret = pwc_set_video_mode(pdev, f->fmt.pix.width, f->fmt.pix.height, - 30, &compression); + pixelformat, 30, &compression); PWC_DEBUG_IOCTL("pwc_set_video_mode(), return=%d\n", ret); - if (ret == 0) { - pdev->pixfmt = pixelformat; - pwc_vidioc_fill_fmt(f, pdev->width, pdev->height, - pdev->pixfmt); - } - + pwc_vidioc_fill_fmt(f, pdev->width, pdev->height, pdev->pixfmt); leave: mutex_unlock(&pdev->udevlock); return ret; @@ -1141,8 +1136,8 @@ static int pwc_s_parm(struct file *file, void *fh, goto leave; } - ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, fps, - &compression); + ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, pdev->pixfmt, + fps, &compression); pwc_g_parm(file, fh, parm); diff --git a/drivers/media/video/pwc/pwc.h b/drivers/media/video/pwc/pwc.h index 0f3bc1b3a971..29b6d3dca83b 100644 --- a/drivers/media/video/pwc/pwc.h +++ b/drivers/media/video/pwc/pwc.h @@ -369,7 +369,7 @@ void pwc_construct(struct pwc_device *pdev); /** Functions in pwc-ctrl.c */ /* Request a certain video mode. Returns < 0 if not possible */ extern int pwc_set_video_mode(struct pwc_device *pdev, int width, int height, - int frames, int *compression); + int pixfmt, int frames, int *compression); extern unsigned int pwc_get_fps(struct pwc_device *pdev, unsigned int index, unsigned int size); extern int pwc_set_leds(struct pwc_device *pdev, int on_value, int off_value); extern int pwc_get_cmos_sensor(struct pwc_device *pdev, int *sensor); From 938d5b9e7c2e20a7e609ad5874c6e7d8d391e6e9 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Tue, 10 Jan 2012 13:14:46 -0300 Subject: [PATCH 348/820] [media] pwc: Avoid sending mode info to the camera when it is not needed Since we always do a set_video_mode on stream start, there is no need to actually send the mode info to the device on a s_fmt / s_parm ioctl. Not doing this saves us doing (slow) usb io. Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-ctrl.c | 29 ++++++++++++++++------------- drivers/media/video/pwc/pwc-if.c | 6 +++--- drivers/media/video/pwc/pwc-v4l.c | 4 ++-- drivers/media/video/pwc/pwc.h | 2 +- 4 files changed, 22 insertions(+), 19 deletions(-) diff --git a/drivers/media/video/pwc/pwc-ctrl.c b/drivers/media/video/pwc/pwc-ctrl.c index 51ab4c570d0b..9c1fb3f07dee 100644 --- a/drivers/media/video/pwc/pwc-ctrl.c +++ b/drivers/media/video/pwc/pwc-ctrl.c @@ -169,10 +169,10 @@ int send_control_msg(struct pwc_device *pdev, } static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, - int frames, int *compression) + int frames, int *compression, int send_to_cam) { unsigned char buf[3]; - int ret, fps; + int fps, ret = 0; struct Nala_table_entry *pEntry; int frames2frames[31] = { /* closest match of framerate */ @@ -207,7 +207,8 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, return -EINVAL; memcpy(buf, pEntry->mode, 3); - ret = send_video_command(pdev, pdev->vendpoint, buf, 3); + if (send_to_cam) + ret = send_video_command(pdev, pdev->vendpoint, buf, 3); if (ret < 0) { PWC_DEBUG_MODULE("Failed to send video command... %d\n", ret); return ret; @@ -246,11 +247,11 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, static int set_video_mode_Timon(struct pwc_device *pdev, int size, int pixfmt, - int frames, int *compression) + int frames, int *compression, int send_to_cam) { unsigned char buf[13]; const struct Timon_table_entry *pChoose; - int ret, fps; + int fps, ret = 0; if (size >= PSZ_MAX || *compression < 0 || *compression > 3) return -EINVAL; @@ -274,7 +275,8 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int pixfmt, return -ENOENT; /* Not supported. */ memcpy(buf, pChoose->mode, 13); - ret = send_video_command(pdev, pdev->vendpoint, buf, 13); + if (send_to_cam) + ret = send_video_command(pdev, pdev->vendpoint, buf, 13); if (ret < 0) return ret; @@ -300,10 +302,10 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int pixfmt, static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int pixfmt, - int frames, int *compression) + int frames, int *compression, int send_to_cam) { const struct Kiara_table_entry *pChoose = NULL; - int fps, ret; + int fps, ret = 0; unsigned char buf[12]; if (size >= PSZ_MAX || *compression < 0 || *compression > 3) @@ -332,7 +334,8 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int pixfmt, memcpy(buf, pChoose->mode, 12); /* Firmware bug: video endpoint is 5, but commands are sent to endpoint 4 */ - ret = send_video_command(pdev, 4 /* pdev->vendpoint */, buf, 12); + if (send_to_cam) + ret = send_video_command(pdev, 4, buf, 12); if (ret < 0) return ret; @@ -358,7 +361,7 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int pixfmt, } int pwc_set_video_mode(struct pwc_device *pdev, int width, int height, - int pixfmt, int frames, int *compression) + int pixfmt, int frames, int *compression, int send_to_cam) { int ret, size; @@ -369,13 +372,13 @@ int pwc_set_video_mode(struct pwc_device *pdev, int width, int height, if (DEVICE_USE_CODEC1(pdev->type)) { ret = set_video_mode_Nala(pdev, size, pixfmt, frames, - compression); + compression, send_to_cam); } else if (DEVICE_USE_CODEC3(pdev->type)) { ret = set_video_mode_Kiara(pdev, size, pixfmt, frames, - compression); + compression, send_to_cam); } else { ret = set_video_mode_Timon(pdev, size, pixfmt, frames, - compression); + compression, send_to_cam); } if (ret < 0) { PWC_ERROR("Failed to set video mode %s@%d fps; return code = %d\n", size2name[size], frames, ret); diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index 1f3386da05fa..23eaceea4862 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -385,8 +385,8 @@ static int pwc_isoc_init(struct pwc_device *pdev) retry: /* We first try with low compression and then retry with a higher compression setting if there is not enough bandwidth. */ - ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, - pdev->pixfmt, pdev->vframes, &compression); + ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, pdev->pixfmt, + pdev->vframes, &compression, 1); /* Get the current alternate interface, adjust packet size */ intf = usb_ifnum_to_if(udev, 0); @@ -1129,7 +1129,7 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id /* Setup intial videomode */ rc = pwc_set_video_mode(pdev, MAX_WIDTH, MAX_HEIGHT, - V4L2_PIX_FMT_YUV420, 30, &compression); + V4L2_PIX_FMT_YUV420, 30, &compression, 1); if (rc) goto err_free_mem; diff --git a/drivers/media/video/pwc/pwc-v4l.c b/drivers/media/video/pwc/pwc-v4l.c index b275fad2927b..46feece38852 100644 --- a/drivers/media/video/pwc/pwc-v4l.c +++ b/drivers/media/video/pwc/pwc-v4l.c @@ -493,7 +493,7 @@ static int pwc_s_fmt_vid_cap(struct file *file, void *fh, struct v4l2_format *f) (pixelformat>>24)&255); ret = pwc_set_video_mode(pdev, f->fmt.pix.width, f->fmt.pix.height, - pixelformat, 30, &compression); + pixelformat, 30, &compression, 0); PWC_DEBUG_IOCTL("pwc_set_video_mode(), return=%d\n", ret); @@ -1137,7 +1137,7 @@ static int pwc_s_parm(struct file *file, void *fh, } ret = pwc_set_video_mode(pdev, pdev->width, pdev->height, pdev->pixfmt, - fps, &compression); + fps, &compression, 0); pwc_g_parm(file, fh, parm); diff --git a/drivers/media/video/pwc/pwc.h b/drivers/media/video/pwc/pwc.h index 29b6d3dca83b..f441999e5bd1 100644 --- a/drivers/media/video/pwc/pwc.h +++ b/drivers/media/video/pwc/pwc.h @@ -369,7 +369,7 @@ void pwc_construct(struct pwc_device *pdev); /** Functions in pwc-ctrl.c */ /* Request a certain video mode. Returns < 0 if not possible */ extern int pwc_set_video_mode(struct pwc_device *pdev, int width, int height, - int pixfmt, int frames, int *compression); + int pixfmt, int frames, int *compression, int send_to_cam); extern unsigned int pwc_get_fps(struct pwc_device *pdev, unsigned int index, unsigned int size); extern int pwc_set_leds(struct pwc_device *pdev, int on_value, int off_value); extern int pwc_get_cmos_sensor(struct pwc_device *pdev, int *sensor); From 1c852201a28601b1379857f615e99071211294c6 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Tue, 10 Jan 2012 13:23:34 -0300 Subject: [PATCH 349/820] [media] pwc: Avoid unnecessarily rebuilding the decoder tables Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-dec23.c | 6 ++++++ drivers/media/video/pwc/pwc-dec23.h | 3 +++ 2 files changed, 9 insertions(+) diff --git a/drivers/media/video/pwc/pwc-dec23.c b/drivers/media/video/pwc/pwc-dec23.c index 6d48c1f90683..98772efc5bde 100644 --- a/drivers/media/video/pwc/pwc-dec23.c +++ b/drivers/media/video/pwc/pwc-dec23.c @@ -301,6 +301,9 @@ void pwc_dec23_init(struct pwc_device *pdev, unsigned char *cmd) mutex_init(&pdec->lock); + if (pdec->last_cmd_valid && pdec->last_cmd == cmd[2]) + return; + if (DEVICE_USE_CODEC3(pdev->type)) { flags = cmd[2] & 0x18; if (flags == 8) @@ -347,6 +350,9 @@ void pwc_dec23_init(struct pwc_device *pdev, unsigned char *cmd) for (i=0; ilast_cmd = cmd[2]; + pdec->last_cmd_valid = 1; } /* diff --git a/drivers/media/video/pwc/pwc-dec23.h b/drivers/media/video/pwc/pwc-dec23.h index a29068ee428b..af31d6047bb2 100644 --- a/drivers/media/video/pwc/pwc-dec23.h +++ b/drivers/media/video/pwc/pwc-dec23.h @@ -31,11 +31,14 @@ struct pwc_dec23_private { struct mutex lock; + unsigned char last_cmd, last_cmd_valid; + unsigned int scalebits; unsigned int nbitsmask, nbits; /* Number of bits of a color in the compressed stream */ unsigned int reservoir; unsigned int nbits_in_reservoir; + const unsigned char *stream; int temp_colors[16]; From 24be689bfbbcd6c047d7918784ff810e97648006 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Tue, 10 Jan 2012 17:02:04 -0300 Subject: [PATCH 350/820] [media] pwc: Use one shared usb command buffer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The pwc driver used to: 1. kmalloc a buffer 2. memcpy data to send over usb there 3. do the usb_control_msg call (which does not work with data on the stack) 4. free the buffer For every usb command send. This patch changes the code to instead malloc a buffer for this purpose once and use it everywhere. [mchehab@redhat.com: Fix a compilation breakage with allyesconfig: drivers/media/video/pwc/pwc-ctrl.c: In function ‘pwc_get_cmos_sensor’: drivers/media/video/pwc/pwc-ctrl.c:546:3: warning: passing argument 4 of ‘recv_control_msg’ makes integer from pointer without a cast [en$ drivers/media/video/pwc/pwc-ctrl.c:107:12: note: expected ‘int’ but argument is of type ‘unsigned char *’ drivers/media/video/pwc/pwc-ctrl.c:546:3: error: too many arguments to function ‘recv_control_msg’ drivers/media/video/pwc/pwc-ctrl.c:107:12: note: declared here] Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-ctrl.c | 154 ++++++++++------------------ drivers/media/video/pwc/pwc-dec1.c | 2 +- drivers/media/video/pwc/pwc-dec1.h | 2 +- drivers/media/video/pwc/pwc-dec23.c | 2 +- drivers/media/video/pwc/pwc-dec23.h | 2 +- drivers/media/video/pwc/pwc-if.c | 10 ++ drivers/media/video/pwc/pwc-v4l.c | 22 ++-- drivers/media/video/pwc/pwc.h | 5 +- 8 files changed, 81 insertions(+), 118 deletions(-) diff --git a/drivers/media/video/pwc/pwc-ctrl.c b/drivers/media/video/pwc/pwc-ctrl.c index 9c1fb3f07dee..1f506fde97d0 100644 --- a/drivers/media/video/pwc/pwc-ctrl.c +++ b/drivers/media/video/pwc/pwc-ctrl.c @@ -104,47 +104,16 @@ static struct Nala_table_entry Nala_table[PSZ_MAX][PWC_FPS_MAX_NALA] = /****************************************************************************/ -static int _send_control_msg(struct pwc_device *pdev, - u8 request, u16 value, int index, void *buf, int buflen) -{ - int rc; - void *kbuf = NULL; - - if (buflen) { - kbuf = kmemdup(buf, buflen, GFP_KERNEL); /* not allowed on stack */ - if (kbuf == NULL) - return -ENOMEM; - } - - rc = usb_control_msg(pdev->udev, usb_sndctrlpipe(pdev->udev, 0), - request, - USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_DEVICE, - value, - index, - kbuf, buflen, USB_CTRL_SET_TIMEOUT); - - kfree(kbuf); - return rc; -} - static int recv_control_msg(struct pwc_device *pdev, - u8 request, u16 value, void *buf, int buflen) + u8 request, u16 value, int recv_count) { int rc; - void *kbuf = kmalloc(buflen, GFP_KERNEL); /* not allowed on stack */ - - if (kbuf == NULL) - return -ENOMEM; rc = usb_control_msg(pdev->udev, usb_rcvctrlpipe(pdev->udev, 0), request, USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_DEVICE, - value, - pdev->vcinterface, - kbuf, buflen, USB_CTRL_GET_TIMEOUT); - memcpy(buf, kbuf, buflen); - kfree(kbuf); - + value, pdev->vcinterface, + pdev->ctrl_buf, recv_count, USB_CTRL_GET_TIMEOUT); if (rc < 0) PWC_ERROR("recv_control_msg error %d req %02x val %04x\n", rc, request, value); @@ -152,26 +121,38 @@ static int recv_control_msg(struct pwc_device *pdev, } static inline int send_video_command(struct pwc_device *pdev, - int index, void *buf, int buflen) + int index, const unsigned char *buf, int buflen) { - return _send_control_msg(pdev, - SET_EP_STREAM_CTL, - VIDEO_OUTPUT_CONTROL_FORMATTER, - index, - buf, buflen); + int rc; + + memcpy(pdev->ctrl_buf, buf, buflen); + + rc = usb_control_msg(pdev->udev, usb_sndctrlpipe(pdev->udev, 0), + SET_EP_STREAM_CTL, + USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_DEVICE, + VIDEO_OUTPUT_CONTROL_FORMATTER, index, + pdev->ctrl_buf, buflen, USB_CTRL_SET_TIMEOUT); + if (rc >= 0) + memcpy(pdev->cmd_buf, buf, buflen); + else + PWC_ERROR("send_video_command error %d\n", rc); + + return rc; } int send_control_msg(struct pwc_device *pdev, u8 request, u16 value, void *buf, int buflen) { - return _send_control_msg(pdev, - request, value, pdev->vcinterface, buf, buflen); + return usb_control_msg(pdev->udev, usb_sndctrlpipe(pdev->udev, 0), + request, + USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_DEVICE, + value, pdev->vcinterface, + buf, buflen, USB_CTRL_SET_TIMEOUT); } static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, int frames, int *compression, int send_to_cam) { - unsigned char buf[3]; int fps, ret = 0; struct Nala_table_entry *pEntry; int frames2frames[31] = @@ -206,18 +187,14 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, if (pEntry->alternate == 0) return -EINVAL; - memcpy(buf, pEntry->mode, 3); if (send_to_cam) - ret = send_video_command(pdev, pdev->vendpoint, buf, 3); - if (ret < 0) { - PWC_DEBUG_MODULE("Failed to send video command... %d\n", ret); + ret = send_video_command(pdev, pdev->vendpoint, + pEntry->mode, 3); + if (ret < 0) return ret; - } - if (pEntry->compressed && pixfmt == V4L2_PIX_FMT_YUV420) - pwc_dec1_init(pdev, buf); - pdev->cmd_len = 3; - memcpy(pdev->cmd_buf, buf, 3); + if (pEntry->compressed && pixfmt == V4L2_PIX_FMT_YUV420) + pwc_dec1_init(pdev, pEntry->mode); /* Set various parameters */ pdev->pixfmt = pixfmt; @@ -249,7 +226,6 @@ static int set_video_mode_Nala(struct pwc_device *pdev, int size, int pixfmt, static int set_video_mode_Timon(struct pwc_device *pdev, int size, int pixfmt, int frames, int *compression, int send_to_cam) { - unsigned char buf[13]; const struct Timon_table_entry *pChoose; int fps, ret = 0; @@ -274,17 +250,14 @@ static int set_video_mode_Timon(struct pwc_device *pdev, int size, int pixfmt, if (pChoose == NULL || pChoose->alternate == 0) return -ENOENT; /* Not supported. */ - memcpy(buf, pChoose->mode, 13); if (send_to_cam) - ret = send_video_command(pdev, pdev->vendpoint, buf, 13); + ret = send_video_command(pdev, pdev->vendpoint, + pChoose->mode, 13); if (ret < 0) return ret; if (pChoose->bandlength > 0 && pixfmt == V4L2_PIX_FMT_YUV420) - pwc_dec23_init(pdev, buf); - - pdev->cmd_len = 13; - memcpy(pdev->cmd_buf, buf, 13); + pwc_dec23_init(pdev, pChoose->mode); /* Set various parameters */ pdev->pixfmt = pixfmt; @@ -306,7 +279,6 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int pixfmt, { const struct Kiara_table_entry *pChoose = NULL; int fps, ret = 0; - unsigned char buf[12]; if (size >= PSZ_MAX || *compression < 0 || *compression > 3) return -EINVAL; @@ -328,22 +300,15 @@ static int set_video_mode_Kiara(struct pwc_device *pdev, int size, int pixfmt, if (pChoose == NULL || pChoose->alternate == 0) return -ENOENT; /* Not supported. */ - PWC_TRACE("Using alternate setting %d.\n", pChoose->alternate); - - /* usb_control_msg won't take staticly allocated arrays as argument?? */ - memcpy(buf, pChoose->mode, 12); - /* Firmware bug: video endpoint is 5, but commands are sent to endpoint 4 */ if (send_to_cam) - ret = send_video_command(pdev, 4, buf, 12); + ret = send_video_command(pdev, 4, pChoose->mode, 12); if (ret < 0) return ret; if (pChoose->bandlength > 0 && pixfmt == V4L2_PIX_FMT_YUV420) - pwc_dec23_init(pdev, buf); + pwc_dec23_init(pdev, pChoose->mode); - pdev->cmd_len = 12; - memcpy(pdev->cmd_buf, buf, 12); /* All set and go */ pdev->pixfmt = pixfmt; pdev->vframes = (fps + 1) * 5; @@ -445,13 +410,12 @@ unsigned int pwc_get_fps(struct pwc_device *pdev, unsigned int index, unsigned i int pwc_get_u8_ctrl(struct pwc_device *pdev, u8 request, u16 value, int *data) { int ret; - u8 buf; - ret = recv_control_msg(pdev, request, value, &buf, sizeof(buf)); + ret = recv_control_msg(pdev, request, value, 1); if (ret < 0) return ret; - *data = buf; + *data = pdev->ctrl_buf[0]; return 0; } @@ -459,7 +423,8 @@ int pwc_set_u8_ctrl(struct pwc_device *pdev, u8 request, u16 value, u8 data) { int ret; - ret = send_control_msg(pdev, request, value, &data, sizeof(data)); + pdev->ctrl_buf[0] = data; + ret = send_control_msg(pdev, request, value, pdev->ctrl_buf, 1); if (ret < 0) return ret; @@ -469,37 +434,34 @@ int pwc_set_u8_ctrl(struct pwc_device *pdev, u8 request, u16 value, u8 data) int pwc_get_s8_ctrl(struct pwc_device *pdev, u8 request, u16 value, int *data) { int ret; - s8 buf; - ret = recv_control_msg(pdev, request, value, &buf, sizeof(buf)); + ret = recv_control_msg(pdev, request, value, 1); if (ret < 0) return ret; - *data = buf; + *data = ((s8 *)pdev->ctrl_buf)[0]; return 0; } int pwc_get_u16_ctrl(struct pwc_device *pdev, u8 request, u16 value, int *data) { int ret; - u8 buf[2]; - ret = recv_control_msg(pdev, request, value, buf, sizeof(buf)); + ret = recv_control_msg(pdev, request, value, 2); if (ret < 0) return ret; - *data = (buf[1] << 8) | buf[0]; + *data = (pdev->ctrl_buf[1] << 8) | pdev->ctrl_buf[0]; return 0; } int pwc_set_u16_ctrl(struct pwc_device *pdev, u8 request, u16 value, u16 data) { int ret; - u8 buf[2]; - buf[0] = data & 0xff; - buf[1] = data >> 8; - ret = send_control_msg(pdev, request, value, buf, sizeof(buf)); + pdev->ctrl_buf[0] = data & 0xff; + pdev->ctrl_buf[1] = data >> 8; + ret = send_control_msg(pdev, request, value, pdev->ctrl_buf, 2); if (ret < 0) return ret; @@ -520,7 +482,6 @@ int pwc_button_ctrl(struct pwc_device *pdev, u16 value) /* POWER */ void pwc_camera_power(struct pwc_device *pdev, int power) { - char buf; int r; if (!pdev->power_save) @@ -530,13 +491,11 @@ void pwc_camera_power(struct pwc_device *pdev, int power) return; /* Not supported by Nala or Timon < release 6 */ if (power) - buf = 0x00; /* active */ + pdev->ctrl_buf[0] = 0x00; /* active */ else - buf = 0xFF; /* power save */ - r = send_control_msg(pdev, - SET_STATUS_CTL, SET_POWER_SAVE_MODE_FORMATTER, - &buf, sizeof(buf)); - + pdev->ctrl_buf[0] = 0xFF; /* power save */ + r = send_control_msg(pdev, SET_STATUS_CTL, + SET_POWER_SAVE_MODE_FORMATTER, pdev->ctrl_buf, 1); if (r < 0) PWC_ERROR("Failed to power %s camera (%d)\n", power ? "on" : "off", r); @@ -544,7 +503,6 @@ void pwc_camera_power(struct pwc_device *pdev, int power) int pwc_set_leds(struct pwc_device *pdev, int on_value, int off_value) { - unsigned char buf[2]; int r; if (pdev->type < 730) @@ -560,11 +518,11 @@ int pwc_set_leds(struct pwc_device *pdev, int on_value, int off_value) if (off_value > 0xff) off_value = 0xff; - buf[0] = on_value; - buf[1] = off_value; + pdev->ctrl_buf[0] = on_value; + pdev->ctrl_buf[1] = off_value; r = send_control_msg(pdev, - SET_STATUS_CTL, LED_FORMATTER, &buf, sizeof(buf)); + SET_STATUS_CTL, LED_FORMATTER, pdev->ctrl_buf, 2); if (r < 0) PWC_ERROR("Failed to set LED on/off time (%d)\n", r); @@ -574,7 +532,6 @@ int pwc_set_leds(struct pwc_device *pdev, int on_value, int off_value) #ifdef CONFIG_USB_PWC_DEBUG int pwc_get_cmos_sensor(struct pwc_device *pdev, int *sensor) { - unsigned char buf; int ret = -1, request; if (pdev->type < 675) @@ -584,14 +541,13 @@ int pwc_get_cmos_sensor(struct pwc_device *pdev, int *sensor) else request = SENSOR_TYPE_FORMATTER2; - ret = recv_control_msg(pdev, - GET_STATUS_CTL, request, &buf, sizeof(buf)); + ret = recv_control_msg(pdev, GET_STATUS_CTL, request, 1); if (ret < 0) return ret; if (pdev->type < 675) - *sensor = buf | 0x100; + *sensor = pdev->ctrl_buf[0] | 0x100; else - *sensor = buf; + *sensor = pdev->ctrl_buf[0]; return 0; } #endif diff --git a/drivers/media/video/pwc/pwc-dec1.c b/drivers/media/video/pwc/pwc-dec1.c index bac0d83fe119..e899036aadf4 100644 --- a/drivers/media/video/pwc/pwc-dec1.c +++ b/drivers/media/video/pwc/pwc-dec1.c @@ -24,7 +24,7 @@ */ #include "pwc.h" -void pwc_dec1_init(struct pwc_device *pdev, void *buffer) +void pwc_dec1_init(struct pwc_device *pdev, const unsigned char *cmd) { struct pwc_dec1_private *pdec = &pdev->dec1; diff --git a/drivers/media/video/pwc/pwc-dec1.h b/drivers/media/video/pwc/pwc-dec1.h index 6e8f3c561c9b..c565ef8f52fb 100644 --- a/drivers/media/video/pwc/pwc-dec1.h +++ b/drivers/media/video/pwc/pwc-dec1.h @@ -34,6 +34,6 @@ struct pwc_dec1_private int version; }; -void pwc_dec1_init(struct pwc_device *pdev, void *buffer); +void pwc_dec1_init(struct pwc_device *pdev, const unsigned char *cmd); #endif diff --git a/drivers/media/video/pwc/pwc-dec23.c b/drivers/media/video/pwc/pwc-dec23.c index 98772efc5bde..3792fedff951 100644 --- a/drivers/media/video/pwc/pwc-dec23.c +++ b/drivers/media/video/pwc/pwc-dec23.c @@ -294,7 +294,7 @@ static unsigned char pwc_crop_table[256 + 2*MAX_OUTER_CROP_VALUE]; /* If the type or the command change, we rebuild the lookup table */ -void pwc_dec23_init(struct pwc_device *pdev, unsigned char *cmd) +void pwc_dec23_init(struct pwc_device *pdev, const unsigned char *cmd) { int flags, version, shift, i; struct pwc_dec23_private *pdec = &pdev->dec23; diff --git a/drivers/media/video/pwc/pwc-dec23.h b/drivers/media/video/pwc/pwc-dec23.h index af31d6047bb2..c655b1c1e6a9 100644 --- a/drivers/media/video/pwc/pwc-dec23.h +++ b/drivers/media/video/pwc/pwc-dec23.h @@ -54,7 +54,7 @@ struct pwc_dec23_private }; -void pwc_dec23_init(struct pwc_device *pdev, unsigned char *cmd); +void pwc_dec23_init(struct pwc_device *pdev, const unsigned char *cmd); void pwc_dec23_decompress(struct pwc_device *pdev, const void *src, void *dst); diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index 23eaceea4862..a07df4e4aa04 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -605,6 +605,7 @@ static void pwc_video_release(struct v4l2_device *v) v4l2_ctrl_handler_free(&pdev->ctrl_handler); + kfree(pdev->ctrl_buf); kfree(pdev); } @@ -1115,6 +1116,14 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id if (hint < MAX_DEV_HINTS) device_hint[hint].pdev = pdev; + /* Allocate USB command buffers */ + pdev->ctrl_buf = kmalloc(sizeof(pdev->cmd_buf), GFP_KERNEL); + if (!pdev->ctrl_buf) { + PWC_ERROR("Oops, could not allocate memory for pwc_device.\n"); + rc = -ENOMEM; + goto err_free_mem; + } + #ifdef CONFIG_USB_PWC_DEBUG /* Query sensor type */ if (pwc_get_cmos_sensor(pdev, &rc) >= 0) { @@ -1199,6 +1208,7 @@ err_free_controls: err_free_mem: if (hint < MAX_DEV_HINTS) device_hint[hint].pdev = NULL; + kfree(pdev->ctrl_buf); kfree(pdev); return rc; } diff --git a/drivers/media/video/pwc/pwc-v4l.c b/drivers/media/video/pwc/pwc-v4l.c index 46feece38852..f495eeb5403a 100644 --- a/drivers/media/video/pwc/pwc-v4l.c +++ b/drivers/media/video/pwc/pwc-v4l.c @@ -772,33 +772,33 @@ static int pwc_set_autogain_expo(struct pwc_device *pdev) static int pwc_set_motor(struct pwc_device *pdev) { int ret; - u8 buf[4]; - buf[0] = 0; + pdev->ctrl_buf[0] = 0; if (pdev->motor_pan_reset->is_new) - buf[0] |= 0x01; + pdev->ctrl_buf[0] |= 0x01; if (pdev->motor_tilt_reset->is_new) - buf[0] |= 0x02; + pdev->ctrl_buf[0] |= 0x02; if (pdev->motor_pan_reset->is_new || pdev->motor_tilt_reset->is_new) { ret = send_control_msg(pdev, SET_MPT_CTL, - PT_RESET_CONTROL_FORMATTER, buf, 1); + PT_RESET_CONTROL_FORMATTER, + pdev->ctrl_buf, 1); if (ret < 0) return ret; } - memset(buf, 0, sizeof(buf)); + memset(pdev->ctrl_buf, 0, 4); if (pdev->motor_pan->is_new) { - buf[0] = pdev->motor_pan->val & 0xFF; - buf[1] = (pdev->motor_pan->val >> 8); + pdev->ctrl_buf[0] = pdev->motor_pan->val & 0xFF; + pdev->ctrl_buf[1] = (pdev->motor_pan->val >> 8); } if (pdev->motor_tilt->is_new) { - buf[2] = pdev->motor_tilt->val & 0xFF; - buf[3] = (pdev->motor_tilt->val >> 8); + pdev->ctrl_buf[2] = pdev->motor_tilt->val & 0xFF; + pdev->ctrl_buf[3] = (pdev->motor_tilt->val >> 8); } if (pdev->motor_pan->is_new || pdev->motor_tilt->is_new) { ret = send_control_msg(pdev, SET_MPT_CTL, PT_RELATIVE_CONTROL_FORMATTER, - buf, sizeof(buf)); + pdev->ctrl_buf, 4); if (ret < 0) return ret; } diff --git a/drivers/media/video/pwc/pwc.h b/drivers/media/video/pwc/pwc.h index f441999e5bd1..e4d4d711dd1f 100644 --- a/drivers/media/video/pwc/pwc.h +++ b/drivers/media/video/pwc/pwc.h @@ -134,9 +134,6 @@ #define DEVICE_USE_CODEC3(x) ((x)>=700) #define DEVICE_USE_CODEC23(x) ((x)>=675) -/* from pwc-dec.h */ -#define PWCX_FLAG_PLANAR 0x0001 - /* Request types: video */ #define SET_LUM_CTL 0x01 #define GET_LUM_CTL 0x02 @@ -250,8 +247,8 @@ struct pwc_device char vmirror; /* for ToUCaM series */ char power_save; /* Do powersaving for this cam */ - int cmd_len; unsigned char cmd_buf[13]; + unsigned char *ctrl_buf; struct urb *urbs[MAX_ISO_BUFS]; char iso_init; From a081c3400ff2695244714293f5f6a4675c3a9493 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Tue, 10 Jan 2012 17:47:25 -0300 Subject: [PATCH 351/820] [media] pwc: Remove dev_hint module parameter The code for this is rather crufty, and being able to tie a device to a specific minor is not really something we want to support in a modern udev based world. Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-if.c | 129 +------------------------------ 1 file changed, 2 insertions(+), 127 deletions(-) diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index a07df4e4aa04..a691cf821747 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -133,12 +133,6 @@ static struct usb_driver pwc_driver = { #endif static int power_save = -1; static int led_on = 100, led_off; /* defaults to LED that is on while in use */ -static struct { - int type; - char serial_number[30]; - int device_node; - struct pwc_device *pdev; -} device_hint[MAX_DEV_HINTS]; /***/ @@ -596,15 +590,8 @@ leave: static void pwc_video_release(struct v4l2_device *v) { struct pwc_device *pdev = container_of(v, struct pwc_device, v4l2_dev); - int hint; - - /* search device_hint[] table if we occupy a slot, by any chance */ - for (hint = 0; hint < MAX_DEV_HINTS; hint++) - if (device_hint[hint].pdev == pdev) - device_hint[hint].pdev = NULL; v4l2_ctrl_handler_free(&pdev->ctrl_handler); - kfree(pdev->ctrl_buf); kfree(pdev); } @@ -805,10 +792,9 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id struct usb_device *udev = interface_to_usbdev(intf); struct pwc_device *pdev = NULL; int vendor_id, product_id, type_id; - int hint, rc; + int rc; int features = 0; int compression = 0; - int video_nr = -1; /* default: use next available device */ int my_power_save = power_save; char serial_number[30], *name; @@ -1098,24 +1084,6 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id pdev->release = le16_to_cpu(udev->descriptor.bcdDevice); PWC_DEBUG_PROBE("Release: %04x\n", pdev->release); - /* Now search device_hint[] table for a match, so we can hint a node number. */ - for (hint = 0; hint < MAX_DEV_HINTS; hint++) { - if (((device_hint[hint].type == -1) || (device_hint[hint].type == pdev->type)) && - (device_hint[hint].pdev == NULL)) { - /* so far, so good... try serial number */ - if ((device_hint[hint].serial_number[0] == '*') || !strcmp(device_hint[hint].serial_number, serial_number)) { - /* match! */ - video_nr = device_hint[hint].device_node; - PWC_DEBUG_PROBE("Found hint, will try to register as /dev/video%d\n", video_nr); - break; - } - } - } - - /* occupy slot */ - if (hint < MAX_DEV_HINTS) - device_hint[hint].pdev = pdev; - /* Allocate USB command buffers */ pdev->ctrl_buf = kmalloc(sizeof(pdev->cmd_buf), GFP_KERNEL); if (!pdev->ctrl_buf) { @@ -1163,7 +1131,7 @@ static int usb_pwc_probe(struct usb_interface *intf, const struct usb_device_id pdev->v4l2_dev.ctrl_handler = &pdev->ctrl_handler; pdev->vdev.v4l2_dev = &pdev->v4l2_dev; - rc = video_register_device(&pdev->vdev, VFL_TYPE_GRABBER, video_nr); + rc = video_register_device(&pdev->vdev, VFL_TYPE_GRABBER, -1); if (rc < 0) { PWC_ERROR("Failed to register as video device (%d).\n", rc); goto err_unregister_v4l2_dev; @@ -1206,8 +1174,6 @@ err_unregister_v4l2_dev: err_free_controls: v4l2_ctrl_handler_free(&pdev->ctrl_handler); err_free_mem: - if (hint < MAX_DEV_HINTS) - device_hint[hint].pdev = NULL; kfree(pdev->ctrl_buf); kfree(pdev); return rc; @@ -1245,22 +1211,18 @@ static void usb_pwc_disconnect(struct usb_interface *intf) static int leds[2] = { -1, -1 }; static unsigned int leds_nargs; -static char *dev_hint[MAX_DEV_HINTS]; -static unsigned int dev_hint_nargs; #ifdef CONFIG_USB_PWC_DEBUG module_param_named(trace, pwc_trace, int, 0644); #endif module_param(power_save, int, 0644); module_param_array(leds, int, &leds_nargs, 0444); -module_param_array(dev_hint, charp, &dev_hint_nargs, 0444); #ifdef CONFIG_USB_PWC_DEBUG MODULE_PARM_DESC(trace, "For debugging purposes"); #endif MODULE_PARM_DESC(power_save, "Turn power saving for new cameras on or off"); MODULE_PARM_DESC(leds, "LED on,off time in milliseconds"); -MODULE_PARM_DESC(dev_hint, "Device node hints"); MODULE_DESCRIPTION("Philips & OEM USB webcam driver"); MODULE_AUTHOR("Luc Saillard "); @@ -1270,105 +1232,18 @@ MODULE_VERSION( PWC_VERSION ); static int __init usb_pwc_init(void) { - int i; - -#ifdef CONFIG_USB_PWC_DEBUG - PWC_INFO("Philips webcam module version " PWC_VERSION " loaded.\n"); - PWC_INFO("Supports Philips PCA645/646, PCVC675/680/690, PCVC720[40]/730/740/750 & PCVC830/840.\n"); - PWC_INFO("Also supports the Askey VC010, various Logitech Quickcams, Samsung MPC-C10 and MPC-C30,\n"); - PWC_INFO("the Creative WebCam 5 & Pro Ex, SOTEC Afina Eye and Visionite VCS-UC300 and VCS-UM100.\n"); - - if (pwc_trace >= 0) { - PWC_DEBUG_MODULE("Trace options: 0x%04x\n", pwc_trace); - } -#endif - if (leds[0] >= 0) led_on = leds[0]; if (leds[1] >= 0) led_off = leds[1]; - /* Big device node whoopla. Basically, it allows you to assign a - device node (/dev/videoX) to a camera, based on its type - & serial number. The format is [type[.serialnumber]:]node. - - Any camera that isn't matched by these rules gets the next - available free device node. - */ - for (i = 0; i < MAX_DEV_HINTS; i++) { - char *s, *colon, *dot; - - /* This loop also initializes the array */ - device_hint[i].pdev = NULL; - s = dev_hint[i]; - if (s != NULL && *s != '\0') { - device_hint[i].type = -1; /* wildcard */ - strcpy(device_hint[i].serial_number, "*"); - - /* parse string: chop at ':' & '/' */ - colon = dot = s; - while (*colon != '\0' && *colon != ':') - colon++; - while (*dot != '\0' && *dot != '.') - dot++; - /* Few sanity checks */ - if (*dot != '\0' && dot > colon) { - PWC_ERROR("Malformed camera hint: the colon must be after the dot.\n"); - return -EINVAL; - } - - if (*colon == '\0') { - /* No colon */ - if (*dot != '\0') { - PWC_ERROR("Malformed camera hint: no colon + device node given.\n"); - return -EINVAL; - } - else { - /* No type or serial number specified, just a number. */ - device_hint[i].device_node = - simple_strtol(s, NULL, 10); - } - } - else { - /* There's a colon, so we have at least a type and a device node */ - device_hint[i].type = - simple_strtol(s, NULL, 10); - device_hint[i].device_node = - simple_strtol(colon + 1, NULL, 10); - if (*dot != '\0') { - /* There's a serial number as well */ - int k; - - dot++; - k = 0; - while (*dot != ':' && k < 29) { - device_hint[i].serial_number[k++] = *dot; - dot++; - } - device_hint[i].serial_number[k] = '\0'; - } - } - PWC_TRACE("device_hint[%d]:\n", i); - PWC_TRACE(" type : %d\n", device_hint[i].type); - PWC_TRACE(" serial# : %s\n", device_hint[i].serial_number); - PWC_TRACE(" node : %d\n", device_hint[i].device_node); - } - else - device_hint[i].type = 0; /* not filled */ - } /* ..for MAX_DEV_HINTS */ - - PWC_DEBUG_PROBE("Registering driver at address 0x%p.\n", &pwc_driver); return usb_register(&pwc_driver); } static void __exit usb_pwc_exit(void) { - PWC_DEBUG_MODULE("Deregistering driver.\n"); usb_deregister(&pwc_driver); - PWC_INFO("Philips webcam module removed.\n"); } module_init(usb_pwc_init); module_exit(usb_pwc_exit); - -/* vim: set cino= formatoptions=croql cindent shiftwidth=8 tabstop=8: */ From c24e13713762a30f22468ac2c3d053abacb7672d Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Tue, 10 Jan 2012 17:51:12 -0300 Subject: [PATCH 352/820] [media] pwc: Simplify leds parameter parsing This reduces our module init to a simple usb_register() call, so that we can make use of the new upcoming macro's for this. Signed-off-by: Hans de Goede Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/pwc/pwc-if.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index a691cf821747..122fbd0081eb 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -132,7 +132,7 @@ static struct usb_driver pwc_driver = { int pwc_trace = PWC_DEBUG_LEVEL; #endif static int power_save = -1; -static int led_on = 100, led_off; /* defaults to LED that is on while in use */ +static int leds[2] = { 100, 0 }; /***/ @@ -737,7 +737,7 @@ static int start_streaming(struct vb2_queue *vq, unsigned int count) /* Turn on camera and set LEDS on */ pwc_camera_power(pdev, 1); - pwc_set_leds(pdev, led_on, led_off); + pwc_set_leds(pdev, leds[0], leds[1]); r = pwc_isoc_init(pdev); if (r) { @@ -1209,7 +1209,6 @@ static void usb_pwc_disconnect(struct usb_interface *intf) * Initialization code & module stuff */ -static int leds[2] = { -1, -1 }; static unsigned int leds_nargs; #ifdef CONFIG_USB_PWC_DEBUG @@ -1232,11 +1231,6 @@ MODULE_VERSION( PWC_VERSION ); static int __init usb_pwc_init(void) { - if (leds[0] >= 0) - led_on = leds[0]; - if (leds[1] >= 0) - led_off = leds[1]; - return usb_register(&pwc_driver); } From 82033bc52abeb266cb3403e783a276d4a3d6cf9f Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 13 Jan 2012 02:28:34 -0300 Subject: [PATCH 353/820] [media] mb86a20s: fix off by one checks Clearly ">=" was intended here instead of ">". Signed-off-by: Dan Carpenter Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/mb86a20s.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/media/dvb/frontends/mb86a20s.c b/drivers/media/dvb/frontends/mb86a20s.c index 7fa3e472cdca..fade566927c3 100644 --- a/drivers/media/dvb/frontends/mb86a20s.c +++ b/drivers/media/dvb/frontends/mb86a20s.c @@ -402,7 +402,7 @@ static int mb86a20s_get_modulation(struct mb86a20s_state *state, [2] = 0x8e, /* Layer C */ }; - if (layer > ARRAY_SIZE(reg)) + if (layer >= ARRAY_SIZE(reg)) return -EINVAL; rc = mb86a20s_writereg(state, 0x6d, reg[layer]); if (rc < 0) @@ -435,7 +435,7 @@ static int mb86a20s_get_fec(struct mb86a20s_state *state, [2] = 0x8f, /* Layer C */ }; - if (layer > ARRAY_SIZE(reg)) + if (layer >= ARRAY_SIZE(reg)) return -EINVAL; rc = mb86a20s_writereg(state, 0x6d, reg[layer]); if (rc < 0) @@ -470,7 +470,7 @@ static int mb86a20s_get_interleaving(struct mb86a20s_state *state, [2] = 0x90, /* Layer C */ }; - if (layer > ARRAY_SIZE(reg)) + if (layer >= ARRAY_SIZE(reg)) return -EINVAL; rc = mb86a20s_writereg(state, 0x6d, reg[layer]); if (rc < 0) @@ -494,7 +494,7 @@ static int mb86a20s_get_segment_count(struct mb86a20s_state *state, [2] = 0x91, /* Layer C */ }; - if (layer > ARRAY_SIZE(reg)) + if (layer >= ARRAY_SIZE(reg)) return -EINVAL; rc = mb86a20s_writereg(state, 0x6d, reg[layer]); if (rc < 0) From 6c2cc5b575ed09d81507ec86cc5cfad022d9236b Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 13 Jan 2012 02:34:00 -0300 Subject: [PATCH 354/820] [media] cx231xx: dereferencing NULL after allocation failure "dev" is NULL here so we should use "nr" instead of "dev->devno". Signed-off-by: Dan Carpenter Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/cx231xx/cx231xx-cards.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/video/cx231xx/cx231xx-cards.c b/drivers/media/video/cx231xx/cx231xx-cards.c index 627829f4d8c9..7d8edb457065 100644 --- a/drivers/media/video/cx231xx/cx231xx-cards.c +++ b/drivers/media/video/cx231xx/cx231xx-cards.c @@ -1052,7 +1052,7 @@ static int cx231xx_usb_probe(struct usb_interface *interface, dev = kzalloc(sizeof(*dev), GFP_KERNEL); if (dev == NULL) { cx231xx_err(DRIVER_NAME ": out of memory!\n"); - clear_bit(dev->devno, &cx231xx_devused); + clear_bit(nr, &cx231xx_devused); return -ENOMEM; } From a7c8aadad39428b64d26c3971d967f8314e2397d Mon Sep 17 00:00:00 2001 From: Miroslav Slugen Date: Sun, 11 Dec 2011 18:57:58 -0300 Subject: [PATCH 355/820] [media] cx23885-dvb: check if dvb_attach() succeded Fix possible null dereference for Leadtek DTV 3200H XC4000 tuner when no firmware file available. Cc: stable@kernel.org Signed-off-by: Miroslav Slugen Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/cx23885/cx23885-dvb.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/media/video/cx23885/cx23885-dvb.c b/drivers/media/video/cx23885/cx23885-dvb.c index af8a225763d3..6835eb1fc093 100644 --- a/drivers/media/video/cx23885/cx23885-dvb.c +++ b/drivers/media/video/cx23885/cx23885-dvb.c @@ -943,6 +943,11 @@ static int dvb_register(struct cx23885_tsport *port) fe = dvb_attach(xc4000_attach, fe0->dvb.frontend, &dev->i2c_bus[1].i2c_adap, &cfg); + if (!fe) { + printk(KERN_ERR "%s/2: xc4000 attach failed\n", + dev->name); + goto frontend_detach; + } } break; case CX23885_BOARD_TBS_6920: From 9ee8537f60590c63d74dc5724c921bc3eb6ef558 Mon Sep 17 00:00:00 2001 From: Miroslav Slugen Date: Sun, 11 Dec 2011 20:19:34 -0300 Subject: [PATCH 356/820] [media] cx23885: Don't duplicate xc4000 entry for radio All radio tuners in cx23885 driver using same address for radio and tuner, so there is no need to probe it twice for same tuner and we can use radio_type UNSET. Be aware radio support in cx23885 is not yet committed, so this is only minor fix for future support. Signed-off-by: Miroslav Slugen Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/cx23885/cx23885-cards.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/media/video/cx23885/cx23885-cards.c b/drivers/media/video/cx23885/cx23885-cards.c index 3c01be999e35..19b5499d2624 100644 --- a/drivers/media/video/cx23885/cx23885-cards.c +++ b/drivers/media/video/cx23885/cx23885-cards.c @@ -213,8 +213,8 @@ struct cx23885_board cx23885_boards[] = { .portc = CX23885_MPEG_DVB, .tuner_type = TUNER_XC4000, .tuner_addr = 0x61, - .radio_type = TUNER_XC4000, - .radio_addr = 0x61, + .radio_type = UNSET, + .radio_addr = ADDR_UNSET, .input = {{ .type = CX23885_VMUX_TELEVISION, .vmux = CX25840_VIN2_CH1 | From b6854e3f31402476bcc9d2f41570389fa491de17 Mon Sep 17 00:00:00 2001 From: Miroslav Slugen Date: Sun, 11 Dec 2011 19:00:06 -0300 Subject: [PATCH 357/820] [media] cx88: fix: don't duplicate xc4000 entry for radio All radio tuners in cx88 driver using same address for radio and tuner, so there is no need to probe it twice for same tuner and we can use radio_type UNSET, this also fix broken radio since kernel 2.6.39-rc1 for those tuners. Cc: stable@kernel.org Signed-off-by: Miroslav Slugen Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/cx88/cx88-cards.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/drivers/media/video/cx88/cx88-cards.c b/drivers/media/video/cx88/cx88-cards.c index 62c7ad050f9b..cbd5d119a2c6 100644 --- a/drivers/media/video/cx88/cx88-cards.c +++ b/drivers/media/video/cx88/cx88-cards.c @@ -1573,8 +1573,8 @@ static const struct cx88_board cx88_boards[] = { .name = "Pinnacle Hybrid PCTV", .tuner_type = TUNER_XC2028, .tuner_addr = 0x61, - .radio_type = TUNER_XC2028, - .radio_addr = 0x61, + .radio_type = UNSET, + .radio_addr = ADDR_UNSET, .input = { { .type = CX88_VMUX_TELEVISION, .vmux = 0, @@ -1611,8 +1611,8 @@ static const struct cx88_board cx88_boards[] = { .name = "Leadtek TV2000 XP Global", .tuner_type = TUNER_XC2028, .tuner_addr = 0x61, - .radio_type = TUNER_XC2028, - .radio_addr = 0x61, + .radio_type = UNSET, + .radio_addr = ADDR_UNSET, .input = { { .type = CX88_VMUX_TELEVISION, .vmux = 0, @@ -2115,8 +2115,8 @@ static const struct cx88_board cx88_boards[] = { .name = "Terratec Cinergy HT PCI MKII", .tuner_type = TUNER_XC2028, .tuner_addr = 0x61, - .radio_type = TUNER_XC2028, - .radio_addr = 0x61, + .radio_type = UNSET, + .radio_addr = ADDR_UNSET, .input = { { .type = CX88_VMUX_TELEVISION, .vmux = 0, @@ -2154,9 +2154,9 @@ static const struct cx88_board cx88_boards[] = { [CX88_BOARD_WINFAST_DTV1800H] = { .name = "Leadtek WinFast DTV1800 Hybrid", .tuner_type = TUNER_XC2028, - .radio_type = TUNER_XC2028, + .radio_type = UNSET, .tuner_addr = 0x61, - .radio_addr = 0x61, + .radio_addr = ADDR_UNSET, /* * GPIO setting * @@ -2195,9 +2195,9 @@ static const struct cx88_board cx88_boards[] = { [CX88_BOARD_WINFAST_DTV1800H_XC4000] = { .name = "Leadtek WinFast DTV1800 H (XC4000)", .tuner_type = TUNER_XC4000, - .radio_type = TUNER_XC4000, + .radio_type = UNSET, .tuner_addr = 0x61, - .radio_addr = 0x61, + .radio_addr = ADDR_UNSET, /* * GPIO setting * @@ -2236,9 +2236,9 @@ static const struct cx88_board cx88_boards[] = { [CX88_BOARD_WINFAST_DTV2000H_PLUS] = { .name = "Leadtek WinFast DTV2000 H PLUS", .tuner_type = TUNER_XC4000, - .radio_type = TUNER_XC4000, + .radio_type = UNSET, .tuner_addr = 0x61, - .radio_addr = 0x61, + .radio_addr = ADDR_UNSET, /* * GPIO * 2: 1: mute audio From cd4ca7afc61d3b18fcd635002459fb6b1d701099 Mon Sep 17 00:00:00 2001 From: Miroslav Slugen Date: Sun, 11 Dec 2011 18:47:32 -0300 Subject: [PATCH 358/820] [media] tuner: Fix numberspace conflict between xc4000 and pti 5nf05 tuners Update xc4000 tuner definition, number 81 is already in use by TUNER_PARTSNIC_PTI_5NF05. Cc: stable@kernel.org Signed-off-by: Miroslav Slugen Signed-off-by: Mauro Carvalho Chehab --- include/media/tuner.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/include/media/tuner.h b/include/media/tuner.h index 89c290b69a5c..29e1920e7339 100644 --- a/include/media/tuner.h +++ b/include/media/tuner.h @@ -127,7 +127,6 @@ #define TUNER_PHILIPS_FMD1216MEX_MK3 78 #define TUNER_PHILIPS_FM1216MK5 79 #define TUNER_PHILIPS_FQ1216LME_MK3 80 /* Active loopthrough, no FM */ -#define TUNER_XC4000 81 /* Xceive Silicon Tuner */ #define TUNER_PARTSNIC_PTI_5NF05 81 #define TUNER_PHILIPS_CU1216L 82 @@ -136,6 +135,8 @@ #define TUNER_PHILIPS_FQ1236_MK5 85 /* NTSC, TDA9885, no FM radio */ #define TUNER_TENA_TNF_5337 86 +#define TUNER_XC4000 87 /* Xceive Silicon Tuner */ + /* tv card specific */ #define TDA9887_PRESENT (1<<0) #define TDA9887_PORT1_INACTIVE (1<<1) From 8a538a8855aa95e5a12ae232a140e6b30c56d0c5 Mon Sep 17 00:00:00 2001 From: Miroslav Slugen Date: Tue, 20 Dec 2011 21:18:38 -0300 Subject: [PATCH 359/820] [media] xc4000: add support for signal strength measures In xc4000 chipsets real signal and noise level is stored in register 0x0A and 0x0B,so we can use those registers to monitor signal strength. I tested this patch on 2 different cards Leadtek DVR3200 and DTV2000H Plus, both with same results, I used special antenna hubs (toner 4x, 6x, 8x and 12x) with mesured signal lost, both registers are in dB value, first represent signal with limit value -113.5dB (should be -114dB) and exactly match with test results. Second represents noise level also in dB and there is no maximum value, but from tests we can drop everything above 32dB which tuner realy can't use, signal was usable till 20dB noise level. In digital mode we can take signal strength but sadly noise level is not relevant and real value is stored in demodulator for now just zl10353, also digital mode is just for testing, because it needs changing other parts of code which reads data only from demodulator. In analog mode I was able to test only FM radio, signal level is not important, it says something about cable and hub losts, but nothing about real quality of reception, so even if we have signal level at minimum 113dB we can still here radio, because of that it is displaied only in debug mode, but for real signal level is used noise register which is again very accurate, radio noise level was betwen 6-20dB for good signal, 20-25dB for medium signal, and above 25dB signal is unusable. For now real benefit of this patch is only for FM radio mode. Signed-off-by: Miroslav Slugen Signed-off-by: Mauro Carvalho Chehab --- drivers/media/common/tuners/xc4000.c | 86 ++++++++++++++++++++++++++++ 1 file changed, 86 insertions(+) diff --git a/drivers/media/common/tuners/xc4000.c b/drivers/media/common/tuners/xc4000.c index d218c1d68c33..68397110b7d9 100644 --- a/drivers/media/common/tuners/xc4000.c +++ b/drivers/media/common/tuners/xc4000.c @@ -154,6 +154,8 @@ struct xc4000_priv { #define XREG_SNR 0x06 #define XREG_VERSION 0x07 #define XREG_PRODUCT_ID 0x08 +#define XREG_SIGNAL_LEVEL 0x0A +#define XREG_NOISE_LEVEL 0x0B /* Basic firmware description. This will remain with @@ -486,6 +488,16 @@ static int xc_get_quality(struct xc4000_priv *priv, u16 *quality) return xc4000_readreg(priv, XREG_QUALITY, quality); } +static int xc_get_signal_level(struct xc4000_priv *priv, u16 *signal) +{ + return xc4000_readreg(priv, XREG_SIGNAL_LEVEL, signal); +} + +static int xc_get_noise_level(struct xc4000_priv *priv, u16 *noise) +{ + return xc4000_readreg(priv, XREG_NOISE_LEVEL, noise); +} + static u16 xc_wait_for_lock(struct xc4000_priv *priv) { u16 lock_state = 0; @@ -1089,6 +1101,8 @@ static void xc_debug_dump(struct xc4000_priv *priv) u32 hsync_freq_hz = 0; u16 frame_lines; u16 quality; + u16 signal = 0; + u16 noise = 0; u8 hw_majorversion = 0, hw_minorversion = 0; u8 fw_majorversion = 0, fw_minorversion = 0; @@ -1119,6 +1133,12 @@ static void xc_debug_dump(struct xc4000_priv *priv) xc_get_quality(priv, &quality); dprintk(1, "*** Quality (0:<8dB, 7:>56dB) = %d\n", quality); + + xc_get_signal_level(priv, &signal); + dprintk(1, "*** Signal level = -%ddB (%d)\n", signal >> 8, signal); + + xc_get_noise_level(priv, &noise); + dprintk(1, "*** Noise level = %ddB (%d)\n", noise >> 8, noise); } static int xc4000_set_params(struct dvb_frontend *fe) @@ -1432,6 +1452,71 @@ fail: return ret; } +static int xc4000_get_signal(struct dvb_frontend *fe, u16 *strength) +{ + struct xc4000_priv *priv = fe->tuner_priv; + u16 value = 0; + int rc; + + mutex_lock(&priv->lock); + rc = xc4000_readreg(priv, XREG_SIGNAL_LEVEL, &value); + mutex_unlock(&priv->lock); + + if (rc < 0) + goto ret; + + /* Informations from real testing of DVB-T and radio part, + coeficient for one dB is 0xff. + */ + tuner_dbg("Signal strength: -%ddB (%05d)\n", value >> 8, value); + + /* all known digital modes */ + if ((priv->video_standard == XC4000_DTV6) || + (priv->video_standard == XC4000_DTV7) || + (priv->video_standard == XC4000_DTV7_8) || + (priv->video_standard == XC4000_DTV8)) + goto digital; + + /* Analog mode has NOISE LEVEL important, signal + depends only on gain of antenna and amplifiers, + but it doesn't tell anything about real quality + of reception. + */ + mutex_lock(&priv->lock); + rc = xc4000_readreg(priv, XREG_NOISE_LEVEL, &value); + mutex_unlock(&priv->lock); + + tuner_dbg("Noise level: %ddB (%05d)\n", value >> 8, value); + + /* highest noise level: 32dB */ + if (value >= 0x2000) { + value = 0; + } else { + value = ~value << 3; + } + + goto ret; + + /* Digital mode has SIGNAL LEVEL important and real + noise level is stored in demodulator registers. + */ +digital: + /* best signal: -50dB */ + if (value <= 0x3200) { + value = 0xffff; + /* minimum: -114dB - should be 0x7200 but real zero is 0x713A */ + } else if (value >= 0x713A) { + value = 0; + } else { + value = ~(value - 0x3200) << 2; + } + +ret: + *strength = value; + + return rc; +} + static int xc4000_get_frequency(struct dvb_frontend *fe, u32 *freq) { struct xc4000_priv *priv = fe->tuner_priv; @@ -1559,6 +1644,7 @@ static const struct dvb_tuner_ops xc4000_tuner_ops = { .set_params = xc4000_set_params, .set_analog_params = xc4000_set_analog_params, .get_frequency = xc4000_get_frequency, + .get_rf_strength = xc4000_get_signal, .get_bandwidth = xc4000_get_bandwidth, .get_status = xc4000_get_status }; From 304bce414efce196f4051f2a5d11418c2c2ef7f9 Mon Sep 17 00:00:00 2001 From: Miroslav Slugen Date: Sun, 11 Dec 2011 20:19:34 -0300 Subject: [PATCH 360/820] [media] Add registers names to XC2028 tuner from datahseet and use them Signed-off-by: Miroslav Slugen Signed-off-by: Mauro Carvalho Chehab --- drivers/media/common/tuners/tuner-xc2028.c | 27 +++++++++++++++++----- 1 file changed, 21 insertions(+), 6 deletions(-) diff --git a/drivers/media/common/tuners/tuner-xc2028.c b/drivers/media/common/tuners/tuner-xc2028.c index 27555995f7e4..b5ee3ebfcfca 100644 --- a/drivers/media/common/tuners/tuner-xc2028.c +++ b/drivers/media/common/tuners/tuner-xc2028.c @@ -24,6 +24,21 @@ #include #include "dvb_frontend.h" +/* Registers (Write-only) */ +#define XREG_INIT 0x00 +#define XREG_RF_FREQ 0x02 +#define XREG_POWER_DOWN 0x08 + +/* Registers (Read-only) */ +#define XREG_FREQ_ERROR 0x01 +#define XREG_LOCK 0x02 +#define XREG_VERSION 0x04 +#define XREG_PRODUCT_ID 0x08 +#define XREG_HSYNC_FREQ 0x10 +#define XREG_FRAME_LINES 0x20 +#define XREG_SNR 0x40 + +#define XREG_ADC_ENV 0x0100 static int debug; module_param(debug, int, 0644); @@ -885,7 +900,7 @@ static int xc2028_signal(struct dvb_frontend *fe, u16 *strength) mutex_lock(&priv->lock); /* Sync Lock Indicator */ - rc = xc2028_get_reg(priv, 0x0002, &frq_lock); + rc = xc2028_get_reg(priv, XREG_LOCK, &frq_lock); if (rc < 0) goto ret; @@ -894,7 +909,7 @@ static int xc2028_signal(struct dvb_frontend *fe, u16 *strength) signal = 1 << 11; /* Get SNR of the video signal */ - rc = xc2028_get_reg(priv, 0x0040, &signal); + rc = xc2028_get_reg(priv, XREG_SNR, &signal); if (rc < 0) goto ret; @@ -1019,9 +1034,9 @@ static int generic_set_freq(struct dvb_frontend *fe, u32 freq /* in HZ */, /* CMD= Set frequency */ if (priv->firm_version < 0x0202) - rc = send_seq(priv, {0x00, 0x02, 0x00, 0x00}); + rc = send_seq(priv, {0x00, XREG_RF_FREQ, 0x00, 0x00}); else - rc = send_seq(priv, {0x80, 0x02, 0x00, 0x00}); + rc = send_seq(priv, {0x80, XREG_RF_FREQ, 0x00, 0x00}); if (rc < 0) goto ret; @@ -1201,9 +1216,9 @@ static int xc2028_sleep(struct dvb_frontend *fe) mutex_lock(&priv->lock); if (priv->firm_version < 0x0202) - rc = send_seq(priv, {0x00, 0x08, 0x00, 0x00}); + rc = send_seq(priv, {0x00, XREG_POWER_DOWN, 0x00, 0x00}); else - rc = send_seq(priv, {0x80, 0x08, 0x00, 0x00}); + rc = send_seq(priv, {0x80, XREG_POWER_DOWN, 0x00, 0x00}); priv->cur_fw.type = 0; /* need firmware reload */ From 764a579fb56a9205a9212250e625efe2b11e3f96 Mon Sep 17 00:00:00 2001 From: Pawel Moll Date: Thu, 15 Dec 2011 14:02:33 +0000 Subject: [PATCH 361/820] ARM: versatile: Add missing ENDPROC to headsmp.S Once the ENDPROC is in place, BSYM() in not longer necessary to get correct pointer to versatile_secondary_startup(). Tested-by: Jon Medhurst Signed-off-by: Pawel Moll Acked-by: Dave Martin --- arch/arm/mach-realview/platsmp.c | 3 +-- arch/arm/mach-vexpress/platsmp.c | 4 +--- arch/arm/plat-versatile/headsmp.S | 1 + 3 files changed, 3 insertions(+), 5 deletions(-) diff --git a/arch/arm/mach-realview/platsmp.c b/arch/arm/mach-realview/platsmp.c index e83c654a58d0..17c878ddbc70 100644 --- a/arch/arm/mach-realview/platsmp.c +++ b/arch/arm/mach-realview/platsmp.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include @@ -75,6 +74,6 @@ void __init platform_smp_prepare_cpus(unsigned int max_cpus) * until it receives a soft interrupt, and then the * secondary CPU branches to this address. */ - __raw_writel(BSYM(virt_to_phys(versatile_secondary_startup)), + __raw_writel(virt_to_phys(versatile_secondary_startup), __io_address(REALVIEW_SYS_FLAGSSET)); } diff --git a/arch/arm/mach-vexpress/platsmp.c b/arch/arm/mach-vexpress/platsmp.c index 2b5f7ac001a3..124ffb169093 100644 --- a/arch/arm/mach-vexpress/platsmp.c +++ b/arch/arm/mach-vexpress/platsmp.c @@ -13,8 +13,6 @@ #include #include -#include - #include #define V2M_PA_CS7 0x10000000 @@ -46,6 +44,6 @@ void __init platform_smp_prepare_cpus(unsigned int max_cpus) * secondary CPU branches to this address. */ writel(~0, MMIO_P2V(V2M_SYS_FLAGSCLR)); - writel(BSYM(virt_to_phys(versatile_secondary_startup)), + writel(virt_to_phys(versatile_secondary_startup), MMIO_P2V(V2M_SYS_FLAGSSET)); } diff --git a/arch/arm/plat-versatile/headsmp.S b/arch/arm/plat-versatile/headsmp.S index d397a1fb2f54..dd703ef09b8d 100644 --- a/arch/arm/plat-versatile/headsmp.S +++ b/arch/arm/plat-versatile/headsmp.S @@ -38,3 +38,4 @@ pen: ldr r7, [r6] .align 1: .long . .long pen_release +ENDPROC(versatile_secondary_startup) From 0397e4505910b44be0066d1b2ce201f0c03bae4f Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 15 Jan 2012 07:25:57 -0300 Subject: [PATCH 362/820] [media] saa7164: remove duplicate initialization These were initialized twice by mistake. They were defined the same way both times so this doesn't change how the code works. Signed-off-by: Dan Carpenter Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/saa7164/saa7164-cards.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/drivers/media/video/saa7164/saa7164-cards.c b/drivers/media/video/saa7164/saa7164-cards.c index 971591d6450f..5b72da5ce418 100644 --- a/drivers/media/video/saa7164/saa7164-cards.c +++ b/drivers/media/video/saa7164/saa7164-cards.c @@ -269,8 +269,6 @@ struct saa7164_board saa7164_boards[] = { .portb = SAA7164_MPEG_DVB, .portc = SAA7164_MPEG_ENCODER, .portd = SAA7164_MPEG_ENCODER, - .portc = SAA7164_MPEG_ENCODER, - .portd = SAA7164_MPEG_ENCODER, .porte = SAA7164_MPEG_VBI, .portf = SAA7164_MPEG_VBI, .chiprev = SAA7164_CHIP_REV3, @@ -333,8 +331,6 @@ struct saa7164_board saa7164_boards[] = { .portd = SAA7164_MPEG_ENCODER, .porte = SAA7164_MPEG_VBI, .portf = SAA7164_MPEG_VBI, - .porte = SAA7164_MPEG_VBI, - .portf = SAA7164_MPEG_VBI, .chiprev = SAA7164_CHIP_REV3, .unit = {{ .id = 0x28, From 8e0f549b23006e90172a4be188ba94f0944f08f8 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Thu, 12 Jan 2012 08:46:23 -0600 Subject: [PATCH 363/820] ARM: msm: Add missing ENDPROC to headsmp.S Once the ENDPROC is in place, BSYM() in not longer necessary to get correct pointer to msm_secondary_startup(). Signed-off-by: Rob Herring Cc: David Brown Cc: Daniel Walker --- arch/arm/mach-msm/headsmp.S | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm/mach-msm/headsmp.S b/arch/arm/mach-msm/headsmp.S index 0c631a9f8647..bcd5af223dea 100644 --- a/arch/arm/mach-msm/headsmp.S +++ b/arch/arm/mach-msm/headsmp.S @@ -34,6 +34,7 @@ pen: ldr r7, [r6] * should now contain the SVC stack for this core */ b secondary_startup +ENDPROC(msm_secondary_startup) .align 1: .long . From 09f759f66b82bbfe21e165acf0e57e7725466312 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Thu, 12 Jan 2012 08:46:23 -0600 Subject: [PATCH 364/820] ARM: ux500: add missing ENDPROC to headsmp.S Once the ENDPROC is in place, BSYM() in not longer necessary to get correct pointer to u8500_secondary_startup(). Signed-off-by: Rob Herring Cc: Srinidhi Kasagar Cc: Linus Walleij --- arch/arm/mach-ux500/headsmp.S | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/mach-ux500/headsmp.S b/arch/arm/mach-ux500/headsmp.S index 64fa451edcfd..08da5589bcd8 100644 --- a/arch/arm/mach-ux500/headsmp.S +++ b/arch/arm/mach-ux500/headsmp.S @@ -32,6 +32,8 @@ pen: ldr r7, [r6] * should now contain the SVC stack for this core */ b secondary_startup +ENDPROC(u8500_secondary_startup) + .align 2 1: .long . .long pen_release From f7597c02a2e6fada7a065b03efe283ae7ef0e0bc Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Mon, 9 Jan 2012 15:39:19 -0600 Subject: [PATCH 365/820] ARM: exynos: remove incorrect BSYM usage BSYM macro is only needed for assembly files and its usage in c files is wrong, so remove it. The linker will correctly set bit 0 for Thumb2 kernels. Signed-off-by: Rob Herring Cc: Dave Martin Cc: Kukjin Kim --- arch/arm/mach-exynos/headsmp.S | 2 ++ arch/arm/mach-exynos/platsmp.c | 5 ++--- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/arm/mach-exynos/headsmp.S b/arch/arm/mach-exynos/headsmp.S index 3cdeb3647542..5364d4bfa8bc 100644 --- a/arch/arm/mach-exynos/headsmp.S +++ b/arch/arm/mach-exynos/headsmp.S @@ -36,6 +36,8 @@ pen: ldr r7, [r6] * should now contain the SVC stack for this core */ b secondary_startup +ENDPROC(exynos4_secondary_startup) + .align 2 1: .long . .long pen_release diff --git a/arch/arm/mach-exynos/platsmp.c b/arch/arm/mach-exynos/platsmp.c index 69ffb2fb3875..b89bfa5b6b71 100644 --- a/arch/arm/mach-exynos/platsmp.c +++ b/arch/arm/mach-exynos/platsmp.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include @@ -163,7 +162,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle) while (time_before(jiffies, timeout)) { smp_rmb(); - __raw_writel(BSYM(virt_to_phys(exynos4_secondary_startup)), + __raw_writel(virt_to_phys(exynos4_secondary_startup), CPU1_BOOT_REG); gic_raise_softirq(cpumask_of(cpu), 1); @@ -218,6 +217,6 @@ void __init platform_smp_prepare_cpus(unsigned int max_cpus) * until it receives a soft interrupt, and then the * secondary CPU branches to this address. */ - __raw_writel(BSYM(virt_to_phys(exynos4_secondary_startup)), + __raw_writel(virt_to_phys(exynos4_secondary_startup), CPU1_BOOT_REG); } From e9ae4d4f028cd9ff9e7d06430cc678d7926df9e0 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 15 Jan 2012 07:32:19 -0300 Subject: [PATCH 366/820] [media] tlg2300: fix up check_firmware() return The caller doesn't check the return value of check_firmware() but static checkers complain. It currently returns negative error codes, or zero or greater on success but since the return type is boolean the values are truncated to one or zero. I've changed it to return an int, negative on error and zero on success. Signed-off-by: Dan Carpenter Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/tlg2300/pd-main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/media/video/tlg2300/pd-main.c b/drivers/media/video/tlg2300/pd-main.c index 129f135d5a5f..c096b3f74200 100644 --- a/drivers/media/video/tlg2300/pd-main.c +++ b/drivers/media/video/tlg2300/pd-main.c @@ -374,7 +374,7 @@ static inline void set_map_flags(struct poseidon *pd, struct usb_device *udev) } #endif -static bool check_firmware(struct usb_device *udev, int *down_firmware) +static int check_firmware(struct usb_device *udev, int *down_firmware) { void *buf; int ret; @@ -398,7 +398,7 @@ static bool check_firmware(struct usb_device *udev, int *down_firmware) *down_firmware = 1; return firmware_download(udev); } - return ret; + return 0; } static int poseidon_probe(struct usb_interface *interface, From faf27976a745ac126943a500c48b30c0dd1082d3 Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Sun, 15 Jan 2012 14:20:50 -0300 Subject: [PATCH 367/820] [media] anysee: do not attach same frontend twice cxd2820r implements only one frontend currently which handles all the standards. Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/anysee.c | 20 +++++++------------- 1 file changed, 7 insertions(+), 13 deletions(-) diff --git a/drivers/media/dvb/dvb-usb/anysee.c b/drivers/media/dvb/dvb-usb/anysee.c index df46015d077a..ecc3addc77ec 100644 --- a/drivers/media/dvb/dvb-usb/anysee.c +++ b/drivers/media/dvb/dvb-usb/anysee.c @@ -877,24 +877,18 @@ static int anysee_frontend_attach(struct dvb_usb_adapter *adap) case ANYSEE_HW_508T2C: /* 20 */ /* E7 T2C */ + if (state->fe_id) + break; + /* enable DVB-T/T2/C demod on IOE[5] */ ret = anysee_wr_reg_mask(adap->dev, REG_IOE, (1 << 5), 0x20); if (ret) goto error; - if (state->fe_id == 0) { - /* DVB-T/T2 */ - adap->fe_adap[state->fe_id].fe = - dvb_attach(cxd2820r_attach, - &anysee_cxd2820r_config, - &adap->dev->i2c_adap, NULL); - } else { - /* DVB-C */ - adap->fe_adap[state->fe_id].fe = - dvb_attach(cxd2820r_attach, - &anysee_cxd2820r_config, - &adap->dev->i2c_adap, adap->fe_adap[0].fe); - } + /* attach demod */ + adap->fe_adap[state->fe_id].fe = dvb_attach(cxd2820r_attach, + &anysee_cxd2820r_config, &adap->dev->i2c_adap, + NULL); state->has_ci = true; From d04ca8df70f0e1c3fe6ee2aeb1114b03a3a3de88 Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Sun, 15 Jan 2012 15:08:30 -0300 Subject: [PATCH 368/820] [media] cxd2820r: do not switch to DVB-T when DVB-C fails Fix another bug introduced by recent multi-frontend to single-frontend change. Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/cxd2820r_core.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/media/dvb/frontends/cxd2820r_core.c b/drivers/media/dvb/frontends/cxd2820r_core.c index 93e1b12e7907..b789a907c3a2 100644 --- a/drivers/media/dvb/frontends/cxd2820r_core.c +++ b/drivers/media/dvb/frontends/cxd2820r_core.c @@ -476,10 +476,10 @@ static enum dvbfe_search cxd2820r_search(struct dvb_frontend *fe) dbg("%s: delsys=%d", __func__, fe->dtv_property_cache.delivery_system); /* switch between DVB-T and DVB-T2 when tune fails */ - if (priv->last_tune_failed && (priv->delivery_system != SYS_DVBC_ANNEX_A)) { + if (priv->last_tune_failed) { if (priv->delivery_system == SYS_DVBT) c->delivery_system = SYS_DVBT2; - else + else if (priv->delivery_system == SYS_DVBT2) c->delivery_system = SYS_DVBT; } From 016961e0d9c5446017610a10fe7db14bf5f5400c Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Sun, 15 Jan 2012 16:39:13 -0300 Subject: [PATCH 369/820] [media] tda18271c2dd: Remove pointless linux/version.h include As pointed out by 'make versioncheck', there's no need for drivers/media/dvb/frontends/tda18271c2dd.c to Signed-off-by: Jesper Juhl Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/tda18271c2dd.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/media/dvb/frontends/tda18271c2dd.c b/drivers/media/dvb/frontends/tda18271c2dd.c index 86da3d816498..ad7c72e8f517 100644 --- a/drivers/media/dvb/frontends/tda18271c2dd.c +++ b/drivers/media/dvb/frontends/tda18271c2dd.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include "dvb_frontend.h" From 1f64972937d84f48a8b4d16e7e0b4a5ee7904c96 Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Sun, 15 Jan 2012 16:50:11 -0300 Subject: [PATCH 370/820] [media] cxd2820r: wait demod lock for DVB-C too Fix yet another bug introduced be recent cxd2820r multi-frontend to single-frontend change. Finally, we have at least almost working picture for DVB-C too. Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/cxd2820r_core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/dvb/frontends/cxd2820r_core.c b/drivers/media/dvb/frontends/cxd2820r_core.c index b789a907c3a2..372a4e700254 100644 --- a/drivers/media/dvb/frontends/cxd2820r_core.c +++ b/drivers/media/dvb/frontends/cxd2820r_core.c @@ -492,6 +492,7 @@ static enum dvbfe_search cxd2820r_search(struct dvb_frontend *fe) /* frontend lock wait loop count */ switch (priv->delivery_system) { case SYS_DVBT: + case SYS_DVBC_ANNEX_A: i = 20; break; case SYS_DVBT2: From 3b6a567afa3250abf280ef535000dc0dc3ed9d52 Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Sun, 15 Jan 2012 18:30:36 -0300 Subject: [PATCH 371/820] [media] cxd2820r: do not allow get_frontend() when demod is not initialized This fixes bug introduced by multi-frontend to single-frontend change. Finally HAS_LOCK is got back! We are not allowed to access hardware in sleep mode... Chip did not like when .get_frontend() reads some registers while chip was sleeping and due to that HAS_LOCK bit was never gained. TODO: We should add logic for dvb-core to drop out illegal calls like that. Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/cxd2820r_core.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/media/dvb/frontends/cxd2820r_core.c b/drivers/media/dvb/frontends/cxd2820r_core.c index 372a4e700254..caae7f79c837 100644 --- a/drivers/media/dvb/frontends/cxd2820r_core.c +++ b/drivers/media/dvb/frontends/cxd2820r_core.c @@ -309,9 +309,14 @@ static int cxd2820r_read_status(struct dvb_frontend *fe, fe_status_t *status) static int cxd2820r_get_frontend(struct dvb_frontend *fe) { + struct cxd2820r_priv *priv = fe->demodulator_priv; int ret; dbg("%s: delsys=%d", __func__, fe->dtv_property_cache.delivery_system); + + if (priv->delivery_system == SYS_UNDEFINED) + return 0; + switch (fe->dtv_property_cache.delivery_system) { case SYS_DVBT: ret = cxd2820r_get_frontend_t(fe); From 0a60cb14a58696100481674682eb07ffb2a8eac7 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Mon, 9 Jan 2012 15:41:40 -0600 Subject: [PATCH 372/820] ARM: imx: remove incorrect BSYM usage BSYM macro is only needed for assembly files and its usage in c files is wrong, so remove it. The linker will correctly set bit 0 for Thumb2 kernels. Signed-off-by: Rob Herring Cc: Sascha Hauer Acked-by: Shawn Guo Acked-by: Dave Martin --- arch/arm/mach-imx/src.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/arm/mach-imx/src.c b/arch/arm/mach-imx/src.c index a8e33681b732..74a4b5d3d1c0 100644 --- a/arch/arm/mach-imx/src.c +++ b/arch/arm/mach-imx/src.c @@ -15,7 +15,6 @@ #include #include #include -#include #define SRC_SCR 0x000 #define SRC_GPR1 0x020 @@ -42,7 +41,7 @@ void imx_enable_cpu(int cpu, bool enable) void imx_set_cpu_jump(int cpu, void *jump_addr) { cpu = cpu_logical_map(cpu); - writel_relaxed(BSYM(virt_to_phys(jump_addr)), + writel_relaxed(virt_to_phys(jump_addr), src_base + SRC_GPR1 + cpu * 8); } From adf55f7f5bde24149455efdee8d055645de96006 Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Mon, 9 Jan 2012 15:41:58 -0600 Subject: [PATCH 373/820] ARM: highbank: remove incorrect BSYM usage BSYM macro is only needed for assembly files and its usage in c files is wrong, so remove it. The linker will correctly set bit 0 for Thumb2 kernels. Signed-off-by: Rob Herring Acked-by: Dave Martin --- arch/arm/mach-highbank/highbank.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/arm/mach-highbank/highbank.c b/arch/arm/mach-highbank/highbank.c index 88660d500f5b..48293f8b4bd3 100644 --- a/arch/arm/mach-highbank/highbank.c +++ b/arch/arm/mach-highbank/highbank.c @@ -25,7 +25,6 @@ #include #include -#include #include #include #include @@ -76,7 +75,7 @@ void highbank_set_cpu_jump(int cpu, void *jump_addr) #ifdef CONFIG_SMP cpu = cpu_logical_map(cpu); #endif - writel(BSYM(virt_to_phys(jump_addr)), HB_JUMP_TABLE_VIRT(cpu)); + writel(virt_to_phys(jump_addr), HB_JUMP_TABLE_VIRT(cpu)); __cpuc_flush_dcache_area(HB_JUMP_TABLE_VIRT(cpu), 16); outer_clean_range(HB_JUMP_TABLE_PHYS(cpu), HB_JUMP_TABLE_PHYS(cpu) + 15); From efb963dcd9854c70667cdba9b5854b7290f1cefd Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Mon, 9 Jan 2012 15:43:20 -0600 Subject: [PATCH 374/820] ARM: make BSYM macro assembly only BSYM macro is only needed for assembly files and its usage in c files is wrong, so only define it for assembly. Signed-off-by: Rob Herring Acked-by: Dave Martin --- arch/arm/include/asm/unified.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/include/asm/unified.h b/arch/arm/include/asm/unified.h index bc631161e9c6..f5989f46b4d2 100644 --- a/arch/arm/include/asm/unified.h +++ b/arch/arm/include/asm/unified.h @@ -37,8 +37,8 @@ #define THUMB(x...) x #ifdef __ASSEMBLY__ #define W(instr) instr.w -#endif #define BSYM(sym) sym + 1 +#endif #else /* !CONFIG_THUMB2_KERNEL */ @@ -49,8 +49,8 @@ #define THUMB(x...) #ifdef __ASSEMBLY__ #define W(instr) instr -#endif #define BSYM(sym) sym +#endif #endif /* CONFIG_THUMB2_KERNEL */ From f7cb8a4be1cb46b03ccf517d505b3474c810853d Mon Sep 17 00:00:00 2001 From: Zeng Zhaoming Date: Mon, 16 Jan 2012 15:18:11 +0800 Subject: [PATCH 375/820] ASoC: sgtl5000: update author email address Update MODULE_AUTHOR email address. Signed-off-by: Zeng Zhaoming Signed-off-by: Mark Brown --- sound/soc/codecs/sgtl5000.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/sgtl5000.c b/sound/soc/codecs/sgtl5000.c index fc9b127206e2..a809d7025c58 100644 --- a/sound/soc/codecs/sgtl5000.c +++ b/sound/soc/codecs/sgtl5000.c @@ -1457,5 +1457,5 @@ static void __exit sgtl5000_exit(void) module_exit(sgtl5000_exit); MODULE_DESCRIPTION("Freescale SGTL5000 ALSA SoC Codec Driver"); -MODULE_AUTHOR("Zeng Zhaoming "); +MODULE_AUTHOR("Zeng Zhaoming "); MODULE_LICENSE("GPL"); From e7c80e2a8b403c10d805f48596efac741cce754d Mon Sep 17 00:00:00 2001 From: Liam Girdwood Date: Mon, 16 Jan 2012 15:23:31 +0000 Subject: [PATCH 376/820] ASoC: dapm - Fix check for codec context in dapm_power_widgets(). Fixes a NULL pointer dereference in dapm_power_widgets() if the dapm context has no codec. Signed-off-by: Liam Girdwood Signed-off-by: Mark Brown --- sound/soc/soc-dapm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/soc-dapm.c b/sound/soc/soc-dapm.c index 3ad1f59b8028..1f55ded4047f 100644 --- a/sound/soc/soc-dapm.c +++ b/sound/soc/soc-dapm.c @@ -1426,7 +1426,7 @@ static int dapm_power_widgets(struct snd_soc_dapm_context *dapm, int event) dapm->target_bias_level = SND_SOC_BIAS_ON; break; case SND_SOC_DAPM_STREAM_STOP: - if (dapm->codec->active) + if (dapm->codec && dapm->codec->active) dapm->target_bias_level = SND_SOC_BIAS_ON; else dapm->target_bias_level = SND_SOC_BIAS_STANDBY; From 675c496ba4e751ea5b895c83219e3bb034f648ba Mon Sep 17 00:00:00 2001 From: Liam Girdwood Date: Mon, 16 Jan 2012 15:25:37 +0000 Subject: [PATCH 377/820] ASoC: core - Free platform DAPM context at platform removal. Fix platform removal by freeing the platform DAPM resources and remove it from the DAPM list. Signed-off-by: Liam Girdwood Signed-off-by: Mark Brown --- sound/soc/soc-core.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index 3986520b4677..b5ecf6d23214 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -907,6 +907,10 @@ static void soc_remove_dai_link(struct snd_soc_card *card, int num, int order) if (err < 0) printk(KERN_ERR "asoc: failed to remove %s\n", platform->name); } + + /* Make sure all DAPM widgets are freed */ + snd_soc_dapm_free(&platform->dapm); + platform->probed = 0; list_del(&platform->card_list); module_put(platform->dev->driver->owner); From c4addcb55417a246b4419cbb075a2e4f412a8645 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A1rton=20N=C3=A9meth?= Date: Sun, 15 Jan 2012 11:32:30 +0100 Subject: [PATCH 378/820] gpio-pch: cleanup NULL pointer checking MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch will remove the following sparse warning ("make C=1"): * warning: Using plain integer as NULL pointer Signed-off-by: Márton Németh Signed-off-by: Grant Likely --- drivers/gpio/gpio-pch.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-pch.c b/drivers/gpio/gpio-pch.c index f0603297f829..7f008fbc4b96 100644 --- a/drivers/gpio/gpio-pch.c +++ b/drivers/gpio/gpio-pch.c @@ -376,7 +376,7 @@ static int __devinit pch_gpio_probe(struct pci_dev *pdev, } chip->base = pci_iomap(pdev, 1, 0); - if (chip->base == 0) { + if (!chip->base) { dev_err(&pdev->dev, "%s : pci_iomap FAILED", __func__); ret = -ENOMEM; goto err_iomap; From cd7bf8a554fdfaadcee6d66900c7710651dc58f2 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A1rton=20N=C3=A9meth?= Date: Sun, 15 Jan 2012 11:32:24 +0100 Subject: [PATCH 379/820] gpio-pch: cleanup __iomem annotation usage MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The __iomem annotation is to be used together with pointers used as iowrite32() parameter. For more details see [1] and [2]. This patch will remove the following sparse warnings ("make C=1"): * warning: incorrect type in assignment (different address spaces) * warning: incorrect type in argument 1 (different address spaces) * warning: incorrect type in argument 2 (different address spaces) References: [1] A new I/O memory access mechanism (Sep 15, 2004) http://lwn.net/Articles/102232/ [2] Being more anal about iospace accesses (Sep 15, 2004) http://lwn.net/Articles/102240/ Signed-off-by: Márton Németh Signed-off-by: Grant Likely --- drivers/gpio/gpio-pch.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-pch.c b/drivers/gpio/gpio-pch.c index 7f008fbc4b96..68fa55e86eb1 100644 --- a/drivers/gpio/gpio-pch.c +++ b/drivers/gpio/gpio-pch.c @@ -231,7 +231,7 @@ static void pch_gpio_setup(struct pch_gpio *chip) static int pch_irq_type(struct irq_data *d, unsigned int type) { u32 im; - u32 *im_reg; + u32 __iomem *im_reg; u32 ien; u32 im_pos; int ch; From 2bd1c85e855c94da564e151cb54553e6cfc2f04b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A1rton=20N=C3=A9meth?= Date: Sun, 15 Jan 2012 10:57:43 +0100 Subject: [PATCH 380/820] gpio-ml-ioh: cleanup NULL pointer checking MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch will remove the following sparse warning ("make C=1"): * warning: Using plain integer as NULL pointer Signed-off-by: Márton Németh Signed-off-by: Grant Likely --- drivers/gpio/gpio-ml-ioh.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-ml-ioh.c b/drivers/gpio/gpio-ml-ioh.c index 461958fc2264..af6d4f352d01 100644 --- a/drivers/gpio/gpio-ml-ioh.c +++ b/drivers/gpio/gpio-ml-ioh.c @@ -428,7 +428,7 @@ static int __devinit ioh_gpio_probe(struct pci_dev *pdev, } base = pci_iomap(pdev, 1, 0); - if (base == 0) { + if (!base) { dev_err(&pdev->dev, "%s : pci_iomap failed", __func__); ret = -ENOMEM; goto err_iomap; From dd9328a6b026060699d4136d8f926e39281c8a18 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A1rton=20N=C3=A9meth?= Date: Sun, 15 Jan 2012 10:57:34 +0100 Subject: [PATCH 381/820] gpio-ml-ioh: cleanup __iomem annotation usage MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The __iomem annotation is to be used together with pointers used in iowrite32() but not for pointers returned by kzalloc(). For more details see [1] and [2]. This patch will remove the following sparse warning (i.e. when copiling with "make C=1"): * warning: incorrect type in assignment (different address spaces) References: [1] A new I/O memory access mechanism (Sep 15, 2004) http://lwn.net/Articles/102232/ [2] Being more anal about iospace accesses (Sep 15, 2004) http://lwn.net/Articles/102240/ Signed-off-by: Márton Németh Signed-off-by: Grant Likely --- drivers/gpio/gpio-ml-ioh.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpio/gpio-ml-ioh.c b/drivers/gpio/gpio-ml-ioh.c index af6d4f352d01..03d6dd5dcb77 100644 --- a/drivers/gpio/gpio-ml-ioh.c +++ b/drivers/gpio/gpio-ml-ioh.c @@ -248,7 +248,7 @@ static void ioh_gpio_setup(struct ioh_gpio *chip, int num_port) static int ioh_irq_type(struct irq_data *d, unsigned int type) { u32 im; - u32 *im_reg; + void __iomem *im_reg; u32 ien; u32 im_pos; int ch; @@ -412,7 +412,7 @@ static int __devinit ioh_gpio_probe(struct pci_dev *pdev, int i, j; struct ioh_gpio *chip; void __iomem *base; - void __iomem *chip_save; + void *chip_save; int irq_base; ret = pci_enable_device(pdev); @@ -521,7 +521,7 @@ static void __devexit ioh_gpio_remove(struct pci_dev *pdev) int err; int i; struct ioh_gpio *chip = pci_get_drvdata(pdev); - void __iomem *chip_save; + void *chip_save; chip_save = chip; From c10076c4304083af15a41f6bc5e657e781c1f9a6 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 13 Jan 2012 21:40:59 -0500 Subject: [PATCH 382/820] tracepoints/module: Fix disabling tracepoints with taint CRAP or OOT Tracepoints are disabled for tainted modules, which is usually because the module is either proprietary or was forced, and we don't want either of them using kernel tracepoints. But, a module can also be tainted by being in the staging directory or compiled out of tree. Either is fine for use with tracepoints, no need to punish them. I found this out when I noticed that my sample trace event module, when done out of tree, stopped working. Cc: stable@vger.kernel.org # 3.2 Cc: Mathieu Desnoyers Cc: Ben Hutchings Cc: Dave Jones Cc: Greg Kroah-Hartman Cc: Rusty Russell Signed-off-by: Steven Rostedt --- kernel/tracepoint.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/kernel/tracepoint.c b/kernel/tracepoint.c index db110b8ae030..f1539decd99d 100644 --- a/kernel/tracepoint.c +++ b/kernel/tracepoint.c @@ -634,10 +634,11 @@ static int tracepoint_module_coming(struct module *mod) int ret = 0; /* - * We skip modules that tain the kernel, especially those with different - * module header (for forced load), to make sure we don't cause a crash. + * We skip modules that taint the kernel, especially those with different + * module headers (for forced load), to make sure we don't cause a crash. + * Staging and out-of-tree GPL modules are fine. */ - if (mod->taints) + if (mod->taints & ~((1 << TAINT_OOT_MODULE) | (1 << TAINT_CRAP))) return 0; mutex_lock(&tracepoints_mutex); tp_mod = kmalloc(sizeof(struct tp_module), GFP_KERNEL); From d8e8ed95cda1dfd6813588333d36552935eba4a1 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Thu, 15 Dec 2011 13:43:24 +1030 Subject: [PATCH 383/820] rcu: Make rcutorture bool parameters really bool (core code) module_param(bool) used to counter-intuitively take an int. In fddd5201 (mid-2009) we allowed bool or int/unsigned int using a messy trick. It's time to remove the int/unsigned int option. For this version it'll simply give a warning, but it'll break next kernel version. This commit makes this change to rcutorture. Signed-off-by: Rusty Russell Signed-off-by: Paul E. McKenney --- kernel/rcutorture.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c index 88f17b8a3b1d..e29edc374cec 100644 --- a/kernel/rcutorture.c +++ b/kernel/rcutorture.c @@ -56,8 +56,8 @@ static int nreaders = -1; /* # reader threads, defaults to 2*ncpus */ static int nfakewriters = 4; /* # fake writer threads */ static int stat_interval; /* Interval between stats, in seconds. */ /* Defaults to "only at end of test". */ -static int verbose; /* Print more debug info. */ -static int test_no_idle_hz; /* Test RCU's support for tickless idle CPUs. */ +static bool verbose; /* Print more debug info. */ +static bool test_no_idle_hz; /* Test RCU's support for tickless idle CPUs. */ static int shuffle_interval = 3; /* Interval between shuffles (in sec)*/ static int stutter = 5; /* Start/stop testing interval (in sec) */ static int irqreader = 1; /* RCU readers from irq (timers). */ From 7061ca3b6c99fc78115560b9a10227c8c5fafc45 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Tue, 20 Dec 2011 08:20:46 -0800 Subject: [PATCH 384/820] sched: Add "const" to is_idle_task() parameter This patch fixes a build warning in -next due to a const pointer being passed to is_idle_task(). Because is_idle_task() does not modify anything, this commit adds the "const" to is_idle_task()'s argument declaration. Reported-by: Stephen Rothwell Signed-off-by: Paul E. McKenney --- include/linux/sched.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 4a7e4d333a27..56fa25a5b1eb 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2074,7 +2074,7 @@ extern struct task_struct *idle_task(int cpu); * is_idle_task - is the specified task an idle task? * @tsk: the task in question. */ -static inline bool is_idle_task(struct task_struct *p) +static inline bool is_idle_task(const struct task_struct *p) { return p->pid == 0; } From 4410030646be072b82ec1892ad5cc7d91af384d8 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 27 Dec 2011 15:04:26 +0100 Subject: [PATCH 385/820] rcu: Add missing __cpuinit annotation in rcutorture code "rcu: Add rcutorture CPU-hotplug capability" adds cpu hotplug operations to the rcutorture code but produces a false positive warning about section mismatches: WARNING: vmlinux.o(.text+0x1e420c): Section mismatch in reference from the function rcu_torture_onoff() to the function .cpuinit.text:cpu_up() The function rcu_torture_onoff() references the function __cpuinit cpu_up(). This is often because rcu_torture_onoff lacks a __cpuinit annotation or the annotation of cpu_up is wrong. This commit therefore adds a __cpuinit annotation so the warning goes away. Signed-off-by: Heiko Carstens Signed-off-by: Paul E. McKenney --- kernel/rcutorture.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c index e29edc374cec..a58ac285fc69 100644 --- a/kernel/rcutorture.c +++ b/kernel/rcutorture.c @@ -1399,7 +1399,7 @@ rcu_torture_shutdown(void *arg) * Execute random CPU-hotplug operations at the interval specified * by the onoff_interval. */ -static int +static int __cpuinit rcu_torture_onoff(void *arg) { int cpu; @@ -1447,7 +1447,7 @@ rcu_torture_onoff(void *arg) return 0; } -static int +static int __cpuinit rcu_torture_onoff_init(void) { if (onoff_interval <= 0) From c9678d867216790f961ea06867b3ccbdb3484542 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= Date: Fri, 13 Jan 2012 22:55:05 +0100 Subject: [PATCH 386/820] MAINTAINERS: move BCMA to keep the list sorted MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Rafał Miłecki Signed-off-by: John W. Linville --- MAINTAINERS | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 1094edf0da17..442d2a70bbd6 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1587,6 +1587,13 @@ L: linux-scsi@vger.kernel.org S: Supported F: drivers/scsi/bnx2fc/ +BROADCOM SPECIFIC AMBA DRIVER (BCMA) +M: Rafał Miłecki +L: linux-wireless@vger.kernel.org +S: Maintained +F: drivers/bcma/ +F: include/linux/bcma/ + BROCADE BFA FC SCSI DRIVER M: Jing Huang L: linux-scsi@vger.kernel.org @@ -6099,13 +6106,6 @@ S: Maintained F: drivers/ssb/ F: include/linux/ssb/ -BROADCOM SPECIFIC AMBA DRIVER (BCMA) -M: Rafał Miłecki -L: linux-wireless@vger.kernel.org -S: Maintained -F: drivers/bcma/ -F: include/linux/bcma/ - SONY VAIO CONTROL DEVICE DRIVER M: Mattia Dongili L: platform-driver-x86@vger.kernel.org From 8675b505a4df40394096948453a17cc5445f6538 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= Date: Fri, 13 Jan 2012 22:55:06 +0100 Subject: [PATCH 387/820] MAINTAINERS: add b43 mailing list MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Rafał Miłecki Signed-off-by: John W. Linville --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 442d2a70bbd6..6e68daa1e64b 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1411,6 +1411,7 @@ F: net/ax25/ B43 WIRELESS DRIVER M: Stefano Brivio L: linux-wireless@vger.kernel.org +L: b43-dev@lists.infradead.org (moderated for non-subscribers) W: http://linuxwireless.org/en/users/Drivers/b43 S: Maintained F: drivers/net/wireless/b43/ From 28e7d218da975f6ae1751e293aed938952c55c98 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= Date: Fri, 13 Jan 2012 23:58:38 +0100 Subject: [PATCH 388/820] bcma: invalidate the mapped core over suspend/resume MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This clears the currently mapped core when suspending, to force re-mapping after resume. Without that we were touching default core registers believing some other core is mapped. Such a behaviour resulted in lockups on some machines. Cc: stable@vger.kernel.org Signed-off-by: Rafał Miłecki Signed-off-by: John W. Linville --- drivers/bcma/host_pci.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/bcma/host_pci.c b/drivers/bcma/host_pci.c index 443b83a2fd7a..c1ca9e3ada5c 100644 --- a/drivers/bcma/host_pci.c +++ b/drivers/bcma/host_pci.c @@ -237,11 +237,14 @@ static void bcma_host_pci_remove(struct pci_dev *dev) #ifdef CONFIG_PM static int bcma_host_pci_suspend(struct pci_dev *dev, pm_message_t state) { + struct bcma_bus *bus = pci_get_drvdata(dev); + /* Host specific */ pci_save_state(dev); pci_disable_device(dev); pci_set_power_state(dev, pci_choose_state(dev, state)); + bus->mapped_core = NULL; return 0; } From 7a532fe7131216a02c81a6c1b1f8632da1195a58 Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Sat, 14 Jan 2012 15:08:34 +0100 Subject: [PATCH 389/820] ath9k_hw: fix interpretation of the rx KeyMiss flag Documentation states that the KeyMiss flag is only valid if RxFrameOK is unset, however empirical evidence has shown that this is false. When KeyMiss is set (and RxFrameOK is 1), the hardware passes a valid frame which has not been decrypted. The driver then falsely marks the frame as decrypted, and when using CCMP this corrupts the rx CCMP PN, leading to connection hangs. Signed-off-by: Felix Fietkau Cc: stable@kernel.org Signed-off-by: John W. Linville --- drivers/net/wireless/ath/ath9k/ar9003_mac.c | 5 +++-- drivers/net/wireless/ath/ath9k/mac.c | 5 +++-- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/ar9003_mac.c b/drivers/net/wireless/ath/ath9k/ar9003_mac.c index 88c81c5706b2..09b8c9dbf78f 100644 --- a/drivers/net/wireless/ath/ath9k/ar9003_mac.c +++ b/drivers/net/wireless/ath/ath9k/ar9003_mac.c @@ -557,10 +557,11 @@ int ath9k_hw_process_rxdesc_edma(struct ath_hw *ah, struct ath_rx_status *rxs, rxs->rs_status |= ATH9K_RXERR_DECRYPT; else if (rxsp->status11 & AR_MichaelErr) rxs->rs_status |= ATH9K_RXERR_MIC; - if (rxsp->status11 & AR_KeyMiss) - rxs->rs_status |= ATH9K_RXERR_KEYMISS; } + if (rxsp->status11 & AR_KeyMiss) + rxs->rs_status |= ATH9K_RXERR_KEYMISS; + return 0; } EXPORT_SYMBOL(ath9k_hw_process_rxdesc_edma); diff --git a/drivers/net/wireless/ath/ath9k/mac.c b/drivers/net/wireless/ath/ath9k/mac.c index fd3f19c2e550..e196aba77acf 100644 --- a/drivers/net/wireless/ath/ath9k/mac.c +++ b/drivers/net/wireless/ath/ath9k/mac.c @@ -618,10 +618,11 @@ int ath9k_hw_rxprocdesc(struct ath_hw *ah, struct ath_desc *ds, rs->rs_status |= ATH9K_RXERR_DECRYPT; else if (ads.ds_rxstatus8 & AR_MichaelErr) rs->rs_status |= ATH9K_RXERR_MIC; - if (ads.ds_rxstatus8 & AR_KeyMiss) - rs->rs_status |= ATH9K_RXERR_KEYMISS; } + if (ads.ds_rxstatus8 & AR_KeyMiss) + rs->rs_status |= ATH9K_RXERR_KEYMISS; + return 0; } EXPORT_SYMBOL(ath9k_hw_rxprocdesc); From 74b8cc3d592e5e884123eb99960a9f064223eaab Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Sat, 14 Jan 2012 21:52:17 +0100 Subject: [PATCH 390/820] Net, mac80211: Fix resource leak in ieee80211_rx_h_mesh_fwding() We may leak the 'fwd_skb' we skb_copy() in ieee80211_rx_h_mesh_fwding() if we take the 'else' branch in the 'if' statement just below. If we take that branch we'll end up returning from the function and since we've not assigned 'fwd_skb' to anything at that point, we leak it when the variable goes out of scope. The simple fix seems to be to just kfree_skb(fwd_skb); just before we return. That is what this patch does. Signed-off-by: Jesper Juhl Signed-off-by: John W. Linville --- net/mac80211/rx.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index f407427c642f..751409120769 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1979,6 +1979,7 @@ ieee80211_rx_h_mesh_fwding(struct ieee80211_rx_data *rx) mesh_path_error_tx(ifmsh->mshcfg.element_ttl, fwd_hdr->addr3, 0, reason, fwd_hdr->addr2, sdata); IEEE80211_IFSTA_MESH_CTR_INC(ifmsh, dropped_frames_no_route); + kfree_skb(fwd_skb); return RX_DROP_MONITOR; } From 331d9301ed4d0fe33e9762c62f2870243bcf77eb Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Mon, 16 Jan 2012 00:47:12 -0200 Subject: [PATCH 391/820] wireless: iwlwifi: iwl-scan.c: Fix build warning MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the following build warning: drivers/net/wireless/iwlwifi/iwl-scan.c: In function ‘iwlagn_request_scan’: drivers/net/wireless/iwlwifi/iwl-scan.c:572: warning: ‘cmd_len’ may be used uninitialized in this function Signed-off-by: Fabio Estevam Acked-by: Wey-Yi Guy Signed-off-by: John W. Linville --- drivers/net/wireless/iwlwifi/iwl-scan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/iwlwifi/iwl-scan.c b/drivers/net/wireless/iwlwifi/iwl-scan.c index 084aa2c4ccfb..a6454726737e 100644 --- a/drivers/net/wireless/iwlwifi/iwl-scan.c +++ b/drivers/net/wireless/iwlwifi/iwl-scan.c @@ -569,7 +569,7 @@ static int iwlagn_request_scan(struct iwl_priv *priv, struct ieee80211_vif *vif) struct iwl_scan_cmd *scan; struct iwl_rxon_context *ctx = &priv->contexts[IWL_RXON_CTX_BSS]; u32 rate_flags = 0; - u16 cmd_len; + u16 cmd_len = 0; u16 rx_chain = 0; enum ieee80211_band band; u8 n_probes = 0; From d32a102819eef99d4d87019266bea208447be0a0 Mon Sep 17 00:00:00 2001 From: Yoni Divinsky Date: Mon, 16 Jan 2012 15:18:59 +0200 Subject: [PATCH 392/820] mac80211: fix tx->skb NULL pointer dereference In function ieee80211_tx_h_encrypt the var info was initialized from tx->skb, since the fucntion is called after the function ieee80211_tx_h_fragment tx->skb is not valid anymore. Signed-off-by: Yoni Divinsky Signed-off-by: John W. Linville --- net/mac80211/tx.c | 10 +--------- net/mac80211/wpa.c | 19 +++++++++++++++++++ net/mac80211/wpa.h | 2 ++ 3 files changed, 22 insertions(+), 9 deletions(-) diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index edcd1c7ab83f..e05667cd5e76 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -1001,8 +1001,6 @@ ieee80211_tx_h_stats(struct ieee80211_tx_data *tx) static ieee80211_tx_result debug_noinline ieee80211_tx_h_encrypt(struct ieee80211_tx_data *tx) { - struct ieee80211_tx_info *info = IEEE80211_SKB_CB(tx->skb); - if (!tx->key) return TX_CONTINUE; @@ -1017,13 +1015,7 @@ ieee80211_tx_h_encrypt(struct ieee80211_tx_data *tx) case WLAN_CIPHER_SUITE_AES_CMAC: return ieee80211_crypto_aes_cmac_encrypt(tx); default: - /* handle hw-only algorithm */ - if (info->control.hw_key) { - ieee80211_tx_set_protected(tx); - return TX_CONTINUE; - } - break; - + return ieee80211_crypto_hw_encrypt(tx); } return TX_DROP; diff --git a/net/mac80211/wpa.c b/net/mac80211/wpa.c index 422b79851ec5..b758350919ff 100644 --- a/net/mac80211/wpa.c +++ b/net/mac80211/wpa.c @@ -643,3 +643,22 @@ ieee80211_crypto_aes_cmac_decrypt(struct ieee80211_rx_data *rx) return RX_CONTINUE; } + +ieee80211_tx_result +ieee80211_crypto_hw_encrypt(struct ieee80211_tx_data *tx) +{ + struct sk_buff *skb; + struct ieee80211_tx_info *info = NULL; + + skb_queue_walk(&tx->skbs, skb) { + info = IEEE80211_SKB_CB(skb); + + /* handle hw-only algorithm */ + if (!info->control.hw_key) + return TX_DROP; + } + + ieee80211_tx_set_protected(tx); + + return TX_CONTINUE; +} diff --git a/net/mac80211/wpa.h b/net/mac80211/wpa.h index baba0608313e..07e33f899c71 100644 --- a/net/mac80211/wpa.h +++ b/net/mac80211/wpa.h @@ -32,5 +32,7 @@ ieee80211_tx_result ieee80211_crypto_aes_cmac_encrypt(struct ieee80211_tx_data *tx); ieee80211_rx_result ieee80211_crypto_aes_cmac_decrypt(struct ieee80211_rx_data *rx); +ieee80211_tx_result +ieee80211_crypto_hw_encrypt(struct ieee80211_tx_data *tx); #endif /* WPA_H */ From 6fef8df1dcb9b586268caff66df1d71ce8610132 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 393/820] Btrfs: get rid of *_alloc_profile fields {data,metadata,system}_alloc_profile fields have been unused for a long time now. Get rid of them. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 3 --- fs/btrfs/disk-io.c | 3 --- fs/btrfs/extent-tree.c | 10 ++++------ fs/btrfs/volumes.c | 6 ++---- 4 files changed, 6 insertions(+), 16 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 67385033323d..f5434ad49b99 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1135,9 +1135,6 @@ struct btrfs_fs_info { u64 avail_data_alloc_bits; u64 avail_metadata_alloc_bits; u64 avail_system_alloc_bits; - u64 data_alloc_profile; - u64 metadata_alloc_profile; - u64 system_alloc_profile; unsigned data_chunk_allocations; unsigned metadata_ratio; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 3f9d5551e582..ce9d0fb3627d 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2321,9 +2321,6 @@ retry_root_backup: fs_info->generation = generation; fs_info->last_trans_committed = generation; - fs_info->data_alloc_profile = (u64)-1; - fs_info->metadata_alloc_profile = (u64)-1; - fs_info->system_alloc_profile = fs_info->metadata_alloc_profile; ret = btrfs_init_space_info(fs_info); if (ret) { diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 8603ee4e3dfd..f0591fd66249 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3067,14 +3067,12 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) static u64 get_alloc_profile(struct btrfs_root *root, u64 flags) { if (flags & BTRFS_BLOCK_GROUP_DATA) - flags |= root->fs_info->avail_data_alloc_bits & - root->fs_info->data_alloc_profile; + flags |= root->fs_info->avail_data_alloc_bits; else if (flags & BTRFS_BLOCK_GROUP_SYSTEM) - flags |= root->fs_info->avail_system_alloc_bits & - root->fs_info->system_alloc_profile; + flags |= root->fs_info->avail_system_alloc_bits; else if (flags & BTRFS_BLOCK_GROUP_METADATA) - flags |= root->fs_info->avail_metadata_alloc_bits & - root->fs_info->metadata_alloc_profile; + flags |= root->fs_info->avail_metadata_alloc_bits; + return btrfs_reduce_alloc_profile(root, flags); } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f4b839fd3c9d..89096f6d6fb4 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2752,8 +2752,7 @@ static noinline int init_first_rw_device(struct btrfs_trans_handle *trans, return ret; alloc_profile = BTRFS_BLOCK_GROUP_METADATA | - (fs_info->metadata_alloc_profile & - fs_info->avail_metadata_alloc_bits); + fs_info->avail_metadata_alloc_bits; alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile); ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size, @@ -2763,8 +2762,7 @@ static noinline int init_first_rw_device(struct btrfs_trans_handle *trans, sys_chunk_offset = chunk_offset + chunk_size; alloc_profile = BTRFS_BLOCK_GROUP_SYSTEM | - (fs_info->system_alloc_profile & - fs_info->avail_system_alloc_bits); + fs_info->avail_system_alloc_bits; alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile); ret = __btrfs_alloc_chunk(trans, extent_root, &sys_map, From 52ba692972532f8d652080214b6599ece3dd51b9 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 394/820] Btrfs: introduce masks for chunk type and profile Chunk's type and profile are encoded in u64 flags field. Introduce masks to easily access them. Also fix the type of BTRFS_BLOCK_GROUP_* constants, it should be ULL. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 24 ++++++++++++++++-------- fs/btrfs/extent-tree.c | 12 +++--------- fs/btrfs/volumes.c | 11 ++--------- 3 files changed, 21 insertions(+), 26 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index f5434ad49b99..4370a56fe81a 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -751,15 +751,23 @@ struct btrfs_csum_item { } __attribute__ ((__packed__)); /* different types of block groups (and chunks) */ -#define BTRFS_BLOCK_GROUP_DATA (1 << 0) -#define BTRFS_BLOCK_GROUP_SYSTEM (1 << 1) -#define BTRFS_BLOCK_GROUP_METADATA (1 << 2) -#define BTRFS_BLOCK_GROUP_RAID0 (1 << 3) -#define BTRFS_BLOCK_GROUP_RAID1 (1 << 4) -#define BTRFS_BLOCK_GROUP_DUP (1 << 5) -#define BTRFS_BLOCK_GROUP_RAID10 (1 << 6) -#define BTRFS_NR_RAID_TYPES 5 +#define BTRFS_BLOCK_GROUP_DATA (1ULL << 0) +#define BTRFS_BLOCK_GROUP_SYSTEM (1ULL << 1) +#define BTRFS_BLOCK_GROUP_METADATA (1ULL << 2) +#define BTRFS_BLOCK_GROUP_RAID0 (1ULL << 3) +#define BTRFS_BLOCK_GROUP_RAID1 (1ULL << 4) +#define BTRFS_BLOCK_GROUP_DUP (1ULL << 5) +#define BTRFS_BLOCK_GROUP_RAID10 (1ULL << 6) +#define BTRFS_NR_RAID_TYPES 5 +#define BTRFS_BLOCK_GROUP_TYPE_MASK (BTRFS_BLOCK_GROUP_DATA | \ + BTRFS_BLOCK_GROUP_SYSTEM | \ + BTRFS_BLOCK_GROUP_METADATA) + +#define BTRFS_BLOCK_GROUP_PROFILE_MASK (BTRFS_BLOCK_GROUP_RAID0 | \ + BTRFS_BLOCK_GROUP_RAID1 | \ + BTRFS_BLOCK_GROUP_DUP | \ + BTRFS_BLOCK_GROUP_RAID10) struct btrfs_block_group_item { __le64 used; __le64 chunk_objectid; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index f0591fd66249..a8d8204188d1 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -618,8 +618,7 @@ static struct btrfs_space_info *__find_space_info(struct btrfs_fs_info *info, struct list_head *head = &info->space_info; struct btrfs_space_info *found; - flags &= BTRFS_BLOCK_GROUP_DATA | BTRFS_BLOCK_GROUP_SYSTEM | - BTRFS_BLOCK_GROUP_METADATA; + flags &= BTRFS_BLOCK_GROUP_TYPE_MASK; rcu_read_lock(); list_for_each_entry_rcu(found, head, list) { @@ -2993,9 +2992,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, INIT_LIST_HEAD(&found->block_groups[i]); init_rwsem(&found->groups_sem); spin_lock_init(&found->lock); - found->flags = flags & (BTRFS_BLOCK_GROUP_DATA | - BTRFS_BLOCK_GROUP_SYSTEM | - BTRFS_BLOCK_GROUP_METADATA); + found->flags = flags & BTRFS_BLOCK_GROUP_TYPE_MASK; found->total_bytes = total_bytes; found->disk_total = total_bytes * factor; found->bytes_used = bytes_used; @@ -3016,10 +3013,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) { - u64 extra_flags = flags & (BTRFS_BLOCK_GROUP_RAID0 | - BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_RAID10 | - BTRFS_BLOCK_GROUP_DUP); + u64 extra_flags = flags & BTRFS_BLOCK_GROUP_PROFILE_MASK; if (extra_flags) { if (flags & BTRFS_BLOCK_GROUP_DATA) fs_info->avail_data_alloc_bits |= extra_flags; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 89096f6d6fb4..d5fdee56777f 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2949,12 +2949,8 @@ again: } } if (rw & REQ_DISCARD) { - if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | - BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_DUP | - BTRFS_BLOCK_GROUP_RAID10)) { + if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) stripes_required = map->num_stripes; - } } if (bbio_ret && (rw & (REQ_WRITE | REQ_DISCARD)) && stripes_allocated < stripes_required) { @@ -2978,10 +2974,7 @@ again: if (rw & REQ_DISCARD) *length = min_t(u64, em->len - offset, *length); - else if (map->type & (BTRFS_BLOCK_GROUP_RAID0 | - BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_RAID10 | - BTRFS_BLOCK_GROUP_DUP)) { + else if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) { /* we limit the length of each bio to what fits in a stripe */ *length = min_t(u64, em->len - offset, map->stripe_len - stripe_offset); From a46d11a8b06dd0431a3888fbc4856ea13a8e634f Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 395/820] Btrfs: add BTRFS_AVAIL_ALLOC_BIT_SINGLE bit Right now on-disk BTRFS_BLOCK_GROUP_* profile bits are used for avail_{data,metadata,system}_alloc_bits fields, which gather info about available allocation profiles in the FS. When chunk is created or read from disk, its profile is OR'ed with the corresponding avail_alloc_bits field. Since SINGLE is denoted by 0 in the on-disk format, currently there is no way to tell when such chunks become avaialble. Restriper needs that information, so add a separate bit for SINGLE profile. This bit is going to be in-memory only, it should never be written out to disk, so it's not a disk format change. However to avoid remappings in future, reserve corresponding on-disk bit. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 15 +++++++++++++++ fs/btrfs/extent-tree.c | 30 +++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 9 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 4370a56fe81a..3f8f11e18b53 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -758,6 +758,7 @@ struct btrfs_csum_item { #define BTRFS_BLOCK_GROUP_RAID1 (1ULL << 4) #define BTRFS_BLOCK_GROUP_DUP (1ULL << 5) #define BTRFS_BLOCK_GROUP_RAID10 (1ULL << 6) +#define BTRFS_BLOCK_GROUP_RESERVED BTRFS_AVAIL_ALLOC_BIT_SINGLE #define BTRFS_NR_RAID_TYPES 5 #define BTRFS_BLOCK_GROUP_TYPE_MASK (BTRFS_BLOCK_GROUP_DATA | \ @@ -768,6 +769,15 @@ struct btrfs_csum_item { BTRFS_BLOCK_GROUP_RAID1 | \ BTRFS_BLOCK_GROUP_DUP | \ BTRFS_BLOCK_GROUP_RAID10) +/* + * We need a bit for restriper to be able to tell when chunks of type + * SINGLE are available. This "extended" profile format is used in + * fs_info->avail_*_alloc_bits (in-memory) and balance item fields + * (on-disk). The corresponding on-disk bit in chunk.type is reserved + * to avoid remappings between two formats in future. + */ +#define BTRFS_AVAIL_ALLOC_BIT_SINGLE (1ULL << 48) + struct btrfs_block_group_item { __le64 used; __le64 chunk_objectid; @@ -1140,6 +1150,11 @@ struct btrfs_fs_info { spinlock_t ref_cache_lock; u64 total_ref_cache_size; + /* + * these three are in extended format (availability of single + * chunks is denoted by BTRFS_AVAIL_ALLOC_BIT_SINGLE bit, other + * types are denoted by corresponding BTRFS_BLOCK_GROUP_* bits) + */ u64 avail_data_alloc_bits; u64 avail_metadata_alloc_bits; u64 avail_system_alloc_bits; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a8d8204188d1..15a22949da17 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3014,16 +3014,24 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) { u64 extra_flags = flags & BTRFS_BLOCK_GROUP_PROFILE_MASK; - if (extra_flags) { - if (flags & BTRFS_BLOCK_GROUP_DATA) - fs_info->avail_data_alloc_bits |= extra_flags; - if (flags & BTRFS_BLOCK_GROUP_METADATA) - fs_info->avail_metadata_alloc_bits |= extra_flags; - if (flags & BTRFS_BLOCK_GROUP_SYSTEM) - fs_info->avail_system_alloc_bits |= extra_flags; - } + + /* chunk -> extended profile */ + if (extra_flags == 0) + extra_flags = BTRFS_AVAIL_ALLOC_BIT_SINGLE; + + if (flags & BTRFS_BLOCK_GROUP_DATA) + fs_info->avail_data_alloc_bits |= extra_flags; + if (flags & BTRFS_BLOCK_GROUP_METADATA) + fs_info->avail_metadata_alloc_bits |= extra_flags; + if (flags & BTRFS_BLOCK_GROUP_SYSTEM) + fs_info->avail_system_alloc_bits |= extra_flags; } +/* + * @flags: available profiles in extended format (see ctree.h) + * + * Returns reduced profile in chunk format. + */ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) { /* @@ -3053,8 +3061,12 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) if ((flags & BTRFS_BLOCK_GROUP_RAID0) && ((flags & BTRFS_BLOCK_GROUP_RAID1) | (flags & BTRFS_BLOCK_GROUP_RAID10) | - (flags & BTRFS_BLOCK_GROUP_DUP))) + (flags & BTRFS_BLOCK_GROUP_DUP))) { flags &= ~BTRFS_BLOCK_GROUP_RAID0; + } + + /* extended -> chunk profile */ + flags &= ~BTRFS_AVAIL_ALLOC_BIT_SINGLE; return flags; } From 10ea00f55a07f8f9536d9112b95108a86f700bab Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 396/820] Btrfs: make avail_*_alloc_bits fields dynamic Currently when new chunks are created respective avail_alloc_bits field is updated to reflect profiles of all chunks present in the system. However when chunks are removed profile bits are never cleared. This patch clears profile bit of respective avail_alloc_bits field when the last chunk with that profile is removed. Restriper needs this to properly operate when "downgrading". Signed-off-by: Ilya Dryomov --- fs/btrfs/extent-tree.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 15a22949da17..946b0671c5d9 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -7469,6 +7469,22 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, return 0; } +static void clear_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) +{ + u64 extra_flags = flags & BTRFS_BLOCK_GROUP_PROFILE_MASK; + + /* chunk -> extended profile */ + if (extra_flags == 0) + extra_flags = BTRFS_AVAIL_ALLOC_BIT_SINGLE; + + if (flags & BTRFS_BLOCK_GROUP_DATA) + fs_info->avail_data_alloc_bits &= ~extra_flags; + if (flags & BTRFS_BLOCK_GROUP_METADATA) + fs_info->avail_metadata_alloc_bits &= ~extra_flags; + if (flags & BTRFS_BLOCK_GROUP_SYSTEM) + fs_info->avail_system_alloc_bits &= ~extra_flags; +} + int btrfs_remove_block_group(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 group_start) { @@ -7479,6 +7495,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, struct btrfs_key key; struct inode *inode; int ret; + int index; int factor; root = root->fs_info->extent_root; @@ -7494,6 +7511,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, free_excluded_extents(root, block_group); memcpy(&key, &block_group->key, sizeof(key)); + index = get_block_group_index(block_group); if (block_group->flags & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)) @@ -7568,6 +7586,8 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, * are still on the list after taking the semaphore */ list_del_init(&block_group->list); + if (list_empty(&block_group->space_info->block_groups[index])) + clear_avail_alloc_bits(root->fs_info, block_group->flags); up_write(&block_group->space_info->groups_sem); if (block_group->cached == BTRFS_CACHE_STARTED) From c9e9f97bdfb64d06e9520f8e4f37674ac21cc9bc Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 397/820] Btrfs: add basic restriper infrastructure Add basic restriper infrastructure: extended balancing ioctl and all related ioctl data structures, add data structure for tracking restriper's state to fs_info, etc. The semantics of the old balancing ioctl are fully preserved. Explicitly disallow any volume operations when balance is in progress. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 6 ++ fs/btrfs/disk-io.c | 4 ++ fs/btrfs/ioctl.c | 135 +++++++++++++++++++++++++++++++++++++++------ fs/btrfs/ioctl.h | 43 +++++++++++++++ fs/btrfs/volumes.c | 120 ++++++++++++++++++++++++++++++++-------- fs/btrfs/volumes.h | 14 ++++- 6 files changed, 281 insertions(+), 41 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3f8f11e18b53..c4d98c8df5c5 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -934,6 +934,7 @@ struct btrfs_block_group_cache { struct reloc_control; struct btrfs_device; struct btrfs_fs_devices; +struct btrfs_balance_control; struct btrfs_delayed_root; struct btrfs_fs_info { u8 fsid[BTRFS_FSID_SIZE]; @@ -1159,6 +1160,11 @@ struct btrfs_fs_info { u64 avail_metadata_alloc_bits; u64 avail_system_alloc_bits; + /* restriper state */ + spinlock_t balance_lock; + struct mutex balance_mutex; + struct btrfs_balance_control *balance_ctl; + unsigned data_chunk_allocations; unsigned metadata_ratio; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index ce9d0fb3627d..190a1b24c902 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2002,6 +2002,10 @@ struct btrfs_root *open_ctree(struct super_block *sb, init_rwsem(&fs_info->scrub_super_lock); fs_info->scrub_workers_refcnt = 0; + spin_lock_init(&fs_info->balance_lock); + mutex_init(&fs_info->balance_mutex); + fs_info->balance_ctl = NULL; + sb->s_blocksize = 4096; sb->s_blocksize_bits = blksize_bits(4096); sb->s_bdi = &fs_info->bdi; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index c04f02c7d5bb..d838d2cfb947 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1203,13 +1203,21 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, if (!capable(CAP_SYS_ADMIN)) return -EPERM; + mutex_lock(&root->fs_info->volume_mutex); + if (root->fs_info->balance_ctl) { + printk(KERN_INFO "btrfs: balance in progress\n"); + ret = -EINVAL; + goto out; + } + vol_args = memdup_user(arg, sizeof(*vol_args)); - if (IS_ERR(vol_args)) - return PTR_ERR(vol_args); + if (IS_ERR(vol_args)) { + ret = PTR_ERR(vol_args); + goto out; + } vol_args->name[BTRFS_PATH_NAME_MAX] = '\0'; - mutex_lock(&root->fs_info->volume_mutex); sizestr = vol_args->name; devstr = strchr(sizestr, ':'); if (devstr) { @@ -1226,7 +1234,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, printk(KERN_INFO "btrfs: resizer unable to find device %llu\n", (unsigned long long)devid); ret = -EINVAL; - goto out_unlock; + goto out_free; } if (!strcmp(sizestr, "max")) new_size = device->bdev->bd_inode->i_size; @@ -1241,7 +1249,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, new_size = memparse(sizestr, NULL); if (new_size == 0) { ret = -EINVAL; - goto out_unlock; + goto out_free; } } @@ -1250,7 +1258,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, if (mod < 0) { if (new_size > old_size) { ret = -EINVAL; - goto out_unlock; + goto out_free; } new_size = old_size - new_size; } else if (mod > 0) { @@ -1259,11 +1267,11 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, if (new_size < 256 * 1024 * 1024) { ret = -EINVAL; - goto out_unlock; + goto out_free; } if (new_size > device->bdev->bd_inode->i_size) { ret = -EFBIG; - goto out_unlock; + goto out_free; } do_div(new_size, root->sectorsize); @@ -1276,7 +1284,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, trans = btrfs_start_transaction(root, 0); if (IS_ERR(trans)) { ret = PTR_ERR(trans); - goto out_unlock; + goto out_free; } ret = btrfs_grow_device(trans, device, new_size); btrfs_commit_transaction(trans, root); @@ -1284,9 +1292,10 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, ret = btrfs_shrink_device(device, new_size); } -out_unlock: - mutex_unlock(&root->fs_info->volume_mutex); +out_free: kfree(vol_args); +out: + mutex_unlock(&root->fs_info->volume_mutex); return ret; } @@ -2052,14 +2061,25 @@ static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg) if (!capable(CAP_SYS_ADMIN)) return -EPERM; + mutex_lock(&root->fs_info->volume_mutex); + if (root->fs_info->balance_ctl) { + printk(KERN_INFO "btrfs: balance in progress\n"); + ret = -EINVAL; + goto out; + } + vol_args = memdup_user(arg, sizeof(*vol_args)); - if (IS_ERR(vol_args)) - return PTR_ERR(vol_args); + if (IS_ERR(vol_args)) { + ret = PTR_ERR(vol_args); + goto out; + } vol_args->name[BTRFS_PATH_NAME_MAX] = '\0'; ret = btrfs_init_new_device(root, vol_args->name); kfree(vol_args); +out: + mutex_unlock(&root->fs_info->volume_mutex); return ret; } @@ -2074,14 +2094,25 @@ static long btrfs_ioctl_rm_dev(struct btrfs_root *root, void __user *arg) if (root->fs_info->sb->s_flags & MS_RDONLY) return -EROFS; + mutex_lock(&root->fs_info->volume_mutex); + if (root->fs_info->balance_ctl) { + printk(KERN_INFO "btrfs: balance in progress\n"); + ret = -EINVAL; + goto out; + } + vol_args = memdup_user(arg, sizeof(*vol_args)); - if (IS_ERR(vol_args)) - return PTR_ERR(vol_args); + if (IS_ERR(vol_args)) { + ret = PTR_ERR(vol_args); + goto out; + } vol_args->name[BTRFS_PATH_NAME_MAX] = '\0'; ret = btrfs_rm_device(root, vol_args->name); kfree(vol_args); +out: + mutex_unlock(&root->fs_info->volume_mutex); return ret; } @@ -3034,6 +3065,76 @@ out: return ret; } +void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, + struct btrfs_ioctl_balance_args *bargs) +{ + struct btrfs_balance_control *bctl = fs_info->balance_ctl; + + bargs->flags = bctl->flags; + + memcpy(&bargs->data, &bctl->data, sizeof(bargs->data)); + memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta)); + memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys)); +} + +static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + struct btrfs_ioctl_balance_args *bargs; + struct btrfs_balance_control *bctl; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (fs_info->sb->s_flags & MS_RDONLY) + return -EROFS; + + mutex_lock(&fs_info->volume_mutex); + mutex_lock(&fs_info->balance_mutex); + + if (arg) { + bargs = memdup_user(arg, sizeof(*bargs)); + if (IS_ERR(bargs)) { + ret = PTR_ERR(bargs); + goto out; + } + } else { + bargs = NULL; + } + + bctl = kzalloc(sizeof(*bctl), GFP_NOFS); + if (!bctl) { + ret = -ENOMEM; + goto out_bargs; + } + + bctl->fs_info = fs_info; + if (arg) { + memcpy(&bctl->data, &bargs->data, sizeof(bctl->data)); + memcpy(&bctl->meta, &bargs->meta, sizeof(bctl->meta)); + memcpy(&bctl->sys, &bargs->sys, sizeof(bctl->sys)); + + bctl->flags = bargs->flags; + } + + ret = btrfs_balance(bctl, bargs); + /* + * bctl is freed in __cancel_balance + */ + if (arg) { + if (copy_to_user(arg, bargs, sizeof(*bargs))) + ret = -EFAULT; + } + +out_bargs: + kfree(bargs); +out: + mutex_unlock(&fs_info->balance_mutex); + mutex_unlock(&fs_info->volume_mutex); + return ret; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -3078,7 +3179,7 @@ long btrfs_ioctl(struct file *file, unsigned int case BTRFS_IOC_DEV_INFO: return btrfs_ioctl_dev_info(root, argp); case BTRFS_IOC_BALANCE: - return btrfs_balance(root->fs_info->dev_root); + return btrfs_ioctl_balance(root, NULL); case BTRFS_IOC_CLONE: return btrfs_ioctl_clone(file, arg, 0, 0, 0); case BTRFS_IOC_CLONE_RANGE: @@ -3110,6 +3211,8 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_scrub_cancel(root, argp); case BTRFS_IOC_SCRUB_PROGRESS: return btrfs_ioctl_scrub_progress(root, argp); + case BTRFS_IOC_BALANCE_V2: + return btrfs_ioctl_balance(root, argp); } return -ENOTTY; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index 252ae9915de8..c8b37d2c0d77 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -109,6 +109,47 @@ struct btrfs_ioctl_fs_info_args { __u64 reserved[124]; /* pad to 1k */ }; +/* + * this is packed, because it should be exactly the same as its disk + * byte order counterpart (struct btrfs_disk_balance_args) + */ +struct btrfs_balance_args { + __u64 profiles; + __u64 usage; + __u64 devid; + __u64 pstart; + __u64 pend; + __u64 vstart; + __u64 vend; + + __u64 target; + + __u64 flags; + + __u64 unused[8]; +} __attribute__ ((__packed__)); + +/* report balance progress to userspace */ +struct btrfs_balance_progress { + __u64 expected; /* estimated # of chunks that will be + * relocated to fulfill the request */ + __u64 considered; /* # of chunks we have considered so far */ + __u64 completed; /* # of chunks relocated so far */ +}; + +struct btrfs_ioctl_balance_args { + __u64 flags; /* in/out */ + __u64 state; /* out */ + + struct btrfs_balance_args data; /* in/out */ + struct btrfs_balance_args meta; /* in/out */ + struct btrfs_balance_args sys; /* in/out */ + + struct btrfs_balance_progress stat; /* out */ + + __u64 unused[72]; /* pad to 1k */ +}; + #define BTRFS_INO_LOOKUP_PATH_MAX 4080 struct btrfs_ioctl_ino_lookup_args { __u64 treeid; @@ -272,6 +313,8 @@ struct btrfs_ioctl_logical_ino_args { struct btrfs_ioctl_dev_info_args) #define BTRFS_IOC_FS_INFO _IOR(BTRFS_IOCTL_MAGIC, 31, \ struct btrfs_ioctl_fs_info_args) +#define BTRFS_IOC_BALANCE_V2 _IOWR(BTRFS_IOCTL_MAGIC, 32, \ + struct btrfs_ioctl_balance_args) #define BTRFS_IOC_INO_PATHS _IOWR(BTRFS_IOCTL_MAGIC, 35, \ struct btrfs_ioctl_ino_path_args) #define BTRFS_IOC_LOGICAL_INO _IOWR(BTRFS_IOCTL_MAGIC, 36, \ diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d5fdee56777f..9fc06e6bc51a 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1282,7 +1282,6 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) bool clear_super = false; mutex_lock(&uuid_mutex); - mutex_lock(&root->fs_info->volume_mutex); all_avail = root->fs_info->avail_data_alloc_bits | root->fs_info->avail_system_alloc_bits | @@ -1452,7 +1451,6 @@ error_close: if (bdev) blkdev_put(bdev, FMODE_READ | FMODE_EXCL); out: - mutex_unlock(&root->fs_info->volume_mutex); mutex_unlock(&uuid_mutex); return ret; error_undo: @@ -1629,7 +1627,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) } filemap_write_and_wait(bdev->bd_inode->i_mapping); - mutex_lock(&root->fs_info->volume_mutex); devices = &root->fs_info->fs_devices->devices; /* @@ -1757,8 +1754,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) ret = btrfs_relocate_sys_chunks(root); BUG_ON(ret); } -out: - mutex_unlock(&root->fs_info->volume_mutex); + return ret; error: blkdev_put(bdev, FMODE_EXCL); @@ -1766,7 +1762,7 @@ error: mutex_unlock(&uuid_mutex); up_write(&sb->s_umount); } - goto out; + return ret; } static noinline int btrfs_update_device(struct btrfs_trans_handle *trans, @@ -2077,6 +2073,35 @@ error: return ret; } +/* + * Should be called with both balance and volume mutexes held to + * serialize other volume operations (add_dev/rm_dev/resize) with + * restriper. Same goes for unset_balance_control. + */ +static void set_balance_control(struct btrfs_balance_control *bctl) +{ + struct btrfs_fs_info *fs_info = bctl->fs_info; + + BUG_ON(fs_info->balance_ctl); + + spin_lock(&fs_info->balance_lock); + fs_info->balance_ctl = bctl; + spin_unlock(&fs_info->balance_lock); +} + +static void unset_balance_control(struct btrfs_fs_info *fs_info) +{ + struct btrfs_balance_control *bctl = fs_info->balance_ctl; + + BUG_ON(!fs_info->balance_ctl); + + spin_lock(&fs_info->balance_lock); + fs_info->balance_ctl = NULL; + spin_unlock(&fs_info->balance_lock); + + kfree(bctl); +} + static u64 div_factor(u64 num, int factor) { if (factor == 10) @@ -2086,29 +2111,23 @@ static u64 div_factor(u64 num, int factor) return num; } -int btrfs_balance(struct btrfs_root *dev_root) +static int __btrfs_balance(struct btrfs_fs_info *fs_info) { - int ret; - struct list_head *devices = &dev_root->fs_info->fs_devices->devices; + struct btrfs_root *chunk_root = fs_info->chunk_root; + struct btrfs_root *dev_root = fs_info->dev_root; + struct list_head *devices; struct btrfs_device *device; u64 old_size; u64 size_to_free; struct btrfs_path *path; struct btrfs_key key; - struct btrfs_root *chunk_root = dev_root->fs_info->chunk_root; - struct btrfs_trans_handle *trans; struct btrfs_key found_key; - - if (dev_root->fs_info->sb->s_flags & MS_RDONLY) - return -EROFS; - - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - - mutex_lock(&dev_root->fs_info->volume_mutex); - dev_root = dev_root->fs_info->dev_root; + struct btrfs_trans_handle *trans; + int ret; + int enospc_errors = 0; /* step one make some room on all the devices */ + devices = &fs_info->fs_devices->devices; list_for_each_entry(device, devices, dev_list) { old_size = device->total_bytes; size_to_free = div_factor(old_size, 1); @@ -2151,12 +2170,14 @@ int btrfs_balance(struct btrfs_root *dev_root) * failed */ if (ret == 0) - break; + BUG(); /* FIXME break ? */ ret = btrfs_previous_item(chunk_root, path, 0, BTRFS_CHUNK_ITEM_KEY); - if (ret) + if (ret) { + ret = 0; break; + } btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]); @@ -2174,12 +2195,63 @@ int btrfs_balance(struct btrfs_root *dev_root) found_key.offset); if (ret && ret != -ENOSPC) goto error; + if (ret == -ENOSPC) + enospc_errors++; key.offset = found_key.offset - 1; } - ret = 0; + error: btrfs_free_path(path); - mutex_unlock(&dev_root->fs_info->volume_mutex); + if (enospc_errors) { + printk(KERN_INFO "btrfs: %d enospc errors during balance\n", + enospc_errors); + if (!ret) + ret = -ENOSPC; + } + + return ret; +} + +static void __cancel_balance(struct btrfs_fs_info *fs_info) +{ + unset_balance_control(fs_info); +} + +void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, + struct btrfs_ioctl_balance_args *bargs); + +/* + * Should be called with both balance and volume mutexes held + */ +int btrfs_balance(struct btrfs_balance_control *bctl, + struct btrfs_ioctl_balance_args *bargs) +{ + struct btrfs_fs_info *fs_info = bctl->fs_info; + int ret; + + if (btrfs_fs_closing(fs_info)) { + ret = -EINVAL; + goto out; + } + + set_balance_control(bctl); + + mutex_unlock(&fs_info->balance_mutex); + + ret = __btrfs_balance(fs_info); + + mutex_lock(&fs_info->balance_mutex); + + if (bargs) { + memset(bargs, 0, sizeof(*bargs)); + update_ioctl_balance_args(fs_info, bargs); + } + + __cancel_balance(fs_info); + + return ret; +out: + kfree(bctl); return ret; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 78f2d4d4f37f..882582385283 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -186,6 +186,17 @@ struct map_lookup { #define map_lookup_size(n) (sizeof(struct map_lookup) + \ (sizeof(struct btrfs_bio_stripe) * (n))) +struct btrfs_balance_args; +struct btrfs_balance_control { + struct btrfs_fs_info *fs_info; + + struct btrfs_balance_args data; + struct btrfs_balance_args meta; + struct btrfs_balance_args sys; + + u64 flags; +}; + int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start, u64 end, u64 *length); @@ -228,7 +239,8 @@ struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid, u8 *uuid, u8 *fsid); int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); -int btrfs_balance(struct btrfs_root *dev_root); +int btrfs_balance(struct btrfs_balance_control *bctl, + struct btrfs_ioctl_balance_args *bargs); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); int find_free_dev_extent(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 num_bytes, From f43ffb60fd94e98be02780944e182ade6653b916 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 398/820] Btrfs: add basic infrastructure for selective balancing This allows to have a separate set of filters for each chunk type (data,meta,sys). The code however is generic and switch on chunk type is only done once. This commit also adds a type filter: it allows to balance for example meta and system chunks w/o touching data ones. Signed-off-by: Ilya Dryomov --- fs/btrfs/ioctl.c | 3 +++ fs/btrfs/volumes.c | 59 ++++++++++++++++++++++++++++++++++++++++++++-- fs/btrfs/volumes.h | 11 +++++++++ 3 files changed, 71 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index d838d2cfb947..29b3a94933f0 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3116,6 +3116,9 @@ static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) memcpy(&bctl->sys, &bargs->sys, sizeof(bctl->sys)); bctl->flags = bargs->flags; + } else { + /* balance everything - no filters */ + bctl->flags |= BTRFS_BALANCE_TYPE_MASK; } ret = btrfs_balance(bctl, bargs); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 9fc06e6bc51a..91bbf6e774c0 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2102,6 +2102,30 @@ static void unset_balance_control(struct btrfs_fs_info *fs_info) kfree(bctl); } +static int should_balance_chunk(struct btrfs_root *root, + struct extent_buffer *leaf, + struct btrfs_chunk *chunk, u64 chunk_offset) +{ + struct btrfs_balance_control *bctl = root->fs_info->balance_ctl; + struct btrfs_balance_args *bargs = NULL; + u64 chunk_type = btrfs_chunk_type(leaf, chunk); + + /* type filter */ + if (!((chunk_type & BTRFS_BLOCK_GROUP_TYPE_MASK) & + (bctl->flags & BTRFS_BALANCE_TYPE_MASK))) { + return 0; + } + + if (chunk_type & BTRFS_BLOCK_GROUP_DATA) + bargs = &bctl->data; + else if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM) + bargs = &bctl->sys; + else if (chunk_type & BTRFS_BLOCK_GROUP_METADATA) + bargs = &bctl->meta; + + return 1; +} + static u64 div_factor(u64 num, int factor) { if (factor == 10) @@ -2119,10 +2143,13 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) struct btrfs_device *device; u64 old_size; u64 size_to_free; + struct btrfs_chunk *chunk; struct btrfs_path *path; struct btrfs_key key; struct btrfs_key found_key; struct btrfs_trans_handle *trans; + struct extent_buffer *leaf; + int slot; int ret; int enospc_errors = 0; @@ -2179,8 +2206,10 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) break; } - btrfs_item_key_to_cpu(path->nodes[0], &found_key, - path->slots[0]); + leaf = path->nodes[0]; + slot = path->slots[0]; + btrfs_item_key_to_cpu(leaf, &found_key, slot); + if (found_key.objectid != key.objectid) break; @@ -2188,7 +2217,14 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) if (found_key.offset == 0) break; + chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk); + + ret = should_balance_chunk(chunk_root, leaf, chunk, + found_key.offset); btrfs_release_path(path); + if (!ret) + goto loop; + ret = btrfs_relocate_chunk(chunk_root, chunk_root->root_key.objectid, found_key.objectid, @@ -2197,6 +2233,7 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) goto error; if (ret == -ENOSPC) enospc_errors++; +loop: key.offset = found_key.offset - 1; } @@ -2227,6 +2264,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs) { struct btrfs_fs_info *fs_info = bctl->fs_info; + u64 allowed; int ret; if (btrfs_fs_closing(fs_info)) { @@ -2234,6 +2272,23 @@ int btrfs_balance(struct btrfs_balance_control *bctl, goto out; } + /* + * In case of mixed groups both data and meta should be picked, + * and identical options should be given for both of them. + */ + allowed = btrfs_super_incompat_flags(fs_info->super_copy); + if ((allowed & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) && + (bctl->flags & (BTRFS_BALANCE_DATA | BTRFS_BALANCE_METADATA))) { + if (!(bctl->flags & BTRFS_BALANCE_DATA) || + !(bctl->flags & BTRFS_BALANCE_METADATA) || + memcmp(&bctl->data, &bctl->meta, sizeof(bctl->data))) { + printk(KERN_ERR "btrfs: with mixed groups data and " + "metadata balance options must be the same\n"); + ret = -EINVAL; + goto out; + } + } + set_balance_control(bctl); mutex_unlock(&fs_info->balance_mutex); diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 882582385283..003e54216069 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -186,6 +186,17 @@ struct map_lookup { #define map_lookup_size(n) (sizeof(struct map_lookup) + \ (sizeof(struct btrfs_bio_stripe) * (n))) +/* + * Restriper's general type filter + */ +#define BTRFS_BALANCE_DATA (1ULL << 0) +#define BTRFS_BALANCE_SYSTEM (1ULL << 1) +#define BTRFS_BALANCE_METADATA (1ULL << 2) + +#define BTRFS_BALANCE_TYPE_MASK (BTRFS_BALANCE_DATA | \ + BTRFS_BALANCE_SYSTEM | \ + BTRFS_BALANCE_METADATA) + struct btrfs_balance_args; struct btrfs_balance_control { struct btrfs_fs_info *fs_info; From ed25e9b26f898d8d63ae4a836489f1923534143b Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 399/820] Btrfs: profiles filter Select chunks based on a given profile mask. Signed-off-by: Ilya Dryomov --- fs/btrfs/volumes.c | 24 ++++++++++++++++++++++++ fs/btrfs/volumes.h | 4 ++++ 2 files changed, 28 insertions(+) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 91bbf6e774c0..447bd422d867 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2102,6 +2102,24 @@ static void unset_balance_control(struct btrfs_fs_info *fs_info) kfree(bctl); } +/* + * Balance filters. Return 1 if chunk should be filtered out + * (should not be balanced). + */ +static int chunk_profiles_filter(u64 chunk_profile, + struct btrfs_balance_args *bargs) +{ + chunk_profile &= BTRFS_BLOCK_GROUP_PROFILE_MASK; + + if (chunk_profile == 0) + chunk_profile = BTRFS_AVAIL_ALLOC_BIT_SINGLE; + + if (bargs->profiles & chunk_profile) + return 0; + + return 1; +} + static int should_balance_chunk(struct btrfs_root *root, struct extent_buffer *leaf, struct btrfs_chunk *chunk, u64 chunk_offset) @@ -2123,6 +2141,12 @@ static int should_balance_chunk(struct btrfs_root *root, else if (chunk_type & BTRFS_BLOCK_GROUP_METADATA) bargs = &bctl->meta; + /* profiles filter */ + if ((bargs->flags & BTRFS_BALANCE_ARGS_PROFILES) && + chunk_profiles_filter(chunk_type, bargs)) { + return 0; + } + return 1; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 003e54216069..fb20d7740440 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -196,6 +196,10 @@ struct map_lookup { #define BTRFS_BALANCE_TYPE_MASK (BTRFS_BALANCE_DATA | \ BTRFS_BALANCE_SYSTEM | \ BTRFS_BALANCE_METADATA) +/* + * Balance filters + */ +#define BTRFS_BALANCE_ARGS_PROFILES (1ULL << 0) struct btrfs_balance_args; struct btrfs_balance_control { From 5ce5b3c0916ba3a2e34cf648b94044adc5ef9e76 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 400/820] Btrfs: usage filter Select chunks that are less than X percent full. Signed-off-by: Ilya Dryomov --- fs/btrfs/volumes.c | 36 ++++++++++++++++++++++++++++++++++++ fs/btrfs/volumes.h | 1 + 2 files changed, 37 insertions(+) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 447bd422d867..b858242374db 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2120,6 +2120,36 @@ static int chunk_profiles_filter(u64 chunk_profile, return 1; } +static u64 div_factor_fine(u64 num, int factor) +{ + if (factor <= 0) + return 0; + if (factor >= 100) + return num; + + num *= factor; + do_div(num, 100); + return num; +} + +static int chunk_usage_filter(struct btrfs_fs_info *fs_info, u64 chunk_offset, + struct btrfs_balance_args *bargs) +{ + struct btrfs_block_group_cache *cache; + u64 chunk_used, user_thresh; + int ret = 1; + + cache = btrfs_lookup_block_group(fs_info, chunk_offset); + chunk_used = btrfs_block_group_used(&cache->item); + + user_thresh = div_factor_fine(cache->key.offset, bargs->usage); + if (chunk_used < user_thresh) + ret = 0; + + btrfs_put_block_group(cache); + return ret; +} + static int should_balance_chunk(struct btrfs_root *root, struct extent_buffer *leaf, struct btrfs_chunk *chunk, u64 chunk_offset) @@ -2147,6 +2177,12 @@ static int should_balance_chunk(struct btrfs_root *root, return 0; } + /* usage filter */ + if ((bargs->flags & BTRFS_BALANCE_ARGS_USAGE) && + chunk_usage_filter(bctl->fs_info, chunk_offset, bargs)) { + return 0; + } + return 1; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index fb20d7740440..eee77fcf812a 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -200,6 +200,7 @@ struct map_lookup { * Balance filters */ #define BTRFS_BALANCE_ARGS_PROFILES (1ULL << 0) +#define BTRFS_BALANCE_ARGS_USAGE (1ULL << 1) struct btrfs_balance_args; struct btrfs_balance_control { From 409d404b461afa9738619f249fd7f62a366b68c2 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:47 +0200 Subject: [PATCH 401/820] Btrfs: devid filter Relocate chunks which have at least one stripe located on a device with devid X. Signed-off-by: Ilya Dryomov --- fs/btrfs/volumes.c | 23 +++++++++++++++++++++++ fs/btrfs/volumes.h | 1 + 2 files changed, 24 insertions(+) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index b858242374db..9ff5cd09a256 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2150,6 +2150,23 @@ static int chunk_usage_filter(struct btrfs_fs_info *fs_info, u64 chunk_offset, return ret; } +static int chunk_devid_filter(struct extent_buffer *leaf, + struct btrfs_chunk *chunk, + struct btrfs_balance_args *bargs) +{ + struct btrfs_stripe *stripe; + int num_stripes = btrfs_chunk_num_stripes(leaf, chunk); + int i; + + for (i = 0; i < num_stripes; i++) { + stripe = btrfs_stripe_nr(chunk, i); + if (btrfs_stripe_devid(leaf, stripe) == bargs->devid) + return 0; + } + + return 1; +} + static int should_balance_chunk(struct btrfs_root *root, struct extent_buffer *leaf, struct btrfs_chunk *chunk, u64 chunk_offset) @@ -2183,6 +2200,12 @@ static int should_balance_chunk(struct btrfs_root *root, return 0; } + /* devid filter */ + if ((bargs->flags & BTRFS_BALANCE_ARGS_DEVID) && + chunk_devid_filter(leaf, chunk, bargs)) { + return 0; + } + return 1; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index eee77fcf812a..7cfec03c29b8 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -201,6 +201,7 @@ struct map_lookup { */ #define BTRFS_BALANCE_ARGS_PROFILES (1ULL << 0) #define BTRFS_BALANCE_ARGS_USAGE (1ULL << 1) +#define BTRFS_BALANCE_ARGS_DEVID (1ULL << 2) struct btrfs_balance_args; struct btrfs_balance_control { From 94e60d5a5c4b98a32b1077dec88df09ada712376 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 402/820] Btrfs: devid subset filter Select chunks which have at least one byte of at least one stripe located on a device with devid X in a given [pstart,pend) physical address range. This filter only works when devid filter is turned on. Signed-off-by: Ilya Dryomov --- fs/btrfs/volumes.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/volumes.h | 1 + 2 files changed, 47 insertions(+) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 9ff5cd09a256..c60071a448a5 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2167,6 +2167,46 @@ static int chunk_devid_filter(struct extent_buffer *leaf, return 1; } +/* [pstart, pend) */ +static int chunk_drange_filter(struct extent_buffer *leaf, + struct btrfs_chunk *chunk, + u64 chunk_offset, + struct btrfs_balance_args *bargs) +{ + struct btrfs_stripe *stripe; + int num_stripes = btrfs_chunk_num_stripes(leaf, chunk); + u64 stripe_offset; + u64 stripe_length; + int factor; + int i; + + if (!(bargs->flags & BTRFS_BALANCE_ARGS_DEVID)) + return 0; + + if (btrfs_chunk_type(leaf, chunk) & (BTRFS_BLOCK_GROUP_DUP | + BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)) + factor = 2; + else + factor = 1; + factor = num_stripes / factor; + + for (i = 0; i < num_stripes; i++) { + stripe = btrfs_stripe_nr(chunk, i); + if (btrfs_stripe_devid(leaf, stripe) != bargs->devid) + continue; + + stripe_offset = btrfs_stripe_offset(leaf, stripe); + stripe_length = btrfs_chunk_length(leaf, chunk); + do_div(stripe_length, factor); + + if (stripe_offset < bargs->pend && + stripe_offset + stripe_length > bargs->pstart) + return 0; + } + + return 1; +} + static int should_balance_chunk(struct btrfs_root *root, struct extent_buffer *leaf, struct btrfs_chunk *chunk, u64 chunk_offset) @@ -2206,6 +2246,12 @@ static int should_balance_chunk(struct btrfs_root *root, return 0; } + /* drange filter, makes sense only with devid filter */ + if ((bargs->flags & BTRFS_BALANCE_ARGS_DRANGE) && + chunk_drange_filter(leaf, chunk, chunk_offset, bargs)) { + return 0; + } + return 1; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 7cfec03c29b8..844b08e388f2 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -202,6 +202,7 @@ struct map_lookup { #define BTRFS_BALANCE_ARGS_PROFILES (1ULL << 0) #define BTRFS_BALANCE_ARGS_USAGE (1ULL << 1) #define BTRFS_BALANCE_ARGS_DEVID (1ULL << 2) +#define BTRFS_BALANCE_ARGS_DRANGE (1ULL << 3) struct btrfs_balance_args; struct btrfs_balance_control { From ea67176ae8c024f64d85ec33873e5eadf1af7247 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 403/820] Btrfs: virtual address space subset filter Select chunks which have at least one byte located inside a given [vstart, vend) virtual address space range. Signed-off-by: Ilya Dryomov --- fs/btrfs/volumes.c | 20 ++++++++++++++++++++ fs/btrfs/volumes.h | 1 + 2 files changed, 21 insertions(+) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index c60071a448a5..e86c9e4fe51e 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2207,6 +2207,20 @@ static int chunk_drange_filter(struct extent_buffer *leaf, return 1; } +/* [vstart, vend) */ +static int chunk_vrange_filter(struct extent_buffer *leaf, + struct btrfs_chunk *chunk, + u64 chunk_offset, + struct btrfs_balance_args *bargs) +{ + if (chunk_offset < bargs->vend && + chunk_offset + btrfs_chunk_length(leaf, chunk) > bargs->vstart) + /* at least part of the chunk is inside this vrange */ + return 0; + + return 1; +} + static int should_balance_chunk(struct btrfs_root *root, struct extent_buffer *leaf, struct btrfs_chunk *chunk, u64 chunk_offset) @@ -2252,6 +2266,12 @@ static int should_balance_chunk(struct btrfs_root *root, return 0; } + /* vrange filter */ + if ((bargs->flags & BTRFS_BALANCE_ARGS_VRANGE) && + chunk_vrange_filter(leaf, chunk, chunk_offset, bargs)) { + return 0; + } + return 1; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 844b08e388f2..eac26c359312 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -203,6 +203,7 @@ struct map_lookup { #define BTRFS_BALANCE_ARGS_USAGE (1ULL << 1) #define BTRFS_BALANCE_ARGS_DEVID (1ULL << 2) #define BTRFS_BALANCE_ARGS_DRANGE (1ULL << 3) +#define BTRFS_BALANCE_ARGS_VRANGE (1ULL << 4) struct btrfs_balance_args; struct btrfs_balance_control { From 70922617b0099f420deceb53d5dc7f4fb30d08d0 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 404/820] Btrfs: do not reduce profile in do_chunk_alloc() Every caller of do_chunk_alloc() feeds it the reduced allocation profile, so stop trying to reduce it one more time. Instead check the validity of the passed profile. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 18 ++++++++++++++++++ fs/btrfs/extent-tree.c | 2 +- 2 files changed, 19 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index c4d98c8df5c5..1e7aea60da2b 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2536,6 +2536,24 @@ static inline void free_fs_info(struct btrfs_fs_info *fs_info) kfree(fs_info->super_for_commit); kfree(fs_info); } +/** + * profile_is_valid - tests whether a given profile is valid and reduced + * @flags: profile to validate + * @extended: if true @flags is treated as an extended profile + */ +static inline int profile_is_valid(u64 flags, int extended) +{ + u64 mask = ~BTRFS_BLOCK_GROUP_PROFILE_MASK; + + flags &= ~BTRFS_BLOCK_GROUP_TYPE_MASK; + if (extended) + mask &= ~BTRFS_AVAIL_ALLOC_BIT_SINGLE; + + if (flags & mask) + return 0; + /* true if zero or exactly one bit set */ + return (flags & (~flags + 1)) == flags; +} /* root-item.c */ int btrfs_find_root_ref(struct btrfs_root *tree_root, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 946b0671c5d9..a1a18ea7b6c6 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3295,7 +3295,7 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, int wait_for_alloc = 0; int ret = 0; - flags = btrfs_reduce_alloc_profile(extent_root, flags); + BUG_ON(!profile_is_valid(flags, 0)); space_info = __find_space_info(extent_root->fs_info, flags); if (!space_info) { From e4d8ec0f65b91bfb4984a4927632ded95f9825ad Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 405/820] Btrfs: implement online profile changing Profile changing is done by launching a balance with BTRFS_BALANCE_CONVERT bits set and target fields of respective btrfs_balance_args structs initialized. Profile reducing code in this case will pick restriper's target profile if it's available instead of doing a blind reduce. If target profile is not yet available it goes back to a plain reduce. Signed-off-by: Ilya Dryomov --- fs/btrfs/extent-tree.c | 56 +++++++++++++++++++++++++++++++++- fs/btrfs/volumes.c | 69 ++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/volumes.h | 5 +++ 3 files changed, 129 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a1a18ea7b6c6..e6a832e3e647 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3030,7 +3030,9 @@ static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) /* * @flags: available profiles in extended format (see ctree.h) * - * Returns reduced profile in chunk format. + * Returns reduced profile in chunk format. If profile changing is in + * progress (either running or paused) picks the target profile (if it's + * already available), otherwise falls back to plain reducing. */ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) { @@ -3042,6 +3044,34 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) u64 num_devices = root->fs_info->fs_devices->rw_devices + root->fs_info->fs_devices->missing_devices; + /* pick restriper's target profile if it's available */ + spin_lock(&root->fs_info->balance_lock); + if (root->fs_info->balance_ctl) { + struct btrfs_balance_control *bctl = root->fs_info->balance_ctl; + u64 tgt = 0; + + if ((flags & BTRFS_BLOCK_GROUP_DATA) && + (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (flags & bctl->data.target)) { + tgt = BTRFS_BLOCK_GROUP_DATA | bctl->data.target; + } else if ((flags & BTRFS_BLOCK_GROUP_SYSTEM) && + (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (flags & bctl->sys.target)) { + tgt = BTRFS_BLOCK_GROUP_SYSTEM | bctl->sys.target; + } else if ((flags & BTRFS_BLOCK_GROUP_METADATA) && + (bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (flags & bctl->meta.target)) { + tgt = BTRFS_BLOCK_GROUP_METADATA | bctl->meta.target; + } + + if (tgt) { + spin_unlock(&root->fs_info->balance_lock); + flags = tgt; + goto out; + } + } + spin_unlock(&root->fs_info->balance_lock); + if (num_devices == 1) flags &= ~(BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID0); if (num_devices < 4) @@ -3065,6 +3095,7 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) flags &= ~BTRFS_BLOCK_GROUP_RAID0; } +out: /* extended -> chunk profile */ flags &= ~BTRFS_AVAIL_ALLOC_BIT_SINGLE; return flags; @@ -6795,6 +6826,29 @@ static u64 update_block_group_flags(struct btrfs_root *root, u64 flags) u64 stripped = BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10; + if (root->fs_info->balance_ctl) { + struct btrfs_balance_control *bctl = root->fs_info->balance_ctl; + u64 tgt = 0; + + /* pick restriper's target profile and return */ + if (flags & BTRFS_BLOCK_GROUP_DATA && + bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) { + tgt = BTRFS_BLOCK_GROUP_DATA | bctl->data.target; + } else if (flags & BTRFS_BLOCK_GROUP_SYSTEM && + bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) { + tgt = BTRFS_BLOCK_GROUP_SYSTEM | bctl->sys.target; + } else if (flags & BTRFS_BLOCK_GROUP_METADATA && + bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) { + tgt = BTRFS_BLOCK_GROUP_METADATA | bctl->meta.target; + } + + if (tgt) { + /* extended -> chunk profile */ + tgt &= ~BTRFS_AVAIL_ALLOC_BIT_SINGLE; + return tgt; + } + } + /* * we add in the count of missing devices because we want * to make sure that any RAID levels on a degraded FS diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index e86c9e4fe51e..f08210e83339 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2438,6 +2438,75 @@ int btrfs_balance(struct btrfs_balance_control *bctl, } } + /* + * Profile changing sanity checks. Skip them if a simple + * balance is requested. + */ + if (!((bctl->data.flags | bctl->sys.flags | bctl->meta.flags) & + BTRFS_BALANCE_ARGS_CONVERT)) + goto do_balance; + + allowed = BTRFS_AVAIL_ALLOC_BIT_SINGLE; + if (fs_info->fs_devices->num_devices == 1) + allowed |= BTRFS_BLOCK_GROUP_DUP; + else if (fs_info->fs_devices->num_devices < 4) + allowed |= (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1); + else + allowed |= (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID10); + + if (!profile_is_valid(bctl->data.target, 1) || + bctl->data.target & ~allowed) { + printk(KERN_ERR "btrfs: unable to start balance with target " + "data profile %llu\n", + (unsigned long long)bctl->data.target); + ret = -EINVAL; + goto out; + } + if (!profile_is_valid(bctl->meta.target, 1) || + bctl->meta.target & ~allowed) { + printk(KERN_ERR "btrfs: unable to start balance with target " + "metadata profile %llu\n", + (unsigned long long)bctl->meta.target); + ret = -EINVAL; + goto out; + } + if (!profile_is_valid(bctl->sys.target, 1) || + bctl->sys.target & ~allowed) { + printk(KERN_ERR "btrfs: unable to start balance with target " + "system profile %llu\n", + (unsigned long long)bctl->sys.target); + ret = -EINVAL; + goto out; + } + + if (bctl->data.target & BTRFS_BLOCK_GROUP_DUP) { + printk(KERN_ERR "btrfs: dup for data is not allowed\n"); + ret = -EINVAL; + goto out; + } + + /* allow to reduce meta or sys integrity only if force set */ + allowed = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID10; + if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (fs_info->avail_system_alloc_bits & allowed) && + !(bctl->sys.target & allowed)) || + ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (fs_info->avail_metadata_alloc_bits & allowed) && + !(bctl->meta.target & allowed))) { + if (bctl->flags & BTRFS_BALANCE_FORCE) { + printk(KERN_INFO "btrfs: force reducing metadata " + "integrity\n"); + } else { + printk(KERN_ERR "btrfs: balance will reduce metadata " + "integrity, use force if you want this\n"); + ret = -EINVAL; + goto out; + } + } + +do_balance: set_balance_control(bctl); mutex_unlock(&fs_info->balance_mutex); diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index eac26c359312..79ee9c324562 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -196,6 +196,9 @@ struct map_lookup { #define BTRFS_BALANCE_TYPE_MASK (BTRFS_BALANCE_DATA | \ BTRFS_BALANCE_SYSTEM | \ BTRFS_BALANCE_METADATA) + +#define BTRFS_BALANCE_FORCE (1ULL << 3) + /* * Balance filters */ @@ -205,6 +208,8 @@ struct map_lookup { #define BTRFS_BALANCE_ARGS_DRANGE (1ULL << 3) #define BTRFS_BALANCE_ARGS_VRANGE (1ULL << 4) +#define BTRFS_BALANCE_ARGS_CONVERT (1ULL << 8) + struct btrfs_balance_args; struct btrfs_balance_control { struct btrfs_fs_info *fs_info; From cfa4c961cc69ffb7bda450972320a25cbd413e19 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 406/820] Btrfs: soft profile changing mode (aka soft convert) When doing convert from one profile to another if soft mode is on restriper won't touch chunks that already have the profile we are converting to. This is useful if e.g. half of the FS was converted earlier. The soft mode switch is (like every other filter) per-type. This means that we can convert for example meta chunks the "hard" way while converting data chunks selectively with soft switch. Signed-off-by: Ilya Dryomov --- fs/btrfs/volumes.c | 23 +++++++++++++++++++++++ fs/btrfs/volumes.h | 6 ++++++ 2 files changed, 29 insertions(+) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f08210e83339..98b4067017ff 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2221,6 +2221,23 @@ static int chunk_vrange_filter(struct extent_buffer *leaf, return 1; } +static int chunk_soft_convert_filter(u64 chunk_profile, + struct btrfs_balance_args *bargs) +{ + if (!(bargs->flags & BTRFS_BALANCE_ARGS_CONVERT)) + return 0; + + chunk_profile &= BTRFS_BLOCK_GROUP_PROFILE_MASK; + + if (chunk_profile == 0) + chunk_profile = BTRFS_AVAIL_ALLOC_BIT_SINGLE; + + if (bargs->target & chunk_profile) + return 1; + + return 0; +} + static int should_balance_chunk(struct btrfs_root *root, struct extent_buffer *leaf, struct btrfs_chunk *chunk, u64 chunk_offset) @@ -2272,6 +2289,12 @@ static int should_balance_chunk(struct btrfs_root *root, return 0; } + /* soft profile changing mode */ + if ((bargs->flags & BTRFS_BALANCE_ARGS_SOFT) && + chunk_soft_convert_filter(chunk_type, bargs)) { + return 0; + } + return 1; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 79ee9c324562..6c143c98017a 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -208,7 +208,13 @@ struct map_lookup { #define BTRFS_BALANCE_ARGS_DRANGE (1ULL << 3) #define BTRFS_BALANCE_ARGS_VRANGE (1ULL << 4) +/* + * Profile changing flags. When SOFT is set we won't relocate chunk if + * it already has the target profile (even though it may be + * half-filled). + */ #define BTRFS_BALANCE_ARGS_CONVERT (1ULL << 8) +#define BTRFS_BALANCE_ARGS_SOFT (1ULL << 9) struct btrfs_balance_args; struct btrfs_balance_control { From 0940ebf6b92ea10a6f30ae5ac3993a3b75745da6 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 407/820] Btrfs: save balance parameters to disk Introduce a new btree objectid for storing balance item. The reason is to be able to resume restriper after a crash with the same parameters. Balance item has a very high objectid and goes into tree of tree roots. The key for the new item is as follows: [ BTRFS_BALANCE_OBJECTID ; BTRFS_BALANCE_ITEM_KEY ; 0 ] Older kernels simply ignore it so it's safe to mount with an older kernel and then go back to the newer one. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 133 ++++++++++++++++++++++++++++++++++++++++++++- fs/btrfs/volumes.c | 99 +++++++++++++++++++++++++++++++++ 2 files changed, 231 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 1e7aea60da2b..9997a59e4f58 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -86,6 +86,9 @@ struct btrfs_ordered_sum; /* holds checksums of all the data extents */ #define BTRFS_CSUM_TREE_OBJECTID 7ULL +/* for storing balance parameters in the root tree */ +#define BTRFS_BALANCE_OBJECTID -4ULL + /* orhpan objectid for tracking unlinked/truncated files */ #define BTRFS_ORPHAN_OBJECTID -5ULL @@ -692,6 +695,54 @@ struct btrfs_root_ref { __le16 name_len; } __attribute__ ((__packed__)); +struct btrfs_disk_balance_args { + /* + * profiles to operate on, single is denoted by + * BTRFS_AVAIL_ALLOC_BIT_SINGLE + */ + __le64 profiles; + + /* usage filter */ + __le64 usage; + + /* devid filter */ + __le64 devid; + + /* devid subset filter [pstart..pend) */ + __le64 pstart; + __le64 pend; + + /* btrfs virtual address space subset filter [vstart..vend) */ + __le64 vstart; + __le64 vend; + + /* + * profile to convert to, single is denoted by + * BTRFS_AVAIL_ALLOC_BIT_SINGLE + */ + __le64 target; + + /* BTRFS_BALANCE_ARGS_* */ + __le64 flags; + + __le64 unused[8]; +} __attribute__ ((__packed__)); + +/* + * store balance parameters to disk so that balance can be properly + * resumed after crash or unmount + */ +struct btrfs_balance_item { + /* BTRFS_BALANCE_* */ + __le64 flags; + + struct btrfs_disk_balance_args data; + struct btrfs_disk_balance_args meta; + struct btrfs_disk_balance_args sys; + + __le64 unused[4]; +} __attribute__ ((__packed__)); + #define BTRFS_FILE_EXTENT_INLINE 0 #define BTRFS_FILE_EXTENT_REG 1 #define BTRFS_FILE_EXTENT_PREALLOC 2 @@ -1409,6 +1460,8 @@ struct btrfs_ioctl_defrag_range_args { #define BTRFS_DEV_ITEM_KEY 216 #define BTRFS_CHUNK_ITEM_KEY 228 +#define BTRFS_BALANCE_ITEM_KEY 248 + /* * string items are for debugging. They just store a short string of * data in the FS @@ -2103,8 +2156,86 @@ BTRFS_SETGET_STACK_FUNCS(backup_bytes_used, struct btrfs_root_backup, BTRFS_SETGET_STACK_FUNCS(backup_num_devices, struct btrfs_root_backup, num_devices, 64); -/* struct btrfs_super_block */ +/* struct btrfs_balance_item */ +BTRFS_SETGET_FUNCS(balance_flags, struct btrfs_balance_item, flags, 64); +static inline void btrfs_balance_data(struct extent_buffer *eb, + struct btrfs_balance_item *bi, + struct btrfs_disk_balance_args *ba) +{ + read_eb_member(eb, bi, struct btrfs_balance_item, data, ba); +} + +static inline void btrfs_set_balance_data(struct extent_buffer *eb, + struct btrfs_balance_item *bi, + struct btrfs_disk_balance_args *ba) +{ + write_eb_member(eb, bi, struct btrfs_balance_item, data, ba); +} + +static inline void btrfs_balance_meta(struct extent_buffer *eb, + struct btrfs_balance_item *bi, + struct btrfs_disk_balance_args *ba) +{ + read_eb_member(eb, bi, struct btrfs_balance_item, meta, ba); +} + +static inline void btrfs_set_balance_meta(struct extent_buffer *eb, + struct btrfs_balance_item *bi, + struct btrfs_disk_balance_args *ba) +{ + write_eb_member(eb, bi, struct btrfs_balance_item, meta, ba); +} + +static inline void btrfs_balance_sys(struct extent_buffer *eb, + struct btrfs_balance_item *bi, + struct btrfs_disk_balance_args *ba) +{ + read_eb_member(eb, bi, struct btrfs_balance_item, sys, ba); +} + +static inline void btrfs_set_balance_sys(struct extent_buffer *eb, + struct btrfs_balance_item *bi, + struct btrfs_disk_balance_args *ba) +{ + write_eb_member(eb, bi, struct btrfs_balance_item, sys, ba); +} + +static inline void +btrfs_disk_balance_args_to_cpu(struct btrfs_balance_args *cpu, + struct btrfs_disk_balance_args *disk) +{ + memset(cpu, 0, sizeof(*cpu)); + + cpu->profiles = le64_to_cpu(disk->profiles); + cpu->usage = le64_to_cpu(disk->usage); + cpu->devid = le64_to_cpu(disk->devid); + cpu->pstart = le64_to_cpu(disk->pstart); + cpu->pend = le64_to_cpu(disk->pend); + cpu->vstart = le64_to_cpu(disk->vstart); + cpu->vend = le64_to_cpu(disk->vend); + cpu->target = le64_to_cpu(disk->target); + cpu->flags = le64_to_cpu(disk->flags); +} + +static inline void +btrfs_cpu_balance_args_to_disk(struct btrfs_disk_balance_args *disk, + struct btrfs_balance_args *cpu) +{ + memset(disk, 0, sizeof(*disk)); + + disk->profiles = cpu_to_le64(cpu->profiles); + disk->usage = cpu_to_le64(cpu->usage); + disk->devid = cpu_to_le64(cpu->devid); + disk->pstart = cpu_to_le64(cpu->pstart); + disk->pend = cpu_to_le64(cpu->pend); + disk->vstart = cpu_to_le64(cpu->vstart); + disk->vend = cpu_to_le64(cpu->vend); + disk->target = cpu_to_le64(cpu->target); + disk->flags = cpu_to_le64(cpu->flags); +} + +/* struct btrfs_super_block */ BTRFS_SETGET_STACK_FUNCS(super_bytenr, struct btrfs_super_block, bytenr, 64); BTRFS_SETGET_STACK_FUNCS(super_flags, struct btrfs_super_block, flags, 64); BTRFS_SETGET_STACK_FUNCS(super_generation, struct btrfs_super_block, diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 98b4067017ff..4c60ca0ae90b 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2073,6 +2073,97 @@ error: return ret; } +static int insert_balance_item(struct btrfs_root *root, + struct btrfs_balance_control *bctl) +{ + struct btrfs_trans_handle *trans; + struct btrfs_balance_item *item; + struct btrfs_disk_balance_args disk_bargs; + struct btrfs_path *path; + struct extent_buffer *leaf; + struct btrfs_key key; + int ret, err; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + btrfs_free_path(path); + return PTR_ERR(trans); + } + + key.objectid = BTRFS_BALANCE_OBJECTID; + key.type = BTRFS_BALANCE_ITEM_KEY; + key.offset = 0; + + ret = btrfs_insert_empty_item(trans, root, path, &key, + sizeof(*item)); + if (ret) + goto out; + + leaf = path->nodes[0]; + item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item); + + memset_extent_buffer(leaf, 0, (unsigned long)item, sizeof(*item)); + + btrfs_cpu_balance_args_to_disk(&disk_bargs, &bctl->data); + btrfs_set_balance_data(leaf, item, &disk_bargs); + btrfs_cpu_balance_args_to_disk(&disk_bargs, &bctl->meta); + btrfs_set_balance_meta(leaf, item, &disk_bargs); + btrfs_cpu_balance_args_to_disk(&disk_bargs, &bctl->sys); + btrfs_set_balance_sys(leaf, item, &disk_bargs); + + btrfs_set_balance_flags(leaf, item, bctl->flags); + + btrfs_mark_buffer_dirty(leaf); +out: + btrfs_free_path(path); + err = btrfs_commit_transaction(trans, root); + if (err && !ret) + ret = err; + return ret; +} + +static int del_balance_item(struct btrfs_root *root) +{ + struct btrfs_trans_handle *trans; + struct btrfs_path *path; + struct btrfs_key key; + int ret, err; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + btrfs_free_path(path); + return PTR_ERR(trans); + } + + key.objectid = BTRFS_BALANCE_OBJECTID; + key.type = BTRFS_BALANCE_ITEM_KEY; + key.offset = 0; + + ret = btrfs_search_slot(trans, root, &key, path, -1, 1); + if (ret < 0) + goto out; + if (ret > 0) { + ret = -ENOENT; + goto out; + } + + ret = btrfs_del_item(trans, root, path); +out: + btrfs_free_path(path); + err = btrfs_commit_transaction(trans, root); + if (err && !ret) + ret = err; + return ret; +} + /* * Should be called with both balance and volume mutexes held to * serialize other volume operations (add_dev/rm_dev/resize) with @@ -2423,7 +2514,11 @@ error: static void __cancel_balance(struct btrfs_fs_info *fs_info) { + int ret; + unset_balance_control(fs_info); + ret = del_balance_item(fs_info->tree_root); + BUG_ON(ret); } void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, @@ -2530,6 +2625,10 @@ int btrfs_balance(struct btrfs_balance_control *bctl, } do_balance: + ret = insert_balance_item(fs_info->tree_root, bctl); + if (ret) + goto out; + set_balance_control(bctl); mutex_unlock(&fs_info->balance_mutex); From 596410151ed71819b9e8a8018c6c9992796b256d Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 408/820] Btrfs: recover balance on mount On mount, if balance item is found, resume balance in a separate kernel thread. Try to be smart to continue roughly where previous balance (or convert) was interrupted. For chunk types that were being converted to some profile we turn on soft convert, in case of a simple balance we turn on usage filter and relocate only less-than-90%-full chunks of that type. These are just heuristics but they help quite a bit, and can be improved in future. Signed-off-by: Ilya Dryomov --- fs/btrfs/disk-io.c | 4 ++ fs/btrfs/volumes.c | 135 ++++++++++++++++++++++++++++++++++++++++++++- fs/btrfs/volumes.h | 2 + 3 files changed, 139 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 190a1b24c902..eb7a11ac5b73 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2427,6 +2427,10 @@ retry_root_backup: if (!err) err = btrfs_orphan_cleanup(fs_info->tree_root); up_read(&fs_info->cleanup_work_sem); + + if (!err) + err = btrfs_recover_balance(fs_info->tree_root); + if (err) { close_ctree(tree_root); return ERR_PTR(err); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 4c60ca0ae90b..17e565388de0 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include "compat.h" #include "ctree.h" @@ -2164,6 +2165,46 @@ out: return ret; } +/* + * This is a heuristic used to reduce the number of chunks balanced on + * resume after balance was interrupted. + */ +static void update_balance_args(struct btrfs_balance_control *bctl) +{ + /* + * Turn on soft mode for chunk types that were being converted. + */ + if (bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) + bctl->data.flags |= BTRFS_BALANCE_ARGS_SOFT; + if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) + bctl->sys.flags |= BTRFS_BALANCE_ARGS_SOFT; + if (bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) + bctl->meta.flags |= BTRFS_BALANCE_ARGS_SOFT; + + /* + * Turn on usage filter if is not already used. The idea is + * that chunks that we have already balanced should be + * reasonably full. Don't do it for chunks that are being + * converted - that will keep us from relocating unconverted + * (albeit full) chunks. + */ + if (!(bctl->data.flags & BTRFS_BALANCE_ARGS_USAGE) && + !(bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT)) { + bctl->data.flags |= BTRFS_BALANCE_ARGS_USAGE; + bctl->data.usage = 90; + } + if (!(bctl->sys.flags & BTRFS_BALANCE_ARGS_USAGE) && + !(bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT)) { + bctl->sys.flags |= BTRFS_BALANCE_ARGS_USAGE; + bctl->sys.usage = 90; + } + if (!(bctl->meta.flags & BTRFS_BALANCE_ARGS_USAGE) && + !(bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT)) { + bctl->meta.flags |= BTRFS_BALANCE_ARGS_USAGE; + bctl->meta.usage = 90; + } +} + /* * Should be called with both balance and volume mutexes held to * serialize other volume operations (add_dev/rm_dev/resize) with @@ -2626,10 +2667,18 @@ int btrfs_balance(struct btrfs_balance_control *bctl, do_balance: ret = insert_balance_item(fs_info->tree_root, bctl); - if (ret) + if (ret && ret != -EEXIST) goto out; - set_balance_control(bctl); + if (!(bctl->flags & BTRFS_BALANCE_RESUME)) { + BUG_ON(ret == -EEXIST); + set_balance_control(bctl); + } else { + BUG_ON(ret != -EEXIST); + spin_lock(&fs_info->balance_lock); + update_balance_args(bctl); + spin_unlock(&fs_info->balance_lock); + } mutex_unlock(&fs_info->balance_mutex); @@ -2646,7 +2695,89 @@ do_balance: return ret; out: + if (bctl->flags & BTRFS_BALANCE_RESUME) + __cancel_balance(fs_info); + else + kfree(bctl); + return ret; +} + +static int balance_kthread(void *data) +{ + struct btrfs_balance_control *bctl = + (struct btrfs_balance_control *)data; + struct btrfs_fs_info *fs_info = bctl->fs_info; + int ret; + + mutex_lock(&fs_info->volume_mutex); + mutex_lock(&fs_info->balance_mutex); + + set_balance_control(bctl); + + printk(KERN_INFO "btrfs: continuing balance\n"); + ret = btrfs_balance(bctl, NULL); + + mutex_unlock(&fs_info->balance_mutex); + mutex_unlock(&fs_info->volume_mutex); + return ret; +} + +int btrfs_recover_balance(struct btrfs_root *tree_root) +{ + struct task_struct *tsk; + struct btrfs_balance_control *bctl; + struct btrfs_balance_item *item; + struct btrfs_disk_balance_args disk_bargs; + struct btrfs_path *path; + struct extent_buffer *leaf; + struct btrfs_key key; + int ret; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + bctl = kzalloc(sizeof(*bctl), GFP_NOFS); + if (!bctl) { + ret = -ENOMEM; + goto out; + } + + key.objectid = BTRFS_BALANCE_OBJECTID; + key.type = BTRFS_BALANCE_ITEM_KEY; + key.offset = 0; + + ret = btrfs_search_slot(NULL, tree_root, &key, path, 0, 0); + if (ret < 0) + goto out_bctl; + if (ret > 0) { /* ret = -ENOENT; */ + ret = 0; + goto out_bctl; + } + + leaf = path->nodes[0]; + item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item); + + bctl->fs_info = tree_root->fs_info; + bctl->flags = btrfs_balance_flags(leaf, item) | BTRFS_BALANCE_RESUME; + + btrfs_balance_data(leaf, item, &disk_bargs); + btrfs_disk_balance_args_to_cpu(&bctl->data, &disk_bargs); + btrfs_balance_meta(leaf, item, &disk_bargs); + btrfs_disk_balance_args_to_cpu(&bctl->meta, &disk_bargs); + btrfs_balance_sys(leaf, item, &disk_bargs); + btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs); + + tsk = kthread_run(balance_kthread, bctl, "btrfs-balance"); + if (IS_ERR(tsk)) + ret = PTR_ERR(tsk); + else + goto out; + +out_bctl: kfree(bctl); +out: + btrfs_free_path(path); return ret; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 6c143c98017a..cd25ea58ec35 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -198,6 +198,7 @@ struct map_lookup { BTRFS_BALANCE_METADATA) #define BTRFS_BALANCE_FORCE (1ULL << 3) +#define BTRFS_BALANCE_RESUME (1ULL << 4) /* * Balance filters @@ -271,6 +272,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); +int btrfs_recover_balance(struct btrfs_root *tree_root); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); int find_free_dev_extent(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 num_bytes, From 9555c6c180600b40f6e86bd4dc53bf47e06ed663 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:48 +0200 Subject: [PATCH 409/820] Btrfs: add skip_balance mount option Since restriper kthread starts involuntarily on mount and can suck cpu and memory bandwidth add a mount option to forcefully skip it. The restriper in that case hangs around in paused state and can be resumed from userspace when it's convenient. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 1 + fs/btrfs/super.c | 11 +++++++++-- fs/btrfs/volumes.c | 10 +++++++--- 3 files changed, 17 insertions(+), 5 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 9997a59e4f58..99eb2bcd9aa7 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1492,6 +1492,7 @@ struct btrfs_ioctl_defrag_range_args { #define BTRFS_MOUNT_AUTO_DEFRAG (1 << 16) #define BTRFS_MOUNT_INODE_MAP_CACHE (1 << 17) #define BTRFS_MOUNT_RECOVERY (1 << 18) +#define BTRFS_MOUNT_SKIP_BALANCE (1 << 19) #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 34a8b6112ea4..063b521e3ded 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -164,8 +164,9 @@ enum { Opt_compress_type, Opt_compress_force, Opt_compress_force_type, Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, - Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, - Opt_inode_cache, Opt_no_space_cache, Opt_recovery, Opt_err, + Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_inode_cache, + Opt_no_space_cache, Opt_recovery, Opt_skip_balance, + Opt_err, }; static match_table_t tokens = { @@ -200,6 +201,7 @@ static match_table_t tokens = { {Opt_inode_cache, "inode_cache"}, {Opt_no_space_cache, "nospace_cache"}, {Opt_recovery, "recovery"}, + {Opt_skip_balance, "skip_balance"}, {Opt_err, NULL}, }; @@ -398,6 +400,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) printk(KERN_INFO "btrfs: enabling auto recovery"); btrfs_set_opt(info->mount_opt, RECOVERY); break; + case Opt_skip_balance: + btrfs_set_opt(info->mount_opt, SKIP_BALANCE); + break; case Opt_err: printk(KERN_INFO "btrfs: unrecognized mount option " "'%s'\n", p); @@ -723,6 +728,8 @@ static int btrfs_show_options(struct seq_file *seq, struct vfsmount *vfs) seq_puts(seq, ",autodefrag"); if (btrfs_test_opt(root, INODE_MAP_CACHE)) seq_puts(seq, ",inode_cache"); + if (btrfs_test_opt(root, SKIP_BALANCE)) + seq_puts(seq, ",skip_balance"); return 0; } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 17e565388de0..e0160607e6e2 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2707,15 +2707,19 @@ static int balance_kthread(void *data) struct btrfs_balance_control *bctl = (struct btrfs_balance_control *)data; struct btrfs_fs_info *fs_info = bctl->fs_info; - int ret; + int ret = 0; mutex_lock(&fs_info->volume_mutex); mutex_lock(&fs_info->balance_mutex); set_balance_control(bctl); - printk(KERN_INFO "btrfs: continuing balance\n"); - ret = btrfs_balance(bctl, NULL); + if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) { + printk(KERN_INFO "btrfs: force skipping balance\n"); + } else { + printk(KERN_INFO "btrfs: continuing balance\n"); + ret = btrfs_balance(bctl, NULL); + } mutex_unlock(&fs_info->balance_mutex); mutex_unlock(&fs_info->volume_mutex); From 837d5b6e46d1a4af5b6cc8f2fe83cb5de79a2961 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:49 +0200 Subject: [PATCH 410/820] Btrfs: allow for pausing restriper Implement an ioctl for pausing restriper. This pauses the relocation, but balance is still considered to be "in progress": balance item is not deleted, other volume operations cannot be started, etc. If paused in the middle of profile changing operation we will continue making allocations with the target profile. Add a hook to close_ctree() to pause restriper and free its data structures on unmount. (It's safe to unmount when restriper is in "paused" state, we will resume with the same parameters on the next mount) Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 4 ++++ fs/btrfs/disk-io.c | 6 ++++++ fs/btrfs/ioctl.c | 28 ++++++++++++++++++++++++- fs/btrfs/ioctl.h | 7 +++++++ fs/btrfs/volumes.c | 51 ++++++++++++++++++++++++++++++++++++++++++++-- fs/btrfs/volumes.h | 1 + 6 files changed, 94 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 99eb2bcd9aa7..1afda75d5414 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1214,7 +1214,10 @@ struct btrfs_fs_info { /* restriper state */ spinlock_t balance_lock; struct mutex balance_mutex; + atomic_t balance_running; + atomic_t balance_pause_req; struct btrfs_balance_control *balance_ctl; + wait_queue_head_t balance_wait_q; unsigned data_chunk_allocations; unsigned metadata_ratio; @@ -2658,6 +2661,7 @@ static inline int btrfs_fs_closing(struct btrfs_fs_info *fs_info) } static inline void free_fs_info(struct btrfs_fs_info *fs_info) { + kfree(fs_info->balance_ctl); kfree(fs_info->delayed_root); kfree(fs_info->extent_root); kfree(fs_info->tree_root); diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index eb7a11ac5b73..8ce837407800 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2004,7 +2004,10 @@ struct btrfs_root *open_ctree(struct super_block *sb, spin_lock_init(&fs_info->balance_lock); mutex_init(&fs_info->balance_mutex); + atomic_set(&fs_info->balance_running, 0); + atomic_set(&fs_info->balance_pause_req, 0); fs_info->balance_ctl = NULL; + init_waitqueue_head(&fs_info->balance_wait_q); sb->s_blocksize = 4096; sb->s_blocksize_bits = blksize_bits(4096); @@ -2980,6 +2983,9 @@ int close_ctree(struct btrfs_root *root) fs_info->closing = 1; smp_mb(); + /* pause restriper - we want to resume on mount */ + btrfs_pause_balance(root->fs_info); + btrfs_scrub_cancel(root); /* wait for any defraggers to finish */ diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 29b3a94933f0..f572c53dda4f 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3072,6 +3072,11 @@ void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, bargs->flags = bctl->flags; + if (atomic_read(&fs_info->balance_running)) + bargs->state |= BTRFS_BALANCE_STATE_RUNNING; + if (atomic_read(&fs_info->balance_pause_req)) + bargs->state |= BTRFS_BALANCE_STATE_PAUSE_REQ; + memcpy(&bargs->data, &bctl->data, sizeof(bargs->data)); memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta)); memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys)); @@ -3103,6 +3108,11 @@ static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) bargs = NULL; } + if (fs_info->balance_ctl) { + ret = -EINPROGRESS; + goto out_bargs; + } + bctl = kzalloc(sizeof(*bctl), GFP_NOFS); if (!bctl) { ret = -ENOMEM; @@ -3123,7 +3133,8 @@ static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) ret = btrfs_balance(bctl, bargs); /* - * bctl is freed in __cancel_balance + * bctl is freed in __cancel_balance or in free_fs_info if + * restriper was paused all the way until unmount */ if (arg) { if (copy_to_user(arg, bargs, sizeof(*bargs))) @@ -3138,6 +3149,19 @@ out: return ret; } +static long btrfs_ioctl_balance_ctl(struct btrfs_root *root, int cmd) +{ + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + switch (cmd) { + case BTRFS_BALANCE_CTL_PAUSE: + return btrfs_pause_balance(root->fs_info); + } + + return -EINVAL; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -3216,6 +3240,8 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_scrub_progress(root, argp); case BTRFS_IOC_BALANCE_V2: return btrfs_ioctl_balance(root, argp); + case BTRFS_IOC_BALANCE_CTL: + return btrfs_ioctl_balance_ctl(root, arg); } return -ENOTTY; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index c8b37d2c0d77..e972e11a8d77 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -109,6 +109,9 @@ struct btrfs_ioctl_fs_info_args { __u64 reserved[124]; /* pad to 1k */ }; +/* balance control ioctl modes */ +#define BTRFS_BALANCE_CTL_PAUSE 1 + /* * this is packed, because it should be exactly the same as its disk * byte order counterpart (struct btrfs_disk_balance_args) @@ -137,6 +140,9 @@ struct btrfs_balance_progress { __u64 completed; /* # of chunks relocated so far */ }; +#define BTRFS_BALANCE_STATE_RUNNING (1ULL << 0) +#define BTRFS_BALANCE_STATE_PAUSE_REQ (1ULL << 1) + struct btrfs_ioctl_balance_args { __u64 flags; /* in/out */ __u64 state; /* out */ @@ -315,6 +321,7 @@ struct btrfs_ioctl_logical_ino_args { struct btrfs_ioctl_fs_info_args) #define BTRFS_IOC_BALANCE_V2 _IOWR(BTRFS_IOCTL_MAGIC, 32, \ struct btrfs_ioctl_balance_args) +#define BTRFS_IOC_BALANCE_CTL _IOW(BTRFS_IOCTL_MAGIC, 33, int) #define BTRFS_IOC_INO_PATHS _IOWR(BTRFS_IOCTL_MAGIC, 35, \ struct btrfs_ioctl_ino_path_args) #define BTRFS_IOC_LOGICAL_INO _IOWR(BTRFS_IOCTL_MAGIC, 36, \ diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index e0160607e6e2..d32660ce753d 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2492,6 +2492,11 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) key.type = BTRFS_CHUNK_ITEM_KEY; while (1) { + if (atomic_read(&fs_info->balance_pause_req)) { + ret = -ECANCELED; + goto error; + } + ret = btrfs_search_slot(NULL, chunk_root, &key, path, 0, 0); if (ret < 0) goto error; @@ -2553,6 +2558,11 @@ error: return ret; } +static inline int balance_need_close(struct btrfs_fs_info *fs_info) +{ + return atomic_read(&fs_info->balance_pause_req) == 0; +} + static void __cancel_balance(struct btrfs_fs_info *fs_info) { int ret; @@ -2575,7 +2585,8 @@ int btrfs_balance(struct btrfs_balance_control *bctl, u64 allowed; int ret; - if (btrfs_fs_closing(fs_info)) { + if (btrfs_fs_closing(fs_info) || + atomic_read(&fs_info->balance_pause_req)) { ret = -EINVAL; goto out; } @@ -2680,18 +2691,25 @@ do_balance: spin_unlock(&fs_info->balance_lock); } + atomic_inc(&fs_info->balance_running); mutex_unlock(&fs_info->balance_mutex); ret = __btrfs_balance(fs_info); mutex_lock(&fs_info->balance_mutex); + atomic_dec(&fs_info->balance_running); if (bargs) { memset(bargs, 0, sizeof(*bargs)); update_ioctl_balance_args(fs_info, bargs); } - __cancel_balance(fs_info); + if ((ret && ret != -ECANCELED && ret != -ENOSPC) || + balance_need_close(fs_info)) { + __cancel_balance(fs_info); + } + + wake_up(&fs_info->balance_wait_q); return ret; out: @@ -2785,6 +2803,35 @@ out: return ret; } +int btrfs_pause_balance(struct btrfs_fs_info *fs_info) +{ + int ret = 0; + + mutex_lock(&fs_info->balance_mutex); + if (!fs_info->balance_ctl) { + mutex_unlock(&fs_info->balance_mutex); + return -ENOTCONN; + } + + if (atomic_read(&fs_info->balance_running)) { + atomic_inc(&fs_info->balance_pause_req); + mutex_unlock(&fs_info->balance_mutex); + + wait_event(fs_info->balance_wait_q, + atomic_read(&fs_info->balance_running) == 0); + + mutex_lock(&fs_info->balance_mutex); + /* we are good with balance_ctl ripped off from under us */ + BUG_ON(atomic_read(&fs_info->balance_running)); + atomic_dec(&fs_info->balance_pause_req); + } else { + ret = -ENOTCONN; + } + + mutex_unlock(&fs_info->balance_mutex); + return ret; +} + /* * shrinking a device means finding all of the device extents past * the new size, and then following the back refs to the chunks. diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index cd25ea58ec35..80953afb12b9 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -273,6 +273,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); int btrfs_recover_balance(struct btrfs_root *tree_root); +int btrfs_pause_balance(struct btrfs_fs_info *fs_info); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); int find_free_dev_extent(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 num_bytes, From a7e99c691af553fc15ac46a51f130b7c59a65f76 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:49 +0200 Subject: [PATCH 411/820] Btrfs: allow for canceling restriper Implement an ioctl for canceling restriper. Currently we wait until relocation of the current block group is finished, in future this can be done by triggering a commit. Balance item is deleted and no memory about the interrupted balance is kept. Signed-off-by: Ilya Dryomov --- fs/btrfs/ctree.h | 1 + fs/btrfs/disk-io.c | 1 + fs/btrfs/ioctl.c | 4 ++++ fs/btrfs/ioctl.h | 2 ++ fs/btrfs/volumes.c | 47 +++++++++++++++++++++++++++++++++++++++++++--- fs/btrfs/volumes.h | 1 + 6 files changed, 53 insertions(+), 3 deletions(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 1afda75d5414..dfc136cc07d7 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1216,6 +1216,7 @@ struct btrfs_fs_info { struct mutex balance_mutex; atomic_t balance_running; atomic_t balance_pause_req; + atomic_t balance_cancel_req; struct btrfs_balance_control *balance_ctl; wait_queue_head_t balance_wait_q; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 8ce837407800..c23b82d8ec08 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2006,6 +2006,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, mutex_init(&fs_info->balance_mutex); atomic_set(&fs_info->balance_running, 0); atomic_set(&fs_info->balance_pause_req, 0); + atomic_set(&fs_info->balance_cancel_req, 0); fs_info->balance_ctl = NULL; init_waitqueue_head(&fs_info->balance_wait_q); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index f572c53dda4f..60852217ce9a 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3076,6 +3076,8 @@ void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, bargs->state |= BTRFS_BALANCE_STATE_RUNNING; if (atomic_read(&fs_info->balance_pause_req)) bargs->state |= BTRFS_BALANCE_STATE_PAUSE_REQ; + if (atomic_read(&fs_info->balance_cancel_req)) + bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ; memcpy(&bargs->data, &bctl->data, sizeof(bargs->data)); memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta)); @@ -3157,6 +3159,8 @@ static long btrfs_ioctl_balance_ctl(struct btrfs_root *root, int cmd) switch (cmd) { case BTRFS_BALANCE_CTL_PAUSE: return btrfs_pause_balance(root->fs_info); + case BTRFS_BALANCE_CTL_CANCEL: + return btrfs_cancel_balance(root->fs_info); } return -EINVAL; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index e972e11a8d77..cd19d10794b9 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -111,6 +111,7 @@ struct btrfs_ioctl_fs_info_args { /* balance control ioctl modes */ #define BTRFS_BALANCE_CTL_PAUSE 1 +#define BTRFS_BALANCE_CTL_CANCEL 2 /* * this is packed, because it should be exactly the same as its disk @@ -142,6 +143,7 @@ struct btrfs_balance_progress { #define BTRFS_BALANCE_STATE_RUNNING (1ULL << 0) #define BTRFS_BALANCE_STATE_PAUSE_REQ (1ULL << 1) +#define BTRFS_BALANCE_STATE_CANCEL_REQ (1ULL << 2) struct btrfs_ioctl_balance_args { __u64 flags; /* in/out */ diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d32660ce753d..c32667318ae4 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2492,7 +2492,8 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) key.type = BTRFS_CHUNK_ITEM_KEY; while (1) { - if (atomic_read(&fs_info->balance_pause_req)) { + if (atomic_read(&fs_info->balance_pause_req) || + atomic_read(&fs_info->balance_cancel_req)) { ret = -ECANCELED; goto error; } @@ -2560,7 +2561,10 @@ error: static inline int balance_need_close(struct btrfs_fs_info *fs_info) { - return atomic_read(&fs_info->balance_pause_req) == 0; + /* cancel requested || normal exit path */ + return atomic_read(&fs_info->balance_cancel_req) || + (atomic_read(&fs_info->balance_pause_req) == 0 && + atomic_read(&fs_info->balance_cancel_req) == 0); } static void __cancel_balance(struct btrfs_fs_info *fs_info) @@ -2586,7 +2590,8 @@ int btrfs_balance(struct btrfs_balance_control *bctl, int ret; if (btrfs_fs_closing(fs_info) || - atomic_read(&fs_info->balance_pause_req)) { + atomic_read(&fs_info->balance_pause_req) || + atomic_read(&fs_info->balance_cancel_req)) { ret = -EINVAL; goto out; } @@ -2832,6 +2837,42 @@ int btrfs_pause_balance(struct btrfs_fs_info *fs_info) return ret; } +int btrfs_cancel_balance(struct btrfs_fs_info *fs_info) +{ + mutex_lock(&fs_info->balance_mutex); + if (!fs_info->balance_ctl) { + mutex_unlock(&fs_info->balance_mutex); + return -ENOTCONN; + } + + atomic_inc(&fs_info->balance_cancel_req); + /* + * if we are running just wait and return, balance item is + * deleted in btrfs_balance in this case + */ + if (atomic_read(&fs_info->balance_running)) { + mutex_unlock(&fs_info->balance_mutex); + wait_event(fs_info->balance_wait_q, + atomic_read(&fs_info->balance_running) == 0); + mutex_lock(&fs_info->balance_mutex); + } else { + /* __cancel_balance needs volume_mutex */ + mutex_unlock(&fs_info->balance_mutex); + mutex_lock(&fs_info->volume_mutex); + mutex_lock(&fs_info->balance_mutex); + + if (fs_info->balance_ctl) + __cancel_balance(fs_info); + + mutex_unlock(&fs_info->volume_mutex); + } + + BUG_ON(fs_info->balance_ctl || atomic_read(&fs_info->balance_running)); + atomic_dec(&fs_info->balance_cancel_req); + mutex_unlock(&fs_info->balance_mutex); + return 0; +} + /* * shrinking a device means finding all of the device extents past * the new size, and then following the back refs to the chunks. diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 80953afb12b9..caa9abd218ef 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -274,6 +274,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl, struct btrfs_ioctl_balance_args *bargs); int btrfs_recover_balance(struct btrfs_root *tree_root); int btrfs_pause_balance(struct btrfs_fs_info *fs_info); +int btrfs_cancel_balance(struct btrfs_fs_info *fs_info); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); int find_free_dev_extent(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 num_bytes, From de322263d3a6d4ffd4ed7c4d0c6536e9497aec9b Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:49 +0200 Subject: [PATCH 412/820] Btrfs: allow for resuming restriper after it was paused Recognize BTRFS_BALANCE_RESUME flag passed from userspace. We use the same heuristics used when recovering balance after a crash to try to start where we left off last time. Signed-off-by: Ilya Dryomov --- fs/btrfs/ioctl.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 60852217ce9a..85e546ffe3c7 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3106,6 +3106,20 @@ static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) ret = PTR_ERR(bargs); goto out; } + + if (bargs->flags & BTRFS_BALANCE_RESUME) { + if (!fs_info->balance_ctl) { + ret = -ENOTCONN; + goto out_bargs; + } + + bctl = fs_info->balance_ctl; + spin_lock(&fs_info->balance_lock); + bctl->flags |= BTRFS_BALANCE_RESUME; + spin_unlock(&fs_info->balance_lock); + + goto do_balance; + } } else { bargs = NULL; } @@ -3133,6 +3147,7 @@ static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) bctl->flags |= BTRFS_BALANCE_TYPE_MASK; } +do_balance: ret = btrfs_balance(bctl, bargs); /* * bctl is freed in __cancel_balance or in free_fs_info if From 19a39dce3b9bf0244d19a446718ad6f7605ff099 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 16 Jan 2012 22:04:49 +0200 Subject: [PATCH 413/820] Btrfs: add balance progress reporting Signed-off-by: Ilya Dryomov --- fs/btrfs/ioctl.c | 45 ++++++++++++++++++++++++++++++++++++++++++++- fs/btrfs/ioctl.h | 2 ++ fs/btrfs/volumes.c | 39 +++++++++++++++++++++++++++++++++++---- fs/btrfs/volumes.h | 3 +++ 4 files changed, 84 insertions(+), 5 deletions(-) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 85e546ffe3c7..1e7a9bac31ab 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3065,7 +3065,7 @@ out: return ret; } -void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, +void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock, struct btrfs_ioctl_balance_args *bargs) { struct btrfs_balance_control *bctl = fs_info->balance_ctl; @@ -3082,6 +3082,14 @@ void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, memcpy(&bargs->data, &bctl->data, sizeof(bargs->data)); memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta)); memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys)); + + if (lock) { + spin_lock(&fs_info->balance_lock); + memcpy(&bargs->stat, &bctl->stat, sizeof(bargs->stat)); + spin_unlock(&fs_info->balance_lock); + } else { + memcpy(&bargs->stat, &bctl->stat, sizeof(bargs->stat)); + } } static long btrfs_ioctl_balance(struct btrfs_root *root, void __user *arg) @@ -3181,6 +3189,39 @@ static long btrfs_ioctl_balance_ctl(struct btrfs_root *root, int cmd) return -EINVAL; } +static long btrfs_ioctl_balance_progress(struct btrfs_root *root, + void __user *arg) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + struct btrfs_ioctl_balance_args *bargs; + int ret = 0; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + mutex_lock(&fs_info->balance_mutex); + if (!fs_info->balance_ctl) { + ret = -ENOTCONN; + goto out; + } + + bargs = kzalloc(sizeof(*bargs), GFP_NOFS); + if (!bargs) { + ret = -ENOMEM; + goto out; + } + + update_ioctl_balance_args(fs_info, 1, bargs); + + if (copy_to_user(arg, bargs, sizeof(*bargs))) + ret = -EFAULT; + + kfree(bargs); +out: + mutex_unlock(&fs_info->balance_mutex); + return ret; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -3261,6 +3302,8 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_balance(root, argp); case BTRFS_IOC_BALANCE_CTL: return btrfs_ioctl_balance_ctl(root, arg); + case BTRFS_IOC_BALANCE_PROGRESS: + return btrfs_ioctl_balance_progress(root, argp); } return -ENOTTY; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index cd19d10794b9..4f69028a68c4 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -324,6 +324,8 @@ struct btrfs_ioctl_logical_ino_args { #define BTRFS_IOC_BALANCE_V2 _IOWR(BTRFS_IOCTL_MAGIC, 32, \ struct btrfs_ioctl_balance_args) #define BTRFS_IOC_BALANCE_CTL _IOW(BTRFS_IOCTL_MAGIC, 33, int) +#define BTRFS_IOC_BALANCE_PROGRESS _IOR(BTRFS_IOCTL_MAGIC, 34, \ + struct btrfs_ioctl_balance_args) #define BTRFS_IOC_INO_PATHS _IOWR(BTRFS_IOCTL_MAGIC, 35, \ struct btrfs_ioctl_ino_path_args) #define BTRFS_IOC_LOGICAL_INO _IOWR(BTRFS_IOCTL_MAGIC, 36, \ diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index c32667318ae4..d73439b4d7da 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2441,6 +2441,7 @@ static u64 div_factor(u64 num, int factor) static int __btrfs_balance(struct btrfs_fs_info *fs_info) { + struct btrfs_balance_control *bctl = fs_info->balance_ctl; struct btrfs_root *chunk_root = fs_info->chunk_root; struct btrfs_root *dev_root = fs_info->dev_root; struct list_head *devices; @@ -2456,6 +2457,7 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) int slot; int ret; int enospc_errors = 0; + bool counting = true; /* step one make some room on all the devices */ devices = &fs_info->fs_devices->devices; @@ -2487,12 +2489,18 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) ret = -ENOMEM; goto error; } + + /* zero out stat counters */ + spin_lock(&fs_info->balance_lock); + memset(&bctl->stat, 0, sizeof(bctl->stat)); + spin_unlock(&fs_info->balance_lock); +again: key.objectid = BTRFS_FIRST_CHUNK_TREE_OBJECTID; key.offset = (u64)-1; key.type = BTRFS_CHUNK_ITEM_KEY; while (1) { - if (atomic_read(&fs_info->balance_pause_req) || + if ((!counting && atomic_read(&fs_info->balance_pause_req)) || atomic_read(&fs_info->balance_cancel_req)) { ret = -ECANCELED; goto error; @@ -2529,24 +2537,47 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) chunk = btrfs_item_ptr(leaf, slot, struct btrfs_chunk); + if (!counting) { + spin_lock(&fs_info->balance_lock); + bctl->stat.considered++; + spin_unlock(&fs_info->balance_lock); + } + ret = should_balance_chunk(chunk_root, leaf, chunk, found_key.offset); btrfs_release_path(path); if (!ret) goto loop; + if (counting) { + spin_lock(&fs_info->balance_lock); + bctl->stat.expected++; + spin_unlock(&fs_info->balance_lock); + goto loop; + } + ret = btrfs_relocate_chunk(chunk_root, chunk_root->root_key.objectid, found_key.objectid, found_key.offset); if (ret && ret != -ENOSPC) goto error; - if (ret == -ENOSPC) + if (ret == -ENOSPC) { enospc_errors++; + } else { + spin_lock(&fs_info->balance_lock); + bctl->stat.completed++; + spin_unlock(&fs_info->balance_lock); + } loop: key.offset = found_key.offset - 1; } + if (counting) { + btrfs_release_path(path); + counting = false; + goto again; + } error: btrfs_free_path(path); if (enospc_errors) { @@ -2576,7 +2607,7 @@ static void __cancel_balance(struct btrfs_fs_info *fs_info) BUG_ON(ret); } -void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, +void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock, struct btrfs_ioctl_balance_args *bargs); /* @@ -2706,7 +2737,7 @@ do_balance: if (bargs) { memset(bargs, 0, sizeof(*bargs)); - update_ioctl_balance_args(fs_info, bargs); + update_ioctl_balance_args(fs_info, 0, bargs); } if ((ret && ret != -ECANCELED && ret != -ENOSPC) || diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index caa9abd218ef..6faec9dd1f93 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -218,6 +218,7 @@ struct map_lookup { #define BTRFS_BALANCE_ARGS_SOFT (1ULL << 9) struct btrfs_balance_args; +struct btrfs_balance_progress; struct btrfs_balance_control { struct btrfs_fs_info *fs_info; @@ -226,6 +227,8 @@ struct btrfs_balance_control { struct btrfs_balance_args sys; u64 flags; + + struct btrfs_balance_progress stat; }; int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start, From 7ad85bb76a61801362701b77c5cee5aa09f35369 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 12 Jan 2012 19:10:12 -0500 Subject: [PATCH 414/820] Btrfs: do not use btrfs_end_transaction_throttle everywhere A user reported a problem where things like open with O_CREAT would take up to 30 seconds when he had nfs activity on the same mount. This is because all of our quick metadata operations, like create, symlink etc all do btrfs_end_transaction_throttle, which if the transaction is blocked will wait for the commit to complete before it returns. This adds a ridiculous amount of latency and isn't really needed. The normal btrfs_end_transaction will mark the transaction as blocked and wake the transaction kthread up if it thinks the transaction needs to end (this being in the running out of global reserve space scenario), and this is all that is really needed since we've already done everything we're going to do, we just need to return. This should help people with the latency they were seeing when using synchronous heavy workloads. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 18 +++++++++--------- fs/btrfs/xattr.c | 2 +- 2 files changed, 10 insertions(+), 10 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index acc4ff39ca4e..5f8ba210c0aa 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2845,7 +2845,7 @@ static void __unlink_end_trans(struct btrfs_trans_handle *trans, BUG_ON(!root->fs_info->enospc_unlink); root->fs_info->enospc_unlink = 0; } - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); } static int btrfs_unlink(struct inode *dir, struct dentry *dentry) @@ -3434,7 +3434,7 @@ static int btrfs_setsize(struct inode *inode, loff_t newsize) i_size_write(inode, newsize); btrfs_ordered_update_i_size(inode, i_size_read(inode), NULL); ret = btrfs_update_inode(trans, root, inode); - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); } else { /* @@ -4655,7 +4655,7 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry, } out_unlock: nr = trans->blocks_used; - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); btrfs_btree_balance_dirty(root, nr); if (drop_inode) { inode_dec_link_count(inode); @@ -4723,7 +4723,7 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry, } out_unlock: nr = trans->blocks_used; - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); if (drop_inode) { inode_dec_link_count(inode); iput(inode); @@ -4782,7 +4782,7 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, } nr = trans->blocks_used; - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); fail: if (drop_inode) { inode_dec_link_count(inode); @@ -4848,7 +4848,7 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) out_fail: nr = trans->blocks_used; - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); if (drop_on_err) iput(inode); btrfs_btree_balance_dirty(root, nr); @@ -6668,7 +6668,7 @@ end_trans: err = ret; nr = trans->blocks_used; - ret = btrfs_end_transaction_throttle(trans, root); + ret = btrfs_end_transaction(trans, root); btrfs_btree_balance_dirty(root, nr); } @@ -7075,7 +7075,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, btrfs_end_log_trans(root); } out_fail: - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); out_notrans: if (old_ino == BTRFS_FIRST_FREE_OBJECTID) up_read(&root->fs_info->subvol_sem); @@ -7247,7 +7247,7 @@ out_unlock: if (!err) d_instantiate(dentry, inode); nr = trans->blocks_used; - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); if (drop_inode) { inode_dec_link_count(inode); iput(inode); diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index 3848b04e310e..e7a5659087e6 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c @@ -200,7 +200,7 @@ int __btrfs_setxattr(struct btrfs_trans_handle *trans, ret = btrfs_update_inode(trans, root, inode); BUG_ON(ret); out: - btrfs_end_transaction_throttle(trans, root); + btrfs_end_transaction(trans, root); return ret; } From f70a9a6b94af86fca069a7552ab672c31b457786 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 12 Jan 2012 19:10:12 -0500 Subject: [PATCH 415/820] Btrfs: fix btrfsck error 400 when truncating a compressed Reproduce steps: # mkfs.btrfs /dev/sdb5 # mount /dev/sdb5 -o compress=lzo /mnt # dd if=/dev/zero of=/mnt/tmpfile bs=128K count=1 # sync # truncate -s 64K /mnt/tmpfile root 5 inode 257 errors 400 This is because of the wrong if condition, which is used to check if we should subtract the bytes of the dropped range from i_blocks/i_bytes of i-node or not. When we truncate a compressed extent, btrfs substracts the bytes of the whole extent, it's wrong. We should substract the real size that we truncate, no matter it is a compressed extent or not. Fix it. Signed-off-by: Miao Xie Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5f8ba210c0aa..946a7f1b3295 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3009,7 +3009,6 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, int pending_del_nr = 0; int pending_del_slot = 0; int extent_type = -1; - int encoding; int ret; int err = 0; u64 ino = btrfs_ino(inode); @@ -3059,7 +3058,6 @@ search_again: leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); found_type = btrfs_key_type(&found_key); - encoding = 0; if (found_key.objectid != ino) break; @@ -3072,10 +3070,6 @@ search_again: fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item); extent_type = btrfs_file_extent_type(leaf, fi); - encoding = btrfs_file_extent_compression(leaf, fi); - encoding |= btrfs_file_extent_encryption(leaf, fi); - encoding |= btrfs_file_extent_other_encoding(leaf, fi); - if (extent_type != BTRFS_FILE_EXTENT_INLINE) { item_end += btrfs_file_extent_num_bytes(leaf, fi); @@ -3103,7 +3097,7 @@ search_again: if (extent_type != BTRFS_FILE_EXTENT_INLINE) { u64 num_dec; extent_start = btrfs_file_extent_disk_bytenr(leaf, fi); - if (!del_item && !encoding) { + if (!del_item) { u64 orig_num_bytes = btrfs_file_extent_num_bytes(leaf, fi); extent_num_bytes = new_size - From ec39e180fd3188c983c94603634bfcd019f42ae7 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 12 Jan 2012 19:10:12 -0500 Subject: [PATCH 416/820] Btrfs: release space on error in page_mkwrite If updating the inode gave us an ENOSPC we were just returning in page_mkwrite, which is a problem since we make our reservation right before trying to update the inode, so fix the out label so that we actually free our reservation. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 946a7f1b3295..85fd86ea9830 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6488,8 +6488,8 @@ out_unlock: if (!ret) return VM_FAULT_LOCKED; unlock_page(page); - btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE); out: + btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE); return ret; } From 45a8090e626ab470c91142954431a93846030b0d Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 12 Jan 2012 19:10:12 -0500 Subject: [PATCH 417/820] Btrfs: don't call btrfs_throttle in file write Btrfs_throttle will make us wait if there is a currently committing transaction until we can open new transactions, which is ridiculous since we don't actually start any transactions within the file write path anyway, so all this does is introduce big latencies if we have a sync/fsync heavy workload going on while somebody else is trying to do work. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/file.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index fc97b00bd871..0f61e11a2998 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1273,7 +1273,6 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, dirty_pages); if (dirty_pages < (root->leafsize >> PAGE_CACHE_SHIFT) + 1) btrfs_btree_balance_dirty(root, 1); - btrfs_throttle(root); pos += copied; num_written += copied; From 3f7de037fb3727b20bc27332cdcf2488b702394c Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 10 Nov 2011 08:29:20 -0500 Subject: [PATCH 418/820] Btrfs: add allocator tracepoints I used these tracepoints when figuring out what the cluster stuff was doing, so add them to mainline in case we need to profile this stuff again. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 9 ++ fs/btrfs/free-space-cache.c | 12 ++- include/trace/events/btrfs.h | 173 +++++++++++++++++++++++++++++++++++ 3 files changed, 193 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a44072a692ab..ad1a20bc834d 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5256,6 +5256,8 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans, ins->objectid = 0; ins->offset = 0; + trace_find_free_extent(orig_root, num_bytes, empty_size, data); + space_info = __find_space_info(root->fs_info, data); if (!space_info) { printk(KERN_ERR "No space info for %llu\n", data); @@ -5432,6 +5434,8 @@ alloc: if (offset) { /* we have a block, we're done */ spin_unlock(&last_ptr->refill_lock); + trace_btrfs_reserve_extent_cluster(root, + block_group, search_start, num_bytes); goto checks; } @@ -5490,6 +5494,9 @@ refill_cluster: if (offset) { /* we found one, proceed */ spin_unlock(&last_ptr->refill_lock); + trace_btrfs_reserve_extent_cluster(root, + block_group, search_start, + num_bytes); goto checks; } } else if (!cached && loop > LOOP_CACHING_NOWAIT @@ -5576,6 +5583,8 @@ checks: ins->objectid = search_start; ins->offset = num_bytes; + trace_btrfs_reserve_extent(orig_root, block_group, + search_start, num_bytes); if (offset < search_start) btrfs_add_free_space(used_block_group, offset, search_start - offset); diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 6c7887a7770c..efe20032e4a1 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -2346,6 +2346,8 @@ again: &entry->offset_index, 1); BUG_ON(ret); + trace_btrfs_setup_cluster(block_group, cluster, + total_found * block_group->sectorsize, 1); return 0; } @@ -2368,6 +2370,7 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, u64 window_start; u64 window_free; u64 max_extent; + u64 total_size = 0; entry = tree_search_offset(ctl, offset, 0, 1); if (!entry) @@ -2433,11 +2436,12 @@ setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, rb_erase(&entry->offset_index, &ctl->free_space_offset); ret = tree_insert_offset(&cluster->root, entry->offset, &entry->offset_index, 0); + total_size += entry->bytes; BUG_ON(ret); } while (node && entry != last); cluster->max_size = max_extent; - + trace_btrfs_setup_cluster(block_group, cluster, total_size, 0); return 0; } @@ -2542,6 +2546,10 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, goto out; } + trace_btrfs_find_cluster(block_group, offset, bytes, empty_size, + min_bytes); + + INIT_LIST_HEAD(&bitmaps); ret = setup_cluster_no_bitmap(block_group, cluster, &bitmaps, offset, bytes + empty_size, cont1_bytes, min_bytes); @@ -2559,6 +2567,8 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, list_add_tail(&cluster->block_group_list, &block_group->cluster_list); cluster->block_group = block_group; + } else { + trace_btrfs_failed_cluster_setup(block_group); } out: spin_unlock(&cluster->lock); diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index b31702ac15be..1750c0e6660c 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -16,6 +16,8 @@ struct btrfs_delayed_ref_node; struct btrfs_delayed_tree_ref; struct btrfs_delayed_data_ref; struct btrfs_delayed_ref_head; +struct btrfs_block_group_cache; +struct btrfs_free_cluster; struct map_lookup; struct extent_buffer; @@ -44,6 +46,15 @@ struct extent_buffer; obj, ((obj >= BTRFS_DATA_RELOC_TREE_OBJECTID) || \ (obj <= BTRFS_CSUM_TREE_OBJECTID )) ? __show_root_type(obj) : "-" +#define BTRFS_GROUP_FLAGS \ + { BTRFS_BLOCK_GROUP_DATA, "DATA"}, \ + { BTRFS_BLOCK_GROUP_SYSTEM, "SYSTEM"}, \ + { BTRFS_BLOCK_GROUP_METADATA, "METADATA"}, \ + { BTRFS_BLOCK_GROUP_RAID0, "RAID0"}, \ + { BTRFS_BLOCK_GROUP_RAID1, "RAID1"}, \ + { BTRFS_BLOCK_GROUP_DUP, "DUP"}, \ + { BTRFS_BLOCK_GROUP_RAID10, "RAID10"} + TRACE_EVENT(btrfs_transaction_commit, TP_PROTO(struct btrfs_root *root), @@ -659,6 +670,168 @@ DEFINE_EVENT(btrfs__reserved_extent, btrfs_reserved_extent_free, TP_ARGS(root, start, len) ); +TRACE_EVENT(find_free_extent, + + TP_PROTO(struct btrfs_root *root, u64 num_bytes, u64 empty_size, + u64 data), + + TP_ARGS(root, num_bytes, empty_size, data), + + TP_STRUCT__entry( + __field( u64, root_objectid ) + __field( u64, num_bytes ) + __field( u64, empty_size ) + __field( u64, data ) + ), + + TP_fast_assign( + __entry->root_objectid = root->root_key.objectid; + __entry->num_bytes = num_bytes; + __entry->empty_size = empty_size; + __entry->data = data; + ), + + TP_printk("root = %Lu(%s), len = %Lu, empty_size = %Lu, " + "flags = %Lu(%s)", show_root_type(__entry->root_objectid), + __entry->num_bytes, __entry->empty_size, __entry->data, + __print_flags((unsigned long)__entry->data, "|", + BTRFS_GROUP_FLAGS)) +); + +DECLARE_EVENT_CLASS(btrfs__reserve_extent, + + TP_PROTO(struct btrfs_root *root, + struct btrfs_block_group_cache *block_group, u64 start, + u64 len), + + TP_ARGS(root, block_group, start, len), + + TP_STRUCT__entry( + __field( u64, root_objectid ) + __field( u64, bg_objectid ) + __field( u64, flags ) + __field( u64, start ) + __field( u64, len ) + ), + + TP_fast_assign( + __entry->root_objectid = root->root_key.objectid; + __entry->bg_objectid = block_group->key.objectid; + __entry->flags = block_group->flags; + __entry->start = start; + __entry->len = len; + ), + + TP_printk("root = %Lu(%s), block_group = %Lu, flags = %Lu(%s), " + "start = %Lu, len = %Lu", + show_root_type(__entry->root_objectid), __entry->bg_objectid, + __entry->flags, __print_flags((unsigned long)__entry->flags, + "|", BTRFS_GROUP_FLAGS), + __entry->start, __entry->len) +); + +DEFINE_EVENT(btrfs__reserve_extent, btrfs_reserve_extent, + + TP_PROTO(struct btrfs_root *root, + struct btrfs_block_group_cache *block_group, u64 start, + u64 len), + + TP_ARGS(root, block_group, start, len) +); + +DEFINE_EVENT(btrfs__reserve_extent, btrfs_reserve_extent_cluster, + + TP_PROTO(struct btrfs_root *root, + struct btrfs_block_group_cache *block_group, u64 start, + u64 len), + + TP_ARGS(root, block_group, start, len) +); + +TRACE_EVENT(btrfs_find_cluster, + + TP_PROTO(struct btrfs_block_group_cache *block_group, u64 start, + u64 bytes, u64 empty_size, u64 min_bytes), + + TP_ARGS(block_group, start, bytes, empty_size, min_bytes), + + TP_STRUCT__entry( + __field( u64, bg_objectid ) + __field( u64, flags ) + __field( u64, start ) + __field( u64, bytes ) + __field( u64, empty_size ) + __field( u64, min_bytes ) + ), + + TP_fast_assign( + __entry->bg_objectid = block_group->key.objectid; + __entry->flags = block_group->flags; + __entry->start = start; + __entry->bytes = bytes; + __entry->empty_size = empty_size; + __entry->min_bytes = min_bytes; + ), + + TP_printk("block_group = %Lu, flags = %Lu(%s), start = %Lu, len = %Lu," + " empty_size = %Lu, min_bytes = %Lu", __entry->bg_objectid, + __entry->flags, + __print_flags((unsigned long)__entry->flags, "|", + BTRFS_GROUP_FLAGS), __entry->start, + __entry->bytes, __entry->empty_size, __entry->min_bytes) +); + +TRACE_EVENT(btrfs_failed_cluster_setup, + + TP_PROTO(struct btrfs_block_group_cache *block_group), + + TP_ARGS(block_group), + + TP_STRUCT__entry( + __field( u64, bg_objectid ) + ), + + TP_fast_assign( + __entry->bg_objectid = block_group->key.objectid; + ), + + TP_printk("block_group = %Lu", __entry->bg_objectid) +); + +TRACE_EVENT(btrfs_setup_cluster, + + TP_PROTO(struct btrfs_block_group_cache *block_group, + struct btrfs_free_cluster *cluster, u64 size, int bitmap), + + TP_ARGS(block_group, cluster, size, bitmap), + + TP_STRUCT__entry( + __field( u64, bg_objectid ) + __field( u64, flags ) + __field( u64, start ) + __field( u64, max_size ) + __field( u64, size ) + __field( int, bitmap ) + ), + + TP_fast_assign( + __entry->bg_objectid = block_group->key.objectid; + __entry->flags = block_group->flags; + __entry->start = cluster->window_start; + __entry->max_size = cluster->max_size; + __entry->size = size; + __entry->bitmap = bitmap; + ), + + TP_printk("block_group = %Lu, flags = %Lu(%s), window_start = %Lu, " + "size = %Lu, max_size = %Lu, bitmap = %d", + __entry->bg_objectid, + __entry->flags, + __print_flags((unsigned long)__entry->flags, "|", + BTRFS_GROUP_FLAGS), __entry->start, + __entry->size, __entry->max_size, __entry->bitmap) +); + #endif /* _TRACE_BTRFS_H */ /* This part must be outside protection */ From 90290e19820e3323ce6b9c2888eeb68bf29c278b Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 2 Dec 2011 15:44:12 -0500 Subject: [PATCH 419/820] Btrfs: protect orphan block rsv with spin_lock We've been seeing warnings coming out of the orphan commit stuff forever from ceph. Turns out it's because we're racing with checking if the orphan block reserve is set, because we clear it outside of the spin_lock. So leave the normal fastpath checks where they are, but take the spin_lock and _recheck_ to make sure we haven't had an orphan block rsv added in the meantime. Then clear the root's orphan block rsv and release the lock. With this patch a user said the warnings went away and they usually showed up pretty soon after he started ceph. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 23 +++++++++++++++++++---- 1 file changed, 19 insertions(+), 4 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 85fd86ea9830..619742d37166 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1951,12 +1951,28 @@ enum btrfs_orphan_cleanup_state { void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans, struct btrfs_root *root) { + struct btrfs_block_rsv *block_rsv; int ret; if (!list_empty(&root->orphan_list) || root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE) return; + spin_lock(&root->orphan_lock); + if (!list_empty(&root->orphan_list)) { + spin_unlock(&root->orphan_lock); + return; + } + + if (root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE) { + spin_unlock(&root->orphan_lock); + return; + } + + block_rsv = root->orphan_block_rsv; + root->orphan_block_rsv = NULL; + spin_unlock(&root->orphan_lock); + if (root->orphan_item_inserted && btrfs_root_refs(&root->root_item) > 0) { ret = btrfs_del_orphan_item(trans, root->fs_info->tree_root, @@ -1965,10 +1981,9 @@ void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans, root->orphan_item_inserted = 0; } - if (root->orphan_block_rsv) { - WARN_ON(root->orphan_block_rsv->size > 0); - btrfs_free_block_rsv(root, root->orphan_block_rsv); - root->orphan_block_rsv = NULL; + if (block_rsv) { + WARN_ON(block_rsv->size > 0); + btrfs_free_block_rsv(root, block_rsv); } } From 8c2a3ca20f6233677ac3222c6506174010eb414f Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 10 Jan 2012 10:31:31 -0500 Subject: [PATCH 420/820] Btrfs: space leak tracepoints This in addition to a script in my btrfs-tracing tree will help track down space leaks when we're getting space left over in block groups on umount. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/delayed-inode.c | 45 ++++++++++++++++++++-------- fs/btrfs/extent-tree.c | 58 +++++++++++++++++++++++++++++++----- fs/btrfs/inode-map.c | 4 +++ fs/btrfs/transaction.c | 2 ++ include/trace/events/btrfs.h | 30 +++++++++++++++++++ 5 files changed, 119 insertions(+), 20 deletions(-) diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index 9c1eccc2c503..fe4cd0f1cef1 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -595,8 +595,12 @@ static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans, num_bytes = btrfs_calc_trans_metadata_size(root, 1); ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes); - if (!ret) + if (!ret) { + trace_btrfs_space_reservation(root->fs_info, "delayed_item", + item->key.objectid, + num_bytes, 1); item->bytes_reserved = num_bytes; + } return ret; } @@ -610,6 +614,9 @@ static void btrfs_delayed_item_release_metadata(struct btrfs_root *root, return; rsv = &root->fs_info->delayed_block_rsv; + trace_btrfs_space_reservation(root->fs_info, "delayed_item", + item->key.objectid, item->bytes_reserved, + 0); btrfs_block_rsv_release(root, rsv, item->bytes_reserved); } @@ -624,7 +631,7 @@ static int btrfs_delayed_inode_reserve_metadata( struct btrfs_block_rsv *dst_rsv; u64 num_bytes; int ret; - int release = false; + bool release = false; src_rsv = trans->block_rsv; dst_rsv = &root->fs_info->delayed_block_rsv; @@ -651,8 +658,13 @@ static int btrfs_delayed_inode_reserve_metadata( */ if (ret == -EAGAIN) ret = -ENOSPC; - if (!ret) + if (!ret) { node->bytes_reserved = num_bytes; + trace_btrfs_space_reservation(root->fs_info, + "delayed_inode", + btrfs_ino(inode), + num_bytes, 1); + } return ret; } else if (src_rsv == &root->fs_info->delalloc_block_rsv) { spin_lock(&BTRFS_I(inode)->lock); @@ -707,11 +719,17 @@ out: * reservation here. I think it may be time for a documentation page on * how block rsvs. work. */ - if (!ret) + if (!ret) { + trace_btrfs_space_reservation(root->fs_info, "delayed_inode", + btrfs_ino(inode), num_bytes, 1); node->bytes_reserved = num_bytes; + } - if (release) + if (release) { + trace_btrfs_space_reservation(root->fs_info, "delalloc", + btrfs_ino(inode), num_bytes, 0); btrfs_block_rsv_release(root, src_rsv, num_bytes); + } return ret; } @@ -725,6 +743,8 @@ static void btrfs_delayed_inode_release_metadata(struct btrfs_root *root, return; rsv = &root->fs_info->delayed_block_rsv; + trace_btrfs_space_reservation(root->fs_info, "delayed_inode", + node->inode_id, node->bytes_reserved, 0); btrfs_block_rsv_release(root, rsv, node->bytes_reserved); node->bytes_reserved = 0; @@ -1372,13 +1392,6 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans, goto release_node; } - ret = btrfs_delayed_item_reserve_metadata(trans, root, delayed_item); - /* - * we have reserved enough space when we start a new transaction, - * so reserving metadata failure is impossible - */ - BUG_ON(ret); - delayed_item->key.objectid = btrfs_ino(dir); btrfs_set_key_type(&delayed_item->key, BTRFS_DIR_INDEX_KEY); delayed_item->key.offset = index; @@ -1391,6 +1404,14 @@ int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans, dir_item->type = type; memcpy((char *)(dir_item + 1), name, name_len); + ret = btrfs_delayed_item_reserve_metadata(trans, root, delayed_item); + /* + * we have reserved enough space when we start a new transaction, + * so reserving metadata failure is impossible + */ + BUG_ON(ret); + + mutex_lock(&delayed_node->mutex); ret = __btrfs_add_delayed_insertion_item(delayed_node, delayed_item); if (unlikely(ret)) { diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index ad1a20bc834d..556f9aa25bb7 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3310,6 +3310,8 @@ commit_trans: return -ENOSPC; } data_sinfo->bytes_may_use += bytes; + trace_btrfs_space_reservation(root->fs_info, "space_info", + (u64)data_sinfo, bytes, 1); spin_unlock(&data_sinfo->lock); return 0; @@ -3329,6 +3331,8 @@ void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes) data_sinfo = BTRFS_I(inode)->space_info; spin_lock(&data_sinfo->lock); data_sinfo->bytes_may_use -= bytes; + trace_btrfs_space_reservation(root->fs_info, "space_info", + (u64)data_sinfo, bytes, 0); spin_unlock(&data_sinfo->lock); } @@ -3686,6 +3690,10 @@ again: if (used <= space_info->total_bytes) { if (used + orig_bytes <= space_info->total_bytes) { space_info->bytes_may_use += orig_bytes; + trace_btrfs_space_reservation(root->fs_info, + "space_info", + (u64)space_info, + orig_bytes, 1); ret = 0; } else { /* @@ -3753,6 +3761,10 @@ again: if (used + num_bytes < space_info->total_bytes + avail) { space_info->bytes_may_use += orig_bytes; + trace_btrfs_space_reservation(root->fs_info, + "space_info", + (u64)space_info, + orig_bytes, 1); ret = 0; } else { wait_ordered = true; @@ -3859,7 +3871,8 @@ static void block_rsv_add_bytes(struct btrfs_block_rsv *block_rsv, spin_unlock(&block_rsv->lock); } -static void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv, +static void block_rsv_release_bytes(struct btrfs_fs_info *fs_info, + struct btrfs_block_rsv *block_rsv, struct btrfs_block_rsv *dest, u64 num_bytes) { struct btrfs_space_info *space_info = block_rsv->space_info; @@ -3895,6 +3908,9 @@ static void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv, if (num_bytes) { spin_lock(&space_info->lock); space_info->bytes_may_use -= num_bytes; + trace_btrfs_space_reservation(fs_info, "space_info", + (u64)space_info, + num_bytes, 0); space_info->reservation_progress++; spin_unlock(&space_info->lock); } @@ -4051,7 +4067,8 @@ void btrfs_block_rsv_release(struct btrfs_root *root, if (global_rsv->full || global_rsv == block_rsv || block_rsv->space_info != global_rsv->space_info) global_rsv = NULL; - block_rsv_release_bytes(block_rsv, global_rsv, num_bytes); + block_rsv_release_bytes(root->fs_info, block_rsv, global_rsv, + num_bytes); } /* @@ -4110,11 +4127,15 @@ static void update_global_block_rsv(struct btrfs_fs_info *fs_info) num_bytes = sinfo->total_bytes - num_bytes; block_rsv->reserved += num_bytes; sinfo->bytes_may_use += num_bytes; + trace_btrfs_space_reservation(fs_info, "space_info", + (u64)sinfo, num_bytes, 1); } if (block_rsv->reserved >= block_rsv->size) { num_bytes = block_rsv->reserved - block_rsv->size; sinfo->bytes_may_use -= num_bytes; + trace_btrfs_space_reservation(fs_info, "space_info", + (u64)sinfo, num_bytes, 0); sinfo->reservation_progress++; block_rsv->reserved = block_rsv->size; block_rsv->full = 1; @@ -4149,7 +4170,8 @@ static void init_global_block_rsv(struct btrfs_fs_info *fs_info) static void release_global_block_rsv(struct btrfs_fs_info *fs_info) { - block_rsv_release_bytes(&fs_info->global_block_rsv, NULL, (u64)-1); + block_rsv_release_bytes(fs_info, &fs_info->global_block_rsv, NULL, + (u64)-1); WARN_ON(fs_info->delalloc_block_rsv.size > 0); WARN_ON(fs_info->delalloc_block_rsv.reserved > 0); WARN_ON(fs_info->trans_block_rsv.size > 0); @@ -4166,6 +4188,8 @@ void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans, if (!trans->bytes_reserved) return; + trace_btrfs_space_reservation(root->fs_info, "transaction", (u64)trans, + trans->bytes_reserved, 0); btrfs_block_rsv_release(root, trans->block_rsv, trans->bytes_reserved); trans->bytes_reserved = 0; } @@ -4183,6 +4207,8 @@ int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans, * when we are truly done with the orphan item. */ u64 num_bytes = btrfs_calc_trans_metadata_size(root, 1); + trace_btrfs_space_reservation(root->fs_info, "orphan", + btrfs_ino(inode), num_bytes, 1); return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes); } @@ -4190,6 +4216,8 @@ void btrfs_orphan_release_metadata(struct inode *inode) { struct btrfs_root *root = BTRFS_I(inode)->root; u64 num_bytes = btrfs_calc_trans_metadata_size(root, 1); + trace_btrfs_space_reservation(root->fs_info, "orphan", + btrfs_ino(inode), num_bytes, 0); btrfs_block_rsv_release(root, root->orphan_block_rsv, num_bytes); } @@ -4370,8 +4398,13 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) if (dropped) to_free += btrfs_calc_trans_metadata_size(root, dropped); - if (to_free) + if (to_free) { btrfs_block_rsv_release(root, block_rsv, to_free); + trace_btrfs_space_reservation(root->fs_info, + "delalloc", + btrfs_ino(inode), + to_free, 0); + } return ret; } @@ -4383,6 +4416,9 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) BTRFS_I(inode)->reserved_extents += nr_extents; spin_unlock(&BTRFS_I(inode)->lock); + if (to_reserve) + trace_btrfs_space_reservation(root->fs_info,"delalloc", + btrfs_ino(inode), to_reserve, 1); block_rsv_add_bytes(block_rsv, to_reserve, 1); return 0; @@ -4412,6 +4448,8 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) if (dropped > 0) to_free += btrfs_calc_trans_metadata_size(root, dropped); + trace_btrfs_space_reservation(root->fs_info, "delalloc", + btrfs_ino(inode), to_free, 0); btrfs_block_rsv_release(root, &root->fs_info->delalloc_block_rsv, to_free); } @@ -4666,7 +4704,10 @@ static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, cache->reserved += num_bytes; space_info->bytes_reserved += num_bytes; if (reserve == RESERVE_ALLOC) { - BUG_ON(space_info->bytes_may_use < num_bytes); + trace_btrfs_space_reservation(cache->fs_info, + "space_info", + (u64)space_info, + num_bytes, 0); space_info->bytes_may_use -= num_bytes; } } @@ -6126,10 +6167,11 @@ use_block_rsv(struct btrfs_trans_handle *trans, return ERR_PTR(-ENOSPC); } -static void unuse_block_rsv(struct btrfs_block_rsv *block_rsv, u32 blocksize) +static void unuse_block_rsv(struct btrfs_fs_info *fs_info, + struct btrfs_block_rsv *block_rsv, u32 blocksize) { block_rsv_add_bytes(block_rsv, blocksize, 0); - block_rsv_release_bytes(block_rsv, NULL, 0); + block_rsv_release_bytes(fs_info, block_rsv, NULL, 0); } /* @@ -6159,7 +6201,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, ret = btrfs_reserve_extent(trans, root, blocksize, blocksize, empty_size, hint, (u64)-1, &ins, 0); if (ret) { - unuse_block_rsv(block_rsv, blocksize); + unuse_block_rsv(root->fs_info, block_rsv, blocksize); return ERR_PTR(ret); } diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c index f8962a957d65..213ffa86ce1b 100644 --- a/fs/btrfs/inode-map.c +++ b/fs/btrfs/inode-map.c @@ -438,6 +438,8 @@ int btrfs_save_ino_cache(struct btrfs_root *root, trans->bytes_reserved); if (ret) goto out; + trace_btrfs_space_reservation(root->fs_info, "ino_cache", (u64)trans, + trans->bytes_reserved, 1); again: inode = lookup_free_ino_inode(root, path); if (IS_ERR(inode) && PTR_ERR(inode) != -ENOENT) { @@ -498,6 +500,8 @@ again: out_put: iput(inode); out_release: + trace_btrfs_space_reservation(root->fs_info, "ino_cache", (u64)trans, + trans->bytes_reserved, 0); btrfs_block_rsv_release(root, trans->block_rsv, trans->bytes_reserved); out: trans->block_rsv = rsv; diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index d5f987b49d70..287a6728b1ad 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -326,6 +326,8 @@ again: } if (num_bytes) { + trace_btrfs_space_reservation(root->fs_info, "transaction", + (u64)h, num_bytes, 1); h->block_rsv = &root->fs_info->trans_block_rsv; h->bytes_reserved = num_bytes; } diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index 1750c0e6660c..84f3001a568d 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -55,6 +55,8 @@ struct extent_buffer; { BTRFS_BLOCK_GROUP_DUP, "DUP"}, \ { BTRFS_BLOCK_GROUP_RAID10, "RAID10"} +#define BTRFS_UUID_SIZE 16 + TRACE_EVENT(btrfs_transaction_commit, TP_PROTO(struct btrfs_root *root), @@ -632,6 +634,34 @@ TRACE_EVENT(btrfs_cow_block, __entry->cow_level) ); +TRACE_EVENT(btrfs_space_reservation, + + TP_PROTO(struct btrfs_fs_info *fs_info, char *type, u64 val, + u64 bytes, int reserve), + + TP_ARGS(fs_info, type, val, bytes, reserve), + + TP_STRUCT__entry( + __array( u8, fsid, BTRFS_UUID_SIZE ) + __string( type, type ) + __field( u64, val ) + __field( u64, bytes ) + __field( int, reserve ) + ), + + TP_fast_assign( + memcpy(__entry->fsid, fs_info->fsid, BTRFS_UUID_SIZE); + __assign_str(type, type); + __entry->val = val; + __entry->bytes = bytes; + __entry->reserve = reserve; + ), + + TP_printk("%pU: %s: %Lu %s %Lu", __entry->fsid, __get_str(type), + __entry->val, __entry->reserve ? "reserve" : "release", + __entry->bytes) +); + DECLARE_EVENT_CLASS(btrfs__reserved_extent, TP_PROTO(struct btrfs_root *root, u64 start, u64 len), From f248679e86fead40cc78e724c7181d6bec1a2046 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 13 Jan 2012 12:09:22 -0500 Subject: [PATCH 421/820] Btrfs: add a delalloc mutex to inodes for delalloc reservations I was using i_mutex for this, but we're getting bogus lockdep warnings by doing that and theres no real way to get rid of those, so just stop using i_mutex to protect delalloc metadata reservations and use a delalloc mutex instead. This shouldn't be contended often at all, only if you are writing and mmap writing to the file at the same time. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/btrfs_inode.h | 3 +++ fs/btrfs/extent-tree.c | 5 +++-- fs/btrfs/inode.c | 11 +---------- fs/btrfs/ioctl.c | 2 -- fs/btrfs/relocation.c | 2 -- 5 files changed, 7 insertions(+), 16 deletions(-) diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index 634608d2a6d0..9b9b15fd5204 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -51,6 +51,9 @@ struct btrfs_inode { /* held while logging the inode in tree-log.c */ struct mutex log_mutex; + /* held while doing delalloc reservations */ + struct mutex delalloc_mutex; + /* used to order data wrt metadata */ struct btrfs_ordered_inode_tree ordered_tree; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 556f9aa25bb7..e0ad5f0f895e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4345,12 +4345,11 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) /* Need to be holding the i_mutex here if we aren't free space cache */ if (btrfs_is_free_space_inode(root, inode)) flush = 0; - else - WARN_ON(!mutex_is_locked(&inode->i_mutex)); if (flush && btrfs_transaction_in_commit(root->fs_info)) schedule_timeout(1); + mutex_lock(&BTRFS_I(inode)->delalloc_mutex); num_bytes = ALIGN(num_bytes, root->sectorsize); spin_lock(&BTRFS_I(inode)->lock); @@ -4405,6 +4404,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) btrfs_ino(inode), to_free, 0); } + mutex_unlock(&BTRFS_I(inode)->delalloc_mutex); return ret; } @@ -4415,6 +4415,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) } BTRFS_I(inode)->reserved_extents += nr_extents; spin_unlock(&BTRFS_I(inode)->lock); + mutex_unlock(&BTRFS_I(inode)->delalloc_mutex); if (to_reserve) trace_btrfs_space_reservation(root->fs_info,"delalloc", diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 619742d37166..5977987abdb1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2239,14 +2239,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) continue; } nr_truncate++; - /* - * Need to hold the imutex for reservation purposes, not - * a huge deal here but I have a WARN_ON in - * btrfs_delalloc_reserve_space to catch offenders. - */ - mutex_lock(&inode->i_mutex); ret = btrfs_truncate(inode); - mutex_unlock(&inode->i_mutex); } else { nr_unlink++; } @@ -6411,10 +6404,7 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) u64 page_start; u64 page_end; - /* Need this to keep space reservations serialized */ - mutex_lock(&inode->i_mutex); ret = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE); - mutex_unlock(&inode->i_mutex); if (!ret) ret = btrfs_update_time(vma->vm_file); if (ret) { @@ -6758,6 +6748,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) extent_io_tree_init(&ei->io_tree, &inode->i_data); extent_io_tree_init(&ei->io_failure_tree, &inode->i_data); mutex_init(&ei->log_mutex); + mutex_init(&ei->delalloc_mutex); btrfs_ordered_inode_tree_init(&ei->ordered_tree); INIT_LIST_HEAD(&ei->i_orphan); INIT_LIST_HEAD(&ei->delalloc_inodes); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 7fdf22c2dc0d..6834be4c8709 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -868,10 +868,8 @@ static int cluster_pages_for_defrag(struct inode *inode, return 0; file_end = (isize - 1) >> PAGE_CACHE_SHIFT; - mutex_lock(&inode->i_mutex); ret = btrfs_delalloc_reserve_space(inode, num_pages << PAGE_CACHE_SHIFT); - mutex_unlock(&inode->i_mutex); if (ret) return ret; again: diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index efe9f792544d..8c1aae2c845d 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -2949,9 +2949,7 @@ static int relocate_file_extent_cluster(struct inode *inode, index = (cluster->start - offset) >> PAGE_CACHE_SHIFT; last_index = (cluster->end - offset) >> PAGE_CACHE_SHIFT; while (index <= last_index) { - mutex_lock(&inode->i_mutex); ret = btrfs_delalloc_reserve_metadata(inode, PAGE_CACHE_SIZE); - mutex_unlock(&inode->i_mutex); if (ret) goto out; From 96bdc7dc61fb1b1e8e858dafb13abee8482ba064 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Mon, 16 Jan 2012 08:13:11 -0500 Subject: [PATCH 422/820] Btrfs: use larger system chunks system chunks by default are very small. This makes them slightly larger and also fixes the conditional checks to make sure we don't allocate a billion of them at once. Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 3 +++ fs/btrfs/volumes.c | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index e0ad5f0f895e..700879ed64cf 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3384,6 +3384,9 @@ static int should_alloc_chunk(struct btrfs_root *root, /* 256MB or 2% of the FS */ thresh = max_t(u64, 256 * 1024 * 1024, div_factor_fine(thresh, 2)); + /* system chunks need a much small threshold */ + if (sinfo->flags & BTRFS_BLOCK_GROUP_SYSTEM) + thresh = 32 * 1024 * 1024; if (num_bytes > thresh && sinfo->bytes_used < div_factor(num_bytes, 8)) return 0; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 59e878f9fdcc..7ffdb154daec 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3166,7 +3166,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, max_stripe_size = 256 * 1024 * 1024; max_chunk_size = max_stripe_size; } else if (type & BTRFS_BLOCK_GROUP_SYSTEM) { - max_stripe_size = 8 * 1024 * 1024; + max_stripe_size = 32 * 1024 * 1024; max_chunk_size = 2 * max_stripe_size; } else { printk(KERN_ERR "btrfs: invalid chunk type 0x%llx requested\n", From da87c937e5a2374686edd58df06cfd5050b125fa Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Mon, 16 Jan 2012 15:17:50 -0600 Subject: [PATCH 423/820] x86/UV2: Fix new UV2 hardware by using native UV2 broadcast mode Update the use of the Broadcast Assist Unit on SGI Altix UV2 to the use of native UV2 mode on new hardware (not the legacy mode). UV2 native mode has a different format for a broadcast message. We also need quick differentiaton between UV1 and UV2. Signed-off-by: Cliff Wickman Link: http://lkml.kernel.org/r/20120116211750.GA5767@sgi.com Cc: Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_bau.h | 93 +++++++++++++++++++++++++++++--- arch/x86/platform/uv/tlb_uv.c | 88 ++++++++++++++++++++++-------- 2 files changed, 151 insertions(+), 30 deletions(-) diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h index 8e862aaf0d90..4a46b27ee9a0 100644 --- a/arch/x86/include/asm/uv/uv_bau.h +++ b/arch/x86/include/asm/uv/uv_bau.h @@ -65,7 +65,7 @@ * UV2: Bit 19 selects between * (0): 10 microsecond timebase and * (1): 80 microseconds - * we're using 655us, similar to UV1: 65 units of 10us + * we're using 560us, similar to UV1: 65 units of 10us */ #define UV1_INTD_SOFT_ACK_TIMEOUT_PERIOD (9UL) #define UV2_INTD_SOFT_ACK_TIMEOUT_PERIOD (15UL) @@ -235,10 +235,10 @@ struct bau_msg_payload { /* - * Message header: 16 bytes (128 bits) (bytes 0x30-0x3f of descriptor) + * UV1 Message header: 16 bytes (128 bits) (bytes 0x30-0x3f of descriptor) * see table 4.2.3.0.1 in broacast_assist spec. */ -struct bau_msg_header { +struct uv1_bau_msg_header { unsigned int dest_subnodeid:6; /* must be 0x10, for the LB */ /* bits 5:0 */ unsigned int base_dest_nasid:15; /* nasid of the first bit */ @@ -317,20 +317,88 @@ struct bau_msg_header { /* bits 127:107 */ }; +/* + * UV2 Message header: 16 bytes (128 bits) (bytes 0x30-0x3f of descriptor) + * see figure 9-2 of harp_sys.pdf + */ +struct uv2_bau_msg_header { + unsigned int base_dest_nasid:15; /* nasid of the first bit */ + /* bits 14:0 */ /* in uvhub map */ + unsigned int dest_subnodeid:5; /* must be 0x10, for the LB */ + /* bits 19:15 */ + unsigned int rsvd_1:1; /* must be zero */ + /* bit 20 */ + /* Address bits 59:21 */ + /* bits 25:2 of address (44:21) are payload */ + /* these next 24 bits become bytes 12-14 of msg */ + /* bits 28:21 land in byte 12 */ + unsigned int replied_to:1; /* sent as 0 by the source to + byte 12 */ + /* bit 21 */ + unsigned int msg_type:3; /* software type of the + message */ + /* bits 24:22 */ + unsigned int canceled:1; /* message canceled, resource + is to be freed*/ + /* bit 25 */ + unsigned int payload_1:3; /* not currently used */ + /* bits 28:26 */ + + /* bits 36:29 land in byte 13 */ + unsigned int payload_2a:3; /* not currently used */ + unsigned int payload_2b:5; /* not currently used */ + /* bits 36:29 */ + + /* bits 44:37 land in byte 14 */ + unsigned int payload_3:8; /* not currently used */ + /* bits 44:37 */ + + unsigned int rsvd_2:7; /* reserved */ + /* bits 51:45 */ + unsigned int swack_flag:1; /* software acknowledge flag */ + /* bit 52 */ + unsigned int rsvd_3a:3; /* must be zero */ + unsigned int rsvd_3b:8; /* must be zero */ + unsigned int rsvd_3c:8; /* must be zero */ + unsigned int rsvd_3d:3; /* must be zero */ + /* bits 74:53 */ + unsigned int fairness:3; /* usually zero */ + /* bits 77:75 */ + + unsigned int sequence:16; /* message sequence number */ + /* bits 93:78 Suppl_A */ + unsigned int chaining:1; /* next descriptor is part of + this activation*/ + /* bit 94 */ + unsigned int multilevel:1; /* multi-level multicast + format */ + /* bit 95 */ + unsigned int rsvd_4:24; /* ordered / source node / + source subnode / aging + must be zero */ + /* bits 119:96 */ + unsigned int command:8; /* message type */ + /* bits 127:120 */ +}; + /* * The activation descriptor: * The format of the message to send, plus all accompanying control * Should be 64 bytes */ struct bau_desc { - struct pnmask distribution; + struct pnmask distribution; /* * message template, consisting of header and payload: */ - struct bau_msg_header header; - struct bau_msg_payload payload; + union bau_msg_header { + struct uv1_bau_msg_header uv1_hdr; + struct uv2_bau_msg_header uv2_hdr; + } header; + + struct bau_msg_payload payload; }; -/* +/* UV1: * -payload-- ---------header------ * bytes 0-11 bits 41-56 bits 58-81 * A B (2) C (3) @@ -340,6 +408,16 @@ struct bau_desc { * bytes 0-11 bytes 12-14 bytes 16-17 (byte 15 filled in by hw as vector) * ------------payload queue----------- */ +/* UV2: + * -payload-- ---------header------ + * bytes 0-11 bits 70-78 bits 21-44 + * A B (2) C (3) + * + * A/B/C are moved to: + * A C B + * bytes 0-11 bytes 12-14 bytes 16-17 (byte 15 filled in by hw as vector) + * ------------payload queue----------- + */ /* * The payload queue on the destination side is an array of these. @@ -511,6 +589,7 @@ struct bau_control { short osnode; short uvhub_cpu; short uvhub; + short uvhub_version; short cpus_in_socket; short cpus_in_uvhub; short partition_base_pnode; diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index 5b552198f774..1341a2e06542 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -573,7 +573,7 @@ static int wait_completion(struct bau_desc *bau_desc, right_shift = ((cpu - UV_CPUS_PER_AS) * UV_ACT_STATUS_SIZE); } - if (is_uv1_hub()) + if (bcp->uvhub_version == 1) return uv1_wait_completion(bau_desc, mmr_offset, right_shift, bcp, try); else @@ -757,15 +757,22 @@ int uv_flush_send_and_wait(struct bau_desc *bau_desc, { int seq_number = 0; int completion_stat = 0; + int uv1 = 0; long try = 0; unsigned long index; cycles_t time1; cycles_t time2; struct ptc_stats *stat = bcp->statp; struct bau_control *hmaster = bcp->uvhub_master; + struct uv1_bau_msg_header *uv1_hdr = NULL; + struct uv2_bau_msg_header *uv2_hdr = NULL; - if (is_uv1_hub()) + if (bcp->uvhub_version == 1) { + uv1 = 1; uv1_throttle(hmaster, stat); + uv1_hdr = &bau_desc->header.uv1_hdr; + } else + uv2_hdr = &bau_desc->header.uv2_hdr; while (hmaster->uvhub_quiesce) cpu_relax(); @@ -773,14 +780,23 @@ int uv_flush_send_and_wait(struct bau_desc *bau_desc, time1 = get_cycles(); do { if (try == 0) { - bau_desc->header.msg_type = MSG_REGULAR; + if (uv1) + uv1_hdr->msg_type = MSG_REGULAR; + else + uv2_hdr->msg_type = MSG_REGULAR; seq_number = bcp->message_number++; } else { - bau_desc->header.msg_type = MSG_RETRY; + if (uv1) + uv1_hdr->msg_type = MSG_RETRY; + else + uv2_hdr->msg_type = MSG_RETRY; stat->s_retry_messages++; } - bau_desc->header.sequence = seq_number; + if (uv1) + uv1_hdr->sequence = seq_number; + else + uv2_hdr->sequence = seq_number; index = (1UL << AS_PUSH_SHIFT) | bcp->uvhub_cpu; bcp->send_message = get_cycles(); @@ -967,7 +983,7 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask, stat->s_ntargself++; bau_desc = bcp->descriptor_base; - bau_desc += ITEMS_PER_DESC * bcp->uvhub_cpu; + bau_desc += (ITEMS_PER_DESC * bcp->uvhub_cpu); bau_uvhubs_clear(&bau_desc->distribution, UV_DISTRIBUTION_SIZE); if (set_distrib_bits(flush_mask, bcp, bau_desc, &locals, &remotes)) return NULL; @@ -1083,7 +1099,7 @@ static void __init enable_timeouts(void) */ mmr_image |= (1L << SOFTACK_MSHIFT); if (is_uv2_hub()) { - mmr_image |= (1L << UV2_LEG_SHFT); + mmr_image &= ~(1L << UV2_LEG_SHFT); mmr_image |= (1L << UV2_EXT_SHFT); } write_mmr_misc_control(pnode, mmr_image); @@ -1432,12 +1448,15 @@ static void activation_descriptor_init(int node, int pnode, int base_pnode) { int i; int cpu; + int uv1 = 0; unsigned long gpa; unsigned long m; unsigned long n; size_t dsize; struct bau_desc *bau_desc; struct bau_desc *bd2; + struct uv1_bau_msg_header *uv1_hdr; + struct uv2_bau_msg_header *uv2_hdr; struct bau_control *bcp; /* @@ -1451,6 +1470,8 @@ static void activation_descriptor_init(int node, int pnode, int base_pnode) gpa = uv_gpa(bau_desc); n = uv_gpa_to_gnode(gpa); m = uv_gpa_to_offset(gpa); + if (is_uv1_hub()) + uv1 = 1; /* the 14-bit pnode */ write_mmr_descriptor_base(pnode, (n << UV_DESC_PSHIFT | m)); @@ -1461,21 +1482,33 @@ static void activation_descriptor_init(int node, int pnode, int base_pnode) */ for (i = 0, bd2 = bau_desc; i < (ADP_SZ * ITEMS_PER_DESC); i++, bd2++) { memset(bd2, 0, sizeof(struct bau_desc)); - bd2->header.swack_flag = 1; - /* - * The base_dest_nasid set in the message header is the nasid - * of the first uvhub in the partition. The bit map will - * indicate destination pnode numbers relative to that base. - * They may not be consecutive if nasid striding is being used. - */ - bd2->header.base_dest_nasid = UV_PNODE_TO_NASID(base_pnode); - bd2->header.dest_subnodeid = UV_LB_SUBNODEID; - bd2->header.command = UV_NET_ENDPOINT_INTD; - bd2->header.int_both = 1; - /* - * all others need to be set to zero: - * fairness chaining multilevel count replied_to - */ + if (uv1) { + uv1_hdr = &bd2->header.uv1_hdr; + uv1_hdr->swack_flag = 1; + /* + * The base_dest_nasid set in the message header + * is the nasid of the first uvhub in the partition. + * The bit map will indicate destination pnode numbers + * relative to that base. They may not be consecutive + * if nasid striding is being used. + */ + uv1_hdr->base_dest_nasid = + UV_PNODE_TO_NASID(base_pnode); + uv1_hdr->dest_subnodeid = UV_LB_SUBNODEID; + uv1_hdr->command = UV_NET_ENDPOINT_INTD; + uv1_hdr->int_both = 1; + /* + * all others need to be set to zero: + * fairness chaining multilevel count replied_to + */ + } else { + uv2_hdr = &bd2->header.uv2_hdr; + uv2_hdr->swack_flag = 1; + uv2_hdr->base_dest_nasid = + UV_PNODE_TO_NASID(base_pnode); + uv2_hdr->dest_subnodeid = UV_LB_SUBNODEID; + uv2_hdr->command = UV_NET_ENDPOINT_INTD; + } } for_each_present_cpu(cpu) { if (pnode != uv_blade_to_pnode(uv_cpu_to_blade_id(cpu))) @@ -1728,6 +1761,14 @@ static int scan_sock(struct socket_desc *sdp, struct uvhub_desc *bdp, bcp->cpus_in_socket = sdp->num_cpus; bcp->socket_master = *smasterp; bcp->uvhub = bdp->uvhub; + if (is_uv1_hub()) + bcp->uvhub_version = 1; + else if (is_uv2_hub()) + bcp->uvhub_version = 2; + else { + printk(KERN_EMERG "uvhub version not 1 or 2\n"); + return 1; + } bcp->uvhub_master = *hmasterp; bcp->uvhub_cpu = uv_cpu_hub_info(cpu)->blade_processor_id; if (bcp->uvhub_cpu >= MAX_CPUS_PER_UVHUB) { @@ -1867,7 +1908,8 @@ static int __init uv_bau_init(void) val = 1L << 63; write_gmmr_activation(pnode, val); mmr = 1; /* should be 1 to broadcast to both sockets */ - write_mmr_data_broadcast(pnode, mmr); + if (!is_uv1_hub()) + write_mmr_data_broadcast(pnode, mmr); } } From d059f9fa84a30e04279c6ff615e9e2cf3b260191 Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Mon, 16 Jan 2012 15:18:48 -0600 Subject: [PATCH 424/820] x86/UV2: Fix BAU destination timeout initialization Move the call to enable_timeouts() forward so that BAU_MISC_CONTROL is initialized before using it in calculate_destination_timeout(). Fix the calculation of a BAU destination timeout for UV2 (in calculate_destination_timeout()). Signed-off-by: Cliff Wickman Link: http://lkml.kernel.org/r/20120116211848.GB5767@sgi.com Cc: Signed-off-by: Ingo Molnar --- arch/x86/platform/uv/tlb_uv.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index 1341a2e06542..c425ff1a9cc3 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -1617,14 +1617,14 @@ static int calculate_destination_timeout(void) ts_ns = base * mult1 * mult2; ret = ts_ns / 1000; } else { - /* 4 bits 0/1 for 10/80us, 3 bits of multiplier */ - mmr_image = uv_read_local_mmr(UVH_AGING_PRESCALE_SEL); + /* 4 bits 0/1 for 10/80us base, 3 bits of multiplier */ + mmr_image = uv_read_local_mmr(UVH_LB_BAU_MISC_CONTROL); mmr_image = (mmr_image & UV_SA_MASK) >> UV_SA_SHFT; if (mmr_image & (1L << UV2_ACK_UNITS_SHFT)) - mult1 = 80; + base = 80; else - mult1 = 10; - base = mmr_image & UV2_ACK_MASK; + base = 10; + mult1 = mmr_image & UV2_ACK_MASK; ret = mult1 * base; } return ret; @@ -1886,6 +1886,8 @@ static int __init uv_bau_init(void) uv_base_pnode = uv_blade_to_pnode(uvhub); } + enable_timeouts(); + if (init_per_cpu(nuvhubs, uv_base_pnode)) { nobau = 1; return 0; @@ -1896,7 +1898,6 @@ static int __init uv_bau_init(void) if (uv_blade_nr_possible_cpus(uvhub)) init_uvhub(uvhub, vector, uv_base_pnode); - enable_timeouts(); alloc_intr_gate(vector, uv_bau_message_intr1); for_each_possible_blade(uvhub) { From c5d35d399e685acccc85a675e8765c26b2a9813a Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Mon, 16 Jan 2012 15:19:47 -0600 Subject: [PATCH 425/820] x86/UV2: Work around BAU bug This patch implements a workaround for a UV2 hardware bug. The bug is a non-atomic update of a memory-mapped register. When hardware message delivery and software message acknowledge occur simultaneously the pending message acknowledge for the arriving message may be lost. This causes the sender's message status to stay busy. Part of the workaround is to not acknowledge a completed message until it is verified that no other message is actually using the resource that is mistakenly recorded in the completed message. Part of the workaround is to test for long elapsed time in such a busy condition, then handle it by using a spare sending descriptor. The stay-busy condition is eventually timed out by hardware, and then the original sending descriptor can be re-used. Most of that logic change is in keeping track of the current descriptor and the state of the spares. The occurrences of the workaround are added to the BAU statistics. Signed-off-by: Cliff Wickman Link: http://lkml.kernel.org/r/20120116211947.GC5767@sgi.com Cc: Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_bau.h | 13 +- arch/x86/platform/uv/tlb_uv.c | 274 +++++++++++++++++++++++++++---- 2 files changed, 254 insertions(+), 33 deletions(-) diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h index 4a46b27ee9a0..1b82f7e87393 100644 --- a/arch/x86/include/asm/uv/uv_bau.h +++ b/arch/x86/include/asm/uv/uv_bau.h @@ -167,6 +167,7 @@ #define FLUSH_RETRY_TIMEOUT 2 #define FLUSH_GIVEUP 3 #define FLUSH_COMPLETE 4 +#define FLUSH_RETRY_BUSYBUG 5 /* * tuning the action when the numalink network is extremely delayed @@ -463,7 +464,6 @@ struct bau_pq_entry { struct msg_desc { struct bau_pq_entry *msg; int msg_slot; - int swack_slot; struct bau_pq_entry *queue_first; struct bau_pq_entry *queue_last; }; @@ -517,6 +517,9 @@ struct ptc_stats { unsigned long s_retry_messages; /* retry broadcasts */ unsigned long s_bau_reenabled; /* for bau enable/disable */ unsigned long s_bau_disabled; /* for bau enable/disable */ + unsigned long s_uv2_wars; /* uv2 workaround, perm. busy */ + unsigned long s_uv2_wars_hw; /* uv2 workaround, hiwater */ + unsigned long s_uv2_war_waits; /* uv2 workaround, long waits */ /* destination statistics */ unsigned long d_alltlb; /* times all tlb's on this cpu were flushed */ @@ -593,6 +596,8 @@ struct bau_control { short cpus_in_socket; short cpus_in_uvhub; short partition_base_pnode; + short using_desc; /* an index, like uvhub_cpu */ + unsigned int inuse_map; unsigned short message_number; unsigned short uvhub_quiesce; short socket_acknowledge_count[DEST_Q_SIZE]; @@ -610,6 +615,7 @@ struct bau_control { int cong_response_us; int cong_reps; int cong_period; + unsigned long clocks_per_100_usec; cycles_t period_time; long period_requests; struct hub_and_pnode *thp; @@ -670,6 +676,11 @@ static inline void write_mmr_sw_ack(unsigned long mr) uv_write_local_mmr(UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_ALIAS, mr); } +static inline void write_gmmr_sw_ack(int pnode, unsigned long mr) +{ + write_gmmr(pnode, UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_ALIAS, mr); +} + static inline unsigned long read_mmr_sw_ack(void) { return read_lmmr(UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE); diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index c425ff1a9cc3..9010ca715c03 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -157,13 +157,14 @@ static int __init uvhub_to_first_apicid(int uvhub) * clear of the Timeout bit (as well) will free the resource. No reply will * be sent (the hardware will only do one reply per message). */ -static void reply_to_message(struct msg_desc *mdp, struct bau_control *bcp) +static void reply_to_message(struct msg_desc *mdp, struct bau_control *bcp, + int do_acknowledge) { unsigned long dw; struct bau_pq_entry *msg; msg = mdp->msg; - if (!msg->canceled) { + if (!msg->canceled && do_acknowledge) { dw = (msg->swack_vec << UV_SW_ACK_NPENDING) | msg->swack_vec; write_mmr_sw_ack(dw); } @@ -212,8 +213,8 @@ static void bau_process_retry_msg(struct msg_desc *mdp, if (mmr & (msg_res << UV_SW_ACK_NPENDING)) { unsigned long mr; /* - * is the resource timed out? - * make everyone ignore the cancelled message. + * Is the resource timed out? + * Make everyone ignore the cancelled message. */ msg2->canceled = 1; stat->d_canceled++; @@ -231,8 +232,8 @@ static void bau_process_retry_msg(struct msg_desc *mdp, * Do all the things a cpu should do for a TLB shootdown message. * Other cpu's may come here at the same time for this message. */ -static void bau_process_message(struct msg_desc *mdp, - struct bau_control *bcp) +static void bau_process_message(struct msg_desc *mdp, struct bau_control *bcp, + int do_acknowledge) { short socket_ack_count = 0; short *sp; @@ -284,8 +285,9 @@ static void bau_process_message(struct msg_desc *mdp, if (msg_ack_count == bcp->cpus_in_uvhub) { /* * All cpus in uvhub saw it; reply + * (unless we are in the UV2 workaround) */ - reply_to_message(mdp, bcp); + reply_to_message(mdp, bcp, do_acknowledge); } } @@ -491,27 +493,138 @@ static int uv1_wait_completion(struct bau_desc *bau_desc, /* * UV2 has an extra bit of status in the ACTIVATION_STATUS_2 register. */ -static unsigned long uv2_read_status(unsigned long offset, int rshft, int cpu) +static unsigned long uv2_read_status(unsigned long offset, int rshft, int desc) { unsigned long descriptor_status; unsigned long descriptor_status2; descriptor_status = ((read_lmmr(offset) >> rshft) & UV_ACT_STATUS_MASK); - descriptor_status2 = (read_mmr_uv2_status() >> cpu) & 0x1UL; + descriptor_status2 = (read_mmr_uv2_status() >> desc) & 0x1UL; descriptor_status = (descriptor_status << 1) | descriptor_status2; return descriptor_status; } +/* + * Return whether the status of the descriptor that is normally used for this + * cpu (the one indexed by its hub-relative cpu number) is busy. + * The status of the original 32 descriptors is always reflected in the 64 + * bits of UVH_LB_BAU_SB_ACTIVATION_STATUS_0. + * The bit provided by the activation_status_2 register is irrelevant to + * the status if it is only being tested for busy or not busy. + */ +int normal_busy(struct bau_control *bcp) +{ + int cpu = bcp->uvhub_cpu; + int mmr_offset; + int right_shift; + + mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_0; + right_shift = cpu * UV_ACT_STATUS_SIZE; + return (((((read_lmmr(mmr_offset) >> right_shift) & + UV_ACT_STATUS_MASK)) << 1) == UV2H_DESC_BUSY); +} + +/* + * Entered when a bau descriptor has gone into a permanent busy wait because + * of a hardware bug. + * Workaround the bug. + */ +int handle_uv2_busy(struct bau_control *bcp) +{ + int busy_one = bcp->using_desc; + int normal = bcp->uvhub_cpu; + int selected = -1; + int i; + unsigned long descriptor_status; + unsigned long status; + int mmr_offset; + struct bau_desc *bau_desc_old; + struct bau_desc *bau_desc_new; + struct bau_control *hmaster = bcp->uvhub_master; + struct ptc_stats *stat = bcp->statp; + cycles_t ttm; + + stat->s_uv2_wars++; + spin_lock(&hmaster->uvhub_lock); + /* try for the original first */ + if (busy_one != normal) { + if (!normal_busy(bcp)) + selected = normal; + } + if (selected < 0) { + /* can't use the normal, select an alternate */ + mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_1; + descriptor_status = read_lmmr(mmr_offset); + + /* scan available descriptors 32-63 */ + for (i = 0; i < UV_CPUS_PER_AS; i++) { + if ((hmaster->inuse_map & (1 << i)) == 0) { + status = ((descriptor_status >> + (i * UV_ACT_STATUS_SIZE)) & + UV_ACT_STATUS_MASK) << 1; + if (status != UV2H_DESC_BUSY) { + selected = i + UV_CPUS_PER_AS; + break; + } + } + } + } + + if (busy_one != normal) + /* mark the busy alternate as not in-use */ + hmaster->inuse_map &= ~(1 << (busy_one - UV_CPUS_PER_AS)); + + if (selected >= 0) { + /* switch to the selected descriptor */ + if (selected != normal) { + /* set the selected alternate as in-use */ + hmaster->inuse_map |= + (1 << (selected - UV_CPUS_PER_AS)); + if (selected > stat->s_uv2_wars_hw) + stat->s_uv2_wars_hw = selected; + } + bau_desc_old = bcp->descriptor_base; + bau_desc_old += (ITEMS_PER_DESC * busy_one); + bcp->using_desc = selected; + bau_desc_new = bcp->descriptor_base; + bau_desc_new += (ITEMS_PER_DESC * selected); + *bau_desc_new = *bau_desc_old; + } else { + /* + * All are busy. Wait for the normal one for this cpu to + * free up. + */ + stat->s_uv2_war_waits++; + spin_unlock(&hmaster->uvhub_lock); + ttm = get_cycles(); + do { + cpu_relax(); + } while (normal_busy(bcp)); + spin_lock(&hmaster->uvhub_lock); + /* switch to the original descriptor */ + bcp->using_desc = normal; + bau_desc_old = bcp->descriptor_base; + bau_desc_old += (ITEMS_PER_DESC * bcp->using_desc); + bcp->using_desc = (ITEMS_PER_DESC * normal); + bau_desc_new = bcp->descriptor_base; + bau_desc_new += (ITEMS_PER_DESC * normal); + *bau_desc_new = *bau_desc_old; /* copy the entire descriptor */ + } + spin_unlock(&hmaster->uvhub_lock); + return FLUSH_RETRY_BUSYBUG; +} + static int uv2_wait_completion(struct bau_desc *bau_desc, unsigned long mmr_offset, int right_shift, struct bau_control *bcp, long try) { unsigned long descriptor_stat; cycles_t ttm; - int cpu = bcp->uvhub_cpu; + int desc = bcp->using_desc; + long busy_reps = 0; struct ptc_stats *stat = bcp->statp; - descriptor_stat = uv2_read_status(mmr_offset, right_shift, cpu); + descriptor_stat = uv2_read_status(mmr_offset, right_shift, desc); /* spin on the status MMR, waiting for it to go idle */ while (descriptor_stat != UV2H_DESC_IDLE) { @@ -542,12 +655,23 @@ static int uv2_wait_completion(struct bau_desc *bau_desc, bcp->conseccompletes = 0; return FLUSH_RETRY_TIMEOUT; } else { + busy_reps++; + if (busy_reps > 1000000) { + /* not to hammer on the clock */ + busy_reps = 0; + ttm = get_cycles(); + if ((ttm - bcp->send_message) > + (bcp->clocks_per_100_usec)) { + return handle_uv2_busy(bcp); + } + } /* * descriptor_stat is still BUSY */ cpu_relax(); } - descriptor_stat = uv2_read_status(mmr_offset, right_shift, cpu); + descriptor_stat = uv2_read_status(mmr_offset, right_shift, + desc); } bcp->conseccompletes++; return FLUSH_COMPLETE; @@ -563,14 +687,14 @@ static int wait_completion(struct bau_desc *bau_desc, { int right_shift; unsigned long mmr_offset; - int cpu = bcp->uvhub_cpu; + int desc = bcp->using_desc; - if (cpu < UV_CPUS_PER_AS) { + if (desc < UV_CPUS_PER_AS) { mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_0; - right_shift = cpu * UV_ACT_STATUS_SIZE; + right_shift = desc * UV_ACT_STATUS_SIZE; } else { mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_1; - right_shift = ((cpu - UV_CPUS_PER_AS) * UV_ACT_STATUS_SIZE); + right_shift = ((desc - UV_CPUS_PER_AS) * UV_ACT_STATUS_SIZE); } if (bcp->uvhub_version == 1) @@ -752,8 +876,7 @@ static void handle_cmplt(int completion_status, struct bau_desc *bau_desc, * Returns 1 if it gives up entirely and the original cpu mask is to be * returned to the kernel. */ -int uv_flush_send_and_wait(struct bau_desc *bau_desc, - struct cpumask *flush_mask, struct bau_control *bcp) +int uv_flush_send_and_wait(struct cpumask *flush_mask, struct bau_control *bcp) { int seq_number = 0; int completion_stat = 0; @@ -766,20 +889,24 @@ int uv_flush_send_and_wait(struct bau_desc *bau_desc, struct bau_control *hmaster = bcp->uvhub_master; struct uv1_bau_msg_header *uv1_hdr = NULL; struct uv2_bau_msg_header *uv2_hdr = NULL; + struct bau_desc *bau_desc; - if (bcp->uvhub_version == 1) { - uv1 = 1; + if (bcp->uvhub_version == 1) uv1_throttle(hmaster, stat); - uv1_hdr = &bau_desc->header.uv1_hdr; - } else - uv2_hdr = &bau_desc->header.uv2_hdr; while (hmaster->uvhub_quiesce) cpu_relax(); time1 = get_cycles(); do { - if (try == 0) { + bau_desc = bcp->descriptor_base; + bau_desc += (ITEMS_PER_DESC * bcp->using_desc); + if (bcp->uvhub_version == 1) { + uv1 = 1; + uv1_hdr = &bau_desc->header.uv1_hdr; + } else + uv2_hdr = &bau_desc->header.uv2_hdr; + if ((try == 0) || (completion_stat == FLUSH_RETRY_BUSYBUG)) { if (uv1) uv1_hdr->msg_type = MSG_REGULAR; else @@ -797,13 +924,14 @@ int uv_flush_send_and_wait(struct bau_desc *bau_desc, uv1_hdr->sequence = seq_number; else uv2_hdr->sequence = seq_number; - index = (1UL << AS_PUSH_SHIFT) | bcp->uvhub_cpu; + index = (1UL << AS_PUSH_SHIFT) | bcp->using_desc; bcp->send_message = get_cycles(); write_mmr_activation(index); try++; completion_stat = wait_completion(bau_desc, bcp, try); + /* UV2: wait_completion() may change the bcp->using_desc */ handle_cmplt(completion_stat, bau_desc, bcp, hmaster, stat); @@ -814,6 +942,7 @@ int uv_flush_send_and_wait(struct bau_desc *bau_desc, } cpu_relax(); } while ((completion_stat == FLUSH_RETRY_PLUGGED) || + (completion_stat == FLUSH_RETRY_BUSYBUG) || (completion_stat == FLUSH_RETRY_TIMEOUT)); time2 = get_cycles(); @@ -828,6 +957,7 @@ int uv_flush_send_and_wait(struct bau_desc *bau_desc, record_send_stats(time1, time2, bcp, stat, completion_stat, try); if (completion_stat == FLUSH_GIVEUP) + /* FLUSH_GIVEUP will fall back to using IPI's for tlb flush */ return 1; return 0; } @@ -983,7 +1113,7 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask, stat->s_ntargself++; bau_desc = bcp->descriptor_base; - bau_desc += (ITEMS_PER_DESC * bcp->uvhub_cpu); + bau_desc += (ITEMS_PER_DESC * bcp->using_desc); bau_uvhubs_clear(&bau_desc->distribution, UV_DISTRIBUTION_SIZE); if (set_distrib_bits(flush_mask, bcp, bau_desc, &locals, &remotes)) return NULL; @@ -996,12 +1126,85 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask, * uv_flush_send_and_wait returns 0 if all cpu's were messaged, * or 1 if it gave up and the original cpumask should be returned. */ - if (!uv_flush_send_and_wait(bau_desc, flush_mask, bcp)) + if (!uv_flush_send_and_wait(flush_mask, bcp)) return NULL; else return cpumask; } +/* + * Search the message queue for any 'other' message with the same software + * acknowledge resource bit vector. + */ +struct bau_pq_entry *find_another_by_swack(struct bau_pq_entry *msg, + struct bau_control *bcp, unsigned char swack_vec) +{ + struct bau_pq_entry *msg_next = msg + 1; + + if (msg_next > bcp->queue_last) + msg_next = bcp->queue_first; + while ((msg_next->swack_vec != 0) && (msg_next != msg)) { + if (msg_next->swack_vec == swack_vec) + return msg_next; + msg_next++; + if (msg_next > bcp->queue_last) + msg_next = bcp->queue_first; + } + return NULL; +} + +/* + * UV2 needs to work around a bug in which an arriving message has not + * set a bit in the UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE register. + * Such a message must be ignored. + */ +void process_uv2_message(struct msg_desc *mdp, struct bau_control *bcp) +{ + unsigned long mmr_image; + unsigned char swack_vec; + struct bau_pq_entry *msg = mdp->msg; + struct bau_pq_entry *other_msg; + + mmr_image = read_mmr_sw_ack(); + swack_vec = msg->swack_vec; + + if ((swack_vec & mmr_image) == 0) { + /* + * This message was assigned a swack resource, but no + * reserved acknowlegment is pending. + * The bug has prevented this message from setting the MMR. + * And no other message has used the same sw_ack resource. + * Do the requested shootdown but do not reply to the msg. + * (the 0 means make no acknowledge) + */ + bau_process_message(mdp, bcp, 0); + return; + } + + /* + * Some message has set the MMR 'pending' bit; it might have been + * another message. Look for that message. + */ + other_msg = find_another_by_swack(msg, bcp, msg->swack_vec); + if (other_msg) { + /* There is another. Do not ack the current one. */ + bau_process_message(mdp, bcp, 0); + /* + * Let the natural processing of that message acknowledge + * it. Don't get the processing of sw_ack's out of order. + */ + return; + } + + /* + * There is no other message using this sw_ack, so it is safe to + * acknowledge it. + */ + bau_process_message(mdp, bcp, 1); + + return; +} + /* * The BAU message interrupt comes here. (registered by set_intr_gate) * See entry_64.S @@ -1038,9 +1241,11 @@ void uv_bau_message_interrupt(struct pt_regs *regs) count++; msgdesc.msg_slot = msg - msgdesc.queue_first; - msgdesc.swack_slot = ffs(msg->swack_vec) - 1; msgdesc.msg = msg; - bau_process_message(&msgdesc, bcp); + if (bcp->uvhub_version == 2) + process_uv2_message(&msgdesc, bcp); + else + bau_process_message(&msgdesc, bcp, 1); msg++; if (msg > msgdesc.queue_last) @@ -1158,7 +1363,7 @@ static int ptc_seq_show(struct seq_file *file, void *data) seq_printf(file, "all one mult none retry canc nocan reset rcan "); seq_printf(file, - "disable enable\n"); + "disable enable wars warshw warwaits\n"); } if (cpu < num_possible_cpus() && cpu_online(cpu)) { stat = &per_cpu(ptcstats, cpu); @@ -1189,8 +1394,10 @@ static int ptc_seq_show(struct seq_file *file, void *data) stat->d_nomsg, stat->d_retries, stat->d_canceled, stat->d_nocanceled, stat->d_resets, stat->d_rcanceled); - seq_printf(file, "%ld %ld\n", - stat->s_bau_disabled, stat->s_bau_reenabled); + seq_printf(file, "%ld %ld %ld %ld %ld\n", + stat->s_bau_disabled, stat->s_bau_reenabled, + stat->s_uv2_wars, stat->s_uv2_wars_hw, + stat->s_uv2_war_waits); } return 0; } @@ -1564,6 +1771,7 @@ static void pq_init(int node, int pnode) write_mmr_payload_first(pnode, pn_first); write_mmr_payload_tail(pnode, first); write_mmr_payload_last(pnode, last); + write_gmmr_sw_ack(pnode, 0xffffUL); /* in effect, all msg_type's are set to MSG_NOOP */ memset(pqp, 0, sizeof(struct bau_pq_entry) * DEST_Q_SIZE); @@ -1651,6 +1859,7 @@ static void __init init_per_cpu_tunables(void) bcp->cong_response_us = congested_respns_us; bcp->cong_reps = congested_reps; bcp->cong_period = congested_period; + bcp->clocks_per_100_usec = usec_2_cycles(100); } } @@ -1771,6 +1980,7 @@ static int scan_sock(struct socket_desc *sdp, struct uvhub_desc *bdp, } bcp->uvhub_master = *hmasterp; bcp->uvhub_cpu = uv_cpu_hub_info(cpu)->blade_processor_id; + bcp->using_desc = bcp->uvhub_cpu; if (bcp->uvhub_cpu >= MAX_CPUS_PER_UVHUB) { printk(KERN_EMERG "%d cpus per uvhub invalid\n", bcp->uvhub_cpu); From 478c6e529e7bd7c6ef8994c55bd252c287c35893 Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Mon, 16 Jan 2012 15:20:50 -0600 Subject: [PATCH 426/820] x86/UV2: Remove stale no-resources test for UV2 BAU This patch removes an unnecessary test for a no-destination-resources-available condition that looks like a destination timeout in UV1, but is separately distinguishable in UV2. Signed-off-by: Cliff Wickman Link: http://lkml.kernel.org/r/20120116212050.GD5767@sgi.com Signed-off-by: Ingo Molnar --- arch/x86/platform/uv/tlb_uv.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index 9010ca715c03..affea509c174 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -642,16 +642,6 @@ static int uv2_wait_completion(struct bau_desc *bau_desc, } else if (descriptor_stat == UV2H_DESC_DEST_TIMEOUT) { stat->s_dtimeout++; ttm = get_cycles(); - /* - * Our retries may be blocked by all destination - * swack resources being consumed, and a timeout - * pending. In that case hardware returns the - * ERROR that looks like a destination timeout. - */ - if (cycles_2_us(ttm - bcp->send_message) < timeout_us) { - bcp->conseccompletes = 0; - return FLUSH_RETRY_PLUGGED; - } bcp->conseccompletes = 0; return FLUSH_RETRY_TIMEOUT; } else { From 88ed9dd7f63c3ae71c1984d99ee2dced0b386dea Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Mon, 16 Jan 2012 15:21:46 -0600 Subject: [PATCH 427/820] x86/UV2: Ack BAU interrupt earlier This patch moves the ack of the BAU interrupt to the beginning of the interrupt handler so that there is less possibility of a lost interrupt and slower response to a shootdown message. Signed-off-by: Cliff Wickman Link: http://lkml.kernel.org/r/20120116212146.GE5767@sgi.com Signed-off-by: Ingo Molnar --- arch/x86/platform/uv/tlb_uv.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index affea509c174..4686bf1e56ec 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -1218,6 +1218,7 @@ void uv_bau_message_interrupt(struct pt_regs *regs) struct ptc_stats *stat; struct msg_desc msgdesc; + ack_APIC_irq(); time_start = get_cycles(); bcp = &per_cpu(bau_control, smp_processor_id()); @@ -1247,8 +1248,6 @@ void uv_bau_message_interrupt(struct pt_regs *regs) stat->d_nomsg++; else if (count > 1) stat->d_multmsg++; - - ack_APIC_irq(); } /* From b54bd9be35f4084edb3eb9ee054a43f722a67483 Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Mon, 16 Jan 2012 15:22:38 -0600 Subject: [PATCH 428/820] x86/UV2: Add accounting for BAU strong nacks This patch adds separate accounting of UV2 message "strong nack's" in the BAU statistics. Signed-off-by: Cliff Wickman Link: http://lkml.kernel.org/r/20120116212238.GF5767@sgi.com Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_bau.h | 1 + arch/x86/platform/uv/tlb_uv.c | 12 +++++++----- 2 files changed, 8 insertions(+), 5 deletions(-) diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h index 1b82f7e87393..becf47b81735 100644 --- a/arch/x86/include/asm/uv/uv_bau.h +++ b/arch/x86/include/asm/uv/uv_bau.h @@ -483,6 +483,7 @@ struct ptc_stats { requests */ unsigned long s_stimeout; /* source side timeouts */ unsigned long s_dtimeout; /* destination side timeouts */ + unsigned long s_strongnacks; /* number of strong nack's */ unsigned long s_time; /* time spent in sending side */ unsigned long s_retriesok; /* successful retries */ unsigned long s_ntargcpu; /* total number of cpu's diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index 4686bf1e56ec..9be4cff00a2d 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -635,13 +635,15 @@ static int uv2_wait_completion(struct bau_desc *bau_desc, * our message and its state will stay IDLE. */ if ((descriptor_stat == UV2H_DESC_SOURCE_TIMEOUT) || - (descriptor_stat == UV2H_DESC_DEST_STRONG_NACK) || (descriptor_stat == UV2H_DESC_DEST_PUT_ERR)) { stat->s_stimeout++; return FLUSH_GIVEUP; + } else if (descriptor_stat == UV2H_DESC_DEST_STRONG_NACK) { + stat->s_strongnacks++; + bcp->conseccompletes = 0; + return FLUSH_GIVEUP; } else if (descriptor_stat == UV2H_DESC_DEST_TIMEOUT) { stat->s_dtimeout++; - ttm = get_cycles(); bcp->conseccompletes = 0; return FLUSH_RETRY_TIMEOUT; } else { @@ -1346,7 +1348,7 @@ static int ptc_seq_show(struct seq_file *file, void *data) seq_printf(file, "remotehub numuvhubs numuvhubs16 numuvhubs8 "); seq_printf(file, - "numuvhubs4 numuvhubs2 numuvhubs1 dto retries rok "); + "numuvhubs4 numuvhubs2 numuvhubs1 dto snacks retries rok "); seq_printf(file, "resetp resett giveup sto bz throt swack recv rtime "); seq_printf(file, @@ -1364,10 +1366,10 @@ static int ptc_seq_show(struct seq_file *file, void *data) stat->s_ntargremotes, stat->s_ntargcpu, stat->s_ntarglocaluvhub, stat->s_ntargremoteuvhub, stat->s_ntarguvhub, stat->s_ntarguvhub16); - seq_printf(file, "%ld %ld %ld %ld %ld ", + seq_printf(file, "%ld %ld %ld %ld %ld %ld ", stat->s_ntarguvhub8, stat->s_ntarguvhub4, stat->s_ntarguvhub2, stat->s_ntarguvhub1, - stat->s_dtimeout); + stat->s_dtimeout, stat->s_strongnacks); seq_printf(file, "%ld %ld %ld %ld %ld %ld %ld %ld ", stat->s_retry_messages, stat->s_retriesok, stat->s_resets_plug, stat->s_resets_timeout, From da4d8b287abe783d30e968155614531a0937d090 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Tue, 29 Nov 2011 22:13:35 +0800 Subject: [PATCH 429/820] ACPICA: Put back the call to acpi_os_validate_address The call to acpi_os_validate_address in acpi_ds_get_region_arguments was removed by mistake in commit 9ad19ac(ACPICA: Split large dsopcode and dsload.c files). Put it back. Cc: stable@vger.kernel.org # 2.6.38+ Reported-and-bisected-by: Luca Tettamanti Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/dsargs.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/drivers/acpi/acpica/dsargs.c b/drivers/acpi/acpica/dsargs.c index 8c7b99728aa2..42163d8db50b 100644 --- a/drivers/acpi/acpica/dsargs.c +++ b/drivers/acpi/acpica/dsargs.c @@ -387,5 +387,29 @@ acpi_status acpi_ds_get_region_arguments(union acpi_operand_object *obj_desc) status = acpi_ds_execute_arguments(node, node->parent, extra_desc->extra.aml_length, extra_desc->extra.aml_start); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + /* Validate the region address/length via the host OS */ + + status = acpi_os_validate_address(obj_desc->region.space_id, + obj_desc->region.address, + (acpi_size) obj_desc->region.length, + acpi_ut_get_node_name(node)); + + if (ACPI_FAILURE(status)) { + /* + * Invalid address/length. We will emit an error message and mark + * the region as invalid, so that it will cause an additional error if + * it is ever used. Then return AE_OK. + */ + ACPI_EXCEPTION((AE_INFO, status, + "During address validation of OpRegion [%4.4s]", + node->name.ascii)); + obj_desc->common.flags |= AOPOBJ_INVALID; + status = AE_OK; + } + return_ACPI_STATUS(status); } From 42f5284306465dc263ce227b0c5bd2c7dcd43cdf Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 28 Nov 2011 09:43:52 +0800 Subject: [PATCH 430/820] ACPICA: Update for GCC 4 Fixes several issues with GCC 4.6 related to the new checks for unused variables. http://www.acpica.org/bugzilla/show_bug.cgi?id=935 Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/utmutex.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/drivers/acpi/acpica/utmutex.c b/drivers/acpi/acpica/utmutex.c index 7d797e2baecd..302c74e26f95 100644 --- a/drivers/acpi/acpica/utmutex.c +++ b/drivers/acpi/acpica/utmutex.c @@ -293,14 +293,10 @@ acpi_status acpi_ut_acquire_mutex(acpi_mutex_handle mutex_id) acpi_status acpi_ut_release_mutex(acpi_mutex_handle mutex_id) { - acpi_thread_id this_thread_id; - ACPI_FUNCTION_NAME(ut_release_mutex); - this_thread_id = acpi_os_get_thread_id(); - ACPI_DEBUG_PRINT((ACPI_DB_MUTEX, "Thread %u releasing Mutex [%s]\n", - (u32)this_thread_id, + (u32)acpi_os_get_thread_id(), acpi_ut_get_mutex_name(mutex_id))); if (mutex_id > ACPI_MAX_MUTEX) { @@ -329,7 +325,8 @@ acpi_status acpi_ut_release_mutex(acpi_mutex_handle mutex_id) * the ACPI subsystem code. */ for (i = mutex_id; i < ACPI_NUM_MUTEX; i++) { - if (acpi_gbl_mutex_info[i].thread_id == this_thread_id) { + if (acpi_gbl_mutex_info[i].thread_id == + acpi_os_get_thread_id()) { if (i == mutex_id) { continue; } From 8931d9ea78848b073bf299594f148b83abde4a5e Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Mon, 28 Nov 2011 09:46:02 +0800 Subject: [PATCH 431/820] ACPICA: Fix to allow region arguments to reference other scopes Allow referenced objects to be in a different scope. http://www.acpica.org/bugzilla/show_bug.cgi?id=937 http://marc.info/?l=linux-acpi&m=131636632718222&w=2 ACPI Error: [RAMB] Namespace lookup failure, AE_NOT_FOUND (20110112/psargs-359) ACPI Exception: AE_NOT_FOUND, Could not execute arguments for [RAMW] (Region) (20110112/nsinit-349) Scope (_SB) { Name (RAMB, 0xDF5A1018) OperationRegion (\RAMW, SystemMemory, RAMB, 0x00010000) } For above ASL code, we need to save scope node(\_SB) to lookup the argument node(\_SB.RAMB). Reported-by: Jim Green Signed-off-by: Lin Ming Signed-off-by: Bob Moore Signed-off-by: Len Brown --- drivers/acpi/acpica/acobject.h | 1 + drivers/acpi/acpica/dsargs.c | 2 +- drivers/acpi/acpica/excreate.c | 6 ++++++ 3 files changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/acpi/acpica/acobject.h b/drivers/acpi/acpica/acobject.h index 1055769f2f01..6d276c20b57b 100644 --- a/drivers/acpi/acpica/acobject.h +++ b/drivers/acpi/acpica/acobject.h @@ -358,6 +358,7 @@ typedef enum { */ struct acpi_object_extra { ACPI_OBJECT_COMMON_HEADER struct acpi_namespace_node *method_REG; /* _REG method for this region (if any) */ + struct acpi_namespace_node *scope_node; void *region_context; /* Region-specific data */ u8 *aml_start; u32 aml_length; diff --git a/drivers/acpi/acpica/dsargs.c b/drivers/acpi/acpica/dsargs.c index 42163d8db50b..d69e4a53175b 100644 --- a/drivers/acpi/acpica/dsargs.c +++ b/drivers/acpi/acpica/dsargs.c @@ -384,7 +384,7 @@ acpi_status acpi_ds_get_region_arguments(union acpi_operand_object *obj_desc) /* Execute the argument AML */ - status = acpi_ds_execute_arguments(node, node->parent, + status = acpi_ds_execute_arguments(node, extra_desc->extra.scope_node, extra_desc->extra.aml_length, extra_desc->extra.aml_start); if (ACPI_FAILURE(status)) { diff --git a/drivers/acpi/acpica/excreate.c b/drivers/acpi/acpica/excreate.c index 110711afada8..8a06dc523af7 100644 --- a/drivers/acpi/acpica/excreate.c +++ b/drivers/acpi/acpica/excreate.c @@ -330,6 +330,12 @@ acpi_ex_create_region(u8 * aml_start, region_obj2 = obj_desc->common.next_object; region_obj2->extra.aml_start = aml_start; region_obj2->extra.aml_length = aml_length; + if (walk_state->scope_info) { + region_obj2->extra.scope_node = + walk_state->scope_info->scope.node; + } else { + region_obj2->extra.scope_node = node; + } /* Init the region from the operands */ From 8519bc9f1e6db43a9b95b70ef0c7b61cb36d58e4 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 28 Nov 2011 10:08:03 +0800 Subject: [PATCH 432/820] ACPICA: Add error msg for unsupported I/O requests (not 8/16/32 bit length) Found during ACPICA debugging. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/hwvalid.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/acpi/acpica/hwvalid.c b/drivers/acpi/acpica/hwvalid.c index 5f1605874655..f8fbea947013 100644 --- a/drivers/acpi/acpica/hwvalid.c +++ b/drivers/acpi/acpica/hwvalid.c @@ -134,6 +134,8 @@ acpi_hw_validate_io_request(acpi_io_address address, u32 bit_width) /* Supported widths are 8/16/32 */ if ((bit_width != 8) && (bit_width != 16) && (bit_width != 32)) { + ACPI_ERROR((AE_INFO, + "Bad BitWidth parameter: %8.8X", bit_width)); return AE_BAD_PARAMETER; } From 46dfb09c024a1a92ead63b40f6993dcb5eb7d153 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 28 Nov 2011 10:10:10 +0800 Subject: [PATCH 433/820] ACPICA: Add error msg for Index/Bank field registers out-of-range Supplement the exception code with an actual message. Found during ACPICA debugging. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/exfldio.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index f915a7f3f921..19df8ce66ee7 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -316,6 +316,7 @@ acpi_ex_access_region(union acpi_operand_object *obj_desc, static u8 acpi_ex_register_overflow(union acpi_operand_object *obj_desc, u64 value) { + ACPI_FUNCTION_NAME(ex_register_overflow); if (obj_desc->common_field.bit_length >= ACPI_INTEGER_BIT_SIZE) { /* @@ -330,6 +331,11 @@ acpi_ex_register_overflow(union acpi_operand_object *obj_desc, u64 value) * The Value is larger than the maximum value that can fit into * the register. */ + ACPI_ERROR((AE_INFO, + "Index value 0x%8.8X%8.8X overflows field width 0x%X", + ACPI_FORMAT_UINT64(value), + obj_desc->common_field.bit_length)); + return (TRUE); } From ec4636669bf9c6ff157121ab42709650a9e0cc2a Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 30 Nov 2011 09:35:05 +0800 Subject: [PATCH 434/820] ACPICA: Do not abort table load on invalid space ID Ignore an invalid space ID during a table load. Instead, detect it if a control method attempts access - then abort the method. http://www.acpica.org/bugzilla/show_bug.cgi?id=925 Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acinterp.h | 2 ++ drivers/acpi/acpica/excreate.c | 23 +++++++++++++---------- drivers/acpi/acpica/exfldio.c | 16 ++++++++++++++-- drivers/acpi/acpica/exutils.c | 25 +++++++++++++++++++++++++ 4 files changed, 54 insertions(+), 12 deletions(-) diff --git a/drivers/acpi/acpica/acinterp.h b/drivers/acpi/acpica/acinterp.h index 3731e1c34b83..4ee9058aba53 100644 --- a/drivers/acpi/acpica/acinterp.h +++ b/drivers/acpi/acpica/acinterp.h @@ -468,6 +468,8 @@ void acpi_ex_eisa_id_to_string(char *dest, u64 compressed_id); void acpi_ex_integer_to_string(char *dest, u64 value); +u8 acpi_is_valid_space_id(u8 space_id); + /* * exregion - default op_region handlers */ diff --git a/drivers/acpi/acpica/excreate.c b/drivers/acpi/acpica/excreate.c index 8a06dc523af7..c66562b4311b 100644 --- a/drivers/acpi/acpica/excreate.c +++ b/drivers/acpi/acpica/excreate.c @@ -267,7 +267,7 @@ acpi_status acpi_ex_create_mutex(struct acpi_walk_state *walk_state) * * PARAMETERS: aml_start - Pointer to the region declaration AML * aml_length - Max length of the declaration AML - * region_space - space_iD for the region + * space_id - Address space ID for the region * walk_state - Current state * * RETURN: Status @@ -279,7 +279,7 @@ acpi_status acpi_ex_create_mutex(struct acpi_walk_state *walk_state) acpi_status acpi_ex_create_region(u8 * aml_start, u32 aml_length, - u8 region_space, struct acpi_walk_state *walk_state) + u8 space_id, struct acpi_walk_state *walk_state) { acpi_status status; union acpi_operand_object *obj_desc; @@ -304,16 +304,19 @@ acpi_ex_create_region(u8 * aml_start, * Space ID must be one of the predefined IDs, or in the user-defined * range */ - if ((region_space >= ACPI_NUM_PREDEFINED_REGIONS) && - (region_space < ACPI_USER_REGION_BEGIN) && - (region_space != ACPI_ADR_SPACE_DATA_TABLE)) { - ACPI_ERROR((AE_INFO, "Invalid AddressSpace type 0x%X", - region_space)); - return_ACPI_STATUS(AE_AML_INVALID_SPACE_ID); + if (!acpi_is_valid_space_id(space_id)) { + /* + * Print an error message, but continue. We don't want to abort + * a table load for this exception. Instead, if the region is + * actually used at runtime, abort the executing method. + */ + ACPI_ERROR((AE_INFO, + "Invalid/unknown Address Space ID: 0x%2.2X", + space_id)); } ACPI_DEBUG_PRINT((ACPI_DB_LOAD, "Region Type - %s (0x%X)\n", - acpi_ut_get_region_name(region_space), region_space)); + acpi_ut_get_region_name(space_id), space_id)); /* Create the region descriptor */ @@ -339,7 +342,7 @@ acpi_ex_create_region(u8 * aml_start, /* Init the region from the operands */ - obj_desc->region.space_id = region_space; + obj_desc->region.space_id = space_id; obj_desc->region.address = 0; obj_desc->region.length = 0; obj_desc->region.node = node; diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index 19df8ce66ee7..2a524fc1e851 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -86,6 +86,7 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc, { acpi_status status = AE_OK; union acpi_operand_object *rgn_desc; + u8 space_id; ACPI_FUNCTION_TRACE_U32(ex_setup_region, field_datum_byte_offset); @@ -101,6 +102,17 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc, return_ACPI_STATUS(AE_AML_OPERAND_TYPE); } + space_id = rgn_desc->region.space_id; + + /* Validate the Space ID */ + + if (!acpi_is_valid_space_id(space_id)) { + ACPI_ERROR((AE_INFO, + "Invalid/unknown Address Space ID: 0x%2.2X", + space_id)); + return_ACPI_STATUS(AE_AML_INVALID_SPACE_ID); + } + /* * If the Region Address and Length have not been previously evaluated, * evaluate them now and save the results. @@ -122,8 +134,8 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc, * Exit now for SMBus or IPMI address space, it has a non-linear * address space and the request cannot be directly validated */ - if (rgn_desc->region.space_id == ACPI_ADR_SPACE_SMBUS || - rgn_desc->region.space_id == ACPI_ADR_SPACE_IPMI) { + if (space_id == ACPI_ADR_SPACE_SMBUS || + space_id == ACPI_ADR_SPACE_IPMI) { /* SMBus or IPMI has a non-linear address space */ diff --git a/drivers/acpi/acpica/exutils.c b/drivers/acpi/acpica/exutils.c index 8ad93146dd32..3993aa5084b6 100644 --- a/drivers/acpi/acpica/exutils.c +++ b/drivers/acpi/acpica/exutils.c @@ -435,4 +435,29 @@ void acpi_ex_integer_to_string(char *out_string, u64 value) } } +/******************************************************************************* + * + * FUNCTION: acpi_is_valid_space_id + * + * PARAMETERS: space_id - ID to be validated + * + * RETURN: TRUE if valid/supported ID. + * + * DESCRIPTION: Validate an operation region space_iD. + * + ******************************************************************************/ + +u8 acpi_is_valid_space_id(u8 space_id) +{ + + if ((space_id >= ACPI_NUM_PREDEFINED_REGIONS) && + (space_id < ACPI_USER_REGION_BEGIN) && + (space_id != ACPI_ADR_SPACE_DATA_TABLE) && + (space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)) { + return (FALSE); + } + + return (TRUE); +} + #endif From 2355e10f07b2c550c671f98b68aedf323c299aeb Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 10:42:09 +0800 Subject: [PATCH 435/820] ACPI 5.0: Basic support for FADT version 5 Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/tbfadt.c | 4 ++-- include/acpi/actbl.h | 22 ++++++++++++++++------ 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/drivers/acpi/acpica/tbfadt.c b/drivers/acpi/acpica/tbfadt.c index 6f5588e62c0a..536fd2bc7a47 100644 --- a/drivers/acpi/acpica/tbfadt.c +++ b/drivers/acpi/acpica/tbfadt.c @@ -277,12 +277,12 @@ void acpi_tb_create_local_fadt(struct acpi_table_header *table, u32 length) { /* * Check if the FADT is larger than the largest table that we expect - * (the ACPI 2.0/3.0 version). If so, truncate the table, and issue + * (the ACPI 5.0 version). If so, truncate the table, and issue * a warning. */ if (length > sizeof(struct acpi_table_fadt)) { ACPI_WARNING((AE_INFO, - "FADT (revision %u) is longer than ACPI 2.0 version, " + "FADT (revision %u) is longer than ACPI 5.0 version, " "truncating length %u to %u", table->revision, length, (u32)sizeof(struct acpi_table_fadt))); diff --git a/include/acpi/actbl.h b/include/acpi/actbl.h index f1380287ed4d..7d0e051b4863 100644 --- a/include/acpi/actbl.h +++ b/include/acpi/actbl.h @@ -255,6 +255,8 @@ struct acpi_table_fadt { struct acpi_generic_address xpm_timer_block; /* 64-bit Extended Power Mgt Timer Ctrl Reg Blk address */ struct acpi_generic_address xgpe0_block; /* 64-bit Extended General Purpose Event 0 Reg Blk address */ struct acpi_generic_address xgpe1_block; /* 64-bit Extended General Purpose Event 1 Reg Blk address */ + struct acpi_generic_address sleep_control; /* 64-bit Sleep Control register */ + struct acpi_generic_address sleep_status; /* 64-bit Sleep Status register */ }; /* Masks for FADT Boot Architecture Flags (boot_flags) */ @@ -264,6 +266,7 @@ struct acpi_table_fadt { #define ACPI_FADT_NO_VGA (1<<2) /* 02: [V4] It is not safe to probe for VGA hardware */ #define ACPI_FADT_NO_MSI (1<<3) /* 03: [V4] Message Signaled Interrupts (MSI) must not be enabled */ #define ACPI_FADT_NO_ASPM (1<<4) /* 04: [V4] PCIe ASPM control must not be enabled */ +#define ACPI_FADT_NO_CMOS_RTC (1<<5) /* 05: [V5] No CMOS real-time clock present */ #define FADT2_REVISION_ID 3 @@ -289,6 +292,8 @@ struct acpi_table_fadt { #define ACPI_FADT_REMOTE_POWER_ON (1<<17) /* 17: [V4] System is compatible with remote power on (ACPI 3.0) */ #define ACPI_FADT_APIC_CLUSTER (1<<18) /* 18: [V4] All local APICs must use cluster model (ACPI 3.0) */ #define ACPI_FADT_APIC_PHYSICAL (1<<19) /* 19: [V4] All local x_aPICs must use physical dest mode (ACPI 3.0) */ +#define ACPI_FADT_HW_REDUCED (1<<20) /* 20: [V5] ACPI hardware is not implemented (ACPI 5.0) */ +#define ACPI_FADT_LOW_POWER_S0 (1<<21) /* 21: [V5] S0 power savings are equal or better than S3 (ACPI 5.0) */ /* Values for preferred_profile (Preferred Power Management Profiles) */ @@ -299,14 +304,16 @@ enum acpi_prefered_pm_profiles { PM_WORKSTATION = 3, PM_ENTERPRISE_SERVER = 4, PM_SOHO_SERVER = 5, - PM_APPLIANCE_PC = 6 + PM_APPLIANCE_PC = 6, + PM_PERFORMANCE_SERVER = 7, + PM_TABLET = 8 }; /* Reset to default packing */ #pragma pack() -#define ACPI_FADT_OFFSET(f) (u8) ACPI_OFFSET (struct acpi_table_fadt, f) +#define ACPI_FADT_OFFSET(f) (u16) ACPI_OFFSET (struct acpi_table_fadt, f) /* * Internal table-related structures @@ -351,12 +358,15 @@ struct acpi_table_desc { * FADT is the bottom line as to what the version really is. * * For reference, the values below are as follows: - * FADT V1 size: 0x74 - * FADT V2 size: 0x84 - * FADT V3+ size: 0xF4 + * FADT V1 size: 0x074 + * FADT V2 size: 0x084 + * FADT V3 size: 0x0F4 + * FADT V4 size: 0x0F4 + * FADT V5 size: 0x10C */ #define ACPI_FADT_V1_SIZE (u32) (ACPI_FADT_OFFSET (flags) + 4) #define ACPI_FADT_V2_SIZE (u32) (ACPI_FADT_OFFSET (reserved4[0]) + 3) -#define ACPI_FADT_V3_SIZE (u32) (sizeof (struct acpi_table_fadt)) +#define ACPI_FADT_V3_SIZE (u32) (ACPI_FADT_OFFSET (sleep_control)) +#define ACPI_FADT_V5_SIZE (u32) (sizeof (struct acpi_table_fadt)) #endif /* __ACTBL_H__ */ From 22e5b40ab21fcac21db0ff25fbb844ffecc73a4a Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 10:57:28 +0800 Subject: [PATCH 436/820] ACPI 5.0: Implement hardware-reduced option If HW-reduced flag is set in the FADT, do not attempt to access or initialize any ACPI hardware, including SCI and global lock. No FACS will be present. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acglobal.h | 13 ++++++++++++- drivers/acpi/acpica/evevent.c | 12 ++++++++++++ drivers/acpi/acpica/evglock.c | 6 ++++++ drivers/acpi/acpica/tbfadt.c | 22 ++++++++++++++++++++-- drivers/acpi/acpica/tbutils.c | 7 +++++++ include/acpi/acpixf.h | 1 + 6 files changed, 58 insertions(+), 3 deletions(-) diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h index 76dc02f15574..25912a93dea7 100644 --- a/drivers/acpi/acpica/acglobal.h +++ b/drivers/acpi/acpica/acglobal.h @@ -140,8 +140,19 @@ u32 acpi_gbl_trace_flags; acpi_name acpi_gbl_trace_method_name; u8 acpi_gbl_system_awake_and_running; +/* + * ACPI 5.0 introduces the concept of a "reduced hardware platform", meaning + * that the ACPI hardware is no longer required. A flag in the FADT indicates + * a reduced HW machine, and that flag is duplicated here for convenience. + */ +u8 acpi_gbl_reduced_hardware; + #endif +/* Do not disassemble buffers to resource descriptors */ + +ACPI_EXTERN u8 ACPI_INIT_GLOBAL(acpi_gbl_no_resource_disassembly, FALSE); + /***************************************************************************** * * Debug support @@ -207,7 +218,7 @@ ACPI_EXTERN struct acpi_rw_lock acpi_gbl_namespace_rw_lock; /***************************************************************************** * - * Mutual exlusion within ACPICA subsystem + * Mutual exclusion within ACPICA subsystem * ****************************************************************************/ diff --git a/drivers/acpi/acpica/evevent.c b/drivers/acpi/acpica/evevent.c index d458b041e651..1f43be470199 100644 --- a/drivers/acpi/acpica/evevent.c +++ b/drivers/acpi/acpica/evevent.c @@ -71,6 +71,12 @@ acpi_status acpi_ev_initialize_events(void) ACPI_FUNCTION_TRACE(ev_initialize_events); + /* If Hardware Reduced flag is set, there are no fixed events */ + + if (acpi_gbl_reduced_hardware) { + return_ACPI_STATUS(AE_OK); + } + /* * Initialize the Fixed and General Purpose Events. This is done prior to * enabling SCIs to prevent interrupts from occurring before the handlers @@ -111,6 +117,12 @@ acpi_status acpi_ev_install_xrupt_handlers(void) ACPI_FUNCTION_TRACE(ev_install_xrupt_handlers); + /* If Hardware Reduced flag is set, there is no ACPI h/w */ + + if (acpi_gbl_reduced_hardware) { + return_ACPI_STATUS(AE_OK); + } + /* Install the SCI handler */ status = acpi_ev_install_sci_handler(); diff --git a/drivers/acpi/acpica/evglock.c b/drivers/acpi/acpica/evglock.c index 56a562a1e5d7..af1e71af1b64 100644 --- a/drivers/acpi/acpica/evglock.c +++ b/drivers/acpi/acpica/evglock.c @@ -70,6 +70,12 @@ acpi_status acpi_ev_init_global_lock_handler(void) ACPI_FUNCTION_TRACE(ev_init_global_lock_handler); + /* If Hardware Reduced flag is set, there is no global lock */ + + if (acpi_gbl_reduced_hardware) { + return_ACPI_STATUS(AE_OK); + } + /* Attempt installation of the global lock handler */ status = acpi_install_fixed_event_handler(ACPI_EVENT_GLOBAL, diff --git a/drivers/acpi/acpica/tbfadt.c b/drivers/acpi/acpica/tbfadt.c index 536fd2bc7a47..1e1a7de34439 100644 --- a/drivers/acpi/acpica/tbfadt.c +++ b/drivers/acpi/acpica/tbfadt.c @@ -253,8 +253,13 @@ void acpi_tb_parse_fadt(u32 table_index) acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.Xdsdt, ACPI_SIG_DSDT, ACPI_TABLE_INDEX_DSDT); - acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT.Xfacs, - ACPI_SIG_FACS, ACPI_TABLE_INDEX_FACS); + /* If Hardware Reduced flag is set, there is no FACS */ + + if (!acpi_gbl_reduced_hardware) { + acpi_tb_install_table((acpi_physical_address) acpi_gbl_FADT. + Xfacs, ACPI_SIG_FACS, + ACPI_TABLE_INDEX_FACS); + } } /******************************************************************************* @@ -297,6 +302,13 @@ void acpi_tb_create_local_fadt(struct acpi_table_header *table, u32 length) ACPI_MEMCPY(&acpi_gbl_FADT, table, ACPI_MIN(length, sizeof(struct acpi_table_fadt))); + /* Take a copy of the Hardware Reduced flag */ + + acpi_gbl_reduced_hardware = FALSE; + if (acpi_gbl_FADT.flags & ACPI_FADT_HW_REDUCED) { + acpi_gbl_reduced_hardware = TRUE; + } + /* Convert the local copy of the FADT to the common internal format */ acpi_tb_convert_fadt(); @@ -502,6 +514,12 @@ static void acpi_tb_validate_fadt(void) acpi_gbl_FADT.Xdsdt = (u64) acpi_gbl_FADT.dsdt; } + /* If Hardware Reduced flag is set, we are all done */ + + if (acpi_gbl_reduced_hardware) { + return; + } + /* Examine all of the 64-bit extended address fields (X fields) */ for (i = 0; i < ACPI_FADT_INFO_ENTRIES; i++) { diff --git a/drivers/acpi/acpica/tbutils.c b/drivers/acpi/acpica/tbutils.c index 0f2d395feaba..da56a1908966 100644 --- a/drivers/acpi/acpica/tbutils.c +++ b/drivers/acpi/acpica/tbutils.c @@ -135,6 +135,13 @@ acpi_status acpi_tb_initialize_facs(void) { acpi_status status; + /* If Hardware Reduced flag is set, there is no FACS */ + + if (acpi_gbl_reduced_hardware) { + acpi_gbl_FACS = NULL; + return (AE_OK); + } + status = acpi_get_table_by_index(ACPI_TABLE_INDEX_FACS, ACPI_CAST_INDIRECT_PTR(struct acpi_table_header, diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index f554a9313b43..9ba2e3aff5c6 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -74,6 +74,7 @@ extern u8 acpi_gbl_disable_auto_repair; extern u32 acpi_current_gpe_count; extern struct acpi_table_fadt acpi_gbl_FADT; extern u8 acpi_gbl_system_awake_and_running; +extern u8 acpi_gbl_reduced_hardware; /* ACPI 5.0 */ extern u32 acpi_rsdt_forced; /* From c5bd6537329e66a8b36234f19a36d94b72d07394 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 11:04:00 +0800 Subject: [PATCH 437/820] ACPI 5.0: Add new/changed tables to headers Adds new file, actbl3.h Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/actbl.h | 1 + include/acpi/actbl1.h | 57 ++++- include/acpi/actbl3.h | 552 ++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 606 insertions(+), 4 deletions(-) create mode 100644 include/acpi/actbl3.h diff --git a/include/acpi/actbl.h b/include/acpi/actbl.h index 7d0e051b4863..8e1b92f6f650 100644 --- a/include/acpi/actbl.h +++ b/include/acpi/actbl.h @@ -349,6 +349,7 @@ struct acpi_table_desc { #include #include +#include /* * Sizes of the various flavors of FADT. We need to look closely diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index 7504bc99b29b..71e747beac8f 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -228,7 +228,8 @@ enum acpi_einj_actions { ACPI_EINJ_EXECUTE_OPERATION = 5, ACPI_EINJ_CHECK_BUSY_STATUS = 6, ACPI_EINJ_GET_COMMAND_STATUS = 7, - ACPI_EINJ_ACTION_RESERVED = 8, /* 8 and greater are reserved */ + ACPI_EINJ_SET_ERROR_TYPE_WITH_ADDRESS = 8, + ACPI_EINJ_ACTION_RESERVED = 9, /* 9 and greater are reserved */ ACPI_EINJ_TRIGGER_ERROR = 0xFF /* Except for this value */ }; @@ -240,7 +241,27 @@ enum acpi_einj_instructions { ACPI_EINJ_WRITE_REGISTER = 2, ACPI_EINJ_WRITE_REGISTER_VALUE = 3, ACPI_EINJ_NOOP = 4, - ACPI_EINJ_INSTRUCTION_RESERVED = 5 /* 5 and greater are reserved */ + ACPI_EINJ_FLUSH_CACHELINE = 5, + ACPI_EINJ_INSTRUCTION_RESERVED = 6 /* 6 and greater are reserved */ +}; + +struct acpi_einj_error_type_with_addr { + u32 error_type; + u32 vendor_struct_offset; + u32 flags; + u32 apic_id; + u64 address; + u64 range; + u32 pcie_id; +}; + +struct acpi_einj_vendor { + u32 length; + u32 pcie_id; + u16 vendor_id; + u16 device_id; + u8 revision_id; + u8 reserved[3]; }; /* EINJ Trigger Error Action Table */ @@ -275,6 +296,7 @@ enum acpi_einj_command_status { #define ACPI_EINJ_PLATFORM_CORRECTABLE (1<<9) #define ACPI_EINJ_PLATFORM_UNCORRECTABLE (1<<10) #define ACPI_EINJ_PLATFORM_FATAL (1<<11) +#define ACPI_EINJ_VENDOR_DEFINED (1<<31) /******************************************************************************* * @@ -631,7 +653,9 @@ enum acpi_madt_type { ACPI_MADT_TYPE_INTERRUPT_SOURCE = 8, ACPI_MADT_TYPE_LOCAL_X2APIC = 9, ACPI_MADT_TYPE_LOCAL_X2APIC_NMI = 10, - ACPI_MADT_TYPE_RESERVED = 11 /* 11 and greater are reserved */ + ACPI_MADT_TYPE_GENERIC_INTERRUPT = 11, + ACPI_MADT_TYPE_GENERIC_DISTRIBUTOR = 12, + ACPI_MADT_TYPE_RESERVED = 13 /* 13 and greater are reserved */ }; /* @@ -752,11 +776,36 @@ struct acpi_madt_local_x2apic_nmi { u8 reserved[3]; }; +/* 11: Generic Interrupt (ACPI 5.0) */ + +struct acpi_madt_generic_interrupt { + struct acpi_subtable_header header; + u16 reserved; /* Reserved - must be zero */ + u32 gic_id; + u32 uid; + u32 flags; + u32 parking_version; + u32 performance_interrupt; + u64 parked_address; + u64 base_address; +}; + +/* 12: Generic Distributor (ACPI 5.0) */ + +struct acpi_madt_generic_distributor { + struct acpi_subtable_header header; + u16 reserved; /* Reserved - must be zero */ + u32 gic_id; + u64 base_address; + u32 global_irq_base; + u32 reserved2; /* Reserved - must be zero */ +}; + /* * Common flags fields for MADT subtables */ -/* MADT Local APIC flags (lapic_flags) */ +/* MADT Local APIC flags (lapic_flags) and GIC flags */ #define ACPI_MADT_ENABLED (1) /* 00: Processor is usable if set */ diff --git a/include/acpi/actbl3.h b/include/acpi/actbl3.h new file mode 100644 index 000000000000..c22ce80e9535 --- /dev/null +++ b/include/acpi/actbl3.h @@ -0,0 +1,552 @@ +/****************************************************************************** + * + * Name: actbl3.h - ACPI Table Definitions + * + *****************************************************************************/ + +/* + * Copyright (C) 2000 - 2011, Intel Corp. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. Redistributions in binary form must reproduce at minimum a disclaimer + * substantially similar to the "NO WARRANTY" disclaimer below + * ("Disclaimer") and any redistribution must be conditioned upon + * including a substantially similar Disclaimer requirement for further + * binary redistribution. + * 3. Neither the names of the above-listed copyright holders nor the names + * of any contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL") version 2 as published by the Free + * Software Foundation. + * + * NO WARRANTY + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, + * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING + * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGES. + */ + +#ifndef __ACTBL3_H__ +#define __ACTBL3_H__ + +/******************************************************************************* + * + * Additional ACPI Tables (3) + * + * These tables are not consumed directly by the ACPICA subsystem, but are + * included here to support device drivers and the AML disassembler. + * + * The tables in this file are fully defined within the ACPI specification. + * + ******************************************************************************/ + +/* + * Values for description table header signatures for tables defined in this + * file. Useful because they make it more difficult to inadvertently type in + * the wrong signature. + */ +#define ACPI_SIG_BGRT "BGRT" /* Boot Graphics Resource Table */ +#define ACPI_SIG_DRTM "DRTM" /* Dynamic Root of Trust for Measurement table */ +#define ACPI_SIG_FPDT "FPDT" /* Firmware Performance Data Table */ +#define ACPI_SIG_GTDT "GTDT" /* Generic Timer Description Table */ +#define ACPI_SIG_MPST "MPST" /* Memory Power State Table */ +#define ACPI_SIG_PCCT "PCCT" /* Platform Communications Channel Table */ +#define ACPI_SIG_PMTT "PMTT" /* Platform Memory Topology Table */ +#define ACPI_SIG_RASF "RASF" /* RAS Feature table */ + +#define ACPI_SIG_S3PT "S3PT" /* S3 Performance (sub)Table */ +#define ACPI_SIG_PCCS "PCC" /* PCC Shared Memory Region */ + +/* Reserved table signatures */ + +#define ACPI_SIG_CSRT "CSRT" /* Core System Resources Table */ +#define ACPI_SIG_DBG2 "DBG2" /* Debug Port table 2 */ +#define ACPI_SIG_MATR "MATR" /* Memory Address Translation Table */ +#define ACPI_SIG_MSDM "MSDM" /* Microsoft Data Management Table */ +#define ACPI_SIG_WPBT "WPBT" /* Windows Platform Binary Table */ + +/* + * All tables must be byte-packed to match the ACPI specification, since + * the tables are provided by the system BIOS. + */ +#pragma pack(1) + +/* + * Note about bitfields: The u8 type is used for bitfields in ACPI tables. + * This is the only type that is even remotely portable. Anything else is not + * portable, so do not use any other bitfield types. + */ + +/******************************************************************************* + * + * BGRT - Boot Graphics Resource Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_bgrt { + struct acpi_table_header header; /* Common ACPI table header */ + u16 version; + u8 status; + u8 image_type; + u64 image_address; + u32 image_offset_x; + u32 image_offset_y; +}; + +/******************************************************************************* + * + * DRTM - Dynamic Root of Trust for Measurement table + * + ******************************************************************************/ + +struct acpi_table_drtm { + struct acpi_table_header header; /* Common ACPI table header */ + u64 entry_base_address; + u64 entry_length; + u32 entry_address32; + u64 entry_address64; + u64 exit_address; + u64 log_area_address; + u32 log_area_length; + u64 arch_dependent_address; + u32 flags; +}; + +/* 1) Validated Tables List */ + +struct acpi_drtm_vtl_list { + u32 validated_table_list_count; +}; + +/* 2) Resources List */ + +struct acpi_drtm_resource_list { + u32 resource_list_count; +}; + +/* 3) Platform-specific Identifiers List */ + +struct acpi_drtm_id_list { + u32 id_list_count; +}; + +/******************************************************************************* + * + * FPDT - Firmware Performance Data Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_fpdt { + struct acpi_table_header header; /* Common ACPI table header */ +}; + +/* FPDT subtable header */ + +struct acpi_fpdt_header { + u16 type; + u8 length; + u8 revision; +}; + +/* Values for Type field above */ + +enum acpi_fpdt_type { + ACPI_FPDT_TYPE_BOOT = 0, + ACPI_FPDT_TYPE_S3PERF = 1, +}; + +/* + * FPDT subtables + */ + +/* 0: Firmware Basic Boot Performance Record */ + +struct acpi_fpdt_boot { + struct acpi_fpdt_header header; + u8 reserved[4]; + u64 reset_end; + u64 load_start; + u64 startup_start; + u64 exit_services_entry; + u64 exit_services_exit; +}; + +/* 1: S3 Performance Table Pointer Record */ + +struct acpi_fpdt_s3pt_ptr { + struct acpi_fpdt_header header; + u8 reserved[4]; + u64 address; +}; + +/* + * S3PT - S3 Performance Table. This table is pointed to by the + * FPDT S3 Pointer Record above. + */ +struct acpi_table_s3pt { + u8 signature[4]; /* "S3PT" */ + u32 length; +}; + +/* + * S3PT Subtables + */ +struct acpi_s3pt_header { + u16 type; + u8 length; + u8 revision; +}; + +/* Values for Type field above */ + +enum acpi_s3pt_type { + ACPI_S3PT_TYPE_RESUME = 0, + ACPI_S3PT_TYPE_SUSPEND = 1, +}; + +struct acpi_s3pt_resume { + struct acpi_s3pt_header header; + u32 resume_count; + u64 full_resume; + u64 average_resume; +}; + +struct acpi_s3pt_suspend { + struct acpi_s3pt_header header; + u64 suspend_start; + u64 suspend_end; +}; + +/******************************************************************************* + * + * GTDT - Generic Timer Description Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_gtdt { + struct acpi_table_header header; /* Common ACPI table header */ + u64 address; + u32 flags; + u32 secure_pl1_interrupt; + u32 secure_pl1_flags; + u32 non_secure_pl1_interrupt; + u32 non_secure_pl1_flags; + u32 virtual_timer_interrupt; + u32 virtual_timer_flags; + u32 non_secure_pl2_interrupt; + u32 non_secure_pl2_flags; +}; + +/* Values for Flags field above */ + +#define ACPI_GTDT_MAPPED_BLOCK_PRESENT 1 + +/* Values for all "TimerFlags" fields above */ + +#define ACPI_GTDT_INTERRUPT_MODE 1 +#define ACPI_GTDT_INTERRUPT_POLARITY 2 + +/******************************************************************************* + * + * MPST - Memory Power State Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +#define ACPI_MPST_CHANNEL_INFO \ + u16 reserved1; \ + u8 channel_id; \ + u8 reserved2; \ + u16 power_node_count; + +/* Main table */ + +struct acpi_table_mpst { + struct acpi_table_header header; /* Common ACPI table header */ + ACPI_MPST_CHANNEL_INFO /* Platform Communication Channel */ +}; + +/* Memory Platform Communication Channel Info */ + +struct acpi_mpst_channel { + ACPI_MPST_CHANNEL_INFO /* Platform Communication Channel */ +}; + +/* Memory Power Node Structure */ + +struct acpi_mpst_power_node { + u8 flags; + u8 reserved1; + u16 node_id; + u32 length; + u64 range_address; + u64 range_length; + u8 num_power_states; + u8 num_physical_components; + u16 reserved2; +}; + +/* Values for Flags field above */ + +#define ACPI_MPST_ENABLED 1 +#define ACPI_MPST_POWER_MANAGED 2 +#define ACPI_MPST_HOT_PLUG_CAPABLE 4 + +/* Memory Power State Structure (follows POWER_NODE above) */ + +struct acpi_mpst_power_state { + u8 power_state; + u8 info_index; +}; + +/* Physical Component ID Structure (follows POWER_STATE above) */ + +struct acpi_mpst_component { + u16 component_id; +}; + +/* Memory Power State Characteristics Structure (follows all POWER_NODEs) */ + +struct acpi_mpst_data_hdr { + u16 characteristics_count; +}; + +struct acpi_mpst_power_data { + u8 revision; + u8 flags; + u16 reserved1; + u32 average_power; + u32 power_saving; + u64 exit_latency; + u64 reserved2; +}; + +/* Values for Flags field above */ + +#define ACPI_MPST_PRESERVE 1 +#define ACPI_MPST_AUTOENTRY 2 +#define ACPI_MPST_AUTOEXIT 4 + +/* Shared Memory Region (not part of an ACPI table) */ + +struct acpi_mpst_shared { + u32 signature; + u16 pcc_command; + u16 pcc_status; + u16 command_register; + u16 status_register; + u16 power_state_id; + u16 power_node_id; + u64 energy_consumed; + u64 average_power; +}; + +/******************************************************************************* + * + * PCCT - Platform Communications Channel Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_pcct { + struct acpi_table_header header; /* Common ACPI table header */ + u32 flags; + u32 latency; + u32 reserved; +}; + +/* Values for Flags field above */ + +#define ACPI_PCCT_DOORBELL 1 + +/* + * PCCT subtables + */ + +/* 0: Generic Communications Subspace */ + +struct acpi_pcct_subspace { + struct acpi_subtable_header header; + u8 reserved[6]; + u64 base_address; + u64 length; + struct acpi_generic_address doorbell_register; + u64 preserve_mask; + u64 write_mask; +}; + +/* + * PCC memory structures (not part of the ACPI table) + */ + +/* Shared Memory Region */ + +struct acpi_pcct_shared_memory { + u32 signature; + u16 command; + u16 status; +}; + +/******************************************************************************* + * + * PMTT - Platform Memory Topology Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_pmtt { + struct acpi_table_header header; /* Common ACPI table header */ + u32 reserved; +}; + +/* Common header for PMTT subtables that follow main table */ + +struct acpi_pmtt_header { + u8 type; + u8 reserved1; + u16 length; + u16 flags; + u16 reserved2; +}; + +/* Values for Type field above */ + +#define ACPI_PMTT_TYPE_SOCKET 0 +#define ACPI_PMTT_TYPE_CONTROLLER 1 +#define ACPI_PMTT_TYPE_DIMM 2 +#define ACPI_PMTT_TYPE_RESERVED 3 /* 0x03-0xFF are reserved */ + +/* Values for Flags field above */ + +#define ACPI_PMTT_TOP_LEVEL 0x0001 +#define ACPI_PMTT_PHYSICAL 0x0002 +#define ACPI_PMTT_MEMORY_TYPE 0x000C + +/* + * PMTT subtables, correspond to Type in struct acpi_pmtt_header + */ + +/* 0: Socket Structure */ + +struct acpi_pmtt_socket { + struct acpi_pmtt_header header; + u16 socket_id; + u16 reserved; +}; + +/* 1: Memory Controller subtable */ + +struct acpi_pmtt_controller { + struct acpi_pmtt_header header; + u32 read_latency; + u32 write_latency; + u32 read_bandwidth; + u32 write_bandwidth; + u16 access_width; + u16 alignment; + u16 reserved; + u16 domain_count; +}; + +/* 1a: Proximity Domain substructure */ + +struct acpi_pmtt_domain { + u32 proximity_domain; +}; + +/* 2: Physical Component Identifier (DIMM) */ + +struct acpi_pmtt_physical_component { + struct acpi_pmtt_header header; + u16 component_id; + u16 reserved; + u32 memory_size; + u32 bios_handle; +}; + +/******************************************************************************* + * + * RASF - RAS Feature Table (ACPI 5.0) + * Version 1 + * + ******************************************************************************/ + +struct acpi_table_rasf { + struct acpi_table_header header; /* Common ACPI table header */ + u8 channel_id[12]; +}; + +/* RASF Platform Communication Channel Shared Memory Region */ + +struct acpi_rasf_shared_memory { + u32 signature; + u16 command; + u16 status; + u64 requested_address; + u64 requested_length; + u64 actual_address; + u64 actual_length; + u16 flags; + u8 speed; +}; + +/* Masks for Flags and Speed fields above */ + +#define ACPI_RASF_SCRUBBER_RUNNING 1 +#define ACPI_RASF_SPEED (7<<1) + +/* Channel Commands */ + +enum acpi_rasf_commands { + ACPI_RASF_GET_RAS_CAPABILITIES = 1, + ACPI_RASF_GET_PATROL_PARAMETERS = 2, + ACPI_RASF_START_PATROL_SCRUBBER = 3, + ACPI_RASF_STOP_PATROL_SCRUBBER = 4 +}; + +/* Channel Command flags */ + +#define ACPI_RASF_GENERATE_SCI (1<<15) + +/* Status values */ + +enum acpi_rasf_status { + ACPI_RASF_SUCCESS = 0, + ACPI_RASF_NOT_VALID = 1, + ACPI_RASF_NOT_SUPPORTED = 2, + ACPI_RASF_BUSY = 3, + ACPI_RASF_FAILED = 4, + ACPI_RASF_ABORTED = 5, + ACPI_RASF_INVALID_DATA = 6 +}; + +/* Status flags */ + +#define ACPI_RASF_COMMAND_COMPLETE (1) +#define ACPI_RASF_SCI_DOORBELL (1<<1) +#define ACPI_RASF_ERROR (1<<2) +#define ACPI_RASF_STATUS (0x1F<<3) + +/* Reset to default packing */ + +#pragma pack() + +#endif /* __ACTBL3_H__ */ From ffef68273b6278e98a99dd4051671d4854b20fe0 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Wed, 16 Nov 2011 11:08:30 +0800 Subject: [PATCH 438/820] ACPI 5.0: New interfaces to allow driver access to AML mutex objects Adds acpi_acquire_mutex, acpi_release_mutex external interfaces. New file, utxfmutex.c. Signed-off-by: Lin Ming Signed-off-by: Bob Moore Signed-off-by: Len Brown --- drivers/acpi/acpica/Makefile | 2 +- drivers/acpi/acpica/utxfmutex.c | 187 ++++++++++++++++++++++++++++++++ include/acpi/acpixf.h | 15 ++- 3 files changed, 201 insertions(+), 3 deletions(-) create mode 100644 drivers/acpi/acpica/utxfmutex.c diff --git a/drivers/acpi/acpica/Makefile b/drivers/acpi/acpica/Makefile index 301bd2d388ad..1bd466113dc8 100644 --- a/drivers/acpi/acpica/Makefile +++ b/drivers/acpi/acpica/Makefile @@ -45,4 +45,4 @@ acpi-y += tbxface.o tbinstal.o tbutils.o tbfind.o tbfadt.o tbxfroot.o acpi-y += utalloc.o utdebug.o uteval.o utinit.o utmisc.o utxface.o \ utcopy.o utdelete.o utglobal.o utmath.o utobject.o \ utstate.o utmutex.o utobject.o utresrc.o utlock.o utids.o \ - utosi.o utxferror.o utdecode.o + utosi.o utxferror.o utdecode.o utxfmutex.o diff --git a/drivers/acpi/acpica/utxfmutex.c b/drivers/acpi/acpica/utxfmutex.c new file mode 100644 index 000000000000..1ea016b228bd --- /dev/null +++ b/drivers/acpi/acpica/utxfmutex.c @@ -0,0 +1,187 @@ +/******************************************************************************* + * + * Module Name: utxfmutex - external AML mutex access functions + * + ******************************************************************************/ + +/* + * Copyright (C) 2000 - 2011, Intel Corp. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. Redistributions in binary form must reproduce at minimum a disclaimer + * substantially similar to the "NO WARRANTY" disclaimer below + * ("Disclaimer") and any redistribution must be conditioned upon + * including a substantially similar Disclaimer requirement for further + * binary redistribution. + * 3. Neither the names of the above-listed copyright holders nor the names + * of any contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL") version 2 as published by the Free + * Software Foundation. + * + * NO WARRANTY + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, + * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING + * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGES. + */ + +#include +#include "accommon.h" +#include "acnamesp.h" + +#define _COMPONENT ACPI_UTILITIES +ACPI_MODULE_NAME("utxfmutex") + +/* Local prototypes */ +static acpi_status +acpi_ut_get_mutex_object(acpi_handle handle, + acpi_string pathname, + union acpi_operand_object **ret_obj); + +/******************************************************************************* + * + * FUNCTION: acpi_ut_get_mutex_object + * + * PARAMETERS: Handle - Mutex or prefix handle (optional) + * Pathname - Mutex pathname (optional) + * ret_obj - Where the mutex object is returned + * + * RETURN: Status + * + * DESCRIPTION: Get an AML mutex object. The mutex node is pointed to by + * Handle:Pathname. Either Handle or Pathname can be NULL, but + * not both. + * + ******************************************************************************/ + +static acpi_status +acpi_ut_get_mutex_object(acpi_handle handle, + acpi_string pathname, + union acpi_operand_object **ret_obj) +{ + struct acpi_namespace_node *mutex_node; + union acpi_operand_object *mutex_obj; + acpi_status status; + + /* Parameter validation */ + + if (!ret_obj || (!handle && !pathname)) { + return (AE_BAD_PARAMETER); + } + + /* Get a the namespace node for the mutex */ + + mutex_node = handle; + if (pathname != NULL) { + status = acpi_get_handle(handle, pathname, + ACPI_CAST_PTR(acpi_handle, + &mutex_node)); + if (ACPI_FAILURE(status)) { + return (status); + } + } + + /* Ensure that we actually have a Mutex object */ + + if (!mutex_node || (mutex_node->type != ACPI_TYPE_MUTEX)) { + return (AE_TYPE); + } + + /* Get the low-level mutex object */ + + mutex_obj = acpi_ns_get_attached_object(mutex_node); + if (!mutex_obj) { + return (AE_NULL_OBJECT); + } + + *ret_obj = mutex_obj; + return (AE_OK); +} + +/******************************************************************************* + * + * FUNCTION: acpi_acquire_mutex + * + * PARAMETERS: Handle - Mutex or prefix handle (optional) + * Pathname - Mutex pathname (optional) + * Timeout - Max time to wait for the lock (millisec) + * + * RETURN: Status + * + * DESCRIPTION: Acquire an AML mutex. This is a device driver interface to + * AML mutex objects, and allows for transaction locking between + * drivers and AML code. The mutex node is pointed to by + * Handle:Pathname. Either Handle or Pathname can be NULL, but + * not both. + * + ******************************************************************************/ + +acpi_status +acpi_acquire_mutex(acpi_handle handle, acpi_string pathname, u16 timeout) +{ + acpi_status status; + union acpi_operand_object *mutex_obj; + + /* Get the low-level mutex associated with Handle:Pathname */ + + status = acpi_ut_get_mutex_object(handle, pathname, &mutex_obj); + if (ACPI_FAILURE(status)) { + return (status); + } + + /* Acquire the OS mutex */ + + status = acpi_os_acquire_mutex(mutex_obj->mutex.os_mutex, timeout); + return (status); +} + +/******************************************************************************* + * + * FUNCTION: acpi_release_mutex + * + * PARAMETERS: Handle - Mutex or prefix handle (optional) + * Pathname - Mutex pathname (optional) + * + * RETURN: Status + * + * DESCRIPTION: Release an AML mutex. This is a device driver interface to + * AML mutex objects, and allows for transaction locking between + * drivers and AML code. The mutex node is pointed to by + * Handle:Pathname. Either Handle or Pathname can be NULL, but + * not both. + * + ******************************************************************************/ + +acpi_status acpi_release_mutex(acpi_handle handle, acpi_string pathname) +{ + acpi_status status; + union acpi_operand_object *mutex_obj; + + /* Get the low-level mutex associated with Handle:Pathname */ + + status = acpi_ut_get_mutex_object(handle, pathname, &mutex_obj); + if (ACPI_FAILURE(status)) { + return (status); + } + + /* Release the OS mutex */ + + acpi_os_release_mutex(mutex_obj->mutex.os_mutex); + return (AE_OK); +} diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 9ba2e3aff5c6..dd86610039b3 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -277,12 +277,23 @@ acpi_status acpi_install_exception_handler(acpi_exception_handler handler); acpi_status acpi_install_interface_handler(acpi_interface_handler handler); /* - * Event interfaces + * Global Lock interfaces */ acpi_status acpi_acquire_global_lock(u16 timeout, u32 * handle); acpi_status acpi_release_global_lock(u32 handle); +/* + * Interfaces to AML mutex objects + */ +acpi_status +acpi_acquire_mutex(acpi_handle handle, acpi_string pathname, u16 timeout); + +acpi_status acpi_release_mutex(acpi_handle handle, acpi_string pathname); + +/* + * Fixed Event interfaces + */ acpi_status acpi_enable_event(u32 event, u32 flags); acpi_status acpi_disable_event(u32 event, u32 flags); @@ -292,7 +303,7 @@ acpi_status acpi_clear_event(u32 event); acpi_status acpi_get_event_status(u32 event, acpi_event_status * event_status); /* - * GPE Interfaces + * General Purpose Event (GPE) Interfaces */ acpi_status acpi_enable_gpe(acpi_handle gpe_device, u32 gpe_number); From 9ce81784c9c0396a6a6be05248928a71134fe60b Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 13:39:07 +0800 Subject: [PATCH 439/820] ACPI 5.0: Implement Connection() and AccessAs() changes Support within the interpreter and operation region dispatch. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acevents.h | 1 + drivers/acpi/acpica/aclocal.h | 6 +- drivers/acpi/acpica/acobject.h | 5 +- drivers/acpi/acpica/acopcode.h | 4 + drivers/acpi/acpica/amlcode.h | 10 +++ drivers/acpi/acpica/dsfield.c | 81 +++++++++++++++---- drivers/acpi/acpica/evregion.c | 29 ++++++- drivers/acpi/acpica/exconfig.c | 6 +- drivers/acpi/acpica/exdump.c | 7 +- drivers/acpi/acpica/exfldio.c | 11 +-- drivers/acpi/acpica/exprep.c | 25 ++++++ drivers/acpi/acpica/psargs.c | 143 +++++++++++++++++++++++++++++---- drivers/acpi/acpica/psopcode.c | 13 ++- drivers/acpi/acpica/pstree.c | 6 ++ include/acpi/actypes.h | 8 ++ 15 files changed, 309 insertions(+), 46 deletions(-) diff --git a/drivers/acpi/acpica/acevents.h b/drivers/acpi/acpica/acevents.h index bea3b4899183..f763b88ed772 100644 --- a/drivers/acpi/acpica/acevents.h +++ b/drivers/acpi/acpica/acevents.h @@ -162,6 +162,7 @@ acpi_status acpi_ev_initialize_op_regions(void); acpi_status acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, + union acpi_operand_object *field_obj, u32 function, u32 region_offset, u32 bit_width, u64 *value); diff --git a/drivers/acpi/acpica/aclocal.h b/drivers/acpi/acpica/aclocal.h index 5552125d8340..52c4ebddbcf1 100644 --- a/drivers/acpi/acpica/aclocal.h +++ b/drivers/acpi/acpica/aclocal.h @@ -53,7 +53,7 @@ typedef u32 acpi_mutex_handle; /* Total number of aml opcodes defined */ -#define AML_NUM_OPCODES 0x7F +#define AML_NUM_OPCODES 0x81 /* Forward declarations */ @@ -249,12 +249,16 @@ struct acpi_create_field_info { struct acpi_namespace_node *field_node; struct acpi_namespace_node *register_node; struct acpi_namespace_node *data_register_node; + struct acpi_namespace_node *connection_node; + u8 *resource_buffer; u32 bank_value; u32 field_bit_position; u32 field_bit_length; + u16 resource_length; u8 field_flags; u8 attribute; u8 field_type; + u8 access_length; }; typedef diff --git a/drivers/acpi/acpica/acobject.h b/drivers/acpi/acpica/acobject.h index 6d276c20b57b..435dd2f847bb 100644 --- a/drivers/acpi/acpica/acobject.h +++ b/drivers/acpi/acpica/acobject.h @@ -254,6 +254,7 @@ ACPI_OBJECT_COMMON_HEADER ACPI_COMMON_NOTIFY_INFO}; u32 base_byte_offset; /* Byte offset within containing object */\ u32 value; /* Value to store into the Bank or Index register */\ u8 start_field_bit_offset;/* Bit offset within first field datum (0-63) */\ + u8 access_length; /* For serial regions/fields */ struct acpi_object_field_common { /* COMMON FIELD (for BUFFER, REGION, BANK, and INDEX fields) */ @@ -261,7 +262,9 @@ struct acpi_object_field_common { /* COMMON FIELD (for BUFFER, REGION, BANK, and }; struct acpi_object_region_field { - ACPI_OBJECT_COMMON_HEADER ACPI_COMMON_FIELD_INFO union acpi_operand_object *region_obj; /* Containing op_region object */ + ACPI_OBJECT_COMMON_HEADER ACPI_COMMON_FIELD_INFO u16 resource_length; + union acpi_operand_object *region_obj; /* Containing op_region object */ + u8 *resource_buffer; /* resource_template for serial regions/fields */ }; struct acpi_object_bank_field { diff --git a/drivers/acpi/acpica/acopcode.h b/drivers/acpi/acpica/acopcode.h index bb2ccfad7376..9896c77e94f9 100644 --- a/drivers/acpi/acpica/acopcode.h +++ b/drivers/acpi/acpica/acopcode.h @@ -93,6 +93,7 @@ #define ARGP_CONCAT_OP ARGP_LIST3 (ARGP_TERMARG, ARGP_TERMARG, ARGP_TARGET) #define ARGP_CONCAT_RES_OP ARGP_LIST3 (ARGP_TERMARG, ARGP_TERMARG, ARGP_TARGET) #define ARGP_COND_REF_OF_OP ARGP_LIST2 (ARGP_SUPERNAME, ARGP_SUPERNAME) +#define ARGP_CONNECTFIELD_OP ARGP_LIST1 (ARGP_NAMESTRING) #define ARGP_CONTINUE_OP ARG_NONE #define ARGP_COPY_OP ARGP_LIST2 (ARGP_TERMARG, ARGP_SIMPLENAME) #define ARGP_CREATE_BIT_FIELD_OP ARGP_LIST3 (ARGP_TERMARG, ARGP_TERMARG, ARGP_NAME) @@ -164,6 +165,7 @@ #define ARGP_RETURN_OP ARGP_LIST1 (ARGP_TERMARG) #define ARGP_REVISION_OP ARG_NONE #define ARGP_SCOPE_OP ARGP_LIST3 (ARGP_PKGLENGTH, ARGP_NAME, ARGP_TERMLIST) +#define ARGP_SERIALFIELD_OP ARGP_LIST1 (ARGP_NAMESTRING) #define ARGP_SHIFT_LEFT_OP ARGP_LIST3 (ARGP_TERMARG, ARGP_TERMARG, ARGP_TARGET) #define ARGP_SHIFT_RIGHT_OP ARGP_LIST3 (ARGP_TERMARG, ARGP_TERMARG, ARGP_TARGET) #define ARGP_SIGNAL_OP ARGP_LIST1 (ARGP_SUPERNAME) @@ -223,6 +225,7 @@ #define ARGI_CONCAT_OP ARGI_LIST3 (ARGI_COMPUTEDATA,ARGI_COMPUTEDATA, ARGI_TARGETREF) #define ARGI_CONCAT_RES_OP ARGI_LIST3 (ARGI_BUFFER, ARGI_BUFFER, ARGI_TARGETREF) #define ARGI_COND_REF_OF_OP ARGI_LIST2 (ARGI_OBJECT_REF, ARGI_TARGETREF) +#define ARGI_CONNECTFIELD_OP ARGI_INVALID_OPCODE #define ARGI_CONTINUE_OP ARGI_INVALID_OPCODE #define ARGI_COPY_OP ARGI_LIST2 (ARGI_ANYTYPE, ARGI_SIMPLE_TARGET) #define ARGI_CREATE_BIT_FIELD_OP ARGI_LIST3 (ARGI_BUFFER, ARGI_INTEGER, ARGI_REFERENCE) @@ -294,6 +297,7 @@ #define ARGI_RETURN_OP ARGI_INVALID_OPCODE #define ARGI_REVISION_OP ARG_NONE #define ARGI_SCOPE_OP ARGI_INVALID_OPCODE +#define ARGI_SERIALFIELD_OP ARGI_INVALID_OPCODE #define ARGI_SHIFT_LEFT_OP ARGI_LIST3 (ARGI_INTEGER, ARGI_INTEGER, ARGI_TARGETREF) #define ARGI_SHIFT_RIGHT_OP ARGI_LIST3 (ARGI_INTEGER, ARGI_INTEGER, ARGI_TARGETREF) #define ARGI_SIGNAL_OP ARGI_LIST1 (ARGI_EVENT) diff --git a/drivers/acpi/acpica/amlcode.h b/drivers/acpi/acpica/amlcode.h index 1077f17859ed..adf7a71ea766 100644 --- a/drivers/acpi/acpica/amlcode.h +++ b/drivers/acpi/acpica/amlcode.h @@ -188,6 +188,14 @@ #define AML_LLESSEQUAL_OP (u16) 0x9294 #define AML_LNOTEQUAL_OP (u16) 0x9293 +/* + * Opcodes for "Field" operators + */ +#define AML_FIELD_OFFSET_OP (u8) 0x00 +#define AML_FIELD_ACCESS_OP (u8) 0x01 +#define AML_FIELD_CONNECTION_OP (u8) 0x02 /* ACPI 5.0 */ +#define AML_FIELD_EXT_ACCESS_OP (u8) 0x03 /* ACPI 5.0 */ + /* * Internal opcodes * Use only "Unknown" AML opcodes, don't attempt to use @@ -202,6 +210,8 @@ #define AML_INT_METHODCALL_OP (u16) 0x0035 #define AML_INT_RETURN_VALUE_OP (u16) 0x0036 #define AML_INT_EVAL_SUBTREE_OP (u16) 0x0037 +#define AML_INT_CONNECTION_OP (u16) 0x0038 +#define AML_INT_EXTACCESSFIELD_OP (u16) 0x0039 #define ARG_NONE 0x0 diff --git a/drivers/acpi/acpica/dsfield.c b/drivers/acpi/acpica/dsfield.c index 34be60c0e448..aa880d992984 100644 --- a/drivers/acpi/acpica/dsfield.c +++ b/drivers/acpi/acpica/dsfield.c @@ -221,6 +221,7 @@ acpi_ds_get_field_names(struct acpi_create_field_info *info, { acpi_status status; u64 position; + union acpi_parse_object *child; ACPI_FUNCTION_TRACE_PTR(ds_get_field_names, info); @@ -232,10 +233,11 @@ acpi_ds_get_field_names(struct acpi_create_field_info *info, while (arg) { /* - * Three types of field elements are handled: - * 1) Offset - specifies a bit offset - * 2) access_as - changes the access mode - * 3) Name - Enters a new named field into the namespace + * Four types of field elements are handled: + * 1) Name - Enters a new named field into the namespace + * 2) Offset - specifies a bit offset + * 3) access_as - changes the access mode/attributes + * 4) Connection - Associate a resource template with the field */ switch (arg->common.aml_opcode) { case AML_INT_RESERVEDFIELD_OP: @@ -253,21 +255,70 @@ acpi_ds_get_field_names(struct acpi_create_field_info *info, break; case AML_INT_ACCESSFIELD_OP: - + case AML_INT_EXTACCESSFIELD_OP: /* - * Get a new access_type and access_attribute -- to be used for all - * field units that follow, until field end or another access_as - * keyword. + * Get new access_type, access_attribute, and access_length fields + * -- to be used for all field units that follow, until the + * end-of-field or another access_as keyword is encountered. + * NOTE. These three bytes are encoded in the integer value + * of the parseop for convenience. * * In field_flags, preserve the flag bits other than the - * ACCESS_TYPE bits + * ACCESS_TYPE bits. */ + + /* access_type (byte_acc, word_acc, etc.) */ + info->field_flags = (u8) ((info-> field_flags & ~(AML_FIELD_ACCESS_TYPE_MASK)) | - ((u8) ((u32) arg->common.value.integer >> 8))); + ((u8)((u32)(arg->common.value.integer & 0x07)))); - info->attribute = (u8) (arg->common.value.integer); + /* access_attribute (attrib_quick, attrib_byte, etc.) */ + + info->attribute = + (u8)((arg->common.value.integer >> 8) & 0xFF); + + /* access_length (for serial/buffer protocols) */ + + info->access_length = + (u8)((arg->common.value.integer >> 16) & 0xFF); + break; + + case AML_INT_CONNECTION_OP: + /* + * Clear any previous connection. New connection is used for all + * fields that follow, similar to access_as + */ + info->resource_buffer = NULL; + info->connection_node = NULL; + + /* + * A Connection() is either an actual resource descriptor (buffer) + * or a named reference to a resource template + */ + child = arg->common.value.arg; + if (child->common.aml_opcode == AML_INT_BYTELIST_OP) { + info->resource_buffer = child->named.data; + info->resource_length = + (u16)child->named.value.integer; + } else { + /* Lookup the Connection() namepath, it should already exist */ + + status = acpi_ns_lookup(walk_state->scope_info, + child->common.value. + name, ACPI_TYPE_ANY, + ACPI_IMODE_EXECUTE, + ACPI_NS_DONT_OPEN_SCOPE, + walk_state, + &info->connection_node); + if (ACPI_FAILURE(status)) { + ACPI_ERROR_NAMESPACE(child->common. + value.name, + status); + return_ACPI_STATUS(status); + } + } break; case AML_INT_NAMEDFIELD_OP: @@ -374,6 +425,8 @@ acpi_ds_create_field(union acpi_parse_object *op, } } + ACPI_MEMSET(&info, 0, sizeof(struct acpi_create_field_info)); + /* Second arg is the field flags */ arg = arg->common.next; @@ -386,7 +439,6 @@ acpi_ds_create_field(union acpi_parse_object *op, info.region_node = region_node; status = acpi_ds_get_field_names(&info, walk_state, arg->common.next); - return_ACPI_STATUS(status); } @@ -474,8 +526,8 @@ acpi_ds_init_field_objects(union acpi_parse_object *op, */ while (arg) { /* - * Ignore OFFSET and ACCESSAS terms here; we are only interested in the - * field names in order to enter them into the namespace. + * Ignore OFFSET/ACCESSAS/CONNECTION terms here; we are only interested + * in the field names in order to enter them into the namespace. */ if (arg->common.aml_opcode == AML_INT_NAMEDFIELD_OP) { status = acpi_ns_lookup(walk_state->scope_info, @@ -651,6 +703,5 @@ acpi_ds_create_index_field(union acpi_parse_object *op, info.region_node = region_node; status = acpi_ds_get_field_names(&info, walk_state, arg->common.next); - return_ACPI_STATUS(status); } diff --git a/drivers/acpi/acpica/evregion.c b/drivers/acpi/acpica/evregion.c index f0edf5c43c03..dcdc7e4a285e 100644 --- a/drivers/acpi/acpica/evregion.c +++ b/drivers/acpi/acpica/evregion.c @@ -329,6 +329,7 @@ acpi_ev_execute_reg_method(union acpi_operand_object *region_obj, u32 function) * FUNCTION: acpi_ev_address_space_dispatch * * PARAMETERS: region_obj - Internal region object + * field_obj - Corresponding field. Can be NULL. * Function - Read or Write operation * region_offset - Where in the region to read or write * bit_width - Field width in bits (8, 16, 32, or 64) @@ -344,6 +345,7 @@ acpi_ev_execute_reg_method(union acpi_operand_object *region_obj, u32 function) acpi_status acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, + union acpi_operand_object *field_obj, u32 function, u32 region_offset, u32 bit_width, u64 *value) { @@ -353,6 +355,7 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, union acpi_operand_object *handler_desc; union acpi_operand_object *region_obj2; void *region_context = NULL; + struct acpi_connection_info *context; ACPI_FUNCTION_TRACE(ev_address_space_dispatch); @@ -375,6 +378,8 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, return_ACPI_STATUS(AE_NOT_EXIST); } + context = handler_desc->address_space.context; + /* * It may be the case that the region has never been initialized. * Some types of regions require special init code @@ -404,8 +409,7 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, acpi_ex_exit_interpreter(); status = region_setup(region_obj, ACPI_REGION_ACTIVATE, - handler_desc->address_space.context, - ®ion_context); + context, ®ion_context); /* Re-enter the interpreter */ @@ -455,6 +459,25 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, acpi_ut_get_region_name(region_obj->region. space_id))); + /* + * Special handling for generic_serial_bus and general_purpose_io: + * There are three extra parameters that must be passed to the + * handler via the context: + * 1) Connection buffer, a resource template from Connection() op. + * 2) Length of the above buffer. + * 3) Actual access length from the access_as() op. + */ + if (((region_obj->region.space_id == ACPI_ADR_SPACE_GSBUS) || + (region_obj->region.space_id == ACPI_ADR_SPACE_GPIO)) && + context && field_obj) { + + /* Get the Connection (resource_template) buffer */ + + context->connection = field_obj->field.resource_buffer; + context->length = field_obj->field.resource_length; + context->access_length = field_obj->field.access_length; + } + if (!(handler_desc->address_space.handler_flags & ACPI_ADDR_HANDLER_DEFAULT_INSTALLED)) { /* @@ -469,7 +492,7 @@ acpi_ev_address_space_dispatch(union acpi_operand_object *region_obj, status = handler(function, (region_obj->region.address + region_offset), - bit_width, value, handler_desc->address_space.context, + bit_width, value, context, region_obj2->extra.region_context); if (ACPI_FAILURE(status)) { diff --git a/drivers/acpi/acpica/exconfig.c b/drivers/acpi/acpica/exconfig.c index 745a42b401f5..efe41f736fab 100644 --- a/drivers/acpi/acpica/exconfig.c +++ b/drivers/acpi/acpica/exconfig.c @@ -297,9 +297,9 @@ acpi_ex_region_read(union acpi_operand_object *obj_desc, u32 length, u8 *buffer) /* Bytewise reads */ for (i = 0; i < length; i++) { - status = acpi_ev_address_space_dispatch(obj_desc, ACPI_READ, - region_offset, 8, - &value); + status = + acpi_ev_address_space_dispatch(obj_desc, NULL, ACPI_READ, + region_offset, 8, &value); if (ACPI_FAILURE(status)) { return status; } diff --git a/drivers/acpi/acpica/exdump.c b/drivers/acpi/acpica/exdump.c index 61b8c0e8b74d..c7064f4d4704 100644 --- a/drivers/acpi/acpica/exdump.c +++ b/drivers/acpi/acpica/exdump.c @@ -192,10 +192,13 @@ static struct acpi_exdump_info acpi_ex_dump_buffer_field[3] = { "Buffer Object"} }; -static struct acpi_exdump_info acpi_ex_dump_region_field[3] = { +static struct acpi_exdump_info acpi_ex_dump_region_field[5] = { {ACPI_EXD_INIT, ACPI_EXD_TABLE_SIZE(acpi_ex_dump_region_field), NULL}, {ACPI_EXD_FIELD, 0, NULL}, - {ACPI_EXD_POINTER, ACPI_EXD_OFFSET(field.region_obj), "Region Object"} + {ACPI_EXD_UINT8, ACPI_EXD_OFFSET(field.access_length), "AccessLength"}, + {ACPI_EXD_POINTER, ACPI_EXD_OFFSET(field.region_obj), "Region Object"}, + {ACPI_EXD_POINTER, ACPI_EXD_OFFSET(field.resource_buffer), + "ResourceBuffer"} }; static struct acpi_exdump_info acpi_ex_dump_bank_field[5] = { diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index 2a524fc1e851..cc9c5dfa4c45 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -283,11 +283,12 @@ acpi_ex_access_region(union acpi_operand_object *obj_desc, /* Invoke the appropriate address_space/op_region handler */ - status = - acpi_ev_address_space_dispatch(rgn_desc, function, region_offset, - ACPI_MUL_8(obj_desc->common_field. - access_byte_width), - value); + status = acpi_ev_address_space_dispatch(rgn_desc, obj_desc, + function, region_offset, + ACPI_MUL_8(obj_desc-> + common_field. + access_byte_width), + value); if (ACPI_FAILURE(status)) { if (status == AE_NOT_IMPLEMENTED) { diff --git a/drivers/acpi/acpica/exprep.c b/drivers/acpi/acpica/exprep.c index cc95e2000406..18b75fed626e 100644 --- a/drivers/acpi/acpica/exprep.c +++ b/drivers/acpi/acpica/exprep.c @@ -47,6 +47,7 @@ #include "acinterp.h" #include "amlcode.h" #include "acnamesp.h" +#include "acdispat.h" #define _COMPONENT ACPI_EXECUTER ACPI_MODULE_NAME("exprep") @@ -455,6 +456,30 @@ acpi_status acpi_ex_prep_field_value(struct acpi_create_field_info *info) obj_desc->field.region_obj = acpi_ns_get_attached_object(info->region_node); + /* Fields specific to generic_serial_bus fields */ + + obj_desc->field.access_length = info->access_length; + + if (info->connection_node) { + second_desc = info->connection_node->object; + if (!(second_desc->common.flags & AOPOBJ_DATA_VALID)) { + status = + acpi_ds_get_buffer_arguments(second_desc); + if (ACPI_FAILURE(status)) { + acpi_ut_delete_object_desc(obj_desc); + return_ACPI_STATUS(status); + } + } + + obj_desc->field.resource_buffer = + second_desc->buffer.pointer; + obj_desc->field.resource_length = + (u16)second_desc->buffer.length; + } else if (info->resource_buffer) { + obj_desc->field.resource_buffer = info->resource_buffer; + obj_desc->field.resource_length = info->resource_length; + } + /* Allow full data read from EC address space */ if ((obj_desc->field.region_obj->region.space_id == diff --git a/drivers/acpi/acpica/psargs.c b/drivers/acpi/acpica/psargs.c index e1fad0ee0136..4037f9032481 100644 --- a/drivers/acpi/acpica/psargs.c +++ b/drivers/acpi/acpica/psargs.c @@ -484,34 +484,54 @@ acpi_ps_get_next_simple_arg(struct acpi_parse_state *parser_state, static union acpi_parse_object *acpi_ps_get_next_field(struct acpi_parse_state *parser_state) { - u32 aml_offset = (u32) - ACPI_PTR_DIFF(parser_state->aml, - parser_state->aml_start); + u32 aml_offset; union acpi_parse_object *field; + union acpi_parse_object *arg = NULL; u16 opcode; u32 name; + u8 access_type; + u8 access_attribute; + u8 access_length; + u32 pkg_length; + u8 *pkg_end; + u32 buffer_length; ACPI_FUNCTION_TRACE(ps_get_next_field); + aml_offset = + (u32)ACPI_PTR_DIFF(parser_state->aml, parser_state->aml_start); + /* Determine field type */ switch (ACPI_GET8(parser_state->aml)) { - default: - - opcode = AML_INT_NAMEDFIELD_OP; - break; - - case 0x00: + case AML_FIELD_OFFSET_OP: opcode = AML_INT_RESERVEDFIELD_OP; parser_state->aml++; break; - case 0x01: + case AML_FIELD_ACCESS_OP: opcode = AML_INT_ACCESSFIELD_OP; parser_state->aml++; break; + + case AML_FIELD_CONNECTION_OP: + + opcode = AML_INT_CONNECTION_OP; + parser_state->aml++; + break; + + case AML_FIELD_EXT_ACCESS_OP: + + opcode = AML_INT_EXTACCESSFIELD_OP; + parser_state->aml++; + break; + + default: + + opcode = AML_INT_NAMEDFIELD_OP; + break; } /* Allocate a new field op */ @@ -549,16 +569,111 @@ static union acpi_parse_object *acpi_ps_get_next_field(struct acpi_parse_state break; case AML_INT_ACCESSFIELD_OP: + case AML_INT_EXTACCESSFIELD_OP: /* * Get access_type and access_attrib and merge into the field Op - * access_type is first operand, access_attribute is second + * access_type is first operand, access_attribute is second. stuff + * these bytes into the node integer value for convenience. */ - field->common.value.integer = - (((u32) ACPI_GET8(parser_state->aml) << 8)); + + /* Get the two bytes (Type/Attribute) */ + + access_type = ACPI_GET8(parser_state->aml); parser_state->aml++; - field->common.value.integer |= ACPI_GET8(parser_state->aml); + access_attribute = ACPI_GET8(parser_state->aml); parser_state->aml++; + + field->common.value.integer = (u8)access_type; + field->common.value.integer |= (u16)(access_attribute << 8); + + /* This opcode has a third byte, access_length */ + + if (opcode == AML_INT_EXTACCESSFIELD_OP) { + access_length = ACPI_GET8(parser_state->aml); + parser_state->aml++; + + field->common.value.integer |= + (u32)(access_length << 16); + } + break; + + case AML_INT_CONNECTION_OP: + + /* + * Argument for Connection operator can be either a Buffer + * (resource descriptor), or a name_string. + */ + if (ACPI_GET8(parser_state->aml) == AML_BUFFER_OP) { + parser_state->aml++; + + pkg_end = parser_state->aml; + pkg_length = + acpi_ps_get_next_package_length(parser_state); + pkg_end += pkg_length; + + if (parser_state->aml < pkg_end) { + + /* Non-empty list */ + + arg = acpi_ps_alloc_op(AML_INT_BYTELIST_OP); + if (!arg) { + return_PTR(NULL); + } + + /* Get the actual buffer length argument */ + + opcode = ACPI_GET8(parser_state->aml); + parser_state->aml++; + + switch (opcode) { + case AML_BYTE_OP: /* AML_BYTEDATA_ARG */ + buffer_length = + ACPI_GET8(parser_state->aml); + parser_state->aml += 1; + break; + + case AML_WORD_OP: /* AML_WORDDATA_ARG */ + buffer_length = + ACPI_GET16(parser_state->aml); + parser_state->aml += 2; + break; + + case AML_DWORD_OP: /* AML_DWORDATA_ARG */ + buffer_length = + ACPI_GET32(parser_state->aml); + parser_state->aml += 4; + break; + + default: + buffer_length = 0; + break; + } + + /* Fill in bytelist data */ + + arg->named.value.size = buffer_length; + arg->named.data = parser_state->aml; + } + + /* Skip to End of byte data */ + + parser_state->aml = pkg_end; + } else { + arg = acpi_ps_alloc_op(AML_INT_NAMEPATH_OP); + if (!arg) { + return_PTR(NULL); + } + + /* Get the Namestring argument */ + + arg->common.value.name = + acpi_ps_get_next_namestring(parser_state); + } + + /* Link the buffer/namestring to parent (CONNECTION_OP) */ + + acpi_ps_append_arg(field, arg); break; default: diff --git a/drivers/acpi/acpica/psopcode.c b/drivers/acpi/acpica/psopcode.c index bed08de7528c..9b3191001b3d 100644 --- a/drivers/acpi/acpica/psopcode.c +++ b/drivers/acpi/acpica/psopcode.c @@ -638,7 +638,16 @@ const struct acpi_opcode_info acpi_gbl_aml_op_info[AML_NUM_OPCODES] = { /* 7E */ ACPI_OP("Timer", ARGP_TIMER_OP, ARGI_TIMER_OP, ACPI_TYPE_ANY, AML_CLASS_EXECUTE, AML_TYPE_EXEC_0A_0T_1R, - AML_FLAGS_EXEC_0A_0T_1R) + AML_FLAGS_EXEC_0A_0T_1R), + +/* ACPI 5.0 opcodes */ + +/* 7F */ ACPI_OP("-ConnectField-", ARGP_CONNECTFIELD_OP, + ARGI_CONNECTFIELD_OP, ACPI_TYPE_ANY, + AML_CLASS_INTERNAL, AML_TYPE_BOGUS, AML_HAS_ARGS), +/* 80 */ ACPI_OP("-ExtAccessField-", ARGP_CONNECTFIELD_OP, + ARGI_CONNECTFIELD_OP, ACPI_TYPE_ANY, + AML_CLASS_INTERNAL, AML_TYPE_BOGUS, 0) /*! [End] no source code translation !*/ }; @@ -657,7 +666,7 @@ static const u8 acpi_gbl_short_op_index[256] = { /* 0x20 */ _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, /* 0x28 */ _UNK, _UNK, _UNK, _UNK, _UNK, 0x63, _PFX, _PFX, /* 0x30 */ 0x67, 0x66, 0x68, 0x65, 0x69, 0x64, 0x6A, 0x7D, -/* 0x38 */ _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, +/* 0x38 */ 0x7F, 0x80, _UNK, _UNK, _UNK, _UNK, _UNK, _UNK, /* 0x40 */ _UNK, _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, /* 0x48 */ _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, /* 0x50 */ _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, _ASC, diff --git a/drivers/acpi/acpica/pstree.c b/drivers/acpi/acpica/pstree.c index f1464c03aa42..04311bbfcccb 100644 --- a/drivers/acpi/acpica/pstree.c +++ b/drivers/acpi/acpica/pstree.c @@ -74,6 +74,12 @@ union acpi_parse_object *acpi_ps_get_arg(union acpi_parse_object *op, u32 argn) ACPI_FUNCTION_ENTRY(); +/* + if (Op->Common.aml_opcode == AML_INT_CONNECTION_OP) + { + return (Op->Common.Value.Arg); + } +*/ /* Get the info structure for this opcode */ op_info = acpi_ps_get_opcode_info(op->common.aml_opcode); diff --git a/include/acpi/actypes.h b/include/acpi/actypes.h index ed73f6705c86..d088c902d6bd 100644 --- a/include/acpi/actypes.h +++ b/include/acpi/actypes.h @@ -957,6 +957,14 @@ acpi_status(*acpi_adr_space_handler) (u32 function, #define ACPI_DEFAULT_HANDLER NULL +/* Special Context data for generic_serial_bus/general_purpose_io (ACPI 5.0) */ + +struct acpi_connection_info { + u8 *connection; + u16 length; + u8 access_length; +}; + typedef acpi_status(*acpi_adr_space_setup) (acpi_handle region_handle, u32 function, From 2da120b6847f85c406f9afa13853e2755684389e Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 14:14:32 +0800 Subject: [PATCH 440/820] ACPI 5.0: Support for GeneralPurposeIo and GenericSerialBus operation region Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acconfig.h | 3 ++- drivers/acpi/acpica/amlcode.h | 17 ++++++++++------- drivers/acpi/acpica/exfield.c | 28 +++++++++++++++++++++------- drivers/acpi/acpica/exfldio.c | 3 ++- drivers/acpi/acpica/utdecode.c | 4 +++- include/acpi/actypes.h | 4 +++- 6 files changed, 41 insertions(+), 18 deletions(-) diff --git a/drivers/acpi/acpica/acconfig.h b/drivers/acpi/acpica/acconfig.h index f895a244ca7e..f467d49c1ace 100644 --- a/drivers/acpi/acpica/acconfig.h +++ b/drivers/acpi/acpica/acconfig.h @@ -202,9 +202,10 @@ #define ACPI_RSDP_CHECKSUM_LENGTH 20 #define ACPI_RSDP_XCHECKSUM_LENGTH 36 -/* SMBus and IPMI bidirectional buffer size */ +/* SMBus, GSBus and IPMI bidirectional buffer size */ #define ACPI_SMBUS_BUFFER_SIZE 34 +#define ACPI_GSBUS_BUFFER_SIZE 34 #define ACPI_IPMI_BUFFER_SIZE 66 /* _sx_d and _sx_w control methods */ diff --git a/drivers/acpi/acpica/amlcode.h b/drivers/acpi/acpica/amlcode.h index adf7a71ea766..c7db80ee61ae 100644 --- a/drivers/acpi/acpica/amlcode.h +++ b/drivers/acpi/acpica/amlcode.h @@ -466,13 +466,16 @@ typedef enum { * access_as keyword */ typedef enum { - AML_FIELD_ATTRIB_SMB_QUICK = 0x02, - AML_FIELD_ATTRIB_SMB_SEND_RCV = 0x04, - AML_FIELD_ATTRIB_SMB_BYTE = 0x06, - AML_FIELD_ATTRIB_SMB_WORD = 0x08, - AML_FIELD_ATTRIB_SMB_BLOCK = 0x0A, - AML_FIELD_ATTRIB_SMB_WORD_CALL = 0x0C, - AML_FIELD_ATTRIB_SMB_BLOCK_CALL = 0x0D + AML_FIELD_ATTRIB_QUICK = 0x02, + AML_FIELD_ATTRIB_SEND_RCV = 0x04, + AML_FIELD_ATTRIB_BYTE = 0x06, + AML_FIELD_ATTRIB_WORD = 0x08, + AML_FIELD_ATTRIB_BLOCK = 0x0A, + AML_FIELD_ATTRIB_MULTIBYTE = 0x0B, + AML_FIELD_ATTRIB_WORD_CALL = 0x0C, + AML_FIELD_ATTRIB_BLOCK_CALL = 0x0D, + AML_FIELD_ATTRIB_RAW_BYTES = 0x0E, + AML_FIELD_ATTRIB_RAW_PROCESS = 0x0F } AML_ACCESS_ATTRIBUTE; /* Bit fields in the AML method_flags byte */ diff --git a/drivers/acpi/acpica/exfield.c b/drivers/acpi/acpica/exfield.c index 0bde2230c028..3aeec028d71c 100644 --- a/drivers/acpi/acpica/exfield.c +++ b/drivers/acpi/acpica/exfield.c @@ -100,18 +100,25 @@ acpi_ex_read_data_from_field(struct acpi_walk_state *walk_state, (obj_desc->field.region_obj->region.space_id == ACPI_ADR_SPACE_SMBUS || obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_GSBUS + || obj_desc->field.region_obj->region.space_id == ACPI_ADR_SPACE_IPMI)) { /* - * This is an SMBus or IPMI read. We must create a buffer to hold + * This is an SMBus, GSBus or IPMI read. We must create a buffer to hold * the data and then directly access the region handler. * - * Note: Smbus protocol value is passed in upper 16-bits of Function + * Note: SMBus and GSBus protocol value is passed in upper 16-bits of Function */ if (obj_desc->field.region_obj->region.space_id == ACPI_ADR_SPACE_SMBUS) { length = ACPI_SMBUS_BUFFER_SIZE; function = ACPI_READ | (obj_desc->field.attribute << 16); + } else if (obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_GSBUS) { + length = ACPI_GSBUS_BUFFER_SIZE; + function = + ACPI_READ | (obj_desc->field.attribute << 16); } else { /* IPMI */ length = ACPI_IPMI_BUFFER_SIZE; @@ -248,21 +255,23 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc, (obj_desc->field.region_obj->region.space_id == ACPI_ADR_SPACE_SMBUS || obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_GSBUS + || obj_desc->field.region_obj->region.space_id == ACPI_ADR_SPACE_IPMI)) { /* - * This is an SMBus or IPMI write. We will bypass the entire field + * This is an SMBus, GSBus or IPMI write. We will bypass the entire field * mechanism and handoff the buffer directly to the handler. For * these address spaces, the buffer is bi-directional; on a write, * return data is returned in the same buffer. * * Source must be a buffer of sufficient size: - * ACPI_SMBUS_BUFFER_SIZE or ACPI_IPMI_BUFFER_SIZE. + * ACPI_SMBUS_BUFFER_SIZE, ACPI_GSBUS_BUFFER_SIZE, or ACPI_IPMI_BUFFER_SIZE. * - * Note: SMBus protocol type is passed in upper 16-bits of Function + * Note: SMBus and GSBus protocol type is passed in upper 16-bits of Function */ if (source_desc->common.type != ACPI_TYPE_BUFFER) { ACPI_ERROR((AE_INFO, - "SMBus or IPMI write requires Buffer, found type %s", + "SMBus/IPMI/GenericSerialBus write requires Buffer, found type %s", acpi_ut_get_object_type_name(source_desc))); return_ACPI_STATUS(AE_AML_OPERAND_TYPE); @@ -273,6 +282,11 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc, length = ACPI_SMBUS_BUFFER_SIZE; function = ACPI_WRITE | (obj_desc->field.attribute << 16); + } else if (obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_GSBUS) { + length = ACPI_GSBUS_BUFFER_SIZE; + function = + ACPI_WRITE | (obj_desc->field.attribute << 16); } else { /* IPMI */ length = ACPI_IPMI_BUFFER_SIZE; @@ -281,7 +295,7 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc, if (source_desc->buffer.length < length) { ACPI_ERROR((AE_INFO, - "SMBus or IPMI write requires Buffer of length %u, found length %u", + "SMBus/IPMI/GenericSerialBus write requires Buffer of length %u, found length %u", length, source_desc->buffer.length)); return_ACPI_STATUS(AE_AML_BUFFER_LIMIT); diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index cc9c5dfa4c45..5b76c11418fa 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -131,10 +131,11 @@ acpi_ex_setup_region(union acpi_operand_object *obj_desc, } /* - * Exit now for SMBus or IPMI address space, it has a non-linear + * Exit now for SMBus, GSBus or IPMI address space, it has a non-linear * address space and the request cannot be directly validated */ if (space_id == ACPI_ADR_SPACE_SMBUS || + space_id == ACPI_ADR_SPACE_GSBUS || space_id == ACPI_ADR_SPACE_IPMI) { /* SMBus or IPMI has a non-linear address space */ diff --git a/drivers/acpi/acpica/utdecode.c b/drivers/acpi/acpica/utdecode.c index 8b087e2d64f4..6fd56d549296 100644 --- a/drivers/acpi/acpica/utdecode.c +++ b/drivers/acpi/acpica/utdecode.c @@ -171,7 +171,9 @@ const char *acpi_gbl_region_types[ACPI_NUM_PREDEFINED_REGIONS] = { "SMBus", "SystemCMOS", "PCIBARTarget", - "IPMI" + "IPMI", + "GeneralPurposeIo", + "GenericSerialBus" }; char *acpi_ut_get_region_name(u8 space_id) diff --git a/include/acpi/actypes.h b/include/acpi/actypes.h index d088c902d6bd..d5dee7ce9474 100644 --- a/include/acpi/actypes.h +++ b/include/acpi/actypes.h @@ -712,8 +712,10 @@ typedef u8 acpi_adr_space_type; #define ACPI_ADR_SPACE_CMOS (acpi_adr_space_type) 5 #define ACPI_ADR_SPACE_PCI_BAR_TARGET (acpi_adr_space_type) 6 #define ACPI_ADR_SPACE_IPMI (acpi_adr_space_type) 7 +#define ACPI_ADR_SPACE_GPIO (acpi_adr_space_type) 8 +#define ACPI_ADR_SPACE_GSBUS (acpi_adr_space_type) 9 -#define ACPI_NUM_PREDEFINED_REGIONS 8 +#define ACPI_NUM_PREDEFINED_REGIONS 10 /* * Special Address Spaces From e0fe0a8d4ed5474261d0ee1452f5d9ae77236958 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Wed, 16 Nov 2011 14:38:13 +0800 Subject: [PATCH 441/820] ACPI 5.0: Support for all new resource descriptors FixedDMA, GPIO descriptors, SerialBus descriptors Signed-off-by: Lin Ming Signed-off-by: Bob Moore Signed-off-by: Len Brown --- drivers/acpi/acpica/Makefile | 2 +- drivers/acpi/acpica/aclocal.h | 6 +- drivers/acpi/acpica/acresrc.h | 109 +++++--- drivers/acpi/acpica/acutils.h | 1 + drivers/acpi/acpica/amlresrc.h | 136 ++++++++++ drivers/acpi/acpica/rscalc.c | 87 ++++++- drivers/acpi/acpica/rscreate.c | 3 +- drivers/acpi/acpica/rsdump.c | 194 ++++++++++++++- drivers/acpi/acpica/rsinfo.c | 56 ++++- drivers/acpi/acpica/rsirq.c | 31 +++ drivers/acpi/acpica/rslist.c | 75 ++++-- drivers/acpi/acpica/rsmisc.c | 267 +++++++++++++++++++- drivers/acpi/acpica/rsserial.c | 441 +++++++++++++++++++++++++++++++++ drivers/acpi/acpica/rsutils.c | 4 + drivers/acpi/acpica/utresrc.c | 276 ++++++++++++++++++--- include/acpi/acrestyp.h | 207 +++++++++++++++- 16 files changed, 1776 insertions(+), 119 deletions(-) create mode 100644 drivers/acpi/acpica/rsserial.c diff --git a/drivers/acpi/acpica/Makefile b/drivers/acpi/acpica/Makefile index 1bd466113dc8..b0af3b83eda8 100644 --- a/drivers/acpi/acpica/Makefile +++ b/drivers/acpi/acpica/Makefile @@ -36,7 +36,7 @@ acpi-y += psargs.o psparse.o psloop.o pstree.o pswalk.o \ psopcode.o psscope.o psutils.o psxface.o acpi-y += rsaddr.o rscreate.o rsinfo.o rsio.o rslist.o rsmisc.o rsxface.o \ - rscalc.o rsirq.o rsmemory.o rsutils.o + rscalc.o rsirq.o rsmemory.o rsutils.o rsserial.o acpi-$(ACPI_FUTURE_USAGE) += rsdump.o diff --git a/drivers/acpi/acpica/aclocal.h b/drivers/acpi/acpica/aclocal.h index 52c4ebddbcf1..60aee56acc5b 100644 --- a/drivers/acpi/acpica/aclocal.h +++ b/drivers/acpi/acpica/aclocal.h @@ -955,7 +955,7 @@ struct acpi_port_info { #define ACPI_RESOURCE_NAME_END_DEPENDENT 0x38 #define ACPI_RESOURCE_NAME_IO 0x40 #define ACPI_RESOURCE_NAME_FIXED_IO 0x48 -#define ACPI_RESOURCE_NAME_RESERVED_S1 0x50 +#define ACPI_RESOURCE_NAME_FIXED_DMA 0x50 #define ACPI_RESOURCE_NAME_RESERVED_S2 0x58 #define ACPI_RESOURCE_NAME_RESERVED_S3 0x60 #define ACPI_RESOURCE_NAME_RESERVED_S4 0x68 @@ -977,7 +977,9 @@ struct acpi_port_info { #define ACPI_RESOURCE_NAME_EXTENDED_IRQ 0x89 #define ACPI_RESOURCE_NAME_ADDRESS64 0x8A #define ACPI_RESOURCE_NAME_EXTENDED_ADDRESS64 0x8B -#define ACPI_RESOURCE_NAME_LARGE_MAX 0x8B +#define ACPI_RESOURCE_NAME_GPIO 0x8C +#define ACPI_RESOURCE_NAME_SERIAL_BUS 0x8E +#define ACPI_RESOURCE_NAME_LARGE_MAX 0x8E /***************************************************************************** * diff --git a/drivers/acpi/acpica/acresrc.h b/drivers/acpi/acpica/acresrc.h index f08b55b7f3a0..05721b023f88 100644 --- a/drivers/acpi/acpica/acresrc.h +++ b/drivers/acpi/acpica/acresrc.h @@ -73,28 +73,40 @@ typedef const struct acpi_rsconvert_info { /* Resource conversion opcodes */ -#define ACPI_RSC_INITGET 0 -#define ACPI_RSC_INITSET 1 -#define ACPI_RSC_FLAGINIT 2 -#define ACPI_RSC_1BITFLAG 3 -#define ACPI_RSC_2BITFLAG 4 -#define ACPI_RSC_COUNT 5 -#define ACPI_RSC_COUNT16 6 -#define ACPI_RSC_LENGTH 7 -#define ACPI_RSC_MOVE8 8 -#define ACPI_RSC_MOVE16 9 -#define ACPI_RSC_MOVE32 10 -#define ACPI_RSC_MOVE64 11 -#define ACPI_RSC_SET8 12 -#define ACPI_RSC_DATA8 13 -#define ACPI_RSC_ADDRESS 14 -#define ACPI_RSC_SOURCE 15 -#define ACPI_RSC_SOURCEX 16 -#define ACPI_RSC_BITMASK 17 -#define ACPI_RSC_BITMASK16 18 -#define ACPI_RSC_EXIT_NE 19 -#define ACPI_RSC_EXIT_LE 20 -#define ACPI_RSC_EXIT_EQ 21 +typedef enum { + ACPI_RSC_INITGET = 0, + ACPI_RSC_INITSET, + ACPI_RSC_FLAGINIT, + ACPI_RSC_1BITFLAG, + ACPI_RSC_2BITFLAG, + ACPI_RSC_3BITFLAG, + ACPI_RSC_ADDRESS, + ACPI_RSC_BITMASK, + ACPI_RSC_BITMASK16, + ACPI_RSC_COUNT, + ACPI_RSC_COUNT16, + ACPI_RSC_COUNT_GPIO_PIN, + ACPI_RSC_COUNT_GPIO_RES, + ACPI_RSC_COUNT_GPIO_VEN, + ACPI_RSC_COUNT_SERIAL_RES, + ACPI_RSC_COUNT_SERIAL_VEN, + ACPI_RSC_DATA8, + ACPI_RSC_EXIT_EQ, + ACPI_RSC_EXIT_LE, + ACPI_RSC_EXIT_NE, + ACPI_RSC_LENGTH, + ACPI_RSC_MOVE_GPIO_PIN, + ACPI_RSC_MOVE_GPIO_RES, + ACPI_RSC_MOVE_SERIAL_RES, + ACPI_RSC_MOVE_SERIAL_VEN, + ACPI_RSC_MOVE8, + ACPI_RSC_MOVE16, + ACPI_RSC_MOVE32, + ACPI_RSC_MOVE64, + ACPI_RSC_SET8, + ACPI_RSC_SOURCE, + ACPI_RSC_SOURCEX +} ACPI_RSCONVERT_OPCODES; /* Resource Conversion sub-opcodes */ @@ -106,6 +118,9 @@ typedef const struct acpi_rsconvert_info { #define ACPI_RS_OFFSET(f) (u8) ACPI_OFFSET (struct acpi_resource,f) #define AML_OFFSET(f) (u8) ACPI_OFFSET (union aml_resource,f) +/* + * Individual entry for the resource dump tables + */ typedef const struct acpi_rsdump_info { u8 opcode; u8 offset; @@ -116,20 +131,25 @@ typedef const struct acpi_rsdump_info { /* Values for the Opcode field above */ -#define ACPI_RSD_TITLE 0 -#define ACPI_RSD_LITERAL 1 -#define ACPI_RSD_STRING 2 -#define ACPI_RSD_UINT8 3 -#define ACPI_RSD_UINT16 4 -#define ACPI_RSD_UINT32 5 -#define ACPI_RSD_UINT64 6 -#define ACPI_RSD_1BITFLAG 7 -#define ACPI_RSD_2BITFLAG 8 -#define ACPI_RSD_SHORTLIST 9 -#define ACPI_RSD_LONGLIST 10 -#define ACPI_RSD_DWORDLIST 11 -#define ACPI_RSD_ADDRESS 12 -#define ACPI_RSD_SOURCE 13 +typedef enum { + ACPI_RSD_TITLE = 0, + ACPI_RSD_1BITFLAG, + ACPI_RSD_2BITFLAG, + ACPI_RSD_3BITFLAG, + ACPI_RSD_ADDRESS, + ACPI_RSD_DWORDLIST, + ACPI_RSD_LITERAL, + ACPI_RSD_LONGLIST, + ACPI_RSD_SHORTLIST, + ACPI_RSD_SHORTLISTX, + ACPI_RSD_SOURCE, + ACPI_RSD_STRING, + ACPI_RSD_UINT8, + ACPI_RSD_UINT16, + ACPI_RSD_UINT32, + ACPI_RSD_UINT64, + ACPI_RSD_WORDLIST +} ACPI_RSDUMP_OPCODES; /* restore default alignment */ @@ -138,13 +158,18 @@ typedef const struct acpi_rsdump_info { /* Resource tables indexed by internal resource type */ extern const u8 acpi_gbl_aml_resource_sizes[]; +extern const u8 acpi_gbl_aml_resource_serial_bus_sizes[]; extern struct acpi_rsconvert_info *acpi_gbl_set_resource_dispatch[]; /* Resource tables indexed by raw AML resource descriptor type */ extern const u8 acpi_gbl_resource_struct_sizes[]; +extern const u8 acpi_gbl_resource_struct_serial_bus_sizes[]; extern struct acpi_rsconvert_info *acpi_gbl_get_resource_dispatch[]; +extern struct acpi_rsconvert_info + *acpi_gbl_convert_resource_serial_bus_dispatch[]; + struct acpi_vendor_walk_info { struct acpi_vendor_uuid *uuid; struct acpi_buffer *buffer; @@ -293,6 +318,11 @@ extern struct acpi_rsconvert_info acpi_rs_convert_address16[]; extern struct acpi_rsconvert_info acpi_rs_convert_ext_irq[]; extern struct acpi_rsconvert_info acpi_rs_convert_address64[]; extern struct acpi_rsconvert_info acpi_rs_convert_ext_address64[]; +extern struct acpi_rsconvert_info acpi_rs_convert_gpio[]; +extern struct acpi_rsconvert_info acpi_rs_convert_fixed_dma[]; +extern struct acpi_rsconvert_info acpi_rs_convert_i2c_serial_bus[]; +extern struct acpi_rsconvert_info acpi_rs_convert_spi_serial_bus[]; +extern struct acpi_rsconvert_info acpi_rs_convert_uart_serial_bus[]; /* These resources require separate get/set tables */ @@ -310,6 +340,7 @@ extern struct acpi_rsconvert_info acpi_rs_set_vendor[]; * rsinfo */ extern struct acpi_rsdump_info *acpi_gbl_dump_resource_dispatch[]; +extern struct acpi_rsdump_info *acpi_gbl_dump_serial_bus_dispatch[]; /* * rsdump @@ -331,6 +362,12 @@ extern struct acpi_rsdump_info acpi_rs_dump_address64[]; extern struct acpi_rsdump_info acpi_rs_dump_ext_address64[]; extern struct acpi_rsdump_info acpi_rs_dump_ext_irq[]; extern struct acpi_rsdump_info acpi_rs_dump_generic_reg[]; +extern struct acpi_rsdump_info acpi_rs_dump_gpio[]; +extern struct acpi_rsdump_info acpi_rs_dump_fixed_dma[]; +extern struct acpi_rsdump_info acpi_rs_dump_common_serial_bus[]; +extern struct acpi_rsdump_info acpi_rs_dump_i2c_serial_bus[]; +extern struct acpi_rsdump_info acpi_rs_dump_spi_serial_bus[]; +extern struct acpi_rsdump_info acpi_rs_dump_uart_serial_bus[]; #endif #endif /* __ACRESRC_H__ */ diff --git a/drivers/acpi/acpica/acutils.h b/drivers/acpi/acpica/acutils.h index 99c140d8e348..ab95d452f87d 100644 --- a/drivers/acpi/acpica/acutils.h +++ b/drivers/acpi/acpica/acutils.h @@ -45,6 +45,7 @@ #define _ACUTILS_H extern const u8 acpi_gbl_resource_aml_sizes[]; +extern const u8 acpi_gbl_resource_aml_serial_bus_sizes[]; /* Strings used by the disassembler and debugger resource dump routines */ diff --git a/drivers/acpi/acpica/amlresrc.h b/drivers/acpi/acpica/amlresrc.h index 59122cde247c..d4c30511aca9 100644 --- a/drivers/acpi/acpica/amlresrc.h +++ b/drivers/acpi/acpica/amlresrc.h @@ -58,29 +58,48 @@ #define ACPI_RESTAG_TYPESPECIFICATTRIBUTES "_ATT" #define ACPI_RESTAG_BASEADDRESS "_BAS" #define ACPI_RESTAG_BUSMASTER "_BM_" /* Master(1), Slave(0) */ +#define ACPI_RESTAG_DEBOUNCETIME "_DBT" #define ACPI_RESTAG_DECODE "_DEC" +#define ACPI_RESTAG_DEVICEPOLARITY "_DPL" #define ACPI_RESTAG_DMA "_DMA" #define ACPI_RESTAG_DMATYPE "_TYP" /* Compatible(0), A(1), B(2), F(3) */ +#define ACPI_RESTAG_DRIVESTRENGTH "_DRS" +#define ACPI_RESTAG_ENDIANNESS "_END" +#define ACPI_RESTAG_FLOWCONTROL "_FLC" #define ACPI_RESTAG_GRANULARITY "_GRA" #define ACPI_RESTAG_INTERRUPT "_INT" #define ACPI_RESTAG_INTERRUPTLEVEL "_LL_" /* active_lo(1), active_hi(0) */ #define ACPI_RESTAG_INTERRUPTSHARE "_SHR" /* Shareable(1), no_share(0) */ #define ACPI_RESTAG_INTERRUPTTYPE "_HE_" /* Edge(1), Level(0) */ +#define ACPI_RESTAG_IORESTRICTION "_IOR" #define ACPI_RESTAG_LENGTH "_LEN" +#define ACPI_RESTAG_LINE "_LIN" #define ACPI_RESTAG_MEMATTRIBUTES "_MTP" /* Memory(0), Reserved(1), ACPI(2), NVS(3) */ #define ACPI_RESTAG_MEMTYPE "_MEM" /* non_cache(0), Cacheable(1) Cache+combine(2), Cache+prefetch(3) */ #define ACPI_RESTAG_MAXADDR "_MAX" #define ACPI_RESTAG_MINADDR "_MIN" #define ACPI_RESTAG_MAXTYPE "_MAF" #define ACPI_RESTAG_MINTYPE "_MIF" +#define ACPI_RESTAG_MODE "_MOD" +#define ACPI_RESTAG_PARITY "_PAR" +#define ACPI_RESTAG_PHASE "_PHA" +#define ACPI_RESTAG_PIN "_PIN" +#define ACPI_RESTAG_PINCONFIG "_PPI" +#define ACPI_RESTAG_POLARITY "_POL" #define ACPI_RESTAG_REGISTERBITOFFSET "_RBO" #define ACPI_RESTAG_REGISTERBITWIDTH "_RBW" #define ACPI_RESTAG_RANGETYPE "_RNG" #define ACPI_RESTAG_READWRITETYPE "_RW_" /* read_only(0), Writeable (1) */ +#define ACPI_RESTAG_LENGTH_RX "_RXL" +#define ACPI_RESTAG_LENGTH_TX "_TXL" +#define ACPI_RESTAG_SLAVEMODE "_SLV" +#define ACPI_RESTAG_SPEED "_SPE" +#define ACPI_RESTAG_STOPBITS "_STB" #define ACPI_RESTAG_TRANSLATION "_TRA" #define ACPI_RESTAG_TRANSTYPE "_TRS" /* Sparse(1), Dense(0) */ #define ACPI_RESTAG_TYPE "_TTP" /* Translation(1), Static (0) */ #define ACPI_RESTAG_XFERTYPE "_SIZ" /* 8(0), 8_and16(1), 16(2) */ +#define ACPI_RESTAG_VENDORDATA "_VEN" /* Default sizes for "small" resource descriptors */ @@ -90,6 +109,7 @@ #define ASL_RDESC_END_DEPEND_SIZE 0x00 #define ASL_RDESC_IO_SIZE 0x07 #define ASL_RDESC_FIXED_IO_SIZE 0x03 +#define ASL_RDESC_FIXED_DMA_SIZE 0x05 #define ASL_RDESC_END_TAG_SIZE 0x01 struct asl_resource_node { @@ -164,6 +184,12 @@ struct aml_resource_end_tag { AML_RESOURCE_SMALL_HEADER_COMMON u8 checksum; }; +struct aml_resource_fixed_dma { + AML_RESOURCE_SMALL_HEADER_COMMON u16 request_lines; + u16 channels; + u8 width; +}; + /* * LARGE descriptors */ @@ -263,6 +289,110 @@ struct aml_resource_generic_register { u64 address; }; +/* Common descriptor for gpio_int and gpio_io (ACPI 5.0) */ + +struct aml_resource_gpio { + AML_RESOURCE_LARGE_HEADER_COMMON u8 revision_id; + u8 connection_type; + u16 flags; + u16 int_flags; + u8 pin_config; + u16 drive_strength; + u16 debounce_timeout; + u16 pin_table_offset; + u8 res_source_index; + u16 res_source_offset; + u16 vendor_offset; + u16 vendor_length; + /* + * Optional fields follow immediately: + * 1) PIN list (Words) + * 2) Resource Source String + * 3) Vendor Data bytes + */ +}; + +#define AML_RESOURCE_GPIO_REVISION 1 /* ACPI 5.0 */ + +/* Values for connection_type above */ + +#define AML_RESOURCE_GPIO_TYPE_INT 0 +#define AML_RESOURCE_GPIO_TYPE_IO 1 +#define AML_RESOURCE_MAX_GPIOTYPE 1 + +/* Common preamble for all serial descriptors (ACPI 5.0) */ + +#define AML_RESOURCE_SERIAL_COMMON \ + u8 revision_id; \ + u8 res_source_index; \ + u8 type; \ + u8 flags; \ + u16 type_specific_flags; \ + u8 type_revision_id; \ + u16 type_data_length; \ + +/* Values for the type field above */ + +#define AML_RESOURCE_I2C_SERIALBUSTYPE 1 +#define AML_RESOURCE_SPI_SERIALBUSTYPE 2 +#define AML_RESOURCE_UART_SERIALBUSTYPE 3 +#define AML_RESOURCE_MAX_SERIALBUSTYPE 3 +#define AML_RESOURCE_VENDOR_SERIALBUSTYPE 192 /* Vendor defined is 0xC0-0xFF (NOT SUPPORTED) */ + +struct aml_resource_common_serialbus { +AML_RESOURCE_LARGE_HEADER_COMMON AML_RESOURCE_SERIAL_COMMON}; + +struct aml_resource_i2c_serialbus { + AML_RESOURCE_LARGE_HEADER_COMMON + AML_RESOURCE_SERIAL_COMMON u32 connection_speed; + u16 slave_address; + /* + * Optional fields follow immediately: + * 1) Vendor Data bytes + * 2) Resource Source String + */ +}; + +#define AML_RESOURCE_I2C_REVISION 1 /* ACPI 5.0 */ +#define AML_RESOURCE_I2C_TYPE_REVISION 1 /* ACPI 5.0 */ +#define AML_RESOURCE_I2C_MIN_DATA_LEN 6 + +struct aml_resource_spi_serialbus { + AML_RESOURCE_LARGE_HEADER_COMMON + AML_RESOURCE_SERIAL_COMMON u32 connection_speed; + u8 data_bit_length; + u8 clock_phase; + u8 clock_polarity; + u16 device_selection; + /* + * Optional fields follow immediately: + * 1) Vendor Data bytes + * 2) Resource Source String + */ +}; + +#define AML_RESOURCE_SPI_REVISION 1 /* ACPI 5.0 */ +#define AML_RESOURCE_SPI_TYPE_REVISION 1 /* ACPI 5.0 */ +#define AML_RESOURCE_SPI_MIN_DATA_LEN 9 + +struct aml_resource_uart_serialbus { + AML_RESOURCE_LARGE_HEADER_COMMON + AML_RESOURCE_SERIAL_COMMON u32 default_baud_rate; + u16 rx_fifo_size; + u16 tx_fifo_size; + u8 parity; + u8 lines_enabled; + /* + * Optional fields follow immediately: + * 1) Vendor Data bytes + * 2) Resource Source String + */ +}; + +#define AML_RESOURCE_UART_REVISION 1 /* ACPI 5.0 */ +#define AML_RESOURCE_UART_TYPE_REVISION 1 /* ACPI 5.0 */ +#define AML_RESOURCE_UART_MIN_DATA_LEN 10 + /* restore default alignment */ #pragma pack() @@ -284,6 +414,7 @@ union aml_resource { struct aml_resource_end_dependent end_dpf; struct aml_resource_io io; struct aml_resource_fixed_io fixed_io; + struct aml_resource_fixed_dma fixed_dma; struct aml_resource_vendor_small vendor_small; struct aml_resource_end_tag end_tag; @@ -299,6 +430,11 @@ union aml_resource { struct aml_resource_address64 address64; struct aml_resource_extended_address64 ext_address64; struct aml_resource_extended_irq extended_irq; + struct aml_resource_gpio gpio; + struct aml_resource_i2c_serialbus i2c_serial_bus; + struct aml_resource_spi_serialbus spi_serial_bus; + struct aml_resource_uart_serialbus uart_serial_bus; + struct aml_resource_common_serialbus common_serial_bus; /* Utility overlays */ diff --git a/drivers/acpi/acpica/rscalc.c b/drivers/acpi/acpica/rscalc.c index 3a8a89ec2ca4..3a29b4dd4453 100644 --- a/drivers/acpi/acpica/rscalc.c +++ b/drivers/acpi/acpica/rscalc.c @@ -313,6 +313,38 @@ acpi_rs_get_aml_length(struct acpi_resource * resource, acpi_size * size_needed) resource_source)); break; + case ACPI_RESOURCE_TYPE_GPIO: + + total_size = + (acpi_rs_length) (total_size + + (resource->data.gpio. + pin_table_length * 2) + + resource->data.gpio. + resource_source.string_length + + resource->data.gpio. + vendor_length); + + break; + + case ACPI_RESOURCE_TYPE_SERIAL_BUS: + + total_size = + acpi_gbl_aml_resource_serial_bus_sizes[resource-> + data. + common_serial_bus. + type]; + + total_size = (acpi_rs_length) (total_size + + resource->data. + i2c_serial_bus. + resource_source. + string_length + + resource->data. + i2c_serial_bus. + vendor_length); + + break; + default: break; } @@ -362,10 +394,11 @@ acpi_rs_get_list_length(u8 * aml_buffer, u32 extra_struct_bytes; u8 resource_index; u8 minimum_aml_resource_length; + union aml_resource *aml_resource; ACPI_FUNCTION_TRACE(rs_get_list_length); - *size_needed = 0; + *size_needed = ACPI_RS_SIZE_MIN; /* Minimum size is one end_tag */ end_aml = aml_buffer + aml_buffer_length; /* Walk the list of AML resource descriptors */ @@ -376,9 +409,15 @@ acpi_rs_get_list_length(u8 * aml_buffer, status = acpi_ut_validate_resource(aml_buffer, &resource_index); if (ACPI_FAILURE(status)) { + /* + * Exit on failure. Cannot continue because the descriptor length + * may be bogus also. + */ return_ACPI_STATUS(status); } + aml_resource = (void *)aml_buffer; + /* Get the resource length and base (minimum) AML size */ resource_length = acpi_ut_get_resource_length(aml_buffer); @@ -422,10 +461,8 @@ acpi_rs_get_list_length(u8 * aml_buffer, case ACPI_RESOURCE_NAME_END_TAG: /* - * End Tag: - * This is the normal exit, add size of end_tag + * End Tag: This is the normal exit */ - *size_needed += ACPI_RS_SIZE_MIN; return_ACPI_STATUS(AE_OK); case ACPI_RESOURCE_NAME_ADDRESS32: @@ -457,6 +494,33 @@ acpi_rs_get_list_length(u8 * aml_buffer, minimum_aml_resource_length); break; + case ACPI_RESOURCE_NAME_GPIO: + + /* Vendor data is optional */ + + if (aml_resource->gpio.vendor_length) { + extra_struct_bytes += + aml_resource->gpio.vendor_offset - + aml_resource->gpio.pin_table_offset + + aml_resource->gpio.vendor_length; + } else { + extra_struct_bytes += + aml_resource->large_header.resource_length + + sizeof(struct aml_resource_large_header) - + aml_resource->gpio.pin_table_offset; + } + break; + + case ACPI_RESOURCE_NAME_SERIAL_BUS: + + minimum_aml_resource_length = + acpi_gbl_resource_aml_serial_bus_sizes + [aml_resource->common_serial_bus.type]; + extra_struct_bytes += + aml_resource->common_serial_bus.resource_length - + minimum_aml_resource_length; + break; + default: break; } @@ -467,9 +531,18 @@ acpi_rs_get_list_length(u8 * aml_buffer, * Important: Round the size up for the appropriate alignment. This * is a requirement on IA64. */ - buffer_size = acpi_gbl_resource_struct_sizes[resource_index] + - extra_struct_bytes; - buffer_size = (u32) ACPI_ROUND_UP_TO_NATIVE_WORD(buffer_size); + if (acpi_ut_get_resource_type(aml_buffer) == + ACPI_RESOURCE_NAME_SERIAL_BUS) { + buffer_size = + acpi_gbl_resource_struct_serial_bus_sizes + [aml_resource->common_serial_bus.type] + + extra_struct_bytes; + } else { + buffer_size = + acpi_gbl_resource_struct_sizes[resource_index] + + extra_struct_bytes; + } + buffer_size = (u32)ACPI_ROUND_UP_TO_NATIVE_WORD(buffer_size); *size_needed += buffer_size; diff --git a/drivers/acpi/acpica/rscreate.c b/drivers/acpi/acpica/rscreate.c index 4ce6e1147e80..f61285d5edc3 100644 --- a/drivers/acpi/acpica/rscreate.c +++ b/drivers/acpi/acpica/rscreate.c @@ -66,9 +66,10 @@ ACPI_MODULE_NAME("rscreate") * of device resources. * ******************************************************************************/ + acpi_status acpi_rs_create_resource_list(union acpi_operand_object *aml_buffer, - struct acpi_buffer *output_buffer) + struct acpi_buffer * output_buffer) { acpi_status status; diff --git a/drivers/acpi/acpica/rsdump.c b/drivers/acpi/acpica/rsdump.c index 33db7520c74b..9969985bf2b7 100644 --- a/drivers/acpi/acpica/rsdump.c +++ b/drivers/acpi/acpica/rsdump.c @@ -61,11 +61,13 @@ static void acpi_rs_out_integer64(char *title, u64 value); static void acpi_rs_out_title(char *title); -static void acpi_rs_dump_byte_list(u16 length, u8 * data); +static void acpi_rs_dump_byte_list(u16 length, u8 *data); -static void acpi_rs_dump_dword_list(u8 length, u32 * data); +static void acpi_rs_dump_word_list(u16 length, u16 *data); -static void acpi_rs_dump_short_byte_list(u8 length, u8 * data); +static void acpi_rs_dump_dword_list(u8 length, u32 *data); + +static void acpi_rs_dump_short_byte_list(u8 length, u8 *data); static void acpi_rs_dump_resource_source(struct acpi_resource_source *resource_source); @@ -309,6 +311,125 @@ struct acpi_rsdump_info acpi_rs_dump_generic_reg[6] = { {ACPI_RSD_UINT64, ACPI_RSD_OFFSET(generic_reg.address), "Address", NULL} }; +struct acpi_rsdump_info acpi_rs_dump_gpio[16] = { + {ACPI_RSD_TITLE, ACPI_RSD_TABLE_SIZE(acpi_rs_dump_gpio), "GPIO", NULL}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(gpio.revision_id), "RevisionId", NULL}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(gpio.connection_type), + "ConnectionType", acpi_gbl_ct_decode}, + {ACPI_RSD_1BITFLAG, ACPI_RSD_OFFSET(gpio.producer_consumer), + "ProducerConsumer", acpi_gbl_consume_decode}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(gpio.pin_config), "PinConfig", + acpi_gbl_ppc_decode}, + {ACPI_RSD_2BITFLAG, ACPI_RSD_OFFSET(gpio.sharable), "Sharable", + acpi_gbl_shr_decode}, + {ACPI_RSD_2BITFLAG, ACPI_RSD_OFFSET(gpio.io_restriction), + "IoRestriction", acpi_gbl_ior_decode}, + {ACPI_RSD_1BITFLAG, ACPI_RSD_OFFSET(gpio.triggering), "Triggering", + acpi_gbl_he_decode}, + {ACPI_RSD_2BITFLAG, ACPI_RSD_OFFSET(gpio.polarity), "Polarity", + acpi_gbl_ll_decode}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(gpio.drive_strength), "DriveStrength", + NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(gpio.debounce_timeout), + "DebounceTimeout", NULL}, + {ACPI_RSD_SOURCE, ACPI_RSD_OFFSET(gpio.resource_source), + "ResourceSource", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(gpio.pin_table_length), + "PinTableLength", NULL}, + {ACPI_RSD_WORDLIST, ACPI_RSD_OFFSET(gpio.pin_table), "PinTable", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(gpio.vendor_length), "VendorLength", + NULL}, + {ACPI_RSD_SHORTLISTX, ACPI_RSD_OFFSET(gpio.vendor_data), "VendorData", + NULL}, +}; + +struct acpi_rsdump_info acpi_rs_dump_fixed_dma[4] = { + {ACPI_RSD_TITLE, ACPI_RSD_TABLE_SIZE(acpi_rs_dump_fixed_dma), + "FixedDma", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(fixed_dma.request_lines), + "RequestLines", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(fixed_dma.channels), "Channels", + NULL}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(fixed_dma.width), "TransferWidth", + acpi_gbl_dts_decode}, +}; + +#define ACPI_RS_DUMP_COMMON_SERIAL_BUS \ + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET (common_serial_bus.revision_id), "RevisionId", NULL}, \ + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET (common_serial_bus.type), "Type", acpi_gbl_sbt_decode}, \ + {ACPI_RSD_1BITFLAG, ACPI_RSD_OFFSET (common_serial_bus.producer_consumer), "ProducerConsumer", acpi_gbl_consume_decode}, \ + {ACPI_RSD_1BITFLAG, ACPI_RSD_OFFSET (common_serial_bus.slave_mode), "SlaveMode", acpi_gbl_sm_decode}, \ + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET (common_serial_bus.type_revision_id), "TypeRevisionId", NULL}, \ + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET (common_serial_bus.type_data_length), "TypeDataLength", NULL}, \ + {ACPI_RSD_SOURCE, ACPI_RSD_OFFSET (common_serial_bus.resource_source), "ResourceSource", NULL}, \ + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET (common_serial_bus.vendor_length), "VendorLength", NULL}, \ + {ACPI_RSD_SHORTLISTX,ACPI_RSD_OFFSET (common_serial_bus.vendor_data), "VendorData", NULL}, + +struct acpi_rsdump_info acpi_rs_dump_common_serial_bus[10] = { + {ACPI_RSD_TITLE, ACPI_RSD_TABLE_SIZE(acpi_rs_dump_common_serial_bus), + "Common Serial Bus", NULL}, + ACPI_RS_DUMP_COMMON_SERIAL_BUS +}; + +struct acpi_rsdump_info acpi_rs_dump_i2c_serial_bus[13] = { + {ACPI_RSD_TITLE, ACPI_RSD_TABLE_SIZE(acpi_rs_dump_i2c_serial_bus), + "I2C Serial Bus", NULL}, + ACPI_RS_DUMP_COMMON_SERIAL_BUS {ACPI_RSD_1BITFLAG, + ACPI_RSD_OFFSET(i2c_serial_bus. + access_mode), + "AccessMode", acpi_gbl_am_decode}, + {ACPI_RSD_UINT32, ACPI_RSD_OFFSET(i2c_serial_bus.connection_speed), + "ConnectionSpeed", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(i2c_serial_bus.slave_address), + "SlaveAddress", NULL}, +}; + +struct acpi_rsdump_info acpi_rs_dump_spi_serial_bus[17] = { + {ACPI_RSD_TITLE, ACPI_RSD_TABLE_SIZE(acpi_rs_dump_spi_serial_bus), + "Spi Serial Bus", NULL}, + ACPI_RS_DUMP_COMMON_SERIAL_BUS {ACPI_RSD_1BITFLAG, + ACPI_RSD_OFFSET(spi_serial_bus. + wire_mode), "WireMode", + acpi_gbl_wm_decode}, + {ACPI_RSD_1BITFLAG, ACPI_RSD_OFFSET(spi_serial_bus.device_polarity), + "DevicePolarity", acpi_gbl_dp_decode}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(spi_serial_bus.data_bit_length), + "DataBitLength", NULL}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(spi_serial_bus.clock_phase), + "ClockPhase", acpi_gbl_cph_decode}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(spi_serial_bus.clock_polarity), + "ClockPolarity", acpi_gbl_cpo_decode}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(spi_serial_bus.device_selection), + "DeviceSelection", NULL}, + {ACPI_RSD_UINT32, ACPI_RSD_OFFSET(spi_serial_bus.connection_speed), + "ConnectionSpeed", NULL}, +}; + +struct acpi_rsdump_info acpi_rs_dump_uart_serial_bus[19] = { + {ACPI_RSD_TITLE, ACPI_RSD_TABLE_SIZE(acpi_rs_dump_uart_serial_bus), + "Uart Serial Bus", NULL}, + ACPI_RS_DUMP_COMMON_SERIAL_BUS {ACPI_RSD_2BITFLAG, + ACPI_RSD_OFFSET(uart_serial_bus. + flow_control), + "FlowControl", acpi_gbl_fc_decode}, + {ACPI_RSD_2BITFLAG, ACPI_RSD_OFFSET(uart_serial_bus.stop_bits), + "StopBits", acpi_gbl_sb_decode}, + {ACPI_RSD_3BITFLAG, ACPI_RSD_OFFSET(uart_serial_bus.data_bits), + "DataBits", acpi_gbl_bpb_decode}, + {ACPI_RSD_1BITFLAG, ACPI_RSD_OFFSET(uart_serial_bus.endian), "Endian", + acpi_gbl_ed_decode}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(uart_serial_bus.parity), "Parity", + acpi_gbl_pt_decode}, + {ACPI_RSD_UINT8, ACPI_RSD_OFFSET(uart_serial_bus.lines_enabled), + "LinesEnabled", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(uart_serial_bus.rx_fifo_size), + "RxFifoSize", NULL}, + {ACPI_RSD_UINT16, ACPI_RSD_OFFSET(uart_serial_bus.tx_fifo_size), + "TxFifoSize", NULL}, + {ACPI_RSD_UINT32, ACPI_RSD_OFFSET(uart_serial_bus.default_baud_rate), + "ConnectionSpeed", NULL}, +}; + /* * Tables used for common address descriptor flag fields */ @@ -413,7 +534,14 @@ acpi_rs_dump_descriptor(void *resource, struct acpi_rsdump_info *table) /* Data items, 8/16/32/64 bit */ case ACPI_RSD_UINT8: - acpi_rs_out_integer8(name, ACPI_GET8(target)); + if (table->pointer) { + acpi_rs_out_string(name, ACPI_CAST_PTR(char, + table-> + pointer + [*target])); + } else { + acpi_rs_out_integer8(name, ACPI_GET8(target)); + } break; case ACPI_RSD_UINT16: @@ -444,6 +572,13 @@ acpi_rs_dump_descriptor(void *resource, struct acpi_rsdump_info *table) 0x03])); break; + case ACPI_RSD_3BITFLAG: + acpi_rs_out_string(name, ACPI_CAST_PTR(char, + table-> + pointer[*target & + 0x07])); + break; + case ACPI_RSD_SHORTLIST: /* * Short byte list (single line output) for DMA and IRQ resources @@ -456,6 +591,20 @@ acpi_rs_dump_descriptor(void *resource, struct acpi_rsdump_info *table) } break; + case ACPI_RSD_SHORTLISTX: + /* + * Short byte list (single line output) for GPIO vendor data + * Note: The list length is obtained from the previous table entry + */ + if (previous_target) { + acpi_rs_out_title(name); + acpi_rs_dump_short_byte_list(*previous_target, + * + (ACPI_CAST_INDIRECT_PTR + (u8, target))); + } + break; + case ACPI_RSD_LONGLIST: /* * Long byte list for Vendor resource data @@ -480,6 +629,18 @@ acpi_rs_dump_descriptor(void *resource, struct acpi_rsdump_info *table) } break; + case ACPI_RSD_WORDLIST: + /* + * Word list for GPIO Pin Table + * Note: The list length is obtained from the previous table entry + */ + if (previous_target) { + acpi_rs_dump_word_list(*previous_target, + *(ACPI_CAST_INDIRECT_PTR + (u16, target))); + } + break; + case ACPI_RSD_ADDRESS: /* * Common flags for all Address resources @@ -627,14 +788,20 @@ void acpi_rs_dump_resource_list(struct acpi_resource *resource_list) /* Dump the resource descriptor */ - acpi_rs_dump_descriptor(&resource_list->data, - acpi_gbl_dump_resource_dispatch[type]); + if (type == ACPI_RESOURCE_TYPE_SERIAL_BUS) { + acpi_rs_dump_descriptor(&resource_list->data, + acpi_gbl_dump_serial_bus_dispatch + [resource_list->data. + common_serial_bus.type]); + } else { + acpi_rs_dump_descriptor(&resource_list->data, + acpi_gbl_dump_resource_dispatch + [type]); + } /* Point to the next resource structure */ - resource_list = - ACPI_ADD_PTR(struct acpi_resource, resource_list, - resource_list->length); + resource_list = ACPI_NEXT_RESOURCE(resource_list); /* Exit when END_TAG descriptor is reached */ @@ -768,4 +935,13 @@ static void acpi_rs_dump_dword_list(u8 length, u32 * data) } } +static void acpi_rs_dump_word_list(u16 length, u16 *data) +{ + u16 i; + + for (i = 0; i < length; i++) { + acpi_os_printf("%25s%2.2X : %4.4X\n", "Word", i, data[i]); + } +} + #endif diff --git a/drivers/acpi/acpica/rsinfo.c b/drivers/acpi/acpica/rsinfo.c index f9ea60872aa4..e5489bf475fa 100644 --- a/drivers/acpi/acpica/rsinfo.c +++ b/drivers/acpi/acpica/rsinfo.c @@ -76,7 +76,10 @@ struct acpi_rsconvert_info *acpi_gbl_set_resource_dispatch[] = { acpi_rs_convert_address64, /* 0x0D, ACPI_RESOURCE_TYPE_ADDRESS64 */ acpi_rs_convert_ext_address64, /* 0x0E, ACPI_RESOURCE_TYPE_EXTENDED_ADDRESS64 */ acpi_rs_convert_ext_irq, /* 0x0F, ACPI_RESOURCE_TYPE_EXTENDED_IRQ */ - acpi_rs_convert_generic_reg /* 0x10, ACPI_RESOURCE_TYPE_GENERIC_REGISTER */ + acpi_rs_convert_generic_reg, /* 0x10, ACPI_RESOURCE_TYPE_GENERIC_REGISTER */ + acpi_rs_convert_gpio, /* 0x11, ACPI_RESOURCE_TYPE_GPIO */ + acpi_rs_convert_fixed_dma, /* 0x12, ACPI_RESOURCE_TYPE_FIXED_DMA */ + NULL, /* 0x13, ACPI_RESOURCE_TYPE_SERIAL_BUS - Use subtype table below */ }; /* Dispatch tables for AML-to-resource (Get Resource) conversion functions */ @@ -94,7 +97,7 @@ struct acpi_rsconvert_info *acpi_gbl_get_resource_dispatch[] = { acpi_rs_convert_end_dpf, /* 0x07, ACPI_RESOURCE_NAME_END_DEPENDENT */ acpi_rs_convert_io, /* 0x08, ACPI_RESOURCE_NAME_IO */ acpi_rs_convert_fixed_io, /* 0x09, ACPI_RESOURCE_NAME_FIXED_IO */ - NULL, /* 0x0A, Reserved */ + acpi_rs_convert_fixed_dma, /* 0x0A, ACPI_RESOURCE_NAME_FIXED_DMA */ NULL, /* 0x0B, Reserved */ NULL, /* 0x0C, Reserved */ NULL, /* 0x0D, Reserved */ @@ -114,7 +117,19 @@ struct acpi_rsconvert_info *acpi_gbl_get_resource_dispatch[] = { acpi_rs_convert_address16, /* 0x08, ACPI_RESOURCE_NAME_ADDRESS16 */ acpi_rs_convert_ext_irq, /* 0x09, ACPI_RESOURCE_NAME_EXTENDED_IRQ */ acpi_rs_convert_address64, /* 0x0A, ACPI_RESOURCE_NAME_ADDRESS64 */ - acpi_rs_convert_ext_address64 /* 0x0B, ACPI_RESOURCE_NAME_EXTENDED_ADDRESS64 */ + acpi_rs_convert_ext_address64, /* 0x0B, ACPI_RESOURCE_NAME_EXTENDED_ADDRESS64 */ + acpi_rs_convert_gpio, /* 0x0C, ACPI_RESOURCE_NAME_GPIO */ + NULL, /* 0x0D, Reserved */ + NULL, /* 0x0E, ACPI_RESOURCE_NAME_SERIAL_BUS - Use subtype table below */ +}; + +/* Subtype table for serial_bus -- I2C, SPI, and UART */ + +struct acpi_rsconvert_info *acpi_gbl_convert_resource_serial_bus_dispatch[] = { + NULL, + acpi_rs_convert_i2c_serial_bus, + acpi_rs_convert_spi_serial_bus, + acpi_rs_convert_uart_serial_bus, }; #ifdef ACPI_FUTURE_USAGE @@ -140,6 +155,16 @@ struct acpi_rsdump_info *acpi_gbl_dump_resource_dispatch[] = { acpi_rs_dump_ext_address64, /* ACPI_RESOURCE_TYPE_EXTENDED_ADDRESS64 */ acpi_rs_dump_ext_irq, /* ACPI_RESOURCE_TYPE_EXTENDED_IRQ */ acpi_rs_dump_generic_reg, /* ACPI_RESOURCE_TYPE_GENERIC_REGISTER */ + acpi_rs_dump_gpio, /* ACPI_RESOURCE_TYPE_GPIO */ + acpi_rs_dump_fixed_dma, /* ACPI_RESOURCE_TYPE_FIXED_DMA */ + NULL, /* ACPI_RESOURCE_TYPE_SERIAL_BUS */ +}; + +struct acpi_rsdump_info *acpi_gbl_dump_serial_bus_dispatch[] = { + NULL, + acpi_rs_dump_i2c_serial_bus, /* AML_RESOURCE_I2C_BUS_TYPE */ + acpi_rs_dump_spi_serial_bus, /* AML_RESOURCE_SPI_BUS_TYPE */ + acpi_rs_dump_uart_serial_bus, /* AML_RESOURCE_UART_BUS_TYPE */ }; #endif @@ -166,7 +191,10 @@ const u8 acpi_gbl_aml_resource_sizes[] = { sizeof(struct aml_resource_address64), /* ACPI_RESOURCE_TYPE_ADDRESS64 */ sizeof(struct aml_resource_extended_address64), /*ACPI_RESOURCE_TYPE_EXTENDED_ADDRESS64 */ sizeof(struct aml_resource_extended_irq), /* ACPI_RESOURCE_TYPE_EXTENDED_IRQ */ - sizeof(struct aml_resource_generic_register) /* ACPI_RESOURCE_TYPE_GENERIC_REGISTER */ + sizeof(struct aml_resource_generic_register), /* ACPI_RESOURCE_TYPE_GENERIC_REGISTER */ + sizeof(struct aml_resource_gpio), /* ACPI_RESOURCE_TYPE_GPIO */ + sizeof(struct aml_resource_fixed_dma), /* ACPI_RESOURCE_TYPE_FIXED_DMA */ + sizeof(struct aml_resource_common_serialbus), /* ACPI_RESOURCE_TYPE_SERIAL_BUS */ }; const u8 acpi_gbl_resource_struct_sizes[] = { @@ -182,7 +210,7 @@ const u8 acpi_gbl_resource_struct_sizes[] = { ACPI_RS_SIZE_MIN, ACPI_RS_SIZE(struct acpi_resource_io), ACPI_RS_SIZE(struct acpi_resource_fixed_io), - 0, + ACPI_RS_SIZE(struct acpi_resource_fixed_dma), 0, 0, 0, @@ -202,5 +230,21 @@ const u8 acpi_gbl_resource_struct_sizes[] = { ACPI_RS_SIZE(struct acpi_resource_address16), ACPI_RS_SIZE(struct acpi_resource_extended_irq), ACPI_RS_SIZE(struct acpi_resource_address64), - ACPI_RS_SIZE(struct acpi_resource_extended_address64) + ACPI_RS_SIZE(struct acpi_resource_extended_address64), + ACPI_RS_SIZE(struct acpi_resource_gpio), + ACPI_RS_SIZE(struct acpi_resource_common_serialbus) +}; + +const u8 acpi_gbl_aml_resource_serial_bus_sizes[] = { + 0, + sizeof(struct aml_resource_i2c_serialbus), + sizeof(struct aml_resource_spi_serialbus), + sizeof(struct aml_resource_uart_serialbus), +}; + +const u8 acpi_gbl_resource_struct_serial_bus_sizes[] = { + 0, + ACPI_RS_SIZE(struct acpi_resource_i2c_serialbus), + ACPI_RS_SIZE(struct acpi_resource_spi_serialbus), + ACPI_RS_SIZE(struct acpi_resource_uart_serialbus), }; diff --git a/drivers/acpi/acpica/rsirq.c b/drivers/acpi/acpica/rsirq.c index 50b8ad211167..3ac9d4a08fcf 100644 --- a/drivers/acpi/acpica/rsirq.c +++ b/drivers/acpi/acpica/rsirq.c @@ -264,3 +264,34 @@ struct acpi_rsconvert_info acpi_rs_convert_dma[6] = { AML_OFFSET(dma.dma_channel_mask), ACPI_RS_OFFSET(data.dma.channel_count)} }; + +/******************************************************************************* + * + * acpi_rs_convert_fixed_dma + * + ******************************************************************************/ + +struct acpi_rsconvert_info acpi_rs_convert_fixed_dma[4] = { + {ACPI_RSC_INITGET, ACPI_RESOURCE_TYPE_FIXED_DMA, + ACPI_RS_SIZE(struct acpi_resource_fixed_dma), + ACPI_RSC_TABLE_SIZE(acpi_rs_convert_fixed_dma)}, + + {ACPI_RSC_INITSET, ACPI_RESOURCE_NAME_FIXED_DMA, + sizeof(struct aml_resource_fixed_dma), + 0}, + + /* + * These fields are contiguous in both the source and destination: + * request_lines + * Channels + */ + + {ACPI_RSC_MOVE16, ACPI_RS_OFFSET(data.fixed_dma.request_lines), + AML_OFFSET(fixed_dma.request_lines), + 2}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.fixed_dma.width), + AML_OFFSET(fixed_dma.width), + 1}, + +}; diff --git a/drivers/acpi/acpica/rslist.c b/drivers/acpi/acpica/rslist.c index 1bfcef736c50..c44635136598 100644 --- a/drivers/acpi/acpica/rslist.c +++ b/drivers/acpi/acpica/rslist.c @@ -70,6 +70,8 @@ acpi_rs_convert_aml_to_resources(u8 * aml, struct acpi_resource **resource_ptr = ACPI_CAST_INDIRECT_PTR(struct acpi_resource, context); struct acpi_resource *resource; + union aml_resource *aml_resource; + struct acpi_rsconvert_info *conversion_table; acpi_status status; ACPI_FUNCTION_TRACE(rs_convert_aml_to_resources); @@ -84,14 +86,37 @@ acpi_rs_convert_aml_to_resources(u8 * aml, "Misaligned resource pointer %p", resource)); } + /* Get the appropriate conversion info table */ + + aml_resource = ACPI_CAST_PTR(union aml_resource, aml); + if (acpi_ut_get_resource_type(aml) == ACPI_RESOURCE_NAME_SERIAL_BUS) { + if (aml_resource->common_serial_bus.type > + AML_RESOURCE_MAX_SERIALBUSTYPE) { + conversion_table = NULL; + } else { + /* This is an I2C, SPI, or UART serial_bus descriptor */ + + conversion_table = + acpi_gbl_convert_resource_serial_bus_dispatch + [aml_resource->common_serial_bus.type]; + } + } else { + conversion_table = + acpi_gbl_get_resource_dispatch[resource_index]; + } + + if (!conversion_table) { + ACPI_ERROR((AE_INFO, + "Invalid/unsupported resource descriptor: Type 0x%2.2X", + resource_index)); + return (AE_AML_INVALID_RESOURCE_TYPE); + } + /* Convert the AML byte stream resource to a local resource struct */ status = - acpi_rs_convert_aml_to_resource(resource, - ACPI_CAST_PTR(union aml_resource, - aml), - acpi_gbl_get_resource_dispatch - [resource_index]); + acpi_rs_convert_aml_to_resource(resource, aml_resource, + conversion_table); if (ACPI_FAILURE(status)) { ACPI_EXCEPTION((AE_INFO, status, "Could not convert AML resource (Type 0x%X)", @@ -106,7 +131,7 @@ acpi_rs_convert_aml_to_resources(u8 * aml, /* Point to the next structure in the output buffer */ - *resource_ptr = ACPI_ADD_PTR(void, resource, resource->length); + *resource_ptr = ACPI_NEXT_RESOURCE(resource); return_ACPI_STATUS(AE_OK); } @@ -135,6 +160,7 @@ acpi_rs_convert_resources_to_aml(struct acpi_resource *resource, { u8 *aml = output_buffer; u8 *end_aml = output_buffer + aml_size_needed; + struct acpi_rsconvert_info *conversion_table; acpi_status status; ACPI_FUNCTION_TRACE(rs_convert_resources_to_aml); @@ -154,11 +180,34 @@ acpi_rs_convert_resources_to_aml(struct acpi_resource *resource, /* Perform the conversion */ - status = acpi_rs_convert_resource_to_aml(resource, ACPI_CAST_PTR(union - aml_resource, - aml), - acpi_gbl_set_resource_dispatch - [resource->type]); + if (resource->type == ACPI_RESOURCE_TYPE_SERIAL_BUS) { + if (resource->data.common_serial_bus.type > + AML_RESOURCE_MAX_SERIALBUSTYPE) { + conversion_table = NULL; + } else { + /* This is an I2C, SPI, or UART serial_bus descriptor */ + + conversion_table = + acpi_gbl_convert_resource_serial_bus_dispatch + [resource->data.common_serial_bus.type]; + } + } else { + conversion_table = + acpi_gbl_set_resource_dispatch[resource->type]; + } + + if (!conversion_table) { + ACPI_ERROR((AE_INFO, + "Invalid/unsupported resource descriptor: Type 0x%2.2X", + resource->type)); + return (AE_AML_INVALID_RESOURCE_TYPE); + } + + status = acpi_rs_convert_resource_to_aml(resource, + ACPI_CAST_PTR(union + aml_resource, + aml), + conversion_table); if (ACPI_FAILURE(status)) { ACPI_EXCEPTION((AE_INFO, status, "Could not convert resource (type 0x%X) to AML", @@ -192,9 +241,7 @@ acpi_rs_convert_resources_to_aml(struct acpi_resource *resource, /* Point to the next input resource descriptor */ - resource = - ACPI_ADD_PTR(struct acpi_resource, resource, - resource->length); + resource = ACPI_NEXT_RESOURCE(resource); } /* Completed buffer, but did not find an end_tag resource descriptor */ diff --git a/drivers/acpi/acpica/rsmisc.c b/drivers/acpi/acpica/rsmisc.c index 410264b22a29..af3dc63163b6 100644 --- a/drivers/acpi/acpica/rsmisc.c +++ b/drivers/acpi/acpica/rsmisc.c @@ -83,6 +83,10 @@ acpi_rs_convert_aml_to_resource(struct acpi_resource *resource, ACPI_FUNCTION_TRACE(rs_convert_aml_to_resource); + if (!info) { + return_ACPI_STATUS(AE_BAD_PARAMETER); + } + if (((acpi_size) resource) & 0x3) { /* Each internal resource struct is expected to be 32-bit aligned */ @@ -101,7 +105,6 @@ acpi_rs_convert_aml_to_resource(struct acpi_resource *resource, * table length (# of table entries) */ count = INIT_TABLE_LENGTH(info); - while (count) { /* * Source is the external AML byte stream buffer, @@ -145,6 +148,14 @@ acpi_rs_convert_aml_to_resource(struct acpi_resource *resource, ((ACPI_GET8(source) >> info->value) & 0x03); break; + case ACPI_RSC_3BITFLAG: + /* + * Mask and shift the flag bits + */ + ACPI_SET8(destination) = (u8) + ((ACPI_GET8(source) >> info->value) & 0x07); + break; + case ACPI_RSC_COUNT: item_count = ACPI_GET8(source); @@ -163,6 +174,69 @@ acpi_rs_convert_aml_to_resource(struct acpi_resource *resource, (info->value * (item_count - 1)); break; + case ACPI_RSC_COUNT_GPIO_PIN: + + target = ACPI_ADD_PTR(void, aml, info->value); + item_count = ACPI_GET16(target) - ACPI_GET16(source); + + resource->length = resource->length + item_count; + item_count = item_count / 2; + ACPI_SET16(destination) = item_count; + break; + + case ACPI_RSC_COUNT_GPIO_VEN: + + item_count = ACPI_GET8(source); + ACPI_SET8(destination) = (u8)item_count; + + resource->length = resource->length + + (info->value * item_count); + break; + + case ACPI_RSC_COUNT_GPIO_RES: + + /* + * Vendor data is optional (length/offset may both be zero) + * Examine vendor data length field first + */ + target = ACPI_ADD_PTR(void, aml, (info->value + 2)); + if (ACPI_GET16(target)) { + + /* Use vendor offset to get resource source length */ + + target = ACPI_ADD_PTR(void, aml, info->value); + item_count = + ACPI_GET16(target) - ACPI_GET16(source); + } else { + /* No vendor data to worry about */ + + item_count = aml->large_header.resource_length + + sizeof(struct aml_resource_large_header) - + ACPI_GET16(source); + } + + resource->length = resource->length + item_count; + ACPI_SET16(destination) = item_count; + break; + + case ACPI_RSC_COUNT_SERIAL_VEN: + + item_count = ACPI_GET16(source) - info->value; + + resource->length = resource->length + item_count; + ACPI_SET16(destination) = item_count; + break; + + case ACPI_RSC_COUNT_SERIAL_RES: + + item_count = (aml_resource_length + + sizeof(struct aml_resource_large_header)) + - ACPI_GET16(source) - info->value; + + resource->length = resource->length + item_count; + ACPI_SET16(destination) = item_count; + break; + case ACPI_RSC_LENGTH: resource->length = resource->length + info->value; @@ -183,6 +257,72 @@ acpi_rs_convert_aml_to_resource(struct acpi_resource *resource, info->opcode); break; + case ACPI_RSC_MOVE_GPIO_PIN: + + /* Generate and set the PIN data pointer */ + + target = (char *)ACPI_ADD_PTR(void, resource, + (resource->length - + item_count * 2)); + *(u16 **)destination = ACPI_CAST_PTR(u16, target); + + /* Copy the PIN data */ + + source = ACPI_ADD_PTR(void, aml, ACPI_GET16(source)); + acpi_rs_move_data(target, source, item_count, + info->opcode); + break; + + case ACPI_RSC_MOVE_GPIO_RES: + + /* Generate and set the resource_source string pointer */ + + target = (char *)ACPI_ADD_PTR(void, resource, + (resource->length - + item_count)); + *(u8 **)destination = ACPI_CAST_PTR(u8, target); + + /* Copy the resource_source string */ + + source = ACPI_ADD_PTR(void, aml, ACPI_GET16(source)); + acpi_rs_move_data(target, source, item_count, + info->opcode); + break; + + case ACPI_RSC_MOVE_SERIAL_VEN: + + /* Generate and set the Vendor Data pointer */ + + target = (char *)ACPI_ADD_PTR(void, resource, + (resource->length - + item_count)); + *(u8 **)destination = ACPI_CAST_PTR(u8, target); + + /* Copy the Vendor Data */ + + source = ACPI_ADD_PTR(void, aml, info->value); + acpi_rs_move_data(target, source, item_count, + info->opcode); + break; + + case ACPI_RSC_MOVE_SERIAL_RES: + + /* Generate and set the resource_source string pointer */ + + target = (char *)ACPI_ADD_PTR(void, resource, + (resource->length - + item_count)); + *(u8 **)destination = ACPI_CAST_PTR(u8, target); + + /* Copy the resource_source string */ + + source = + ACPI_ADD_PTR(void, aml, + (ACPI_GET16(source) + info->value)); + acpi_rs_move_data(target, source, item_count, + info->opcode); + break; + case ACPI_RSC_SET8: ACPI_MEMSET(destination, info->aml_offset, info->value); @@ -219,13 +359,18 @@ acpi_rs_convert_aml_to_resource(struct acpi_resource *resource, * Optional resource_source (Index and String). This is the more * complicated case used by the Interrupt() macro */ - target = - ACPI_ADD_PTR(char, resource, - info->aml_offset + (item_count * 4)); + target = ACPI_ADD_PTR(char, resource, + info->aml_offset + + (item_count * 4)); resource->length += acpi_rs_get_resource_source(aml_resource_length, - (acpi_rs_length) (((item_count - 1) * sizeof(u32)) + info->value), destination, aml, target); + (acpi_rs_length) + (((item_count - + 1) * sizeof(u32)) + + info->value), + destination, aml, + target); break; case ACPI_RSC_BITMASK: @@ -327,6 +472,7 @@ acpi_rs_convert_resource_to_aml(struct acpi_resource *resource, { void *source = NULL; void *destination; + char *target; acpi_rsdesc_size aml_length = 0; u8 count; u16 temp16 = 0; @@ -334,6 +480,10 @@ acpi_rs_convert_resource_to_aml(struct acpi_resource *resource, ACPI_FUNCTION_TRACE(rs_convert_resource_to_aml); + if (!info) { + return_ACPI_STATUS(AE_BAD_PARAMETER); + } + /* * First table entry must be ACPI_RSC_INITxxx and must contain the * table length (# of table entries) @@ -383,6 +533,14 @@ acpi_rs_convert_resource_to_aml(struct acpi_resource *resource, ((ACPI_GET8(source) & 0x03) << info->value); break; + case ACPI_RSC_3BITFLAG: + /* + * Mask and shift the flag bits + */ + ACPI_SET8(destination) |= (u8) + ((ACPI_GET8(source) & 0x07) << info->value); + break; + case ACPI_RSC_COUNT: item_count = ACPI_GET8(source); @@ -400,6 +558,63 @@ acpi_rs_convert_resource_to_aml(struct acpi_resource *resource, acpi_rs_set_resource_length(aml_length, aml); break; + case ACPI_RSC_COUNT_GPIO_PIN: + + item_count = ACPI_GET16(source); + ACPI_SET16(destination) = (u16)aml_length; + + aml_length = (u16)(aml_length + item_count * 2); + target = ACPI_ADD_PTR(void, aml, info->value); + ACPI_SET16(target) = (u16)aml_length; + acpi_rs_set_resource_length(aml_length, aml); + break; + + case ACPI_RSC_COUNT_GPIO_VEN: + + item_count = ACPI_GET16(source); + ACPI_SET16(destination) = (u16)item_count; + + aml_length = + (u16)(aml_length + (info->value * item_count)); + acpi_rs_set_resource_length(aml_length, aml); + break; + + case ACPI_RSC_COUNT_GPIO_RES: + + /* Set resource source string length */ + + item_count = ACPI_GET16(source); + ACPI_SET16(destination) = (u16)aml_length; + + /* Compute offset for the Vendor Data */ + + aml_length = (u16)(aml_length + item_count); + target = ACPI_ADD_PTR(void, aml, info->value); + + /* Set vendor offset only if there is vendor data */ + + if (resource->data.gpio.vendor_length) { + ACPI_SET16(target) = (u16)aml_length; + } + + acpi_rs_set_resource_length(aml_length, aml); + break; + + case ACPI_RSC_COUNT_SERIAL_VEN: + + item_count = ACPI_GET16(source); + ACPI_SET16(destination) = item_count + info->value; + aml_length = (u16)(aml_length + item_count); + acpi_rs_set_resource_length(aml_length, aml); + break; + + case ACPI_RSC_COUNT_SERIAL_RES: + + item_count = ACPI_GET16(source); + aml_length = (u16)(aml_length + item_count); + acpi_rs_set_resource_length(aml_length, aml); + break; + case ACPI_RSC_LENGTH: acpi_rs_set_resource_length(info->value, aml); @@ -417,6 +632,48 @@ acpi_rs_convert_resource_to_aml(struct acpi_resource *resource, info->opcode); break; + case ACPI_RSC_MOVE_GPIO_PIN: + + destination = (char *)ACPI_ADD_PTR(void, aml, + ACPI_GET16 + (destination)); + source = *(u16 **)source; + acpi_rs_move_data(destination, source, item_count, + info->opcode); + break; + + case ACPI_RSC_MOVE_GPIO_RES: + + /* Used for both resource_source string and vendor_data */ + + destination = (char *)ACPI_ADD_PTR(void, aml, + ACPI_GET16 + (destination)); + source = *(u8 **)source; + acpi_rs_move_data(destination, source, item_count, + info->opcode); + break; + + case ACPI_RSC_MOVE_SERIAL_VEN: + + destination = (char *)ACPI_ADD_PTR(void, aml, + (aml_length - + item_count)); + source = *(u8 **)source; + acpi_rs_move_data(destination, source, item_count, + info->opcode); + break; + + case ACPI_RSC_MOVE_SERIAL_RES: + + destination = (char *)ACPI_ADD_PTR(void, aml, + (aml_length - + item_count)); + source = *(u8 **)source; + acpi_rs_move_data(destination, source, item_count, + info->opcode); + break; + case ACPI_RSC_ADDRESS: /* Set the Resource Type, General Flags, and Type-Specific Flags */ diff --git a/drivers/acpi/acpica/rsserial.c b/drivers/acpi/acpica/rsserial.c new file mode 100644 index 000000000000..d0b8e121843a --- /dev/null +++ b/drivers/acpi/acpica/rsserial.c @@ -0,0 +1,441 @@ +/******************************************************************************* + * + * Module Name: rsserial - GPIO/serial_bus resource descriptors + * + ******************************************************************************/ + +/* + * Copyright (C) 2000 - 2011, Intel Corp. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. Redistributions in binary form must reproduce at minimum a disclaimer + * substantially similar to the "NO WARRANTY" disclaimer below + * ("Disclaimer") and any redistribution must be conditioned upon + * including a substantially similar Disclaimer requirement for further + * binary redistribution. + * 3. Neither the names of the above-listed copyright holders nor the names + * of any contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL") version 2 as published by the Free + * Software Foundation. + * + * NO WARRANTY + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, + * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING + * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGES. + */ + +#include +#include "accommon.h" +#include "acresrc.h" + +#define _COMPONENT ACPI_RESOURCES +ACPI_MODULE_NAME("rsserial") + +/******************************************************************************* + * + * acpi_rs_convert_gpio + * + ******************************************************************************/ +struct acpi_rsconvert_info acpi_rs_convert_gpio[17] = { + {ACPI_RSC_INITGET, ACPI_RESOURCE_TYPE_GPIO, + ACPI_RS_SIZE(struct acpi_resource_gpio), + ACPI_RSC_TABLE_SIZE(acpi_rs_convert_gpio)}, + + {ACPI_RSC_INITSET, ACPI_RESOURCE_NAME_GPIO, + sizeof(struct aml_resource_gpio), + 0}, + + /* + * These fields are contiguous in both the source and destination: + * revision_id + * connection_type + */ + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.gpio.revision_id), + AML_OFFSET(gpio.revision_id), + 2}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.gpio.producer_consumer), + AML_OFFSET(gpio.flags), + 0}, + + {ACPI_RSC_2BITFLAG, ACPI_RS_OFFSET(data.gpio.sharable), + AML_OFFSET(gpio.int_flags), + 3}, + + {ACPI_RSC_2BITFLAG, ACPI_RS_OFFSET(data.gpio.io_restriction), + AML_OFFSET(gpio.int_flags), + 0}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.gpio.triggering), + AML_OFFSET(gpio.int_flags), + 0}, + + {ACPI_RSC_2BITFLAG, ACPI_RS_OFFSET(data.gpio.polarity), + AML_OFFSET(gpio.int_flags), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.gpio.pin_config), + AML_OFFSET(gpio.pin_config), + 1}, + + /* + * These fields are contiguous in both the source and destination: + * drive_strength + * debounce_timeout + */ + {ACPI_RSC_MOVE16, ACPI_RS_OFFSET(data.gpio.drive_strength), + AML_OFFSET(gpio.drive_strength), + 2}, + + /* Pin Table */ + + {ACPI_RSC_COUNT_GPIO_PIN, ACPI_RS_OFFSET(data.gpio.pin_table_length), + AML_OFFSET(gpio.pin_table_offset), + AML_OFFSET(gpio.res_source_offset)}, + + {ACPI_RSC_MOVE_GPIO_PIN, ACPI_RS_OFFSET(data.gpio.pin_table), + AML_OFFSET(gpio.pin_table_offset), + 0}, + + /* Resource Source */ + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.gpio.resource_source.index), + AML_OFFSET(gpio.res_source_index), + 1}, + + {ACPI_RSC_COUNT_GPIO_RES, + ACPI_RS_OFFSET(data.gpio.resource_source.string_length), + AML_OFFSET(gpio.res_source_offset), + AML_OFFSET(gpio.vendor_offset)}, + + {ACPI_RSC_MOVE_GPIO_RES, + ACPI_RS_OFFSET(data.gpio.resource_source.string_ptr), + AML_OFFSET(gpio.res_source_offset), + 0}, + + /* Vendor Data */ + + {ACPI_RSC_COUNT_GPIO_VEN, ACPI_RS_OFFSET(data.gpio.vendor_length), + AML_OFFSET(gpio.vendor_length), + 1}, + + {ACPI_RSC_MOVE_GPIO_RES, ACPI_RS_OFFSET(data.gpio.vendor_data), + AML_OFFSET(gpio.vendor_offset), + 0}, +}; + +/******************************************************************************* + * + * acpi_rs_convert_i2c_serial_bus + * + ******************************************************************************/ + +struct acpi_rsconvert_info acpi_rs_convert_i2c_serial_bus[16] = { + {ACPI_RSC_INITGET, ACPI_RESOURCE_TYPE_SERIAL_BUS, + ACPI_RS_SIZE(struct acpi_resource_i2c_serialbus), + ACPI_RSC_TABLE_SIZE(acpi_rs_convert_i2c_serial_bus)}, + + {ACPI_RSC_INITSET, ACPI_RESOURCE_NAME_SERIAL_BUS, + sizeof(struct aml_resource_i2c_serialbus), + 0}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.common_serial_bus.revision_id), + AML_OFFSET(common_serial_bus.revision_id), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.common_serial_bus.type), + AML_OFFSET(common_serial_bus.type), + 1}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.common_serial_bus.slave_mode), + AML_OFFSET(common_serial_bus.flags), + 0}, + + {ACPI_RSC_1BITFLAG, + ACPI_RS_OFFSET(data.common_serial_bus.producer_consumer), + AML_OFFSET(common_serial_bus.flags), + 1}, + + {ACPI_RSC_MOVE8, + ACPI_RS_OFFSET(data.common_serial_bus.type_revision_id), + AML_OFFSET(common_serial_bus.type_revision_id), + 1}, + + {ACPI_RSC_MOVE16, + ACPI_RS_OFFSET(data.common_serial_bus.type_data_length), + AML_OFFSET(common_serial_bus.type_data_length), + 1}, + + /* Vendor data */ + + {ACPI_RSC_COUNT_SERIAL_VEN, + ACPI_RS_OFFSET(data.common_serial_bus.vendor_length), + AML_OFFSET(common_serial_bus.type_data_length), + AML_RESOURCE_I2C_MIN_DATA_LEN}, + + {ACPI_RSC_MOVE_SERIAL_VEN, + ACPI_RS_OFFSET(data.common_serial_bus.vendor_data), + 0, + sizeof(struct aml_resource_i2c_serialbus)}, + + /* Resource Source */ + + {ACPI_RSC_MOVE8, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.index), + AML_OFFSET(common_serial_bus.res_source_index), + 1}, + + {ACPI_RSC_COUNT_SERIAL_RES, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.string_length), + AML_OFFSET(common_serial_bus.type_data_length), + sizeof(struct aml_resource_common_serialbus)}, + + {ACPI_RSC_MOVE_SERIAL_RES, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.string_ptr), + AML_OFFSET(common_serial_bus.type_data_length), + sizeof(struct aml_resource_common_serialbus)}, + + /* I2C bus type specific */ + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.i2c_serial_bus.access_mode), + AML_OFFSET(i2c_serial_bus.type_specific_flags), + 0}, + + {ACPI_RSC_MOVE32, ACPI_RS_OFFSET(data.i2c_serial_bus.connection_speed), + AML_OFFSET(i2c_serial_bus.connection_speed), + 1}, + + {ACPI_RSC_MOVE16, ACPI_RS_OFFSET(data.i2c_serial_bus.slave_address), + AML_OFFSET(i2c_serial_bus.slave_address), + 1}, +}; + +/******************************************************************************* + * + * acpi_rs_convert_spi_serial_bus + * + ******************************************************************************/ + +struct acpi_rsconvert_info acpi_rs_convert_spi_serial_bus[20] = { + {ACPI_RSC_INITGET, ACPI_RESOURCE_TYPE_SERIAL_BUS, + ACPI_RS_SIZE(struct acpi_resource_spi_serialbus), + ACPI_RSC_TABLE_SIZE(acpi_rs_convert_spi_serial_bus)}, + + {ACPI_RSC_INITSET, ACPI_RESOURCE_NAME_SERIAL_BUS, + sizeof(struct aml_resource_spi_serialbus), + 0}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.common_serial_bus.revision_id), + AML_OFFSET(common_serial_bus.revision_id), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.common_serial_bus.type), + AML_OFFSET(common_serial_bus.type), + 1}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.common_serial_bus.slave_mode), + AML_OFFSET(common_serial_bus.flags), + 0}, + + {ACPI_RSC_1BITFLAG, + ACPI_RS_OFFSET(data.common_serial_bus.producer_consumer), + AML_OFFSET(common_serial_bus.flags), + 1}, + + {ACPI_RSC_MOVE8, + ACPI_RS_OFFSET(data.common_serial_bus.type_revision_id), + AML_OFFSET(common_serial_bus.type_revision_id), + 1}, + + {ACPI_RSC_MOVE16, + ACPI_RS_OFFSET(data.common_serial_bus.type_data_length), + AML_OFFSET(common_serial_bus.type_data_length), + 1}, + + /* Vendor data */ + + {ACPI_RSC_COUNT_SERIAL_VEN, + ACPI_RS_OFFSET(data.common_serial_bus.vendor_length), + AML_OFFSET(common_serial_bus.type_data_length), + AML_RESOURCE_SPI_MIN_DATA_LEN}, + + {ACPI_RSC_MOVE_SERIAL_VEN, + ACPI_RS_OFFSET(data.common_serial_bus.vendor_data), + 0, + sizeof(struct aml_resource_spi_serialbus)}, + + /* Resource Source */ + + {ACPI_RSC_MOVE8, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.index), + AML_OFFSET(common_serial_bus.res_source_index), + 1}, + + {ACPI_RSC_COUNT_SERIAL_RES, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.string_length), + AML_OFFSET(common_serial_bus.type_data_length), + sizeof(struct aml_resource_common_serialbus)}, + + {ACPI_RSC_MOVE_SERIAL_RES, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.string_ptr), + AML_OFFSET(common_serial_bus.type_data_length), + sizeof(struct aml_resource_common_serialbus)}, + + /* Spi bus type specific */ + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.spi_serial_bus.wire_mode), + AML_OFFSET(spi_serial_bus.type_specific_flags), + 0}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.spi_serial_bus.device_polarity), + AML_OFFSET(spi_serial_bus.type_specific_flags), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.spi_serial_bus.data_bit_length), + AML_OFFSET(spi_serial_bus.data_bit_length), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.spi_serial_bus.clock_phase), + AML_OFFSET(spi_serial_bus.clock_phase), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.spi_serial_bus.clock_polarity), + AML_OFFSET(spi_serial_bus.clock_polarity), + 1}, + + {ACPI_RSC_MOVE16, ACPI_RS_OFFSET(data.spi_serial_bus.device_selection), + AML_OFFSET(spi_serial_bus.device_selection), + 1}, + + {ACPI_RSC_MOVE32, ACPI_RS_OFFSET(data.spi_serial_bus.connection_speed), + AML_OFFSET(spi_serial_bus.connection_speed), + 1}, +}; + +/******************************************************************************* + * + * acpi_rs_convert_uart_serial_bus + * + ******************************************************************************/ + +struct acpi_rsconvert_info acpi_rs_convert_uart_serial_bus[22] = { + {ACPI_RSC_INITGET, ACPI_RESOURCE_TYPE_SERIAL_BUS, + ACPI_RS_SIZE(struct acpi_resource_uart_serialbus), + ACPI_RSC_TABLE_SIZE(acpi_rs_convert_uart_serial_bus)}, + + {ACPI_RSC_INITSET, ACPI_RESOURCE_NAME_SERIAL_BUS, + sizeof(struct aml_resource_uart_serialbus), + 0}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.common_serial_bus.revision_id), + AML_OFFSET(common_serial_bus.revision_id), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.common_serial_bus.type), + AML_OFFSET(common_serial_bus.type), + 1}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.common_serial_bus.slave_mode), + AML_OFFSET(common_serial_bus.flags), + 0}, + + {ACPI_RSC_1BITFLAG, + ACPI_RS_OFFSET(data.common_serial_bus.producer_consumer), + AML_OFFSET(common_serial_bus.flags), + 1}, + + {ACPI_RSC_MOVE8, + ACPI_RS_OFFSET(data.common_serial_bus.type_revision_id), + AML_OFFSET(common_serial_bus.type_revision_id), + 1}, + + {ACPI_RSC_MOVE16, + ACPI_RS_OFFSET(data.common_serial_bus.type_data_length), + AML_OFFSET(common_serial_bus.type_data_length), + 1}, + + /* Vendor data */ + + {ACPI_RSC_COUNT_SERIAL_VEN, + ACPI_RS_OFFSET(data.common_serial_bus.vendor_length), + AML_OFFSET(common_serial_bus.type_data_length), + AML_RESOURCE_UART_MIN_DATA_LEN}, + + {ACPI_RSC_MOVE_SERIAL_VEN, + ACPI_RS_OFFSET(data.common_serial_bus.vendor_data), + 0, + sizeof(struct aml_resource_uart_serialbus)}, + + /* Resource Source */ + + {ACPI_RSC_MOVE8, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.index), + AML_OFFSET(common_serial_bus.res_source_index), + 1}, + + {ACPI_RSC_COUNT_SERIAL_RES, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.string_length), + AML_OFFSET(common_serial_bus.type_data_length), + sizeof(struct aml_resource_common_serialbus)}, + + {ACPI_RSC_MOVE_SERIAL_RES, + ACPI_RS_OFFSET(data.common_serial_bus.resource_source.string_ptr), + AML_OFFSET(common_serial_bus.type_data_length), + sizeof(struct aml_resource_common_serialbus)}, + + /* Uart bus type specific */ + + {ACPI_RSC_2BITFLAG, ACPI_RS_OFFSET(data.uart_serial_bus.flow_control), + AML_OFFSET(uart_serial_bus.type_specific_flags), + 0}, + + {ACPI_RSC_2BITFLAG, ACPI_RS_OFFSET(data.uart_serial_bus.stop_bits), + AML_OFFSET(uart_serial_bus.type_specific_flags), + 2}, + + {ACPI_RSC_3BITFLAG, ACPI_RS_OFFSET(data.uart_serial_bus.data_bits), + AML_OFFSET(uart_serial_bus.type_specific_flags), + 4}, + + {ACPI_RSC_1BITFLAG, ACPI_RS_OFFSET(data.uart_serial_bus.endian), + AML_OFFSET(uart_serial_bus.type_specific_flags), + 7}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.uart_serial_bus.parity), + AML_OFFSET(uart_serial_bus.parity), + 1}, + + {ACPI_RSC_MOVE8, ACPI_RS_OFFSET(data.uart_serial_bus.lines_enabled), + AML_OFFSET(uart_serial_bus.lines_enabled), + 1}, + + {ACPI_RSC_MOVE16, ACPI_RS_OFFSET(data.uart_serial_bus.rx_fifo_size), + AML_OFFSET(uart_serial_bus.rx_fifo_size), + 1}, + + {ACPI_RSC_MOVE16, ACPI_RS_OFFSET(data.uart_serial_bus.tx_fifo_size), + AML_OFFSET(uart_serial_bus.tx_fifo_size), + 1}, + + {ACPI_RSC_MOVE32, + ACPI_RS_OFFSET(data.uart_serial_bus.default_baud_rate), + AML_OFFSET(uart_serial_bus.default_baud_rate), + 1}, +}; diff --git a/drivers/acpi/acpica/rsutils.c b/drivers/acpi/acpica/rsutils.c index 231811e56939..4409db8209d7 100644 --- a/drivers/acpi/acpica/rsutils.c +++ b/drivers/acpi/acpica/rsutils.c @@ -144,6 +144,9 @@ acpi_rs_move_data(void *destination, void *source, u16 item_count, u8 move_type) * since there are no alignment or endian issues */ case ACPI_RSC_MOVE8: + case ACPI_RSC_MOVE_GPIO_RES: + case ACPI_RSC_MOVE_SERIAL_VEN: + case ACPI_RSC_MOVE_SERIAL_RES: ACPI_MEMCPY(destination, source, item_count); return; @@ -153,6 +156,7 @@ acpi_rs_move_data(void *destination, void *source, u16 item_count, u8 move_type) * misaligned memory transfers */ case ACPI_RSC_MOVE16: + case ACPI_RSC_MOVE_GPIO_PIN: ACPI_MOVE_16_TO_16(&ACPI_CAST_PTR(u16, destination)[i], &ACPI_CAST_PTR(u16, source)[i]); break; diff --git a/drivers/acpi/acpica/utresrc.c b/drivers/acpi/acpica/utresrc.c index 6ffd3a8bdaa5..cd7fbbf57f30 100644 --- a/drivers/acpi/acpica/utresrc.c +++ b/drivers/acpi/acpica/utresrc.c @@ -43,7 +43,7 @@ #include #include "accommon.h" -#include "amlresrc.h" +#include "acresrc.h" #define _COMPONENT ACPI_UTILITIES ACPI_MODULE_NAME("utresrc") @@ -154,6 +154,138 @@ const char *acpi_gbl_typ_decode[] = { "TypeF" }; +const char *acpi_gbl_ppc_decode[] = { + "PullDefault", + "PullUp", + "PullDown", + "PullNone" +}; + +const char *acpi_gbl_ior_decode[] = { + "IoRestrictionNone", + "IoRestrictionInputOnly", + "IoRestrictionOutputOnly", + "IoRestrictionNoneAndPreserve" +}; + +const char *acpi_gbl_dts_decode[] = { + "Width8bit", + "Width16bit", + "Width32bit", + "Width64bit", + "Width128bit", + "Width256bit", +}; + +/* GPIO connection type */ + +const char *acpi_gbl_ct_decode[] = { + "Interrupt", + "I/O" +}; + +/* Serial bus type */ + +const char *acpi_gbl_sbt_decode[] = { + "/* UNKNOWN serial bus type */", + "I2C", + "SPI", + "UART" +}; + +/* I2C serial bus access mode */ + +const char *acpi_gbl_am_decode[] = { + "AddressingMode7Bit", + "AddressingMode10Bit" +}; + +/* I2C serial bus slave mode */ + +const char *acpi_gbl_sm_decode[] = { + "ControllerInitiated", + "DeviceInitiated" +}; + +/* SPI serial bus wire mode */ + +const char *acpi_gbl_wm_decode[] = { + "FourWireMode", + "ThreeWireMode" +}; + +/* SPI serial clock phase */ + +const char *acpi_gbl_cph_decode[] = { + "ClockPhaseFirst", + "ClockPhaseSecond" +}; + +/* SPI serial bus clock polarity */ + +const char *acpi_gbl_cpo_decode[] = { + "ClockPolarityLow", + "ClockPolarityHigh" +}; + +/* SPI serial bus device polarity */ + +const char *acpi_gbl_dp_decode[] = { + "PolarityLow", + "PolarityHigh" +}; + +/* UART serial bus endian */ + +const char *acpi_gbl_ed_decode[] = { + "LittleEndian", + "BigEndian" +}; + +/* UART serial bus bits per byte */ + +const char *acpi_gbl_bpb_decode[] = { + "DataBitsFive", + "DataBitsSix", + "DataBitsSeven", + "DataBitsEight", + "DataBitsNine", + "/* UNKNOWN Bits per byte */", + "/* UNKNOWN Bits per byte */", + "/* UNKNOWN Bits per byte */" +}; + +/* UART serial bus stop bits */ + +const char *acpi_gbl_sb_decode[] = { + "StopBitsNone", + "StopBitsOne", + "StopBitsOnePlusHalf", + "StopBitsTwo" +}; + +/* UART serial bus flow control */ + +const char *acpi_gbl_fc_decode[] = { + "FlowControlNone", + "FlowControlHardware", + "FlowControlXON", + "/* UNKNOWN flow control keyword */" +}; + +/* UART serial bus parity type */ + +const char *acpi_gbl_pt_decode[] = { + "ParityTypeNone", + "ParityTypeEven", + "ParityTypeOdd", + "ParityTypeMark", + "ParityTypeSpace", + "/* UNKNOWN parity keyword */", + "/* UNKNOWN parity keyword */", + "/* UNKNOWN parity keyword */" +}; + #endif /* @@ -173,7 +305,7 @@ const u8 acpi_gbl_resource_aml_sizes[] = { ACPI_AML_SIZE_SMALL(struct aml_resource_end_dependent), ACPI_AML_SIZE_SMALL(struct aml_resource_io), ACPI_AML_SIZE_SMALL(struct aml_resource_fixed_io), - 0, + ACPI_AML_SIZE_SMALL(struct aml_resource_fixed_dma), 0, 0, 0, @@ -193,7 +325,17 @@ const u8 acpi_gbl_resource_aml_sizes[] = { ACPI_AML_SIZE_LARGE(struct aml_resource_address16), ACPI_AML_SIZE_LARGE(struct aml_resource_extended_irq), ACPI_AML_SIZE_LARGE(struct aml_resource_address64), - ACPI_AML_SIZE_LARGE(struct aml_resource_extended_address64) + ACPI_AML_SIZE_LARGE(struct aml_resource_extended_address64), + ACPI_AML_SIZE_LARGE(struct aml_resource_gpio), + 0, + ACPI_AML_SIZE_LARGE(struct aml_resource_common_serialbus), +}; + +const u8 acpi_gbl_resource_aml_serial_bus_sizes[] = { + 0, + ACPI_AML_SIZE_LARGE(struct aml_resource_i2c_serialbus), + ACPI_AML_SIZE_LARGE(struct aml_resource_spi_serialbus), + ACPI_AML_SIZE_LARGE(struct aml_resource_uart_serialbus), }; /* @@ -209,35 +351,49 @@ static const u8 acpi_gbl_resource_types[] = { 0, 0, 0, - ACPI_SMALL_VARIABLE_LENGTH, - ACPI_FIXED_LENGTH, - ACPI_SMALL_VARIABLE_LENGTH, - ACPI_FIXED_LENGTH, - ACPI_FIXED_LENGTH, - ACPI_FIXED_LENGTH, + ACPI_SMALL_VARIABLE_LENGTH, /* 04 IRQ */ + ACPI_FIXED_LENGTH, /* 05 DMA */ + ACPI_SMALL_VARIABLE_LENGTH, /* 06 start_dependent_functions */ + ACPI_FIXED_LENGTH, /* 07 end_dependent_functions */ + ACPI_FIXED_LENGTH, /* 08 IO */ + ACPI_FIXED_LENGTH, /* 09 fixed_iO */ + ACPI_FIXED_LENGTH, /* 0_a fixed_dMA */ 0, 0, 0, - 0, - ACPI_VARIABLE_LENGTH, - ACPI_FIXED_LENGTH, + ACPI_VARIABLE_LENGTH, /* 0_e vendor_short */ + ACPI_FIXED_LENGTH, /* 0_f end_tag */ /* Large descriptors */ 0, - ACPI_FIXED_LENGTH, - ACPI_FIXED_LENGTH, + ACPI_FIXED_LENGTH, /* 01 Memory24 */ + ACPI_FIXED_LENGTH, /* 02 generic_register */ 0, - ACPI_VARIABLE_LENGTH, - ACPI_FIXED_LENGTH, - ACPI_FIXED_LENGTH, - ACPI_VARIABLE_LENGTH, - ACPI_VARIABLE_LENGTH, - ACPI_VARIABLE_LENGTH, - ACPI_VARIABLE_LENGTH, - ACPI_FIXED_LENGTH + ACPI_VARIABLE_LENGTH, /* 04 vendor_long */ + ACPI_FIXED_LENGTH, /* 05 Memory32 */ + ACPI_FIXED_LENGTH, /* 06 memory32_fixed */ + ACPI_VARIABLE_LENGTH, /* 07 Dword* address */ + ACPI_VARIABLE_LENGTH, /* 08 Word* address */ + ACPI_VARIABLE_LENGTH, /* 09 extended_iRQ */ + ACPI_VARIABLE_LENGTH, /* 0_a Qword* address */ + ACPI_FIXED_LENGTH, /* 0_b Extended* address */ + ACPI_VARIABLE_LENGTH, /* 0_c Gpio* */ + 0, + ACPI_VARIABLE_LENGTH /* 0_e *serial_bus */ }; +/* + * For the i_aSL compiler/disassembler, we don't want any error messages + * because the disassembler uses the resource validation code to determine + * if Buffer objects are actually Resource Templates. + */ +#ifdef ACPI_ASL_COMPILER +#define ACPI_RESOURCE_ERROR(plist) +#else +#define ACPI_RESOURCE_ERROR(plist) ACPI_ERROR(plist) +#endif + /******************************************************************************* * * FUNCTION: acpi_ut_walk_aml_resources @@ -265,6 +421,7 @@ acpi_ut_walk_aml_resources(u8 * aml, u8 resource_index; u32 length; u32 offset = 0; + u8 end_tag[2] = { 0x79, 0x00 }; ACPI_FUNCTION_TRACE(ut_walk_aml_resources); @@ -286,6 +443,10 @@ acpi_ut_walk_aml_resources(u8 * aml, status = acpi_ut_validate_resource(aml, &resource_index); if (ACPI_FAILURE(status)) { + /* + * Exit on failure. Cannot continue because the descriptor length + * may be bogus also. + */ return_ACPI_STATUS(status); } @@ -300,7 +461,7 @@ acpi_ut_walk_aml_resources(u8 * aml, user_function(aml, length, offset, resource_index, context); if (ACPI_FAILURE(status)) { - return (status); + return_ACPI_STATUS(status); } } @@ -333,7 +494,19 @@ acpi_ut_walk_aml_resources(u8 * aml, /* Did not find an end_tag descriptor */ - return (AE_AML_NO_RESOURCE_END_TAG); + if (user_function) { + + /* Insert an end_tag anyway. acpi_rs_get_list_length always leaves room */ + + (void)acpi_ut_validate_resource(end_tag, &resource_index); + status = + user_function(end_tag, 2, offset, resource_index, context); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + } + + return_ACPI_STATUS(AE_AML_NO_RESOURCE_END_TAG); } /******************************************************************************* @@ -354,6 +527,7 @@ acpi_ut_walk_aml_resources(u8 * aml, acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) { + union aml_resource *aml_resource; u8 resource_type; u8 resource_index; acpi_rs_length resource_length; @@ -375,7 +549,7 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) /* Verify the large resource type (name) against the max */ if (resource_type > ACPI_RESOURCE_NAME_LARGE_MAX) { - return (AE_AML_INVALID_RESOURCE_TYPE); + goto invalid_resource; } /* @@ -392,15 +566,17 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) ((resource_type & ACPI_RESOURCE_NAME_SMALL_MASK) >> 3); } - /* Check validity of the resource type, zero indicates name is invalid */ - + /* + * Check validity of the resource type, via acpi_gbl_resource_types. Zero + * indicates an invalid resource. + */ if (!acpi_gbl_resource_types[resource_index]) { - return (AE_AML_INVALID_RESOURCE_TYPE); + goto invalid_resource; } /* - * 2) Validate the resource_length field. This ensures that the length - * is at least reasonable, and guarantees that it is non-zero. + * Validate the resource_length field. This ensures that the length + * is at least reasonable, and guarantees that it is non-zero. */ resource_length = acpi_ut_get_resource_length(aml); minimum_resource_length = acpi_gbl_resource_aml_sizes[resource_index]; @@ -413,7 +589,7 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) /* Fixed length resource, length must match exactly */ if (resource_length != minimum_resource_length) { - return (AE_AML_BAD_RESOURCE_LENGTH); + goto bad_resource_length; } break; @@ -422,7 +598,7 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) /* Variable length resource, length must be at least the minimum */ if (resource_length < minimum_resource_length) { - return (AE_AML_BAD_RESOURCE_LENGTH); + goto bad_resource_length; } break; @@ -432,7 +608,7 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) if ((resource_length > minimum_resource_length) || (resource_length < (minimum_resource_length - 1))) { - return (AE_AML_BAD_RESOURCE_LENGTH); + goto bad_resource_length; } break; @@ -440,7 +616,23 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) /* Shouldn't happen (because of validation earlier), but be sure */ - return (AE_AML_INVALID_RESOURCE_TYPE); + goto invalid_resource; + } + + aml_resource = ACPI_CAST_PTR(union aml_resource, aml); + if (resource_type == ACPI_RESOURCE_NAME_SERIAL_BUS) { + + /* Validate the bus_type field */ + + if ((aml_resource->common_serial_bus.type == 0) || + (aml_resource->common_serial_bus.type > + AML_RESOURCE_MAX_SERIALBUSTYPE)) { + ACPI_RESOURCE_ERROR((AE_INFO, + "Invalid/unsupported SerialBus resource descriptor: BusType 0x%2.2X", + aml_resource->common_serial_bus. + type)); + return (AE_AML_INVALID_RESOURCE_TYPE); + } } /* Optionally return the resource table index */ @@ -450,6 +642,22 @@ acpi_status acpi_ut_validate_resource(void *aml, u8 * return_index) } return (AE_OK); + + invalid_resource: + + ACPI_RESOURCE_ERROR((AE_INFO, + "Invalid/unsupported resource descriptor: Type 0x%2.2X", + resource_type)); + return (AE_AML_INVALID_RESOURCE_TYPE); + + bad_resource_length: + + ACPI_RESOURCE_ERROR((AE_INFO, + "Invalid resource descriptor length: Type " + "0x%2.2X, Length 0x%4.4X, MinLength 0x%4.4X", + resource_type, resource_length, + minimum_resource_length)); + return (AE_AML_BAD_RESOURCE_LENGTH); } /******************************************************************************* diff --git a/include/acpi/acrestyp.h b/include/acpi/acrestyp.h index 0a66cc45dd6b..3506e39a66b1 100644 --- a/include/acpi/acrestyp.h +++ b/include/acpi/acrestyp.h @@ -61,11 +61,14 @@ typedef u32 acpi_rsdesc_size; /* Max Resource Descriptor size is (Length+3) = (6 #define ACPI_WRITE_COMBINING_MEMORY (u8) 0x02 #define ACPI_PREFETCHABLE_MEMORY (u8) 0x03 +/*! [Begin] no source code translation */ /* * IO Attributes - * The ISA IO ranges are: n000-n0_fFh, n400-n4_fFh, n800-n8_fFh, n_c00-n_cFFh. - * The non-ISA IO ranges are: n100-n3_fFh, n500-n7_fFh, n900-n_bFFh, n_cd0-n_fFFh. + * The ISA IO ranges are: n000-n0FFh, n400-n4FFh, n800-n8FFh, nC00-nCFFh. + * The non-ISA IO ranges are: n100-n3FFh, n500-n7FFh, n900-nBFFh, nCD0-nFFFh. */ +/*! [End] no source code translation !*/ + #define ACPI_NON_ISA_ONLY_RANGES (u8) 0x01 #define ACPI_ISA_ONLY_RANGES (u8) 0x02 #define ACPI_ENTIRE_RANGE (ACPI_NON_ISA_ONLY_RANGES | ACPI_ISA_ONLY_RANGES) @@ -81,16 +84,26 @@ typedef u32 acpi_rsdesc_size; /* Max Resource Descriptor size is (Length+3) = (6 #define ACPI_DECODE_16 (u8) 0x01 /* 16-bit IO address decode */ /* - * IRQ Attributes + * Interrupt attributes - used in multiple descriptors */ + +/* Triggering */ + #define ACPI_LEVEL_SENSITIVE (u8) 0x00 #define ACPI_EDGE_SENSITIVE (u8) 0x01 +/* Polarity */ + #define ACPI_ACTIVE_HIGH (u8) 0x00 #define ACPI_ACTIVE_LOW (u8) 0x01 +#define ACPI_ACTIVE_BOTH (u8) 0x02 + +/* Sharing */ #define ACPI_EXCLUSIVE (u8) 0x00 #define ACPI_SHARED (u8) 0x01 +#define ACPI_EXCLUSIVE_AND_WAKE (u8) 0x02 +#define ACPI_SHARED_AND_WAKE (u8) 0x03 /* * DMA Attributes @@ -127,6 +140,8 @@ typedef u32 acpi_rsdesc_size; /* Max Resource Descriptor size is (Length+3) = (6 #define ACPI_POS_DECODE (u8) 0x00 #define ACPI_SUB_DECODE (u8) 0x01 +/* Producer/Consumer */ + #define ACPI_PRODUCER (u8) 0x00 #define ACPI_CONSUMER (u8) 0x01 @@ -192,6 +207,21 @@ struct acpi_resource_fixed_io { u8 address_length; }; +struct acpi_resource_fixed_dma { + u16 request_lines; + u16 channels; + u8 width; +}; + +/* Values for Width field above */ + +#define ACPI_DMA_WIDTH8 0 +#define ACPI_DMA_WIDTH16 1 +#define ACPI_DMA_WIDTH32 2 +#define ACPI_DMA_WIDTH64 3 +#define ACPI_DMA_WIDTH128 4 +#define ACPI_DMA_WIDTH256 5 + struct acpi_resource_vendor { u16 byte_length; u8 byte_data[1]; @@ -329,6 +359,166 @@ struct acpi_resource_generic_register { u64 address; }; +struct acpi_resource_gpio { + u8 revision_id; + u8 connection_type; + u8 producer_consumer; /* For values, see Producer/Consumer above */ + u8 pin_config; + u8 sharable; /* For values, see Interrupt Attributes above */ + u8 io_restriction; + u8 triggering; /* For values, see Interrupt Attributes above */ + u8 polarity; /* For values, see Interrupt Attributes above */ + u16 drive_strength; + u16 debounce_timeout; + u16 pin_table_length; + u16 vendor_length; + struct acpi_resource_source resource_source; + u16 *pin_table; + u8 *vendor_data; +}; + +/* Values for GPIO connection_type field above */ + +#define ACPI_RESOURCE_GPIO_TYPE_INT 0 +#define ACPI_RESOURCE_GPIO_TYPE_IO 1 + +/* Values for pin_config field above */ + +#define ACPI_PIN_CONFIG_DEFAULT 0 +#define ACPI_PIN_CONFIG_PULLUP 1 +#define ACPI_PIN_CONFIG_PULLDOWN 2 +#define ACPI_PIN_CONFIG_NOPULL 3 + +/* Values for io_restriction field above */ + +#define ACPI_IO_RESTRICT_NONE 0 +#define ACPI_IO_RESTRICT_INPUT 1 +#define ACPI_IO_RESTRICT_OUTPUT 2 +#define ACPI_IO_RESTRICT_NONE_PRESERVE 3 + +/* Common structure for I2C, SPI, and UART serial descriptors */ + +#define ACPI_RESOURCE_SERIAL_COMMON \ + u8 revision_id; \ + u8 type; \ + u8 producer_consumer; /* For values, see Producer/Consumer above */\ + u8 slave_mode; \ + u8 type_revision_id; \ + u16 type_data_length; \ + u16 vendor_length; \ + struct acpi_resource_source resource_source; \ + u8 *vendor_data; + +struct acpi_resource_common_serialbus { +ACPI_RESOURCE_SERIAL_COMMON}; + +/* Values for the Type field above */ + +#define ACPI_RESOURCE_SERIAL_TYPE_I2C 1 +#define ACPI_RESOURCE_SERIAL_TYPE_SPI 2 +#define ACPI_RESOURCE_SERIAL_TYPE_UART 3 + +/* Values for slave_mode field above */ + +#define ACPI_CONTROLLER_INITIATED 0 +#define ACPI_DEVICE_INITIATED 1 + +struct acpi_resource_i2c_serialbus { + ACPI_RESOURCE_SERIAL_COMMON u8 access_mode; + u16 slave_address; + u32 connection_speed; +}; + +/* Values for access_mode field above */ + +#define ACPI_I2C_7BIT_MODE 0 +#define ACPI_I2C_10BIT_MODE 1 + +struct acpi_resource_spi_serialbus { + ACPI_RESOURCE_SERIAL_COMMON u8 wire_mode; + u8 device_polarity; + u8 data_bit_length; + u8 clock_phase; + u8 clock_polarity; + u16 device_selection; + u32 connection_speed; +}; + +/* Values for wire_mode field above */ + +#define ACPI_SPI_4WIRE_MODE 0 +#define ACPI_SPI_3WIRE_MODE 1 + +/* Values for device_polarity field above */ + +#define ACPI_SPI_ACTIVE_LOW 0 +#define ACPI_SPI_ACTIVE_HIGH 1 + +/* Values for clock_phase field above */ + +#define ACPI_SPI_FIRST_PHASE 0 +#define ACPI_SPI_SECOND_PHASE 1 + +/* Values for clock_polarity field above */ + +#define ACPI_SPI_START_LOW 0 +#define ACPI_SPI_START_HIGH 1 + +struct acpi_resource_uart_serialbus { + ACPI_RESOURCE_SERIAL_COMMON u8 endian; + u8 data_bits; + u8 stop_bits; + u8 flow_control; + u8 parity; + u8 lines_enabled; + u16 rx_fifo_size; + u16 tx_fifo_size; + u32 default_baud_rate; +}; + +/* Values for Endian field above */ + +#define ACPI_UART_LITTLE_ENDIAN 0 +#define ACPI_UART_BIG_ENDIAN 1 + +/* Values for data_bits field above */ + +#define ACPI_UART_5_DATA_BITS 0 +#define ACPI_UART_6_DATA_BITS 1 +#define ACPI_UART_7_DATA_BITS 2 +#define ACPI_UART_8_DATA_BITS 3 +#define ACPI_UART_9_DATA_BITS 4 + +/* Values for stop_bits field above */ + +#define ACPI_UART_NO_STOP_BITS 0 +#define ACPI_UART_1_STOP_BIT 1 +#define ACPI_UART_1P5_STOP_BITS 2 +#define ACPI_UART_2_STOP_BITS 3 + +/* Values for flow_control field above */ + +#define ACPI_UART_FLOW_CONTROL_NONE 0 +#define ACPI_UART_FLOW_CONTROL_HW 1 +#define ACPI_UART_FLOW_CONTROL_XON_XOFF 2 + +/* Values for Parity field above */ + +#define ACPI_UART_PARITY_NONE 0 +#define ACPI_UART_PARITY_EVEN 1 +#define ACPI_UART_PARITY_ODD 2 +#define ACPI_UART_PARITY_MARK 3 +#define ACPI_UART_PARITY_SPACE 4 + +/* Values for lines_enabled bitfield above */ + +#define ACPI_UART_CARRIER_DETECT (1<<2) +#define ACPI_UART_RING_INDICATOR (1<<3) +#define ACPI_UART_DATA_SET_READY (1<<4) +#define ACPI_UART_DATA_TERMINAL_READY (1<<5) +#define ACPI_UART_CLEAR_TO_SEND (1<<6) +#define ACPI_UART_REQUEST_TO_SEND (1<<7) + /* ACPI_RESOURCE_TYPEs */ #define ACPI_RESOURCE_TYPE_IRQ 0 @@ -348,7 +538,10 @@ struct acpi_resource_generic_register { #define ACPI_RESOURCE_TYPE_EXTENDED_ADDRESS64 14 /* ACPI 3.0 */ #define ACPI_RESOURCE_TYPE_EXTENDED_IRQ 15 #define ACPI_RESOURCE_TYPE_GENERIC_REGISTER 16 -#define ACPI_RESOURCE_TYPE_MAX 16 +#define ACPI_RESOURCE_TYPE_GPIO 17 /* ACPI 5.0 */ +#define ACPI_RESOURCE_TYPE_FIXED_DMA 18 /* ACPI 5.0 */ +#define ACPI_RESOURCE_TYPE_SERIAL_BUS 19 /* ACPI 5.0 */ +#define ACPI_RESOURCE_TYPE_MAX 19 /* Master union for resource descriptors */ @@ -358,6 +551,7 @@ union acpi_resource_data { struct acpi_resource_start_dependent start_dpf; struct acpi_resource_io io; struct acpi_resource_fixed_io fixed_io; + struct acpi_resource_fixed_dma fixed_dma; struct acpi_resource_vendor vendor; struct acpi_resource_vendor_typed vendor_typed; struct acpi_resource_end_tag end_tag; @@ -370,6 +564,11 @@ union acpi_resource_data { struct acpi_resource_extended_address64 ext_address64; struct acpi_resource_extended_irq extended_irq; struct acpi_resource_generic_register generic_reg; + struct acpi_resource_gpio gpio; + struct acpi_resource_i2c_serialbus i2c_serial_bus; + struct acpi_resource_spi_serialbus spi_serial_bus; + struct acpi_resource_uart_serialbus uart_serial_bus; + struct acpi_resource_common_serialbus common_serial_bus; /* Common fields */ From a91cdde20a870bd773d605c764ed211539bf3020 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 14:46:57 +0800 Subject: [PATCH 442/820] ACPI 5.0: New interface, acpi_get_event_resources Executes _AEI and formats the result, similar to acpi_get_current_resources, etc. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/acresrc.h | 4 +++ drivers/acpi/acpica/rsutils.c | 50 +++++++++++++++++++++++++++++++++++ drivers/acpi/acpica/rsxface.c | 40 ++++++++++++++++++++++++++++ include/acpi/acpixf.h | 4 +++ 4 files changed, 98 insertions(+) diff --git a/drivers/acpi/acpica/acresrc.h b/drivers/acpi/acpica/acresrc.h index 05721b023f88..61b2e049d8d6 100644 --- a/drivers/acpi/acpica/acresrc.h +++ b/drivers/acpi/acpica/acresrc.h @@ -215,6 +215,10 @@ acpi_status acpi_rs_set_srs_method_data(struct acpi_namespace_node *node, struct acpi_buffer *ret_buffer); +acpi_status +acpi_rs_get_aei_method_data(struct acpi_namespace_node *node, + struct acpi_buffer *ret_buffer); + /* * rscalc */ diff --git a/drivers/acpi/acpica/rsutils.c b/drivers/acpi/acpica/rsutils.c index 4409db8209d7..54a9a974b624 100644 --- a/drivers/acpi/acpica/rsutils.c +++ b/drivers/acpi/acpica/rsutils.c @@ -592,6 +592,56 @@ acpi_rs_get_prs_method_data(struct acpi_namespace_node *node, } #endif /* ACPI_FUTURE_USAGE */ +/******************************************************************************* + * + * FUNCTION: acpi_rs_get_aei_method_data + * + * PARAMETERS: Node - Device node + * ret_buffer - Pointer to a buffer structure for the + * results + * + * RETURN: Status + * + * DESCRIPTION: This function is called to get the _AEI value of an object + * contained in an object specified by the handle passed in + * + * If the function fails an appropriate status will be returned + * and the contents of the callers buffer is undefined. + * + ******************************************************************************/ + +acpi_status +acpi_rs_get_aei_method_data(struct acpi_namespace_node *node, + struct acpi_buffer *ret_buffer) +{ + union acpi_operand_object *obj_desc; + acpi_status status; + + ACPI_FUNCTION_TRACE(rs_get_aei_method_data); + + /* Parameters guaranteed valid by caller */ + + /* Execute the method, no parameters */ + + status = acpi_ut_evaluate_object(node, METHOD_NAME__AEI, + ACPI_BTYPE_BUFFER, &obj_desc); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + /* + * Make the call to create a resource linked list from the + * byte stream buffer that comes back from the _CRS method + * execution. + */ + status = acpi_rs_create_resource_list(obj_desc, ret_buffer); + + /* On exit, we must delete the object returned by evaluate_object */ + + acpi_ut_remove_reference(obj_desc); + return_ACPI_STATUS(status); +} + /******************************************************************************* * * FUNCTION: acpi_rs_get_method_data diff --git a/drivers/acpi/acpica/rsxface.c b/drivers/acpi/acpica/rsxface.c index fe86b37b16ce..45a9e5e4e4db 100644 --- a/drivers/acpi/acpica/rsxface.c +++ b/drivers/acpi/acpica/rsxface.c @@ -307,6 +307,46 @@ acpi_set_current_resources(acpi_handle device_handle, ACPI_EXPORT_SYMBOL(acpi_set_current_resources) +/******************************************************************************* + * + * FUNCTION: acpi_get_event_resources + * + * PARAMETERS: device_handle - Handle to the device object for the + * device we are getting resources + * in_buffer - Pointer to a buffer containing the + * resources to be set for the device + * + * RETURN: Status + * + * DESCRIPTION: This function is called to get the event resources for a + * specific device. The caller must first acquire a handle for + * the desired device. The resource data is passed to the routine + * the buffer pointed to by the in_buffer variable. Uses the + * _AEI method. + * + ******************************************************************************/ +acpi_status +acpi_get_event_resources(acpi_handle device_handle, + struct acpi_buffer *ret_buffer) +{ + acpi_status status; + struct acpi_namespace_node *node; + + ACPI_FUNCTION_TRACE(acpi_get_event_resources); + + /* Validate parameters then dispatch to internal routine */ + + status = acpi_rs_validate_parameters(device_handle, ret_buffer, &node); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + status = acpi_rs_get_aei_method_data(node, ret_buffer); + return_ACPI_STATUS(status); +} + +ACPI_EXPORT_SYMBOL(acpi_get_event_resources) + /****************************************************************************** * * FUNCTION: acpi_resource_to_address64 diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index dd86610039b3..726d937ef5b1 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -357,6 +357,10 @@ acpi_status acpi_get_possible_resources(acpi_handle device, struct acpi_buffer *ret_buffer); #endif +acpi_status +acpi_get_event_resources(acpi_handle device_handle, + struct acpi_buffer *ret_buffer); + acpi_status acpi_walk_resources(acpi_handle device, char *name, From 0e243178047c0219b3367dd44f81040826b7ea83 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 14:51:01 +0800 Subject: [PATCH 443/820] ACPI 5.0: New interface, acpi_buffer_to_resource This interface converts an AML buffer to an internal ACPI_RESOURCE. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/rscreate.c | 64 ++++++++++++++++++++++++++++++++++ include/acpi/acpixf.h | 5 +++ 2 files changed, 69 insertions(+) diff --git a/drivers/acpi/acpica/rscreate.c b/drivers/acpi/acpica/rscreate.c index f61285d5edc3..06264e8b58d9 100644 --- a/drivers/acpi/acpica/rscreate.c +++ b/drivers/acpi/acpica/rscreate.c @@ -49,6 +49,70 @@ #define _COMPONENT ACPI_RESOURCES ACPI_MODULE_NAME("rscreate") +/******************************************************************************* + * + * FUNCTION: acpi_buffer_to_resource + * + * PARAMETERS: aml_buffer - Pointer to the resource byte stream + * aml_buffer_length - Length of the aml_buffer + * resource_ptr - Where the converted resource is returned + * + * RETURN: Status + * + * DESCRIPTION: Convert a raw AML buffer to a resource list + * + ******************************************************************************/ +acpi_status +acpi_buffer_to_resource(u8 *aml_buffer, + u16 aml_buffer_length, + struct acpi_resource **resource_ptr) +{ + acpi_status status; + acpi_size list_size_needed; + void *resource; + void *current_resource_ptr; + + /* + * Note: we allow AE_AML_NO_RESOURCE_END_TAG, since an end tag + * is not required here. + */ + + /* Get the required length for the converted resource */ + + status = acpi_rs_get_list_length(aml_buffer, aml_buffer_length, + &list_size_needed); + if (status == AE_AML_NO_RESOURCE_END_TAG) { + status = AE_OK; + } + if (ACPI_FAILURE(status)) { + return (status); + } + + /* Allocate a buffer for the converted resource */ + + resource = ACPI_ALLOCATE_ZEROED(list_size_needed); + current_resource_ptr = resource; + if (!resource) { + return (AE_NO_MEMORY); + } + + /* Perform the AML-to-Resource conversion */ + + status = acpi_ut_walk_aml_resources(aml_buffer, aml_buffer_length, + acpi_rs_convert_aml_to_resources, + ¤t_resource_ptr); + if (status == AE_AML_NO_RESOURCE_END_TAG) { + status = AE_OK; + } + if (ACPI_FAILURE(status)) { + ACPI_FREE(resource); + } else { + *resource_ptr = resource; + } + + return (status); +} + /******************************************************************************* * * FUNCTION: acpi_rs_create_resource_list diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 726d937ef5b1..5d8df1dfb203 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -376,6 +376,11 @@ acpi_status acpi_resource_to_address64(struct acpi_resource *resource, struct acpi_resource_address64 *out); +acpi_status +acpi_buffer_to_resource(u8 *aml_buffer, + u16 aml_buffer_length, + struct acpi_resource **resource_ptr); + /* * Hardware (ACPI device) interfaces */ From 7fce7a4b0db7d90db553faf568853d56cd908c39 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Wed, 16 Nov 2011 14:59:17 +0800 Subject: [PATCH 444/820] ACPI 5.0: Add new predefined names Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/aclocal.h | 3 ++- drivers/acpi/acpica/acpredef.h | 39 ++++++++++++++++++++++++++++++++- drivers/acpi/acpica/nspredef.c | 29 ++++++++++++++++++++++++ drivers/acpi/acpica/nsrepair.c | 1 + drivers/acpi/acpica/nsrepair2.c | 5 +++-- 5 files changed, 73 insertions(+), 4 deletions(-) diff --git a/drivers/acpi/acpica/aclocal.h b/drivers/acpi/acpica/aclocal.h index 60aee56acc5b..2ddd61d07a06 100644 --- a/drivers/acpi/acpica/aclocal.h +++ b/drivers/acpi/acpica/aclocal.h @@ -319,7 +319,8 @@ struct acpi_name_info { /* * Used for ACPI_PTYPE1_FIXED, ACPI_PTYPE1_VAR, ACPI_PTYPE2, - * ACPI_PTYPE2_MIN, ACPI_PTYPE2_PKG_COUNT, ACPI_PTYPE2_COUNT + * ACPI_PTYPE2_MIN, ACPI_PTYPE2_PKG_COUNT, ACPI_PTYPE2_COUNT, + * ACPI_PTYPE2_FIX_VAR */ struct acpi_package_info { u8 type; diff --git a/drivers/acpi/acpica/acpredef.h b/drivers/acpi/acpica/acpredef.h index c445cca490ea..ecac9cf38769 100644 --- a/drivers/acpi/acpica/acpredef.h +++ b/drivers/acpi/acpica/acpredef.h @@ -94,6 +94,14 @@ * ACPI_PTYPE2_REV_FIXED: Revision at start, each subpackage is Fixed-length * (Used for _ART, _FPS) * + * ACPI_PTYPE2_FIX_VAR: Each subpackage consists of some fixed-length elements + * followed by an optional element + * object type + * count + * object type + * count = 0 (optional) + * (Used for _DLM) + * *****************************************************************************/ enum acpi_return_package_types { @@ -105,7 +113,8 @@ enum acpi_return_package_types { ACPI_PTYPE2_PKG_COUNT = 6, ACPI_PTYPE2_FIXED = 7, ACPI_PTYPE2_MIN = 8, - ACPI_PTYPE2_REV_FIXED = 9 + ACPI_PTYPE2_REV_FIXED = 9, + ACPI_PTYPE2_FIX_VAR = 10 }; #ifdef ACPI_CREATE_PREDEFINED_TABLE @@ -154,6 +163,7 @@ static const union acpi_predefined_info predefined_names[] = {{"_AC8", 0, ACPI_RTYPE_INTEGER}}, {{"_AC9", 0, ACPI_RTYPE_INTEGER}}, {{"_ADR", 0, ACPI_RTYPE_INTEGER}}, + {{"_AEI", 0, ACPI_RTYPE_BUFFER}}, {{"_AL0", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Refs) */ {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_REFERENCE, 0,0}, 0,0}}, @@ -229,6 +239,13 @@ static const union acpi_predefined_info predefined_names[] = {{"_CID", 0, ACPI_RTYPE_INTEGER | ACPI_RTYPE_STRING | ACPI_RTYPE_PACKAGE}}, /* Variable-length (Ints/Strs) */ {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_INTEGER | ACPI_RTYPE_STRING, 0,0}, 0,0}}, + {{"_CLS", 0, ACPI_RTYPE_PACKAGE}}, /* Fixed-length (3 Int) */ + {{{ACPI_PTYPE1_FIXED, ACPI_RTYPE_INTEGER, 3, 0}, 0, 0}}, + + {{"_CPC", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Ints/Bufs) */ + {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_INTEGER | ACPI_RTYPE_BUFFER, 0, 0}, 0, + 0}}, + {{"_CRS", 0, ACPI_RTYPE_BUFFER}}, {{"_CRT", 0, ACPI_RTYPE_INTEGER}}, {{"_CSD", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (1 Int(n), n-1 Int) */ @@ -237,12 +254,21 @@ static const union acpi_predefined_info predefined_names[] = {{"_CST", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (1 Int(n), n Pkg (1 Buf/3 Int) */ {{{ACPI_PTYPE2_PKG_COUNT,ACPI_RTYPE_BUFFER, 1, ACPI_RTYPE_INTEGER}, 3,0}}, + {{"_CWS", 1, ACPI_RTYPE_INTEGER}}, {{"_DCK", 1, ACPI_RTYPE_INTEGER}}, {{"_DCS", 0, ACPI_RTYPE_INTEGER}}, {{"_DDC", 1, ACPI_RTYPE_INTEGER | ACPI_RTYPE_BUFFER}}, {{"_DDN", 0, ACPI_RTYPE_STRING}}, + {{"_DEP", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Refs) */ + {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_REFERENCE, 0, 0}, 0, 0}}, + {{"_DGS", 0, ACPI_RTYPE_INTEGER}}, {{"_DIS", 0, 0}}, + + {{"_DLM", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Pkgs) each (1 Ref, 0/1 Optional Buf/Ref) */ + {{{ACPI_PTYPE2_FIX_VAR, ACPI_RTYPE_REFERENCE, 1, + ACPI_RTYPE_REFERENCE | ACPI_RTYPE_BUFFER}, 0, 0}}, + {{"_DMA", 0, ACPI_RTYPE_BUFFER}}, {{"_DOD", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Ints) */ {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_INTEGER, 0,0}, 0,0}}, @@ -262,6 +288,7 @@ static const union acpi_predefined_info predefined_names[] = {{"_EJ3", 1, 0}}, {{"_EJ4", 1, 0}}, {{"_EJD", 0, ACPI_RTYPE_STRING}}, + {{"_EVT", 1, 0}}, {{"_FDE", 0, ACPI_RTYPE_BUFFER}}, {{"_FDI", 0, ACPI_RTYPE_PACKAGE}}, /* Fixed-length (16 Int) */ {{{ACPI_PTYPE1_FIXED, ACPI_RTYPE_INTEGER, 16,0}, 0,0}}, @@ -281,14 +308,17 @@ static const union acpi_predefined_info predefined_names[] = {{{ACPI_PTYPE1_FIXED, ACPI_RTYPE_INTEGER, 3, 0}, 0, 0}}, {{"_GAI", 0, ACPI_RTYPE_INTEGER}}, + {{"_GCP", 0, ACPI_RTYPE_INTEGER}}, {{"_GHL", 0, ACPI_RTYPE_INTEGER}}, {{"_GLK", 0, ACPI_RTYPE_INTEGER}}, {{"_GPD", 0, ACPI_RTYPE_INTEGER}}, {{"_GPE", 0, ACPI_RTYPE_INTEGER}}, /* _GPE method, not _GPE scope */ + {{"_GRT", 0, ACPI_RTYPE_BUFFER}}, {{"_GSB", 0, ACPI_RTYPE_INTEGER}}, {{"_GTF", 0, ACPI_RTYPE_BUFFER}}, {{"_GTM", 0, ACPI_RTYPE_BUFFER}}, {{"_GTS", 1, 0}}, + {{"_GWS", 1, ACPI_RTYPE_INTEGER}}, {{"_HID", 0, ACPI_RTYPE_INTEGER | ACPI_RTYPE_STRING}}, {{"_HOT", 0, ACPI_RTYPE_INTEGER}}, {{"_HPP", 0, ACPI_RTYPE_PACKAGE}}, /* Fixed-length (4 Int) */ @@ -303,6 +333,7 @@ static const union acpi_predefined_info predefined_names[] = {{"_HPX", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Pkgs) each (var Ints) */ {{{ACPI_PTYPE2_MIN, ACPI_RTYPE_INTEGER, 5,0}, 0,0}}, + {{"_HRV", 0, ACPI_RTYPE_INTEGER}}, {{"_IFT", 0, ACPI_RTYPE_INTEGER}}, /* See IPMI spec */ {{"_INI", 0, 0}}, {{"_IRC", 0, 0}}, @@ -361,6 +392,9 @@ static const union acpi_predefined_info predefined_names[] = {{"_PR3", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Refs) */ {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_REFERENCE, 0, 0}, 0, 0}}, + {{"_PRE", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Refs) */ + {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_REFERENCE, 0, 0}, 0, 0}}, + {{"_PRL", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Refs) */ {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_REFERENCE, 0, 0}, 0, 0}}, @@ -391,6 +425,7 @@ static const union acpi_predefined_info predefined_names[] = {{"_PSD", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Pkgs) each (5 Int) with count */ {{{ACPI_PTYPE2_COUNT, ACPI_RTYPE_INTEGER,0,0}, 0,0}}, + {{"_PSE", 1, 0}}, {{"_PSL", 0, ACPI_RTYPE_PACKAGE}}, /* Variable-length (Refs) */ {{{ACPI_PTYPE1_VAR, ACPI_RTYPE_REFERENCE, 0,0}, 0,0}}, @@ -457,6 +492,7 @@ static const union acpi_predefined_info predefined_names[] = {{"_SLI", 0, ACPI_RTYPE_BUFFER}}, {{"_SPD", 1, ACPI_RTYPE_INTEGER}}, {{"_SRS", 1, 0}}, + {{"_SRT", 1, ACPI_RTYPE_INTEGER}}, {{"_SRV", 0, ACPI_RTYPE_INTEGER}}, /* See IPMI spec */ {{"_SST", 1, 0}}, {{"_STA", 0, ACPI_RTYPE_INTEGER}}, @@ -464,6 +500,7 @@ static const union acpi_predefined_info predefined_names[] = {{"_STP", 2, ACPI_RTYPE_INTEGER}}, {{"_STR", 0, ACPI_RTYPE_BUFFER}}, {{"_STV", 2, ACPI_RTYPE_INTEGER}}, + {{"_SUB", 0, ACPI_RTYPE_STRING}}, {{"_SUN", 0, ACPI_RTYPE_INTEGER}}, {{"_SWS", 0, ACPI_RTYPE_INTEGER}}, {{"_TC1", 0, ACPI_RTYPE_INTEGER}}, diff --git a/drivers/acpi/acpica/nspredef.c b/drivers/acpi/acpica/nspredef.c index c845c8089f39..db80366b3dd9 100644 --- a/drivers/acpi/acpica/nspredef.c +++ b/drivers/acpi/acpica/nspredef.c @@ -620,6 +620,7 @@ acpi_ns_check_package(struct acpi_predefined_data *data, case ACPI_PTYPE2_FIXED: case ACPI_PTYPE2_MIN: case ACPI_PTYPE2_COUNT: + case ACPI_PTYPE2_FIX_VAR: /* * These types all return a single Package that consists of a @@ -759,6 +760,34 @@ acpi_ns_check_package_list(struct acpi_predefined_data *data, } break; + case ACPI_PTYPE2_FIX_VAR: + /* + * Each subpackage has a fixed number of elements and an + * optional element + */ + expected_count = + package->ret_info.count1 + package->ret_info.count2; + if (sub_package->package.count < expected_count) { + goto package_too_small; + } + + status = + acpi_ns_check_package_elements(data, sub_elements, + package->ret_info. + object_type1, + package->ret_info. + count1, + package->ret_info. + object_type2, + sub_package->package. + count - + package->ret_info. + count1, 0); + if (ACPI_FAILURE(status)) { + return (status); + } + break; + case ACPI_PTYPE2_FIXED: /* Each sub-package has a fixed length */ diff --git a/drivers/acpi/acpica/nsrepair.c b/drivers/acpi/acpica/nsrepair.c index ac7b854b0bd7..dd0a0bdc4fc7 100644 --- a/drivers/acpi/acpica/nsrepair.c +++ b/drivers/acpi/acpica/nsrepair.c @@ -634,6 +634,7 @@ acpi_ns_remove_null_elements(struct acpi_predefined_data *data, case ACPI_PTYPE2_FIXED: case ACPI_PTYPE2_MIN: case ACPI_PTYPE2_REV_FIXED: + case ACPI_PTYPE2_FIX_VAR: break; default: diff --git a/drivers/acpi/acpica/nsrepair2.c b/drivers/acpi/acpica/nsrepair2.c index 024c4f263f87..3ee2b22a9f7d 100644 --- a/drivers/acpi/acpica/nsrepair2.c +++ b/drivers/acpi/acpica/nsrepair2.c @@ -467,11 +467,12 @@ acpi_ns_repair_HID(struct acpi_predefined_data *data, } /* - * Copy and uppercase the string. From the ACPI specification: + * Copy and uppercase the string. From the ACPI 5.0 specification: * * A valid PNP ID must be of the form "AAA####" where A is an uppercase * letter and # is a hex digit. A valid ACPI ID must be of the form - * "ACPI####" where # is a hex digit. + * "NNNN####" where N is an uppercase letter or decimal digit, and + * # is a hex digit. */ for (dest = new_string->string.pointer; *source; dest++, source++) { *dest = (char)ACPI_TOUPPER(*source); From 25169b77f7c33b363db7d099d67491e977cdefc7 Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Wed, 16 Nov 2011 17:03:54 +0800 Subject: [PATCH 445/820] ACPI 5.0: Allow _AEI method in walk resources _AEI contains a resource template, this change adds support for the walk resources function. Signed-off-by: Zhang Rui Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/rsxface.c | 10 ++++++---- include/acpi/acnames.h | 1 + 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/acpi/acpica/rsxface.c b/drivers/acpi/acpica/rsxface.c index 45a9e5e4e4db..f6cceaccc3e9 100644 --- a/drivers/acpi/acpica/rsxface.c +++ b/drivers/acpi/acpica/rsxface.c @@ -526,8 +526,9 @@ acpi_rs_match_vendor_resource(struct acpi_resource *resource, void *context) * * PARAMETERS: device_handle - Handle to the device object for the * device we are querying - * Name - Method name of the resources we want - * (METHOD_NAME__CRS or METHOD_NAME__PRS) + * Name - Method name of the resources we want. + * (METHOD_NAME__CRS, METHOD_NAME__PRS, or + * METHOD_NAME__AEI) * user_function - Called for each resource * Context - Passed to user_function * @@ -554,11 +555,12 @@ acpi_walk_resources(acpi_handle device_handle, if (!device_handle || !user_function || !name || (!ACPI_COMPARE_NAME(name, METHOD_NAME__CRS) && - !ACPI_COMPARE_NAME(name, METHOD_NAME__PRS))) { + !ACPI_COMPARE_NAME(name, METHOD_NAME__PRS) && + !ACPI_COMPARE_NAME(name, METHOD_NAME__AEI))) { return_ACPI_STATUS(AE_BAD_PARAMETER); } - /* Get the _CRS or _PRS resource list */ + /* Get the _CRS/_PRS/_AEI resource list */ buffer.length = ACPI_ALLOCATE_LOCAL_BUFFER; status = acpi_rs_get_method_data(device_handle, name, &buffer); diff --git a/include/acpi/acnames.h b/include/acpi/acnames.h index fc1575fd4596..5b5af0d30a97 100644 --- a/include/acpi/acnames.h +++ b/include/acpi/acnames.h @@ -58,6 +58,7 @@ #define METHOD_NAME__PRT "_PRT" #define METHOD_NAME__CRS "_CRS" #define METHOD_NAME__PRS "_PRS" +#define METHOD_NAME__AEI "_AEI" #define METHOD_NAME__PRW "_PRW" #define METHOD_NAME__SRS "_SRS" From 5dc46c3f13b2f538402a09083bdf67d47836fd81 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Tue, 29 Nov 2011 14:16:05 +0800 Subject: [PATCH 446/820] ACPICA: Clean up Makefile Make the various files in alphabetical order to simplify addition of new files. Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/Makefile | 157 +++++++++++++++++++++++++++++------ 1 file changed, 133 insertions(+), 24 deletions(-) diff --git a/drivers/acpi/acpica/Makefile b/drivers/acpi/acpica/Makefile index b0af3b83eda8..5f7d0cd11002 100644 --- a/drivers/acpi/acpica/Makefile +++ b/drivers/acpi/acpica/Makefile @@ -8,41 +8,150 @@ ccflags-$(CONFIG_ACPI_DEBUG) += -DACPI_DEBUG_OUTPUT # use acpi.o to put all files here into acpi.o modparam namespace obj-y += acpi.o -acpi-y := dsfield.o dsmthdat.o dsopcode.o dswexec.o dswscope.o \ - dsmethod.o dsobject.o dsutils.o dswload.o dswstate.o \ - dsinit.o dsargs.o dscontrol.o dswload2.o +acpi-y := \ + dsargs.o \ + dscontrol.o \ + dsfield.o \ + dsinit.o \ + dsmethod.o \ + dsmthdat.o \ + dsobject.o \ + dsopcode.o \ + dsutils.o \ + dswexec.o \ + dswload.o \ + dswload2.o \ + dswscope.o \ + dswstate.o -acpi-y += evevent.o evregion.o evsci.o evxfevnt.o \ - evmisc.o evrgnini.o evxface.o evxfregn.o \ - evgpe.o evgpeblk.o evgpeinit.o evgpeutil.o evxfgpe.o evglock.o +acpi-y += \ + evevent.o \ + evgpe.o \ + evgpeblk.o \ + evgpeinit.o \ + evgpeutil.o \ + evglock.o \ + evmisc.o \ + evregion.o \ + evrgnini.o \ + evsci.o \ + evxface.o \ + evxfevnt.o \ + evxfgpe.o \ + evxfregn.o -acpi-y += exconfig.o exfield.o exnames.o exoparg6.o exresolv.o exstorob.o\ - exconvrt.o exfldio.o exoparg1.o exprep.o exresop.o exsystem.o\ - excreate.o exmisc.o exoparg2.o exregion.o exstore.o exutils.o \ - exdump.o exmutex.o exoparg3.o exresnte.o exstoren.o exdebug.o +acpi-y += \ + exconfig.o \ + exconvrt.o \ + excreate.o \ + exdebug.o \ + exdump.o \ + exfield.o \ + exfldio.o \ + exmutex.o \ + exnames.o \ + exoparg1.o \ + exoparg2.o \ + exoparg3.o \ + exoparg6.o \ + exprep.o \ + exmisc.o \ + exregion.o \ + exresnte.o \ + exresolv.o \ + exresop.o \ + exstore.o \ + exstoren.o \ + exstorob.o \ + exsystem.o \ + exutils.o -acpi-y += hwacpi.o hwgpe.o hwregs.o hwsleep.o hwxface.o hwvalid.o hwpci.o +acpi-y += \ + hwacpi.o \ + hwgpe.o \ + hwpci.o \ + hwregs.o \ + hwsleep.o \ + hwvalid.o \ + hwxface.o acpi-$(ACPI_FUTURE_USAGE) += hwtimer.o -acpi-y += nsaccess.o nsload.o nssearch.o nsxfeval.o \ - nsalloc.o nseval.o nsnames.o nsutils.o nsxfname.o \ - nsdump.o nsinit.o nsobject.o nswalk.o nsxfobj.o \ - nsparse.o nspredef.o nsrepair.o nsrepair2.o +acpi-y += \ + nsaccess.o \ + nsalloc.o \ + nsdump.o \ + nseval.o \ + nsinit.o \ + nsload.o \ + nsnames.o \ + nsobject.o \ + nsparse.o \ + nspredef.o \ + nsrepair.o \ + nsrepair2.o \ + nssearch.o \ + nsutils.o \ + nswalk.o \ + nsxfeval.o \ + nsxfname.o \ + nsxfobj.o acpi-$(ACPI_FUTURE_USAGE) += nsdumpdv.o -acpi-y += psargs.o psparse.o psloop.o pstree.o pswalk.o \ - psopcode.o psscope.o psutils.o psxface.o +acpi-y += \ + psargs.o \ + psloop.o \ + psopcode.o \ + psparse.o \ + psscope.o \ + pstree.o \ + psutils.o \ + pswalk.o \ + psxface.o -acpi-y += rsaddr.o rscreate.o rsinfo.o rsio.o rslist.o rsmisc.o rsxface.o \ - rscalc.o rsirq.o rsmemory.o rsutils.o rsserial.o +acpi-y += \ + rsaddr.o \ + rscalc.o \ + rscreate.o \ + rsinfo.o \ + rsio.o \ + rsirq.o \ + rslist.o \ + rsmemory.o \ + rsmisc.o \ + rsserial.o \ + rsutils.o \ + rsxface.o acpi-$(ACPI_FUTURE_USAGE) += rsdump.o -acpi-y += tbxface.o tbinstal.o tbutils.o tbfind.o tbfadt.o tbxfroot.o +acpi-y += \ + tbfadt.o \ + tbfind.o \ + tbinstal.o \ + tbutils.o \ + tbxface.o \ + tbxfroot.o -acpi-y += utalloc.o utdebug.o uteval.o utinit.o utmisc.o utxface.o \ - utcopy.o utdelete.o utglobal.o utmath.o utobject.o \ - utstate.o utmutex.o utobject.o utresrc.o utlock.o utids.o \ - utosi.o utxferror.o utdecode.o utxfmutex.o +acpi-y += \ + utalloc.o \ + utcopy.o \ + utdebug.o \ + utdecode.o \ + utdelete.o \ + uteval.o \ + utglobal.o \ + utids.o \ + utinit.o \ + utlock.o \ + utmath.o \ + utmisc.o \ + utmutex.o \ + utobject.o \ + utosi.o \ + utresrc.o \ + utstate.o \ + utxface.o \ + utxferror.o \ + utxfmutex.o From 240c320c59aba93605bff761ceb698df3ee90761 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Mon, 28 Nov 2011 10:18:24 +0800 Subject: [PATCH 447/820] ACPICA: Update to version 20111123 This version contains full support for the ACPI 5.0 specification. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/acpixf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 5d8df1dfb203..30f5bef4f14d 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -47,7 +47,7 @@ /* Current ACPICA subsystem version in YYYYMMDD format */ -#define ACPI_CA_VERSION 0x20110623 +#define ACPI_CA_VERSION 0x20111123 #include "actypes.h" #include "actbl.h" From ecafe6faa294697305453b5a6af7760b94622787 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Thu, 12 Jan 2012 11:09:58 +0800 Subject: [PATCH 448/820] ACPICA: Update for larger ACPI 5 FADT size FADT is now larger than 256 bytes, so all FADT offsets must be changed from 8 bits to 16 bits. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/tbfadt.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/acpi/acpica/tbfadt.c b/drivers/acpi/acpica/tbfadt.c index 1e1a7de34439..ff1f7f041f59 100644 --- a/drivers/acpi/acpica/tbfadt.c +++ b/drivers/acpi/acpica/tbfadt.c @@ -63,14 +63,15 @@ static void acpi_tb_setup_fadt_registers(void); typedef struct acpi_fadt_info { char *name; - u8 address64; - u8 address32; - u8 length; + u16 address64; + u16 address32; + u16 length; u8 default_length; u8 type; } acpi_fadt_info; +#define ACPI_FADT_OPTIONAL 0 #define ACPI_FADT_REQUIRED 1 #define ACPI_FADT_SEPARATE_LENGTH 2 @@ -87,7 +88,7 @@ static struct acpi_fadt_info fadt_info_table[] = { ACPI_FADT_OFFSET(pm1b_event_block), ACPI_FADT_OFFSET(pm1_event_length), ACPI_PM1_REGISTER_WIDTH * 2, /* Enable + Status register */ - 0}, + ACPI_FADT_OPTIONAL}, {"Pm1aControlBlock", ACPI_FADT_OFFSET(xpm1a_control_block), @@ -101,7 +102,7 @@ static struct acpi_fadt_info fadt_info_table[] = { ACPI_FADT_OFFSET(pm1b_control_block), ACPI_FADT_OFFSET(pm1_control_length), ACPI_PM1_REGISTER_WIDTH, - 0}, + ACPI_FADT_OPTIONAL}, {"Pm2ControlBlock", ACPI_FADT_OFFSET(xpm2_control_block), @@ -139,7 +140,7 @@ static struct acpi_fadt_info fadt_info_table[] = { typedef struct acpi_fadt_pm_info { struct acpi_generic_address *target; - u8 source; + u16 source; u8 register_num; } acpi_fadt_pm_info; From f654c0fefa8c16d439185b61442710fadc167e78 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Thu, 12 Jan 2012 13:10:32 +0800 Subject: [PATCH 449/820] ACPICA: Add support for region address conflict checking Allows drivers to determine if any memory or I/O addresses will conflict with addresses used by ACPI operation regions. Introduces a new interface, acpi_check_address_range. http://marc.info/?t=132251388700002&r=1&w=2 Reported-and-tested-by: Luca Tettamanti Signed-off-by: Lin Ming Signed-off-by: Bob Moore Signed-off-by: Len Brown --- drivers/acpi/acpica/Makefile | 1 + drivers/acpi/acpica/acconfig.h | 4 + drivers/acpi/acpica/acglobal.h | 2 + drivers/acpi/acpica/aclocal.h | 9 + drivers/acpi/acpica/acutils.h | 18 ++ drivers/acpi/acpica/dsargs.c | 30 ++-- drivers/acpi/acpica/utaddress.c | 294 ++++++++++++++++++++++++++++++++ drivers/acpi/acpica/utdelete.c | 13 +- drivers/acpi/acpica/utglobal.c | 6 + drivers/acpi/acpica/utinit.c | 1 + drivers/acpi/acpica/utxface.c | 38 +++++ drivers/acpi/osl.c | 202 ++-------------------- include/acpi/acpiosxf.h | 7 - include/acpi/acpixf.h | 5 + 14 files changed, 408 insertions(+), 222 deletions(-) create mode 100644 drivers/acpi/acpica/utaddress.c diff --git a/drivers/acpi/acpica/Makefile b/drivers/acpi/acpica/Makefile index 5f7d0cd11002..0ca208b6dcf0 100644 --- a/drivers/acpi/acpica/Makefile +++ b/drivers/acpi/acpica/Makefile @@ -135,6 +135,7 @@ acpi-y += \ tbxfroot.o acpi-y += \ + utaddress.o \ utalloc.o \ utcopy.o \ utdebug.o \ diff --git a/drivers/acpi/acpica/acconfig.h b/drivers/acpi/acpica/acconfig.h index f467d49c1ace..4925aa5817f4 100644 --- a/drivers/acpi/acpica/acconfig.h +++ b/drivers/acpi/acpica/acconfig.h @@ -123,6 +123,10 @@ #define ACPI_MAX_SLEEP 2000 /* Two seconds */ +/* Address Range lists are per-space_id (Memory and I/O only) */ + +#define ACPI_ADDRESS_RANGE_MAX 2 + /****************************************************************************** * * ACPI Specification constants (Do not change unless the specification changes) diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h index 25912a93dea7..dd5a3e341214 100644 --- a/drivers/acpi/acpica/acglobal.h +++ b/drivers/acpi/acpica/acglobal.h @@ -306,6 +306,8 @@ ACPI_EXTERN u8 acpi_gbl_acpi_hardware_present; ACPI_EXTERN u8 acpi_gbl_events_initialized; ACPI_EXTERN u8 acpi_gbl_osi_data; ACPI_EXTERN struct acpi_interface_info *acpi_gbl_supported_interfaces; +ACPI_EXTERN struct acpi_address_range + *acpi_gbl_address_range_list[ACPI_ADDRESS_RANGE_MAX]; #ifndef DEFINE_ACPI_GLOBALS diff --git a/drivers/acpi/acpica/aclocal.h b/drivers/acpi/acpica/aclocal.h index 2ddd61d07a06..d3b6f29cc41b 100644 --- a/drivers/acpi/acpica/aclocal.h +++ b/drivers/acpi/acpica/aclocal.h @@ -630,6 +630,15 @@ union acpi_generic_state { typedef acpi_status(*ACPI_EXECUTE_OP) (struct acpi_walk_state * walk_state); +/* Address Range info block */ + +struct acpi_address_range { + struct acpi_address_range *next; + struct acpi_namespace_node *region_node; + acpi_physical_address start_address; + acpi_physical_address end_address; +}; + /***************************************************************************** * * Parser typedefs and structs diff --git a/drivers/acpi/acpica/acutils.h b/drivers/acpi/acpica/acutils.h index ab95d452f87d..57df29360b9f 100644 --- a/drivers/acpi/acpica/acutils.h +++ b/drivers/acpi/acpica/acutils.h @@ -579,6 +579,24 @@ acpi_ut_create_list(char *list_name, #endif /* ACPI_DBG_TRACK_ALLOCATIONS */ +/* + * utaddress - address range check + */ +acpi_status +acpi_ut_add_address_range(acpi_adr_space_type space_id, + acpi_physical_address address, + u32 length, struct acpi_namespace_node *region_node); + +void +acpi_ut_remove_address_range(acpi_adr_space_type space_id, + struct acpi_namespace_node *region_node); + +u32 +acpi_ut_check_address_range(acpi_adr_space_type space_id, + acpi_physical_address address, u32 length, u8 warn); + +void acpi_ut_delete_address_lists(void); + /* * utxferror - various error/warning output functions */ diff --git a/drivers/acpi/acpica/dsargs.c b/drivers/acpi/acpica/dsargs.c index d69e4a53175b..19c2be941e78 100644 --- a/drivers/acpi/acpica/dsargs.c +++ b/drivers/acpi/acpica/dsargs.c @@ -250,6 +250,13 @@ acpi_ds_get_bank_field_arguments(union acpi_operand_object *obj_desc) status = acpi_ds_execute_arguments(node, node->parent, extra_desc->extra.aml_length, extra_desc->extra.aml_start); + if (ACPI_FAILURE(status)) { + return_ACPI_STATUS(status); + } + + status = acpi_ut_add_address_range(obj_desc->region.space_id, + obj_desc->region.address, + obj_desc->region.length, node); return_ACPI_STATUS(status); } @@ -391,25 +398,8 @@ acpi_status acpi_ds_get_region_arguments(union acpi_operand_object *obj_desc) return_ACPI_STATUS(status); } - /* Validate the region address/length via the host OS */ - - status = acpi_os_validate_address(obj_desc->region.space_id, - obj_desc->region.address, - (acpi_size) obj_desc->region.length, - acpi_ut_get_node_name(node)); - - if (ACPI_FAILURE(status)) { - /* - * Invalid address/length. We will emit an error message and mark - * the region as invalid, so that it will cause an additional error if - * it is ever used. Then return AE_OK. - */ - ACPI_EXCEPTION((AE_INFO, status, - "During address validation of OpRegion [%4.4s]", - node->name.ascii)); - obj_desc->common.flags |= AOPOBJ_INVALID; - status = AE_OK; - } - + status = acpi_ut_add_address_range(obj_desc->region.space_id, + obj_desc->region.address, + obj_desc->region.length, node); return_ACPI_STATUS(status); } diff --git a/drivers/acpi/acpica/utaddress.c b/drivers/acpi/acpica/utaddress.c new file mode 100644 index 000000000000..67932aebe6dd --- /dev/null +++ b/drivers/acpi/acpica/utaddress.c @@ -0,0 +1,294 @@ +/****************************************************************************** + * + * Module Name: utaddress - op_region address range check + * + *****************************************************************************/ + +/* + * Copyright (C) 2000 - 2012, Intel Corp. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. Redistributions in binary form must reproduce at minimum a disclaimer + * substantially similar to the "NO WARRANTY" disclaimer below + * ("Disclaimer") and any redistribution must be conditioned upon + * including a substantially similar Disclaimer requirement for further + * binary redistribution. + * 3. Neither the names of the above-listed copyright holders nor the names + * of any contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL") version 2 as published by the Free + * Software Foundation. + * + * NO WARRANTY + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * HOLDERS OR CONTRIBUTORS BE LIABLE FOR SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, + * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING + * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGES. + */ + +#include +#include "accommon.h" +#include "acnamesp.h" + +#define _COMPONENT ACPI_UTILITIES +ACPI_MODULE_NAME("utaddress") + +/******************************************************************************* + * + * FUNCTION: acpi_ut_add_address_range + * + * PARAMETERS: space_id - Address space ID + * Address - op_region start address + * Length - op_region length + * region_node - op_region namespace node + * + * RETURN: Status + * + * DESCRIPTION: Add the Operation Region address range to the global list. + * The only supported Space IDs are Memory and I/O. Called when + * the op_region address/length operands are fully evaluated. + * + * MUTEX: Locks the namespace + * + * NOTE: Because this interface is only called when an op_region argument + * list is evaluated, there cannot be any duplicate region_nodes. + * Duplicate Address/Length values are allowed, however, so that multiple + * address conflicts can be detected. + * + ******************************************************************************/ +acpi_status +acpi_ut_add_address_range(acpi_adr_space_type space_id, + acpi_physical_address address, + u32 length, struct acpi_namespace_node *region_node) +{ + struct acpi_address_range *range_info; + acpi_status status; + + ACPI_FUNCTION_TRACE(ut_add_address_range); + + if ((space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) && + (space_id != ACPI_ADR_SPACE_SYSTEM_IO)) { + return_ACPI_STATUS(AE_OK); + } + + /* Allocate/init a new info block, add it to the appropriate list */ + + range_info = ACPI_ALLOCATE(sizeof(struct acpi_address_range)); + if (!range_info) { + return_ACPI_STATUS(AE_NO_MEMORY); + } + + range_info->start_address = address; + range_info->end_address = (address + length - 1); + range_info->region_node = region_node; + + status = acpi_ut_acquire_mutex(ACPI_MTX_NAMESPACE); + if (ACPI_FAILURE(status)) { + ACPI_FREE(range_info); + return_ACPI_STATUS(status); + } + + range_info->next = acpi_gbl_address_range_list[space_id]; + acpi_gbl_address_range_list[space_id] = range_info; + + ACPI_DEBUG_PRINT((ACPI_DB_NAMES, + "\nAdded [%4.4s] address range: 0x%p-0x%p\n", + acpi_ut_get_node_name(range_info->region_node), + ACPI_CAST_PTR(void, address), + ACPI_CAST_PTR(void, range_info->end_address))); + + (void)acpi_ut_release_mutex(ACPI_MTX_NAMESPACE); + return_ACPI_STATUS(AE_OK); +} + +/******************************************************************************* + * + * FUNCTION: acpi_ut_remove_address_range + * + * PARAMETERS: space_id - Address space ID + * region_node - op_region namespace node + * + * RETURN: None + * + * DESCRIPTION: Remove the Operation Region from the global list. The only + * supported Space IDs are Memory and I/O. Called when an + * op_region is deleted. + * + * MUTEX: Assumes the namespace is locked + * + ******************************************************************************/ + +void +acpi_ut_remove_address_range(acpi_adr_space_type space_id, + struct acpi_namespace_node *region_node) +{ + struct acpi_address_range *range_info; + struct acpi_address_range *prev; + + ACPI_FUNCTION_TRACE(ut_remove_address_range); + + if ((space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) && + (space_id != ACPI_ADR_SPACE_SYSTEM_IO)) { + return_VOID; + } + + /* Get the appropriate list head and check the list */ + + range_info = prev = acpi_gbl_address_range_list[space_id]; + while (range_info) { + if (range_info->region_node == region_node) { + if (range_info == prev) { /* Found at list head */ + acpi_gbl_address_range_list[space_id] = + range_info->next; + } else { + prev->next = range_info->next; + } + + ACPI_DEBUG_PRINT((ACPI_DB_NAMES, + "\nRemoved [%4.4s] address range: 0x%p-0x%p\n", + acpi_ut_get_node_name(range_info-> + region_node), + ACPI_CAST_PTR(void, + range_info-> + start_address), + ACPI_CAST_PTR(void, + range_info-> + end_address))); + + ACPI_FREE(range_info); + return_VOID; + } + + prev = range_info; + range_info = range_info->next; + } + + return_VOID; +} + +/******************************************************************************* + * + * FUNCTION: acpi_ut_check_address_range + * + * PARAMETERS: space_id - Address space ID + * Address - Start address + * Length - Length of address range + * Warn - TRUE if warning on overlap desired + * + * RETURN: Count of the number of conflicts detected. Zero is always + * returned for Space IDs other than Memory or I/O. + * + * DESCRIPTION: Check if the input address range overlaps any of the + * ASL operation region address ranges. The only supported + * Space IDs are Memory and I/O. + * + * MUTEX: Assumes the namespace is locked. + * + ******************************************************************************/ + +u32 +acpi_ut_check_address_range(acpi_adr_space_type space_id, + acpi_physical_address address, u32 length, u8 warn) +{ + struct acpi_address_range *range_info; + acpi_physical_address end_address; + char *pathname; + u32 overlap_count = 0; + + ACPI_FUNCTION_TRACE(ut_check_address_range); + + if ((space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) && + (space_id != ACPI_ADR_SPACE_SYSTEM_IO)) { + return_UINT32(0); + } + + range_info = acpi_gbl_address_range_list[space_id]; + end_address = address + length - 1; + + /* Check entire list for all possible conflicts */ + + while (range_info) { + /* + * Check if the requested Address/Length overlaps this address_range. + * Four cases to consider: + * + * 1) Input address/length is contained completely in the address range + * 2) Input address/length overlaps range at the range start + * 3) Input address/length overlaps range at the range end + * 4) Input address/length completely encompasses the range + */ + if ((address <= range_info->end_address) && + (end_address >= range_info->start_address)) { + + /* Found an address range overlap */ + + overlap_count++; + if (warn) { /* Optional warning message */ + pathname = + acpi_ns_get_external_pathname(range_info-> + region_node); + + ACPI_WARNING((AE_INFO, + "0x%p-0x%p %s conflicts with Region %s %d", + ACPI_CAST_PTR(void, address), + ACPI_CAST_PTR(void, end_address), + acpi_ut_get_region_name(space_id), + pathname, overlap_count)); + ACPI_FREE(pathname); + } + } + + range_info = range_info->next; + } + + return_UINT32(overlap_count); +} + +/******************************************************************************* + * + * FUNCTION: acpi_ut_delete_address_lists + * + * PARAMETERS: None + * + * RETURN: None + * + * DESCRIPTION: Delete all global address range lists (called during + * subsystem shutdown). + * + ******************************************************************************/ + +void acpi_ut_delete_address_lists(void) +{ + struct acpi_address_range *next; + struct acpi_address_range *range_info; + int i; + + /* Delete all elements in all address range lists */ + + for (i = 0; i < ACPI_ADDRESS_RANGE_MAX; i++) { + next = acpi_gbl_address_range_list[i]; + + while (next) { + range_info = next; + next = range_info->next; + ACPI_FREE(range_info); + } + + acpi_gbl_address_range_list[i] = NULL; + } +} diff --git a/drivers/acpi/acpica/utdelete.c b/drivers/acpi/acpica/utdelete.c index 31f5a7832ef1..93ec06b0fe25 100644 --- a/drivers/acpi/acpica/utdelete.c +++ b/drivers/acpi/acpica/utdelete.c @@ -215,11 +215,14 @@ static void acpi_ut_delete_internal_obj(union acpi_operand_object *object) ACPI_DEBUG_PRINT((ACPI_DB_ALLOCATIONS, "***** Region %p\n", object)); - /* Invalidate the region address/length via the host OS */ - - acpi_os_invalidate_address(object->region.space_id, - object->region.address, - (acpi_size) object->region.length); + /* + * Update address_range list. However, only permanent regions + * are installed in this list. (Not created within a method) + */ + if (!(object->region.node->flags & ANOBJ_TEMPORARY)) { + acpi_ut_remove_address_range(object->region.space_id, + object->region.node); + } second_desc = acpi_ns_get_secondary_object(object); if (second_desc) { diff --git a/drivers/acpi/acpica/utglobal.c b/drivers/acpi/acpica/utglobal.c index ffba0a39c3e8..3cf758bab242 100644 --- a/drivers/acpi/acpica/utglobal.c +++ b/drivers/acpi/acpica/utglobal.c @@ -264,6 +264,12 @@ acpi_status acpi_ut_init_globals(void) return_ACPI_STATUS(status); } + /* Address Range lists */ + + for (i = 0; i < ACPI_ADDRESS_RANGE_MAX; i++) { + acpi_gbl_address_range_list[i] = NULL; + } + /* Mutex locked flags */ for (i = 0; i < ACPI_NUM_MUTEX; i++) { diff --git a/drivers/acpi/acpica/utinit.c b/drivers/acpi/acpica/utinit.c index 191b6828cce9..cab61a05d21d 100644 --- a/drivers/acpi/acpica/utinit.c +++ b/drivers/acpi/acpica/utinit.c @@ -92,6 +92,7 @@ static void acpi_ut_terminate(void) gpe_xrupt_info = next_gpe_xrupt_info; } + acpi_ut_delete_address_lists(); return_VOID; } diff --git a/drivers/acpi/acpica/utxface.c b/drivers/acpi/acpica/utxface.c index 420ebfe08c72..15cbddebb6b4 100644 --- a/drivers/acpi/acpica/utxface.c +++ b/drivers/acpi/acpica/utxface.c @@ -48,6 +48,7 @@ #include "acnamesp.h" #include "acdebug.h" #include "actables.h" +#include "acinterp.h" #define _COMPONENT ACPI_UTILITIES ACPI_MODULE_NAME("utxface") @@ -640,4 +641,41 @@ acpi_status acpi_install_interface_handler(acpi_interface_handler handler) } ACPI_EXPORT_SYMBOL(acpi_install_interface_handler) + +/***************************************************************************** + * + * FUNCTION: acpi_check_address_range + * + * PARAMETERS: space_id - Address space ID + * Address - Start address + * Length - Length + * Warn - TRUE if warning on overlap desired + * + * RETURN: Count of the number of conflicts detected. + * + * DESCRIPTION: Check if the input address range overlaps any of the + * ASL operation region address ranges. + * + ****************************************************************************/ +u32 +acpi_check_address_range(acpi_adr_space_type space_id, + acpi_physical_address address, + acpi_size length, u8 warn) +{ + u32 overlaps; + acpi_status status; + + status = acpi_ut_acquire_mutex(ACPI_MTX_NAMESPACE); + if (ACPI_FAILURE(status)) { + return (0); + } + + overlaps = acpi_ut_check_address_range(space_id, address, + (u32)length, warn); + + (void)acpi_ut_release_mutex(ACPI_MTX_NAMESPACE); + return (overlaps); +} + +ACPI_EXPORT_SYMBOL(acpi_check_address_range) #endif /* !ACPI_ASL_COMPILER */ diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index f31c5c5f1b7e..3e57fbdf50a3 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -83,19 +83,6 @@ static struct workqueue_struct *kacpi_notify_wq; struct workqueue_struct *kacpi_hotplug_wq; EXPORT_SYMBOL(kacpi_hotplug_wq); -struct acpi_res_list { - resource_size_t start; - resource_size_t end; - acpi_adr_space_type resource_type; /* IO port, System memory, ...*/ - char name[5]; /* only can have a length of 4 chars, make use of this - one instead of res->name, no need to kalloc then */ - struct list_head resource_list; - int count; -}; - -static LIST_HEAD(resource_list_head); -static DEFINE_SPINLOCK(acpi_res_lock); - /* * This list of permanent mappings is for memory that may be accessed from * interrupt context, where we can't do the ioremap(). @@ -1278,44 +1265,28 @@ __setup("acpi_enforce_resources=", acpi_enforce_resources_setup); * drivers */ int acpi_check_resource_conflict(const struct resource *res) { - struct acpi_res_list *res_list_elem; - int ioport = 0, clash = 0; + acpi_adr_space_type space_id; + acpi_size length; + u8 warn = 0; + int clash = 0; if (acpi_enforce_resources == ENFORCE_RESOURCES_NO) return 0; if (!(res->flags & IORESOURCE_IO) && !(res->flags & IORESOURCE_MEM)) return 0; - ioport = res->flags & IORESOURCE_IO; + if (res->flags & IORESOURCE_IO) + space_id = ACPI_ADR_SPACE_SYSTEM_IO; + else + space_id = ACPI_ADR_SPACE_SYSTEM_MEMORY; - spin_lock(&acpi_res_lock); - list_for_each_entry(res_list_elem, &resource_list_head, - resource_list) { - if (ioport && (res_list_elem->resource_type - != ACPI_ADR_SPACE_SYSTEM_IO)) - continue; - if (!ioport && (res_list_elem->resource_type - != ACPI_ADR_SPACE_SYSTEM_MEMORY)) - continue; - - if (res->end < res_list_elem->start - || res_list_elem->end < res->start) - continue; - clash = 1; - break; - } - spin_unlock(&acpi_res_lock); + length = res->end - res->start + 1; + if (acpi_enforce_resources != ENFORCE_RESOURCES_NO) + warn = 1; + clash = acpi_check_address_range(space_id, res->start, length, warn); if (clash) { if (acpi_enforce_resources != ENFORCE_RESOURCES_NO) { - printk(KERN_WARNING "ACPI: resource %s %pR" - " conflicts with ACPI region %s " - "[%s 0x%zx-0x%zx]\n", - res->name, res, res_list_elem->name, - (res_list_elem->resource_type == - ACPI_ADR_SPACE_SYSTEM_IO) ? "io" : "mem", - (size_t) res_list_elem->start, - (size_t) res_list_elem->end); if (acpi_enforce_resources == ENFORCE_RESOURCES_LAX) printk(KERN_NOTICE "ACPI: This conflict may" " cause random problems and system" @@ -1467,155 +1438,6 @@ acpi_status acpi_os_release_object(acpi_cache_t * cache, void *object) kmem_cache_free(cache, object); return (AE_OK); } - -static inline int acpi_res_list_add(struct acpi_res_list *res) -{ - struct acpi_res_list *res_list_elem; - - list_for_each_entry(res_list_elem, &resource_list_head, - resource_list) { - - if (res->resource_type == res_list_elem->resource_type && - res->start == res_list_elem->start && - res->end == res_list_elem->end) { - - /* - * The Region(addr,len) already exist in the list, - * just increase the count - */ - - res_list_elem->count++; - return 0; - } - } - - res->count = 1; - list_add(&res->resource_list, &resource_list_head); - return 1; -} - -static inline void acpi_res_list_del(struct acpi_res_list *res) -{ - struct acpi_res_list *res_list_elem; - - list_for_each_entry(res_list_elem, &resource_list_head, - resource_list) { - - if (res->resource_type == res_list_elem->resource_type && - res->start == res_list_elem->start && - res->end == res_list_elem->end) { - - /* - * If the res count is decreased to 0, - * remove and free it - */ - - if (--res_list_elem->count == 0) { - list_del(&res_list_elem->resource_list); - kfree(res_list_elem); - } - return; - } - } -} - -acpi_status -acpi_os_invalidate_address( - u8 space_id, - acpi_physical_address address, - acpi_size length) -{ - struct acpi_res_list res; - - switch (space_id) { - case ACPI_ADR_SPACE_SYSTEM_IO: - case ACPI_ADR_SPACE_SYSTEM_MEMORY: - /* Only interference checks against SystemIO and SystemMemory - are needed */ - res.start = address; - res.end = address + length - 1; - res.resource_type = space_id; - spin_lock(&acpi_res_lock); - acpi_res_list_del(&res); - spin_unlock(&acpi_res_lock); - break; - case ACPI_ADR_SPACE_PCI_CONFIG: - case ACPI_ADR_SPACE_EC: - case ACPI_ADR_SPACE_SMBUS: - case ACPI_ADR_SPACE_CMOS: - case ACPI_ADR_SPACE_PCI_BAR_TARGET: - case ACPI_ADR_SPACE_DATA_TABLE: - case ACPI_ADR_SPACE_FIXED_HARDWARE: - break; - } - return AE_OK; -} - -/****************************************************************************** - * - * FUNCTION: acpi_os_validate_address - * - * PARAMETERS: space_id - ACPI space ID - * address - Physical address - * length - Address length - * - * RETURN: AE_OK if address/length is valid for the space_id. Otherwise, - * should return AE_AML_ILLEGAL_ADDRESS. - * - * DESCRIPTION: Validate a system address via the host OS. Used to validate - * the addresses accessed by AML operation regions. - * - *****************************************************************************/ - -acpi_status -acpi_os_validate_address ( - u8 space_id, - acpi_physical_address address, - acpi_size length, - char *name) -{ - struct acpi_res_list *res; - int added; - if (acpi_enforce_resources == ENFORCE_RESOURCES_NO) - return AE_OK; - - switch (space_id) { - case ACPI_ADR_SPACE_SYSTEM_IO: - case ACPI_ADR_SPACE_SYSTEM_MEMORY: - /* Only interference checks against SystemIO and SystemMemory - are needed */ - res = kzalloc(sizeof(struct acpi_res_list), GFP_KERNEL); - if (!res) - return AE_OK; - /* ACPI names are fixed to 4 bytes, still better use strlcpy */ - strlcpy(res->name, name, 5); - res->start = address; - res->end = address + length - 1; - res->resource_type = space_id; - spin_lock(&acpi_res_lock); - added = acpi_res_list_add(res); - spin_unlock(&acpi_res_lock); - pr_debug("%s %s resource: start: 0x%llx, end: 0x%llx, " - "name: %s\n", added ? "Added" : "Already exist", - (space_id == ACPI_ADR_SPACE_SYSTEM_IO) - ? "SystemIO" : "System Memory", - (unsigned long long)res->start, - (unsigned long long)res->end, - res->name); - if (!added) - kfree(res); - break; - case ACPI_ADR_SPACE_PCI_CONFIG: - case ACPI_ADR_SPACE_EC: - case ACPI_ADR_SPACE_SMBUS: - case ACPI_ADR_SPACE_CMOS: - case ACPI_ADR_SPACE_PCI_BAR_TARGET: - case ACPI_ADR_SPACE_DATA_TABLE: - case ACPI_ADR_SPACE_FIXED_HARDWARE: - break; - } - return AE_OK; -} #endif acpi_status __init acpi_os_initialize(void) diff --git a/include/acpi/acpiosxf.h b/include/acpi/acpiosxf.h index 83062ed0ef2f..2fe8639b3ae7 100644 --- a/include/acpi/acpiosxf.h +++ b/include/acpi/acpiosxf.h @@ -238,13 +238,6 @@ acpi_os_write_pci_configuration(struct acpi_pci_id *pci_id, /* * Miscellaneous */ -acpi_status -acpi_os_validate_address(u8 space_id, acpi_physical_address address, - acpi_size length, char *name); -acpi_status -acpi_os_invalidate_address(u8 space_id, acpi_physical_address address, - acpi_size length); - u64 acpi_os_get_timer(void); acpi_status acpi_os_signal(u32 function, void *info); diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index 30f5bef4f14d..a4cfded67c08 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -112,6 +112,11 @@ acpi_status acpi_install_interface(acpi_string interface_name); acpi_status acpi_remove_interface(acpi_string interface_name); +u32 +acpi_check_address_range(acpi_adr_space_type space_id, + acpi_physical_address address, + acpi_size length, u8 warn); + /* * ACPI Memory management */ From 77848130e53b06c22fe37a7b6acbb82bb3e9bfba Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Thu, 12 Jan 2012 13:27:23 +0800 Subject: [PATCH 450/820] ACPICA: Update all copyrights to 2012 Update all copyrights to 2012. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/acpica/accommon.h | 2 +- drivers/acpi/acpica/acconfig.h | 2 +- drivers/acpi/acpica/acdebug.h | 2 +- drivers/acpi/acpica/acdispat.h | 2 +- drivers/acpi/acpica/acevents.h | 2 +- drivers/acpi/acpica/acglobal.h | 2 +- drivers/acpi/acpica/achware.h | 2 +- drivers/acpi/acpica/acinterp.h | 2 +- drivers/acpi/acpica/aclocal.h | 2 +- drivers/acpi/acpica/acmacros.h | 2 +- drivers/acpi/acpica/acnamesp.h | 2 +- drivers/acpi/acpica/acobject.h | 2 +- drivers/acpi/acpica/acopcode.h | 2 +- drivers/acpi/acpica/acparser.h | 2 +- drivers/acpi/acpica/acpredef.h | 2 +- drivers/acpi/acpica/acresrc.h | 2 +- drivers/acpi/acpica/acstruct.h | 2 +- drivers/acpi/acpica/actables.h | 2 +- drivers/acpi/acpica/acutils.h | 2 +- drivers/acpi/acpica/amlcode.h | 2 +- drivers/acpi/acpica/amlresrc.h | 2 +- drivers/acpi/acpica/dsargs.c | 2 +- drivers/acpi/acpica/dscontrol.c | 2 +- drivers/acpi/acpica/dsfield.c | 2 +- drivers/acpi/acpica/dsinit.c | 2 +- drivers/acpi/acpica/dsmethod.c | 2 +- drivers/acpi/acpica/dsmthdat.c | 2 +- drivers/acpi/acpica/dsobject.c | 2 +- drivers/acpi/acpica/dsopcode.c | 2 +- drivers/acpi/acpica/dsutils.c | 2 +- drivers/acpi/acpica/dswexec.c | 2 +- drivers/acpi/acpica/dswload.c | 2 +- drivers/acpi/acpica/dswload2.c | 2 +- drivers/acpi/acpica/dswscope.c | 2 +- drivers/acpi/acpica/dswstate.c | 2 +- drivers/acpi/acpica/evevent.c | 2 +- drivers/acpi/acpica/evglock.c | 2 +- drivers/acpi/acpica/evgpe.c | 2 +- drivers/acpi/acpica/evgpeblk.c | 2 +- drivers/acpi/acpica/evgpeinit.c | 2 +- drivers/acpi/acpica/evgpeutil.c | 2 +- drivers/acpi/acpica/evmisc.c | 2 +- drivers/acpi/acpica/evregion.c | 2 +- drivers/acpi/acpica/evrgnini.c | 2 +- drivers/acpi/acpica/evsci.c | 2 +- drivers/acpi/acpica/evxface.c | 2 +- drivers/acpi/acpica/evxfevnt.c | 2 +- drivers/acpi/acpica/evxfgpe.c | 2 +- drivers/acpi/acpica/evxfregn.c | 2 +- drivers/acpi/acpica/exconfig.c | 2 +- drivers/acpi/acpica/exconvrt.c | 2 +- drivers/acpi/acpica/excreate.c | 2 +- drivers/acpi/acpica/exdebug.c | 2 +- drivers/acpi/acpica/exdump.c | 2 +- drivers/acpi/acpica/exfield.c | 2 +- drivers/acpi/acpica/exfldio.c | 2 +- drivers/acpi/acpica/exmisc.c | 2 +- drivers/acpi/acpica/exmutex.c | 2 +- drivers/acpi/acpica/exnames.c | 2 +- drivers/acpi/acpica/exoparg1.c | 2 +- drivers/acpi/acpica/exoparg2.c | 2 +- drivers/acpi/acpica/exoparg3.c | 2 +- drivers/acpi/acpica/exoparg6.c | 2 +- drivers/acpi/acpica/exprep.c | 2 +- drivers/acpi/acpica/exregion.c | 2 +- drivers/acpi/acpica/exresnte.c | 2 +- drivers/acpi/acpica/exresolv.c | 2 +- drivers/acpi/acpica/exresop.c | 2 +- drivers/acpi/acpica/exstore.c | 2 +- drivers/acpi/acpica/exstoren.c | 2 +- drivers/acpi/acpica/exstorob.c | 2 +- drivers/acpi/acpica/exsystem.c | 2 +- drivers/acpi/acpica/exutils.c | 2 +- drivers/acpi/acpica/hwacpi.c | 2 +- drivers/acpi/acpica/hwgpe.c | 2 +- drivers/acpi/acpica/hwpci.c | 2 +- drivers/acpi/acpica/hwregs.c | 2 +- drivers/acpi/acpica/hwsleep.c | 2 +- drivers/acpi/acpica/hwtimer.c | 2 +- drivers/acpi/acpica/hwvalid.c | 2 +- drivers/acpi/acpica/hwxface.c | 2 +- drivers/acpi/acpica/nsaccess.c | 2 +- drivers/acpi/acpica/nsalloc.c | 2 +- drivers/acpi/acpica/nsdump.c | 2 +- drivers/acpi/acpica/nsdumpdv.c | 2 +- drivers/acpi/acpica/nseval.c | 2 +- drivers/acpi/acpica/nsinit.c | 2 +- drivers/acpi/acpica/nsload.c | 2 +- drivers/acpi/acpica/nsnames.c | 2 +- drivers/acpi/acpica/nsobject.c | 2 +- drivers/acpi/acpica/nsparse.c | 2 +- drivers/acpi/acpica/nspredef.c | 2 +- drivers/acpi/acpica/nsrepair.c | 2 +- drivers/acpi/acpica/nsrepair2.c | 2 +- drivers/acpi/acpica/nssearch.c | 2 +- drivers/acpi/acpica/nsutils.c | 2 +- drivers/acpi/acpica/nswalk.c | 2 +- drivers/acpi/acpica/nsxfeval.c | 2 +- drivers/acpi/acpica/nsxfname.c | 2 +- drivers/acpi/acpica/nsxfobj.c | 2 +- drivers/acpi/acpica/psargs.c | 2 +- drivers/acpi/acpica/psloop.c | 2 +- drivers/acpi/acpica/psopcode.c | 2 +- drivers/acpi/acpica/psparse.c | 2 +- drivers/acpi/acpica/psscope.c | 2 +- drivers/acpi/acpica/pstree.c | 2 +- drivers/acpi/acpica/psutils.c | 2 +- drivers/acpi/acpica/pswalk.c | 2 +- drivers/acpi/acpica/psxface.c | 2 +- drivers/acpi/acpica/rsaddr.c | 2 +- drivers/acpi/acpica/rscalc.c | 2 +- drivers/acpi/acpica/rscreate.c | 2 +- drivers/acpi/acpica/rsdump.c | 2 +- drivers/acpi/acpica/rsinfo.c | 2 +- drivers/acpi/acpica/rsio.c | 2 +- drivers/acpi/acpica/rsirq.c | 2 +- drivers/acpi/acpica/rslist.c | 2 +- drivers/acpi/acpica/rsmemory.c | 2 +- drivers/acpi/acpica/rsmisc.c | 2 +- drivers/acpi/acpica/rsserial.c | 2 +- drivers/acpi/acpica/rsutils.c | 2 +- drivers/acpi/acpica/rsxface.c | 2 +- drivers/acpi/acpica/tbfadt.c | 2 +- drivers/acpi/acpica/tbfind.c | 2 +- drivers/acpi/acpica/tbinstal.c | 2 +- drivers/acpi/acpica/tbutils.c | 2 +- drivers/acpi/acpica/tbxface.c | 2 +- drivers/acpi/acpica/tbxfroot.c | 2 +- drivers/acpi/acpica/utalloc.c | 2 +- drivers/acpi/acpica/utcopy.c | 2 +- drivers/acpi/acpica/utdebug.c | 2 +- drivers/acpi/acpica/utdecode.c | 2 +- drivers/acpi/acpica/utdelete.c | 2 +- drivers/acpi/acpica/uteval.c | 2 +- drivers/acpi/acpica/utglobal.c | 2 +- drivers/acpi/acpica/utids.c | 2 +- drivers/acpi/acpica/utinit.c | 2 +- drivers/acpi/acpica/utlock.c | 2 +- drivers/acpi/acpica/utmath.c | 2 +- drivers/acpi/acpica/utmisc.c | 2 +- drivers/acpi/acpica/utmutex.c | 2 +- drivers/acpi/acpica/utobject.c | 2 +- drivers/acpi/acpica/utosi.c | 2 +- drivers/acpi/acpica/utresrc.c | 2 +- drivers/acpi/acpica/utstate.c | 2 +- drivers/acpi/acpica/utxface.c | 2 +- drivers/acpi/acpica/utxferror.c | 2 +- drivers/acpi/acpica/utxfmutex.c | 2 +- 148 files changed, 148 insertions(+), 148 deletions(-) diff --git a/drivers/acpi/acpica/accommon.h b/drivers/acpi/acpica/accommon.h index e0ba17f0a7c8..a44bd424f9f4 100644 --- a/drivers/acpi/acpica/accommon.h +++ b/drivers/acpi/acpica/accommon.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acconfig.h b/drivers/acpi/acpica/acconfig.h index 4925aa5817f4..1f30af613e87 100644 --- a/drivers/acpi/acpica/acconfig.h +++ b/drivers/acpi/acpica/acconfig.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acdebug.h b/drivers/acpi/acpica/acdebug.h index eb0b1f8dee6d..deaa81979561 100644 --- a/drivers/acpi/acpica/acdebug.h +++ b/drivers/acpi/acpica/acdebug.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acdispat.h b/drivers/acpi/acpica/acdispat.h index 2d1b7ffa377a..5935ba6707e2 100644 --- a/drivers/acpi/acpica/acdispat.h +++ b/drivers/acpi/acpica/acdispat.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acevents.h b/drivers/acpi/acpica/acevents.h index f763b88ed772..c53caa521a30 100644 --- a/drivers/acpi/acpica/acevents.h +++ b/drivers/acpi/acpica/acevents.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acglobal.h b/drivers/acpi/acpica/acglobal.h index dd5a3e341214..261fd5eb2b68 100644 --- a/drivers/acpi/acpica/acglobal.h +++ b/drivers/acpi/acpica/acglobal.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/achware.h b/drivers/acpi/acpica/achware.h index e7213beaafc7..677793e938f5 100644 --- a/drivers/acpi/acpica/achware.h +++ b/drivers/acpi/acpica/achware.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acinterp.h b/drivers/acpi/acpica/acinterp.h index 4ee9058aba53..eb308635da72 100644 --- a/drivers/acpi/acpica/acinterp.h +++ b/drivers/acpi/acpica/acinterp.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/aclocal.h b/drivers/acpi/acpica/aclocal.h index d3b6f29cc41b..3f24068837d5 100644 --- a/drivers/acpi/acpica/aclocal.h +++ b/drivers/acpi/acpica/aclocal.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acmacros.h b/drivers/acpi/acpica/acmacros.h index b7491ee1fba6..ef338a96f5b2 100644 --- a/drivers/acpi/acpica/acmacros.h +++ b/drivers/acpi/acpica/acmacros.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acnamesp.h b/drivers/acpi/acpica/acnamesp.h index 79a598c67fe3..2c9e0f049523 100644 --- a/drivers/acpi/acpica/acnamesp.h +++ b/drivers/acpi/acpica/acnamesp.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acobject.h b/drivers/acpi/acpica/acobject.h index 435dd2f847bb..c065078ca83b 100644 --- a/drivers/acpi/acpica/acobject.h +++ b/drivers/acpi/acpica/acobject.h @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acopcode.h b/drivers/acpi/acpica/acopcode.h index 9896c77e94f9..9440d053fbb3 100644 --- a/drivers/acpi/acpica/acopcode.h +++ b/drivers/acpi/acpica/acopcode.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acparser.h b/drivers/acpi/acpica/acparser.h index 5ea1e06afa20..b725d780d34d 100644 --- a/drivers/acpi/acpica/acparser.h +++ b/drivers/acpi/acpica/acparser.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acpredef.h b/drivers/acpi/acpica/acpredef.h index ecac9cf38769..bbb34c9be4e8 100644 --- a/drivers/acpi/acpica/acpredef.h +++ b/drivers/acpi/acpica/acpredef.h @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acresrc.h b/drivers/acpi/acpica/acresrc.h index 61b2e049d8d6..0347d0993497 100644 --- a/drivers/acpi/acpica/acresrc.h +++ b/drivers/acpi/acpica/acresrc.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acstruct.h b/drivers/acpi/acpica/acstruct.h index 1623b245dde2..0404df605bc1 100644 --- a/drivers/acpi/acpica/acstruct.h +++ b/drivers/acpi/acpica/acstruct.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/actables.h b/drivers/acpi/acpica/actables.h index 967f08124eba..d5bec304c823 100644 --- a/drivers/acpi/acpica/actables.h +++ b/drivers/acpi/acpica/actables.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/acutils.h b/drivers/acpi/acpica/acutils.h index 57df29360b9f..925ccf22101b 100644 --- a/drivers/acpi/acpica/acutils.h +++ b/drivers/acpi/acpica/acutils.h @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/amlcode.h b/drivers/acpi/acpica/amlcode.h index c7db80ee61ae..905280fec0fa 100644 --- a/drivers/acpi/acpica/amlcode.h +++ b/drivers/acpi/acpica/amlcode.h @@ -7,7 +7,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/amlresrc.h b/drivers/acpi/acpica/amlresrc.h index d4c30511aca9..7b2128f274e7 100644 --- a/drivers/acpi/acpica/amlresrc.h +++ b/drivers/acpi/acpica/amlresrc.h @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsargs.c b/drivers/acpi/acpica/dsargs.c index 19c2be941e78..80eb1900297f 100644 --- a/drivers/acpi/acpica/dsargs.c +++ b/drivers/acpi/acpica/dsargs.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dscontrol.c b/drivers/acpi/acpica/dscontrol.c index 26c49fff58da..effe4ca1133f 100644 --- a/drivers/acpi/acpica/dscontrol.c +++ b/drivers/acpi/acpica/dscontrol.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsfield.c b/drivers/acpi/acpica/dsfield.c index aa880d992984..cd243cf2cab2 100644 --- a/drivers/acpi/acpica/dsfield.c +++ b/drivers/acpi/acpica/dsfield.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsinit.c b/drivers/acpi/acpica/dsinit.c index a7718bf2b9a1..9e5ac7f780a7 100644 --- a/drivers/acpi/acpica/dsinit.c +++ b/drivers/acpi/acpica/dsinit.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsmethod.c b/drivers/acpi/acpica/dsmethod.c index 5d797751e205..00f5dab5bcc0 100644 --- a/drivers/acpi/acpica/dsmethod.c +++ b/drivers/acpi/acpica/dsmethod.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsmthdat.c b/drivers/acpi/acpica/dsmthdat.c index 905ce29a92e1..b40bd507be5d 100644 --- a/drivers/acpi/acpica/dsmthdat.c +++ b/drivers/acpi/acpica/dsmthdat.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsobject.c b/drivers/acpi/acpica/dsobject.c index f42e17e5c252..d7045ca3e32a 100644 --- a/drivers/acpi/acpica/dsobject.c +++ b/drivers/acpi/acpica/dsobject.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsopcode.c b/drivers/acpi/acpica/dsopcode.c index c627a288e027..e5eff7585102 100644 --- a/drivers/acpi/acpica/dsopcode.c +++ b/drivers/acpi/acpica/dsopcode.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dsutils.c b/drivers/acpi/acpica/dsutils.c index 2c477ce172fa..1abcda31037f 100644 --- a/drivers/acpi/acpica/dsutils.c +++ b/drivers/acpi/acpica/dsutils.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dswexec.c b/drivers/acpi/acpica/dswexec.c index fe40e4c6554f..642f3c053e87 100644 --- a/drivers/acpi/acpica/dswexec.c +++ b/drivers/acpi/acpica/dswexec.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dswload.c b/drivers/acpi/acpica/dswload.c index 324acec1179a..552aa3a50c84 100644 --- a/drivers/acpi/acpica/dswload.c +++ b/drivers/acpi/acpica/dswload.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dswload2.c b/drivers/acpi/acpica/dswload2.c index 976318138c56..ae7147724763 100644 --- a/drivers/acpi/acpica/dswload2.c +++ b/drivers/acpi/acpica/dswload2.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dswscope.c b/drivers/acpi/acpica/dswscope.c index 76a661fc1e09..9e9490a9cbf0 100644 --- a/drivers/acpi/acpica/dswscope.c +++ b/drivers/acpi/acpica/dswscope.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/dswstate.c b/drivers/acpi/acpica/dswstate.c index a6c374ef9914..c9c2ac13e7cc 100644 --- a/drivers/acpi/acpica/dswstate.c +++ b/drivers/acpi/acpica/dswstate.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evevent.c b/drivers/acpi/acpica/evevent.c index 1f43be470199..6729ebe2f1e6 100644 --- a/drivers/acpi/acpica/evevent.c +++ b/drivers/acpi/acpica/evevent.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evglock.c b/drivers/acpi/acpica/evglock.c index af1e71af1b64..5e5683cb1f0d 100644 --- a/drivers/acpi/acpica/evglock.c +++ b/drivers/acpi/acpica/evglock.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evgpe.c b/drivers/acpi/acpica/evgpe.c index 65c79add3b19..9e88cb6fb25e 100644 --- a/drivers/acpi/acpica/evgpe.c +++ b/drivers/acpi/acpica/evgpe.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evgpeblk.c b/drivers/acpi/acpica/evgpeblk.c index ca2c41a53311..be75339cd5dd 100644 --- a/drivers/acpi/acpica/evgpeblk.c +++ b/drivers/acpi/acpica/evgpeblk.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evgpeinit.c b/drivers/acpi/acpica/evgpeinit.c index ce9aa9f9a972..adf7494da9db 100644 --- a/drivers/acpi/acpica/evgpeinit.c +++ b/drivers/acpi/acpica/evgpeinit.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evgpeutil.c b/drivers/acpi/acpica/evgpeutil.c index 80a81d0c4a80..25073932aa10 100644 --- a/drivers/acpi/acpica/evgpeutil.c +++ b/drivers/acpi/acpica/evgpeutil.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evmisc.c b/drivers/acpi/acpica/evmisc.c index d0b331844427..84966f416463 100644 --- a/drivers/acpi/acpica/evmisc.c +++ b/drivers/acpi/acpica/evmisc.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evregion.c b/drivers/acpi/acpica/evregion.c index dcdc7e4a285e..1b0180a1b798 100644 --- a/drivers/acpi/acpica/evregion.c +++ b/drivers/acpi/acpica/evregion.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evrgnini.c b/drivers/acpi/acpica/evrgnini.c index 55a5d35ef34a..819c17f5897a 100644 --- a/drivers/acpi/acpica/evrgnini.c +++ b/drivers/acpi/acpica/evrgnini.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evsci.c b/drivers/acpi/acpica/evsci.c index 2ebd40e1a3ef..26065c612e76 100644 --- a/drivers/acpi/acpica/evsci.c +++ b/drivers/acpi/acpica/evsci.c @@ -6,7 +6,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evxface.c b/drivers/acpi/acpica/evxface.c index f4f523bf5939..61944e89565a 100644 --- a/drivers/acpi/acpica/evxface.c +++ b/drivers/acpi/acpica/evxface.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evxfevnt.c b/drivers/acpi/acpica/evxfevnt.c index 20516e599476..1768bbec1002 100644 --- a/drivers/acpi/acpica/evxfevnt.c +++ b/drivers/acpi/acpica/evxfevnt.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evxfgpe.c b/drivers/acpi/acpica/evxfgpe.c index f06a3ee356ba..33388fd69df4 100644 --- a/drivers/acpi/acpica/evxfgpe.c +++ b/drivers/acpi/acpica/evxfgpe.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/evxfregn.c b/drivers/acpi/acpica/evxfregn.c index aee887e3ca5c..6019208cd4b6 100644 --- a/drivers/acpi/acpica/evxfregn.c +++ b/drivers/acpi/acpica/evxfregn.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exconfig.c b/drivers/acpi/acpica/exconfig.c index efe41f736fab..c86d44e41bc8 100644 --- a/drivers/acpi/acpica/exconfig.c +++ b/drivers/acpi/acpica/exconfig.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exconvrt.c b/drivers/acpi/acpica/exconvrt.c index 74162a11817d..e385436bd424 100644 --- a/drivers/acpi/acpica/exconvrt.c +++ b/drivers/acpi/acpica/exconvrt.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/excreate.c b/drivers/acpi/acpica/excreate.c index c66562b4311b..3f5bc998c1cb 100644 --- a/drivers/acpi/acpica/excreate.c +++ b/drivers/acpi/acpica/excreate.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exdebug.c b/drivers/acpi/acpica/exdebug.c index c7a2f1edd282..e211e9c19215 100644 --- a/drivers/acpi/acpica/exdebug.c +++ b/drivers/acpi/acpica/exdebug.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exdump.c b/drivers/acpi/acpica/exdump.c index c7064f4d4704..2a6ac0a3bc1e 100644 --- a/drivers/acpi/acpica/exdump.c +++ b/drivers/acpi/acpica/exdump.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exfield.c b/drivers/acpi/acpica/exfield.c index 3aeec028d71c..dc092f5b35d6 100644 --- a/drivers/acpi/acpica/exfield.c +++ b/drivers/acpi/acpica/exfield.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exfldio.c b/drivers/acpi/acpica/exfldio.c index 5b76c11418fa..149de45fdadd 100644 --- a/drivers/acpi/acpica/exfldio.c +++ b/drivers/acpi/acpica/exfldio.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exmisc.c b/drivers/acpi/acpica/exmisc.c index 703d88ed0b3d..0a0893310348 100644 --- a/drivers/acpi/acpica/exmisc.c +++ b/drivers/acpi/acpica/exmisc.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exmutex.c b/drivers/acpi/acpica/exmutex.c index be1c56ead653..60933e9dc3c0 100644 --- a/drivers/acpi/acpica/exmutex.c +++ b/drivers/acpi/acpica/exmutex.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exnames.c b/drivers/acpi/acpica/exnames.c index 49ec049c157e..fcc75fa27d32 100644 --- a/drivers/acpi/acpica/exnames.c +++ b/drivers/acpi/acpica/exnames.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exoparg1.c b/drivers/acpi/acpica/exoparg1.c index 236ead14b7f7..9ba8c73cea16 100644 --- a/drivers/acpi/acpica/exoparg1.c +++ b/drivers/acpi/acpica/exoparg1.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exoparg2.c b/drivers/acpi/acpica/exoparg2.c index 2571b4a310f4..879e8a277b94 100644 --- a/drivers/acpi/acpica/exoparg2.c +++ b/drivers/acpi/acpica/exoparg2.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exoparg3.c b/drivers/acpi/acpica/exoparg3.c index 1b48d9d28c9a..71fcc65c9ffa 100644 --- a/drivers/acpi/acpica/exoparg3.c +++ b/drivers/acpi/acpica/exoparg3.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exoparg6.c b/drivers/acpi/acpica/exoparg6.c index f4a2787e8e92..0786b8659061 100644 --- a/drivers/acpi/acpica/exoparg6.c +++ b/drivers/acpi/acpica/exoparg6.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exprep.c b/drivers/acpi/acpica/exprep.c index 18b75fed626e..30157f5a12d7 100644 --- a/drivers/acpi/acpica/exprep.c +++ b/drivers/acpi/acpica/exprep.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exregion.c b/drivers/acpi/acpica/exregion.c index f0d5e14f1f2c..12d51df6d3bf 100644 --- a/drivers/acpi/acpica/exregion.c +++ b/drivers/acpi/acpica/exregion.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exresnte.c b/drivers/acpi/acpica/exresnte.c index 55997e46948b..fa50e77e64a8 100644 --- a/drivers/acpi/acpica/exresnte.c +++ b/drivers/acpi/acpica/exresnte.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exresolv.c b/drivers/acpi/acpica/exresolv.c index db502cd7d934..6e335dc34528 100644 --- a/drivers/acpi/acpica/exresolv.c +++ b/drivers/acpi/acpica/exresolv.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exresop.c b/drivers/acpi/acpica/exresop.c index e3bb00ccdff5..a67b1d925ddd 100644 --- a/drivers/acpi/acpica/exresop.c +++ b/drivers/acpi/acpica/exresop.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exstore.c b/drivers/acpi/acpica/exstore.c index c0c8842dd344..c6cf843cc4c9 100644 --- a/drivers/acpi/acpica/exstore.c +++ b/drivers/acpi/acpica/exstore.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exstoren.c b/drivers/acpi/acpica/exstoren.c index a979017d56b8..b35bed52e061 100644 --- a/drivers/acpi/acpica/exstoren.c +++ b/drivers/acpi/acpica/exstoren.c @@ -7,7 +7,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exstorob.c b/drivers/acpi/acpica/exstorob.c index dc665cc554de..65a45d8335c8 100644 --- a/drivers/acpi/acpica/exstorob.c +++ b/drivers/acpi/acpica/exstorob.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exsystem.c b/drivers/acpi/acpica/exsystem.c index df66e7b686be..191a12945226 100644 --- a/drivers/acpi/acpica/exsystem.c +++ b/drivers/acpi/acpica/exsystem.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/exutils.c b/drivers/acpi/acpica/exutils.c index 3993aa5084b6..eb6798ba8b59 100644 --- a/drivers/acpi/acpica/exutils.c +++ b/drivers/acpi/acpica/exutils.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwacpi.c b/drivers/acpi/acpica/hwacpi.c index fc380d3d45ab..d21ec5f0b3a9 100644 --- a/drivers/acpi/acpica/hwacpi.c +++ b/drivers/acpi/acpica/hwacpi.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwgpe.c b/drivers/acpi/acpica/hwgpe.c index f610d88a66be..1a6894afef79 100644 --- a/drivers/acpi/acpica/hwgpe.c +++ b/drivers/acpi/acpica/hwgpe.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwpci.c b/drivers/acpi/acpica/hwpci.c index 050fd227951b..1455ddcdc32c 100644 --- a/drivers/acpi/acpica/hwpci.c +++ b/drivers/acpi/acpica/hwpci.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwregs.c b/drivers/acpi/acpica/hwregs.c index cc70f3fdcdd1..4ea4eeb51bfd 100644 --- a/drivers/acpi/acpica/hwregs.c +++ b/drivers/acpi/acpica/hwregs.c @@ -7,7 +7,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwsleep.c b/drivers/acpi/acpica/hwsleep.c index d52da3073650..3c4a922a9fc2 100644 --- a/drivers/acpi/acpica/hwsleep.c +++ b/drivers/acpi/acpica/hwsleep.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwtimer.c b/drivers/acpi/acpica/hwtimer.c index 50d21c40b5c1..d4973d9da9f1 100644 --- a/drivers/acpi/acpica/hwtimer.c +++ b/drivers/acpi/acpica/hwtimer.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwvalid.c b/drivers/acpi/acpica/hwvalid.c index f8fbea947013..6e5c43a60bb7 100644 --- a/drivers/acpi/acpica/hwvalid.c +++ b/drivers/acpi/acpica/hwvalid.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/hwxface.c b/drivers/acpi/acpica/hwxface.c index c2793a82f120..7bdbbcf35f85 100644 --- a/drivers/acpi/acpica/hwxface.c +++ b/drivers/acpi/acpica/hwxface.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsaccess.c b/drivers/acpi/acpica/nsaccess.c index d93172fd15a8..61623f3f6826 100644 --- a/drivers/acpi/acpica/nsaccess.c +++ b/drivers/acpi/acpica/nsaccess.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsalloc.c b/drivers/acpi/acpica/nsalloc.c index 1d0ef15d158f..7c3d3ceb98b3 100644 --- a/drivers/acpi/acpica/nsalloc.c +++ b/drivers/acpi/acpica/nsalloc.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsdump.c b/drivers/acpi/acpica/nsdump.c index b683cc2ff9d3..b7f2b3be79ac 100644 --- a/drivers/acpi/acpica/nsdump.c +++ b/drivers/acpi/acpica/nsdump.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsdumpdv.c b/drivers/acpi/acpica/nsdumpdv.c index 2ed294b7a4db..30ea5bc53a78 100644 --- a/drivers/acpi/acpica/nsdumpdv.c +++ b/drivers/acpi/acpica/nsdumpdv.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nseval.c b/drivers/acpi/acpica/nseval.c index c1bd02b1a058..f375cb82e321 100644 --- a/drivers/acpi/acpica/nseval.c +++ b/drivers/acpi/acpica/nseval.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsinit.c b/drivers/acpi/acpica/nsinit.c index fd7c6380e294..9d84ec2f0211 100644 --- a/drivers/acpi/acpica/nsinit.c +++ b/drivers/acpi/acpica/nsinit.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsload.c b/drivers/acpi/acpica/nsload.c index 5f7dc691c183..5cbf15ffe7d8 100644 --- a/drivers/acpi/acpica/nsload.c +++ b/drivers/acpi/acpica/nsload.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsnames.c b/drivers/acpi/acpica/nsnames.c index d5fa520c3de5..b20e7c8c3ffb 100644 --- a/drivers/acpi/acpica/nsnames.c +++ b/drivers/acpi/acpica/nsnames.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsobject.c b/drivers/acpi/acpica/nsobject.c index 3bb8bf105ea2..dd77a3ce6e50 100644 --- a/drivers/acpi/acpica/nsobject.c +++ b/drivers/acpi/acpica/nsobject.c @@ -6,7 +6,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsparse.c b/drivers/acpi/acpica/nsparse.c index b3234fa795b8..ec7ba2d3463c 100644 --- a/drivers/acpi/acpica/nsparse.c +++ b/drivers/acpi/acpica/nsparse.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nspredef.c b/drivers/acpi/acpica/nspredef.c index db80366b3dd9..bbe46a447d34 100644 --- a/drivers/acpi/acpica/nspredef.c +++ b/drivers/acpi/acpica/nspredef.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsrepair.c b/drivers/acpi/acpica/nsrepair.c index dd0a0bdc4fc7..9c35d20eb52b 100644 --- a/drivers/acpi/acpica/nsrepair.c +++ b/drivers/acpi/acpica/nsrepair.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsrepair2.c b/drivers/acpi/acpica/nsrepair2.c index 3ee2b22a9f7d..726bc8e687f7 100644 --- a/drivers/acpi/acpica/nsrepair2.c +++ b/drivers/acpi/acpica/nsrepair2.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nssearch.c b/drivers/acpi/acpica/nssearch.c index 28b0d7a62b99..507043d66114 100644 --- a/drivers/acpi/acpica/nssearch.c +++ b/drivers/acpi/acpica/nssearch.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsutils.c b/drivers/acpi/acpica/nsutils.c index cb1b104a69a2..a535b7afda5c 100644 --- a/drivers/acpi/acpica/nsutils.c +++ b/drivers/acpi/acpica/nsutils.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nswalk.c b/drivers/acpi/acpica/nswalk.c index 345f0c3c6ad2..f69895a54895 100644 --- a/drivers/acpi/acpica/nswalk.c +++ b/drivers/acpi/acpica/nswalk.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsxfeval.c b/drivers/acpi/acpica/nsxfeval.c index e7f016d1b226..71d15f61807b 100644 --- a/drivers/acpi/acpica/nsxfeval.c +++ b/drivers/acpi/acpica/nsxfeval.c @@ -6,7 +6,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsxfname.c b/drivers/acpi/acpica/nsxfname.c index 83bf93024303..af401c9c4dfc 100644 --- a/drivers/acpi/acpica/nsxfname.c +++ b/drivers/acpi/acpica/nsxfname.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/nsxfobj.c b/drivers/acpi/acpica/nsxfobj.c index 57e6d825ed84..880a605cee20 100644 --- a/drivers/acpi/acpica/nsxfobj.c +++ b/drivers/acpi/acpica/nsxfobj.c @@ -6,7 +6,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psargs.c b/drivers/acpi/acpica/psargs.c index 4037f9032481..5ac36aba507c 100644 --- a/drivers/acpi/acpica/psargs.c +++ b/drivers/acpi/acpica/psargs.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psloop.c b/drivers/acpi/acpica/psloop.c index 01dd70d1de51..9547ad8a620b 100644 --- a/drivers/acpi/acpica/psloop.c +++ b/drivers/acpi/acpica/psloop.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psopcode.c b/drivers/acpi/acpica/psopcode.c index 9b3191001b3d..a0226fdcf75c 100644 --- a/drivers/acpi/acpica/psopcode.c +++ b/drivers/acpi/acpica/psopcode.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psparse.c b/drivers/acpi/acpica/psparse.c index 9bb0cbd37b5e..2ff9c35a1968 100644 --- a/drivers/acpi/acpica/psparse.c +++ b/drivers/acpi/acpica/psparse.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psscope.c b/drivers/acpi/acpica/psscope.c index a5faa1323a02..c872aa4b926e 100644 --- a/drivers/acpi/acpica/psscope.c +++ b/drivers/acpi/acpica/psscope.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/pstree.c b/drivers/acpi/acpica/pstree.c index 04311bbfcccb..2b03cdbbe1c0 100644 --- a/drivers/acpi/acpica/pstree.c +++ b/drivers/acpi/acpica/pstree.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psutils.c b/drivers/acpi/acpica/psutils.c index 7eda78503422..13bb131ae125 100644 --- a/drivers/acpi/acpica/psutils.c +++ b/drivers/acpi/acpica/psutils.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/pswalk.c b/drivers/acpi/acpica/pswalk.c index 3312d6368bf1..ab96cf47896d 100644 --- a/drivers/acpi/acpica/pswalk.c +++ b/drivers/acpi/acpica/pswalk.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/psxface.c b/drivers/acpi/acpica/psxface.c index 8086805d4494..9d98c5ff66a5 100644 --- a/drivers/acpi/acpica/psxface.c +++ b/drivers/acpi/acpica/psxface.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsaddr.c b/drivers/acpi/acpica/rsaddr.c index 9e66f9078426..a0305652394f 100644 --- a/drivers/acpi/acpica/rsaddr.c +++ b/drivers/acpi/acpica/rsaddr.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rscalc.c b/drivers/acpi/acpica/rscalc.c index 3a29b4dd4453..3c6df4b7eb2d 100644 --- a/drivers/acpi/acpica/rscalc.c +++ b/drivers/acpi/acpica/rscalc.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rscreate.c b/drivers/acpi/acpica/rscreate.c index 06264e8b58d9..46d6eb38ae66 100644 --- a/drivers/acpi/acpica/rscreate.c +++ b/drivers/acpi/acpica/rscreate.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsdump.c b/drivers/acpi/acpica/rsdump.c index 9969985bf2b7..b4c581132393 100644 --- a/drivers/acpi/acpica/rsdump.c +++ b/drivers/acpi/acpica/rsdump.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsinfo.c b/drivers/acpi/acpica/rsinfo.c index e5489bf475fa..a9fa5158200b 100644 --- a/drivers/acpi/acpica/rsinfo.c +++ b/drivers/acpi/acpica/rsinfo.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsio.c b/drivers/acpi/acpica/rsio.c index 0c7efef008be..f6a081057a22 100644 --- a/drivers/acpi/acpica/rsio.c +++ b/drivers/acpi/acpica/rsio.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsirq.c b/drivers/acpi/acpica/rsirq.c index 3ac9d4a08fcf..e23a9ec248cb 100644 --- a/drivers/acpi/acpica/rsirq.c +++ b/drivers/acpi/acpica/rsirq.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rslist.c b/drivers/acpi/acpica/rslist.c index c44635136598..9be129f5d6f4 100644 --- a/drivers/acpi/acpica/rslist.c +++ b/drivers/acpi/acpica/rslist.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsmemory.c b/drivers/acpi/acpica/rsmemory.c index 7cc6d8625f1e..4fd611ad02b4 100644 --- a/drivers/acpi/acpica/rsmemory.c +++ b/drivers/acpi/acpica/rsmemory.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsmisc.c b/drivers/acpi/acpica/rsmisc.c index af3dc63163b6..8073b371cc7c 100644 --- a/drivers/acpi/acpica/rsmisc.c +++ b/drivers/acpi/acpica/rsmisc.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsserial.c b/drivers/acpi/acpica/rsserial.c index d0b8e121843a..9aa5e689b444 100644 --- a/drivers/acpi/acpica/rsserial.c +++ b/drivers/acpi/acpica/rsserial.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsutils.c b/drivers/acpi/acpica/rsutils.c index 54a9a974b624..433a375deb93 100644 --- a/drivers/acpi/acpica/rsutils.c +++ b/drivers/acpi/acpica/rsutils.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/rsxface.c b/drivers/acpi/acpica/rsxface.c index f6cceaccc3e9..f58c098c7aeb 100644 --- a/drivers/acpi/acpica/rsxface.c +++ b/drivers/acpi/acpica/rsxface.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/tbfadt.c b/drivers/acpi/acpica/tbfadt.c index ff1f7f041f59..c5d870406f41 100644 --- a/drivers/acpi/acpica/tbfadt.c +++ b/drivers/acpi/acpica/tbfadt.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/tbfind.c b/drivers/acpi/acpica/tbfind.c index a55cb2bb5abb..4903e36ea75a 100644 --- a/drivers/acpi/acpica/tbfind.c +++ b/drivers/acpi/acpica/tbfind.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/tbinstal.c b/drivers/acpi/acpica/tbinstal.c index 62365f6075dd..1aecf7baa4e0 100644 --- a/drivers/acpi/acpica/tbinstal.c +++ b/drivers/acpi/acpica/tbinstal.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/tbutils.c b/drivers/acpi/acpica/tbutils.c index da56a1908966..09ca39e14337 100644 --- a/drivers/acpi/acpica/tbutils.c +++ b/drivers/acpi/acpica/tbutils.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/tbxface.c b/drivers/acpi/acpica/tbxface.c index e7d13f5d3f2d..abcc6412c244 100644 --- a/drivers/acpi/acpica/tbxface.c +++ b/drivers/acpi/acpica/tbxface.c @@ -6,7 +6,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/tbxfroot.c b/drivers/acpi/acpica/tbxfroot.c index 7eb6c6cc1edf..4258f647ca3d 100644 --- a/drivers/acpi/acpica/tbxfroot.c +++ b/drivers/acpi/acpica/tbxfroot.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utalloc.c b/drivers/acpi/acpica/utalloc.c index 0a697351cf69..9982d2ea66fb 100644 --- a/drivers/acpi/acpica/utalloc.c +++ b/drivers/acpi/acpica/utalloc.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utcopy.c b/drivers/acpi/acpica/utcopy.c index aded299a2fa8..3317c0a406ee 100644 --- a/drivers/acpi/acpica/utcopy.c +++ b/drivers/acpi/acpica/utcopy.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utdebug.c b/drivers/acpi/acpica/utdebug.c index a1f8d7509e66..a0998a886318 100644 --- a/drivers/acpi/acpica/utdebug.c +++ b/drivers/acpi/acpica/utdebug.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utdecode.c b/drivers/acpi/acpica/utdecode.c index 6fd56d549296..d42ede5260c7 100644 --- a/drivers/acpi/acpica/utdecode.c +++ b/drivers/acpi/acpica/utdecode.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utdelete.c b/drivers/acpi/acpica/utdelete.c index 93ec06b0fe25..2a6c3e183697 100644 --- a/drivers/acpi/acpica/utdelete.c +++ b/drivers/acpi/acpica/utdelete.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/uteval.c b/drivers/acpi/acpica/uteval.c index 18f73c9d10bc..479f32b33415 100644 --- a/drivers/acpi/acpica/uteval.c +++ b/drivers/acpi/acpica/uteval.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utglobal.c b/drivers/acpi/acpica/utglobal.c index 3cf758bab242..4153584cf526 100644 --- a/drivers/acpi/acpica/utglobal.c +++ b/drivers/acpi/acpica/utglobal.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utids.c b/drivers/acpi/acpica/utids.c index b679ea693545..c92eb1d93785 100644 --- a/drivers/acpi/acpica/utids.c +++ b/drivers/acpi/acpica/utids.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utinit.c b/drivers/acpi/acpica/utinit.c index cab61a05d21d..8359c0c5dc98 100644 --- a/drivers/acpi/acpica/utinit.c +++ b/drivers/acpi/acpica/utinit.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utlock.c b/drivers/acpi/acpica/utlock.c index f6bb75c6faf5..155fd786d0f2 100644 --- a/drivers/acpi/acpica/utlock.c +++ b/drivers/acpi/acpica/utlock.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utmath.c b/drivers/acpi/acpica/utmath.c index ce481da9bb45..2491a552b0e6 100644 --- a/drivers/acpi/acpica/utmath.c +++ b/drivers/acpi/acpica/utmath.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utmisc.c b/drivers/acpi/acpica/utmisc.c index c33a852d4f42..86f19db74e05 100644 --- a/drivers/acpi/acpica/utmisc.c +++ b/drivers/acpi/acpica/utmisc.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utmutex.c b/drivers/acpi/acpica/utmutex.c index 302c74e26f95..43174df33121 100644 --- a/drivers/acpi/acpica/utmutex.c +++ b/drivers/acpi/acpica/utmutex.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utobject.c b/drivers/acpi/acpica/utobject.c index 188340a017b4..b112744fc9ae 100644 --- a/drivers/acpi/acpica/utobject.c +++ b/drivers/acpi/acpica/utobject.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utosi.c b/drivers/acpi/acpica/utosi.c index 1fb10cb8f11d..2360cf70c18c 100644 --- a/drivers/acpi/acpica/utosi.c +++ b/drivers/acpi/acpica/utosi.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utresrc.c b/drivers/acpi/acpica/utresrc.c index cd7fbbf57f30..9d441ea70305 100644 --- a/drivers/acpi/acpica/utresrc.c +++ b/drivers/acpi/acpica/utresrc.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utstate.c b/drivers/acpi/acpica/utstate.c index 30c21e1a9360..4267477c2797 100644 --- a/drivers/acpi/acpica/utstate.c +++ b/drivers/acpi/acpica/utstate.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utxface.c b/drivers/acpi/acpica/utxface.c index 15cbddebb6b4..644e8c8ebc4b 100644 --- a/drivers/acpi/acpica/utxface.c +++ b/drivers/acpi/acpica/utxface.c @@ -5,7 +5,7 @@ *****************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utxferror.c b/drivers/acpi/acpica/utxferror.c index 8d0245ec4315..52b568af1819 100644 --- a/drivers/acpi/acpica/utxferror.c +++ b/drivers/acpi/acpica/utxferror.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without diff --git a/drivers/acpi/acpica/utxfmutex.c b/drivers/acpi/acpica/utxfmutex.c index 1ea016b228bd..1427d191d15a 100644 --- a/drivers/acpi/acpica/utxfmutex.c +++ b/drivers/acpi/acpica/utxfmutex.c @@ -5,7 +5,7 @@ ******************************************************************************/ /* - * Copyright (C) 2000 - 2011, Intel Corp. + * Copyright (C) 2000 - 2012, Intel Corp. * All rights reserved. * * Redistribution and use in source and binary forms, with or without From 5d1f86572685d577b76142b7c61453000e1a5e26 Mon Sep 17 00:00:00 2001 From: Bob Moore Date: Thu, 12 Jan 2012 13:29:41 +0800 Subject: [PATCH 451/820] ACPICA: Update to version 20120111 Version 20120111. Signed-off-by: Bob Moore Signed-off-by: Lin Ming Signed-off-by: Len Brown --- include/acpi/acpixf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/acpi/acpixf.h b/include/acpi/acpixf.h index a4cfded67c08..dceeaf51d706 100644 --- a/include/acpi/acpixf.h +++ b/include/acpi/acpixf.h @@ -47,7 +47,7 @@ /* Current ACPICA subsystem version in YYYYMMDD format */ -#define ACPI_CA_VERSION 0x20111123 +#define ACPI_CA_VERSION 0x20120111 #include "actypes.h" #include "actbl.h" From 04c25997c97e57d95d1f5a7a6e5fff2411dbd48b Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:40 +0800 Subject: [PATCH 452/820] ACPI, Add 64bit read/write support to atomicio on i386 There is no 64bit read/write support in ACPI atomicio because readq/writeq is used to implement 64bit read/write, but readq/writeq is not available on i386. This patch implement 64bit read/write support in atomicio via two readl/writel. Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/atomicio.c | 36 ++++++++++++++++++++++++++++++------ 1 file changed, 30 insertions(+), 6 deletions(-) diff --git a/drivers/acpi/atomicio.c b/drivers/acpi/atomicio.c index cfc0cc10af39..1016f186c17c 100644 --- a/drivers/acpi/atomicio.c +++ b/drivers/acpi/atomicio.c @@ -260,6 +260,21 @@ int acpi_post_unmap_gar(struct acpi_generic_address *reg) } EXPORT_SYMBOL_GPL(acpi_post_unmap_gar); +#ifdef readq +static inline u64 read64(const volatile void __iomem *addr) +{ + return readq(addr); +} +#else +static inline u64 read64(const volatile void __iomem *addr) +{ + u64 l, h; + l = readl(addr); + h = readl(addr+4); + return l | (h << 32); +} +#endif + /* * Can be used in atomic (including NMI) or process context. RCU read * lock can only be released after the IO memory area accessing. @@ -280,11 +295,9 @@ static int acpi_atomic_read_mem(u64 paddr, u64 *val, u32 width) case 32: *val = readl(addr); break; -#ifdef readq case 64: - *val = readq(addr); + *val = read64(addr); break; -#endif default: return -EINVAL; } @@ -293,6 +306,19 @@ static int acpi_atomic_read_mem(u64 paddr, u64 *val, u32 width) return 0; } +#ifdef writeq +static inline void write64(u64 val, volatile void __iomem *addr) +{ + writeq(val, addr); +} +#else +static inline void write64(u64 val, volatile void __iomem *addr) +{ + writel(val, addr); + writel(val>>32, addr+4); +} +#endif + static int acpi_atomic_write_mem(u64 paddr, u64 val, u32 width) { void __iomem *addr; @@ -309,11 +335,9 @@ static int acpi_atomic_write_mem(u64 paddr, u64 val, u32 width) case 32: writel(val, addr); break; -#ifdef writeq case 64: - writeq(val, addr); + write64(val, addr); break; -#endif default: return -EINVAL; } From a654e5ee4f2213844d23361eda4955fe9efaf35f Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:41 +0800 Subject: [PATCH 453/820] ACPI, APEI, GHES: Add PCIe AER recovery support aer_recover_queue() is called when recoverable PCIe AER errors are notified by firmware to do the recovery work. Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/ghes.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c index b8e08cb67a18..511b971d1148 100644 --- a/drivers/acpi/apei/ghes.c +++ b/drivers/acpi/apei/ghes.c @@ -45,6 +45,8 @@ #include #include #include +#include +#include #include #include #include @@ -476,6 +478,27 @@ static void ghes_do_proc(const struct acpi_hest_generic_status *estatus) } #endif } +#ifdef CONFIG_ACPI_APEI_PCIEAER + else if (!uuid_le_cmp(*(uuid_le *)gdata->section_type, + CPER_SEC_PCIE)) { + struct cper_sec_pcie *pcie_err; + pcie_err = (struct cper_sec_pcie *)(gdata+1); + if (sev == GHES_SEV_RECOVERABLE && + sec_sev == GHES_SEV_RECOVERABLE && + pcie_err->validation_bits & CPER_PCIE_VALID_DEVICE_ID && + pcie_err->validation_bits & CPER_PCIE_VALID_AER_INFO) { + unsigned int devfn; + int aer_severity; + devfn = PCI_DEVFN(pcie_err->device_id.device, + pcie_err->device_id.function); + aer_severity = cper_severity_to_aer(sev); + aer_recover_queue(pcie_err->device_id.segment, + pcie_err->device_id.bus, + devfn, aer_severity); + } + + } +#endif } } From 46b91e379f7180b482b789fbe615946d91e3a07f Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 8 Dec 2011 11:25:42 +0800 Subject: [PATCH 454/820] ACPI, APEI, Print resource errors in conventional format Use the normal %pR-like format for MMIO and I/O port ranges. Signed-off-by: Bjorn Helgaas Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/apei-base.c | 8 ++++---- drivers/acpi/apei/einj.c | 11 ++++++----- 2 files changed, 10 insertions(+), 9 deletions(-) diff --git a/drivers/acpi/apei/apei-base.c b/drivers/acpi/apei/apei-base.c index 61540360d5ce..3492896b96f2 100644 --- a/drivers/acpi/apei/apei-base.c +++ b/drivers/acpi/apei/apei-base.c @@ -460,9 +460,9 @@ int apei_resources_request(struct apei_resources *resources, desc); if (!r) { pr_err(APEI_PFX - "Can not request iomem region <%016llx-%016llx> for GARs.\n", + "Can not request [mem %#010llx-%#010llx] for %s registers\n", (unsigned long long)res->start, - (unsigned long long)res->end); + (unsigned long long)res->end - 1, desc); res_bak = res; goto err_unmap_iomem; } @@ -472,9 +472,9 @@ int apei_resources_request(struct apei_resources *resources, r = request_region(res->start, res->end - res->start, desc); if (!r) { pr_err(APEI_PFX - "Can not request ioport region <%016llx-%016llx> for GARs.\n", + "Can not request [io %#06llx-%#06llx] for %s registers\n", (unsigned long long)res->start, - (unsigned long long)res->end); + (unsigned long long)res->end - 1, desc); res_bak = res; goto err_unmap_ioport; } diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index 589b96c38704..d25390274d7b 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -209,9 +209,10 @@ static int __einj_error_trigger(u64 trigger_paddr) "APEI EINJ Trigger Table"); if (!r) { pr_err(EINJ_PFX - "Can not request iomem region <%016llx-%016llx> for Trigger table.\n", + "Can not request [mem %#010llx-%#010llx] for Trigger table\n", (unsigned long long)trigger_paddr, - (unsigned long long)trigger_paddr+sizeof(*trigger_tab)); + (unsigned long long)trigger_paddr + + sizeof(*trigger_tab) - 1); goto out; } trigger_tab = ioremap_cache(trigger_paddr, sizeof(*trigger_tab)); @@ -232,9 +233,9 @@ static int __einj_error_trigger(u64 trigger_paddr) "APEI EINJ Trigger Table"); if (!r) { pr_err(EINJ_PFX -"Can not request iomem region <%016llx-%016llx> for Trigger Table Entry.\n", - (unsigned long long)trigger_paddr+sizeof(*trigger_tab), - (unsigned long long)trigger_paddr + table_size); +"Can not request [mem %#010llx-%#010llx] for Trigger Table Entry\n", + (unsigned long long)trigger_paddr + sizeof(*trigger_tab), + (unsigned long long)trigger_paddr + table_size - 1); goto out_rel_header; } iounmap(trigger_tab); From ad6861547b52ad7c31eacc336b79ac91d7fded75 Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:43 +0800 Subject: [PATCH 455/820] ACPI, APEI, Remove table not found message Because APEI tables are optional, these message may confuse users, for example, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/599715 Reported-by: Bjorn Helgaas Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/einj.c | 5 ++--- drivers/acpi/apei/erst.c | 5 ++--- drivers/acpi/apei/hest.c | 5 ++--- 3 files changed, 6 insertions(+), 9 deletions(-) diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index d25390274d7b..43eeb2e6e635 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -466,10 +466,9 @@ static int __init einj_init(void) status = acpi_get_table(ACPI_SIG_EINJ, 0, (struct acpi_table_header **)&einj_tab); - if (status == AE_NOT_FOUND) { - pr_info(EINJ_PFX "Table is not found!\n"); + if (status == AE_NOT_FOUND) return -ENODEV; - } else if (ACPI_FAILURE(status)) { + else if (ACPI_FAILURE(status)) { const char *msg = acpi_format_exception(status); pr_err(EINJ_PFX "Failed to get table, %s\n", msg); return -EINVAL; diff --git a/drivers/acpi/apei/erst.c b/drivers/acpi/apei/erst.c index 631b9477b99c..8e8d786c5d23 100644 --- a/drivers/acpi/apei/erst.c +++ b/drivers/acpi/apei/erst.c @@ -1125,10 +1125,9 @@ static int __init erst_init(void) status = acpi_get_table(ACPI_SIG_ERST, 0, (struct acpi_table_header **)&erst_tab); - if (status == AE_NOT_FOUND) { - pr_info(ERST_PFX "Table is not found!\n"); + if (status == AE_NOT_FOUND) goto err; - } else if (ACPI_FAILURE(status)) { + else if (ACPI_FAILURE(status)) { const char *msg = acpi_format_exception(status); pr_err(ERST_PFX "Failed to get table, %s\n", msg); rc = -EINVAL; diff --git a/drivers/acpi/apei/hest.c b/drivers/acpi/apei/hest.c index 05fee06f4d6e..f709269d4932 100644 --- a/drivers/acpi/apei/hest.c +++ b/drivers/acpi/apei/hest.c @@ -221,10 +221,9 @@ void __init acpi_hest_init(void) status = acpi_get_table(ACPI_SIG_HEST, 0, (struct acpi_table_header **)&hest_tab); - if (status == AE_NOT_FOUND) { - pr_info(HEST_PFX "Table not found.\n"); + if (status == AE_NOT_FOUND) goto err; - } else if (ACPI_FAILURE(status)) { + else if (ACPI_FAILURE(status)) { const char *msg = acpi_format_exception(status); pr_err(HEST_PFX "Failed to get table, %s\n", msg); rc = -EINVAL; From 5ba82ab534a325d310fe02af1c149f1072792c7b Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:44 +0800 Subject: [PATCH 456/820] ACPI, APEI, GHES, Distinguish interleaved error report in kernel log In most cases, printk only guarantees messages from different printk calling will not be interleaved between each other. But, one APEI GHES hardware error report will involve multiple printk calling, normally each for one line. So it is possible that the hardware error report comes from different generic hardware error source will be interleaved. In this patch, a sequence number is prefixed to each line of error report. So that, even if they are interleaved, they still can be distinguished by the prefixed sequence number. Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/ghes.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c index 511b971d1148..9dcb2d86aeab 100644 --- a/drivers/acpi/apei/ghes.c +++ b/drivers/acpi/apei/ghes.c @@ -506,16 +506,22 @@ static void __ghes_print_estatus(const char *pfx, const struct acpi_hest_generic *generic, const struct acpi_hest_generic_status *estatus) { + static atomic_t seqno; + unsigned int curr_seqno; + char pfx_seq[64]; + if (pfx == NULL) { if (ghes_severity(estatus->error_severity) <= GHES_SEV_CORRECTED) - pfx = KERN_WARNING HW_ERR; + pfx = KERN_WARNING; else - pfx = KERN_ERR HW_ERR; + pfx = KERN_ERR; } + curr_seqno = atomic_inc_return(&seqno); + snprintf(pfx_seq, sizeof(pfx_seq), "%s{%u}" HW_ERR, pfx, curr_seqno); printk("%s""Hardware error from APEI Generic Hardware Error Source: %d\n", - pfx, generic->header.source_id); - apei_estatus_print(pfx, estatus); + pfx_seq, generic->header.source_id); + apei_estatus_print(pfx_seq, estatus); } static int ghes_print_estatus(const char *pfx, @@ -798,7 +804,7 @@ static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs) if (sev_global >= GHES_SEV_PANIC) { oops_begin(); - __ghes_print_estatus(KERN_EMERG HW_ERR, ghes_global->generic, + __ghes_print_estatus(KERN_EMERG, ghes_global->generic, ghes_global->estatus); /* reboot to log the error! */ if (panic_timeout == 0) From 46d12f0bcb17b2de89a059114349d472b7eb1783 Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:45 +0800 Subject: [PATCH 457/820] ACPI, APEI, Printk queued error record before panic Because printk is not safe inside NMI handler, the recoverable error records received in NMI handler will be queued to be printked in a delayed IRQ context via irq_work. If a fatal error occurs after the recoverable error and before the irq_work processed, we lost a error report. To solve the issue, the queued error records are printked in NMI handler if system will go panic. Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/ghes.c | 59 ++++++++++++++++++++++++++++++++-------- 1 file changed, 47 insertions(+), 12 deletions(-) diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c index 9dcb2d86aeab..aaf36090de1e 100644 --- a/drivers/acpi/apei/ghes.c +++ b/drivers/acpi/apei/ghes.c @@ -740,26 +740,34 @@ static int ghes_notify_sci(struct notifier_block *this, return ret; } -static void ghes_proc_in_irq(struct irq_work *irq_work) +static struct llist_node *llist_nodes_reverse(struct llist_node *llnode) { - struct llist_node *llnode, *next, *tail = NULL; - struct ghes_estatus_node *estatus_node; - struct acpi_hest_generic *generic; - struct acpi_hest_generic_status *estatus; - u32 len, node_len; + struct llist_node *next, *tail = NULL; - /* - * Because the time order of estatus in list is reversed, - * revert it back to proper order. - */ - llnode = llist_del_all(&ghes_estatus_llist); while (llnode) { next = llnode->next; llnode->next = tail; tail = llnode; llnode = next; } - llnode = tail; + + return tail; +} + +static void ghes_proc_in_irq(struct irq_work *irq_work) +{ + struct llist_node *llnode, *next; + struct ghes_estatus_node *estatus_node; + struct acpi_hest_generic *generic; + struct acpi_hest_generic_status *estatus; + u32 len, node_len; + + llnode = llist_del_all(&ghes_estatus_llist); + /* + * Because the time order of estatus in list is reversed, + * revert it back to proper order. + */ + llnode = llist_nodes_reverse(llnode); while (llnode) { next = llnode->next; estatus_node = llist_entry(llnode, struct ghes_estatus_node, @@ -779,6 +787,32 @@ static void ghes_proc_in_irq(struct irq_work *irq_work) } } +static void ghes_print_queued_estatus(void) +{ + struct llist_node *llnode; + struct ghes_estatus_node *estatus_node; + struct acpi_hest_generic *generic; + struct acpi_hest_generic_status *estatus; + u32 len, node_len; + + llnode = llist_del_all(&ghes_estatus_llist); + /* + * Because the time order of estatus in list is reversed, + * revert it back to proper order. + */ + llnode = llist_nodes_reverse(llnode); + while (llnode) { + estatus_node = llist_entry(llnode, struct ghes_estatus_node, + llnode); + estatus = GHES_ESTATUS_FROM_NODE(estatus_node); + len = apei_estatus_len(estatus); + node_len = GHES_ESTATUS_NODE_LEN(len); + generic = estatus_node->generic; + ghes_print_estatus(NULL, generic, estatus); + llnode = llnode->next; + } +} + static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs) { struct ghes *ghes, *ghes_global = NULL; @@ -804,6 +838,7 @@ static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs) if (sev_global >= GHES_SEV_PANIC) { oops_begin(); + ghes_print_queued_estatus(); __ghes_print_estatus(KERN_EMERG, ghes_global->generic, ghes_global->estatus); /* reboot to log the error! */ From 76da3fb3575e39fb23b2c072997ccd1187a2ce9d Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:46 +0800 Subject: [PATCH 458/820] ACPI, Add RAM mapping support to ACPI atomic IO support On one of our testing machine, the following EINJ command lines: # echo 0x10000000 > param1 # echo 0xfffffffffffff000 > param2 # echo 0x8 > error_type # echo 1 > error_inject Will get: echo: write error: Input/output error The EIO comes from: rc = apei_exec_pre_map_gars(&trigger_ctx); The root cause is as follow. Normally, ACPI atomic IO support is used to access IO memory. But in EINJ of that machine, it is used to access RAM to trigger the injected error. And the ioremap() called by apei_exec_pre_map_gars() can not map the RAM. This patch add RAM mapping support to ACPI atomic IO support to satisfy EINJ requirement. Signed-off-by: Huang Ying Tested-by: Tony Luck Signed-off-by: Len Brown --- drivers/acpi/atomicio.c | 41 +++++++++++++++++++++++++++++++++++++---- 1 file changed, 37 insertions(+), 4 deletions(-) diff --git a/drivers/acpi/atomicio.c b/drivers/acpi/atomicio.c index 1016f186c17c..d4a5b3d3657b 100644 --- a/drivers/acpi/atomicio.c +++ b/drivers/acpi/atomicio.c @@ -32,6 +32,8 @@ #include #include #include +#include +#include #include #define ACPI_PFX "ACPI: " @@ -97,6 +99,37 @@ static void __iomem *__acpi_try_ioremap(phys_addr_t paddr, return NULL; } +#ifndef CONFIG_IA64 +#define should_use_kmap(pfn) page_is_ram(pfn) +#else +/* ioremap will take care of cache attributes */ +#define should_use_kmap(pfn) 0 +#endif + +static void __iomem *acpi_map(phys_addr_t pg_off, unsigned long pg_sz) +{ + unsigned long pfn; + + pfn = pg_off >> PAGE_SHIFT; + if (should_use_kmap(pfn)) { + if (pg_sz > PAGE_SIZE) + return NULL; + return (void __iomem __force *)kmap(pfn_to_page(pfn)); + } else + return ioremap(pg_off, pg_sz); +} + +static void acpi_unmap(phys_addr_t pg_off, void __iomem *vaddr) +{ + unsigned long pfn; + + pfn = pg_off >> PAGE_SHIFT; + if (page_is_ram(pfn)) + kunmap(pfn_to_page(pfn)); + else + iounmap(vaddr); +} + /* * Used to pre-map the specified IO memory area. First try to find * whether the area is already pre-mapped, if it is, increase the @@ -119,7 +152,7 @@ static void __iomem *acpi_pre_map(phys_addr_t paddr, pg_off = paddr & PAGE_MASK; pg_sz = ((paddr + size + PAGE_SIZE - 1) & PAGE_MASK) - pg_off; - vaddr = ioremap(pg_off, pg_sz); + vaddr = acpi_map(pg_off, pg_sz); if (!vaddr) return NULL; map = kmalloc(sizeof(*map), GFP_KERNEL); @@ -135,7 +168,7 @@ static void __iomem *acpi_pre_map(phys_addr_t paddr, vaddr = __acpi_try_ioremap(paddr, size); if (vaddr) { spin_unlock_irqrestore(&acpi_iomaps_lock, flags); - iounmap(map->vaddr); + acpi_unmap(pg_off, map->vaddr); kfree(map); return vaddr; } @@ -144,7 +177,7 @@ static void __iomem *acpi_pre_map(phys_addr_t paddr, return map->vaddr + (paddr - map->paddr); err_unmap: - iounmap(vaddr); + acpi_unmap(pg_off, vaddr); return NULL; } @@ -177,7 +210,7 @@ static void acpi_post_unmap(phys_addr_t paddr, unsigned long size) return; synchronize_rcu(); - iounmap(map->vaddr); + acpi_unmap(map->paddr, map->vaddr); kfree(map); } From fdea163d8c17ba08814142259a467ba3e899010d Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:47 +0800 Subject: [PATCH 459/820] ACPI, APEI, EINJ, Fix resource conflict on some machine Some APEI firmware implementation will access injected address specified in param1 to trigger the error when injecting memory error. This will cause resource conflict with RAM. On one of our testing machine, if injecting at memory address 0x10000000, the following error will be reported in dmesg: APEI: Can not request iomem region <0000000010000000-0000000010000008> for GARs. This patch removes the injecting memory address range from trigger table resources to avoid conflict. Signed-off-by: Huang Ying Tested-by: Tony Luck Signed-off-by: Len Brown --- drivers/acpi/apei/apei-base.c | 11 +++++++++++ drivers/acpi/apei/apei-internal.h | 3 +++ drivers/acpi/apei/einj.c | 24 ++++++++++++++++++++++-- 3 files changed, 36 insertions(+), 2 deletions(-) diff --git a/drivers/acpi/apei/apei-base.c b/drivers/acpi/apei/apei-base.c index 3492896b96f2..f2c5062e2b32 100644 --- a/drivers/acpi/apei/apei-base.c +++ b/drivers/acpi/apei/apei-base.c @@ -421,6 +421,17 @@ static int apei_resources_merge(struct apei_resources *resources1, return 0; } +int apei_resources_add(struct apei_resources *resources, + unsigned long start, unsigned long size, + bool iomem) +{ + if (iomem) + return apei_res_add(&resources->iomem, start, size); + else + return apei_res_add(&resources->ioport, start, size); +} +EXPORT_SYMBOL_GPL(apei_resources_add); + /* * EINJ has two groups of GARs (EINJ table entry and trigger table * entry), so common resources are subtracted from the trigger table diff --git a/drivers/acpi/apei/apei-internal.h b/drivers/acpi/apei/apei-internal.h index f57050e7a5e7..d778edd34fba 100644 --- a/drivers/acpi/apei/apei-internal.h +++ b/drivers/acpi/apei/apei-internal.h @@ -95,6 +95,9 @@ static inline void apei_resources_init(struct apei_resources *resources) } void apei_resources_fini(struct apei_resources *resources); +int apei_resources_add(struct apei_resources *resources, + unsigned long start, unsigned long size, + bool iomem); int apei_resources_sub(struct apei_resources *resources1, struct apei_resources *resources2); int apei_resources_request(struct apei_resources *resources, diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index 43eeb2e6e635..4fdc8a3b4f6c 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -195,7 +195,8 @@ static int einj_check_trigger_header(struct acpi_einj_trigger *trigger_tab) } /* Execute instructions in trigger error action table */ -static int __einj_error_trigger(u64 trigger_paddr) +static int __einj_error_trigger(u64 trigger_paddr, u32 type, + u64 param1, u64 param2) { struct acpi_einj_trigger *trigger_tab = NULL; struct apei_exec_context trigger_ctx; @@ -256,6 +257,25 @@ static int __einj_error_trigger(u64 trigger_paddr) rc = apei_resources_sub(&trigger_resources, &einj_resources); if (rc) goto out_fini; + /* + * Some firmware will access target address specified in + * param1 to trigger the error when injecting memory error. + * This will cause resource conflict with regular memory. So + * remove it from trigger table resources. + */ + if (param_extension && (type & 0x0038) && param2) { + struct apei_resources addr_resources; + apei_resources_init(&addr_resources); + rc = apei_resources_add(&addr_resources, + param1 & param2, + ~param2 + 1, true); + if (rc) + goto out_fini; + rc = apei_resources_sub(&trigger_resources, &addr_resources); + apei_resources_fini(&addr_resources); + if (rc) + goto out_fini; + } rc = apei_resources_request(&trigger_resources, "APEI EINJ Trigger"); if (rc) goto out_fini; @@ -325,7 +345,7 @@ static int __einj_error_inject(u32 type, u64 param1, u64 param2) if (rc) return rc; trigger_paddr = apei_exec_ctx_get_output(&ctx); - rc = __einj_error_trigger(trigger_paddr); + rc = __einj_error_trigger(trigger_paddr, type, param1, param2); if (rc) return rc; rc = apei_exec_run_optional(&ctx, ACPI_EINJ_END_OPERATION); From b4e008dc53a31cb4bf6a12d9dbaf1d5c6070a838 Mon Sep 17 00:00:00 2001 From: "Xiao, Hui" Date: Thu, 8 Dec 2011 11:25:48 +0800 Subject: [PATCH 460/820] ACPI, APEI, EINJ, Refine the fix of resource conflict Current fix for resource conflict is to remove the address region from trigger resource, which is highly relies on valid user input. This patch is trying to avoid such potential issues by fetching the exact address region from trigger action table entry. Signed-off-by: Xiao, Hui Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/einj.c | 38 ++++++++++++++++++++++++++++++++------ 1 file changed, 32 insertions(+), 6 deletions(-) diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index 4fdc8a3b4f6c..6e6512e68a2d 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -194,6 +194,26 @@ static int einj_check_trigger_header(struct acpi_einj_trigger *trigger_tab) return 0; } +static struct acpi_generic_address *einj_get_trigger_parameter_region( + struct acpi_einj_trigger *trigger_tab, u64 param1, u64 param2) +{ + int i; + struct acpi_whea_header *entry; + + entry = (struct acpi_whea_header *) + ((char *)trigger_tab + sizeof(struct acpi_einj_trigger)); + for (i = 0; i < trigger_tab->entry_count; i++) { + if (entry->action == ACPI_EINJ_TRIGGER_ERROR && + entry->instruction == ACPI_EINJ_WRITE_REGISTER_VALUE && + entry->register_region.space_id == + ACPI_ADR_SPACE_SYSTEM_MEMORY && + (entry->register_region.address & param2) == (param1 & param2)) + return &entry->register_region; + entry++; + } + + return NULL; +} /* Execute instructions in trigger error action table */ static int __einj_error_trigger(u64 trigger_paddr, u32 type, u64 param1, u64 param2) @@ -205,6 +225,7 @@ static int __einj_error_trigger(u64 trigger_paddr, u32 type, struct resource *r; u32 table_size; int rc = -EIO; + struct acpi_generic_address *trigger_param_region = NULL; r = request_mem_region(trigger_paddr, sizeof(*trigger_tab), "APEI EINJ Trigger Table"); @@ -266,12 +287,17 @@ static int __einj_error_trigger(u64 trigger_paddr, u32 type, if (param_extension && (type & 0x0038) && param2) { struct apei_resources addr_resources; apei_resources_init(&addr_resources); - rc = apei_resources_add(&addr_resources, - param1 & param2, - ~param2 + 1, true); - if (rc) - goto out_fini; - rc = apei_resources_sub(&trigger_resources, &addr_resources); + trigger_param_region = einj_get_trigger_parameter_region( + trigger_tab, param1, param2); + if (trigger_param_region) { + rc = apei_resources_add(&addr_resources, + trigger_param_region->address, + trigger_param_region->bit_width/8, true); + if (rc) + goto out_fini; + rc = apei_resources_sub(&trigger_resources, + &addr_resources); + } apei_resources_fini(&addr_resources); if (rc) goto out_fini; From b54ac6d2a25084667da781c7ca2cebef52a2bcdd Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:49 +0800 Subject: [PATCH 461/820] ACPI, Record ACPI NVS regions Some firmware will access memory in ACPI NVS region via APEI. That is, instructions in APEI ERST/EINJ table will read/write ACPI NVS region. The original resource conflict checking in APEI code will check memory/ioport accessed by APEI via general resource management mechanism. But ACPI NVS region is marked as busy already, so that the false resource conflict will prevent APEI ERST/EINJ to work. To fix this, this patch record ACPI NVS regions, so that we can avoid request resources for memory region inside it. Signed-off-by: Huang Ying Signed-off-by: Len Brown --- arch/x86/kernel/e820.c | 4 ++-- drivers/acpi/Makefile | 3 ++- drivers/acpi/nvs.c | 53 +++++++++++++++++++++++++++++++++++++++++- include/linux/acpi.h | 20 +++++++++++----- 4 files changed, 70 insertions(+), 10 deletions(-) diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 303a0e48f076..51c3b186e5b9 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -714,7 +714,7 @@ void __init e820_mark_nosave_regions(unsigned long limit_pfn) } #endif -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_ACPI /** * Mark ACPI NVS memory region, so that we can save/restore it during * hibernation and the subsequent resume. @@ -727,7 +727,7 @@ static int __init e820_mark_nvs_memory(void) struct e820entry *ei = &e820.map[i]; if (ei->type == E820_NVS) - suspend_nvs_register(ei->addr, ei->size); + acpi_nvs_register(ei->addr, ei->size); } return 0; diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile index ecb26b4f29a0..c07f44f05f9d 100644 --- a/drivers/acpi/Makefile +++ b/drivers/acpi/Makefile @@ -20,11 +20,12 @@ obj-y += acpi.o \ # All the builtin files are in the "acpi." module_param namespace. acpi-y += osl.o utils.o reboot.o acpi-y += atomicio.o +acpi-y += nvs.o # sleep related files acpi-y += wakeup.o acpi-y += sleep.o -acpi-$(CONFIG_ACPI_SLEEP) += proc.o nvs.o +acpi-$(CONFIG_ACPI_SLEEP) += proc.o # diff --git a/drivers/acpi/nvs.c b/drivers/acpi/nvs.c index 096787b43c96..7a2035fa8c71 100644 --- a/drivers/acpi/nvs.c +++ b/drivers/acpi/nvs.c @@ -15,6 +15,56 @@ #include #include +/* ACPI NVS regions, APEI may use it */ + +struct nvs_region { + __u64 phys_start; + __u64 size; + struct list_head node; +}; + +static LIST_HEAD(nvs_region_list); + +#ifdef CONFIG_ACPI_SLEEP +static int suspend_nvs_register(unsigned long start, unsigned long size); +#else +static inline int suspend_nvs_register(unsigned long a, unsigned long b) +{ + return 0; +} +#endif + +int acpi_nvs_register(__u64 start, __u64 size) +{ + struct nvs_region *region; + + region = kmalloc(sizeof(*region), GFP_KERNEL); + if (!region) + return -ENOMEM; + region->phys_start = start; + region->size = size; + list_add_tail(®ion->node, &nvs_region_list); + + return suspend_nvs_register(start, size); +} + +int acpi_nvs_for_each_region(int (*func)(__u64 start, __u64 size, void *data), + void *data) +{ + int rc; + struct nvs_region *region; + + list_for_each_entry(region, &nvs_region_list, node) { + rc = func(region->phys_start, region->size, data); + if (rc) + return rc; + } + + return 0; +} + + +#ifdef CONFIG_ACPI_SLEEP /* * Platforms, like ACPI, may want us to save some memory used by them during * suspend and to restore the contents of this memory during the subsequent @@ -41,7 +91,7 @@ static LIST_HEAD(nvs_list); * things so that the data from page-aligned addresses in this region will * be copied into separate RAM pages. */ -int suspend_nvs_register(unsigned long start, unsigned long size) +static int suspend_nvs_register(unsigned long start, unsigned long size) { struct nvs_page *entry, *next; @@ -159,3 +209,4 @@ void suspend_nvs_restore(void) if (entry->data) memcpy(entry->kaddr, entry->data, entry->size); } +#endif diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 6001b4da39dd..26b75442ff7a 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -306,6 +306,11 @@ extern acpi_status acpi_pci_osc_control_set(acpi_handle handle, u32 *mask, u32 req); extern void acpi_early_init(void); +extern int acpi_nvs_register(__u64 start, __u64 size); + +extern int acpi_nvs_for_each_region(int (*func)(__u64, __u64, void *), + void *data); + #else /* !CONFIG_ACPI */ #define acpi_disabled 1 @@ -348,15 +353,18 @@ static inline int acpi_table_parse(char *id, { return -1; } -#endif /* !CONFIG_ACPI */ -#ifdef CONFIG_ACPI_SLEEP -int suspend_nvs_register(unsigned long start, unsigned long size); -#else -static inline int suspend_nvs_register(unsigned long a, unsigned long b) +static inline int acpi_nvs_register(__u64 start, __u64 size) { return 0; } -#endif + +static inline int acpi_nvs_for_each_region(int (*func)(__u64, __u64, void *), + void *data) +{ + return 0; +} + +#endif /* !CONFIG_ACPI */ #endif /*_LINUX_ACPI_H*/ From 4134b8c8811f23aa8a281db50dcee64dda414736 Mon Sep 17 00:00:00 2001 From: Huang Ying Date: Thu, 8 Dec 2011 11:25:50 +0800 Subject: [PATCH 462/820] ACPI, APEI, Resolve false conflict between ACPI NVS and APEI Some firmware will access memory in ACPI NVS region via APEI. That is, instructions in APEI ERST/EINJ table will read/write ACPI NVS region. The original resource conflict checking in APEI code will check memory/ioport accessed by APEI via general resource management mech. But ACPI NVS region is marked as busy already, so that the false resource conflict will prevent APEI ERST/EINJ to work. To fix this, this patch excludes ACPI NVS regions when APEI components request resources. So that they will not conflict with ACPI NVS regions. Reported-and-tested-by: Pavel Ivanov Signed-off-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/apei-base.c | 29 ++++++++++++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) diff --git a/drivers/acpi/apei/apei-base.c b/drivers/acpi/apei/apei-base.c index f2c5062e2b32..4abb6c74a938 100644 --- a/drivers/acpi/apei/apei-base.c +++ b/drivers/acpi/apei/apei-base.c @@ -449,8 +449,19 @@ int apei_resources_sub(struct apei_resources *resources1, } EXPORT_SYMBOL_GPL(apei_resources_sub); +static int apei_get_nvs_callback(__u64 start, __u64 size, void *data) +{ + struct apei_resources *resources = data; + return apei_res_add(&resources->iomem, start, size); +} + +static int apei_get_nvs_resources(struct apei_resources *resources) +{ + return acpi_nvs_for_each_region(apei_get_nvs_callback, resources); +} + /* - * IO memory/port rersource management mechanism is used to check + * IO memory/port resource management mechanism is used to check * whether memory/port area used by GARs conflicts with normal memory * or IO memory/port of devices. */ @@ -459,12 +470,26 @@ int apei_resources_request(struct apei_resources *resources, { struct apei_res *res, *res_bak = NULL; struct resource *r; + struct apei_resources nvs_resources; int rc; rc = apei_resources_sub(resources, &apei_resources_all); if (rc) return rc; + /* + * Some firmware uses ACPI NVS region, that has been marked as + * busy, so exclude it from APEI resources to avoid false + * conflict. + */ + apei_resources_init(&nvs_resources); + rc = apei_get_nvs_resources(&nvs_resources); + if (rc) + goto res_fini; + rc = apei_resources_sub(resources, &nvs_resources); + if (rc) + goto res_fini; + rc = -EINVAL; list_for_each_entry(res, &resources->iomem, list) { r = request_mem_region(res->start, res->end - res->start, @@ -511,6 +536,8 @@ err_unmap_iomem: break; release_mem_region(res->start, res->end - res->start); } +res_fini: + apei_resources_fini(&nvs_resources); return rc; } EXPORT_SYMBOL_GPL(apei_resources_request); From 8df0eb7c9d96f9e82f233ee8b74e0f0c8471f868 Mon Sep 17 00:00:00 2001 From: Kurt Garloff Date: Tue, 17 Jan 2012 04:18:02 -0500 Subject: [PATCH 463/820] ACPI: Store SRAT table revision In SRAT v1, we had 8bit proximity domain (PXM) fields; SRAT v2 provides 32bits for these. The new fields were reserved before. According to the ACPI spec, the OS must disregrard reserved fields. In order to know whether or not, we must know what version the SRAT table has. This patch stores the SRAT table revision for later consumption by arch specific __init functions. Signed-off-by: Kurt Garloff Signed-off-by: Len Brown --- drivers/acpi/numa.c | 6 ++++++ include/acpi/acpi_numa.h | 1 + 2 files changed, 7 insertions(+) diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c index 3b5c3189fd99..e56f3be7b07d 100644 --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -45,6 +45,8 @@ static int pxm_to_node_map[MAX_PXM_DOMAINS] static int node_to_pxm_map[MAX_NUMNODES] = { [0 ... MAX_NUMNODES - 1] = PXM_INVAL }; +unsigned char acpi_srat_revision __initdata; + int pxm_to_node(int pxm) { if (pxm < 0) @@ -255,9 +257,13 @@ acpi_parse_memory_affinity(struct acpi_subtable_header * header, static int __init acpi_parse_srat(struct acpi_table_header *table) { + struct acpi_table_srat *srat; if (!table) return -EINVAL; + srat = (struct acpi_table_srat *)table; + acpi_srat_revision = srat->header.revision; + /* Real work done in acpi_table_parse_srat below. */ return 0; diff --git a/include/acpi/acpi_numa.h b/include/acpi/acpi_numa.h index 173972672175..451823cb8837 100644 --- a/include/acpi/acpi_numa.h +++ b/include/acpi/acpi_numa.h @@ -15,6 +15,7 @@ extern int pxm_to_node(int); extern int node_to_pxm(int); extern void __acpi_map_pxm_to_node(int, int); extern int acpi_map_pxm_to_node(int); +extern unsigned char acpi_srat_revision; #endif /* CONFIG_ACPI_NUMA */ #endif /* __ACP_NUMA_H */ From cd298f60a2451a16e0f077404bf69b62ec868733 Mon Sep 17 00:00:00 2001 From: Kurt Garloff Date: Tue, 17 Jan 2012 04:20:31 -0500 Subject: [PATCH 464/820] ACPI, x86: Use SRAT table rev to use 8bit or 32bit PXM fields (x86/x86-64) In SRAT v1, we had 8bit proximity domain (PXM) fields; SRAT v2 provides 32bits for these. The new fields were reserved before. According to the ACPI spec, the OS must disregrard reserved fields. x86/x86-64 was rather inconsistent prior to this patch; it used 8 bits for the pxm field in cpu_affinity, but 32 bits in mem_affinity. This patch makes it consistent: Either use 8 bits consistently (SRAT rev 1 or lower) or 32 bits (SRAT rev 2 or higher). cc: x86@kernel.org Signed-off-by: Kurt Garloff Signed-off-by: Len Brown --- arch/x86/mm/srat.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/x86/mm/srat.c b/arch/x86/mm/srat.c index 81dbfdeb080d..7efd0c615d58 100644 --- a/arch/x86/mm/srat.c +++ b/arch/x86/mm/srat.c @@ -104,6 +104,8 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) if ((pa->flags & ACPI_SRAT_CPU_ENABLED) == 0) return; pxm = pa->proximity_domain_lo; + if (acpi_srat_revision >= 2) + pxm |= *((unsigned int*)pa->proximity_domain_hi) << 8; node = setup_node(pxm); if (node < 0) { printk(KERN_ERR "SRAT: Too many proximity domains %x\n", pxm); @@ -155,6 +157,8 @@ acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) start = ma->base_address; end = start + ma->length; pxm = ma->proximity_domain; + if (acpi_srat_revision <= 1) + pxm &= 0xff; node = setup_node(pxm); if (node < 0) { printk(KERN_ERR "SRAT: Too many proximity domains.\n"); From 9f10f6a520deb3639fac78d81151a3ade88b4e7f Mon Sep 17 00:00:00 2001 From: Kurt Garloff Date: Tue, 17 Jan 2012 04:21:49 -0500 Subject: [PATCH 465/820] ACPI, ia64: Use SRAT table rev to use 8bit or 16/32bit PXM fields (ia64) In SRAT v1, we had 8bit proximity domain (PXM) fields; SRAT v2 provides 32bits for these. The new fields were reserved before. According to the ACPI spec, the OS must disregrard reserved fields. ia64 did handle the PXM fields almost consistently, but depending on sgi's sn2 platform. This patch leaves the sn2 logic in, but does also use 16/32 bits for PXM if the SRAT has rev 2 or higher. The patch also adds __init to the two pxm accessor functions, as they access __initdata now and are called from an __init function only anyway. Note that the code only uses 16 bits for the PXM field in the processor proximity field; the patch does not address this as 16 bits are more than enough. Signed-off-by: Kurt Garloff Signed-off-by: Len Brown --- arch/ia64/kernel/acpi.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c index bfb4d01e0e51..5207035dc061 100644 --- a/arch/ia64/kernel/acpi.c +++ b/arch/ia64/kernel/acpi.c @@ -429,22 +429,24 @@ static u32 __devinitdata pxm_flag[PXM_FLAG_LEN]; static struct acpi_table_slit __initdata *slit_table; cpumask_t early_cpu_possible_map = CPU_MASK_NONE; -static int get_processor_proximity_domain(struct acpi_srat_cpu_affinity *pa) +static int __init +get_processor_proximity_domain(struct acpi_srat_cpu_affinity *pa) { int pxm; pxm = pa->proximity_domain_lo; - if (ia64_platform_is("sn2")) + if (ia64_platform_is("sn2") || acpi_srat_revision >= 2) pxm += pa->proximity_domain_hi[0] << 8; return pxm; } -static int get_memory_proximity_domain(struct acpi_srat_mem_affinity *ma) +static int __init +get_memory_proximity_domain(struct acpi_srat_mem_affinity *ma) { int pxm; pxm = ma->proximity_domain; - if (!ia64_platform_is("sn2")) + if (!ia64_platform_is("sn2") && acpi_srat_revision <= 1) pxm &= 0xff; return pxm; From bc9ffce27962c0c5fdc6adf74790ea0fcbe4a99c Mon Sep 17 00:00:00 2001 From: Myron Stowe Date: Mon, 7 Nov 2011 16:23:27 -0700 Subject: [PATCH 466/820] ACPI: Fix possible alignment issues with GAS 'address' references Generic Address Structures (GAS) may reside within ACPI tables which are byte aligned. This patch copies GAS 'address' references to a local variable, which will be naturally aligned, to be used going forward. ACPI Generic Address Structure (GAS) reference: ACPI Specification, Revision 4.0, Section 5.2.3.1, "Generic Address Structure" Signed-off-by: Myron Stowe Signed-off-by: Len Brown --- drivers/acpi/osl.c | 38 ++++++++++++++++++++++++-------------- 1 file changed, 24 insertions(+), 14 deletions(-) diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index f31c5c5f1b7e..2e285cdbefb1 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -166,17 +166,21 @@ static u32 acpi_osi_handler(acpi_string interface, u32 supported) return supported; } -static void __init acpi_request_region (struct acpi_generic_address *addr, +static void __init acpi_request_region (struct acpi_generic_address *gas, unsigned int length, char *desc) { - if (!addr->address || !length) + u64 addr; + + /* Handle possible alignment issues */ + memcpy(&addr, &gas->address, sizeof(addr)); + if (!addr || !length) return; /* Resources are never freed */ - if (addr->space_id == ACPI_ADR_SPACE_SYSTEM_IO) - request_region(addr->address, length, desc); - else if (addr->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) - request_mem_region(addr->address, length, desc); + if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_IO) + request_region(addr, length, desc); + else if (gas->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) + request_mem_region(addr, length, desc); } static int __init acpi_reserve_resources(void) @@ -427,35 +431,41 @@ void __init early_acpi_os_unmap_memory(void __iomem *virt, acpi_size size) __acpi_unmap_table(virt, size); } -static int acpi_os_map_generic_address(struct acpi_generic_address *addr) +static int acpi_os_map_generic_address(struct acpi_generic_address *gas) { + u64 addr; void __iomem *virt; - if (addr->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) + if (gas->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) return 0; - if (!addr->address || !addr->bit_width) + /* Handle possible alignment issues */ + memcpy(&addr, &gas->address, sizeof(addr)); + if (!addr || !gas->bit_width) return -EINVAL; - virt = acpi_os_map_memory(addr->address, addr->bit_width / 8); + virt = acpi_os_map_memory(addr, gas->bit_width / 8); if (!virt) return -EIO; return 0; } -static void acpi_os_unmap_generic_address(struct acpi_generic_address *addr) +static void acpi_os_unmap_generic_address(struct acpi_generic_address *gas) { + u64 addr; struct acpi_ioremap *map; - if (addr->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) + if (gas->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) return; - if (!addr->address || !addr->bit_width) + /* Handle possible alignment issues */ + memcpy(&addr, &gas->address, sizeof(addr)); + if (!addr || !gas->bit_width) return; mutex_lock(&acpi_ioremap_lock); - map = acpi_map_lookup(addr->address, addr->bit_width / 8); + map = acpi_map_lookup(addr, gas->bit_width / 8); if (!map) { mutex_unlock(&acpi_ioremap_lock); return; From 6f68c91c55ea3576d366797fa8d45e31c4aa79f8 Mon Sep 17 00:00:00 2001 From: Myron Stowe Date: Mon, 7 Nov 2011 16:23:34 -0700 Subject: [PATCH 467/820] ACPI: Export interfaces for ioremapping/iounmapping ACPI registers Export remapping and unmapping interfaces - acpi_os_map_generic_address() and acpi_os_unmap_generic_address() - for ACPI generic registers that are backed by memory mapped I/O (MMIO). The acpi_os_map_generic_address() and acpi_os_unmap_generic_address() declarations may more properly belong in include/acpi/acpiosxf.h next to acpi_os_read_memory() but I believe that would require the ACPI CA making them an official part of the ACPI CA - OS interface. ACPI Generic Address Structure (GAS) reference (ACPI's fixed/generic hardware registers use the GAS format): ACPI Specification, Revision 4.0, Section 5.2.3.1, "Generic Address Structure" Signed-off-by: Myron Stowe Acked-by: Rafael J. Wysocki Signed-off-by: Len Brown --- drivers/acpi/osl.c | 6 ++++-- include/linux/acpi_io.h | 3 +++ 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 2e285cdbefb1..b11f2676f7c9 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -431,7 +431,7 @@ void __init early_acpi_os_unmap_memory(void __iomem *virt, acpi_size size) __acpi_unmap_table(virt, size); } -static int acpi_os_map_generic_address(struct acpi_generic_address *gas) +int acpi_os_map_generic_address(struct acpi_generic_address *gas) { u64 addr; void __iomem *virt; @@ -450,8 +450,9 @@ static int acpi_os_map_generic_address(struct acpi_generic_address *gas) return 0; } +EXPORT_SYMBOL(acpi_os_map_generic_address); -static void acpi_os_unmap_generic_address(struct acpi_generic_address *gas) +void acpi_os_unmap_generic_address(struct acpi_generic_address *gas) { u64 addr; struct acpi_ioremap *map; @@ -475,6 +476,7 @@ static void acpi_os_unmap_generic_address(struct acpi_generic_address *gas) acpi_os_map_cleanup(map); } +EXPORT_SYMBOL(acpi_os_unmap_generic_address); #ifdef ACPI_FUTURE_USAGE acpi_status diff --git a/include/linux/acpi_io.h b/include/linux/acpi_io.h index 4afd7102459d..b0ffa219993e 100644 --- a/include/linux/acpi_io.h +++ b/include/linux/acpi_io.h @@ -12,4 +12,7 @@ static inline void __iomem *acpi_os_ioremap(acpi_physical_address phys, void __iomem *acpi_os_get_iomem(acpi_physical_address phys, unsigned int size); +int acpi_os_map_generic_address(struct acpi_generic_address *addr); +void acpi_os_unmap_generic_address(struct acpi_generic_address *addr); + #endif From 700130b41f4ee54520ac2ef2f7f1d072789711a4 Mon Sep 17 00:00:00 2001 From: Myron Stowe Date: Mon, 7 Nov 2011 16:23:41 -0700 Subject: [PATCH 468/820] ACPI APEI: Convert atomicio routines APEI needs memory access in interrupt context. The obvious choice is acpi_read(), but originally it couldn't be used in interrupt context because it makes temporary mappings with ioremap(). Therefore, we added drivers/acpi/atomicio.c, which provides: acpi_pre_map_gar() -- ioremap in process context acpi_atomic_read() -- memory access in interrupt context acpi_post_unmap_gar() -- iounmap Later we added acpi_os_map_generic_address() (2971852) and enhanced acpi_read() so it works in interrupt context as long as the address has been previously mapped (620242a). Now this sequence: acpi_os_map_generic_address() -- ioremap in process context acpi_read()/apei_read() -- now OK in interrupt context acpi_os_unmap_generic_address() is equivalent to what atomicio.c provides. This patch introduces apei_read() and apei_write(), which currently are functional equivalents of acpi_read() and acpi_write(). This is mainly proactive, to prevent APEI breakages if acpi_read() and acpi_write() are ever augmented to support the 'bit_offset' field of GAS, as APEI's __apei_exec_write_register() precludes splitting up functionality related to 'bit_offset' and APEI's 'mask' (see its APEI_EXEC_PRESERVE_REGISTER block). With apei_read() and apei_write() in place, usages of atomicio routines are converted to apei_read()/apei_write() and existing calls within osl.c and the CA, based on the re-factoring that was done in an earlier patch series - http://marc.info/?l=linux-acpi&m=128769263327206&w=2: acpi_pre_map_gar() --> acpi_os_map_generic_address() acpi_post_unmap_gar() --> acpi_os_unmap_generic_address() acpi_atomic_read() --> apei_read() acpi_atomic_write() --> apei_write() Note that acpi_read() and acpi_write() currently use 'bit_width' for accessing GARs which seems incorrect. 'bit_width' is the size of the register, while 'access_width' is the size of the access the processor must generate on the bus. The 'access_width' may be larger, for example, if the hardware only supports 32-bit or 64-bit reads. I wanted to minimize any possible impacts with this patch series so I did *not* change this behavior. Signed-off-by: Myron Stowe Signed-off-by: Len Brown --- drivers/acpi/apei/apei-base.c | 102 ++++++++++++++++++++++++++++-- drivers/acpi/apei/apei-internal.h | 3 + drivers/acpi/apei/ghes.c | 10 +-- 3 files changed, 104 insertions(+), 11 deletions(-) diff --git a/drivers/acpi/apei/apei-base.c b/drivers/acpi/apei/apei-base.c index 4abb6c74a938..e45350cb6ac8 100644 --- a/drivers/acpi/apei/apei-base.c +++ b/drivers/acpi/apei/apei-base.c @@ -34,13 +34,13 @@ #include #include #include +#include #include #include #include #include #include #include -#include #include "apei-internal.h" @@ -70,7 +70,7 @@ int __apei_exec_read_register(struct acpi_whea_header *entry, u64 *val) { int rc; - rc = acpi_atomic_read(val, &entry->register_region); + rc = apei_read(val, &entry->register_region); if (rc) return rc; *val >>= entry->register_region.bit_offset; @@ -116,13 +116,13 @@ int __apei_exec_write_register(struct acpi_whea_header *entry, u64 val) val <<= entry->register_region.bit_offset; if (entry->flags & APEI_EXEC_PRESERVE_REGISTER) { u64 valr = 0; - rc = acpi_atomic_read(&valr, &entry->register_region); + rc = apei_read(&valr, &entry->register_region); if (rc) return rc; valr &= ~(entry->mask << entry->register_region.bit_offset); val |= valr; } - rc = acpi_atomic_write(val, &entry->register_region); + rc = apei_write(val, &entry->register_region); return rc; } @@ -243,7 +243,7 @@ static int pre_map_gar_callback(struct apei_exec_context *ctx, u8 ins = entry->instruction; if (ctx->ins_table[ins].flags & APEI_EXEC_INS_ACCESS_REGISTER) - return acpi_pre_map_gar(&entry->register_region); + return acpi_os_map_generic_address(&entry->register_region); return 0; } @@ -276,7 +276,7 @@ static int post_unmap_gar_callback(struct apei_exec_context *ctx, u8 ins = entry->instruction; if (ctx->ins_table[ins].flags & APEI_EXEC_INS_ACCESS_REGISTER) - acpi_post_unmap_gar(&entry->register_region); + acpi_os_unmap_generic_address(&entry->register_region); return 0; } @@ -591,6 +591,96 @@ static int apei_check_gar(struct acpi_generic_address *reg, u64 *paddr) return 0; } +/* read GAR in interrupt (including NMI) or process context */ +int apei_read(u64 *val, struct acpi_generic_address *reg) +{ + int rc; + u64 address; + u32 tmp, width = reg->bit_width; + acpi_status status; + + rc = apei_check_gar(reg, &address); + if (rc) + return rc; + + if (width == 64) + width = 32; /* Break into two 32-bit transfers */ + + *val = 0; + switch(reg->space_id) { + case ACPI_ADR_SPACE_SYSTEM_MEMORY: + status = acpi_os_read_memory((acpi_physical_address) + address, &tmp, width); + if (ACPI_FAILURE(status)) + return -EIO; + *val = tmp; + + if (reg->bit_width == 64) { + /* Read the top 32 bits */ + status = acpi_os_read_memory((acpi_physical_address) + (address + 4), &tmp, 32); + if (ACPI_FAILURE(status)) + return -EIO; + *val |= ((u64)tmp << 32); + } + break; + case ACPI_ADR_SPACE_SYSTEM_IO: + status = acpi_os_read_port(address, (u32 *)val, reg->bit_width); + if (ACPI_FAILURE(status)) + return -EIO; + break; + default: + return -EINVAL; + } + + return 0; +} +EXPORT_SYMBOL_GPL(apei_read); + +/* write GAR in interrupt (including NMI) or process context */ +int apei_write(u64 val, struct acpi_generic_address *reg) +{ + int rc; + u64 address; + u32 width = reg->bit_width; + acpi_status status; + + rc = apei_check_gar(reg, &address); + if (rc) + return rc; + + if (width == 64) + width = 32; /* Break into two 32-bit transfers */ + + switch (reg->space_id) { + case ACPI_ADR_SPACE_SYSTEM_MEMORY: + status = acpi_os_write_memory((acpi_physical_address) + address, ACPI_LODWORD(val), + width); + if (ACPI_FAILURE(status)) + return -EIO; + + if (reg->bit_width == 64) { + status = acpi_os_write_memory((acpi_physical_address) + (address + 4), + ACPI_HIDWORD(val), 32); + if (ACPI_FAILURE(status)) + return -EIO; + } + break; + case ACPI_ADR_SPACE_SYSTEM_IO: + status = acpi_os_write_port(address, val, reg->bit_width); + if (ACPI_FAILURE(status)) + return -EIO; + break; + default: + return -EINVAL; + } + + return 0; +} +EXPORT_SYMBOL_GPL(apei_write); + static int collect_res_callback(struct apei_exec_context *ctx, struct acpi_whea_header *entry, void *data) diff --git a/drivers/acpi/apei/apei-internal.h b/drivers/acpi/apei/apei-internal.h index d778edd34fba..cca240a33038 100644 --- a/drivers/acpi/apei/apei-internal.h +++ b/drivers/acpi/apei/apei-internal.h @@ -68,6 +68,9 @@ static inline int apei_exec_run_optional(struct apei_exec_context *ctx, u8 actio /* IP has been set in instruction function */ #define APEI_EXEC_SET_IP 1 +int apei_read(u64 *val, struct acpi_generic_address *reg); +int apei_write(u64 val, struct acpi_generic_address *reg); + int __apei_exec_read_register(struct acpi_whea_header *entry, u64 *val); int __apei_exec_write_register(struct acpi_whea_header *entry, u64 val); int apei_exec_read_register(struct apei_exec_context *ctx, diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c index aaf36090de1e..b3207e16670e 100644 --- a/drivers/acpi/apei/ghes.c +++ b/drivers/acpi/apei/ghes.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include #include @@ -48,7 +49,6 @@ #include #include #include -#include #include #include #include @@ -301,7 +301,7 @@ static struct ghes *ghes_new(struct acpi_hest_generic *generic) if (!ghes) return ERR_PTR(-ENOMEM); ghes->generic = generic; - rc = acpi_pre_map_gar(&generic->error_status_address); + rc = acpi_os_map_generic_address(&generic->error_status_address); if (rc) goto err_free; error_block_length = generic->error_block_length; @@ -321,7 +321,7 @@ static struct ghes *ghes_new(struct acpi_hest_generic *generic) return ghes; err_unmap: - acpi_post_unmap_gar(&generic->error_status_address); + acpi_os_unmap_generic_address(&generic->error_status_address); err_free: kfree(ghes); return ERR_PTR(rc); @@ -330,7 +330,7 @@ err_free: static void ghes_fini(struct ghes *ghes) { kfree(ghes->estatus); - acpi_post_unmap_gar(&ghes->generic->error_status_address); + acpi_os_unmap_generic_address(&ghes->generic->error_status_address); } enum { @@ -401,7 +401,7 @@ static int ghes_read_estatus(struct ghes *ghes, int silent) u32 len; int rc; - rc = acpi_atomic_read(&buf_paddr, &g->error_status_address); + rc = apei_read(&buf_paddr, &g->error_status_address); if (rc) { if (!silent && printk_ratelimit()) pr_warning(FW_WARN GHES_PFX From 5ee71535440f034de1196b11f78cef81c4025c2b Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 16 Jan 2012 11:57:18 -0800 Subject: [PATCH 469/820] x86/kconfig: Move the ZONE_DMA entry under a menu Move the ZONE_DMA kconfig symbol under a menu item instead of having it listed before everything else in "make {xconfig | gconfig | nconfig | menuconfig}". This drops the first line of the top-level kernel config menu (in 3.2) below and moves it under "Processor type and features". [*] DMA memory allocation support General setup ---> [*] Enable loadable module support ---> [*] Enable the block layer ---> Processor type and features ---> Power management and ACPI options ---> Bus options (PCI etc.) ---> Executable file formats / Emulations ---> Signed-off-by: Randy Dunlap Acked-by: David Rientjes Cc: Linus Torvalds Cc: linux-mm@kvack.org Link: http://lkml.kernel.org/r/4F14811E.6090107@xenotime.net Signed-off-by: Ingo Molnar Cc: David Rientjes --- arch/x86/Kconfig | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 5731eb70e0a0..db190faffba1 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -120,16 +120,6 @@ config HAVE_LATENCYTOP_SUPPORT config MMU def_bool y -config ZONE_DMA - bool "DMA memory allocation support" if EXPERT - default y - help - DMA memory allocation support allows devices with less than 32-bit - addressing to allocate within the first 16MB of address space. - Disable if no such devices will be used. - - If unsure, say Y. - config SBUS bool @@ -253,6 +243,16 @@ source "kernel/Kconfig.freezer" menu "Processor type and features" +config ZONE_DMA + bool "DMA memory allocation support" if EXPERT + default y + help + DMA memory allocation support allows devices with less than 32-bit + addressing to allocate within the first 16MB of address space. + Disable if no such devices will be used. + + If unsure, say Y. + source "kernel/time/Kconfig" config SMP From 088067f4f14d6ee5c6a196b015a560cbe7744224 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Tue, 17 Jan 2012 10:39:05 +0100 Subject: [PATCH 470/820] netfilter: ipset: autoload set type modules safely Jan Engelhardt noticed when userspace requests a set type unknown to the kernel, it can lead to a loop due to the unsafe type module loading. The issue is fixed in this patch. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipset/ip_set_core.c | 36 ++++++++++++++++++++++--------- 1 file changed, 26 insertions(+), 10 deletions(-) diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index 86137b558f45..0f8e5f2fa1ba 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -77,35 +77,42 @@ find_set_type(const char *name, u8 family, u8 revision) } /* Unlock, try to load a set type module and lock again */ -static int -try_to_load_type(const char *name) +static bool +load_settype(const char *name) { nfnl_unlock(); pr_debug("try to load ip_set_%s\n", name); if (request_module("ip_set_%s", name) < 0) { pr_warning("Can't find ip_set type %s\n", name); nfnl_lock(); - return -IPSET_ERR_FIND_TYPE; + return false; } nfnl_lock(); - return -EAGAIN; + return true; } /* Find a set type and reference it */ +#define find_set_type_get(name, family, revision, found) \ + __find_set_type_get(name, family, revision, found, false) + static int -find_set_type_get(const char *name, u8 family, u8 revision, - struct ip_set_type **found) +__find_set_type_get(const char *name, u8 family, u8 revision, + struct ip_set_type **found, bool retry) { struct ip_set_type *type; int err; + if (retry && !load_settype(name)) + return -IPSET_ERR_FIND_TYPE; + rcu_read_lock(); *found = find_set_type(name, family, revision); if (*found) { err = !try_module_get((*found)->me) ? -EFAULT : 0; goto unlock; } - /* Make sure the type is loaded but we don't support the revision */ + /* Make sure the type is already loaded + * but we don't support the revision */ list_for_each_entry_rcu(type, &ip_set_type_list, list) if (STREQ(type->name, name)) { err = -IPSET_ERR_FIND_TYPE; @@ -113,7 +120,8 @@ find_set_type_get(const char *name, u8 family, u8 revision, } rcu_read_unlock(); - return try_to_load_type(name); + return retry ? -IPSET_ERR_FIND_TYPE : + __find_set_type_get(name, family, revision, found, true); unlock: rcu_read_unlock(); @@ -124,12 +132,19 @@ unlock: * If we succeeded, the supported minimal and maximum revisions are * filled out. */ +#define find_set_type_minmax(name, family, min, max) \ + __find_set_type_minmax(name, family, min, max, false) + static int -find_set_type_minmax(const char *name, u8 family, u8 *min, u8 *max) +__find_set_type_minmax(const char *name, u8 family, u8 *min, u8 *max, + bool retry) { struct ip_set_type *type; bool found = false; + if (retry && !load_settype(name)) + return -IPSET_ERR_FIND_TYPE; + *min = 255; *max = 0; rcu_read_lock(); list_for_each_entry_rcu(type, &ip_set_type_list, list) @@ -145,7 +160,8 @@ find_set_type_minmax(const char *name, u8 family, u8 *min, u8 *max) if (found) return 0; - return try_to_load_type(name); + return retry ? -IPSET_ERR_FIND_TYPE : + __find_set_type_minmax(name, family, min, max, true); } #define family_name(f) ((f) == AF_INET ? "inet" : \ From be94db9ddac0d6ea76f479b04fb3a43314654c9f Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Sat, 14 Jan 2012 16:42:13 +0100 Subject: [PATCH 471/820] netfilter: ipset: dumping error triggered removing references twice If there was a dumping error in the middle, the set-specific variable was not zeroed out and thus the 'done' function of the dumping wrongly tried to release the already released reference of the set. The already released reference was caught by __ip_set_put and triggered a kernel BUG message. Reported by Jean-Philippe Menil. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipset/ip_set_core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index 0f8e5f2fa1ba..32dbf0fa89db 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -1142,6 +1142,7 @@ release_refcount: if (ret || !cb->args[2]) { pr_debug("release set %s\n", ip_set_list[index]->name); ip_set_put_byindex(index); + cb->args[2] = 0; } out: if (nlh) { From 39a74fdedd1c1461d6fb6d330b5266886513c98f Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Tue, 10 Jan 2012 15:48:19 -0800 Subject: [PATCH 472/820] intel_idle: fix API misuse smp_call_function() only lets all other CPUs execute a specific function, while we expect all CPUs do in intel_idle. Without the fix, we could have one cpu which has auto_demotion enabled or has no broadcast timer setup. Usually we don't see impact because auto demotion just harms power and the intel_idle init is called in CPU 0, where boradcast timer delivers interrupt, but this still could be a problem. Cc: stable@vger.kernel.org Signed-off-by: Shaohua Li Signed-off-by: Andrew Morton Signed-off-by: Len Brown --- drivers/idle/intel_idle.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 5d2f8e13cf0e..af0b4a5d52f2 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -394,7 +394,7 @@ static int intel_idle_probe(void) if (boot_cpu_has(X86_FEATURE_ARAT)) /* Always Reliable APIC Timer */ lapic_timer_reliable_states = LAPIC_TIMER_ALWAYS_RELIABLE; else { - smp_call_function(__setup_broadcast_timer, (void *)true, 1); + on_each_cpu(__setup_broadcast_timer, (void *)true, 1); register_cpu_notifier(&setup_broadcast_notifier); } @@ -471,7 +471,7 @@ static int intel_idle_cpuidle_driver_init(void) } if (auto_demotion_disable_flags) - smp_call_function(auto_demotion_disable, NULL, 1); + on_each_cpu(auto_demotion_disable, NULL, 1); return 0; } @@ -568,7 +568,7 @@ static void __exit intel_idle_exit(void) cpuidle_unregister_driver(&intel_idle_driver); if (lapic_timer_reliable_states != LAPIC_TIMER_ALWAYS_RELIABLE) { - smp_call_function(__setup_broadcast_timer, (void *)false, 1); + on_each_cpu(__setup_broadcast_timer, (void *)false, 1); unregister_cpu_notifier(&setup_broadcast_notifier); } From d640113fe80e45ebd4a5b420b220d3f6bf37f682 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Tue, 13 Dec 2011 09:36:03 +0800 Subject: [PATCH 473/820] ACPI: processor: fix acpi_get_cpuid for UP processor For UP processor, it is likely that no _MAT method or MADT table defined. So currently acpi_get_cpuid(...) always return -1 for UP processor. This is wrong. It should return valid value for CPU0. In the other hand, BIOS may define multiple CPU handles even for UP processor, for example Scope (_PR) { Processor (CPU0, 0x00, 0x00000410, 0x06) {} Processor (CPU1, 0x01, 0x00000410, 0x06) {} Processor (CPU2, 0x02, 0x00000410, 0x06) {} Processor (CPU3, 0x03, 0x00000410, 0x06) {} } We should only return valid value for CPU0's acpi handle. And return invalid value for others. http://marc.info/?t=132329819900003&r=1&w=2 Cc: stable@vger.kernel.org Reported-and-tested-by: wallak@free.fr Signed-off-by: Lin Ming Signed-off-by: Len Brown --- drivers/acpi/processor_core.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-) diff --git a/drivers/acpi/processor_core.c b/drivers/acpi/processor_core.c index 3a0428e8435c..c850de4c9a14 100644 --- a/drivers/acpi/processor_core.c +++ b/drivers/acpi/processor_core.c @@ -173,8 +173,30 @@ int acpi_get_cpuid(acpi_handle handle, int type, u32 acpi_id) apic_id = map_mat_entry(handle, type, acpi_id); if (apic_id == -1) apic_id = map_madt_entry(type, acpi_id); - if (apic_id == -1) - return apic_id; + if (apic_id == -1) { + /* + * On UP processor, there is no _MAT or MADT table. + * So above apic_id is always set to -1. + * + * BIOS may define multiple CPU handles even for UP processor. + * For example, + * + * Scope (_PR) + * { + * Processor (CPU0, 0x00, 0x00000410, 0x06) {} + * Processor (CPU1, 0x01, 0x00000410, 0x06) {} + * Processor (CPU2, 0x02, 0x00000410, 0x06) {} + * Processor (CPU3, 0x03, 0x00000410, 0x06) {} + * } + * + * Ignores apic_id and always return 0 for CPU0's handle. + * Return -1 for other CPU's handle. + */ + if (acpi_id == 0) + return acpi_id; + else + return apic_id; + } #ifdef CONFIG_SMP for_each_possible_cpu(i) { From 3333ea7804201918aa241641cf8493a62b8ad527 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Thu, 17 Nov 2011 23:37:00 +0100 Subject: [PATCH 474/820] ACPI processor: Fix error path, also remove sysdev link Signed-off-by: Thomas Renninger Signed-off-by: Len Brown --- drivers/acpi/processor_driver.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index 9d7bc9f6b6cc..90719d125e74 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -539,6 +539,7 @@ err_thermal_unregister: thermal_cooling_device_unregister(pr->cdev); err_power_exit: acpi_processor_power_exit(pr, device); + sysfs_remove_link(&device->dev.kobj, "sysdev"); err_free_cpumask: free_cpumask_var(pr->throttling.shared_cpu_map); From a34e2a784f8d5b4d26fb3cc378e04296cdc7ad23 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Tue, 17 Jan 2012 09:45:48 -0200 Subject: [PATCH 475/820] [media] DocBook/dvbproperty.xml: Fix the units for DTV_FREQUENCY The units for DTV_FREQUENCY are kHz for satellital delivery systems (DVB-S/DVB-S2/DVB-TURBO/ISDB-S). Fix it at the API spec. Signed-off-by: Mauro Carvalho Chehab --- Documentation/DocBook/media/dvb/dvbproperty.xml | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/Documentation/DocBook/media/dvb/dvbproperty.xml b/Documentation/DocBook/media/dvb/dvbproperty.xml index ffee1fbbc001..049664192bb7 100644 --- a/Documentation/DocBook/media/dvb/dvbproperty.xml +++ b/Documentation/DocBook/media/dvb/dvbproperty.xml @@ -163,14 +163,16 @@ get/set up to 64 properties. The actual meaning of each property is described on
<constant>DTV_FREQUENCY</constant> - Central frequency of the channel, in HZ. + Central frequency of the channel. Notes: - 1)For ISDB-T, the channels are usually transmitted with an offset of 143kHz. + 1)For satellital delivery systems, it is measured in kHz. + For the other ones, it is measured in Hz. + 2)For ISDB-T, the channels are usually transmitted with an offset of 143kHz. E.g. a valid frequncy could be 474143 kHz. The stepping is bound to the bandwidth of the channel which is 6MHz. - 2)As in ISDB-Tsb the channel consists of only one or three segments the + 3)As in ISDB-Tsb the channel consists of only one or three segments the frequency step is 429kHz, 3*429 respectively. As for ISDB-T the central frequency of the channel is expected.
From b5dad9fdd8e2c102e1a0f8490c55dbda198d412c Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Tue, 17 Jan 2012 09:49:28 -0200 Subject: [PATCH 476/820] [media] DocBook/dvbproperty.xml: Fix ISDB-T delivery system parameters The ISDB-T differs on its way to implement the hierarchical transmissions: instead of using a low-priority/high-priority FEC codes, it does that by using different layers, each layer with their groups of segments. So, those parameters don't make sense for ISDB-T. Signed-off-by: Mauro Carvalho Chehab --- Documentation/DocBook/media/dvb/dvbproperty.xml | 3 --- 1 file changed, 3 deletions(-) diff --git a/Documentation/DocBook/media/dvb/dvbproperty.xml b/Documentation/DocBook/media/dvb/dvbproperty.xml index 049664192bb7..2de2bb326906 100644 --- a/Documentation/DocBook/media/dvb/dvbproperty.xml +++ b/Documentation/DocBook/media/dvb/dvbproperty.xml @@ -740,11 +740,8 @@ typedef enum fe_hierarchy { DTV_MODULATION DTV_BANDWIDTH_HZ DTV_INVERSION - DTV_CODE_RATE_HP - DTV_CODE_RATE_LP DTV_GUARD_INTERVAL DTV_TRANSMISSION_MODE - DTV_HIERARCHY DTV_ISDBT_LAYER_ENABLED DTV_ISDBT_PARTIAL_RECEPTION DTV_ISDBT_SOUND_BROADCASTING From 7bb0f088f8dd1d60b8f5743471cc3db3f820df59 Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Tue, 17 Jan 2012 10:00:41 -0200 Subject: [PATCH 477/820] [media] DocBook/dvbproperty.xml: Remove DTV_MODULATION from ISDB-T On ISDB-T, each layer can have its own independent modulation, applied to the carriers that belong to the segments associated with them. So, there's no sense to define a global modulation parameter. Signed-off-by: Mauro Carvalho Chehab --- Documentation/DocBook/media/dvb/dvbproperty.xml | 1 - 1 file changed, 1 deletion(-) diff --git a/Documentation/DocBook/media/dvb/dvbproperty.xml b/Documentation/DocBook/media/dvb/dvbproperty.xml index 2de2bb326906..c7a4ca517859 100644 --- a/Documentation/DocBook/media/dvb/dvbproperty.xml +++ b/Documentation/DocBook/media/dvb/dvbproperty.xml @@ -737,7 +737,6 @@ typedef enum fe_hierarchy { DTV_TUNE DTV_CLEAR DTV_FREQUENCY - DTV_MODULATION DTV_BANDWIDTH_HZ DTV_INVERSION DTV_GUARD_INTERVAL From 5d2031f2395728c765cf2bd7c1ebb6d51c4fe076 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 13 Jan 2012 23:58:39 +0100 Subject: [PATCH 478/820] bcma: convert suspend/resume to pm_ops MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit .. and avoid doing the unnecessary PCI operations - the PCI layer will do them for us. Signed-off-by: Linus Torvalds Signed-off-by: Rafał Miłecki Signed-off-by: John W. Linville --- drivers/bcma/host_pci.c | 44 ++++++++++++++++------------------------- 1 file changed, 17 insertions(+), 27 deletions(-) diff --git a/drivers/bcma/host_pci.c b/drivers/bcma/host_pci.c index c1ca9e3ada5c..86dfeea77e1a 100644 --- a/drivers/bcma/host_pci.c +++ b/drivers/bcma/host_pci.c @@ -235,41 +235,32 @@ static void bcma_host_pci_remove(struct pci_dev *dev) } #ifdef CONFIG_PM -static int bcma_host_pci_suspend(struct pci_dev *dev, pm_message_t state) +static int bcma_host_pci_suspend(struct device *dev) { - struct bcma_bus *bus = pci_get_drvdata(dev); - - /* Host specific */ - pci_save_state(dev); - pci_disable_device(dev); - pci_set_power_state(dev, pci_choose_state(dev, state)); + struct pci_dev *pdev = to_pci_dev(dev); + struct bcma_bus *bus = pci_get_drvdata(pdev); bus->mapped_core = NULL; + return 0; } -static int bcma_host_pci_resume(struct pci_dev *dev) +static int bcma_host_pci_resume(struct device *dev) { - struct bcma_bus *bus = pci_get_drvdata(dev); - int err; + struct pci_dev *pdev = to_pci_dev(dev); + struct bcma_bus *bus = pci_get_drvdata(pdev); - /* Host specific */ - pci_set_power_state(dev, 0); - err = pci_enable_device(dev); - if (err) - return err; - pci_restore_state(dev); - - /* Bus specific */ - err = bcma_bus_resume(bus); - if (err) - return err; - - return 0; + return bcma_bus_resume(bus); } + +static SIMPLE_DEV_PM_OPS(bcma_pm_ops, bcma_host_pci_suspend, + bcma_host_pci_resume); +#define BCMA_PM_OPS (&bcma_pm_ops) + #else /* CONFIG_PM */ -# define bcma_host_pci_suspend NULL -# define bcma_host_pci_resume NULL + +#define BCMA_PM_OPS NULL + #endif /* CONFIG_PM */ static DEFINE_PCI_DEVICE_TABLE(bcma_pci_bridge_tbl) = { @@ -287,8 +278,7 @@ static struct pci_driver bcma_pci_bridge_driver = { .id_table = bcma_pci_bridge_tbl, .probe = bcma_host_pci_probe, .remove = bcma_host_pci_remove, - .suspend = bcma_host_pci_suspend, - .resume = bcma_host_pci_resume, + .driver.pm = BCMA_PM_OPS, }; int __init bcma_host_pci_init(void) From 685a4ef0df23b97eea1842c2698da9a2e35fc976 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 13 Jan 2012 23:58:40 +0100 Subject: [PATCH 479/820] bcma: add stub for bcma_bus_suspend() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit .. and connect it up with the pci host bcma driver. Now, the next step is to connect those bcma bus-level suspend/resume functions to the actual bcma device suspend resume functions. Signed-off-by: Linus Torvalds Signed-off-by: Rafał Miłecki Signed-off-by: John W. Linville --- drivers/bcma/bcma_private.h | 1 + drivers/bcma/host_pci.c | 2 +- drivers/bcma/main.c | 5 +++++ 3 files changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/bcma/bcma_private.h b/drivers/bcma/bcma_private.h index fda56bde36b8..0def898a1d15 100644 --- a/drivers/bcma/bcma_private.h +++ b/drivers/bcma/bcma_private.h @@ -19,6 +19,7 @@ int __init bcma_bus_early_register(struct bcma_bus *bus, struct bcma_device *core_cc, struct bcma_device *core_mips); #ifdef CONFIG_PM +int bcma_bus_suspend(struct bcma_bus *bus); int bcma_bus_resume(struct bcma_bus *bus); #endif diff --git a/drivers/bcma/host_pci.c b/drivers/bcma/host_pci.c index 86dfeea77e1a..f59244e33971 100644 --- a/drivers/bcma/host_pci.c +++ b/drivers/bcma/host_pci.c @@ -242,7 +242,7 @@ static int bcma_host_pci_suspend(struct device *dev) bus->mapped_core = NULL; - return 0; + return bcma_bus_suspend(bus); } static int bcma_host_pci_resume(struct device *dev) diff --git a/drivers/bcma/main.c b/drivers/bcma/main.c index 10f92b371e58..b711d9d634a7 100644 --- a/drivers/bcma/main.c +++ b/drivers/bcma/main.c @@ -241,6 +241,11 @@ int __init bcma_bus_early_register(struct bcma_bus *bus, } #ifdef CONFIG_PM +int bcma_bus_suspend(struct bcma_bus *bus) +{ + return 0; +} + int bcma_bus_resume(struct bcma_bus *bus) { struct bcma_device *core; From 7d5869e78f4c9d32f834dadefbb7dcb3c9d4d85f Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 13 Jan 2012 23:58:41 +0100 Subject: [PATCH 480/820] bcma: connect the bcma bus suspend/resume to the bcma driver suspend/resume Now the low-level driver actually gets informed that it is getting suspended and resumed. Signed-off-by: Linus Torvalds Signed-off-by: John W. Linville --- drivers/bcma/main.c | 19 +++++++++++++++++++ .../wireless/brcm80211/brcmsmac/mac80211_if.c | 2 +- include/linux/bcma/bcma.h | 2 +- 3 files changed, 21 insertions(+), 2 deletions(-) diff --git a/drivers/bcma/main.c b/drivers/bcma/main.c index b711d9d634a7..febbc0a1222a 100644 --- a/drivers/bcma/main.c +++ b/drivers/bcma/main.c @@ -243,6 +243,16 @@ int __init bcma_bus_early_register(struct bcma_bus *bus, #ifdef CONFIG_PM int bcma_bus_suspend(struct bcma_bus *bus) { + struct bcma_device *core; + + list_for_each_entry(core, &bus->cores, list) { + struct device_driver *drv = core->dev.driver; + if (drv) { + struct bcma_driver *adrv = container_of(drv, struct bcma_driver, drv); + if (adrv->suspend) + adrv->suspend(core); + } + } return 0; } @@ -257,6 +267,15 @@ int bcma_bus_resume(struct bcma_bus *bus) bcma_core_chipcommon_init(&bus->drv_cc); } + list_for_each_entry(core, &bus->cores, list) { + struct device_driver *drv = core->dev.driver; + if (drv) { + struct bcma_driver *adrv = container_of(drv, struct bcma_driver, drv); + if (adrv->resume) + adrv->resume(core); + } + } + return 0; } #endif diff --git a/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c b/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c index d106576ce338..213130afdaf7 100644 --- a/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c +++ b/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c @@ -1135,7 +1135,7 @@ static int brcms_pci_suspend(struct pci_dev *pdev) return pci_set_power_state(pdev, PCI_D3hot); } -static int brcms_suspend(struct bcma_device *pdev, pm_message_t state) +static int brcms_suspend(struct bcma_device *pdev) { struct brcms_info *wl; struct ieee80211_hw *hw; diff --git a/include/linux/bcma/bcma.h b/include/linux/bcma/bcma.h index f4b8346b1a33..83c209f39493 100644 --- a/include/linux/bcma/bcma.h +++ b/include/linux/bcma/bcma.h @@ -162,7 +162,7 @@ struct bcma_driver { int (*probe)(struct bcma_device *dev); void (*remove)(struct bcma_device *dev); - int (*suspend)(struct bcma_device *dev, pm_message_t state); + int (*suspend)(struct bcma_device *dev); int (*resume)(struct bcma_device *dev); void (*shutdown)(struct bcma_device *dev); From 7e9e7fa414e47fccc6127f23bc866cbcc7e27dcb Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 13 Jan 2012 23:58:42 +0100 Subject: [PATCH 481/820] brcmsmac: remove PCI suspend/resume from bcma driver The brcmsmac driver isn't a PCI driver any more, it's a bcma one. The PCI device has been resumed by the PCI driver (the generic PCI layer, really), we should be resuming just our own driver state. Also add pr_debug() calls to show that we now actually get the suspend/resume events. Signed-off-by: Linus Torvalds Signed-off-by: John W. Linville --- .../wireless/brcm80211/brcmsmac/mac80211_if.c | 38 ++----------------- 1 file changed, 3 insertions(+), 35 deletions(-) diff --git a/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c b/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c index 213130afdaf7..448ab9c4eb47 100644 --- a/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c +++ b/drivers/net/wireless/brcm80211/brcmsmac/mac80211_if.c @@ -1128,13 +1128,6 @@ static int __devinit brcms_bcma_probe(struct bcma_device *pdev) return 0; } -static int brcms_pci_suspend(struct pci_dev *pdev) -{ - pci_save_state(pdev); - pci_disable_device(pdev); - return pci_set_power_state(pdev, PCI_D3hot); -} - static int brcms_suspend(struct bcma_device *pdev) { struct brcms_info *wl; @@ -1153,40 +1146,15 @@ static int brcms_suspend(struct bcma_device *pdev) wl->pub->hw_up = false; spin_unlock_bh(&wl->lock); - /* temporarily do suspend ourselves */ - return brcms_pci_suspend(pdev->bus->host_pci); -} - -static int brcms_pci_resume(struct pci_dev *pdev) -{ - int err = 0; - uint val; - - err = pci_set_power_state(pdev, PCI_D0); - if (err) - return err; - - pci_restore_state(pdev); - - err = pci_enable_device(pdev); - if (err) - return err; - - pci_set_master(pdev); - - pci_read_config_dword(pdev, 0x40, &val); - if ((val & 0x0000ff00) != 0) - pci_write_config_dword(pdev, 0x40, val & 0xffff00ff); + pr_debug("brcms_suspend ok\n"); return 0; } static int brcms_resume(struct bcma_device *pdev) { - /* - * just do pci resume for now until bcma supports it. - */ - return brcms_pci_resume(pdev->bus->host_pci); + pr_debug("brcms_resume ok\n"); + return 0; } static struct bcma_driver brcms_bcma_driver = { From 1886e5d2c694e7fb59434c717e704e7fd8475d2e Mon Sep 17 00:00:00 2001 From: Thadeu Lima de Souza Cascardo Date: Fri, 13 Jan 2012 08:06:32 +0000 Subject: [PATCH 482/820] ehea: make some functions and variables static Some functions and variables in ehea are only used in their own file, so they should be static. One particular function had a very generic name, print_error_data. Signed-off-by: Thadeu Lima de Souza Cascardo Signed-off-by: David S. Miller --- drivers/net/ethernet/ibm/ehea/ehea_ethtool.c | 2 +- drivers/net/ethernet/ibm/ehea/ehea_main.c | 32 ++++++++++---------- drivers/net/ethernet/ibm/ehea/ehea_qmr.c | 14 ++++----- 3 files changed, 23 insertions(+), 25 deletions(-) diff --git a/drivers/net/ethernet/ibm/ehea/ehea_ethtool.c b/drivers/net/ethernet/ibm/ehea/ehea_ethtool.c index 05b7359bde8d..6bdd8e36e564 100644 --- a/drivers/net/ethernet/ibm/ehea/ehea_ethtool.c +++ b/drivers/net/ethernet/ibm/ehea/ehea_ethtool.c @@ -263,7 +263,7 @@ static void ehea_get_ethtool_stats(struct net_device *dev, data[i++] = atomic_read(&port->port_res[k].swqe_avail); } -const struct ethtool_ops ehea_ethtool_ops = { +static const struct ethtool_ops ehea_ethtool_ops = { .get_settings = ehea_get_settings, .get_drvinfo = ehea_get_drvinfo, .get_msglevel = ehea_get_msglevel, diff --git a/drivers/net/ethernet/ibm/ehea/ehea_main.c b/drivers/net/ethernet/ibm/ehea/ehea_main.c index 3554414eb5e2..5d5fb2627184 100644 --- a/drivers/net/ethernet/ibm/ehea/ehea_main.c +++ b/drivers/net/ethernet/ibm/ehea/ehea_main.c @@ -94,8 +94,8 @@ static int port_name_cnt; static LIST_HEAD(adapter_list); static unsigned long ehea_driver_flags; static DEFINE_MUTEX(dlpar_mem_lock); -struct ehea_fw_handle_array ehea_fw_handles; -struct ehea_bcmc_reg_array ehea_bcmc_regs; +static struct ehea_fw_handle_array ehea_fw_handles; +static struct ehea_bcmc_reg_array ehea_bcmc_regs; static int __devinit ehea_probe_adapter(struct platform_device *dev, @@ -133,7 +133,7 @@ void ehea_dump(void *adr, int len, char *msg) } } -void ehea_schedule_port_reset(struct ehea_port *port) +static void ehea_schedule_port_reset(struct ehea_port *port) { if (!test_bit(__EHEA_DISABLE_PORT_RESET, &port->flags)) schedule_work(&port->reset_task); @@ -1404,7 +1404,7 @@ out: return ret; } -int ehea_gen_smrs(struct ehea_port_res *pr) +static int ehea_gen_smrs(struct ehea_port_res *pr) { int ret; struct ehea_adapter *adapter = pr->port->adapter; @@ -1426,7 +1426,7 @@ out: return -EIO; } -int ehea_rem_smrs(struct ehea_port_res *pr) +static int ehea_rem_smrs(struct ehea_port_res *pr) { if ((ehea_rem_mr(&pr->send_mr)) || (ehea_rem_mr(&pr->recv_mr))) @@ -2190,7 +2190,7 @@ out: return err; } -int ehea_activate_qp(struct ehea_adapter *adapter, struct ehea_qp *qp) +static int ehea_activate_qp(struct ehea_adapter *adapter, struct ehea_qp *qp) { int ret = -EIO; u64 hret; @@ -2531,7 +2531,7 @@ static void ehea_flush_sq(struct ehea_port *port) } } -int ehea_stop_qps(struct net_device *dev) +static int ehea_stop_qps(struct net_device *dev) { struct ehea_port *port = netdev_priv(dev); struct ehea_adapter *adapter = port->adapter; @@ -2600,7 +2600,7 @@ out: return ret; } -void ehea_update_rqs(struct ehea_qp *orig_qp, struct ehea_port_res *pr) +static void ehea_update_rqs(struct ehea_qp *orig_qp, struct ehea_port_res *pr) { struct ehea_qp qp = *orig_qp; struct ehea_qp_init_attr *init_attr = &qp.init_attr; @@ -2633,7 +2633,7 @@ void ehea_update_rqs(struct ehea_qp *orig_qp, struct ehea_port_res *pr) } } -int ehea_restart_qps(struct net_device *dev) +static int ehea_restart_qps(struct net_device *dev) { struct ehea_port *port = netdev_priv(dev); struct ehea_adapter *adapter = port->adapter; @@ -2824,7 +2824,7 @@ static void ehea_tx_watchdog(struct net_device *dev) ehea_schedule_port_reset(port); } -int ehea_sense_adapter_attr(struct ehea_adapter *adapter) +static int ehea_sense_adapter_attr(struct ehea_adapter *adapter) { struct hcp_query_ehea *cb; u64 hret; @@ -2852,7 +2852,7 @@ out: return ret; } -int ehea_get_jumboframe_status(struct ehea_port *port, int *jumbo) +static int ehea_get_jumboframe_status(struct ehea_port *port, int *jumbo) { struct hcp_ehea_port_cb4 *cb4; u64 hret; @@ -2966,7 +2966,7 @@ static const struct net_device_ops ehea_netdev_ops = { .ndo_tx_timeout = ehea_tx_watchdog, }; -struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter, +static struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter, u32 logical_port_id, struct device_node *dn) { @@ -3237,7 +3237,7 @@ static ssize_t ehea_remove_port(struct device *dev, static DEVICE_ATTR(probe_port, S_IWUSR, NULL, ehea_probe_port); static DEVICE_ATTR(remove_port, S_IWUSR, NULL, ehea_remove_port); -int ehea_create_device_sysfs(struct platform_device *dev) +static int ehea_create_device_sysfs(struct platform_device *dev) { int ret = device_create_file(&dev->dev, &dev_attr_probe_port); if (ret) @@ -3248,7 +3248,7 @@ out: return ret; } -void ehea_remove_device_sysfs(struct platform_device *dev) +static void ehea_remove_device_sysfs(struct platform_device *dev) { device_remove_file(&dev->dev, &dev_attr_probe_port); device_remove_file(&dev->dev, &dev_attr_remove_port); @@ -3379,7 +3379,7 @@ static int __devexit ehea_remove(struct platform_device *dev) return 0; } -void ehea_crash_handler(void) +static void ehea_crash_handler(void) { int i; @@ -3491,7 +3491,7 @@ static ssize_t ehea_show_capabilities(struct device_driver *drv, static DRIVER_ATTR(capabilities, S_IRUSR | S_IRGRP | S_IROTH, ehea_show_capabilities, NULL); -int __init ehea_module_init(void) +static int __init ehea_module_init(void) { int ret; diff --git a/drivers/net/ethernet/ibm/ehea/ehea_qmr.c b/drivers/net/ethernet/ibm/ehea/ehea_qmr.c index 95b9f4fa811e..c25b05b94daa 100644 --- a/drivers/net/ethernet/ibm/ehea/ehea_qmr.c +++ b/drivers/net/ethernet/ibm/ehea/ehea_qmr.c @@ -34,9 +34,7 @@ #include "ehea_phyp.h" #include "ehea_qmr.h" -struct ehea_bmap *ehea_bmap = NULL; - - +static struct ehea_bmap *ehea_bmap; static void *hw_qpageit_get_inc(struct hw_queue *queue) { @@ -212,7 +210,7 @@ out_nomem: return NULL; } -u64 ehea_destroy_cq_res(struct ehea_cq *cq, u64 force) +static u64 ehea_destroy_cq_res(struct ehea_cq *cq, u64 force) { u64 hret; u64 adapter_handle = cq->adapter->handle; @@ -337,7 +335,7 @@ struct ehea_eqe *ehea_poll_eq(struct ehea_eq *eq) return eqe; } -u64 ehea_destroy_eq_res(struct ehea_eq *eq, u64 force) +static u64 ehea_destroy_eq_res(struct ehea_eq *eq, u64 force) { u64 hret; unsigned long flags; @@ -381,7 +379,7 @@ int ehea_destroy_eq(struct ehea_eq *eq) /** * allocates memory for a queue and registers pages in phyp */ -int ehea_qp_alloc_register(struct ehea_qp *qp, struct hw_queue *hw_queue, +static int ehea_qp_alloc_register(struct ehea_qp *qp, struct hw_queue *hw_queue, int nr_pages, int wqe_size, int act_nr_sges, struct ehea_adapter *adapter, int h_call_q_selector) { @@ -516,7 +514,7 @@ out_freemem: return NULL; } -u64 ehea_destroy_qp_res(struct ehea_qp *qp, u64 force) +static u64 ehea_destroy_qp_res(struct ehea_qp *qp, u64 force) { u64 hret; struct ehea_qp_init_attr *qp_attr = &qp->init_attr; @@ -976,7 +974,7 @@ int ehea_gen_smr(struct ehea_adapter *adapter, struct ehea_mr *old_mr, return 0; } -void print_error_data(u64 *data) +static void print_error_data(u64 *data) { int length; u64 type = EHEA_BMASK_GET(ERROR_DATA_TYPE, data[2]); From 795d9a2538b205d9329f34798ec96503a07b7919 Mon Sep 17 00:00:00 2001 From: Hiroaki SHIMODA Date: Sat, 14 Jan 2012 07:10:21 +0000 Subject: [PATCH 483/820] bql: Fix inconsistency between file mode and attr method. There is no store() method for inflight attribute in the tx-/byte_queue_limits sysfs directory. So remove S_IWUSR bit. Signed-off-by: Hiroaki SHIMODA Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/net-sysfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index f3dbd4f596a4..a1727cda03d7 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -929,7 +929,7 @@ static ssize_t bql_show_inflight(struct netdev_queue *queue, } static struct netdev_queue_attribute bql_inflight_attribute = - __ATTR(inflight, S_IRUGO | S_IWUSR, bql_show_inflight, NULL); + __ATTR(inflight, S_IRUGO, bql_show_inflight, NULL); #define BQL_ATTR(NAME, FIELD) \ static ssize_t bql_show_ ## NAME(struct netdev_queue *queue, \ From 8d9d399f14ea65aeb50c7404e986bceede93bb99 Mon Sep 17 00:00:00 2001 From: Devendra Naga Date: Sat, 14 Jan 2012 08:16:21 +0000 Subject: [PATCH 484/820] net: remove version.h includes in net/openvswitch/ remove version.h includes in net/openswitch/ as reported by make versioncheck. Signed-off-by: Devendra Naga Signed-off-by: David S. Miller --- net/openvswitch/datapath.c | 1 - net/openvswitch/datapath.h | 1 - net/openvswitch/vport-internal_dev.c | 1 - net/openvswitch/vport.c | 1 - 4 files changed, 4 deletions(-) diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c index 9a2725114e99..d9d01cbf0fcb 100644 --- a/net/openvswitch/datapath.c +++ b/net/openvswitch/datapath.c @@ -36,7 +36,6 @@ #include #include #include -#include #include #include #include diff --git a/net/openvswitch/datapath.h b/net/openvswitch/datapath.h index 5b9f884b7055..c73370cc1f02 100644 --- a/net/openvswitch/datapath.h +++ b/net/openvswitch/datapath.h @@ -25,7 +25,6 @@ #include #include #include -#include #include "flow.h" diff --git a/net/openvswitch/vport-internal_dev.c b/net/openvswitch/vport-internal_dev.c index 8fc28b86f2b3..322b8d206693 100644 --- a/net/openvswitch/vport-internal_dev.c +++ b/net/openvswitch/vport-internal_dev.c @@ -23,7 +23,6 @@ #include #include #include -#include #include "datapath.h" #include "vport-internal_dev.h" diff --git a/net/openvswitch/vport.c b/net/openvswitch/vport.c index 7f0ef3794c51..6c066ba25dc7 100644 --- a/net/openvswitch/vport.c +++ b/net/openvswitch/vport.c @@ -27,7 +27,6 @@ #include #include #include -#include #include "vport.h" #include "vport-internal_dev.h" From 5addc0de28f5e286f9d121112c4222250807b5a5 Mon Sep 17 00:00:00 2001 From: Alexandre Oliva Date: Mon, 16 Jan 2012 14:00:12 -0500 Subject: [PATCH 485/820] brcmfmac: work-around gcc 4.7 build issue Alexandre Oliva says: "It's an issue brought about by GCC 4.7's partial-inlining, that ends up splitting the udelay function just at the wrong spot, in such a way that some sanity checks for constants fails, and we end up calling bad_udelay. This patch fixes the problem. Feel free to push it upstream if it makes sense to you." Signed-off-by: John W. Linville --- drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c b/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c index f23b0c3e4ea3..bf11850a20f1 100644 --- a/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c +++ b/drivers/net/wireless/brcm80211/brcmfmac/wl_cfg80211.c @@ -2475,7 +2475,7 @@ static s32 brcmf_init_iscan(struct brcmf_cfg80211_priv *cfg_priv) return err; } -static void brcmf_delay(u32 ms) +static __always_inline void brcmf_delay(u32 ms) { if (ms < 1000 / HZ) { cond_resched(); From d981e05908f85beb68ae0d935e59c5e399d2136e Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 17 Jan 2012 10:33:31 +0300 Subject: [PATCH 486/820] mwl8k: fix condition in mwl8k_cmd_encryption_remove_key() The intent here was to check whether key->cipher was WEP40 or WEP104. We do a similar check correctly in several other places in this file. The current condition is always true. Signed-off-by: Dan Carpenter Signed-off-by: John W. Linville --- drivers/net/wireless/mwl8k.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/mwl8k.c b/drivers/net/wireless/mwl8k.c index 7becea3dec65..d187ea24798e 100644 --- a/drivers/net/wireless/mwl8k.c +++ b/drivers/net/wireless/mwl8k.c @@ -4066,7 +4066,7 @@ static int mwl8k_cmd_encryption_remove_key(struct ieee80211_hw *hw, goto done; if (key->cipher == WLAN_CIPHER_SUITE_WEP40 || - WLAN_CIPHER_SUITE_WEP104) + key->cipher == WLAN_CIPHER_SUITE_WEP104) mwl8k_vif->wep_key_conf[key->keyidx].enabled = 0; cmd->action = cpu_to_le32(MWL8K_ENCR_REMOVE_KEY); From 544204733a20200951c271f81c74930ba5424638 Mon Sep 17 00:00:00 2001 From: Helmut Schaa Date: Tue, 17 Jan 2012 09:22:49 +0100 Subject: [PATCH 487/820] mac80211: Fix possible race between sta_unblock and network softirq All other code paths in sta_unblock synchronize with the network softirq by using local_bh_disable/enable. Do the same around ieee80211_sta_ps_deliver_wakeup. Signed-off-by: Helmut Schaa Signed-off-by: John W. Linville --- net/mac80211/sta_info.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index 3c428d4839c7..ff11f6bf8266 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -238,9 +238,11 @@ static void sta_unblock(struct work_struct *wk) if (sta->dead) return; - if (!test_sta_flag(sta, WLAN_STA_PS_STA)) + if (!test_sta_flag(sta, WLAN_STA_PS_STA)) { + local_bh_disable(); ieee80211_sta_ps_deliver_wakeup(sta); - else if (test_and_clear_sta_flag(sta, WLAN_STA_PSPOLL)) { + local_bh_enable(); + } else if (test_and_clear_sta_flag(sta, WLAN_STA_PSPOLL)) { clear_sta_flag(sta, WLAN_STA_PS_DRIVER); local_bh_disable(); From 604c4ef1c453a1b2ea2cdf04d2b49afec421ebfa Mon Sep 17 00:00:00 2001 From: Yogesh Ashok Powar Date: Tue, 17 Jan 2012 15:45:15 +0530 Subject: [PATCH 488/820] mwl8k: Fixing Sparse ENDIAN CHECK warning Fixing following sparse warning >drivers/net/wireless/mwl8k.c:2780:15: warning: incorrect type in assignment (different base types) >drivers/net/wireless/mwl8k.c:2780:15: expected restricted unsigned short [usertype] channel >drivers/net/wireless/mwl8k.c:2780:15: got unsigned short [unsigned] [usertype] hw_value Signed-off-by: Yogesh Ashok Powar Signed-off-by: John W. Linville --- drivers/net/wireless/mwl8k.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wireless/mwl8k.c b/drivers/net/wireless/mwl8k.c index d187ea24798e..dd5aeaff44ba 100644 --- a/drivers/net/wireless/mwl8k.c +++ b/drivers/net/wireless/mwl8k.c @@ -2777,7 +2777,7 @@ static int mwl8k_cmd_tx_power(struct ieee80211_hw *hw, else if (channel->band == IEEE80211_BAND_5GHZ) cmd->band = cpu_to_le16(0x4); - cmd->channel = channel->hw_value; + cmd->channel = cpu_to_le16(channel->hw_value); if (conf->channel_type == NL80211_CHAN_NO_HT || conf->channel_type == NL80211_CHAN_HT20) { From 319d3b9c97b5e3191e419bb95496bf08ec50f096 Mon Sep 17 00:00:00 2001 From: Glauber Costa Date: Sun, 15 Jan 2012 22:04:39 +0000 Subject: [PATCH 489/820] net: move sock_update_memcg outside of CONFIG_INET Although only used currently for tcp sockets, this function is now used in common sock code (for sock_clone()) Commit 475f1b52645a29936b9df1d8fcd45f7e56bd4a9f moved the declaration of sock_update_clone() to inside sock.c, but this only fixes the problem when CONFIG_CGROUP_MEM_RES_CTLR_KMEM is also not defined. This patch here is verified to fix both problems, although reverting the previous one is not necessary. Signed-off-by: Glauber Costa CC: David S. Miller CC: Stephen Rothwell Reported-by: Randy Dunlap Acked-by: Randy Dunlap Signed-off-by: David S. Miller --- mm/memcontrol.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 602207be9853..3dbff4dcde35 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -373,7 +373,6 @@ static void mem_cgroup_put(struct mem_cgroup *memcg); /* Writing them here to avoid exposing memcg's inner layout */ #ifdef CONFIG_CGROUP_MEM_RES_CTLR_KMEM -#ifdef CONFIG_INET #include #include @@ -420,6 +419,7 @@ void sock_release_memcg(struct sock *sk) } } +#ifdef CONFIG_INET struct cg_proto *tcp_proto_cgroup(struct mem_cgroup *memcg) { if (!memcg || mem_cgroup_is_root(memcg)) From 27a429383b399fc0f68bea55c422d7179f3c8b05 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 16 Jan 2012 04:35:50 +0000 Subject: [PATCH 490/820] bridge: BH already disabled in br_fdb_cleanup() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit br_fdb_cleanup() is run from timer interrupt, BH already masked. Signed-off-by: Eric Dumazet CC: Stephen Hemminger CC: Štefan Gula Acked-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/bridge/br_fdb.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/bridge/br_fdb.c b/net/bridge/br_fdb.c index f963f6b1884f..5ba0c844d508 100644 --- a/net/bridge/br_fdb.c +++ b/net/bridge/br_fdb.c @@ -146,7 +146,7 @@ void br_fdb_cleanup(unsigned long _data) unsigned long next_timer = jiffies + br->ageing_time; int i; - spin_lock_bh(&br->hash_lock); + spin_lock(&br->hash_lock); for (i = 0; i < BR_HASH_SIZE; i++) { struct net_bridge_fdb_entry *f; struct hlist_node *h, *n; @@ -162,7 +162,7 @@ void br_fdb_cleanup(unsigned long _data) next_timer = this_timer; } } - spin_unlock_bh(&br->hash_lock); + spin_unlock(&br->hash_lock); mod_timer(&br->gc_timer, round_jiffies_up(next_timer)); } From e52ac3398c3d772d372b9b62ab408fd5eec96840 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Mon, 16 Jan 2012 12:38:59 +0000 Subject: [PATCH 491/820] net: Use device model to get driver name in skb_gso_segment() ethtool operations generally require the caller to hold RTNL and are not safe to call in atomic context. The device model provides this information for most devices; we'll only lose it for some old ISA drivers. Signed-off-by: Ben Hutchings Acked-by: Herbert Xu Signed-off-by: David S. Miller --- net/core/dev.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index f494675471a9..7e6b7dcaacde 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1962,13 +1962,13 @@ struct sk_buff *skb_gso_segment(struct sk_buff *skb, if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) { struct net_device *dev = skb->dev; - struct ethtool_drvinfo info = {}; + const char *driver = ""; - if (dev && dev->ethtool_ops && dev->ethtool_ops->get_drvinfo) - dev->ethtool_ops->get_drvinfo(dev, &info); + if (dev && dev->dev.parent) + driver = dev_driver_string(dev->dev.parent); WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d ip_summed=%d\n", - info.driver, dev ? &dev->features : NULL, + driver, dev ? &dev->features : NULL, skb->sk ? &skb->sk->sk_route_caps : NULL, skb->len, skb->data_len, skb->ip_summed); From 34aa6f1400810890636ba0b170effbfa71eacec7 Mon Sep 17 00:00:00 2001 From: Nobuhiro Iwamatsu Date: Mon, 16 Jan 2012 16:50:16 +0000 Subject: [PATCH 492/820] net: sh-eth: Fix build error by the value which is not defined ----- drivers/net/ethernet/renesas/sh_eth.c:1706: error: 'pdid' undeclared (first use in this function) drivers/net/ethernet/renesas/sh_eth.c:1706: error: (Each undeclared identifier is reported only once drivers/net/ethernet/renesas/sh_eth.c:1706: error: for each function it appears in.) make[5]: *** [drivers/net/ethernet/renesas/sh_eth.o] Error 1 ----- Signed-off-by: Nobuhiro Iwamatsu CC: Florian Fainelli Signed-off-by: David S. Miller --- drivers/net/ethernet/renesas/sh_eth.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/renesas/sh_eth.c b/drivers/net/ethernet/renesas/sh_eth.c index 6ece4295d78f..813d41c4a845 100644 --- a/drivers/net/ethernet/renesas/sh_eth.c +++ b/drivers/net/ethernet/renesas/sh_eth.c @@ -1703,7 +1703,7 @@ static int sh_mdio_init(struct net_device *ndev, int id, mdp->mii_bus->name = "sh_mii"; mdp->mii_bus->parent = &ndev->dev; snprintf(mdp->mii_bus->id, MII_BUS_ID_SIZE, "%s-%x", - mdp->pdev->name, pdid); + mdp->pdev->name, id); /* PHY IRQ */ mdp->mii_bus->irq = kmalloc(sizeof(int)*PHY_MAX_ADDR, GFP_KERNEL); From 2b2d465631db59c144f537a9bef47a883c55d670 Mon Sep 17 00:00:00 2001 From: Shan Wei Date: Mon, 16 Jan 2012 18:34:24 +0000 Subject: [PATCH 493/820] net: kill duplicate included header For net part, remove duplicate included header. Signed-off-by: Shan Wei Signed-off-by: David S. Miller --- net/caif/caif_usb.c | 1 - net/ipv4/udp_diag.c | 1 - net/openvswitch/flow.c | 1 - 3 files changed, 3 deletions(-) diff --git a/net/caif/caif_usb.c b/net/caif/caif_usb.c index 5fc9eca8cd41..3d737fbcb182 100644 --- a/net/caif/caif_usb.c +++ b/net/caif/caif_usb.c @@ -11,7 +11,6 @@ #include #include #include -#include #include #include #include diff --git a/net/ipv4/udp_diag.c b/net/ipv4/udp_diag.c index e5e18cb8a586..8a949f19deb6 100644 --- a/net/ipv4/udp_diag.c +++ b/net/ipv4/udp_diag.c @@ -15,7 +15,6 @@ #include #include #include -#include #include static int sk_diag_dump(struct sock *sk, struct sk_buff *skb, diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c index fe7f020a843e..1252c3081ef1 100644 --- a/net/openvswitch/flow.c +++ b/net/openvswitch/flow.c @@ -32,7 +32,6 @@ #include #include #include -#include #include #include #include From 747465ef7a082033e086dedc8189febfda43b015 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 16 Jan 2012 19:27:39 +0000 Subject: [PATCH 494/820] net: fix some sparse errors make C=2 CF="-D__CHECK_ENDIAN__" M=net And fix flowi4_init_output() prototype for sport Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/flow.h | 2 +- net/core/secure_seq.c | 2 +- net/ipv4/inetpeer.c | 2 +- net/ipv4/ipconfig.c | 8 ++++---- net/ipv4/ping.c | 27 ++++++++++++++------------- net/ipv6/datagram.c | 2 +- 6 files changed, 22 insertions(+), 21 deletions(-) diff --git a/include/net/flow.h b/include/net/flow.h index da1f064a81b3..9b582437fbea 100644 --- a/include/net/flow.h +++ b/include/net/flow.h @@ -78,7 +78,7 @@ static inline void flowi4_init_output(struct flowi4 *fl4, int oif, __u32 mark, __u8 tos, __u8 scope, __u8 proto, __u8 flags, __be32 daddr, __be32 saddr, - __be16 dport, __be32 sport) + __be16 dport, __be16 sport) { fl4->flowi4_oif = oif; fl4->flowi4_iif = 0; diff --git a/net/core/secure_seq.c b/net/core/secure_seq.c index 6fd44606fdd1..99b2596531bb 100644 --- a/net/core/secure_seq.c +++ b/net/core/secure_seq.c @@ -46,7 +46,7 @@ __u32 secure_tcpv6_sequence_number(const __be32 *saddr, const __be32 *daddr, memcpy(hash, saddr, 16); for (i = 0; i < 4; i++) - secret[i] = net_secret[i] + daddr[i]; + secret[i] = net_secret[i] + (__force u32)daddr[i]; secret[4] = net_secret[4] + (((__force u16)sport << 16) + (__force u16)dport); for (i = 5; i < MD5_MESSAGE_BYTES / 4; i++) diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index 86f13c67ea85..986a02bda660 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -136,7 +136,7 @@ static int addr_compare(const struct inetpeer_addr *a, for (i = 0; i < n; i++) { if (a->addr.a6[i] == b->addr.a6[i]) continue; - if (a->addr.a6[i] < b->addr.a6[i]) + if ((__force u32)a->addr.a6[i] < (__force u32)b->addr.a6[i]) return -1; return 1; } diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index 7e4ec9fc2cef..6e412a60a91f 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -141,7 +141,7 @@ __be32 ic_servaddr = NONE; /* Boot server IP address */ __be32 root_server_addr = NONE; /* Address of NFS server */ u8 root_server_path[256] = { 0, }; /* Path to mount as root */ -u32 ic_dev_xid; /* Device under configuration */ +__be32 ic_dev_xid; /* Device under configuration */ /* vendor class identifier */ static char vendor_class_identifier[253] __initdata; @@ -859,9 +859,9 @@ static int __init ic_bootp_string(char *dest, char *src, int len, int max) */ static void __init ic_do_bootp_ext(u8 *ext) { - u8 servers; - int i; - u16 mtu; + u8 servers; + int i; + __be16 mtu; #ifdef IPCONFIG_DEBUG u8 *c; diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index 43d4c3b22369..aea5a199c37a 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -140,13 +140,14 @@ static void ping_v4_unhash(struct sock *sk) write_lock_bh(&ping_table.lock); hlist_nulls_del(&sk->sk_nulls_node); sock_put(sk); - isk->inet_num = isk->inet_sport = 0; + isk->inet_num = 0; + isk->inet_sport = 0; sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); write_unlock_bh(&ping_table.lock); } } -static struct sock *ping_v4_lookup(struct net *net, u32 saddr, u32 daddr, +static struct sock *ping_v4_lookup(struct net *net, __be32 saddr, __be32 daddr, u16 ident, int dif) { struct hlist_nulls_head *hslot = ping_hashslot(&ping_table, net, ident); @@ -154,15 +155,15 @@ static struct sock *ping_v4_lookup(struct net *net, u32 saddr, u32 daddr, struct inet_sock *isk; struct hlist_nulls_node *hnode; - pr_debug("try to find: num = %d, daddr = %ld, dif = %d\n", - (int)ident, (unsigned long)daddr, dif); + pr_debug("try to find: num = %d, daddr = %pI4, dif = %d\n", + (int)ident, &daddr, dif); read_lock_bh(&ping_table.lock); ping_portaddr_for_each_entry(sk, hnode, hslot) { isk = inet_sk(sk); - pr_debug("found: %p: num = %d, daddr = %ld, dif = %d\n", sk, - (int)isk->inet_num, (unsigned long)isk->inet_rcv_saddr, + pr_debug("found: %p: num = %d, daddr = %pI4, dif = %d\n", sk, + (int)isk->inet_num, &isk->inet_rcv_saddr, sk->sk_bound_dev_if); pr_debug("iterate\n"); @@ -254,7 +255,7 @@ static int ping_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) sk, addr->sin_addr.s_addr, ntohs(addr->sin_port)); chk_addr_ret = inet_addr_type(sock_net(sk), addr->sin_addr.s_addr); - if (addr->sin_addr.s_addr == INADDR_ANY) + if (addr->sin_addr.s_addr == htonl(INADDR_ANY)) chk_addr_ret = RTN_LOCAL; if ((sysctl_ip_nonlocal_bind == 0 && @@ -278,9 +279,9 @@ static int ping_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) goto out; } - pr_debug("after bind(): num = %d, daddr = %ld, dif = %d\n", + pr_debug("after bind(): num = %d, daddr = %pI4, dif = %d\n", (int)isk->inet_num, - (unsigned long) isk->inet_rcv_saddr, + &isk->inet_rcv_saddr, (int)sk->sk_bound_dev_if); err = 0; @@ -407,7 +408,7 @@ out: struct pingfakehdr { struct icmphdr icmph; struct iovec *iov; - u32 wcheck; + __wsum wcheck; }; static int ping_getfrag(void *from, char * to, @@ -459,7 +460,7 @@ static int ping_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, struct rtable *rt = NULL; struct ip_options_data opt_copy; int free = 0; - u32 saddr, daddr, faddr; + __be32 saddr, daddr, faddr; u8 tos; int err; @@ -696,8 +697,8 @@ void ping_rcv(struct sk_buff *skb) struct net *net = dev_net(skb->dev); struct iphdr *iph = ip_hdr(skb); struct icmphdr *icmph = icmp_hdr(skb); - u32 saddr = iph->saddr; - u32 daddr = iph->daddr; + __be32 saddr = iph->saddr; + __be32 daddr = iph->daddr; /* We assume the packet has already been checked by icmp_rcv */ diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index ae08aee1773c..251e7cd75e89 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -575,7 +575,7 @@ int datagram_recv_ctl(struct sock *sk, struct msghdr *msg, struct sk_buff *skb) } if (np->rxopt.bits.rxorigdstaddr) { struct sockaddr_in6 sin6; - u16 *ports = (u16 *) skb_transport_header(skb); + __be16 *ports = (__be16 *) skb_transport_header(skb); if (skb_transport_offset(skb) + 4 <= skb->len) { /* All current transport protocols have the port numbers in the From 11b2ec6b739ee90211dc6f6942e2ba3a141434a8 Mon Sep 17 00:00:00 2001 From: Yaniv Rosner Date: Tue, 17 Jan 2012 02:33:25 +0000 Subject: [PATCH 495/820] bnx2x: Fix Super-Isolate mode for BCM84833 The Super-Isolate mode comes to isolate the BCM84833 PHY from the outside world. Not doing it correctly, made link partner see the link before the driver was loaded. This patch also involves SPIROM version fixes since it is used to determine whether the common init of the PHY was already executed, and the common init of this PHY is partially responsible for setting the Super-Isolate mode. Signed-off-by: Yaniv Rosner Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- .../net/ethernet/broadcom/bnx2x/bnx2x_link.c | 264 ++++++++++-------- 1 file changed, 144 insertions(+), 120 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c index 4df9505b67b6..3b184c2a8553 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c @@ -9266,62 +9266,68 @@ static void bnx2x_8727_link_reset(struct bnx2x_phy *phy, /* BCM8481/BCM84823/BCM84833 PHY SECTION */ /******************************************************************/ static void bnx2x_save_848xx_spirom_version(struct bnx2x_phy *phy, - struct link_params *params) + struct bnx2x *bp, + u8 port) { u16 val, fw_ver1, fw_ver2, cnt; - u8 port; - struct bnx2x *bp = params->bp; - port = params->port; + if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { + bnx2x_cl45_read(bp, phy, MDIO_CTL_DEVAD, 0x400f, &fw_ver1); + bnx2x_save_spirom_version(bp, port, + ((fw_ver1 & 0xf000)>>5) | (fw_ver1 & 0x7f), + phy->ver_addr); + } else { + /* For 32-bit registers in 848xx, access via MDIO2ARM i/f. */ + /* (1) set reg 0xc200_0014(SPI_BRIDGE_CTRL_2) to 0x03000000 */ + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA819, 0x0014); + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81A, 0xc200); + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81B, 0x0000); + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81C, 0x0300); + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA817, 0x0009); - /* For the 32 bits registers in 848xx, access via MDIO2ARM interface.*/ - /* (1) set register 0xc200_0014(SPI_BRIDGE_CTRL_2) to 0x03000000 */ - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA819, 0x0014); - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81A, 0xc200); - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81B, 0x0000); - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81C, 0x0300); - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA817, 0x0009); + for (cnt = 0; cnt < 100; cnt++) { + bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA818, &val); + if (val & 1) + break; + udelay(5); + } + if (cnt == 100) { + DP(NETIF_MSG_LINK, "Unable to read 848xx " + "phy fw version(1)\n"); + bnx2x_save_spirom_version(bp, port, 0, + phy->ver_addr); + return; + } - for (cnt = 0; cnt < 100; cnt++) { - bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA818, &val); - if (val & 1) - break; - udelay(5); - } - if (cnt == 100) { - DP(NETIF_MSG_LINK, "Unable to read 848xx phy fw version(1)\n"); - bnx2x_save_spirom_version(bp, port, 0, + + /* 2) read register 0xc200_0000 (SPI_FW_STATUS) */ + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA819, 0x0000); + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81A, 0xc200); + bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA817, 0x000A); + for (cnt = 0; cnt < 100; cnt++) { + bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA818, &val); + if (val & 1) + break; + udelay(5); + } + if (cnt == 100) { + DP(NETIF_MSG_LINK, "Unable to read 848xx phy fw " + "version(2)\n"); + bnx2x_save_spirom_version(bp, port, 0, + phy->ver_addr); + return; + } + + /* lower 16 bits of the register SPI_FW_STATUS */ + bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA81B, &fw_ver1); + /* upper 16 bits of register SPI_FW_STATUS */ + bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA81C, &fw_ver2); + + bnx2x_save_spirom_version(bp, port, (fw_ver2<<16) | fw_ver1, phy->ver_addr); - return; } - - /* 2) read register 0xc200_0000 (SPI_FW_STATUS) */ - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA819, 0x0000); - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA81A, 0xc200); - bnx2x_cl45_write(bp, phy, MDIO_PMA_DEVAD, 0xA817, 0x000A); - for (cnt = 0; cnt < 100; cnt++) { - bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA818, &val); - if (val & 1) - break; - udelay(5); - } - if (cnt == 100) { - DP(NETIF_MSG_LINK, "Unable to read 848xx phy fw version(2)\n"); - bnx2x_save_spirom_version(bp, port, 0, - phy->ver_addr); - return; - } - - /* lower 16 bits of the register SPI_FW_STATUS */ - bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA81B, &fw_ver1); - /* upper 16 bits of register SPI_FW_STATUS */ - bnx2x_cl45_read(bp, phy, MDIO_PMA_DEVAD, 0xA81C, &fw_ver2); - - bnx2x_save_spirom_version(bp, port, (fw_ver2<<16) | fw_ver1, - phy->ver_addr); } - static void bnx2x_848xx_set_led(struct bnx2x *bp, struct bnx2x_phy *phy) { @@ -9392,10 +9398,13 @@ static int bnx2x_848xx_cmn_config_init(struct bnx2x_phy *phy, u16 tmp_req_line_speed; tmp_req_line_speed = phy->req_line_speed; - if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) + if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { if (phy->req_line_speed == SPEED_10000) phy->req_line_speed = SPEED_AUTO_NEG; - + } else { + /* Save spirom version */ + bnx2x_save_848xx_spirom_version(phy, bp, params->port); + } /* * This phy uses the NIG latch mechanism since link indication * arrives through its LED4 and not via its LASI signal, so we @@ -9539,9 +9548,6 @@ static int bnx2x_848xx_cmn_config_init(struct bnx2x_phy *phy, MDIO_AN_REG_8481_10GBASE_T_AN_CTRL, 1); - /* Save spirom version */ - bnx2x_save_848xx_spirom_version(phy, params); - phy->req_line_speed = tmp_req_line_speed; return 0; @@ -9749,17 +9755,7 @@ static int bnx2x_848x3_config_init(struct bnx2x_phy *phy, /* Wait for GPHY to come out of reset */ msleep(50); - if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { - /* Bring PHY out of super isolate mode */ - bnx2x_cl45_read(bp, phy, - MDIO_CTL_DEVAD, - MDIO_84833_TOP_CFG_XGPHY_STRAP1, &val); - val &= ~MDIO_84833_SUPER_ISOLATE; - bnx2x_cl45_write(bp, phy, - MDIO_CTL_DEVAD, - MDIO_84833_TOP_CFG_XGPHY_STRAP1, val); - bnx2x_84833_pair_swap_cfg(phy, params, vars); - } else { + if (phy->type != PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { /* * BCM84823 requires that XGXS links up first @ 10G for normal * behavior. @@ -9816,24 +9812,27 @@ static int bnx2x_848x3_config_init(struct bnx2x_phy *phy, DP(NETIF_MSG_LINK, "Multi_phy config = 0x%x, Media control = 0x%x\n", params->multi_phy_config, val); - /* AutogrEEEn */ - if (params->feature_config_flags & - FEATURE_CONFIG_AUTOGREEEN_ENABLED) - cmd_args[0] = 0x2; - else - cmd_args[0] = 0x0; + if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { + bnx2x_84833_pair_swap_cfg(phy, params, vars); - cmd_args[1] = 0x0; - cmd_args[2] = PHY84833_CONSTANT_LATENCY + 1; - cmd_args[3] = PHY84833_CONSTANT_LATENCY; - rc = bnx2x_84833_cmd_hdlr(phy, params, - PHY84833_CMD_SET_EEE_MODE, cmd_args); - if (rc != 0) - DP(NETIF_MSG_LINK, "Cfg AutogrEEEn failed.\n"); + /* AutogrEEEn */ + if (params->feature_config_flags & + FEATURE_CONFIG_AUTOGREEEN_ENABLED) + cmd_args[0] = 0x2; + else + cmd_args[0] = 0x0; + cmd_args[1] = 0x0; + cmd_args[2] = PHY84833_CONSTANT_LATENCY + 1; + cmd_args[3] = PHY84833_CONSTANT_LATENCY; + rc = bnx2x_84833_cmd_hdlr(phy, params, + PHY84833_CMD_SET_EEE_MODE, cmd_args); + if (rc != 0) + DP(NETIF_MSG_LINK, "Cfg AutogrEEEn failed.\n"); + } if (initialize) rc = bnx2x_848xx_cmn_config_init(phy, params, vars); else - bnx2x_save_848xx_spirom_version(phy, params); + bnx2x_save_848xx_spirom_version(phy, bp, params->port); /* 84833 PHY has a better feature and doesn't need to support this. */ if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84823) { cms_enable = REG_RD(bp, params->shmem_base + @@ -9851,6 +9850,16 @@ static int bnx2x_848x3_config_init(struct bnx2x_phy *phy, MDIO_CTL_REG_84823_USER_CTRL_REG, val); } + if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { + /* Bring PHY out of super isolate mode as the final step. */ + bnx2x_cl45_read(bp, phy, + MDIO_CTL_DEVAD, + MDIO_84833_TOP_CFG_XGPHY_STRAP1, &val); + val &= ~MDIO_84833_SUPER_ISOLATE; + bnx2x_cl45_write(bp, phy, + MDIO_CTL_DEVAD, + MDIO_84833_TOP_CFG_XGPHY_STRAP1, val); + } return rc; } @@ -9988,10 +9997,11 @@ static void bnx2x_848x3_link_reset(struct bnx2x_phy *phy, } else { bnx2x_cl45_read(bp, phy, MDIO_CTL_DEVAD, - 0x400f, &val16); + MDIO_84833_TOP_CFG_XGPHY_STRAP1, &val16); + val16 |= MDIO_84833_SUPER_ISOLATE; bnx2x_cl45_write(bp, phy, - MDIO_PMA_DEVAD, - MDIO_PMA_REG_CTRL, 0x800); + MDIO_CTL_DEVAD, + MDIO_84833_TOP_CFG_XGPHY_STRAP1, val16); } } @@ -12333,55 +12343,69 @@ static int bnx2x_84833_common_init_phy(struct bnx2x *bp, u32 chip_id) { u8 reset_gpios; - struct bnx2x_phy phy; - u32 shmem_base, shmem2_base, cnt; - s8 port = 0; - u16 val; - reset_gpios = bnx2x_84833_get_reset_gpios(bp, shmem_base_path, chip_id); bnx2x_set_mult_gpio(bp, reset_gpios, MISC_REGISTERS_GPIO_OUTPUT_LOW); udelay(10); bnx2x_set_mult_gpio(bp, reset_gpios, MISC_REGISTERS_GPIO_OUTPUT_HIGH); DP(NETIF_MSG_LINK, "84833 reset pulse on pin values 0x%x\n", reset_gpios); - for (port = PORT_MAX - 1; port >= PORT_0; port--) { - /* This PHY is for E2 and E3. */ - shmem_base = shmem_base_path[port]; - shmem2_base = shmem2_base_path[port]; - /* Extract the ext phy address for the port */ - if (bnx2x_populate_phy(bp, phy_index, shmem_base, shmem2_base, - 0, &phy) != - 0) { - DP(NETIF_MSG_LINK, "populate_phy failed\n"); - return -EINVAL; - } - - /* Wait for FW completing its initialization. */ - for (cnt = 0; cnt < 1000; cnt++) { - bnx2x_cl45_read(bp, &phy, - MDIO_PMA_DEVAD, - MDIO_PMA_REG_CTRL, &val); - if (!(val & (1<<15))) - break; - msleep(1); - } - if (cnt >= 1000) - DP(NETIF_MSG_LINK, - "84833 Cmn reset timeout (%d)\n", port); - - /* Put the port in super isolate mode. */ - bnx2x_cl45_read(bp, &phy, - MDIO_CTL_DEVAD, - MDIO_84833_TOP_CFG_XGPHY_STRAP1, &val); - val |= MDIO_84833_SUPER_ISOLATE; - bnx2x_cl45_write(bp, &phy, - MDIO_CTL_DEVAD, - MDIO_84833_TOP_CFG_XGPHY_STRAP1, val); - } - return 0; } +static int bnx2x_84833_pre_init_phy(struct bnx2x *bp, + struct bnx2x_phy *phy) +{ + u16 val, cnt; + /* Wait for FW completing its initialization. */ + for (cnt = 0; cnt < 1500; cnt++) { + bnx2x_cl45_read(bp, phy, + MDIO_PMA_DEVAD, + MDIO_PMA_REG_CTRL, &val); + if (!(val & (1<<15))) + break; + msleep(1); + } + if (cnt >= 1500) { + DP(NETIF_MSG_LINK, "84833 reset timeout\n"); + return -EINVAL; + } + + /* Put the port in super isolate mode. */ + bnx2x_cl45_read(bp, phy, + MDIO_CTL_DEVAD, + MDIO_84833_TOP_CFG_XGPHY_STRAP1, &val); + val |= MDIO_84833_SUPER_ISOLATE; + bnx2x_cl45_write(bp, phy, + MDIO_CTL_DEVAD, + MDIO_84833_TOP_CFG_XGPHY_STRAP1, val); + + /* Save spirom version */ + bnx2x_save_848xx_spirom_version(phy, bp, PORT_0); + return 0; +} + +int bnx2x_pre_init_phy(struct bnx2x *bp, + u32 shmem_base, + u32 shmem2_base, + u32 chip_id) +{ + int rc = 0; + struct bnx2x_phy phy; + bnx2x_set_mdio_clk(bp, chip_id, PORT_0); + if (bnx2x_populate_phy(bp, EXT_PHY1, shmem_base, shmem2_base, + PORT_0, &phy)) { + DP(NETIF_MSG_LINK, "populate_phy failed\n"); + return -EINVAL; + } + switch (phy.type) { + case PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833: + rc = bnx2x_84833_pre_init_phy(bp, &phy); + break; + default: + break; + } + return rc; +} static int bnx2x_ext_phy_common_init(struct bnx2x *bp, u32 shmem_base_path[], u32 shmem2_base_path[], u8 phy_index, From 127302bb42257eef5d357722d670c4ac53327088 Mon Sep 17 00:00:00 2001 From: Yaniv Rosner Date: Tue, 17 Jan 2012 02:33:26 +0000 Subject: [PATCH 496/820] bnx2x: Fix PFC setting on BCM57840 This patch handles the second port of a path in a 4-port device of BCM57840. Signed-off-by: Yaniv Rosner Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- .../net/ethernet/broadcom/bnx2x/bnx2x_link.c | 22 ++++++++++--------- .../net/ethernet/broadcom/bnx2x/bnx2x_reg.h | 1 + 2 files changed, 13 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c index 3b184c2a8553..cc806378464d 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c @@ -2502,7 +2502,7 @@ static void bnx2x_update_pfc_nig(struct link_params *params, struct bnx2x_nig_brb_pfc_port_params *nig_params) { u32 xcm_mask = 0, ppp_enable = 0, pause_enable = 0, llfc_out_en = 0; - u32 llfc_enable = 0, xcm0_out_en = 0, p0_hwpfc_enable = 0; + u32 llfc_enable = 0, xcm_out_en = 0, hwpfc_enable = 0; u32 pkt_priority_to_cos = 0; struct bnx2x *bp = params->bp; u8 port = params->port; @@ -2516,9 +2516,8 @@ static void bnx2x_update_pfc_nig(struct link_params *params, * MAC control frames (that are not pause packets) * will be forwarded to the XCM. */ - xcm_mask = REG_RD(bp, - port ? NIG_REG_LLH1_XCM_MASK : - NIG_REG_LLH0_XCM_MASK); + xcm_mask = REG_RD(bp, port ? NIG_REG_LLH1_XCM_MASK : + NIG_REG_LLH0_XCM_MASK); /* * nig params will override non PFC params, since it's possible to * do transition from PFC to SAFC @@ -2533,8 +2532,8 @@ static void bnx2x_update_pfc_nig(struct link_params *params, ppp_enable = 1; xcm_mask &= ~(port ? NIG_LLH1_XCM_MASK_REG_LLH1_XCM_MASK_BCN : NIG_LLH0_XCM_MASK_REG_LLH0_XCM_MASK_BCN); - xcm0_out_en = 0; - p0_hwpfc_enable = 1; + xcm_out_en = 0; + hwpfc_enable = 1; } else { if (nig_params) { llfc_out_en = nig_params->llfc_out_en; @@ -2545,7 +2544,7 @@ static void bnx2x_update_pfc_nig(struct link_params *params, xcm_mask |= (port ? NIG_LLH1_XCM_MASK_REG_LLH1_XCM_MASK_BCN : NIG_LLH0_XCM_MASK_REG_LLH0_XCM_MASK_BCN); - xcm0_out_en = 1; + xcm_out_en = 1; } if (CHIP_IS_E3(bp)) @@ -2564,13 +2563,16 @@ static void bnx2x_update_pfc_nig(struct link_params *params, REG_WR(bp, port ? NIG_REG_LLH1_XCM_MASK : NIG_REG_LLH0_XCM_MASK, xcm_mask); - REG_WR(bp, NIG_REG_LLFC_EGRESS_SRC_ENABLE_0, 0x7); + REG_WR(bp, port ? NIG_REG_LLFC_EGRESS_SRC_ENABLE_1 : + NIG_REG_LLFC_EGRESS_SRC_ENABLE_0, 0x7); /* output enable for RX_XCM # IF */ - REG_WR(bp, NIG_REG_XCM0_OUT_EN, xcm0_out_en); + REG_WR(bp, port ? NIG_REG_XCM1_OUT_EN : + NIG_REG_XCM0_OUT_EN, xcm_out_en); /* HW PFC TX enable */ - REG_WR(bp, NIG_REG_P0_HWPFC_ENABLE, p0_hwpfc_enable); + REG_WR(bp, port ? NIG_REG_P1_HWPFC_ENABLE : + NIG_REG_P0_HWPFC_ENABLE, hwpfc_enable); if (nig_params) { u8 i = 0; diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h index 44609de4e5dc..dddbcf6e154e 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_reg.h @@ -2176,6 +2176,7 @@ * set to 0x345678021. This is a new register (with 2_) added in E3 B0 to * accommodate the 9 input clients to ETS arbiter. */ #define NIG_REG_P0_TX_ARB_PRIORITY_CLIENT2_MSB 0x18684 +#define NIG_REG_P1_HWPFC_ENABLE 0x181d0 #define NIG_REG_P1_MAC_IN_EN 0x185c0 /* [RW 1] Output enable for TX MAC interface */ #define NIG_REG_P1_MAC_OUT_EN 0x185c4 From 75318327802235ecd7e90b0760cceb994bf975ca Mon Sep 17 00:00:00 2001 From: Yaniv Rosner Date: Tue, 17 Jan 2012 02:33:27 +0000 Subject: [PATCH 497/820] bnx2x: Remove 100Mb force speed for BCM84833 Remove unsupported speed of 100Mb force for BCM84833 due to hardware limitation. Signed-off-by: Yaniv Rosner Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- .../ethernet/broadcom/bnx2x/bnx2x_ethtool.c | 7 ++++++- .../net/ethernet/broadcom/bnx2x/bnx2x_link.c | 20 ++++++++++++++----- 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c index a688b9d975a2..f99c6e312a5d 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c @@ -365,13 +365,18 @@ static int bnx2x_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) DP(NETIF_MSG_LINK, "cfg_idx = %x\n", cfg_idx); if (cmd->autoneg == AUTONEG_ENABLE) { + u32 an_supported_speed = bp->port.supported[cfg_idx]; + if (bp->link_params.phy[EXT_PHY1].type == + PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) + an_supported_speed |= (SUPPORTED_100baseT_Half | + SUPPORTED_100baseT_Full); if (!(bp->port.supported[cfg_idx] & SUPPORTED_Autoneg)) { DP(NETIF_MSG_LINK, "Autoneg not supported\n"); return -EINVAL; } /* advertise the requested speed and duplex if supported */ - if (cmd->advertising & ~(bp->port.supported[cfg_idx])) { + if (cmd->advertising & ~an_supported_speed) { DP(NETIF_MSG_LINK, "Advertisement parameters " "are not supported\n"); return -EINVAL; diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c index cc806378464d..1d1a8093e62e 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c @@ -9454,13 +9454,10 @@ static int bnx2x_848xx_cmn_config_init(struct bnx2x_phy *phy, an_1000_val); /* set 100 speed advertisement */ - if (((phy->req_line_speed == SPEED_AUTO_NEG) && + if ((phy->req_line_speed == SPEED_AUTO_NEG) && (phy->speed_cap_mask & (PORT_HW_CFG_SPEED_CAPABILITY_D0_100M_FULL | - PORT_HW_CFG_SPEED_CAPABILITY_D0_100M_HALF)) && - (phy->supported & - (SUPPORTED_100baseT_Half | - SUPPORTED_100baseT_Full)))) { + PORT_HW_CFG_SPEED_CAPABILITY_D0_100M_HALF))) { an_10_100_val |= (1<<7); /* Enable autoneg and restart autoneg for legacy speeds */ autoneg_val |= (1<<9 | 1<<12); @@ -11528,6 +11525,19 @@ static int bnx2x_populate_ext_phy(struct bnx2x *bp, } phy->mdio_ctrl = bnx2x_get_emac_base(bp, mdc_mdio_access, port); + if ((phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) && + (phy->ver_addr)) { + /* + * Remove 100Mb link supported for BCM84833 when phy fw + * version lower than or equal to 1.39 + */ + u32 raw_ver = REG_RD(bp, phy->ver_addr); + if (((raw_ver & 0x7F) <= 39) && + (((raw_ver & 0xF80) >> 7) <= 1)) + phy->supported &= ~(SUPPORTED_100baseT_Half | + SUPPORTED_100baseT_Full); + } + /* * In case mdc/mdio_access of the external phy is different than the * mdc/mdio access of the XGXS, a HW lock must be taken in each access From 096b9527db77defb2501509f2517d2ab24300d9c Mon Sep 17 00:00:00 2001 From: Yaniv Rosner Date: Tue, 17 Jan 2012 02:33:28 +0000 Subject: [PATCH 498/820] bnx2x: Remove AutoGrEEEn for BCM84833 Disable the autoGrEEEn feature for BCM84833. Signed-off-by: Yaniv Rosner Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c index 1d1a8093e62e..a6c48d61081a 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c @@ -9814,12 +9814,8 @@ static int bnx2x_848x3_config_init(struct bnx2x_phy *phy, if (phy->type == PORT_HW_CFG_XGXS_EXT_PHY_TYPE_BCM84833) { bnx2x_84833_pair_swap_cfg(phy, params, vars); - /* AutogrEEEn */ - if (params->feature_config_flags & - FEATURE_CONFIG_AUTOGREEEN_ENABLED) - cmd_args[0] = 0x2; - else - cmd_args[0] = 0x0; + /* Keep AutogrEEEn disabled. */ + cmd_args[0] = 0x0; cmd_args[1] = 0x0; cmd_args[2] = PHY84833_CONSTANT_LATENCY + 1; cmd_args[3] = PHY84833_CONSTANT_LATENCY; From 6ab48a5c86ce778188c173818cb2f1644526e962 Mon Sep 17 00:00:00 2001 From: Yaniv Rosner Date: Tue, 17 Jan 2012 02:33:29 +0000 Subject: [PATCH 499/820] bnx2x: Disable AN KR work-around for BCM57810 Disable the work-around for the autoneg KR of the BCM57810 in case the Warpcore version is 0xD108 and above, which fixes this problem. Signed-off-by: Yaniv Rosner Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c index a6c48d61081a..2091e5dbbcdd 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c @@ -3763,7 +3763,15 @@ static void bnx2x_warpcore_enable_AN_KR(struct bnx2x_phy *phy, /* Advertise pause */ bnx2x_ext_phy_set_pause(params, phy, vars); - vars->rx_tx_asic_rst = MAX_KR_LINK_RETRY; + /* + * Set KR Autoneg Work-Around flag for Warpcore version older than D108 + */ + bnx2x_cl45_read(bp, phy, MDIO_WC_DEVAD, + MDIO_WC_REG_UC_INFO_B1_VERSION, &val16); + if (val16 < 0xd108) { + DP(NETIF_MSG_LINK, "Enable AN KR work-around\n"); + vars->rx_tx_asic_rst = MAX_KR_LINK_RETRY; + } bnx2x_cl45_read(bp, phy, MDIO_WC_DEVAD, MDIO_WC_REG_DIGITAL5_MISC7, &val16); From 3371bb3f7ed11b7b7473a202e2713bde50dc01c0 Mon Sep 17 00:00:00 2001 From: "sjur.brandeland@stericsson.com" Date: Tue, 17 Jan 2012 03:03:13 +0000 Subject: [PATCH 500/820] caif: Fix typo in Vendor/Product-ID for CAIF modems MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix typo for the Vendor/Product Id for ST-Ericsson CAIF modems. Discovery is based on fixed USB vendor 0x04cc (ST-Ericsson), product-id 0x230f (NCM). Signed-off-by: Sjur Brændeland Signed-off-by: David S. Miller --- net/caif/caif_usb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/caif/caif_usb.c b/net/caif/caif_usb.c index 3d737fbcb182..fd7cbf5aa895 100644 --- a/net/caif/caif_usb.c +++ b/net/caif/caif_usb.c @@ -26,7 +26,7 @@ MODULE_LICENSE("GPL"); #define CFUSB_ALIGNMENT 4 /* Number of bytes to align. */ #define CFUSB_MAX_HEADLEN (CFUSB_PAD_DESCR_SZ + CFUSB_ALIGNMENT-1) #define STE_USB_VID 0x04cc /* USB Product ID for ST-Ericsson */ -#define STE_USB_PID_CAIF 0x2306 /* Product id for CAIF Modems */ +#define STE_USB_PID_CAIF 0x230f /* Product id for CAIF Modems */ struct cfusbl { struct cflayer layer; From 59f608d84f937cb5e2fd3a0b9f848f200d47ba85 Mon Sep 17 00:00:00 2001 From: "sjur.brandeland@stericsson.com" Date: Tue, 17 Jan 2012 03:03:14 +0000 Subject: [PATCH 501/820] caif: Remove bad WARN_ON in caif_dev MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Remove WARN_ON and bad handling of SKB without destructor callback in caif_flow_cb. SKB without destructor cannot be handled as an error case. Signed-off-by: Sjur Brændeland Signed-off-by: David S. Miller --- net/caif/caif_dev.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/net/caif/caif_dev.c b/net/caif/caif_dev.c index 61570ee76fe6..673728add60b 100644 --- a/net/caif/caif_dev.c +++ b/net/caif/caif_dev.c @@ -146,15 +146,17 @@ void caif_flow_cb(struct sk_buff *skb) spin_lock_bh(&caifd->flow_lock); send_xoff = caifd->xoff; caifd->xoff = 0; - if (!WARN_ON(caifd->xoff_skb_dtor == NULL)) { - WARN_ON(caifd->xoff_skb != skb); - dtor = caifd->xoff_skb_dtor; - caifd->xoff_skb = NULL; - caifd->xoff_skb_dtor = NULL; - } + dtor = caifd->xoff_skb_dtor; + + if (WARN_ON(caifd->xoff_skb != skb)) + skb = NULL; + + caifd->xoff_skb = NULL; + caifd->xoff_skb_dtor = NULL; + spin_unlock_bh(&caifd->flow_lock); - if (dtor) + if (dtor && skb) dtor(skb); if (send_xoff) From 986b2f2c21b308e21160a40359b81370f8fee6ec Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 17 Jan 2012 16:28:59 +0000 Subject: [PATCH 502/820] ASoC: Wait for WM8993 FLL to stabilise Ensure the FLL is locked before we return from set_fll(). Signed-off-by: Mark Brown --- sound/soc/codecs/wm8993.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/sound/soc/codecs/wm8993.c b/sound/soc/codecs/wm8993.c index 2b40c93601ed..7c7fd925db8d 100644 --- a/sound/soc/codecs/wm8993.c +++ b/sound/soc/codecs/wm8993.c @@ -444,6 +444,12 @@ static int _wm8993_set_fll(struct snd_soc_codec *codec, int fll_id, int source, /* Enable the FLL */ snd_soc_write(codec, WM8993_FLL_CONTROL_1, reg1 | WM8993_FLL_ENA); + /* Both overestimates */ + if (Fref < 1000000) + msleep(3); + else + msleep(1); + dev_dbg(codec->dev, "FLL enabled at %dHz->%dHz\n", Fref, Fout); wm8993->fll_fref = Fref; From 951880e634a79884236a575b896abf55c39ae0bf Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 17 Jan 2012 10:19:41 -0800 Subject: [PATCH 503/820] Revert "capabitlies: ns_capable can use the cap helpers rather than lsm call" This reverts commit d2a7009f0bb03fa22ad08dd25472efa0568126b9. J. R. Okajima explains: "After this commit, I am afraid access(2) on NFS may not work correctly. The scenario based upon my guess. - access(2) overrides the credentials. - calls inode_permission() -- ... -- generic_permission() -- ns_capable(). - while the old ns_capable() calls security_capable(current_cred()), the new ns_capable() calls has_ns_capability(current) -- security_capable(__task_cred(t)). current_cred() returns current->cred which is effective (overridden) credentials, but __task_cred(current) returns current->real_cred (the NFSD's credential). And the overridden credentials by access(2) lost." Requested-by: J. R. Okajima Acked-by: Eric Paris Signed-off-by: Linus Torvalds --- kernel/capability.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/capability.c b/kernel/capability.c index 0fcf1c14a297..3f1adb6c6470 100644 --- a/kernel/capability.c +++ b/kernel/capability.c @@ -384,7 +384,7 @@ bool ns_capable(struct user_namespace *ns, int cap) BUG(); } - if (has_ns_capability(current, ns, cap)) { + if (security_capable(current_cred(), ns, cap) == 0) { current->flags |= PF_SUPERPRIV; return true; } From 51dcb19aaf9448f6547f653b60a9f083845aad4a Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Tue, 17 Jan 2012 16:20:37 -0200 Subject: [PATCH 504/820] [media] dvb_frontend: Don't call get_frontend() if idle If the frontend is in idle state, don't call get_frontend. Calling get_frontend() when the device is not tuned may result in wrong parameters to be returned to the userspace. I was tempted to not call get_frontend() at all, except inside the dvb frontend thread, but this won't work for all cases. The ISDB-T specs (ABNT NBR 15601 and ARIB STD-B31) allow the broadcaster to dynamically change the channel specs at runtime. That means that an ISDB-T optimized application may want/need to monitor the TMCC tables, decoded at the frontends via get_frontend call. So, let's do the simpler change here. Eventually, the logic could be changed to work only if the device is tuned and has lock, but, even so, the lock is also standard-dependent. For ISDB-T, the right lock to wait is that the demod has TMCC lock. So, drivers may need to implement some logic to detect if the get_frontend info was retrieved or not. Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-core/dvb_frontend.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/media/dvb/dvb-core/dvb_frontend.c b/drivers/media/dvb/dvb-core/dvb_frontend.c index f5fa7aa89618..2998f38d64a2 100644 --- a/drivers/media/dvb/dvb-core/dvb_frontend.c +++ b/drivers/media/dvb/dvb-core/dvb_frontend.c @@ -1744,6 +1744,7 @@ static int dvb_frontend_ioctl_properties(struct file *file, { struct dvb_device *dvbdev = file->private_data; struct dvb_frontend *fe = dvbdev->priv; + struct dvb_frontend_private *fepriv = fe->frontend_priv; struct dtv_frontend_properties *c = &fe->dtv_property_cache; int err = 0; @@ -1810,9 +1811,14 @@ static int dvb_frontend_ioctl_properties(struct file *file, /* * Fills the cache out struct with the cache contents, plus - * the data retrieved from get_frontend. + * the data retrieved from get_frontend, if the frontend + * is not idle. Otherwise, returns the cached content */ - dtv_get_frontend(fe, NULL); + if (fepriv->state != FESTATE_IDLE) { + err = dtv_get_frontend(fe, NULL); + if (err < 0) + goto out; + } for (i = 0; i < tvps->num; i++) { err = dtv_property_process_get(fe, c, tvp + i, file); if (err < 0) From 20c300b10c358daa507be335aec6aa3987ef425a Mon Sep 17 00:00:00 2001 From: Konstantin Khlebnikov Date: Tue, 17 Jan 2012 12:54:01 +0400 Subject: [PATCH 505/820] tty: remove unused tty_driver->termios_locked This field is unused since 2.6.28 (commit fe6e29fdb1a7: "tty: simplify ktermios allocation", to be exact) Signed-off-by: Konstantin Khlebnikov Signed-off-by: Linus Torvalds --- include/linux/tty_driver.h | 1 - 1 file changed, 1 deletion(-) diff --git a/include/linux/tty_driver.h b/include/linux/tty_driver.h index ecdaeb98b293..5cf685086dd3 100644 --- a/include/linux/tty_driver.h +++ b/include/linux/tty_driver.h @@ -312,7 +312,6 @@ struct tty_driver { */ struct tty_struct **ttys; struct ktermios **termios; - struct ktermios **termios_locked; void *driver_state; /* From ba8438fb4e85bfe6cf8d3149fe63b85e49fdf217 Mon Sep 17 00:00:00 2001 From: Michael Neuling Date: Mon, 12 Dec 2011 09:49:14 +1100 Subject: [PATCH 506/820] powerpc: fix compile error with 85xx/p1022_ds.c Current linux-next compiled with mpc85xx_defconfig causes this: arch/powerpc/platforms/85xx/p1022_ds.c:341:14: error: 'udbg_progress' undeclared here (not in a function) Add include to fix this. Signed-off-by: Michael Neuling Acked-by: Timur Tabi Signed-off-by: Kumar Gala --- arch/powerpc/platforms/85xx/p1022_ds.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/powerpc/platforms/85xx/p1022_ds.c b/arch/powerpc/platforms/85xx/p1022_ds.c index bb3d84f4046f..b0984ada3f83 100644 --- a/arch/powerpc/platforms/85xx/p1022_ds.c +++ b/arch/powerpc/platforms/85xx/p1022_ds.c @@ -25,6 +25,7 @@ #include #include +#include #include #include "smp.h" From 0cf572dc00cd36250af9260377a0b5faac9b3284 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Thu, 12 Jan 2012 10:55:14 +0100 Subject: [PATCH 507/820] arch/powerpc/sysdev/fsl_pci.c: add missing iounmap Add missing iounmap in error handling code, in a case where the function already preforms iounmap on some other execution path. A simplified version of the semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @@ expression e; statement S,S1; int ret; @@ e = \(ioremap\|ioremap_nocache\)(...) ... when != iounmap(e) if (<+...e...+>) S ... when any when != iounmap(e) *if (...) { ... when != iounmap(e) return ...; } ... when any iounmap(e); // Signed-off-by: Julia Lawall Signed-off-by: Kumar Gala --- arch/powerpc/sysdev/fsl_pci.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/sysdev/fsl_pci.c b/arch/powerpc/sysdev/fsl_pci.c index 3b61e8cf3421..30eb17ecad49 100644 --- a/arch/powerpc/sysdev/fsl_pci.c +++ b/arch/powerpc/sysdev/fsl_pci.c @@ -205,12 +205,12 @@ static void __init setup_pci_atmu(struct pci_controller *hose, if (paddr_hi == paddr_lo) { pr_err("%s: No outbound window space\n", name); - return ; + goto out; } if (paddr_lo == 0) { pr_err("%s: No space for inbound window\n", name); - return ; + goto out; } /* setup PCSRBAR/PEXCSRBAR */ @@ -357,6 +357,7 @@ static void __init setup_pci_atmu(struct pci_controller *hose, (u64)hose->dma_window_size); } +out: iounmap(pci); } From 63ff07beaebef2a82da41cf04053ae88b579226d Mon Sep 17 00:00:00 2001 From: Yanmin Zhang Date: Tue, 10 Jan 2012 15:48:21 -0800 Subject: [PATCH 508/820] intel_idle: remove redundant local_irq_disable() call irq disabling happens earlier in process_32.c:cpu_idle. Basically, cpuidle_state->enter is called, cpu irq is disabled. cpuidle_state->enter would turn on irq when exiting. intel_idle doesn't follow this assumption. Although it doesn't cause real issue, it misleads developers. Remove the call to local_irq_disable() at entry. [akpm@linux-foundation.org: add comment] Signed-off-by: Mingming Zhang Signed-off-by: Andrew Morton Signed-off-by: Len Brown --- drivers/idle/intel_idle.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index af0b4a5d52f2..920c61c730a1 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -232,6 +232,7 @@ static int get_driver_data(int cstate) * @drv: cpuidle driver * @index: index of cpuidle state * + * Must be called under local_irq_disable(). */ static int intel_idle(struct cpuidle_device *dev, struct cpuidle_driver *drv, int index) @@ -247,8 +248,6 @@ static int intel_idle(struct cpuidle_device *dev, cstate = (((eax) >> MWAIT_SUBSTATE_SIZE) & MWAIT_CSTATE_MASK) + 1; - local_irq_disable(); - /* * leave_mm() to avoid costly and often unnecessary wakeups * for flushing the user TLB's associated with the active mm. From 2e92c7ad8f269c2b5b7f2a4763675f55f00b75f5 Mon Sep 17 00:00:00 2001 From: Masanari Iida Date: Thu, 15 Dec 2011 01:18:52 +0900 Subject: [PATCH 509/820] ACPI: kernel-parameters.txt : Add intel_idle.max_cstate Add missing intel_idle.max_cstate in kernel-parameters.txt Signed-off-by Masanari Iida Signed-off-by: Len Brown --- Documentation/kernel-parameters.txt | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 81c287fad79d..c6a56d8b901c 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1035,6 +1035,11 @@ bytes respectively. Such letter suffixes can also be entirely omitted. By default, super page will be supported if Intel IOMMU has the capability. With this option, super page will not be supported. + + intel_idle.max_cstate= [KNL,HW,ACPI,X86] + 0 disables intel_idle and fall back on acpi_idle. + 1 to 6 specify maximum depth of C-state. + intremap= [X86-64, Intel-IOMMU] on enable Interrupt Remapping (default) off disable Interrupt Remapping From 95e3ec11491d0cbce9fcdf1cc17a527c114c7dcf Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 15 Dec 2011 13:03:14 +0000 Subject: [PATCH 510/820] intel_idle: Fix a cast to pointer from integer of different size warning in intel_idle Fix the following warning: drivers/idle/intel_idle.c: In function 'intel_idle_cpuidle_devices_init': drivers/idle/intel_idle.c:518:5: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] By making get_driver_data() return a long instead of an int. Signed-off-by: David Howells Signed-off-by: Len Brown --- drivers/idle/intel_idle.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 920c61c730a1..18f9ad814104 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -197,7 +197,7 @@ static struct cpuidle_state atom_cstates[MWAIT_MAX_NUM_CSTATES] = { .enter = &intel_idle }, }; -static int get_driver_data(int cstate) +static long get_driver_data(int cstate) { int driver_data; switch (cstate) { From 5c2a9f06a9cd7194f884cdc88144866235dec07d Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Sun, 4 Dec 2011 22:17:29 +0100 Subject: [PATCH 511/820] intel idle: Make idle driver more robust kvm -cpu host passes the original cpuid info to the guest. Latest kvm version seem to return true for mwait_leaf cpuid function on recent Intel CPUs. But it does not return mwait C-states (mwait_substates), instead zero is returned. While real CPUs seem to always return non-zero values, the intel idle driver should not get active in kvm (mwait_substates == 0) case and bail out. Otherwise a Null pointer exception will happen later when the cpuidle subsystem tries to get active: [0.984807] BUG: unable to handle kernel NULL pointer dereference at (null) [0.984807] IP: [<(null)>] (null) ... [0.984807][] ? cpuidle_idle_call+0xb4/0x340 [0.984807][] ? __atomic_notifier_call_chain+0x4c/0x70 [0.984807][] ? cpu_idle+0x78/0xd0 Reference: https://bugzilla.novell.com/show_bug.cgi?id=726296 Cc: stable@vger.kernel.org Signed-off-by: Thomas Renninger CC: Bruno Friedmann Signed-off-by: Len Brown --- drivers/idle/intel_idle.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 18f9ad814104..91aff180ac47 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -347,7 +347,8 @@ static int intel_idle_probe(void) cpuid(CPUID_MWAIT_LEAF, &eax, &ebx, &ecx, &mwait_substates); if (!(ecx & CPUID5_ECX_EXTENSIONS_SUPPORTED) || - !(ecx & CPUID5_ECX_INTERRUPT_BREAK)) + !(ecx & CPUID5_ECX_INTERRUPT_BREAK) || + !mwait_substates) return -ENODEV; pr_debug(PREFIX "MWAIT substates: 0x%x\n", mwait_substates); From 5e7590d40dc59d5c1889d9e70c9da1f1df3918c6 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Tue, 17 Jan 2012 17:35:22 +0100 Subject: [PATCH 512/820] ACPI processor: Remove unneeded cpuidle_unregister_driver call Since commit 46bcfad7a819bd17ac4e831b04405152d59784ab registering and unregistering cpuidle is done in processor_idle.c. Unregistering via: acpi_bus_unregister_driver(&acpi_processor_driver) -> acpi_processor_remove() -> acpi_processor_power_exit() Remove not needed cpuidle_unregister_driver() call from acpi_processor_exit Signed-off-by: Thomas Renninger CC: Deepthi Dharwar Signed-off-by: Len Brown --- drivers/acpi/processor_driver.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index 90719d125e74..3616ce5509c8 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -828,8 +828,6 @@ static void __exit acpi_processor_exit(void) acpi_bus_unregister_driver(&acpi_processor_driver); - cpuidle_unregister_driver(&acpi_idle_driver); - return; } From ce79dac861e0d9a473d9923391bdbaad83c1c57f Mon Sep 17 00:00:00 2001 From: Ulrich Drepper Date: Tue, 17 Jan 2012 14:14:02 -0500 Subject: [PATCH 513/820] x86, opcode: ANDN and Group 17 in x86-opcode-map.txt The Intel documentation at http://software.intel.com/file/36945 shows the ANDN opcode and Group 17 with encoding f2 and f3 encoding respectively. The current version of x86-opcode-map.txt shows them with f3 and f4. Unless someone can point to documentation which shows the currently used encoding the following patch be applied. Signed-off-by: Ulrich Drepper Link: http://lkml.kernel.org/r/CAOPLpQdq5SuVo9=023CYhbFLAX9rONyjmYq7jJkqc5xwctW5eA@mail.gmail.com Signed-off-by: H. Peter Anvin --- arch/x86/lib/x86-opcode-map.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/lib/x86-opcode-map.txt b/arch/x86/lib/x86-opcode-map.txt index 5b83c51c12e0..4c8010d4f5e6 100644 --- a/arch/x86/lib/x86-opcode-map.txt +++ b/arch/x86/lib/x86-opcode-map.txt @@ -729,8 +729,8 @@ de: VAESDEC Vdq,Hdq,Wdq (66),(v1) df: VAESDECLAST Vdq,Hdq,Wdq (66),(v1) f0: MOVBE Gy,My | MOVBE Gw,Mw (66) | CRC32 Gd,Eb (F2) f1: MOVBE My,Gy | MOVBE Mw,Gw (66) | CRC32 Gd,Ey (F2) -f3: ANDN Gy,By,Ey (v) -f4: Grp17 (1A) +f2: ANDN Gy,By,Ey (v) +f3: Grp17 (1A) f5: BZHI Gy,Ey,By (v) | PEXT Gy,By,Ey (F3),(v) | PDEP Gy,By,Ey (F2),(v) f6: MULX By,Gy,rDX,Ey (F2),(v) f7: BEXTR Gy,Ey,By (v) | SHLX Gy,Ey,By (66),(v) | SARX Gy,Ey,By (F3),(v) | SHRX Gy,Ey,By (F2),(v) From 54b466e44b1c7809144bbd8cd6be3f85877ca46f Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 17 Jan 2012 21:26:11 +0100 Subject: [PATCH 514/820] cfq-iosched: fix use-after-free of cfqq With the changes in life time management between the cfq IO contexts and the cfq queues, we now risk having cfqd->active_queue being freed when cfq_slice_expired() is being called. cfq_preempt_queue() caches this queue and uses it after calling said function, causing a use-after-free condition. This triggers the following oops, when cfqq_type() attempts to dereference it: BUG: unable to handle kernel paging request at ffff8800746c4f0c IP: [] cfqq_type+0xb/0x20 PGD 18d4063 PUD 1fe15067 PMD 1ffb9067 PTE 80000000746c4160 Oops: 0000 [#1] SMP DEBUG_PAGEALLOC CPU 3 Modules linked in: Pid: 1, comm: init Not tainted 3.2.0-josef+ #367 Bochs Bochs RIP: 0010:[] [] cfqq_type+0xb/0x20 RSP: 0018:ffff880079c11778 EFLAGS: 00010046 RAX: 0000000000000000 RBX: ffff880076f3df08 RCX: 0000000000000000 RDX: 0000000000000006 RSI: ffff880074271888 RDI: ffff8800746c4f08 RBP: ffff880079c11778 R08: 0000000000000078 R09: 0000000000000001 R10: 09f911029d74e35b R11: 09f911029d74e35b R12: ffff880076f337f0 R13: ffff8800746c4f08 R14: ffff8800746c4f08 R15: 0000000000000002 FS: 00007f62fd44f700(0000) GS:ffff88007cd80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffff8800746c4f0c CR3: 0000000076c21000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process init (pid: 1, threadinfo ffff880079c10000, task ffff880079c0a040) Stack: ffff880079c117c8 ffffffff812683d8 ffff880079c117a8 ffffffff8125de43 ffff8800744fcf48 ffff880074b43e98 ffff8800770c8828 ffff880074b43e98 0000000000000003 0000000000000000 ffff880079c117f8 ffffffff81254149 Call Trace: [] cfq_insert_request+0x3f5/0x47c [] ? blk_recount_segments+0x20/0x31 [] __elv_add_request+0x1ca/0x200 [] blk_queue_bio+0x2ef/0x312 [] generic_make_request+0x9f/0xe0 [] submit_bio+0xbf/0xca [] submit_bh+0xdf/0xfe [] ext3_bread+0x50/0x99 [] dx_probe+0x38/0x291 [] ext3_dx_find_entry+0x58/0x219 [] ext3_find_entry+0xb0/0x406 [] ? cache_alloc_debugcheck_after.isra.46+0x14d/0x1a0 [] ? kmem_cache_alloc+0xef/0x191 [] ext3_lookup+0x39/0xe1 [] d_alloc_and_lookup+0x45/0x6c [] do_lookup+0x1e4/0x2f5 [] link_path_walk+0x1a4/0x6ef [] path_lookupat+0x59/0x5ea [] ? __strncpy_from_user+0x30/0x5a [] do_path_lookup+0x23/0x59 [] user_path_at_empty+0x53/0x99 [] ? remove_wait_queue+0x51/0x56 [] user_path_at+0x11/0x13 [] vfs_fstatat+0x3a/0x64 [] vfs_stat+0x1b/0x1d [] sys_newstat+0x1a/0x33 [] ? task_stopped_code+0x42/0x42 [] system_call_fastpath+0x16/0x1b Code: 89 e6 48 89 c7 e8 fa ca fe ff 85 c0 74 06 4c 89 2b 41 b6 01 5b 44 89 f0 41 5c 41 5d 41 5e 5d c3 55 48 89 e5 66 66 66 66 90 31 c0 <8b> 57 04 f6 c6 01 74 0b 83 e2 20 83 fa 01 19 c0 83 c0 02 5d c3 RIP [] cfqq_type+0xb/0x20 RSP CR2: ffff8800746c4f0c Get rid of the caching of cfqd->active_queue, and reorder the check so that it happens before we expire the active queue. Thanks to Tejun for pin pointing the error location. Reported-by: Chris Mason Tested-by: Chris Mason Signed-off-by: Jens Axboe --- block/cfq-iosched.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 163263ddd381..ee55019066a1 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -3117,18 +3117,17 @@ cfq_should_preempt(struct cfq_data *cfqd, struct cfq_queue *new_cfqq, */ static void cfq_preempt_queue(struct cfq_data *cfqd, struct cfq_queue *cfqq) { - struct cfq_queue *old_cfqq = cfqd->active_queue; - cfq_log_cfqq(cfqd, cfqq, "preempt"); - cfq_slice_expired(cfqd, 1); /* * workload type is changed, don't save slice, otherwise preempt * doesn't happen */ - if (cfqq_type(old_cfqq) != cfqq_type(cfqq)) + if (cfqq_type(cfqd->active_queue) != cfqq_type(cfqq)) cfqq->cfqg->saved_workload_slice = 0; + cfq_slice_expired(cfqd, 1); + /* * Put the new queue at the front of the of the current list, * so we know that it will be selected next. From 36c92474498ad6cc3afb24b91a67a444d79978fe Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 17 Jan 2012 07:57:56 +0000 Subject: [PATCH 515/820] net: WARN if skb_checksum_help() is called on skb requiring segmentation skb_checksum_help() has never done anything useful with skbs that require segmentation. Setting skb->ip_summed = CHECKSUM_NONE makes them invalid and provokes a later WARNing in skb_gso_segment(). Passing such an skb to skb_checksum_help() indicates a bug, so we should warn about it immediately. Move the warning from skb_gso_segment() into a shared function, and add gso_type and gso_size to it. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller --- net/core/dev.c | 31 +++++++++++++++++++------------ 1 file changed, 19 insertions(+), 12 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 7e6b7dcaacde..17db2f2e5236 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1887,6 +1887,22 @@ void skb_set_dev(struct sk_buff *skb, struct net_device *dev) EXPORT_SYMBOL(skb_set_dev); #endif /* CONFIG_NET_NS */ +static void skb_warn_bad_offload(const struct sk_buff *skb) +{ + struct net_device *dev = skb->dev; + const char *driver = ""; + + if (dev && dev->dev.parent) + driver = dev_driver_string(dev->dev.parent); + + WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d gso_size=%d " + "gso_type=%d ip_summed=%d\n", + driver, dev ? &dev->features : NULL, + skb->sk ? &skb->sk->sk_route_caps : NULL, + skb->len, skb->data_len, skb_shinfo(skb)->gso_size, + skb_shinfo(skb)->gso_type, skb->ip_summed); +} + /* * Invalidate hardware checksum when packet is to be mangled, and * complete checksum manually on outgoing path. @@ -1900,8 +1916,8 @@ int skb_checksum_help(struct sk_buff *skb) goto out_set_summed; if (unlikely(skb_shinfo(skb)->gso_size)) { - /* Let GSO fix up the checksum. */ - goto out_set_summed; + skb_warn_bad_offload(skb); + return -EINVAL; } offset = skb_checksum_start_offset(skb); @@ -1961,16 +1977,7 @@ struct sk_buff *skb_gso_segment(struct sk_buff *skb, __skb_pull(skb, skb->mac_len); if (unlikely(skb->ip_summed != CHECKSUM_PARTIAL)) { - struct net_device *dev = skb->dev; - const char *driver = ""; - - if (dev && dev->dev.parent) - driver = dev_driver_string(dev->dev.parent); - - WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d ip_summed=%d\n", - driver, dev ? &dev->features : NULL, - skb->sk ? &skb->sk->sk_route_caps : NULL, - skb->len, skb->data_len, skb->ip_summed); + skb_warn_bad_offload(skb); if (skb_header_cloned(skb) && (err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC))) From 65e9d2faab70d07b9a38ac6ed298f191d24541fc Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20Miros=C5=82aw?= Date: Tue, 17 Jan 2012 10:00:40 +0000 Subject: [PATCH 516/820] net: fix NULL-deref in WARN() in skb_gso_segment() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Bug was introduced in commit c8f44affb7244f2ac3e703cab13d55ede27621bb. Signed-off-by: Michał Mirosław Signed-off-by: David S. Miller --- net/core/dev.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 17db2f2e5236..115dee1d985d 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1889,6 +1889,7 @@ EXPORT_SYMBOL(skb_set_dev); static void skb_warn_bad_offload(const struct sk_buff *skb) { + static const netdev_features_t null_features = 0; struct net_device *dev = skb->dev; const char *driver = ""; @@ -1897,8 +1898,8 @@ static void skb_warn_bad_offload(const struct sk_buff *skb) WARN(1, "%s: caps=(%pNF, %pNF) len=%d data_len=%d gso_size=%d " "gso_type=%d ip_summed=%d\n", - driver, dev ? &dev->features : NULL, - skb->sk ? &skb->sk->sk_route_caps : NULL, + driver, dev ? &dev->features : &null_features, + skb->sk ? &skb->sk->sk_route_caps : &null_features, skb->len, skb->data_len, skb_shinfo(skb)->gso_size, skb_shinfo(skb)->gso_type, skb->ip_summed); } From 10ec1bb7e9eb462548f14dd53c73e927a3ddf31c Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 17 Jan 2012 10:48:43 +0000 Subject: [PATCH 517/820] inetpeer: initialize ->redirect_genid in inet_getpeer() kmemcheck complains that ->redirect_genid doesn't get initialized. Presumably it should be set to zero. Signed-off-by: Dan Carpenter Signed-off-by: David S. Miller --- net/ipv4/inetpeer.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index 986a02bda660..bf4a9c4808e1 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -447,6 +447,7 @@ relookup: p->rate_last = 0; p->pmtu_expires = 0; p->pmtu_orig = 0; + p->redirect_genid = 0; memset(&p->redirect_learned, 0, sizeof(p->redirect_learned)); From 8096b1ebb59b94b3bc6abb6b7d121419e83447ba Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:07 +0000 Subject: [PATCH 518/820] xfs: remove the if_ext_max field in struct xfs_ifork We spent a lot of effort to maintain this field, but it always equals to the fork size divided by the constant size of an extent. The prime use of it is to assert that the two stay in sync. Just divide the fork size by the extent size in the few places that we actually use it and remove the overhead of maintaining it. Also introduce a few helpers to consolidate the places where we actually care about the value. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_attr_leaf.c | 9 ---- fs/xfs/xfs_bmap.c | 101 +++++++++++++++++++--------------------- fs/xfs/xfs_dfrag.c | 43 +++++++++-------- fs/xfs/xfs_iget.c | 2 - fs/xfs/xfs_inode.c | 30 ++++-------- fs/xfs/xfs_inode.h | 4 +- fs/xfs/xfs_inode_item.c | 2 - fs/xfs/xfs_trace.h | 5 +- 8 files changed, 81 insertions(+), 115 deletions(-) diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index c1b55e596551..d25eafd4d28d 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -271,10 +271,6 @@ xfs_attr_shortform_add(xfs_da_args_t *args, int forkoff) dp = args->dp; mp = dp->i_mount; dp->i_d.di_forkoff = forkoff; - dp->i_df.if_ext_max = - XFS_IFORK_DSIZE(dp) / (uint)sizeof(xfs_bmbt_rec_t); - dp->i_afp->if_ext_max = - XFS_IFORK_ASIZE(dp) / (uint)sizeof(xfs_bmbt_rec_t); ifp = dp->i_afp; ASSERT(ifp->if_flags & XFS_IFINLINE); @@ -326,7 +322,6 @@ xfs_attr_fork_reset( ASSERT(ip->i_d.di_anextents == 0); ASSERT(ip->i_afp == NULL); - ip->i_df.if_ext_max = XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); } @@ -389,10 +384,6 @@ xfs_attr_shortform_remove(xfs_da_args_t *args) (args->op_flags & XFS_DA_OP_ADDNAME) || !(mp->m_flags & XFS_MOUNT_ATTR2) || dp->i_d.di_format == XFS_DINODE_FMT_BTREE); - dp->i_afp->if_ext_max = - XFS_IFORK_ASIZE(dp) / (uint)sizeof(xfs_bmbt_rec_t); - dp->i_df.if_ext_max = - XFS_IFORK_DSIZE(dp) / (uint)sizeof(xfs_bmbt_rec_t); xfs_trans_log_inode(args->trans, dp, XFS_ILOG_CORE | XFS_ILOG_ADATA); } diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index d0ab78837057..7a888ca2f7f6 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -249,7 +249,27 @@ xfs_bmbt_lookup_ge( } /* -* Update the record referred to by cur to the value given + * Check if the inode needs to be converted to btree format. + */ +static inline bool xfs_bmap_needs_btree(struct xfs_inode *ip, int whichfork) +{ + return XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS && + XFS_IFORK_NEXTENTS(ip, whichfork) > + XFS_IFORK_MAXEXT(ip, whichfork); +} + +/* + * Check if the inode should be converted to extent format. + */ +static inline bool xfs_bmap_wants_extents(struct xfs_inode *ip, int whichfork) +{ + return XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE && + XFS_IFORK_NEXTENTS(ip, whichfork) <= + XFS_IFORK_MAXEXT(ip, whichfork); +} + +/* + * Update the record referred to by cur to the value given * by [off, bno, len, state]. * This either works (return 0) or gets an EFSCORRUPTED error. */ @@ -683,8 +703,8 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - if (bma->ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS && - bma->ip->i_d.di_nextents > bma->ip->i_df.if_ext_max) { + + if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) { error = xfs_bmap_extents_to_btree(bma->tp, bma->ip, bma->firstblock, bma->flist, &bma->cur, 1, &tmp_rval, XFS_DATA_FORK); @@ -767,8 +787,8 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - if (bma->ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS && - bma->ip->i_d.di_nextents > bma->ip->i_df.if_ext_max) { + + if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) { error = xfs_bmap_extents_to_btree(bma->tp, bma->ip, bma->firstblock, bma->flist, &bma->cur, 1, &tmp_rval, XFS_DATA_FORK); @@ -836,8 +856,8 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - if (bma->ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS && - bma->ip->i_d.di_nextents > bma->ip->i_df.if_ext_max) { + + if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) { error = xfs_bmap_extents_to_btree(bma->tp, bma->ip, bma->firstblock, bma->flist, &bma->cur, 1, &tmp_rval, XFS_DATA_FORK); @@ -884,8 +904,7 @@ xfs_bmap_add_extent_delay_real( } /* convert to a btree if necessary */ - if (XFS_IFORK_FORMAT(bma->ip, XFS_DATA_FORK) == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(bma->ip, XFS_DATA_FORK) > ifp->if_ext_max) { + if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) { int tmp_logflags; /* partial log flag return val */ ASSERT(bma->cur == NULL); @@ -1421,8 +1440,7 @@ xfs_bmap_add_extent_unwritten_real( } /* convert to a btree if necessary */ - if (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) > ifp->if_ext_max) { + if (xfs_bmap_needs_btree(ip, XFS_DATA_FORK)) { int tmp_logflags; /* partial log flag return val */ ASSERT(cur == NULL); @@ -1812,8 +1830,7 @@ xfs_bmap_add_extent_hole_real( } /* convert to a btree if necessary */ - if (XFS_IFORK_FORMAT(bma->ip, whichfork) == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(bma->ip, whichfork) > ifp->if_ext_max) { + if (xfs_bmap_needs_btree(bma->ip, whichfork)) { int tmp_logflags; /* partial log flag return val */ ASSERT(bma->cur == NULL); @@ -3037,8 +3054,7 @@ xfs_bmap_extents_to_btree( ifp = XFS_IFORK_PTR(ip, whichfork); ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS); - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); + /* * Make space in the inode incore. */ @@ -3184,13 +3200,8 @@ xfs_bmap_forkoff_reset( ip->i_d.di_format != XFS_DINODE_FMT_BTREE) { uint dfl_forkoff = xfs_default_attroffset(ip) >> 3; - if (dfl_forkoff > ip->i_d.di_forkoff) { + if (dfl_forkoff > ip->i_d.di_forkoff) ip->i_d.di_forkoff = dfl_forkoff; - ip->i_df.if_ext_max = - XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t); - ip->i_afp->if_ext_max = - XFS_IFORK_ASIZE(ip) / sizeof(xfs_bmbt_rec_t); - } } } @@ -3430,8 +3441,6 @@ xfs_bmap_add_attrfork( int error; /* error return value */ ASSERT(XFS_IFORK_Q(ip) == 0); - ASSERT(ip->i_df.if_ext_max == - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t)); mp = ip->i_mount; ASSERT(!XFS_NOT_DQATTACHED(mp, ip)); @@ -3486,12 +3495,9 @@ xfs_bmap_add_attrfork( error = XFS_ERROR(EINVAL); goto error1; } - ip->i_df.if_ext_max = - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); + ASSERT(ip->i_afp == NULL); ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP); - ip->i_afp->if_ext_max = - XFS_IFORK_ASIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); ip->i_afp->if_flags = XFS_IFEXTENTS; logflags = 0; xfs_bmap_init(&flist, &firstblock); @@ -3535,20 +3541,17 @@ xfs_bmap_add_attrfork( } else spin_unlock(&mp->m_sb_lock); } - if ((error = xfs_bmap_finish(&tp, &flist, &committed))) + + error = xfs_bmap_finish(&tp, &flist, &committed); + if (error) goto error2; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); - ASSERT(ip->i_df.if_ext_max == - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t)); - return error; + return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); error2: xfs_bmap_cancel(&flist); error1: xfs_iunlock(ip, XFS_ILOCK_EXCL); error0: xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); - ASSERT(ip->i_df.if_ext_max == - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t)); return error; } @@ -4379,8 +4382,6 @@ xfs_bmapi_read( XFS_STATS_INC(xs_blk_mapr); ifp = XFS_IFORK_PTR(ip, whichfork); - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); if (!(ifp->if_flags & XFS_IFEXTENTS)) { error = xfs_iread_extents(NULL, ip, whichfork); @@ -4871,8 +4872,6 @@ xfs_bmapi_write( return XFS_ERROR(EIO); ifp = XFS_IFORK_PTR(ip, whichfork); - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); XFS_STATS_INC(xs_blk_mapw); @@ -4981,8 +4980,7 @@ xfs_bmapi_write( /* * Transform from btree to extents, give it cur. */ - if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE && - XFS_IFORK_NEXTENTS(ip, whichfork) <= ifp->if_ext_max) { + if (xfs_bmap_wants_extents(ip, whichfork)) { int tmp_logflags = 0; ASSERT(bma.cur); @@ -4992,10 +4990,10 @@ xfs_bmapi_write( if (error) goto error0; } - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); + ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE || - XFS_IFORK_NEXTENTS(ip, whichfork) > ifp->if_ext_max); + XFS_IFORK_NEXTENTS(ip, whichfork) > + XFS_IFORK_MAXEXT(ip, whichfork)); error = 0; error0: /* @@ -5095,8 +5093,7 @@ xfs_bunmapi( ASSERT(len > 0); ASSERT(nexts >= 0); - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); + if (!(ifp->if_flags & XFS_IFEXTENTS) && (error = xfs_iread_extents(tp, ip, whichfork))) return error; @@ -5322,7 +5319,8 @@ xfs_bunmapi( */ if (!wasdel && xfs_trans_get_block_res(tp) == 0 && XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(ip, whichfork) >= ifp->if_ext_max && + XFS_IFORK_NEXTENTS(ip, whichfork) >= /* Note the >= */ + XFS_IFORK_MAXEXT(ip, whichfork) && del.br_startoff > got.br_startoff && del.br_startoff + del.br_blockcount < got.br_startoff + got.br_blockcount) { @@ -5353,13 +5351,11 @@ nodelete: } } *done = bno == (xfs_fileoff_t)-1 || bno < start || lastx < 0; - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); + /* * Convert to a btree if necessary. */ - if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(ip, whichfork) > ifp->if_ext_max) { + if (xfs_bmap_needs_btree(ip, whichfork)) { ASSERT(cur == NULL); error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist, &cur, 0, &tmp_logflags, whichfork); @@ -5370,8 +5366,7 @@ nodelete: /* * transform from btree to extents, give it cur */ - else if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE && - XFS_IFORK_NEXTENTS(ip, whichfork) <= ifp->if_ext_max) { + else if (xfs_bmap_wants_extents(ip, whichfork)) { ASSERT(cur != NULL); error = xfs_bmap_btree_to_extents(tp, ip, cur, &tmp_logflags, whichfork); @@ -5382,8 +5377,6 @@ nodelete: /* * transform from extents to local? */ - ASSERT(ifp->if_ext_max == - XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); error = 0; error0: /* diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index 654dc6f05bac..dd974a55c77d 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -163,12 +163,14 @@ xfs_swap_extents_check_format( /* Check temp in extent form to max in target */ if (tip->i_d.di_format == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK) > ip->i_df.if_ext_max) + XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK) > + XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)) return EINVAL; /* Check target in extent form to max in temp */ if (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS && - XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) > tip->i_df.if_ext_max) + XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) > + XFS_IFORK_MAXEXT(tip, XFS_DATA_FORK)) return EINVAL; /* @@ -180,18 +182,25 @@ xfs_swap_extents_check_format( * (a common defrag case) which will occur when the temp inode is in * extent format... */ - if (tip->i_d.di_format == XFS_DINODE_FMT_BTREE && - ((XFS_IFORK_BOFF(ip) && - tip->i_df.if_broot_bytes > XFS_IFORK_BOFF(ip)) || - XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK) <= ip->i_df.if_ext_max)) - return EINVAL; + if (tip->i_d.di_format == XFS_DINODE_FMT_BTREE) { + if (XFS_IFORK_BOFF(ip) && + tip->i_df.if_broot_bytes > XFS_IFORK_BOFF(ip)) + return EINVAL; + if (XFS_IFORK_NEXTENTS(tip, XFS_DATA_FORK) <= + XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)) + return EINVAL; + } /* Reciprocal target->temp btree format checks */ - if (ip->i_d.di_format == XFS_DINODE_FMT_BTREE && - ((XFS_IFORK_BOFF(tip) && - ip->i_df.if_broot_bytes > XFS_IFORK_BOFF(tip)) || - XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) <= tip->i_df.if_ext_max)) - return EINVAL; + if (ip->i_d.di_format == XFS_DINODE_FMT_BTREE) { + if (XFS_IFORK_BOFF(tip) && + ip->i_df.if_broot_bytes > XFS_IFORK_BOFF(tip)) + return EINVAL; + + if (XFS_IFORK_NEXTENTS(ip, XFS_DATA_FORK) <= + XFS_IFORK_MAXEXT(tip, XFS_DATA_FORK)) + return EINVAL; + } return 0; } @@ -348,16 +357,6 @@ xfs_swap_extents( *ifp = *tifp; /* struct copy */ *tifp = *tempifp; /* struct copy */ - /* - * Fix the in-memory data fork values that are dependent on the fork - * offset in the inode. We can't assume they remain the same as attr2 - * has dynamic fork offsets. - */ - ifp->if_ext_max = XFS_IFORK_SIZE(ip, XFS_DATA_FORK) / - (uint)sizeof(xfs_bmbt_rec_t); - tifp->if_ext_max = XFS_IFORK_SIZE(tip, XFS_DATA_FORK) / - (uint)sizeof(xfs_bmbt_rec_t); - /* * Fix the on-disk inode values */ diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 3960a066d7ff..f180ce896cd7 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -450,8 +450,6 @@ again: *ipp = ip; - ASSERT(ip->i_df.if_ext_max == - XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t)); /* * If we have a real type for an on-disk inode, we can set ops(&unlock) * now. If it's a new inode being created, xfs_ialloc will handle it. diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index ccd619a993f6..96b29e3286db 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -299,11 +299,8 @@ xfs_iformat( { xfs_attr_shortform_t *atp; int size; - int error; + int error = 0; xfs_fsize_t di_size; - ip->i_df.if_ext_max = - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); - error = 0; if (unlikely(be32_to_cpu(dip->di_nextents) + be16_to_cpu(dip->di_anextents) > @@ -409,10 +406,10 @@ xfs_iformat( } if (!XFS_DFORK_Q(dip)) return 0; + ASSERT(ip->i_afp == NULL); ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP | KM_NOFS); - ip->i_afp->if_ext_max = - XFS_IFORK_ASIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); + switch (dip->di_aformat) { case XFS_DINODE_FMT_LOCAL: atp = (xfs_attr_shortform_t *)XFS_DFORK_APTR(dip); @@ -604,10 +601,11 @@ xfs_iformat_btree( * or the number of extents is greater than the number of * blocks. */ - if (unlikely(XFS_IFORK_NEXTENTS(ip, whichfork) <= ifp->if_ext_max - || XFS_BMDR_SPACE_CALC(nrecs) > - XFS_DFORK_SIZE(dip, ip->i_mount, whichfork) - || XFS_IFORK_NEXTENTS(ip, whichfork) > ip->i_d.di_nblocks)) { + if (unlikely(XFS_IFORK_NEXTENTS(ip, whichfork) <= + XFS_IFORK_MAXEXT(ip, whichfork) || + XFS_BMDR_SPACE_CALC(nrecs) > + XFS_DFORK_SIZE(dip, ip->i_mount, whichfork) || + XFS_IFORK_NEXTENTS(ip, whichfork) > ip->i_d.di_nblocks)) { xfs_warn(ip->i_mount, "corrupt inode %Lu (btree).", (unsigned long long) ip->i_ino); XFS_CORRUPTION_ERROR("xfs_iformat_btree", XFS_ERRLEVEL_LOW, @@ -835,12 +833,6 @@ xfs_iread( * with the uninitialized part of it. */ ip->i_d.di_mode = 0; - /* - * Initialize the per-fork minima and maxima for a new - * inode here. xfs_iformat will do it for old inodes. - */ - ip->i_df.if_ext_max = - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); } /* @@ -1740,8 +1732,6 @@ xfs_ifree( ip->i_d.di_flags = 0; ip->i_d.di_dmevmask = 0; ip->i_d.di_forkoff = 0; /* mark the attr fork not in use */ - ip->i_df.if_ext_max = - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); ip->i_d.di_format = XFS_DINODE_FMT_EXTENTS; ip->i_d.di_aformat = XFS_DINODE_FMT_EXTENTS; /* @@ -2408,7 +2398,7 @@ xfs_iflush( ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); ASSERT(!completion_done(&ip->i_flush)); ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE || - ip->i_d.di_nextents > ip->i_df.if_ext_max); + ip->i_d.di_nextents > XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)); iip = ip->i_itemp; mp = ip->i_mount; @@ -2524,7 +2514,7 @@ xfs_iflush_int( ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); ASSERT(!completion_done(&ip->i_flush)); ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE || - ip->i_d.di_nextents > ip->i_df.if_ext_max); + ip->i_d.di_nextents > XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)); iip = ip->i_itemp; mp = ip->i_mount; diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index c1574721a191..47497e11c111 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -66,7 +66,6 @@ typedef struct xfs_ifork { struct xfs_btree_block *if_broot; /* file's incore btree root */ short if_broot_bytes; /* bytes allocated for root */ unsigned char if_flags; /* per-fork flags */ - unsigned char if_ext_max; /* max # of extent records */ union { xfs_bmbt_rec_host_t *if_extents;/* linear map file exts */ xfs_ext_irec_t *if_ext_irec; /* irec map file exts */ @@ -206,7 +205,8 @@ typedef struct xfs_icdinode { ((w) == XFS_DATA_FORK ? \ ((ip)->i_d.di_nextents = (n)) : \ ((ip)->i_d.di_anextents = (n))) - +#define XFS_IFORK_MAXEXT(ip, w) \ + (XFS_IFORK_SIZE(ip, w) / sizeof(xfs_bmbt_rec_t)) #ifdef __KERNEL__ diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index cfd6c7f8cc3c..2b6b4fcef49e 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -79,8 +79,6 @@ xfs_inode_item_size( break; case XFS_DINODE_FMT_BTREE: - ASSERT(ip->i_df.if_ext_max == - XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t)); iip->ili_format.ilf_fields &= ~(XFS_ILOG_DDATA | XFS_ILOG_DEXT | XFS_ILOG_DEV | XFS_ILOG_UUID); diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 297f9fa6fb64..81efa0416173 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -1568,7 +1568,6 @@ DECLARE_EVENT_CLASS(xfs_swap_extent_class, __field(xfs_ino_t, ino) __field(int, format) __field(int, nex) - __field(int, max_nex) __field(int, broot_size) __field(int, fork_off) ), @@ -1578,18 +1577,16 @@ DECLARE_EVENT_CLASS(xfs_swap_extent_class, __entry->ino = ip->i_ino; __entry->format = ip->i_d.di_format; __entry->nex = ip->i_d.di_nextents; - __entry->max_nex = ip->i_df.if_ext_max; __entry->broot_size = ip->i_df.if_broot_bytes; __entry->fork_off = XFS_IFORK_BOFF(ip); ), TP_printk("dev %d:%d ino 0x%llx (%s), %s format, num_extents %d, " - "Max in-fork extents %d, broot size %d, fork offset %d", + "broot size %d, fork offset %d", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __print_symbolic(__entry->which, XFS_SWAPEXT_INODES), __print_symbolic(__entry->format, XFS_INODE_FORMAT_STR), __entry->nex, - __entry->max_nex, __entry->broot_size, __entry->fork_off) ) From 49e4c70e52a2bc2090e5a4e003e2888af21d6a2b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:08 +0000 Subject: [PATCH 519/820] xfs: make i_flags an unsigned long To be used for bit wakeup i_flags needs to be an unsigned long or we'll run into trouble on big endian systems. Because of the 1-byte i_update field right after it this actually causes a fairly large size increase on its own (4 or 8 bytes), but that increase will be more than offset by the next two patches. Signed-off-by: Christoph Hellwig Reviewed-by: Alex Elder Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_inode.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 47497e11c111..be8dc0c2cf52 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -242,7 +242,7 @@ typedef struct xfs_inode { wait_queue_head_t i_ipin_wait; /* inode pinning wait queue */ spinlock_t i_flags_lock; /* inode i_flags lock */ /* Miscellaneous state. */ - unsigned short i_flags; /* see defined flags below */ + unsigned long i_flags; /* see defined flags below */ unsigned char i_update_core; /* timestamps/size is dirty */ unsigned int i_delayed_blks; /* count of delay alloc blks */ From 474fce067521a40dbacc722e8ba119e81c2d31bf Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:09 +0000 Subject: [PATCH 520/820] xfs: replace i_flock with a sleeping bitlock We almost never block on i_flock, the exception is synchronous inode flushing. Instead of bloating the inode with a 16/24-byte completion that we abuse as a semaphore just implement it as a bitlock that uses a bit waitqueue for the rare sleeping path. This primarily is a tradeoff between a much smaller inode and a faster non-blocking path vs faster wakeups, and we are much better off with the former. A small downside is that we will lose lockdep checking for i_flock, but given that it's always taken inside the ilock that should be acceptable. Note that for example the inode writeback locking is implemented in a very similar way. Signed-off-by: Christoph Hellwig Reviewed-by: Alex Elder Signed-off-by: Ben Myers --- fs/xfs/xfs_iget.c | 20 +++++++++-- fs/xfs/xfs_inode.c | 4 +-- fs/xfs/xfs_inode.h | 78 ++++++++++++++++++++++++++--------------- fs/xfs/xfs_inode_item.c | 4 +-- fs/xfs/xfs_super.c | 7 ---- fs/xfs/xfs_sync.c | 9 +++-- 6 files changed, 76 insertions(+), 46 deletions(-) diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index f180ce896cd7..a7cf7139f9ad 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -77,7 +77,7 @@ xfs_inode_alloc( ASSERT(atomic_read(&ip->i_pincount) == 0); ASSERT(!spin_is_locked(&ip->i_flags_lock)); - ASSERT(completion_done(&ip->i_flush)); + ASSERT(!xfs_isiflocked(ip)); ASSERT(ip->i_ino == 0); mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); @@ -150,7 +150,7 @@ xfs_inode_free( /* asserts to verify all state is correct here */ ASSERT(atomic_read(&ip->i_pincount) == 0); ASSERT(!spin_is_locked(&ip->i_flags_lock)); - ASSERT(completion_done(&ip->i_flush)); + ASSERT(!xfs_isiflocked(ip)); /* * Because we use RCU freeing we need to ensure the inode always @@ -713,3 +713,19 @@ xfs_isilocked( return 0; } #endif + +void +__xfs_iflock( + struct xfs_inode *ip) +{ + wait_queue_head_t *wq = bit_waitqueue(&ip->i_flags, __XFS_IFLOCK_BIT); + DEFINE_WAIT_BIT(wait, &ip->i_flags, __XFS_IFLOCK_BIT); + + do { + prepare_to_wait_exclusive(wq, &wait.wait, TASK_UNINTERRUPTIBLE); + if (xfs_isiflocked(ip)) + io_schedule(); + } while (!xfs_iflock_nowait(ip)); + + finish_wait(wq, &wait.wait); +} diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 96b29e3286db..eeb60d31b086 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -2396,7 +2396,7 @@ xfs_iflush( XFS_STATS_INC(xs_iflush_count); ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); - ASSERT(!completion_done(&ip->i_flush)); + ASSERT(xfs_isiflocked(ip)); ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE || ip->i_d.di_nextents > XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)); @@ -2512,7 +2512,7 @@ xfs_iflush_int( #endif ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); - ASSERT(!completion_done(&ip->i_flush)); + ASSERT(xfs_isiflocked(ip)); ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE || ip->i_d.di_nextents > XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK)); diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index be8dc0c2cf52..960d2a89b3ac 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -237,7 +237,6 @@ typedef struct xfs_inode { struct xfs_inode_log_item *i_itemp; /* logging information */ mrlock_t i_lock; /* inode lock */ mrlock_t i_iolock; /* inode IO lock */ - struct completion i_flush; /* inode flush completion q */ atomic_t i_pincount; /* inode pin count */ wait_queue_head_t i_ipin_wait; /* inode pinning wait queue */ spinlock_t i_flags_lock; /* inode i_flags lock */ @@ -324,6 +323,19 @@ xfs_iflags_test_and_clear(xfs_inode_t *ip, unsigned short flags) return ret; } +static inline int +xfs_iflags_test_and_set(xfs_inode_t *ip, unsigned short flags) +{ + int ret; + + spin_lock(&ip->i_flags_lock); + ret = ip->i_flags & flags; + if (!ret) + ip->i_flags |= flags; + spin_unlock(&ip->i_flags_lock); + return ret; +} + /* * Project quota id helpers (previously projid was 16bit only * and using two 16bit values to hold new 32bit projid was chosen @@ -343,36 +355,18 @@ xfs_set_projid(struct xfs_inode *ip, ip->i_d.di_projid_lo = (__uint16_t) (projid & 0xffff); } -/* - * Manage the i_flush queue embedded in the inode. This completion - * queue synchronizes processes attempting to flush the in-core - * inode back to disk. - */ -static inline void xfs_iflock(xfs_inode_t *ip) -{ - wait_for_completion(&ip->i_flush); -} - -static inline int xfs_iflock_nowait(xfs_inode_t *ip) -{ - return try_wait_for_completion(&ip->i_flush); -} - -static inline void xfs_ifunlock(xfs_inode_t *ip) -{ - complete(&ip->i_flush); -} - /* * In-core inode flags. */ -#define XFS_IRECLAIM 0x0001 /* started reclaiming this inode */ -#define XFS_ISTALE 0x0002 /* inode has been staled */ -#define XFS_IRECLAIMABLE 0x0004 /* inode can be reclaimed */ -#define XFS_INEW 0x0008 /* inode has just been allocated */ -#define XFS_IFILESTREAM 0x0010 /* inode is in a filestream directory */ -#define XFS_ITRUNCATED 0x0020 /* truncated down so flush-on-close */ -#define XFS_IDIRTY_RELEASE 0x0040 /* dirty release already seen */ +#define XFS_IRECLAIM (1 << 0) /* started reclaiming this inode */ +#define XFS_ISTALE (1 << 1) /* inode has been staled */ +#define XFS_IRECLAIMABLE (1 << 2) /* inode can be reclaimed */ +#define XFS_INEW (1 << 3) /* inode has just been allocated */ +#define XFS_IFILESTREAM (1 << 4) /* inode is in a filestream dir. */ +#define XFS_ITRUNCATED (1 << 5) /* truncated down so flush-on-close */ +#define XFS_IDIRTY_RELEASE (1 << 6) /* dirty release already seen */ +#define __XFS_IFLOCK_BIT 7 /* inode is being flushed right now */ +#define XFS_IFLOCK (1 << __XFS_IFLOCK_BIT) /* * Per-lifetime flags need to be reset when re-using a reclaimable inode during @@ -384,6 +378,34 @@ static inline void xfs_ifunlock(xfs_inode_t *ip) XFS_IDIRTY_RELEASE | XFS_ITRUNCATED | \ XFS_IFILESTREAM); +/* + * Synchronize processes attempting to flush the in-core inode back to disk. + */ + +extern void __xfs_iflock(struct xfs_inode *ip); + +static inline int xfs_iflock_nowait(struct xfs_inode *ip) +{ + return !xfs_iflags_test_and_set(ip, XFS_IFLOCK); +} + +static inline void xfs_iflock(struct xfs_inode *ip) +{ + if (!xfs_iflock_nowait(ip)) + __xfs_iflock(ip); +} + +static inline void xfs_ifunlock(struct xfs_inode *ip) +{ + xfs_iflags_clear(ip, XFS_IFLOCK); + wake_up_bit(&ip->i_flags, __XFS_IFLOCK_BIT); +} + +static inline int xfs_isiflocked(struct xfs_inode *ip) +{ + return xfs_iflags_test(ip, XFS_IFLOCK); +} + /* * Flags for inode locking. * Bit ranges: 1<<1 - 1<<16-1 -- iolock/ilock modes (bitfield) diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index 2b6b4fcef49e..c8d4ce0efd5a 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -717,7 +717,7 @@ xfs_inode_item_pushbuf( * If a flush is not in progress anymore, chances are that the * inode was taken off the AIL. So, just get out. */ - if (completion_done(&ip->i_flush) || + if (!xfs_isiflocked(ip) || !(lip->li_flags & XFS_LI_IN_AIL)) { xfs_iunlock(ip, XFS_ILOCK_SHARED); return true; @@ -750,7 +750,7 @@ xfs_inode_item_push( struct xfs_inode *ip = iip->ili_inode; ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED)); - ASSERT(!completion_done(&ip->i_flush)); + ASSERT(xfs_isiflocked(ip)); /* * Since we were able to lock the inode's flush lock and diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 281961c1d81a..6851fa7b1afa 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -829,13 +829,6 @@ xfs_fs_inode_init_once( atomic_set(&ip->i_pincount, 0); spin_lock_init(&ip->i_flags_lock); init_waitqueue_head(&ip->i_ipin_wait); - /* - * Because we want to use a counting completion, complete - * the flush completion once to allow a single access to - * the flush completion without blocking. - */ - init_completion(&ip->i_flush); - complete(&ip->i_flush); mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER, "xfsino", ip->i_ino); diff --git a/fs/xfs/xfs_sync.c b/fs/xfs/xfs_sync.c index 72c01a1c16e7..40b75eecd2b4 100644 --- a/fs/xfs/xfs_sync.c +++ b/fs/xfs/xfs_sync.c @@ -707,14 +707,13 @@ xfs_reclaim_inode_grab( return 1; /* - * do some unlocked checks first to avoid unnecessary lock traffic. - * The first is a flush lock check, the second is a already in reclaim - * check. Only do these checks if we are not going to block on locks. + * If we are asked for non-blocking operation, do unlocked checks to + * see if the inode already is being flushed or in reclaim to avoid + * lock traffic. */ if ((flags & SYNC_TRYLOCK) && - (!ip->i_flush.done || __xfs_iflags_test(ip, XFS_IRECLAIM))) { + __xfs_iflags_test(ip, XFS_IFLOCK | XFS_IRECLAIM)) return 1; - } /* * The radix tree lock here protects a thread in xfs_iget from racing From f392e6319a4e9a028b0c8b48f000bb01d660ad53 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:10 +0000 Subject: [PATCH 521/820] xfs: replace i_pin_wait with a bit waitqueue Replace i_pin_wait, which is only used during synchronous inode flushing with a bit waitqueue. This trades off a much smaller inode against slightly slower wakeup performance, and saves 12 (32-bit) or 20 (64-bit) bytes in the XFS inode. Reviewed-by: Alex Elder Reviewed-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Ben Myers --- fs/xfs/xfs_inode.c | 27 +++++++++++++++++++++------ fs/xfs/xfs_inode.h | 3 ++- fs/xfs/xfs_inode_item.c | 2 +- fs/xfs/xfs_super.c | 1 - 4 files changed, 24 insertions(+), 9 deletions(-) diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index eeb60d31b086..62603369b523 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -2037,7 +2037,7 @@ xfs_idestroy_fork( * once someone is waiting for it to be unpinned. */ static void -xfs_iunpin_nowait( +xfs_iunpin( struct xfs_inode *ip) { ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); @@ -2049,14 +2049,29 @@ xfs_iunpin_nowait( } +static void +__xfs_iunpin_wait( + struct xfs_inode *ip) +{ + wait_queue_head_t *wq = bit_waitqueue(&ip->i_flags, __XFS_IPINNED_BIT); + DEFINE_WAIT_BIT(wait, &ip->i_flags, __XFS_IPINNED_BIT); + + xfs_iunpin(ip); + + do { + prepare_to_wait(wq, &wait.wait, TASK_UNINTERRUPTIBLE); + if (xfs_ipincount(ip)) + io_schedule(); + } while (xfs_ipincount(ip)); + finish_wait(wq, &wait.wait); +} + void xfs_iunpin_wait( struct xfs_inode *ip) { - if (xfs_ipincount(ip)) { - xfs_iunpin_nowait(ip); - wait_event(ip->i_ipin_wait, (xfs_ipincount(ip) == 0)); - } + if (xfs_ipincount(ip)) + __xfs_iunpin_wait(ip); } /* @@ -2415,7 +2430,7 @@ xfs_iflush( * out for us if they occur after the log force completes. */ if (!(flags & SYNC_WAIT) && xfs_ipincount(ip)) { - xfs_iunpin_nowait(ip); + xfs_iunpin(ip); xfs_ifunlock(ip); return EAGAIN; } diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 960d2a89b3ac..4acbe740be46 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -238,7 +238,6 @@ typedef struct xfs_inode { mrlock_t i_lock; /* inode lock */ mrlock_t i_iolock; /* inode IO lock */ atomic_t i_pincount; /* inode pin count */ - wait_queue_head_t i_ipin_wait; /* inode pinning wait queue */ spinlock_t i_flags_lock; /* inode i_flags lock */ /* Miscellaneous state. */ unsigned long i_flags; /* see defined flags below */ @@ -367,6 +366,8 @@ xfs_set_projid(struct xfs_inode *ip, #define XFS_IDIRTY_RELEASE (1 << 6) /* dirty release already seen */ #define __XFS_IFLOCK_BIT 7 /* inode is being flushed right now */ #define XFS_IFLOCK (1 << __XFS_IFLOCK_BIT) +#define __XFS_IPINNED_BIT 8 /* wakeup key for zero pin count */ +#define XFS_IPINNED (1 << __XFS_IPINNED_BIT) /* * Per-lifetime flags need to be reset when re-using a reclaimable inode during diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index c8d4ce0efd5a..91d71dcd4852 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -555,7 +555,7 @@ xfs_inode_item_unpin( trace_xfs_inode_unpin(ip, _RET_IP_); ASSERT(atomic_read(&ip->i_pincount) > 0); if (atomic_dec_and_test(&ip->i_pincount)) - wake_up(&ip->i_ipin_wait); + wake_up_bit(&ip->i_flags, __XFS_IPINNED_BIT); } /* diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 6851fa7b1afa..ee5b695c99a7 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -828,7 +828,6 @@ xfs_fs_inode_init_once( /* xfs inode */ atomic_set(&ip->i_pincount, 0); spin_lock_init(&ip->i_flags_lock); - init_waitqueue_head(&ip->i_ipin_wait); mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER, "xfsino", ip->i_ino); From ce7ae151ddada3dbf67301464343c154903166b3 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:11 +0000 Subject: [PATCH 522/820] xfs: remove the i_size field in struct xfs_inode There is no fundamental need to keep an in-memory inode size copy in the XFS inode. We already have the on-disk value in the dinode, and the separate in-memory copy that we need for regular files only in the XFS inode. Remove the xfs_inode i_size field and change the XFS_ISIZE macro to use the VFS inode i_size field for regular files. Switch code that was directly accessing the i_size field in the xfs_inode to XFS_ISIZE, or in cases where we are limited to regular files direct access of the VFS inode i_size field. This also allows dropping some fairly complicated code in the write path which dealt with keeping the xfs_inode i_size uptodate with the VFS i_size that is getting updated inside ->write_end. Note that we do not bother resetting the VFS i_size when truncating a file that gets freed to zero as there is no point in doing so because the VFS inode is no longer in use at this point. Just relax the assert in xfs_ifree to only check the on-disk size instead. Reviewed-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Ben Myers --- fs/xfs/xfs_aops.c | 2 +- fs/xfs/xfs_bmap.c | 15 ++++++-------- fs/xfs/xfs_file.c | 45 ++++++++++------------------------------ fs/xfs/xfs_fs_subr.c | 2 +- fs/xfs/xfs_iget.c | 1 - fs/xfs/xfs_inode.c | 8 ++----- fs/xfs/xfs_inode.h | 16 ++++++++++---- fs/xfs/xfs_iomap.c | 12 +++++------ fs/xfs/xfs_iops.c | 3 +-- fs/xfs/xfs_qm_syscalls.c | 1 - fs/xfs/xfs_trace.h | 2 +- fs/xfs/xfs_vnodeops.c | 31 ++++++++++++++------------- 12 files changed, 56 insertions(+), 82 deletions(-) diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 574d4ee9b625..4d27ea117e0e 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -111,7 +111,7 @@ xfs_ioend_new_eof( xfs_fsize_t bsize; bsize = ioend->io_offset + ioend->io_size; - isize = MAX(ip->i_size, ip->i_new_size); + isize = MAX(i_size_read(VFS_I(ip)), ip->i_new_size); isize = MIN(isize, bsize); return isize > ip->i_d.di_size ? isize : 0; } diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 7a888ca2f7f6..188ef2fbd628 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -3997,11 +3997,8 @@ xfs_bmap_one_block( xfs_bmbt_irec_t s; /* internal version of extent */ #ifndef DEBUG - if (whichfork == XFS_DATA_FORK) { - return S_ISREG(ip->i_d.di_mode) ? - (ip->i_size == ip->i_mount->m_sb.sb_blocksize) : - (ip->i_d.di_size == ip->i_mount->m_sb.sb_blocksize); - } + if (whichfork == XFS_DATA_FORK) + return XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize; #endif /* !DEBUG */ if (XFS_IFORK_NEXTENTS(ip, whichfork) != 1) return 0; @@ -4013,7 +4010,7 @@ xfs_bmap_one_block( xfs_bmbt_get_all(ep, &s); rval = s.br_startoff == 0 && s.br_blockcount == 1; if (rval && whichfork == XFS_DATA_FORK) - ASSERT(ip->i_size == ip->i_mount->m_sb.sb_blocksize); + ASSERT(XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize); return rval; } @@ -5427,7 +5424,7 @@ xfs_getbmapx_fix_eof_hole( if (startblock == HOLESTARTBLOCK) { mp = ip->i_mount; out->bmv_block = -1; - fixlen = XFS_FSB_TO_BB(mp, XFS_B_TO_FSB(mp, ip->i_size)); + fixlen = XFS_FSB_TO_BB(mp, XFS_B_TO_FSB(mp, XFS_ISIZE(ip))); fixlen -= out->bmv_offset; if (prealloced && out->bmv_offset + out->bmv_length == end) { /* Came to hole at EOF. Trim it. */ @@ -5515,7 +5512,7 @@ xfs_getbmap( fixlen = XFS_MAXIOFFSET(mp); } else { prealloced = 0; - fixlen = ip->i_size; + fixlen = XFS_ISIZE(ip); } } @@ -5544,7 +5541,7 @@ xfs_getbmap( xfs_ilock(ip, XFS_IOLOCK_SHARED); if (whichfork == XFS_DATA_FORK && !(iflags & BMV_IF_DELALLOC)) { - if (ip->i_delayed_blks || ip->i_size > ip->i_d.di_size) { + if (ip->i_delayed_blks || XFS_ISIZE(ip) > ip->i_d.di_size) { error = xfs_flush_pages(ip, 0, -1, 0, FI_REMAPF); if (error) goto out_unlock_iolock; diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index f675f3d9d7b3..86d5dc260464 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -327,7 +327,7 @@ xfs_file_aio_read( mp->m_rtdev_targp : mp->m_ddev_targp; if ((iocb->ki_pos & target->bt_smask) || (size & target->bt_smask)) { - if (iocb->ki_pos == ip->i_size) + if (iocb->ki_pos == i_size_read(inode)) return 0; return -XFS_ERROR(EINVAL); } @@ -412,30 +412,6 @@ xfs_file_splice_read( return ret; } -STATIC void -xfs_aio_write_isize_update( - struct inode *inode, - loff_t *ppos, - ssize_t bytes_written) -{ - struct xfs_inode *ip = XFS_I(inode); - xfs_fsize_t isize = i_size_read(inode); - - if (bytes_written > 0) - XFS_STATS_ADD(xs_write_bytes, bytes_written); - - if (unlikely(bytes_written < 0 && bytes_written != -EFAULT && - *ppos > isize)) - *ppos = isize; - - if (*ppos > ip->i_size) { - xfs_rw_ilock(ip, XFS_ILOCK_EXCL); - if (*ppos > ip->i_size) - ip->i_size = *ppos; - xfs_rw_iunlock(ip, XFS_ILOCK_EXCL); - } -} - /* * If this was a direct or synchronous I/O that failed (such as ENOSPC) then * part of the I/O may have been written to disk before the error occurred. In @@ -451,8 +427,8 @@ xfs_aio_write_newsize_update( xfs_rw_ilock(ip, XFS_ILOCK_EXCL); if (new_size == ip->i_new_size) ip->i_new_size = 0; - if (ip->i_d.di_size > ip->i_size) - ip->i_d.di_size = ip->i_size; + if (ip->i_d.di_size > i_size_read(VFS_I(ip))) + ip->i_d.di_size = i_size_read(VFS_I(ip)); xfs_rw_iunlock(ip, XFS_ILOCK_EXCL); } } @@ -492,15 +468,16 @@ xfs_file_splice_write( new_size = *ppos + count; xfs_ilock(ip, XFS_ILOCK_EXCL); - if (new_size > ip->i_size) + if (new_size > i_size_read(inode)) ip->i_new_size = new_size; xfs_iunlock(ip, XFS_ILOCK_EXCL); trace_xfs_file_splice_write(ip, count, *ppos, ioflags); ret = generic_file_splice_write(pipe, outfilp, ppos, count, flags); + if (ret > 0) + XFS_STATS_ADD(xs_write_bytes, ret); - xfs_aio_write_isize_update(inode, ppos, ret); xfs_aio_write_newsize_update(ip, new_size); xfs_iunlock(ip, XFS_IOLOCK_EXCL); return ret; @@ -728,14 +705,14 @@ restart: * values are still valid. */ if ((ip->i_new_size && *pos > ip->i_new_size) || - (!ip->i_new_size && *pos > ip->i_size)) { + (!ip->i_new_size && *pos > i_size_read(inode))) { if (*iolock == XFS_IOLOCK_SHARED) { xfs_rw_iunlock(ip, XFS_ILOCK_EXCL | *iolock); *iolock = XFS_IOLOCK_EXCL; xfs_rw_ilock(ip, XFS_ILOCK_EXCL | *iolock); goto restart; } - error = -xfs_zero_eof(ip, *pos, ip->i_size); + error = -xfs_zero_eof(ip, *pos, i_size_read(inode)); } /* @@ -744,7 +721,7 @@ restart: * ip->i_new_size if this IO ends beyond any other in-flight writes. */ new_size = *pos + *count; - if (new_size > ip->i_size) { + if (new_size > i_size_read(inode)) { if (new_size > ip->i_new_size) ip->i_new_size = new_size; *new_sizep = new_size; @@ -957,11 +934,11 @@ xfs_file_aio_write( ret = xfs_file_buffered_aio_write(iocb, iovp, nr_segs, pos, ocount, &new_size, &iolock); - xfs_aio_write_isize_update(inode, &iocb->ki_pos, ret); - if (ret <= 0) goto out_unlock; + XFS_STATS_ADD(xs_write_bytes, ret); + /* Handle various SYNC-type writes */ if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) { loff_t end = pos + ret - 1; diff --git a/fs/xfs/xfs_fs_subr.c b/fs/xfs/xfs_fs_subr.c index ed88ed16811c..652b875a9d4c 100644 --- a/fs/xfs/xfs_fs_subr.c +++ b/fs/xfs/xfs_fs_subr.c @@ -90,7 +90,7 @@ xfs_wait_on_pages( if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) { return -filemap_fdatawait_range(mapping, first, - last == -1 ? ip->i_size - 1 : last); + last == -1 ? XFS_ISIZE(ip) - 1 : last); } return 0; } diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index a7cf7139f9ad..3b5b78aa3b87 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -94,7 +94,6 @@ xfs_inode_alloc( ip->i_update_core = 0; ip->i_delayed_blks = 0; memset(&ip->i_d, 0, sizeof(xfs_icdinode_t)); - ip->i_size = 0; ip->i_new_size = 0; return ip; diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 62603369b523..b21022499c2e 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -347,7 +347,6 @@ xfs_iformat( return XFS_ERROR(EFSCORRUPTED); } ip->i_d.di_size = 0; - ip->i_size = 0; ip->i_df.if_u2.if_rdev = xfs_dinode_get_rdev(dip); break; @@ -853,7 +852,6 @@ xfs_iread( } ip->i_delayed_blks = 0; - ip->i_size = ip->i_d.di_size; /* * Mark the buffer containing the inode as something to keep @@ -1043,7 +1041,6 @@ xfs_ialloc( } ip->i_d.di_size = 0; - ip->i_size = 0; ip->i_d.di_nextents = 0; ASSERT(ip->i_d.di_nblocks == 0); @@ -1198,7 +1195,7 @@ xfs_itruncate_extents( int done = 0; ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL)); - ASSERT(new_size <= ip->i_size); + ASSERT(new_size <= XFS_ISIZE(ip)); ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES); ASSERT(ip->i_itemp != NULL); ASSERT(ip->i_itemp->ili_lock_flags == 0); @@ -1712,8 +1709,7 @@ xfs_ifree( ASSERT(ip->i_d.di_nlink == 0); ASSERT(ip->i_d.di_nextents == 0); ASSERT(ip->i_d.di_anextents == 0); - ASSERT((ip->i_d.di_size == 0 && ip->i_size == 0) || - (!S_ISREG(ip->i_d.di_mode))); + ASSERT(ip->i_d.di_size == 0 || !S_ISREG(ip->i_d.di_mode)); ASSERT(ip->i_d.di_nblocks == 0); /* diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 4acbe740be46..cd99e43fa8f0 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -246,16 +246,12 @@ typedef struct xfs_inode { xfs_icdinode_t i_d; /* most of ondisk inode */ - xfs_fsize_t i_size; /* in-memory size */ xfs_fsize_t i_new_size; /* size when write completes */ /* VFS inode */ struct inode i_vnode; /* embedded VFS inode */ } xfs_inode_t; -#define XFS_ISIZE(ip) S_ISREG((ip)->i_d.di_mode) ? \ - (ip)->i_size : (ip)->i_d.di_size; - /* Convert from vfs inode to xfs inode */ static inline struct xfs_inode *XFS_I(struct inode *inode) { @@ -268,6 +264,18 @@ static inline struct inode *VFS_I(struct xfs_inode *ip) return &ip->i_vnode; } +/* + * For regular files we only update the on-disk filesize when actually + * writing data back to disk. Until then only the copy in the VFS inode + * is uptodate. + */ +static inline xfs_fsize_t XFS_ISIZE(struct xfs_inode *ip) +{ + if (S_ISREG(ip->i_d.di_mode)) + return i_size_read(VFS_I(ip)); + return ip->i_d.di_size; +} + /* * i_flags helper functions */ diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index a27a44659da6..246c7d57c6f9 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -74,7 +74,7 @@ xfs_iomap_eof_align_last_fsb( else if (mp->m_dalign) align = mp->m_dalign; - if (align && ip->i_size >= XFS_FSB_TO_B(mp, align)) + if (align && XFS_ISIZE(ip) >= XFS_FSB_TO_B(mp, align)) new_last_fsb = roundup_64(*last_fsb, align); } @@ -154,7 +154,7 @@ xfs_iomap_write_direct( offset_fsb = XFS_B_TO_FSBT(mp, offset); last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count))); - if ((offset + count) > ip->i_size) { + if ((offset + count) > XFS_ISIZE(ip)) { error = xfs_iomap_eof_align_last_fsb(mp, ip, extsz, &last_fsb); if (error) goto error_out; @@ -211,7 +211,7 @@ xfs_iomap_write_direct( xfs_trans_ijoin(tp, ip, 0); bmapi_flag = 0; - if (offset < ip->i_size || extsz) + if (offset < XFS_ISIZE(ip) || extsz) bmapi_flag |= XFS_BMAPI_PREALLOC; /* @@ -286,7 +286,7 @@ xfs_iomap_eof_want_preallocate( int found_delalloc = 0; *prealloc = 0; - if ((offset + count) <= ip->i_size) + if (offset + count <= XFS_ISIZE(ip)) return 0; /* @@ -340,7 +340,7 @@ xfs_iomap_prealloc_size( * if we pass in alloc_blocks = 0. Hence the "+ 1" to * ensure we always pass in a non-zero value. */ - alloc_blocks = XFS_B_TO_FSB(mp, ip->i_size) + 1; + alloc_blocks = XFS_B_TO_FSB(mp, XFS_ISIZE(ip)) + 1; alloc_blocks = XFS_FILEOFF_MIN(MAXEXTLEN, rounddown_pow_of_two(alloc_blocks)); @@ -564,7 +564,7 @@ xfs_iomap_write_allocate( * back.... */ nimaps = 1; - end_fsb = XFS_B_TO_FSB(mp, ip->i_size); + end_fsb = XFS_B_TO_FSB(mp, XFS_ISIZE(ip)); error = xfs_bmap_last_offset(NULL, ip, &last_block, XFS_DATA_FORK); if (error) diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index f02eaa298d3c..ab302539e5b9 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -778,7 +778,7 @@ xfs_setattr_size( lock_flags |= XFS_IOLOCK_EXCL; xfs_ilock(ip, lock_flags); - oldsize = ip->i_size; + oldsize = inode->i_size; newsize = iattr->ia_size; /* @@ -897,7 +897,6 @@ xfs_setattr_size( * they get written to. */ ip->i_d.di_size = newsize; - ip->i_size = newsize; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); if (newsize <= oldsize) { diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 27378650b5cb..eafbcff81f3a 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -265,7 +265,6 @@ xfs_qm_scall_trunc_qfile( xfs_trans_ijoin(tp, ip, 0); ip->i_d.di_size = 0; - ip->i_size = 0; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, 0); diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 81efa0416173..2aabcc9c507e 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -1038,7 +1038,7 @@ DECLARE_EVENT_CLASS(xfs_simple_io_class, TP_fast_assign( __entry->dev = VFS_I(ip)->i_sb->s_dev; __entry->ino = ip->i_ino; - __entry->isize = ip->i_size; + __entry->isize = VFS_I(ip)->i_size; __entry->disize = ip->i_d.di_size; __entry->new_size = ip->i_new_size; __entry->offset = offset; diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 96ff03421753..0cf52da9d246 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -175,7 +175,7 @@ xfs_free_eofblocks( * Figure out if there are any blocks beyond the end * of the file. If not, then there is nothing to do. */ - end_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)ip->i_size)); + end_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)XFS_ISIZE(ip)); last_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)XFS_MAXIOFFSET(mp)); if (last_fsb <= end_fsb) return 0; @@ -233,7 +233,7 @@ xfs_free_eofblocks( * may be full of holes (ie NULL files bug). */ error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, - ip->i_size); + XFS_ISIZE(ip)); if (error) { /* * If we get an error at this point we simply don't @@ -547,8 +547,8 @@ xfs_release( return 0; if ((S_ISREG(ip->i_d.di_mode) && - ((ip->i_size > 0) || (VN_CACHED(VFS_I(ip)) > 0 || - ip->i_delayed_blks > 0)) && + (VFS_I(ip)->i_size > 0 || + (VN_CACHED(VFS_I(ip)) > 0 || ip->i_delayed_blks > 0)) && (ip->i_df.if_flags & XFS_IFEXTENTS)) && (!(ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)))) { @@ -625,7 +625,7 @@ xfs_inactive( * only one with a reference to the inode. */ truncate = ((ip->i_d.di_nlink == 0) && - ((ip->i_d.di_size != 0) || (ip->i_size != 0) || + ((ip->i_d.di_size != 0) || XFS_ISIZE(ip) != 0 || (ip->i_d.di_nextents > 0) || (ip->i_delayed_blks > 0)) && S_ISREG(ip->i_d.di_mode)); @@ -639,12 +639,12 @@ xfs_inactive( if (ip->i_d.di_nlink != 0) { if ((S_ISREG(ip->i_d.di_mode) && - ((ip->i_size > 0) || (VN_CACHED(VFS_I(ip)) > 0 || - ip->i_delayed_blks > 0)) && - (ip->i_df.if_flags & XFS_IFEXTENTS) && - (!(ip->i_d.di_flags & + (VFS_I(ip)->i_size > 0 || + (VN_CACHED(VFS_I(ip)) > 0 || ip->i_delayed_blks > 0)) && + (ip->i_df.if_flags & XFS_IFEXTENTS) && + (!(ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)) || - (ip->i_delayed_blks != 0)))) { + ip->i_delayed_blks != 0))) { error = xfs_free_eofblocks(mp, ip, 0); if (error) return VN_INACTIVE_CACHE; @@ -678,7 +678,6 @@ xfs_inactive( xfs_trans_ijoin(tp, ip, 0); ip->i_d.di_size = 0; - ip->i_size = 0; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, 0); @@ -1974,11 +1973,11 @@ xfs_zero_remaining_bytes( * since nothing can read beyond eof. The space will * be zeroed when the file is extended anyway. */ - if (startoff >= ip->i_size) + if (startoff >= XFS_ISIZE(ip)) return 0; - if (endoff > ip->i_size) - endoff = ip->i_size; + if (endoff > XFS_ISIZE(ip)) + endoff = XFS_ISIZE(ip); bp = xfs_buf_get_uncached(XFS_IS_REALTIME_INODE(ip) ? mp->m_rtdev_targp : mp->m_ddev_targp, @@ -2273,7 +2272,7 @@ xfs_change_file_space( bf->l_start += offset; break; case 2: /*SEEK_END*/ - bf->l_start += ip->i_size; + bf->l_start += XFS_ISIZE(ip); break; default: return XFS_ERROR(EINVAL); @@ -2290,7 +2289,7 @@ xfs_change_file_space( bf->l_whence = 0; startoffset = bf->l_start; - fsize = ip->i_size; + fsize = XFS_ISIZE(ip); /* * XFS_IOC_RESVSP and XFS_IOC_UNRESVSP will reserve or unreserve From 2813d682e8e6a278f94817429afd46b30875bb6e Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:12 +0000 Subject: [PATCH 523/820] xfs: remove the i_new_size field in struct xfs_inode Now that we use the VFS i_size field throughout XFS there is no need for the i_new_size field any more given that the VFS i_size field gets updated in ->write_end before unlocking the page, and thus is always uptodate when writeback could see a page. Removing i_new_size also has the advantage that we will never have to trim back di_size during a failed buffered write, given that it never gets updated past i_size. Note that currently the generic direct I/O code only updates i_size after calling our end_io handler, which requires a small workaround to make sure di_size actually makes it to disk. I hope to fix this properly in the generic code. A downside is that we lose the support for parallel non-overlapping O_DIRECT appending writes that recently was added. I don't think keeping the complex and fragile i_new_size infrastructure for this is a good tradeoff - if we really care about parallel appending writers we should investigate turning the iolock into a range lock, which would also allow for parallel non-overlapping buffered writers. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_aops.c | 29 ++++++++++--------- fs/xfs/xfs_file.c | 72 +++++++--------------------------------------- fs/xfs/xfs_iget.c | 1 - fs/xfs/xfs_inode.h | 2 -- fs/xfs/xfs_trace.h | 18 +++--------- 5 files changed, 30 insertions(+), 92 deletions(-) diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 4d27ea117e0e..74b9baf36ac3 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -111,8 +111,7 @@ xfs_ioend_new_eof( xfs_fsize_t bsize; bsize = ioend->io_offset + ioend->io_size; - isize = MAX(i_size_read(VFS_I(ip)), ip->i_new_size); - isize = MIN(isize, bsize); + isize = MIN(i_size_read(VFS_I(ip)), bsize); return isize > ip->i_d.di_size ? isize : 0; } @@ -126,11 +125,7 @@ static inline bool xfs_ioend_is_append(struct xfs_ioend *ioend) } /* - * Update on-disk file size now that data has been written to disk. The - * current in-memory file size is i_size. If a write is beyond eof i_new_size - * will be the intended file size until i_size is updated. If this write does - * not extend all the way to the valid file size then restrict this update to - * the end of the write. + * Update on-disk file size now that data has been written to disk. * * This function does not block as blocking on the inode lock in IO completion * can lead to IO completion order dependency deadlocks.. If it can't get the @@ -1278,6 +1273,15 @@ xfs_end_io_direct_write( { struct xfs_ioend *ioend = iocb->private; + /* + * While the generic direct I/O code updates the inode size, it does + * so only after the end_io handler is called, which means our + * end_io handler thinks the on-disk size is outside the in-core + * size. To prevent this just update it a little bit earlier here. + */ + if (offset + size > i_size_read(ioend->io_inode)) + i_size_write(ioend->io_inode, offset + size); + /* * blockdev_direct_IO can return an error even after the I/O * completion handler was called. Thus we need to protect @@ -1340,12 +1344,11 @@ xfs_vm_write_failed( if (to > inode->i_size) { /* - * punch out the delalloc blocks we have already allocated. We - * don't call xfs_setattr() to do this as we may be in the - * middle of a multi-iovec write and so the vfs inode->i_size - * will not match the xfs ip->i_size and so it will zero too - * much. Hence we jus truncate the page cache to zero what is - * necessary and punch the delalloc blocks directly. + * Punch out the delalloc blocks we have already allocated. + * + * Don't bother with xfs_setattr given that nothing can have + * made it to disk yet as the page is still locked at this + * point. */ struct xfs_inode *ip = XFS_I(inode); xfs_fileoff_t start_fsb; diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 86d5dc260464..632313926788 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -412,27 +412,6 @@ xfs_file_splice_read( return ret; } -/* - * If this was a direct or synchronous I/O that failed (such as ENOSPC) then - * part of the I/O may have been written to disk before the error occurred. In - * this case the on-disk file size may have been adjusted beyond the in-memory - * file size and now needs to be truncated back. - */ -STATIC void -xfs_aio_write_newsize_update( - struct xfs_inode *ip, - xfs_fsize_t new_size) -{ - if (new_size == ip->i_new_size) { - xfs_rw_ilock(ip, XFS_ILOCK_EXCL); - if (new_size == ip->i_new_size) - ip->i_new_size = 0; - if (ip->i_d.di_size > i_size_read(VFS_I(ip))) - ip->i_d.di_size = i_size_read(VFS_I(ip)); - xfs_rw_iunlock(ip, XFS_ILOCK_EXCL); - } -} - /* * xfs_file_splice_write() does not use xfs_rw_ilock() because * generic_file_splice_write() takes the i_mutex itself. This, in theory, @@ -451,7 +430,6 @@ xfs_file_splice_write( { struct inode *inode = outfilp->f_mapping->host; struct xfs_inode *ip = XFS_I(inode); - xfs_fsize_t new_size; int ioflags = 0; ssize_t ret; @@ -465,20 +443,12 @@ xfs_file_splice_write( xfs_ilock(ip, XFS_IOLOCK_EXCL); - new_size = *ppos + count; - - xfs_ilock(ip, XFS_ILOCK_EXCL); - if (new_size > i_size_read(inode)) - ip->i_new_size = new_size; - xfs_iunlock(ip, XFS_ILOCK_EXCL); - trace_xfs_file_splice_write(ip, count, *ppos, ioflags); ret = generic_file_splice_write(pipe, outfilp, ppos, count, flags); if (ret > 0) XFS_STATS_ADD(xs_write_bytes, ret); - xfs_aio_write_newsize_update(ip, new_size); xfs_iunlock(ip, XFS_IOLOCK_EXCL); return ret; } @@ -673,16 +643,13 @@ xfs_file_aio_write_checks( struct file *file, loff_t *pos, size_t *count, - xfs_fsize_t *new_sizep, int *iolock) { struct inode *inode = file->f_mapping->host; struct xfs_inode *ip = XFS_I(inode); - xfs_fsize_t new_size; int error = 0; xfs_rw_ilock(ip, XFS_ILOCK_EXCL); - *new_sizep = 0; restart: error = generic_write_checks(file, pos, count, S_ISBLK(inode->i_mode)); if (error) { @@ -697,15 +664,13 @@ restart: /* * If the offset is beyond the size of the file, we need to zero any * blocks that fall between the existing EOF and the start of this - * write. There is no need to issue zeroing if another in-flght IO ends - * at or before this one If zeronig is needed and we are currently - * holding the iolock shared, we need to update it to exclusive which - * involves dropping all locks and relocking to maintain correct locking - * order. If we do this, restart the function to ensure all checks and - * values are still valid. + * write. If zeroing is needed and we are currently holding the + * iolock shared, we need to update it to exclusive which involves + * dropping all locks and relocking to maintain correct locking order. + * If we do this, restart the function to ensure all checks and values + * are still valid. */ - if ((ip->i_new_size && *pos > ip->i_new_size) || - (!ip->i_new_size && *pos > i_size_read(inode))) { + if (*pos > i_size_read(inode)) { if (*iolock == XFS_IOLOCK_SHARED) { xfs_rw_iunlock(ip, XFS_ILOCK_EXCL | *iolock); *iolock = XFS_IOLOCK_EXCL; @@ -714,19 +679,6 @@ restart: } error = -xfs_zero_eof(ip, *pos, i_size_read(inode)); } - - /* - * If this IO extends beyond EOF, we may need to update ip->i_new_size. - * We have already zeroed space beyond EOF (if necessary). Only update - * ip->i_new_size if this IO ends beyond any other in-flight writes. - */ - new_size = *pos + *count; - if (new_size > i_size_read(inode)) { - if (new_size > ip->i_new_size) - ip->i_new_size = new_size; - *new_sizep = new_size; - } - xfs_rw_iunlock(ip, XFS_ILOCK_EXCL); if (error) return error; @@ -772,7 +724,6 @@ xfs_file_dio_aio_write( unsigned long nr_segs, loff_t pos, size_t ocount, - xfs_fsize_t *new_size, int *iolock) { struct file *file = iocb->ki_filp; @@ -817,7 +768,7 @@ xfs_file_dio_aio_write( xfs_rw_ilock(ip, *iolock); } - ret = xfs_file_aio_write_checks(file, &pos, &count, new_size, iolock); + ret = xfs_file_aio_write_checks(file, &pos, &count, iolock); if (ret) return ret; @@ -855,7 +806,6 @@ xfs_file_buffered_aio_write( unsigned long nr_segs, loff_t pos, size_t ocount, - xfs_fsize_t *new_size, int *iolock) { struct file *file = iocb->ki_filp; @@ -869,7 +819,7 @@ xfs_file_buffered_aio_write( *iolock = XFS_IOLOCK_EXCL; xfs_rw_ilock(ip, *iolock); - ret = xfs_file_aio_write_checks(file, &pos, &count, new_size, iolock); + ret = xfs_file_aio_write_checks(file, &pos, &count, iolock); if (ret) return ret; @@ -909,7 +859,6 @@ xfs_file_aio_write( ssize_t ret; int iolock; size_t ocount = 0; - xfs_fsize_t new_size = 0; XFS_STATS_INC(xs_write_calls); @@ -929,10 +878,10 @@ xfs_file_aio_write( if (unlikely(file->f_flags & O_DIRECT)) ret = xfs_file_dio_aio_write(iocb, iovp, nr_segs, pos, - ocount, &new_size, &iolock); + ocount, &iolock); else ret = xfs_file_buffered_aio_write(iocb, iovp, nr_segs, pos, - ocount, &new_size, &iolock); + ocount, &iolock); if (ret <= 0) goto out_unlock; @@ -953,7 +902,6 @@ xfs_file_aio_write( } out_unlock: - xfs_aio_write_newsize_update(ip, new_size); xfs_rw_iunlock(ip, iolock); return ret; } diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 3b5b78aa3b87..8c3e46394d48 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -94,7 +94,6 @@ xfs_inode_alloc( ip->i_update_core = 0; ip->i_delayed_blks = 0; memset(&ip->i_d, 0, sizeof(xfs_icdinode_t)); - ip->i_new_size = 0; return ip; } diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index cd99e43fa8f0..2f27b7454085 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -246,8 +246,6 @@ typedef struct xfs_inode { xfs_icdinode_t i_d; /* most of ondisk inode */ - xfs_fsize_t i_new_size; /* size when write completes */ - /* VFS inode */ struct inode i_vnode; /* embedded VFS inode */ } xfs_inode_t; diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 2aabcc9c507e..6b6df5802e95 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -891,7 +891,6 @@ DECLARE_EVENT_CLASS(xfs_file_class, __field(dev_t, dev) __field(xfs_ino_t, ino) __field(xfs_fsize_t, size) - __field(xfs_fsize_t, new_size) __field(loff_t, offset) __field(size_t, count) __field(int, flags) @@ -900,17 +899,15 @@ DECLARE_EVENT_CLASS(xfs_file_class, __entry->dev = VFS_I(ip)->i_sb->s_dev; __entry->ino = ip->i_ino; __entry->size = ip->i_d.di_size; - __entry->new_size = ip->i_new_size; __entry->offset = offset; __entry->count = count; __entry->flags = flags; ), - TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " + TP_printk("dev %d:%d ino 0x%llx size 0x%llx " "offset 0x%llx count 0x%zx ioflags %s", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->size, - __entry->new_size, __entry->offset, __entry->count, __print_flags(__entry->flags, "|", XFS_IO_FLAGS)) @@ -978,7 +975,6 @@ DECLARE_EVENT_CLASS(xfs_imap_class, __field(dev_t, dev) __field(xfs_ino_t, ino) __field(loff_t, size) - __field(loff_t, new_size) __field(loff_t, offset) __field(size_t, count) __field(int, type) @@ -990,7 +986,6 @@ DECLARE_EVENT_CLASS(xfs_imap_class, __entry->dev = VFS_I(ip)->i_sb->s_dev; __entry->ino = ip->i_ino; __entry->size = ip->i_d.di_size; - __entry->new_size = ip->i_new_size; __entry->offset = offset; __entry->count = count; __entry->type = type; @@ -998,13 +993,11 @@ DECLARE_EVENT_CLASS(xfs_imap_class, __entry->startblock = irec ? irec->br_startblock : 0; __entry->blockcount = irec ? irec->br_blockcount : 0; ), - TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " - "offset 0x%llx count %zd type %s " - "startoff 0x%llx startblock %lld blockcount 0x%llx", + TP_printk("dev %d:%d ino 0x%llx size 0x%llx offset 0x%llx count %zd " + "type %s startoff 0x%llx startblock %lld blockcount 0x%llx", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->size, - __entry->new_size, __entry->offset, __entry->count, __print_symbolic(__entry->type, XFS_IO_TYPES), @@ -1031,7 +1024,6 @@ DECLARE_EVENT_CLASS(xfs_simple_io_class, __field(xfs_ino_t, ino) __field(loff_t, isize) __field(loff_t, disize) - __field(loff_t, new_size) __field(loff_t, offset) __field(size_t, count) ), @@ -1040,17 +1032,15 @@ DECLARE_EVENT_CLASS(xfs_simple_io_class, __entry->ino = ip->i_ino; __entry->isize = VFS_I(ip)->i_size; __entry->disize = ip->i_d.di_size; - __entry->new_size = ip->i_new_size; __entry->offset = offset; __entry->count = count; ), - TP_printk("dev %d:%d ino 0x%llx isize 0x%llx disize 0x%llx new_size 0x%llx " + TP_printk("dev %d:%d ino 0x%llx isize 0x%llx disize 0x%llx " "offset 0x%llx count %zd", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->isize, __entry->disize, - __entry->new_size, __entry->offset, __entry->count) ); From 5bf1f26227a59b9634e95eb3c7c012b766e5e6a0 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:13 +0000 Subject: [PATCH 524/820] xfs: always return with the iolock held from xfs_file_aio_write_checks While xfs_iunlock is fine with 0 lockflags the calling conventions are much cleaner if xfs_file_aio_write_checks never returns without the iolock held. Reviewed-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Ben Myers --- fs/xfs/xfs_file.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 632313926788..134ff2fe4f4d 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -636,7 +636,9 @@ out_lock: /* * Common pre-write limit and setup checks. * - * Returns with iolock held according to @iolock. + * Called with the iolocked held either shared and exclusive according to + * @iolock, and returns with it held. Might upgrade the iolock to exclusive + * if called for a direct write beyond i_size. */ STATIC ssize_t xfs_file_aio_write_checks( @@ -653,8 +655,7 @@ xfs_file_aio_write_checks( restart: error = generic_write_checks(file, pos, count, S_ISBLK(inode->i_mode)); if (error) { - xfs_rw_iunlock(ip, XFS_ILOCK_EXCL | *iolock); - *iolock = 0; + xfs_rw_iunlock(ip, XFS_ILOCK_EXCL); return error; } From d060646436233912178e6b9e3a7f30a41214220f Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 18 Dec 2011 20:00:14 +0000 Subject: [PATCH 525/820] xfs: cleanup xfs_file_aio_write With all the size field updates out of the way xfs_file_aio_write can be further simplified by pushing all iolock handling into xfs_file_dio_aio_write and xfs_file_buffered_aio_write and using the generic generic_write_sync helper for synchronous writes. Reviewed-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Ben Myers --- fs/xfs/xfs_file.c | 82 +++++++++++++++++++++-------------------------- 1 file changed, 37 insertions(+), 45 deletions(-) diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 134ff2fe4f4d..7e5bc872f2b4 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -724,8 +724,7 @@ xfs_file_dio_aio_write( const struct iovec *iovp, unsigned long nr_segs, loff_t pos, - size_t ocount, - int *iolock) + size_t ocount) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -735,10 +734,10 @@ xfs_file_dio_aio_write( ssize_t ret = 0; size_t count = ocount; int unaligned_io = 0; + int iolock; struct xfs_buftarg *target = XFS_IS_REALTIME_INODE(ip) ? mp->m_rtdev_targp : mp->m_ddev_targp; - *iolock = 0; if ((pos & target->bt_smask) || (count & target->bt_smask)) return -XFS_ERROR(EINVAL); @@ -753,31 +752,31 @@ xfs_file_dio_aio_write( * EOF zeroing cases and fill out the new inode size as appropriate. */ if (unaligned_io || mapping->nrpages) - *iolock = XFS_IOLOCK_EXCL; + iolock = XFS_IOLOCK_EXCL; else - *iolock = XFS_IOLOCK_SHARED; - xfs_rw_ilock(ip, *iolock); + iolock = XFS_IOLOCK_SHARED; + xfs_rw_ilock(ip, iolock); /* * Recheck if there are cached pages that need invalidate after we got * the iolock to protect against other threads adding new pages while * we were waiting for the iolock. */ - if (mapping->nrpages && *iolock == XFS_IOLOCK_SHARED) { - xfs_rw_iunlock(ip, *iolock); - *iolock = XFS_IOLOCK_EXCL; - xfs_rw_ilock(ip, *iolock); + if (mapping->nrpages && iolock == XFS_IOLOCK_SHARED) { + xfs_rw_iunlock(ip, iolock); + iolock = XFS_IOLOCK_EXCL; + xfs_rw_ilock(ip, iolock); } - ret = xfs_file_aio_write_checks(file, &pos, &count, iolock); + ret = xfs_file_aio_write_checks(file, &pos, &count, &iolock); if (ret) - return ret; + goto out; if (mapping->nrpages) { ret = -xfs_flushinval_pages(ip, (pos & PAGE_CACHE_MASK), -1, FI_REMAPF_LOCKED); if (ret) - return ret; + goto out; } /* @@ -786,15 +785,18 @@ xfs_file_dio_aio_write( */ if (unaligned_io) inode_dio_wait(inode); - else if (*iolock == XFS_IOLOCK_EXCL) { + else if (iolock == XFS_IOLOCK_EXCL) { xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL); - *iolock = XFS_IOLOCK_SHARED; + iolock = XFS_IOLOCK_SHARED; } trace_xfs_file_direct_write(ip, count, iocb->ki_pos, 0); ret = generic_file_direct_write(iocb, iovp, &nr_segs, pos, &iocb->ki_pos, count, ocount); +out: + xfs_rw_iunlock(ip, iolock); + /* No fallback to buffered IO on errors for XFS. */ ASSERT(ret < 0 || ret == count); return ret; @@ -806,8 +808,7 @@ xfs_file_buffered_aio_write( const struct iovec *iovp, unsigned long nr_segs, loff_t pos, - size_t ocount, - int *iolock) + size_t ocount) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -815,14 +816,14 @@ xfs_file_buffered_aio_write( struct xfs_inode *ip = XFS_I(inode); ssize_t ret; int enospc = 0; + int iolock = XFS_IOLOCK_EXCL; size_t count = ocount; - *iolock = XFS_IOLOCK_EXCL; - xfs_rw_ilock(ip, *iolock); + xfs_rw_ilock(ip, iolock); - ret = xfs_file_aio_write_checks(file, &pos, &count, iolock); + ret = xfs_file_aio_write_checks(file, &pos, &count, &iolock); if (ret) - return ret; + goto out; /* We can write back this queue in page reclaim */ current->backing_dev_info = mapping->backing_dev_info; @@ -836,13 +837,15 @@ write_retry: * page locks and retry *once* */ if (ret == -ENOSPC && !enospc) { - ret = -xfs_flush_pages(ip, 0, -1, 0, FI_NONE); - if (ret) - return ret; enospc = 1; - goto write_retry; + ret = -xfs_flush_pages(ip, 0, -1, 0, FI_NONE); + if (!ret) + goto write_retry; } + current->backing_dev_info = NULL; +out: + xfs_rw_iunlock(ip, iolock); return ret; } @@ -858,7 +861,6 @@ xfs_file_aio_write( struct inode *inode = mapping->host; struct xfs_inode *ip = XFS_I(inode); ssize_t ret; - int iolock; size_t ocount = 0; XFS_STATS_INC(xs_write_calls); @@ -878,32 +880,22 @@ xfs_file_aio_write( return -EIO; if (unlikely(file->f_flags & O_DIRECT)) - ret = xfs_file_dio_aio_write(iocb, iovp, nr_segs, pos, - ocount, &iolock); + ret = xfs_file_dio_aio_write(iocb, iovp, nr_segs, pos, ocount); else ret = xfs_file_buffered_aio_write(iocb, iovp, nr_segs, pos, - ocount, &iolock); + ocount); - if (ret <= 0) - goto out_unlock; + if (ret > 0) { + ssize_t err; - XFS_STATS_ADD(xs_write_bytes, ret); + XFS_STATS_ADD(xs_write_bytes, ret); - /* Handle various SYNC-type writes */ - if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) { - loff_t end = pos + ret - 1; - int error; - - xfs_rw_iunlock(ip, iolock); - error = xfs_file_fsync(file, pos, end, - (file->f_flags & __O_SYNC) ? 0 : 1); - xfs_rw_ilock(ip, iolock); - if (error) - ret = error; + /* Handle various SYNC-type writes */ + err = generic_write_sync(file, pos, ret); + if (err < 0) + ret = err; } -out_unlock: - xfs_rw_iunlock(ip, iolock); return ret; } From 5ef30ee53b187786e64bdc1f8109e39d17f2ce58 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:05 -0500 Subject: [PATCH 526/820] audit: make filetype matching consistent with other filters Every other filter that matches part of the inodes list collected by audit will match against any of the inodes on that list. The filetype matching however had a strange way of doing things. It allowed userspace to indicated if it should match on the first of the second name collected by the kernel. Name collection ordering seems like a kernel internal and making userspace rules get that right just seems like a bad idea. As it turns out the userspace audit writers had no idea it was doing this and thus never overloaded the value field. The kernel always checked the first name collected which for the tested rules was always correct. This patch just makes the filetype matching like the major, minor, inode, and LSM rules in that it will match against any of the names collected. It also changes the rule validation to reject the old unused rule types. Noone knew it was there. Noone used it. Why keep around the extra code? Signed-off-by: Eric Paris --- kernel/auditfilter.c | 4 ++-- kernel/auditsc.c | 19 +++++++++---------- 2 files changed, 11 insertions(+), 12 deletions(-) diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index f8277c80d678..d94dde82c3c8 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -385,7 +385,7 @@ static struct audit_entry *audit_rule_to_entry(struct audit_rule *rule) goto exit_free; break; case AUDIT_FILETYPE: - if ((f->val & ~S_IFMT) > S_IFMT) + if (f->val & ~S_IFMT) goto exit_free; break; case AUDIT_INODE: @@ -536,7 +536,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, goto exit_free; break; case AUDIT_FILETYPE: - if ((f->val & ~S_IFMT) > S_IFMT) + if (f->val & ~S_IFMT) goto exit_free; break; default: diff --git a/kernel/auditsc.c b/kernel/auditsc.c index e7fe2b0d29b3..a09c50317059 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -305,21 +305,20 @@ static int audit_match_perm(struct audit_context *ctx, int mask) } } -static int audit_match_filetype(struct audit_context *ctx, int which) +static int audit_match_filetype(struct audit_context *ctx, int val) { - unsigned index = which & ~S_IFMT; - umode_t mode = which & S_IFMT; + int index; + umode_t mode = (umode_t)val; if (unlikely(!ctx)) return 0; - if (index >= ctx->name_count) - return 0; - if (ctx->names[index].ino == -1) - return 0; - if ((ctx->names[index].mode ^ mode) & S_IFMT) - return 0; - return 1; + for (index = 0; index < ctx->name_count; index++) { + if ((ctx->names[index].ino != -1) && + ((ctx->names[index].mode & S_IFMT) == mode)) + return 1; + } + return 0; } /* From 5195d8e217a78697152d64fc09a16e063a022465 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:05 -0500 Subject: [PATCH 527/820] audit: dynamically allocate audit_names when not enough space is in the names array This patch does 2 things. First it reduces the number of audit_names allocated in every audit context from 20 to 5. 5 should be enough for all 'normal' syscalls (rename being the worst). Some syscalls can still touch more the 5 inodes such as mount. When rpc filesystem is mounted it will create inodes and those can exceed 5. To handle that problem this patch will dynamically allocate audit_names if it needs more than 5. This should decrease the typicall memory usage while still supporting all the possible kernel operations. Signed-off-by: Eric Paris --- kernel/auditsc.c | 403 +++++++++++++++++++++++++---------------------- 1 file changed, 215 insertions(+), 188 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index a09c50317059..1a92d61ddd27 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -71,8 +71,9 @@ #include "audit.h" /* AUDIT_NAMES is the number of slots we reserve in the audit_context - * for saving names from getname(). */ -#define AUDIT_NAMES 20 + * for saving names from getname(). If we get more names we will allocate + * a name dynamically and also add those to the list anchored by names_list. */ +#define AUDIT_NAMES 5 /* Indicates that audit should log the full pathname. */ #define AUDIT_NAME_FULL -1 @@ -101,9 +102,8 @@ struct audit_cap_data { * * Further, in fs/namei.c:path_lookup() we store the inode and device. */ struct audit_names { + struct list_head list; /* audit_context->names_list */ const char *name; - int name_len; /* number of name's characters to log */ - unsigned name_put; /* call __putname() for this name */ unsigned long ino; dev_t dev; umode_t mode; @@ -113,6 +113,14 @@ struct audit_names { u32 osid; struct audit_cap_data fcap; unsigned int fcap_ver; + int name_len; /* number of name's characters to log */ + bool name_put; /* call __putname() for this name */ + /* + * This was an allocated audit_names and not from the array of + * names allocated in the task audit context. Thus this name + * should be freed on syscall exit + */ + bool should_free; }; struct audit_aux_data { @@ -174,8 +182,17 @@ struct audit_context { long return_code;/* syscall return code */ u64 prio; int return_valid; /* return code is valid */ - int name_count; - struct audit_names names[AUDIT_NAMES]; + /* + * The names_list is the list of all audit_names collected during this + * syscall. The first AUDIT_NAMES entries in the names_list will + * actually be from the preallocated_names array for performance + * reasons. Except during allocation they should never be referenced + * through the preallocated_names array and should only be found/used + * by running the names_list. + */ + struct audit_names preallocated_names[AUDIT_NAMES]; + int name_count; /* total records in names_list */ + struct list_head names_list; /* anchor for struct audit_names->list */ char * filterkey; /* key for rule that triggered record */ struct path pwd; struct audit_context *previous; /* For nested syscalls */ @@ -307,17 +324,18 @@ static int audit_match_perm(struct audit_context *ctx, int mask) static int audit_match_filetype(struct audit_context *ctx, int val) { - int index; + struct audit_names *n; umode_t mode = (umode_t)val; if (unlikely(!ctx)) return 0; - for (index = 0; index < ctx->name_count; index++) { - if ((ctx->names[index].ino != -1) && - ((ctx->names[index].mode & S_IFMT) == mode)) + list_for_each_entry(n, &ctx->names_list, list) { + if ((n->ino != -1) && + ((n->mode & S_IFMT) == mode)) return 1; } + return 0; } @@ -456,13 +474,14 @@ static int audit_filter_rules(struct task_struct *tsk, bool task_creation) { const struct cred *cred; - int i, j, need_sid = 1; + int i, need_sid = 1; u32 sid; cred = rcu_dereference_check(tsk->cred, tsk == current || task_creation); for (i = 0; i < rule->field_count; i++) { struct audit_field *f = &rule->fields[i]; + struct audit_names *n; int result = 0; switch (f->type) { @@ -525,8 +544,8 @@ static int audit_filter_rules(struct task_struct *tsk, result = audit_comparator(MAJOR(name->dev), f->op, f->val); else if (ctx) { - for (j = 0; j < ctx->name_count; j++) { - if (audit_comparator(MAJOR(ctx->names[j].dev), f->op, f->val)) { + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_comparator(MAJOR(n->dev), f->op, f->val)) { ++result; break; } @@ -538,8 +557,8 @@ static int audit_filter_rules(struct task_struct *tsk, result = audit_comparator(MINOR(name->dev), f->op, f->val); else if (ctx) { - for (j = 0; j < ctx->name_count; j++) { - if (audit_comparator(MINOR(ctx->names[j].dev), f->op, f->val)) { + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_comparator(MINOR(n->dev), f->op, f->val)) { ++result; break; } @@ -550,8 +569,8 @@ static int audit_filter_rules(struct task_struct *tsk, if (name) result = (name->ino == f->val); else if (ctx) { - for (j = 0; j < ctx->name_count; j++) { - if (audit_comparator(ctx->names[j].ino, f->op, f->val)) { + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_comparator(n->ino, f->op, f->val)) { ++result; break; } @@ -606,11 +625,10 @@ static int audit_filter_rules(struct task_struct *tsk, name->osid, f->type, f->op, f->lsm_rule, ctx); } else if (ctx) { - for (j = 0; j < ctx->name_count; j++) { - if (security_audit_rule_match( - ctx->names[j].osid, - f->type, f->op, - f->lsm_rule, ctx)) { + list_for_each_entry(n, &ctx->names_list, list) { + if (security_audit_rule_match(n->osid, f->type, + f->op, f->lsm_rule, + ctx)) { ++result; break; } @@ -721,40 +739,53 @@ static enum audit_state audit_filter_syscall(struct task_struct *tsk, return AUDIT_BUILD_CONTEXT; } -/* At syscall exit time, this filter is called if any audit_names[] have been +/* + * Given an audit_name check the inode hash table to see if they match. + * Called holding the rcu read lock to protect the use of audit_inode_hash + */ +static int audit_filter_inode_name(struct task_struct *tsk, + struct audit_names *n, + struct audit_context *ctx) { + int word, bit; + int h = audit_hash_ino((u32)n->ino); + struct list_head *list = &audit_inode_hash[h]; + struct audit_entry *e; + enum audit_state state; + + word = AUDIT_WORD(ctx->major); + bit = AUDIT_BIT(ctx->major); + + if (list_empty(list)) + return 0; + + list_for_each_entry_rcu(e, list, list) { + if ((e->rule.mask[word] & bit) == bit && + audit_filter_rules(tsk, &e->rule, ctx, n, &state, false)) { + ctx->current_state = state; + return 1; + } + } + + return 0; +} + +/* At syscall exit time, this filter is called if any audit_names have been * collected during syscall processing. We only check rules in sublists at hash - * buckets applicable to the inode numbers in audit_names[]. + * buckets applicable to the inode numbers in audit_names. * Regarding audit_state, same rules apply as for audit_filter_syscall(). */ void audit_filter_inodes(struct task_struct *tsk, struct audit_context *ctx) { - int i; - struct audit_entry *e; - enum audit_state state; + struct audit_names *n; if (audit_pid && tsk->tgid == audit_pid) return; rcu_read_lock(); - for (i = 0; i < ctx->name_count; i++) { - int word = AUDIT_WORD(ctx->major); - int bit = AUDIT_BIT(ctx->major); - struct audit_names *n = &ctx->names[i]; - int h = audit_hash_ino((u32)n->ino); - struct list_head *list = &audit_inode_hash[h]; - if (list_empty(list)) - continue; - - list_for_each_entry_rcu(e, list, list) { - if ((e->rule.mask[word] & bit) == bit && - audit_filter_rules(tsk, &e->rule, ctx, n, - &state, false)) { - rcu_read_unlock(); - ctx->current_state = state; - return; - } - } + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_filter_inode_name(tsk, n, ctx)) + break; } rcu_read_unlock(); } @@ -798,7 +829,7 @@ static inline struct audit_context *audit_get_context(struct task_struct *tsk, static inline void audit_free_names(struct audit_context *context) { - int i; + struct audit_names *n, *next; #if AUDIT_DEBUG == 2 if (context->put_count + context->ino_count != context->name_count) { @@ -809,10 +840,9 @@ static inline void audit_free_names(struct audit_context *context) context->serial, context->major, context->in_syscall, context->name_count, context->put_count, context->ino_count); - for (i = 0; i < context->name_count; i++) { + list_for_each_entry(n, &context->names_list, list) { printk(KERN_ERR "names[%d] = %p = %s\n", i, - context->names[i].name, - context->names[i].name ?: "(null)"); + n->name, n->name ?: "(null)"); } dump_stack(); return; @@ -823,9 +853,12 @@ static inline void audit_free_names(struct audit_context *context) context->ino_count = 0; #endif - for (i = 0; i < context->name_count; i++) { - if (context->names[i].name && context->names[i].name_put) - __putname(context->names[i].name); + list_for_each_entry_safe(n, next, &context->names_list, list) { + list_del(&n->list); + if (n->name && n->name_put) + __putname(n->name); + if (n->should_free) + kfree(n); } context->name_count = 0; path_put(&context->pwd); @@ -863,6 +896,7 @@ static inline struct audit_context *audit_alloc_context(enum audit_state state) return NULL; audit_zero_context(context, state); INIT_LIST_HEAD(&context->killed_trees); + INIT_LIST_HEAD(&context->names_list); return context; } @@ -1323,6 +1357,68 @@ static void show_special(struct audit_context *context, int *call_panic) audit_log_end(ab); } +static void audit_log_name(struct audit_context *context, struct audit_names *n, + int record_num, int *call_panic) +{ + struct audit_buffer *ab; + ab = audit_log_start(context, GFP_KERNEL, AUDIT_PATH); + if (!ab) + return; /* audit_panic has been called */ + + audit_log_format(ab, "item=%d", record_num); + + if (n->name) { + switch (n->name_len) { + case AUDIT_NAME_FULL: + /* log the full path */ + audit_log_format(ab, " name="); + audit_log_untrustedstring(ab, n->name); + break; + case 0: + /* name was specified as a relative path and the + * directory component is the cwd */ + audit_log_d_path(ab, "name=", &context->pwd); + break; + default: + /* log the name's directory component */ + audit_log_format(ab, " name="); + audit_log_n_untrustedstring(ab, n->name, + n->name_len); + } + } else + audit_log_format(ab, " name=(null)"); + + if (n->ino != (unsigned long)-1) { + audit_log_format(ab, " inode=%lu" + " dev=%02x:%02x mode=%#ho" + " ouid=%u ogid=%u rdev=%02x:%02x", + n->ino, + MAJOR(n->dev), + MINOR(n->dev), + n->mode, + n->uid, + n->gid, + MAJOR(n->rdev), + MINOR(n->rdev)); + } + if (n->osid != 0) { + char *ctx = NULL; + u32 len; + if (security_secid_to_secctx( + n->osid, &ctx, &len)) { + audit_log_format(ab, " osid=%u", n->osid); + *call_panic = 2; + } else { + audit_log_format(ab, " obj=%s", ctx); + security_release_secctx(ctx, len); + } + } + + audit_log_fcaps(ab, n); + + audit_log_end(ab); +} + static void audit_log_exit(struct audit_context *context, struct task_struct *tsk) { const struct cred *cred; @@ -1330,6 +1426,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts struct audit_buffer *ab; struct audit_aux_data *aux; const char *tty; + struct audit_names *n; /* tsk == current */ context->pid = tsk->pid; @@ -1469,66 +1566,10 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts audit_log_end(ab); } } - for (i = 0; i < context->name_count; i++) { - struct audit_names *n = &context->names[i]; - ab = audit_log_start(context, GFP_KERNEL, AUDIT_PATH); - if (!ab) - continue; /* audit_panic has been called */ - - audit_log_format(ab, "item=%d", i); - - if (n->name) { - switch(n->name_len) { - case AUDIT_NAME_FULL: - /* log the full path */ - audit_log_format(ab, " name="); - audit_log_untrustedstring(ab, n->name); - break; - case 0: - /* name was specified as a relative path and the - * directory component is the cwd */ - audit_log_d_path(ab, "name=", &context->pwd); - break; - default: - /* log the name's directory component */ - audit_log_format(ab, " name="); - audit_log_n_untrustedstring(ab, n->name, - n->name_len); - } - } else - audit_log_format(ab, " name=(null)"); - - if (n->ino != (unsigned long)-1) { - audit_log_format(ab, " inode=%lu" - " dev=%02x:%02x mode=%#ho" - " ouid=%u ogid=%u rdev=%02x:%02x", - n->ino, - MAJOR(n->dev), - MINOR(n->dev), - n->mode, - n->uid, - n->gid, - MAJOR(n->rdev), - MINOR(n->rdev)); - } - if (n->osid != 0) { - char *ctx = NULL; - u32 len; - if (security_secid_to_secctx( - n->osid, &ctx, &len)) { - audit_log_format(ab, " osid=%u", n->osid); - call_panic = 2; - } else { - audit_log_format(ab, " obj=%s", ctx); - security_release_secctx(ctx, len); - } - } - - audit_log_fcaps(ab, n); - - audit_log_end(ab); - } + i = 0; + list_for_each_entry(n, &context->names_list, list) + audit_log_name(context, n, i++, &call_panic); /* Send end of event record to help user space know we are finished */ ab = audit_log_start(context, GFP_KERNEL, AUDIT_EOE); @@ -1820,6 +1861,30 @@ retry: #endif } +static struct audit_names *audit_alloc_name(struct audit_context *context) +{ + struct audit_names *aname; + + if (context->name_count < AUDIT_NAMES) { + aname = &context->preallocated_names[context->name_count]; + memset(aname, 0, sizeof(*aname)); + } else { + aname = kzalloc(sizeof(*aname), GFP_NOFS); + if (!aname) + return NULL; + aname->should_free = true; + } + + aname->ino = (unsigned long)-1; + list_add_tail(&aname->list, &context->names_list); + + context->name_count++; +#if AUDIT_DEBUG + context->ino_count++; +#endif + return aname; +} + /** * audit_getname - add a name to the list * @name: name to add @@ -1830,6 +1895,7 @@ retry: void __audit_getname(const char *name) { struct audit_context *context = current->audit_context; + struct audit_names *n; if (IS_ERR(name) || !name) return; @@ -1842,13 +1908,15 @@ void __audit_getname(const char *name) #endif return; } - BUG_ON(context->name_count >= AUDIT_NAMES); - context->names[context->name_count].name = name; - context->names[context->name_count].name_len = AUDIT_NAME_FULL; - context->names[context->name_count].name_put = 1; - context->names[context->name_count].ino = (unsigned long)-1; - context->names[context->name_count].osid = 0; - ++context->name_count; + + n = audit_alloc_name(context); + if (!n) + return; + + n->name = name; + n->name_len = AUDIT_NAME_FULL; + n->name_put = true; + if (!context->pwd.dentry) get_fs_pwd(current->fs, &context->pwd); } @@ -1870,12 +1938,13 @@ void audit_putname(const char *name) printk(KERN_ERR "%s:%d(:%d): __putname(%p)\n", __FILE__, __LINE__, context->serial, name); if (context->name_count) { + struct audit_names *n; int i; - for (i = 0; i < context->name_count; i++) + + list_for_each_entry(n, &context->names_list, list) printk(KERN_ERR "name[%d] = %p = %s\n", i, - context->names[i].name, - context->names[i].name ?: "(null)"); - } + n->name, n->name ?: "(null)"); + } #endif __putname(name); } @@ -1896,39 +1965,11 @@ void audit_putname(const char *name) #endif } -static int audit_inc_name_count(struct audit_context *context, - const struct inode *inode) -{ - if (context->name_count >= AUDIT_NAMES) { - if (inode) - printk(KERN_DEBUG "audit: name_count maxed, losing inode data: " - "dev=%02x:%02x, inode=%lu\n", - MAJOR(inode->i_sb->s_dev), - MINOR(inode->i_sb->s_dev), - inode->i_ino); - - else - printk(KERN_DEBUG "name_count maxed, losing inode data\n"); - return 1; - } - context->name_count++; -#if AUDIT_DEBUG - context->ino_count++; -#endif - return 0; -} - - static inline int audit_copy_fcaps(struct audit_names *name, const struct dentry *dentry) { struct cpu_vfs_cap_data caps; int rc; - memset(&name->fcap.permitted, 0, sizeof(kernel_cap_t)); - memset(&name->fcap.inheritable, 0, sizeof(kernel_cap_t)); - name->fcap.fE = 0; - name->fcap_ver = 0; - if (!dentry) return 0; @@ -1968,30 +2009,25 @@ static void audit_copy_inode(struct audit_names *name, const struct dentry *dent */ void __audit_inode(const char *name, const struct dentry *dentry) { - int idx; struct audit_context *context = current->audit_context; const struct inode *inode = dentry->d_inode; + struct audit_names *n; if (!context->in_syscall) return; - if (context->name_count - && context->names[context->name_count-1].name - && context->names[context->name_count-1].name == name) - idx = context->name_count - 1; - else if (context->name_count > 1 - && context->names[context->name_count-2].name - && context->names[context->name_count-2].name == name) - idx = context->name_count - 2; - else { - /* FIXME: how much do we care about inodes that have no - * associated name? */ - if (audit_inc_name_count(context, inode)) - return; - idx = context->name_count - 1; - context->names[idx].name = NULL; + + list_for_each_entry_reverse(n, &context->names_list, list) { + if (n->name && (n->name == name)) + goto out; } + + /* unable to find the name from a previous getname() */ + n = audit_alloc_name(context); + if (!n) + return; +out: handle_path(dentry); - audit_copy_inode(&context->names[idx], dentry, inode); + audit_copy_inode(n, dentry, inode); } /** @@ -2010,11 +2046,11 @@ void __audit_inode(const char *name, const struct dentry *dentry) void __audit_inode_child(const struct dentry *dentry, const struct inode *parent) { - int idx; struct audit_context *context = current->audit_context; const char *found_parent = NULL, *found_child = NULL; const struct inode *inode = dentry->d_inode; const char *dname = dentry->d_name.name; + struct audit_names *n; int dirlen = 0; if (!context->in_syscall) @@ -2024,9 +2060,7 @@ void __audit_inode_child(const struct dentry *dentry, handle_one(inode); /* parent is more likely, look for it first */ - for (idx = 0; idx < context->name_count; idx++) { - struct audit_names *n = &context->names[idx]; - + list_for_each_entry(n, &context->names_list, list) { if (!n->name) continue; @@ -2039,9 +2073,7 @@ void __audit_inode_child(const struct dentry *dentry, } /* no matching parent, look for matching child */ - for (idx = 0; idx < context->name_count; idx++) { - struct audit_names *n = &context->names[idx]; - + list_for_each_entry(n, &context->names_list, list) { if (!n->name) continue; @@ -2059,34 +2091,29 @@ void __audit_inode_child(const struct dentry *dentry, add_names: if (!found_parent) { - if (audit_inc_name_count(context, parent)) + n = audit_alloc_name(context); + if (!n) return; - idx = context->name_count - 1; - context->names[idx].name = NULL; - audit_copy_inode(&context->names[idx], NULL, parent); + audit_copy_inode(n, NULL, parent); } if (!found_child) { - if (audit_inc_name_count(context, inode)) + n = audit_alloc_name(context); + if (!n) return; - idx = context->name_count - 1; /* Re-use the name belonging to the slot for a matching parent * directory. All names for this context are relinquished in * audit_free_names() */ if (found_parent) { - context->names[idx].name = found_parent; - context->names[idx].name_len = AUDIT_NAME_FULL; + n->name = found_parent; + n->name_len = AUDIT_NAME_FULL; /* don't call __putname() */ - context->names[idx].name_put = 0; - } else { - context->names[idx].name = NULL; + n->name_put = false; } if (inode) - audit_copy_inode(&context->names[idx], NULL, inode); - else - context->names[idx].ino = (unsigned long)-1; + audit_copy_inode(n, NULL, inode); } } EXPORT_SYMBOL_GPL(__audit_inode_child); From 3035c51e8ac0512686ceb9f2bd1d13bdc6e4fb29 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:05 -0500 Subject: [PATCH 528/820] audit: drop the meaningless and format breaking word 'user' userspace audit messages look like so: type=USER msg=audit(1271170549.415:24710): user pid=14722 uid=0 auid=500 ses=1 subj=unconfined_u:unconfined_r:auditctl_t:s0-s0:c0.c1023 msg='' That third field just says 'user'. That's useless and doesn't follow the key=value pair we are trying to enforce. We already know it came from the user based on the record type. Kill that word. Die. Signed-off-by: Eric Paris --- kernel/audit.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/audit.c b/kernel/audit.c index 2c1d6ab7106e..00efe4758c86 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -631,7 +631,7 @@ static int audit_log_common_recv_msg(struct audit_buffer **ab, u16 msg_type, } *ab = audit_log_start(NULL, GFP_KERNEL, msg_type); - audit_log_format(*ab, "user pid=%d uid=%u auid=%u ses=%u", + audit_log_format(*ab, "pid=%d uid=%u auid=%u ses=%u", pid, uid, auid, ses); if (sid) { rc = security_secid_to_secctx(sid, &ctx, &len); From 16c174bd95cb07c9d0ad3fcd8c70f9cea7214c9d Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:05 -0500 Subject: [PATCH 529/820] audit: check current inode and containing object when filtering on major and minor The audit system has the ability to filter on the major and minor number of the device containing the inode being operated upon. Lets say that /dev/sda1 has major,minor 8,1 and that we mount /dev/sda1 on /boot. Now lets say we add a watch with a filter on 8,1. If we proceed to open an inode inside /boot, such as /vboot/vmlinuz, we will match the major,minor filter. Lets instead assume that one were to use a tool like debugfs and were to open /dev/sda1 directly and to modify it's contents. We might hope that this would also be logged, but it isn't. The rules will check the major,minor of the device containing /dev/sda1. In other words the rule would match on the major/minor of the tmpfs mounted at /dev. I believe these rules should trigger on either device. The man page is devoid of useful information about the intended semantics. It only seems logical that if you want to know everything that happened on a major,minor that would include things that happened to the device itself... Signed-off-by: Eric Paris --- kernel/auditsc.c | 24 ++++++++++++++---------- 1 file changed, 14 insertions(+), 10 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 1a92d61ddd27..7c495147c3d9 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -540,12 +540,14 @@ static int audit_filter_rules(struct task_struct *tsk, } break; case AUDIT_DEVMAJOR: - if (name) - result = audit_comparator(MAJOR(name->dev), - f->op, f->val); - else if (ctx) { + if (name) { + if (audit_comparator(MAJOR(name->dev), f->op, f->val) || + audit_comparator(MAJOR(name->rdev), f->op, f->val)) + ++result; + } else if (ctx) { list_for_each_entry(n, &ctx->names_list, list) { - if (audit_comparator(MAJOR(n->dev), f->op, f->val)) { + if (audit_comparator(MAJOR(n->dev), f->op, f->val) || + audit_comparator(MAJOR(n->rdev), f->op, f->val)) { ++result; break; } @@ -553,12 +555,14 @@ static int audit_filter_rules(struct task_struct *tsk, } break; case AUDIT_DEVMINOR: - if (name) - result = audit_comparator(MINOR(name->dev), - f->op, f->val); - else if (ctx) { + if (name) { + if (audit_comparator(MINOR(name->dev), f->op, f->val) || + audit_comparator(MINOR(name->rdev), f->op, f->val)) + ++result; + } else if (ctx) { list_for_each_entry(n, &ctx->names_list, list) { - if (audit_comparator(MINOR(n->dev), f->op, f->val)) { + if (audit_comparator(MINOR(n->dev), f->op, f->val) || + audit_comparator(MINOR(n->rdev), f->op, f->val)) { ++result; break; } From 85e7bac33b8d5edafc4e219c7dfdb3d48e0b4e31 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:05 -0500 Subject: [PATCH 530/820] seccomp: audit abnormal end to a process due to seccomp The audit system likes to collect information about processes that end abnormally (SIGSEGV) as this may me useful intrusion detection information. This patch adds audit support to collect information when seccomp forces a task to exit because of misbehavior in a similar way. Signed-off-by: Eric Paris --- include/linux/audit.h | 8 +++++++ kernel/auditsc.c | 50 +++++++++++++++++++++++++------------------ kernel/seccomp.c | 2 ++ 3 files changed, 39 insertions(+), 21 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 426ab9f4dd85..6e1c533f9b46 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -430,6 +430,7 @@ extern void audit_putname(const char *name); extern void __audit_inode(const char *name, const struct dentry *dentry); extern void __audit_inode_child(const struct dentry *dentry, const struct inode *parent); +extern void __audit_seccomp(unsigned long syscall); extern void __audit_ptrace(struct task_struct *t); static inline int audit_dummy_context(void) @@ -453,6 +454,12 @@ static inline void audit_inode_child(const struct dentry *dentry, } void audit_core_dumps(long signr); +static inline void audit_seccomp(unsigned long syscall) +{ + if (unlikely(!audit_dummy_context())) + __audit_seccomp(syscall); +} + static inline void audit_ptrace(struct task_struct *t) { if (unlikely(!audit_dummy_context())) @@ -558,6 +565,7 @@ extern int audit_signals; #define audit_inode(n,d) do { (void)(d); } while (0) #define audit_inode_child(i,p) do { ; } while (0) #define audit_core_dumps(i) do { ; } while (0) +#define audit_seccomp(i) do { ; } while (0) #define auditsc_get_stamp(c,t,s) (0) #define audit_get_loginuid(t) (-1) #define audit_get_sessionid(t) (-1) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 7c495147c3d9..e9bcb93800d8 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -2529,6 +2529,25 @@ void __audit_mmap_fd(int fd, int flags) context->type = AUDIT_MMAP; } +static void audit_log_abend(struct audit_buffer *ab, char *reason, long signr) +{ + uid_t auid, uid; + gid_t gid; + unsigned int sessionid; + + auid = audit_get_loginuid(current); + sessionid = audit_get_sessionid(current); + current_uid_gid(&uid, &gid); + + audit_log_format(ab, "auid=%u uid=%u gid=%u ses=%u", + auid, uid, gid, sessionid); + audit_log_task_context(ab); + audit_log_format(ab, " pid=%d comm=", current->pid); + audit_log_untrustedstring(ab, current->comm); + audit_log_format(ab, " reason="); + audit_log_string(ab, reason); + audit_log_format(ab, " sig=%ld", signr); +} /** * audit_core_dumps - record information about processes that end abnormally * @signr: signal value @@ -2539,10 +2558,6 @@ void __audit_mmap_fd(int fd, int flags) void audit_core_dumps(long signr) { struct audit_buffer *ab; - u32 sid; - uid_t auid = audit_get_loginuid(current), uid; - gid_t gid; - unsigned int sessionid = audit_get_sessionid(current); if (!audit_enabled) return; @@ -2551,24 +2566,17 @@ void audit_core_dumps(long signr) return; ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_ANOM_ABEND); - current_uid_gid(&uid, &gid); - audit_log_format(ab, "auid=%u uid=%u gid=%u ses=%u", - auid, uid, gid, sessionid); - security_task_getsecid(current, &sid); - if (sid) { - char *ctx = NULL; - u32 len; + audit_log_abend(ab, "memory violation", signr); + audit_log_end(ab); +} - if (security_secid_to_secctx(sid, &ctx, &len)) - audit_log_format(ab, " ssid=%u", sid); - else { - audit_log_format(ab, " subj=%s", ctx); - security_release_secctx(ctx, len); - } - } - audit_log_format(ab, " pid=%d comm=", current->pid); - audit_log_untrustedstring(ab, current->comm); - audit_log_format(ab, " sig=%ld", signr); +void __audit_seccomp(unsigned long syscall) +{ + struct audit_buffer *ab; + + ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_ANOM_ABEND); + audit_log_abend(ab, "seccomp", SIGKILL); + audit_log_format(ab, " syscall=%ld", syscall); audit_log_end(ab); } diff --git a/kernel/seccomp.c b/kernel/seccomp.c index 57d4b13b631d..e8d76c5895ea 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -6,6 +6,7 @@ * This defines a simple but solid secure-computing mode. */ +#include #include #include #include @@ -54,6 +55,7 @@ void __secure_computing(int this_syscall) #ifdef SECCOMP_DEBUG dump_stack(); #endif + audit_seccomp(this_syscall); do_exit(SIGKILL); } From d7e7528bcd456f5c36ad4a202ccfb43c5aa98bc4 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:06 -0500 Subject: [PATCH 531/820] Audit: push audit success and retcode into arch ptrace.h The audit system previously expected arches calling to audit_syscall_exit to supply as arguments if the syscall was a success and what the return code was. Audit also provides a helper AUDITSC_RESULT which was supposed to simplify things by converting from negative retcodes to an audit internal magic value stating success or failure. This helper was wrong and could indicate that a valid pointer returned to userspace was a failed syscall. The fix is to fix the layering foolishness. We now pass audit_syscall_exit a struct pt_reg and it in turns calls back into arch code to collect the return value and to determine if the syscall was a success or failure. We also define a generic is_syscall_success() macro which determines success/failure based on if the value is < -MAX_ERRNO. This works for arches like x86 which do not use a separate mechanism to indicate syscall failure. We make both the is_syscall_success() and regs_return_value() static inlines instead of macros. The reason is because the audit function must take a void* for the regs. (uml calls theirs struct uml_pt_regs instead of just struct pt_regs so audit_syscall_exit can't take a struct pt_regs). Since the audit function takes a void* we need to use static inlines to cast it back to the arch correct structure to dereference it. The other major change is that on some arches, like ia64, MIPS and ppc, we change regs_return_value() to give us the negative value on syscall failure. THE only other user of this macro, kretprobe_example.c, won't notice and it makes the value signed consistently for the audit functions across all archs. In arch/sh/kernel/ptrace_64.c I see that we were using regs[9] in the old audit code as the return value. But the ptrace_64.h code defined the macro regs_return_value() as regs[3]. I have no idea which one is correct, but this patch now uses the regs_return_value() function, so it now uses regs[3]. For powerpc we previously used regs->result but now use the regs_return_value() function which uses regs->gprs[3]. regs->gprs[3] is always positive so the regs_return_value(), much like ia64 makes it negative before calling the audit code when appropriate. Signed-off-by: Eric Paris Acked-by: H. Peter Anvin [for x86 portion] Acked-by: Tony Luck [for ia64] Acked-by: Richard Weinberger [for uml] Acked-by: David S. Miller [for sparc] Acked-by: Ralf Baechle [for mips] Acked-by: Benjamin Herrenschmidt [for ppc] --- arch/ia64/include/asm/ptrace.h | 13 ++++++++++++- arch/ia64/kernel/ptrace.c | 9 +-------- arch/microblaze/include/asm/ptrace.h | 5 +++++ arch/microblaze/kernel/ptrace.c | 3 +-- arch/mips/include/asm/ptrace.h | 14 +++++++++++++- arch/mips/kernel/ptrace.c | 4 +--- arch/powerpc/include/asm/ptrace.h | 13 ++++++++++++- arch/powerpc/kernel/ptrace.c | 4 +--- arch/s390/include/asm/ptrace.h | 6 +++++- arch/s390/kernel/ptrace.c | 4 +--- arch/sh/include/asm/ptrace_32.h | 5 ++++- arch/sh/include/asm/ptrace_64.h | 5 ++++- arch/sh/kernel/ptrace_32.c | 4 +--- arch/sh/kernel/ptrace_64.c | 4 +--- arch/sparc/include/asm/ptrace.h | 10 +++++++++- arch/sparc/kernel/ptrace_64.c | 11 +---------- arch/um/kernel/ptrace.c | 4 ++-- arch/x86/ia32/ia32entry.S | 10 +++++----- arch/x86/kernel/entry_32.S | 8 ++++---- arch/x86/kernel/entry_64.S | 10 +++++----- arch/x86/kernel/ptrace.c | 3 +-- arch/x86/kernel/vm86_32.c | 4 ++-- arch/x86/um/shared/sysdep/ptrace.h | 5 +++++ include/linux/audit.h | 22 ++++++++++++++-------- include/linux/ptrace.h | 10 ++++++++++ kernel/auditsc.c | 16 ++++++++++++---- 26 files changed, 132 insertions(+), 74 deletions(-) diff --git a/arch/ia64/include/asm/ptrace.h b/arch/ia64/include/asm/ptrace.h index f5cb27614e35..68c98f5b3ca6 100644 --- a/arch/ia64/include/asm/ptrace.h +++ b/arch/ia64/include/asm/ptrace.h @@ -246,7 +246,18 @@ static inline unsigned long user_stack_pointer(struct pt_regs *regs) return regs->ar_bspstore; } -#define regs_return_value(regs) ((regs)->r8) +static inline int is_syscall_success(struct pt_regs *regs) +{ + return regs->r10 != -1; +} + +static inline long regs_return_value(struct pt_regs *regs) +{ + if (is_syscall_success(regs)) + return regs->r8; + else + return -regs->r8; +} /* Conserve space in histogram by encoding slot bits in address * bits 2 and 3 rather than bits 0 and 1. diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 8848f43d819e..2c154088cce7 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -1268,14 +1268,7 @@ syscall_trace_leave (long arg0, long arg1, long arg2, long arg3, { int step; - if (unlikely(current->audit_context)) { - int success = AUDITSC_RESULT(regs.r10); - long result = regs.r8; - - if (success != AUDITSC_SUCCESS) - result = -result; - audit_syscall_exit(success, result); - } + audit_syscall_exit(®s); step = test_thread_flag(TIF_SINGLESTEP); if (step || test_thread_flag(TIF_SYSCALL_TRACE)) diff --git a/arch/microblaze/include/asm/ptrace.h b/arch/microblaze/include/asm/ptrace.h index 816bee64b196..94e92c805859 100644 --- a/arch/microblaze/include/asm/ptrace.h +++ b/arch/microblaze/include/asm/ptrace.h @@ -61,6 +61,11 @@ struct pt_regs { #define instruction_pointer(regs) ((regs)->pc) #define profile_pc(regs) instruction_pointer(regs) +static inline long regs_return_value(struct pt_regs *regs) +{ + return regs->r3; +} + #else /* __KERNEL__ */ /* pt_regs offsets used by gdbserver etc in ptrace syscalls */ diff --git a/arch/microblaze/kernel/ptrace.c b/arch/microblaze/kernel/ptrace.c index 043cb58f9c44..f564b1bfd386 100644 --- a/arch/microblaze/kernel/ptrace.c +++ b/arch/microblaze/kernel/ptrace.c @@ -159,8 +159,7 @@ asmlinkage void do_syscall_trace_leave(struct pt_regs *regs) { int step; - if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(regs->r3), regs->r3); + audit_syscall_exit(regs); step = test_thread_flag(TIF_SINGLESTEP); if (step || test_thread_flag(TIF_SYSCALL_TRACE)) diff --git a/arch/mips/include/asm/ptrace.h b/arch/mips/include/asm/ptrace.h index de39b1f343ea..7d409505df2d 100644 --- a/arch/mips/include/asm/ptrace.h +++ b/arch/mips/include/asm/ptrace.h @@ -137,7 +137,19 @@ extern int ptrace_set_watch_regs(struct task_struct *child, */ #define user_mode(regs) (((regs)->cp0_status & KU_MASK) == KU_USER) -#define regs_return_value(_regs) ((_regs)->regs[2]) +static inline int is_syscall_success(struct pt_regs *regs) +{ + return !regs->regs[7]; +} + +static inline long regs_return_value(struct pt_regs *regs) +{ + if (is_syscall_success(regs)) + return regs->regs[2]; + else + return -regs->regs[2]; +} + #define instruction_pointer(regs) ((regs)->cp0_epc) #define profile_pc(regs) instruction_pointer(regs) diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index 4e6ea1ffad46..ab0f1963a7bd 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -572,9 +572,7 @@ out: */ asmlinkage void syscall_trace_leave(struct pt_regs *regs) { - if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(regs->regs[7]), - -regs->regs[2]); + audit_syscall_exit(regs); if (!(current->ptrace & PT_PTRACED)) return; diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h index 48223f9b8728..78a205162fd7 100644 --- a/arch/powerpc/include/asm/ptrace.h +++ b/arch/powerpc/include/asm/ptrace.h @@ -86,7 +86,18 @@ struct pt_regs { #define instruction_pointer(regs) ((regs)->nip) #define user_stack_pointer(regs) ((regs)->gpr[1]) #define kernel_stack_pointer(regs) ((regs)->gpr[1]) -#define regs_return_value(regs) ((regs)->gpr[3]) +static inline int is_syscall_success(struct pt_regs *regs) +{ + return !(regs->ccr & 0x10000000); +} + +static inline long regs_return_value(struct pt_regs *regs) +{ + if (is_syscall_success(regs)) + return regs->gpr[3]; + else + return -regs->gpr[3]; +} #ifdef CONFIG_SMP extern unsigned long profile_pc(struct pt_regs *regs); diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c index 5de73dbd15c7..09d31c12a5e3 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -1748,9 +1748,7 @@ void do_syscall_trace_leave(struct pt_regs *regs) { int step; - if (unlikely(current->audit_context)) - audit_syscall_exit((regs->ccr&0x10000000)?AUDITSC_FAILURE:AUDITSC_SUCCESS, - regs->result); + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_exit(regs, regs->result); diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h index 56da355678f4..aeb77f017985 100644 --- a/arch/s390/include/asm/ptrace.h +++ b/arch/s390/include/asm/ptrace.h @@ -541,9 +541,13 @@ struct user_regs_struct #define user_mode(regs) (((regs)->psw.mask & PSW_MASK_PSTATE) != 0) #define instruction_pointer(regs) ((regs)->psw.addr & PSW_ADDR_INSN) #define user_stack_pointer(regs)((regs)->gprs[15]) -#define regs_return_value(regs)((regs)->gprs[2]) #define profile_pc(regs) instruction_pointer(regs) +static inline long regs_return_value(struct pt_regs *regs) +{ + return regs->gprs[2]; +} + int regs_query_register_offset(const char *name); const char *regs_query_register_name(unsigned int offset); unsigned long regs_get_register(struct pt_regs *regs, unsigned int offset); diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 573bc29551ef..f52758600980 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -751,9 +751,7 @@ asmlinkage long do_syscall_trace_enter(struct pt_regs *regs) asmlinkage void do_syscall_trace_exit(struct pt_regs *regs) { - if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(regs->gprs[2]), - regs->gprs[2]); + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_exit(regs, regs->gprs[2]); diff --git a/arch/sh/include/asm/ptrace_32.h b/arch/sh/include/asm/ptrace_32.h index 6c2239cca1a2..2d3e906aa722 100644 --- a/arch/sh/include/asm/ptrace_32.h +++ b/arch/sh/include/asm/ptrace_32.h @@ -76,7 +76,10 @@ struct pt_dspregs { #ifdef __KERNEL__ #define MAX_REG_OFFSET offsetof(struct pt_regs, tra) -#define regs_return_value(_regs) ((_regs)->regs[0]) +static inline long regs_return_value(struct pt_regs *regs) +{ + return regs->regs[0]; +} #endif /* __KERNEL__ */ diff --git a/arch/sh/include/asm/ptrace_64.h b/arch/sh/include/asm/ptrace_64.h index bf9be7764d69..eb3fcceaf64b 100644 --- a/arch/sh/include/asm/ptrace_64.h +++ b/arch/sh/include/asm/ptrace_64.h @@ -13,7 +13,10 @@ struct pt_regs { #ifdef __KERNEL__ #define MAX_REG_OFFSET offsetof(struct pt_regs, tregs[7]) -#define regs_return_value(_regs) ((_regs)->regs[3]) +static inline long regs_return_value(struct pt_regs *regs) +{ + return regs->regs[3]; +} #endif /* __KERNEL__ */ diff --git a/arch/sh/kernel/ptrace_32.c b/arch/sh/kernel/ptrace_32.c index 92b3c276339a..c0b5c179d27b 100644 --- a/arch/sh/kernel/ptrace_32.c +++ b/arch/sh/kernel/ptrace_32.c @@ -530,9 +530,7 @@ asmlinkage void do_syscall_trace_leave(struct pt_regs *regs) { int step; - if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(regs->regs[0]), - regs->regs[0]); + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_exit(regs, regs->regs[0]); diff --git a/arch/sh/kernel/ptrace_64.c b/arch/sh/kernel/ptrace_64.c index c8f97649f354..ba720d686435 100644 --- a/arch/sh/kernel/ptrace_64.c +++ b/arch/sh/kernel/ptrace_64.c @@ -548,9 +548,7 @@ asmlinkage void do_syscall_trace_leave(struct pt_regs *regs) { int step; - if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(regs->regs[9]), - regs->regs[9]); + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_exit(regs, regs->regs[9]); diff --git a/arch/sparc/include/asm/ptrace.h b/arch/sparc/include/asm/ptrace.h index a0e1bcf843a1..c00c3b5c2806 100644 --- a/arch/sparc/include/asm/ptrace.h +++ b/arch/sparc/include/asm/ptrace.h @@ -207,7 +207,15 @@ do { current_thread_info()->syscall_noerror = 1; \ #define instruction_pointer(regs) ((regs)->tpc) #define instruction_pointer_set(regs, val) ((regs)->tpc = (val)) #define user_stack_pointer(regs) ((regs)->u_regs[UREG_FP]) -#define regs_return_value(regs) ((regs)->u_regs[UREG_I0]) +static inline int is_syscall_success(struct pt_regs *regs) +{ + return !(regs->tstate & (TSTATE_XCARRY | TSTATE_ICARRY)); +} + +static inline long regs_return_value(struct pt_regs *regs) +{ + return regs->u_regs[UREG_I0]; +} #ifdef CONFIG_SMP extern unsigned long profile_pc(struct pt_regs *); #else diff --git a/arch/sparc/kernel/ptrace_64.c b/arch/sparc/kernel/ptrace_64.c index 96ee50a80661..c73c8c50f117 100644 --- a/arch/sparc/kernel/ptrace_64.c +++ b/arch/sparc/kernel/ptrace_64.c @@ -1086,17 +1086,8 @@ asmlinkage int syscall_trace_enter(struct pt_regs *regs) asmlinkage void syscall_trace_leave(struct pt_regs *regs) { -#ifdef CONFIG_AUDITSYSCALL - if (unlikely(current->audit_context)) { - unsigned long tstate = regs->tstate; - int result = AUDITSC_SUCCESS; + audit_syscall_exit(regs); - if (unlikely(tstate & (TSTATE_XCARRY | TSTATE_ICARRY))) - result = AUDITSC_FAILURE; - - audit_syscall_exit(result, regs->u_regs[UREG_I0]); - } -#endif if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_exit(regs, regs->u_regs[UREG_G1]); diff --git a/arch/um/kernel/ptrace.c b/arch/um/kernel/ptrace.c index c9da32b0c707..2ccf25c42feb 100644 --- a/arch/um/kernel/ptrace.c +++ b/arch/um/kernel/ptrace.c @@ -175,8 +175,8 @@ void syscall_trace(struct uml_pt_regs *regs, int entryexit) UPT_SYSCALL_ARG2(regs), UPT_SYSCALL_ARG3(regs), UPT_SYSCALL_ARG4(regs)); - else audit_syscall_exit(AUDITSC_RESULT(UPT_SYSCALL_RET(regs)), - UPT_SYSCALL_RET(regs)); + else + audit_syscall_exit(regs); } /* Fake a debug trap */ diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index 3e274564f6bf..64ced0b8f8fd 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -14,6 +14,7 @@ #include #include #include +#include /* Avoid __ASSEMBLER__'ifying just for this. */ #include @@ -208,12 +209,11 @@ sysexit_from_sys_call: TRACE_IRQS_ON sti movl %eax,%esi /* second arg, syscall return value */ - cmpl $0,%eax /* is it < 0? */ - setl %al /* 1 if so, 0 if not */ + cmpl $-MAX_ERRNO,%eax /* is it an error ? */ + setbe %al /* 1 if so, 0 if not */ movzbl %al,%edi /* zero-extend that into %edi */ - inc %edi /* first arg, 0->1(AUDITSC_SUCCESS), 1->2(AUDITSC_FAILURE) */ - call audit_syscall_exit - movl RAX-ARGOFFSET(%rsp),%eax /* reload syscall return value */ + call __audit_syscall_exit + movq RAX-ARGOFFSET(%rsp),%rax /* reload syscall return value */ movl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT),%edi cli TRACE_IRQS_OFF diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 22d0e21b4dd7..a22facf06f0e 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -42,6 +42,7 @@ */ #include +#include #include #include #include @@ -466,11 +467,10 @@ sysexit_audit: TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_ANY) movl %eax,%edx /* second arg, syscall return value */ - cmpl $0,%eax /* is it < 0? */ - setl %al /* 1 if so, 0 if not */ + cmpl $-MAX_ERRNO,%eax /* is it an error ? */ + setbe %al /* 1 if so, 0 if not */ movzbl %al,%eax /* zero-extend that */ - inc %eax /* first arg, 0->1(AUDITSC_SUCCESS), 1->2(AUDITSC_FAILURE) */ - call audit_syscall_exit + call __audit_syscall_exit DISABLE_INTERRUPTS(CLBR_ANY) TRACE_IRQS_OFF movl TI_flags(%ebp), %ecx diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index a20e1cb9dc87..e51393dd93a3 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -55,6 +55,7 @@ #include #include #include +#include /* Avoid __ASSEMBLER__'ifying just for this. */ #include @@ -563,17 +564,16 @@ auditsys: jmp system_call_fastpath /* - * Return fast path for syscall audit. Call audit_syscall_exit() + * Return fast path for syscall audit. Call __audit_syscall_exit() * directly and then jump back to the fast path with TIF_SYSCALL_AUDIT * masked off. */ sysret_audit: movq RAX-ARGOFFSET(%rsp),%rsi /* second arg, syscall return value */ - cmpq $0,%rsi /* is it < 0? */ - setl %al /* 1 if so, 0 if not */ + cmpq $-MAX_ERRNO,%rsi /* is it < -MAX_ERRNO? */ + setbe %al /* 1 if so, 0 if not */ movzbl %al,%edi /* zero-extend that into %edi */ - inc %edi /* first arg, 0->1(AUDITSC_SUCCESS), 1->2(AUDITSC_FAILURE) */ - call audit_syscall_exit + call __audit_syscall_exit movl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT),%edi jmp sysret_check #endif /* CONFIG_AUDITSYSCALL */ diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 89a04c7b5bb6..8b0218758775 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -1414,8 +1414,7 @@ void syscall_trace_leave(struct pt_regs *regs) { bool step; - if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax); + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_exit(regs, regs->ax); diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index 863f8753ab0a..af17e1c966dc 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -335,9 +335,9 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk if (info->flags & VM86_SCREEN_BITMAP) mark_screen_rdonly(tsk->mm); - /*call audit_syscall_exit since we do not exit via the normal paths */ + /*call __audit_syscall_exit since we do not exit via the normal paths */ if (unlikely(current->audit_context)) - audit_syscall_exit(AUDITSC_RESULT(0), 0); + __audit_syscall_exit(1, 0); __asm__ __volatile__( "movl %0,%%esp\n\t" diff --git a/arch/x86/um/shared/sysdep/ptrace.h b/arch/x86/um/shared/sysdep/ptrace.h index 711b1621747f..5ef9344a8b24 100644 --- a/arch/x86/um/shared/sysdep/ptrace.h +++ b/arch/x86/um/shared/sysdep/ptrace.h @@ -3,3 +3,8 @@ #else #include "ptrace_64.h" #endif + +static inline long regs_return_value(struct uml_pt_regs *regs) +{ + return UPT_SYSCALL_RET(regs); +} diff --git a/include/linux/audit.h b/include/linux/audit.h index 6e1c533f9b46..3d65e4b3ba06 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -26,6 +26,7 @@ #include #include +#include /* The netlink messages for the audit system is divided into blocks: * 1000 - 1099 are for commanding the audit system @@ -408,10 +409,6 @@ struct audit_field { void *lsm_rule; }; -#define AUDITSC_INVALID 0 -#define AUDITSC_SUCCESS 1 -#define AUDITSC_FAILURE 2 -#define AUDITSC_RESULT(x) ( ((long)(x))<0?AUDITSC_FAILURE:AUDITSC_SUCCESS ) extern int __init audit_register_class(int class, unsigned *list); extern int audit_classify_syscall(int abi, unsigned syscall); extern int audit_classify_arch(int arch); @@ -424,7 +421,7 @@ extern void audit_free(struct task_struct *task); extern void audit_syscall_entry(int arch, int major, unsigned long a0, unsigned long a1, unsigned long a2, unsigned long a3); -extern void audit_syscall_exit(int failed, long return_code); +extern void __audit_syscall_exit(int ret_success, long ret_value); extern void __audit_getname(const char *name); extern void audit_putname(const char *name); extern void __audit_inode(const char *name, const struct dentry *dentry); @@ -438,6 +435,15 @@ static inline int audit_dummy_context(void) void *p = current->audit_context; return !p || *(int *)p; } +static inline void audit_syscall_exit(void *pt_regs) +{ + if (unlikely(current->audit_context)) { + int success = is_syscall_success(pt_regs); + int return_code = regs_return_value(pt_regs); + + __audit_syscall_exit(success, return_code); + } +} static inline void audit_getname(const char *name) { if (unlikely(!audit_dummy_context())) @@ -551,12 +557,12 @@ static inline void audit_mmap_fd(int fd, int flags) extern int audit_n_rules; extern int audit_signals; -#else +#else /* CONFIG_AUDITSYSCALL */ #define audit_finish_fork(t) #define audit_alloc(t) ({ 0; }) #define audit_free(t) do { ; } while (0) #define audit_syscall_entry(ta,a,b,c,d,e) do { ; } while (0) -#define audit_syscall_exit(f,r) do { ; } while (0) +#define audit_syscall_exit(r) do { ; } while (0) #define audit_dummy_context() 1 #define audit_getname(n) do { ; } while (0) #define audit_putname(n) do { ; } while (0) @@ -587,7 +593,7 @@ extern int audit_signals; #define audit_ptrace(t) ((void)0) #define audit_n_rules 0 #define audit_signals 0 -#endif +#endif /* CONFIG_AUDITSYSCALL */ #ifdef CONFIG_AUDIT /* These are defined in audit.c */ diff --git a/include/linux/ptrace.h b/include/linux/ptrace.h index 800f113bea66..dd4cefa6519d 100644 --- a/include/linux/ptrace.h +++ b/include/linux/ptrace.h @@ -112,6 +112,7 @@ #include /* For unlikely. */ #include /* For struct task_struct. */ +#include /* for IS_ERR_VALUE */ extern long arch_ptrace(struct task_struct *child, long request, @@ -265,6 +266,15 @@ static inline void ptrace_release_task(struct task_struct *task) #define force_successful_syscall_return() do { } while (0) #endif +#ifndef is_syscall_success +/* + * On most systems we can tell if a syscall is a success based on if the retval + * is an error value. On some systems like ia64 and powerpc they have different + * indicators of success/failure and must define their own. + */ +#define is_syscall_success(regs) (!IS_ERR_VALUE((unsigned long)(regs_return_value(regs)))) +#endif + /* * should define the following things inside #ifdef __KERNEL__. * diff --git a/kernel/auditsc.c b/kernel/auditsc.c index e9bcb93800d8..3d2853808185 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -70,6 +70,11 @@ #include "audit.h" +/* flags stating the success for a syscall */ +#define AUDITSC_INVALID 0 +#define AUDITSC_SUCCESS 1 +#define AUDITSC_FAILURE 2 + /* AUDIT_NAMES is the number of slots we reserve in the audit_context * for saving names from getname(). If we get more names we will allocate * a name dynamically and also add those to the list anchored by names_list. */ @@ -1724,8 +1729,7 @@ void audit_finish_fork(struct task_struct *child) /** * audit_syscall_exit - deallocate audit context after a system call - * @valid: success/failure flag - * @return_code: syscall return value + * @pt_regs: syscall registers * * Tear down after system call. If the audit context has been marked as * auditable (either because of the AUDIT_RECORD_CONTEXT state from @@ -1733,13 +1737,17 @@ void audit_finish_fork(struct task_struct *child) * message), then write out the syscall information. In call cases, * free the names stored from getname(). */ -void audit_syscall_exit(int valid, long return_code) +void __audit_syscall_exit(int success, long return_code) { struct task_struct *tsk = current; struct audit_context *context; - context = audit_get_context(tsk, valid, return_code); + if (success) + success = AUDITSC_SUCCESS; + else + success = AUDITSC_FAILURE; + context = audit_get_context(tsk, success, return_code); if (likely(!context)) return; From f031cd25568a390dc2c9c3a4015054183753449a Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:06 -0500 Subject: [PATCH 532/820] audit: ia32entry.S sign extend error codes when calling 64 bit code In the ia32entry syscall exit audit fastpath we have assembly code which calls __audit_syscall_exit directly. This code was, however, zeroes the upper 32 bits of the return code. It then proceeded to call code which expects longs to be 64bits long. In order to handle code which expects longs to be 64bit we sign extend the return code if that code is an error. Thus the __audit_syscall_exit function can correctly handle using the values in snprintf("%ld"). This fixes the regression introduced in 5cbf1565f29eb57a86a. Old record: type=SYSCALL msg=audit(1306197182.256:281): arch=40000003 syscall=192 success=no exit=4294967283 New record: type=SYSCALL msg=audit(1306197182.256:281): arch=40000003 syscall=192 success=no exit=-13 Signed-off-by: Eric Paris Acked-by: H. Peter Anvin --- arch/x86/ia32/ia32entry.S | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index 64ced0b8f8fd..025f0f01d254 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -210,7 +210,9 @@ sysexit_from_sys_call: sti movl %eax,%esi /* second arg, syscall return value */ cmpl $-MAX_ERRNO,%eax /* is it an error ? */ - setbe %al /* 1 if so, 0 if not */ + jbe 1f + movslq %eax, %rsi /* if error sign extend to 64 bits */ +1: setbe %al /* 1 if error, 0 if not */ movzbl %al,%edi /* zero-extend that into %edi */ call __audit_syscall_exit movq RAX-ARGOFFSET(%rsp),%rax /* reload syscall return value */ From b05d8447e7821695bc2fa3359431f7a664232743 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:06 -0500 Subject: [PATCH 533/820] audit: inline audit_syscall_entry to reduce burden on archs Every arch calls: if (unlikely(current->audit_context)) audit_syscall_entry() which requires knowledge about audit (the existance of audit_context) in the arch code. Just do it all in static inline in audit.h so that arch's can remain blissfully ignorant. Signed-off-by: Eric Paris --- arch/ia64/kernel/ptrace.c | 9 +-------- arch/microblaze/kernel/ptrace.c | 6 ++---- arch/mips/kernel/ptrace.c | 7 +++---- arch/powerpc/kernel/ptrace.c | 26 ++++++++++++-------------- arch/s390/kernel/ptrace.c | 11 +++++------ arch/sh/kernel/ptrace_32.c | 7 +++---- arch/sh/kernel/ptrace_64.c | 7 +++---- arch/sparc/kernel/ptrace_64.c | 17 ++++++++--------- arch/um/kernel/ptrace.c | 20 +++++++++----------- arch/x86/ia32/ia32entry.S | 2 +- arch/x86/kernel/entry_32.S | 2 +- arch/x86/kernel/entry_64.S | 4 ++-- arch/x86/kernel/ptrace.c | 22 ++++++++++------------ arch/xtensa/kernel/ptrace.c | 3 +-- include/linux/audit.h | 13 ++++++++++--- kernel/auditsc.c | 2 +- 16 files changed, 72 insertions(+), 86 deletions(-) diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 2c154088cce7..dad91661ddf9 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -1246,15 +1246,8 @@ syscall_trace_enter (long arg0, long arg1, long arg2, long arg3, if (test_thread_flag(TIF_RESTORE_RSE)) ia64_sync_krbs(); - if (unlikely(current->audit_context)) { - long syscall; - int arch; - syscall = regs.r15; - arch = AUDIT_ARCH_IA64; - - audit_syscall_entry(arch, syscall, arg0, arg1, arg2, arg3); - } + audit_syscall_entry(AUDIT_ARCH_IA64, regs.r15, arg0, arg1, arg2, arg3); return 0; } diff --git a/arch/microblaze/kernel/ptrace.c b/arch/microblaze/kernel/ptrace.c index f564b1bfd386..6eb2aa927d89 100644 --- a/arch/microblaze/kernel/ptrace.c +++ b/arch/microblaze/kernel/ptrace.c @@ -147,10 +147,8 @@ asmlinkage long do_syscall_trace_enter(struct pt_regs *regs) */ ret = -1L; - if (unlikely(current->audit_context)) - audit_syscall_entry(EM_MICROBLAZE, regs->r12, - regs->r5, regs->r6, - regs->r7, regs->r8); + audit_syscall_entry(EM_MICROBLAZE, regs->r12, regs->r5, regs->r6, + regs->r7, regs->r8); return ret ?: regs->r12; } diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index ab0f1963a7bd..7786b608d932 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -560,10 +560,9 @@ asmlinkage void syscall_trace_enter(struct pt_regs *regs) } out: - if (unlikely(current->audit_context)) - audit_syscall_entry(audit_arch(), regs->regs[2], - regs->regs[4], regs->regs[5], - regs->regs[6], regs->regs[7]); + audit_syscall_entry(audit_arch(), regs->regs[2], + regs->regs[4], regs->regs[5], + regs->regs[6], regs->regs[7]); } /* diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c index 09d31c12a5e3..5b43325402bc 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -1724,22 +1724,20 @@ long do_syscall_trace_enter(struct pt_regs *regs) if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->gpr[0]); - if (unlikely(current->audit_context)) { #ifdef CONFIG_PPC64 - if (!is_32bit_task()) - audit_syscall_entry(AUDIT_ARCH_PPC64, - regs->gpr[0], - regs->gpr[3], regs->gpr[4], - regs->gpr[5], regs->gpr[6]); - else + if (!is_32bit_task()) + audit_syscall_entry(AUDIT_ARCH_PPC64, + regs->gpr[0], + regs->gpr[3], regs->gpr[4], + regs->gpr[5], regs->gpr[6]); + else #endif - audit_syscall_entry(AUDIT_ARCH_PPC, - regs->gpr[0], - regs->gpr[3] & 0xffffffff, - regs->gpr[4] & 0xffffffff, - regs->gpr[5] & 0xffffffff, - regs->gpr[6] & 0xffffffff); - } + audit_syscall_entry(AUDIT_ARCH_PPC, + regs->gpr[0], + regs->gpr[3] & 0xffffffff, + regs->gpr[4] & 0xffffffff, + regs->gpr[5] & 0xffffffff, + regs->gpr[6] & 0xffffffff); return ret ?: regs->gpr[0]; } diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index f52758600980..9d82ed4bcb27 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -740,12 +740,11 @@ asmlinkage long do_syscall_trace_enter(struct pt_regs *regs) if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->gprs[2]); - if (unlikely(current->audit_context)) - audit_syscall_entry(is_compat_task() ? - AUDIT_ARCH_S390 : AUDIT_ARCH_S390X, - regs->gprs[2], regs->orig_gpr2, - regs->gprs[3], regs->gprs[4], - regs->gprs[5]); + audit_syscall_entry(is_compat_task() ? + AUDIT_ARCH_S390 : AUDIT_ARCH_S390X, + regs->gprs[2], regs->orig_gpr2, + regs->gprs[3], regs->gprs[4], + regs->gprs[5]); return ret ?: regs->gprs[2]; } diff --git a/arch/sh/kernel/ptrace_32.c b/arch/sh/kernel/ptrace_32.c index c0b5c179d27b..a3e651563763 100644 --- a/arch/sh/kernel/ptrace_32.c +++ b/arch/sh/kernel/ptrace_32.c @@ -518,10 +518,9 @@ asmlinkage long do_syscall_trace_enter(struct pt_regs *regs) if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->regs[0]); - if (unlikely(current->audit_context)) - audit_syscall_entry(audit_arch(), regs->regs[3], - regs->regs[4], regs->regs[5], - regs->regs[6], regs->regs[7]); + audit_syscall_entry(audit_arch(), regs->regs[3], + regs->regs[4], regs->regs[5], + regs->regs[6], regs->regs[7]); return ret ?: regs->regs[0]; } diff --git a/arch/sh/kernel/ptrace_64.c b/arch/sh/kernel/ptrace_64.c index ba720d686435..3d0080b5c976 100644 --- a/arch/sh/kernel/ptrace_64.c +++ b/arch/sh/kernel/ptrace_64.c @@ -536,10 +536,9 @@ asmlinkage long long do_syscall_trace_enter(struct pt_regs *regs) if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->regs[9]); - if (unlikely(current->audit_context)) - audit_syscall_entry(audit_arch(), regs->regs[1], - regs->regs[2], regs->regs[3], - regs->regs[4], regs->regs[5]); + audit_syscall_entry(audit_arch(), regs->regs[1], + regs->regs[2], regs->regs[3], + regs->regs[4], regs->regs[5]); return ret ?: regs->regs[9]; } diff --git a/arch/sparc/kernel/ptrace_64.c b/arch/sparc/kernel/ptrace_64.c index c73c8c50f117..9388844cd88c 100644 --- a/arch/sparc/kernel/ptrace_64.c +++ b/arch/sparc/kernel/ptrace_64.c @@ -1071,15 +1071,14 @@ asmlinkage int syscall_trace_enter(struct pt_regs *regs) if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->u_regs[UREG_G1]); - if (unlikely(current->audit_context) && !ret) - audit_syscall_entry((test_thread_flag(TIF_32BIT) ? - AUDIT_ARCH_SPARC : - AUDIT_ARCH_SPARC64), - regs->u_regs[UREG_G1], - regs->u_regs[UREG_I0], - regs->u_regs[UREG_I1], - regs->u_regs[UREG_I2], - regs->u_regs[UREG_I3]); + audit_syscall_entry((test_thread_flag(TIF_32BIT) ? + AUDIT_ARCH_SPARC : + AUDIT_ARCH_SPARC64), + regs->u_regs[UREG_G1], + regs->u_regs[UREG_I0], + regs->u_regs[UREG_I1], + regs->u_regs[UREG_I2], + regs->u_regs[UREG_I3]); return ret; } diff --git a/arch/um/kernel/ptrace.c b/arch/um/kernel/ptrace.c index 2ccf25c42feb..06b190390505 100644 --- a/arch/um/kernel/ptrace.c +++ b/arch/um/kernel/ptrace.c @@ -167,17 +167,15 @@ void syscall_trace(struct uml_pt_regs *regs, int entryexit) int is_singlestep = (current->ptrace & PT_DTRACE) && entryexit; int tracesysgood; - if (unlikely(current->audit_context)) { - if (!entryexit) - audit_syscall_entry(HOST_AUDIT_ARCH, - UPT_SYSCALL_NR(regs), - UPT_SYSCALL_ARG1(regs), - UPT_SYSCALL_ARG2(regs), - UPT_SYSCALL_ARG3(regs), - UPT_SYSCALL_ARG4(regs)); - else - audit_syscall_exit(regs); - } + if (!entryexit) + audit_syscall_entry(HOST_AUDIT_ARCH, + UPT_SYSCALL_NR(regs), + UPT_SYSCALL_ARG1(regs), + UPT_SYSCALL_ARG2(regs), + UPT_SYSCALL_ARG3(regs), + UPT_SYSCALL_ARG4(regs)); + else + audit_syscall_exit(regs); /* Fake a debug trap */ if (is_singlestep) diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index 025f0f01d254..cecfd9a8f734 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -192,7 +192,7 @@ sysexit_from_sys_call: movl %ebx,%edx /* 3rd arg: 1st syscall arg */ movl %eax,%esi /* 2nd arg: syscall number */ movl $AUDIT_ARCH_I386,%edi /* 1st arg: audit arch */ - call audit_syscall_entry + call __audit_syscall_entry movl RAX-ARGOFFSET(%rsp),%eax /* reload syscall number */ cmpq $(IA32_NR_syscalls-1),%rax ja ia32_badsys diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index a22facf06f0e..1ccd742eba1b 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -456,7 +456,7 @@ sysenter_audit: movl %ebx,%ecx /* 3rd arg: 1st syscall arg */ movl %eax,%edx /* 2nd arg: syscall number */ movl $AUDIT_ARCH_I386,%eax /* 1st arg: audit arch */ - call audit_syscall_entry + call __audit_syscall_entry pushl_cfi %ebx movl PT_EAX(%esp),%eax /* reload syscall number */ jmp sysenter_do_call diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index e51393dd93a3..1ca66b650123 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -549,7 +549,7 @@ badsys: #ifdef CONFIG_AUDITSYSCALL /* * Fast path for syscall audit without full syscall trace. - * We just call audit_syscall_entry() directly, and then + * We just call __audit_syscall_entry() directly, and then * jump back to the normal fast path. */ auditsys: @@ -559,7 +559,7 @@ auditsys: movq %rdi,%rdx /* 3rd arg: 1st syscall arg */ movq %rax,%rsi /* 2nd arg: syscall number */ movl $AUDIT_ARCH_X86_64,%edi /* 1st arg: audit arch */ - call audit_syscall_entry + call __audit_syscall_entry LOAD_ARGS 0 /* reload call-clobbered registers */ jmp system_call_fastpath diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 8b0218758775..50267386b766 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -1392,20 +1392,18 @@ long syscall_trace_enter(struct pt_regs *regs) if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) trace_sys_enter(regs, regs->orig_ax); - if (unlikely(current->audit_context)) { - if (IS_IA32) - audit_syscall_entry(AUDIT_ARCH_I386, - regs->orig_ax, - regs->bx, regs->cx, - regs->dx, regs->si); + if (IS_IA32) + audit_syscall_entry(AUDIT_ARCH_I386, + regs->orig_ax, + regs->bx, regs->cx, + regs->dx, regs->si); #ifdef CONFIG_X86_64 - else - audit_syscall_entry(AUDIT_ARCH_X86_64, - regs->orig_ax, - regs->di, regs->si, - regs->dx, regs->r10); + else + audit_syscall_entry(AUDIT_ARCH_X86_64, + regs->orig_ax, + regs->di, regs->si, + regs->dx, regs->r10); #endif - } return ret ?: regs->orig_ax; } diff --git a/arch/xtensa/kernel/ptrace.c b/arch/xtensa/kernel/ptrace.c index a0d042aa2967..2dff698ab02e 100644 --- a/arch/xtensa/kernel/ptrace.c +++ b/arch/xtensa/kernel/ptrace.c @@ -334,8 +334,7 @@ void do_syscall_trace_enter(struct pt_regs *regs) do_syscall_trace(); #if 0 - if (unlikely(current->audit_context)) - audit_syscall_entry(current, AUDIT_ARCH_XTENSA..); + audit_syscall_entry(current, AUDIT_ARCH_XTENSA..); #endif } diff --git a/include/linux/audit.h b/include/linux/audit.h index 3d65e4b3ba06..f56ce2669b83 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -418,9 +418,9 @@ extern int audit_classify_arch(int arch); extern void audit_finish_fork(struct task_struct *child); extern int audit_alloc(struct task_struct *task); extern void audit_free(struct task_struct *task); -extern void audit_syscall_entry(int arch, - int major, unsigned long a0, unsigned long a1, - unsigned long a2, unsigned long a3); +extern void __audit_syscall_entry(int arch, + int major, unsigned long a0, unsigned long a1, + unsigned long a2, unsigned long a3); extern void __audit_syscall_exit(int ret_success, long ret_value); extern void __audit_getname(const char *name); extern void audit_putname(const char *name); @@ -435,6 +435,13 @@ static inline int audit_dummy_context(void) void *p = current->audit_context; return !p || *(int *)p; } +static inline void audit_syscall_entry(int arch, int major, unsigned long a0, + unsigned long a1, unsigned long a2, + unsigned long a3) +{ + if (unlikely(!audit_dummy_context())) + __audit_syscall_entry(arch, major, a0, a1, a2, a3); +} static inline void audit_syscall_exit(void *pt_regs) { if (unlikely(current->audit_context)) { diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 3d2853808185..b408100dd6ef 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1632,7 +1632,7 @@ void audit_free(struct task_struct *tsk) * will only be written if another part of the kernel requests that it * be written). */ -void audit_syscall_entry(int arch, int major, +void __audit_syscall_entry(int arch, int major, unsigned long a1, unsigned long a2, unsigned long a3, unsigned long a4) { From 997f5b6444f4608692ec807fb802fd9767c80e76 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:06 -0500 Subject: [PATCH 534/820] audit: remove AUDIT_SETUP_CONTEXT as it isn't used Audit contexts have 3 states. Disabled, which doesn't collect anything, build, which collects info but might not emit it, and record, which collects and emits. There is a 4th state, setup, which isn't used. Get rid of it. Signed-off-by: Eric Paris --- kernel/audit.h | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/kernel/audit.h b/kernel/audit.h index 91e7071c4d2c..816766803371 100644 --- a/kernel/audit.h +++ b/kernel/audit.h @@ -36,12 +36,8 @@ enum audit_state { AUDIT_DISABLED, /* Do not create per-task audit_context. * No syscall-specific audit records can * be generated. */ - AUDIT_SETUP_CONTEXT, /* Create the per-task audit_context, - * but don't necessarily fill it in at - * syscall entry time (i.e., filter - * instead). */ AUDIT_BUILD_CONTEXT, /* Create the per-task audit_context, - * and always fill it in at syscall + * and fill it in at syscall * entry time. This makes a full * syscall record available if some * other part of the kernel decides it From 56179a6ec65a56e0279a58e35cb450d38f061b94 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:06 -0500 Subject: [PATCH 535/820] audit: drop some potentially inadvisable likely notations The audit code makes heavy use of likely() and unlikely() macros, but they don't always make sense. Drop any that seem questionable and let the computer do it's thing. Signed-off-by: Eric Paris --- kernel/auditsc.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index b408100dd6ef..d7382c2aaa9e 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -805,7 +805,7 @@ static inline struct audit_context *audit_get_context(struct task_struct *tsk, { struct audit_context *context = tsk->audit_context; - if (likely(!context)) + if (!context) return NULL; context->return_valid = return_valid; @@ -928,7 +928,7 @@ int audit_alloc(struct task_struct *tsk) return 0; /* Return if not auditing. */ state = audit_filter_task(tsk, &key); - if (likely(state == AUDIT_DISABLED)) + if (state == AUDIT_DISABLED) return 0; if (!(context = audit_alloc_context(state))) { @@ -1599,7 +1599,7 @@ void audit_free(struct task_struct *tsk) struct audit_context *context; context = audit_get_context(tsk, 0, 0); - if (likely(!context)) + if (!context) return; /* Check for system calls that do not go through the exit @@ -1640,7 +1640,7 @@ void __audit_syscall_entry(int arch, int major, struct audit_context *context = tsk->audit_context; enum audit_state state; - if (unlikely(!context)) + if (!context) return; /* @@ -1697,7 +1697,7 @@ void __audit_syscall_entry(int arch, int major, context->prio = 0; state = audit_filter_syscall(tsk, context, &audit_filter_list[AUDIT_FILTER_ENTRY]); } - if (likely(state == AUDIT_DISABLED)) + if (state == AUDIT_DISABLED) return; context->serial = 0; @@ -1748,7 +1748,7 @@ void __audit_syscall_exit(int success, long return_code) success = AUDITSC_FAILURE; context = audit_get_context(tsk, success, return_code); - if (likely(!context)) + if (!context) return; if (context->in_syscall && context->current_state == AUDIT_RECORD_CONTEXT) From 07c49417877f8658a6aa0ad9b4e21e4fd4df11b6 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 536/820] audit: inline checks for not needing to collect aux records A number of audit hooks make function calls before they determine that auxilary records do not need to be collected. Do those checks as static inlines since the most common case is going to be that records are not needed and we can skip the function call overhead. Signed-off-by: Eric Paris --- include/linux/audit.h | 23 ++++++++++++++++++++--- kernel/auditsc.c | 15 +++------------ 2 files changed, 23 insertions(+), 15 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index f56ce2669b83..cf16faff6b8a 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -489,9 +489,9 @@ extern int audit_set_loginuid(struct task_struct *task, uid_t loginuid); extern void audit_log_task_context(struct audit_buffer *ab); extern void __audit_ipc_obj(struct kern_ipc_perm *ipcp); extern void __audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, umode_t mode); -extern int audit_bprm(struct linux_binprm *bprm); -extern void audit_socketcall(int nargs, unsigned long *args); -extern int audit_sockaddr(int len, void *addr); +extern int __audit_bprm(struct linux_binprm *bprm); +extern void __audit_socketcall(int nargs, unsigned long *args); +extern int __audit_sockaddr(int len, void *addr); extern void __audit_fd_pair(int fd1, int fd2); extern int audit_set_macxattr(const char *name); extern void __audit_mq_open(int oflag, umode_t mode, struct mq_attr *attr); @@ -519,6 +519,23 @@ static inline void audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid if (unlikely(!audit_dummy_context())) __audit_ipc_set_perm(qbytes, uid, gid, mode); } +static inline int audit_bprm(struct linux_binprm *bprm) +{ + if (unlikely(!audit_dummy_context())) + return __audit_bprm(bprm); + return 0; +} +static inline void audit_socketcall(int nargs, unsigned long *args) +{ + if (unlikely(!audit_dummy_context())) + __audit_socketcall(nargs, args); +} +static inline int audit_sockaddr(int len, void *addr) +{ + if (unlikely(!audit_dummy_context())) + return __audit_sockaddr(len, addr); + return 0; +} static inline void audit_mq_open(int oflag, umode_t mode, struct mq_attr *attr) { if (unlikely(!audit_dummy_context())) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index d7382c2aaa9e..e1062f66b01b 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -2309,14 +2309,11 @@ void __audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, umode_t mo context->ipc.has_perm = 1; } -int audit_bprm(struct linux_binprm *bprm) +int __audit_bprm(struct linux_binprm *bprm) { struct audit_aux_data_execve *ax; struct audit_context *context = current->audit_context; - if (likely(!audit_enabled || !context || context->dummy)) - return 0; - ax = kmalloc(sizeof(*ax), GFP_KERNEL); if (!ax) return -ENOMEM; @@ -2337,13 +2334,10 @@ int audit_bprm(struct linux_binprm *bprm) * @args: args array * */ -void audit_socketcall(int nargs, unsigned long *args) +void __audit_socketcall(int nargs, unsigned long *args) { struct audit_context *context = current->audit_context; - if (likely(!context || context->dummy)) - return; - context->type = AUDIT_SOCKETCALL; context->socketcall.nargs = nargs; memcpy(context->socketcall.args, args, nargs * sizeof(unsigned long)); @@ -2369,13 +2363,10 @@ void __audit_fd_pair(int fd1, int fd2) * * Returns 0 for success or NULL context or < 0 on error. */ -int audit_sockaddr(int len, void *a) +int __audit_sockaddr(int len, void *a) { struct audit_context *context = current->audit_context; - if (likely(!context || context->dummy)) - return 0; - if (!context->sockaddr) { void *p = kmalloc(sizeof(struct sockaddr_storage), GFP_KERNEL); if (!p) From 38cdce53daa0408a61fe6d86fe48f31515c9b840 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 537/820] audit: drop audit_set_macxattr as it doesn't do anything unused. deleted. Signed-off-by: Eric Paris --- include/linux/audit.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index cf16faff6b8a..4f1efe3e8616 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -493,7 +493,6 @@ extern int __audit_bprm(struct linux_binprm *bprm); extern void __audit_socketcall(int nargs, unsigned long *args); extern int __audit_sockaddr(int len, void *addr); extern void __audit_fd_pair(int fd1, int fd2); -extern int audit_set_macxattr(const char *name); extern void __audit_mq_open(int oflag, umode_t mode, struct mq_attr *attr); extern void __audit_mq_sendrecv(mqd_t mqdes, size_t msg_len, unsigned int msg_prio, const struct timespec *abs_timeout); extern void __audit_mq_notify(mqd_t mqdes, const struct sigevent *notification); @@ -606,7 +605,6 @@ extern int audit_signals; #define audit_socketcall(n,a) ((void)0) #define audit_fd_pair(n,a) ((void)0) #define audit_sockaddr(len, addr) ({ 0; }) -#define audit_set_macxattr(n) do { ; } while (0) #define audit_mq_open(o,m,a) ((void)0) #define audit_mq_sendrecv(d,l,p,t) ((void)0) #define audit_mq_notify(d,n) ((void)0) From a4ff8dba7d8ce5ceb43fb27df66292251cc73bdc Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 538/820] audit: inline audit_free to simplify the look of generic code make the conditional a static inline instead of doing it in generic code. Signed-off-by: Eric Paris --- include/linux/audit.h | 7 ++++++- kernel/auditsc.c | 2 +- kernel/exit.c | 3 +-- 3 files changed, 8 insertions(+), 4 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 4f1efe3e8616..8eb8bda749b3 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -417,7 +417,7 @@ extern int audit_classify_arch(int arch); /* Public API */ extern void audit_finish_fork(struct task_struct *child); extern int audit_alloc(struct task_struct *task); -extern void audit_free(struct task_struct *task); +extern void __audit_free(struct task_struct *task); extern void __audit_syscall_entry(int arch, int major, unsigned long a0, unsigned long a1, unsigned long a2, unsigned long a3); @@ -435,6 +435,11 @@ static inline int audit_dummy_context(void) void *p = current->audit_context; return !p || *(int *)p; } +static inline void audit_free(struct task_struct *task) +{ + if (unlikely(task->audit_context)) + __audit_free(task); +} static inline void audit_syscall_entry(int arch, int major, unsigned long a0, unsigned long a1, unsigned long a2, unsigned long a3) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index e1062f66b01b..7aaeb38b262a 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1594,7 +1594,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts * * Called from copy_process and do_exit */ -void audit_free(struct task_struct *tsk) +void __audit_free(struct task_struct *tsk) { struct audit_context *context; diff --git a/kernel/exit.c b/kernel/exit.c index 94ed6e20bb53..88dcbbc446f7 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -964,8 +964,7 @@ NORET_TYPE void do_exit(long code) acct_collect(code, group_dead); if (group_dead) tty_audit_exit(); - if (unlikely(tsk->audit_context)) - audit_free(tsk); + audit_free(tsk); tsk->exit_code = code; taskstats_exit(tsk, group_dead); From 7ff68e53ece8c175d2951bb8a30b3cce8f9c5579 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 539/820] audit: reject entry,always rules We deprecated entry,always rules a long time ago. Reject those rules as invalid. Signed-off-by: Eric Paris --- kernel/auditfilter.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index d94dde82c3c8..903caa269b5c 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -235,13 +235,15 @@ static inline struct audit_entry *audit_to_entry_common(struct audit_rule *rule) switch(listnr) { default: goto exit_err; - case AUDIT_FILTER_USER: - case AUDIT_FILTER_TYPE: #ifdef CONFIG_AUDITSYSCALL case AUDIT_FILTER_ENTRY: + if (rule->action == AUDIT_ALWAYS) + goto exit_err; case AUDIT_FILTER_EXIT: case AUDIT_FILTER_TASK: #endif + case AUDIT_FILTER_USER: + case AUDIT_FILTER_TYPE: ; } if (unlikely(rule->action == AUDIT_POSSIBLE)) { From 6422e78de6880c66a82af512d9bd0c85eb62e661 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 540/820] audit: remove audit_finish_fork as it can't be called Audit entry,always rules are not allowed and are automatically changed in exit,always rules in userspace. The kernel refuses to load such rules. Thus a task in the middle of a syscall (and thus in audit_finish_fork()) can only be in one of two states: AUDIT_BUILD_CONTEXT or AUDIT_DISABLED. Since the current task cannot be in AUDIT_RECORD_CONTEXT we aren't every going to actually use the code in audit_finish_fork() since it will return without doing anything. Thus drop the code. Signed-off-by: Eric Paris --- include/linux/audit.h | 2 -- kernel/auditsc.c | 20 -------------------- kernel/fork.c | 2 -- 3 files changed, 24 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 8eb8bda749b3..67b66c37a254 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -415,7 +415,6 @@ extern int audit_classify_arch(int arch); #ifdef CONFIG_AUDITSYSCALL /* These are defined in auditsc.c */ /* Public API */ -extern void audit_finish_fork(struct task_struct *child); extern int audit_alloc(struct task_struct *task); extern void __audit_free(struct task_struct *task); extern void __audit_syscall_entry(int arch, @@ -586,7 +585,6 @@ static inline void audit_mmap_fd(int fd, int flags) extern int audit_n_rules; extern int audit_signals; #else /* CONFIG_AUDITSYSCALL */ -#define audit_finish_fork(t) #define audit_alloc(t) ({ 0; }) #define audit_free(t) do { ; } while (0) #define audit_syscall_entry(ta,a,b,c,d,e) do { ; } while (0) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 7aaeb38b262a..4d8920f5ab88 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1707,26 +1707,6 @@ void __audit_syscall_entry(int arch, int major, context->ppid = 0; } -void audit_finish_fork(struct task_struct *child) -{ - struct audit_context *ctx = current->audit_context; - struct audit_context *p = child->audit_context; - if (!p || !ctx) - return; - if (!ctx->in_syscall || ctx->current_state != AUDIT_RECORD_CONTEXT) - return; - p->arch = ctx->arch; - p->major = ctx->major; - memcpy(p->argv, ctx->argv, sizeof(ctx->argv)); - p->ctime = ctx->ctime; - p->dummy = ctx->dummy; - p->in_syscall = ctx->in_syscall; - p->filterkey = kstrdup(ctx->filterkey, GFP_KERNEL); - p->ppid = current->pid; - p->prio = ctx->prio; - p->current_state = ctx->current_state; -} - /** * audit_syscall_exit - deallocate audit context after a system call * @pt_regs: syscall registers diff --git a/kernel/fork.c b/kernel/fork.c index 443f5125f11e..c1e5c21f48c1 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1525,8 +1525,6 @@ long do_fork(unsigned long clone_flags, init_completion(&vfork); } - audit_finish_fork(p); - /* * We set PF_STARTING at creation in case tracing wants to * use this to distinguish a fully live task from one that From efaffd6e4417860c67576ac760dd6e8bbd15f006 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 541/820] audit: allow matching on obj_uid Allow syscall exit filter matching based on the uid of the owner of an inode used in a syscall. aka: auditctl -a always,exit -S open -F obj_uid=0 -F perm=wa Signed-off-by: Eric Paris --- include/linux/audit.h | 1 + kernel/auditfilter.c | 1 + kernel/auditsc.c | 12 ++++++++++++ 3 files changed, 14 insertions(+) diff --git a/include/linux/audit.h b/include/linux/audit.h index 67b66c37a254..55cb3daaf474 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -223,6 +223,7 @@ #define AUDIT_PERM 106 #define AUDIT_DIR 107 #define AUDIT_FILETYPE 108 +#define AUDIT_OBJ_UID 109 #define AUDIT_ARG0 200 #define AUDIT_ARG1 (AUDIT_ARG0+1) diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index 903caa269b5c..13e997423dcd 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -461,6 +461,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, case AUDIT_ARG1: case AUDIT_ARG2: case AUDIT_ARG3: + case AUDIT_OBJ_UID: break; case AUDIT_ARCH: entry->rule.arch_f = f; diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 4d8920f5ab88..5cf3ecc01517 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -586,6 +586,18 @@ static int audit_filter_rules(struct task_struct *tsk, } } break; + case AUDIT_OBJ_UID: + if (name) { + result = audit_comparator(name->uid, f->op, f->val); + } else if (ctx) { + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_comparator(n->uid, f->op, f->val)) { + ++result; + break; + } + } + } + break; case AUDIT_WATCH: if (name) result = audit_watch_compare(rule->watch, name->ino, name->dev); From 54d3218b31aee5bc9c859ae60fbde933d922448b Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:07 -0500 Subject: [PATCH 542/820] audit: allow audit matching on inode gid Much like the ability to filter audit on the uid of an inode collected, we should be able to filter on the gid of the inode. Signed-off-by: Eric Paris --- include/linux/audit.h | 1 + kernel/auditfilter.c | 1 + kernel/auditsc.c | 12 ++++++++++++ 3 files changed, 14 insertions(+) diff --git a/include/linux/audit.h b/include/linux/audit.h index 55cb3daaf474..e36aa37c88af 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -224,6 +224,7 @@ #define AUDIT_DIR 107 #define AUDIT_FILETYPE 108 #define AUDIT_OBJ_UID 109 +#define AUDIT_OBJ_GID 110 #define AUDIT_ARG0 200 #define AUDIT_ARG1 (AUDIT_ARG0+1) diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index 13e997423dcd..f10605c787e6 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -462,6 +462,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, case AUDIT_ARG2: case AUDIT_ARG3: case AUDIT_OBJ_UID: + case AUDIT_OBJ_GID: break; case AUDIT_ARCH: entry->rule.arch_f = f; diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 5cf3ecc01517..87b375fb12ff 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -598,6 +598,18 @@ static int audit_filter_rules(struct task_struct *tsk, } } break; + case AUDIT_OBJ_GID: + if (name) { + result = audit_comparator(name->gid, f->op, f->val); + } else if (ctx) { + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_comparator(n->gid, f->op, f->val)) { + ++result; + break; + } + } + } + break; case AUDIT_WATCH: if (name) result = audit_watch_compare(rule->watch, name->ino, name->dev); From 0a300be6d5be8f66cd96609334710c268d0bfdce Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:08 -0500 Subject: [PATCH 543/820] audit: remove task argument to audit_set_loginuid The function always deals with current. Don't expose an option pretending one can use it for something. You can't. Signed-off-by: Eric Paris --- fs/proc/base.c | 2 +- include/linux/audit.h | 2 +- kernel/auditsc.c | 6 +++--- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index 8173dfd89cb2..e3cbebbabebd 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1228,7 +1228,7 @@ static ssize_t proc_loginuid_write(struct file * file, const char __user * buf, goto out_free_page; } - length = audit_set_loginuid(current, loginuid); + length = audit_set_loginuid(loginuid); if (likely(length == 0)) length = count; diff --git a/include/linux/audit.h b/include/linux/audit.h index e36aa37c88af..7cbd6fe41573 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -489,7 +489,7 @@ static inline void audit_ptrace(struct task_struct *t) extern unsigned int audit_serial(void); extern int auditsc_get_stamp(struct audit_context *ctx, struct timespec *t, unsigned int *serial); -extern int audit_set_loginuid(struct task_struct *task, uid_t loginuid); +extern int audit_set_loginuid(uid_t loginuid); #define audit_get_loginuid(t) ((t)->loginuid) #define audit_get_sessionid(t) ((t)->sessionid) extern void audit_log_task_context(struct audit_buffer *ab); diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 87b375fb12ff..9d6dd7d869c0 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -2163,16 +2163,16 @@ int auditsc_get_stamp(struct audit_context *ctx, static atomic_t session_id = ATOMIC_INIT(0); /** - * audit_set_loginuid - set a task's audit_context loginuid - * @task: task whose audit context is being modified + * audit_set_loginuid - set current task's audit_context loginuid * @loginuid: loginuid value * * Returns 0. * * Called (set) from fs/proc/base.c::proc_loginuid_write(). */ -int audit_set_loginuid(struct task_struct *task, uid_t loginuid) +int audit_set_loginuid(uid_t loginuid) { + struct task_struct *task = current; unsigned int sessionid = atomic_inc_return(&session_id); struct audit_context *context = task->audit_context; From 633b45454503489209b0d9a45f9e3cd1b852c614 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:08 -0500 Subject: [PATCH 544/820] audit: only allow tasks to set their loginuid if it is -1 At the moment we allow tasks to set their loginuid if they have CAP_AUDIT_CONTROL. In reality we want tasks to set the loginuid when they log in and it be impossible to ever reset. We had to make it mutable even after it was once set (with the CAP) because on update and admin might have to restart sshd. Now sshd would get his loginuid and the next user which logged in using ssh would not be able to set his loginuid. Systemd has changed how userspace works and allowed us to make the kernel work the way it should. With systemd users (even admins) are not supposed to restart services directly. The system will restart the service for them. Thus since systemd is going to loginuid==-1, sshd would get -1, and sshd would be allowed to set a new loginuid without special permissions. If an admin in this system were to manually start an sshd he is inserting himself into the system chain of trust and thus, logically, it's his loginuid that should be used! Since we have old systems I make this a Kconfig option. Signed-off-by: Eric Paris --- fs/proc/base.c | 3 --- init/Kconfig | 14 ++++++++++++++ kernel/auditsc.c | 11 ++++++++++- 3 files changed, 24 insertions(+), 4 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index e3cbebbabebd..482df23036b5 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1197,9 +1197,6 @@ static ssize_t proc_loginuid_write(struct file * file, const char __user * buf, ssize_t length; uid_t loginuid; - if (!capable(CAP_AUDIT_CONTROL)) - return -EPERM; - rcu_read_lock(); if (current != pid_task(proc_pid(inode), PIDTYPE_PID)) { rcu_read_unlock(); diff --git a/init/Kconfig b/init/Kconfig index a075765d5fbe..5ad8b775f2ac 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -372,6 +372,20 @@ config AUDIT_TREE depends on AUDITSYSCALL select FSNOTIFY +config AUDIT_LOGINUID_IMMUTABLE + bool "Make audit loginuid immutable" + depends on AUDIT + help + The config option toggles if a task setting it's loginuid requires + CAP_SYS_AUDITCONTROL or if that task should require no special permissions + but should instead only allow setting its loginuid if it was never + previously set. On systems which use systemd or a similar central + process to restart login services this should be set to true. On older + systems in which an admin would typically have to directly stop and + start processes this should be set to false. Setting this to true allows + one to drop potentially dangerous capabilites from the login tasks, + but may not be backwards compatible with older init systems. + source "kernel/irq/Kconfig" menu "RCU Subsystem" diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 9d6dd7d869c0..bd084a13c719 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -2173,9 +2173,18 @@ static atomic_t session_id = ATOMIC_INIT(0); int audit_set_loginuid(uid_t loginuid) { struct task_struct *task = current; - unsigned int sessionid = atomic_inc_return(&session_id); struct audit_context *context = task->audit_context; + unsigned int sessionid; +#ifdef CONFIG_AUDIT_LOGINUID_IMMUTABLE + if (task->loginuid != -1) + return -EPERM; +#else /* CONFIG_AUDIT_LOGINUID_IMMUTABLE */ + if (!capable(CAP_AUDIT_CONTROL)) + return -EPERM; +#endif /* CONFIG_AUDIT_LOGINUID_IMMUTABLE */ + + sessionid = atomic_inc_return(&session_id); if (context && context->in_syscall) { struct audit_buffer *ab; From 4043cde8ecf7f7d880eb1133c201a3d392fd68c3 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:08 -0500 Subject: [PATCH 545/820] audit: do not call audit_getname on error Just a code cleanup really. We don't need to make a function call just for it to return on error. This also makes the VFS function even easier to follow and removes a conditional on a hot path. Signed-off-by: Eric Paris --- fs/namei.c | 24 +++++++++++------------- kernel/auditsc.c | 3 --- 2 files changed, 11 insertions(+), 16 deletions(-) diff --git a/fs/namei.c b/fs/namei.c index c283a1ec008e..208c6aa4a989 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -140,21 +140,19 @@ static int do_getname(const char __user *filename, char *page) static char *getname_flags(const char __user *filename, int flags, int *empty) { - char *tmp, *result; + char *result = __getname(); + int retval; - result = ERR_PTR(-ENOMEM); - tmp = __getname(); - if (tmp) { - int retval = do_getname(filename, tmp); + if (!result) + return ERR_PTR(-ENOMEM); - result = tmp; - if (retval < 0) { - if (retval == -ENOENT && empty) - *empty = 1; - if (retval != -ENOENT || !(flags & LOOKUP_EMPTY)) { - __putname(tmp); - result = ERR_PTR(retval); - } + retval = do_getname(filename, result); + if (retval < 0) { + if (retval == -ENOENT && empty) + *empty = 1; + if (retval != -ENOENT || !(flags & LOOKUP_EMPTY)) { + __putname(result); + return ERR_PTR(retval); } } audit_getname(result); diff --git a/kernel/auditsc.c b/kernel/auditsc.c index bd084a13c719..9161e70a4379 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1913,9 +1913,6 @@ void __audit_getname(const char *name) struct audit_context *context = current->audit_context; struct audit_names *n; - if (IS_ERR(name) || !name) - return; - if (!context->in_syscall) { #if AUDIT_DEBUG == 2 printk(KERN_ERR "%s:%d(:%d): ignoring getname(%p)\n", From 29ef73b7a823b77a7cd0bdd7d7cded3fb6c2587b Mon Sep 17 00:00:00 2001 From: Nathaniel Husted Date: Tue, 3 Jan 2012 14:23:09 -0500 Subject: [PATCH 546/820] Kernel: Audit Support For The ARM Platform This patch provides functionality to audit system call events on the ARM platform. The implementation was based off the structure of the MIPS platform and information in this (http://lists.fedoraproject.org/pipermail/arm/2009-October/000382.html) mailing list thread. The required audit_syscall_exit and audit_syscall_entry checks were added to ptrace using the standard registers for system call values (r0 through r3). A thread information flag was added for auditing (TIF_SYSCALL_AUDIT) and a meta-flag was added (_TIF_SYSCALL_WORK) to simplify modifications to the syscall entry/exit. Now, if either the TRACE flag is set or the AUDIT flag is set, the syscall_trace function will be executed. The prober changes were made to Kconfig to allow CONFIG_AUDITSYSCALL to be enabled. Due to platform availability limitations, this patch was only tested on the Android platform running the modified "android-goldfish-2.6.29" kernel. A test compile was performed using Code Sourcery's cross-compilation toolset and the current linux-3.0 stable kernel. The changes compile without error. I'm hoping, due to the simple modifications, the patch is "obviously correct". Signed-off-by: Nathaniel Husted Signed-off-by: Eric Paris --- arch/arm/include/asm/kprobes.h | 1 - arch/arm/include/asm/ptrace.h | 5 +++++ arch/arm/include/asm/thread_info.h | 6 ++++++ arch/arm/kernel/entry-common.S | 4 ++-- arch/arm/kernel/ptrace.c | 16 +++++++++++----- init/Kconfig | 2 +- 6 files changed, 25 insertions(+), 9 deletions(-) diff --git a/arch/arm/include/asm/kprobes.h b/arch/arm/include/asm/kprobes.h index feec86768f9c..f82ec22eeb11 100644 --- a/arch/arm/include/asm/kprobes.h +++ b/arch/arm/include/asm/kprobes.h @@ -24,7 +24,6 @@ #define MAX_INSN_SIZE 2 #define MAX_STACK_SIZE 64 /* 32 would probably be OK */ -#define regs_return_value(regs) ((regs)->ARM_r0) #define flush_insn_slot(p) do { } while (0) #define kretprobe_blacklist_size 0 diff --git a/arch/arm/include/asm/ptrace.h b/arch/arm/include/asm/ptrace.h index 96187ff58c24..451808ba1211 100644 --- a/arch/arm/include/asm/ptrace.h +++ b/arch/arm/include/asm/ptrace.h @@ -189,6 +189,11 @@ static inline int valid_user_regs(struct pt_regs *regs) return 0; } +static inline long regs_return_value(struct pt_regs *regs) +{ + return regs->ARM_r0; +} + #define instruction_pointer(regs) (regs)->ARM_pc #ifdef CONFIG_SMP diff --git a/arch/arm/include/asm/thread_info.h b/arch/arm/include/asm/thread_info.h index 0f30c3a78fc1..d4c24d412a8d 100644 --- a/arch/arm/include/asm/thread_info.h +++ b/arch/arm/include/asm/thread_info.h @@ -129,6 +129,7 @@ extern void vfp_flush_hwstate(struct thread_info *); /* * thread information flags: * TIF_SYSCALL_TRACE - syscall trace active + * TIF_SYSCAL_AUDIT - syscall auditing active * TIF_SIGPENDING - signal pending * TIF_NEED_RESCHED - rescheduling necessary * TIF_NOTIFY_RESUME - callback before returning to user @@ -139,6 +140,7 @@ extern void vfp_flush_hwstate(struct thread_info *); #define TIF_NEED_RESCHED 1 #define TIF_NOTIFY_RESUME 2 /* callback before returning to user */ #define TIF_SYSCALL_TRACE 8 +#define TIF_SYSCALL_AUDIT 9 #define TIF_POLLING_NRFLAG 16 #define TIF_USING_IWMMXT 17 #define TIF_MEMDIE 18 /* is terminating due to OOM killer */ @@ -149,11 +151,15 @@ extern void vfp_flush_hwstate(struct thread_info *); #define _TIF_NEED_RESCHED (1 << TIF_NEED_RESCHED) #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) +#define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT) #define _TIF_POLLING_NRFLAG (1 << TIF_POLLING_NRFLAG) #define _TIF_USING_IWMMXT (1 << TIF_USING_IWMMXT) #define _TIF_RESTORE_SIGMASK (1 << TIF_RESTORE_SIGMASK) #define _TIF_SECCOMP (1 << TIF_SECCOMP) +/* Checks for any syscall work in entry-common.S */ +#define _TIF_SYSCALL_WORK (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT) + /* * Change these and you break ASM code in entry-common.S */ diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index b2a27b6b0046..520889cf1b5b 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -87,7 +87,7 @@ ENTRY(ret_from_fork) get_thread_info tsk ldr r1, [tsk, #TI_FLAGS] @ check for syscall tracing mov why, #1 - tst r1, #_TIF_SYSCALL_TRACE @ are we tracing syscalls? + tst r1, #_TIF_SYSCALL_WORK @ are we tracing syscalls? beq ret_slow_syscall mov r1, sp mov r0, #1 @ trace exit [IP = 1] @@ -443,7 +443,7 @@ ENTRY(vector_swi) 1: #endif - tst r10, #_TIF_SYSCALL_TRACE @ are we tracing syscalls? + tst r10, #_TIF_SYSCALL_WORK @ are we tracing syscalls? bne __sys_trace cmp scno, #NR_syscalls @ check upper syscall limit diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index 483727ad6892..e1d5e1929fbd 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -906,11 +906,6 @@ asmlinkage int syscall_trace(int why, struct pt_regs *regs, int scno) { unsigned long ip; - if (!test_thread_flag(TIF_SYSCALL_TRACE)) - return scno; - if (!(current->ptrace & PT_PTRACED)) - return scno; - /* * Save IP. IP is used to denote syscall entry/exit: * IP = 0 -> entry, = 1 -> exit @@ -918,6 +913,17 @@ asmlinkage int syscall_trace(int why, struct pt_regs *regs, int scno) ip = regs->ARM_ip; regs->ARM_ip = why; + if (!ip) + audit_syscall_exit(regs); + else + audit_syscall_entry(AUDIT_ARCH_ARMEB, scno, regs->ARM_r0, + regs->ARM_r1, regs->ARM_r2, regs->ARM_r3); + + if (!test_thread_flag(TIF_SYSCALL_TRACE)) + return scno; + if (!(current->ptrace & PT_PTRACED)) + return scno; + current_thread_info()->syscall = scno; /* the 0x80 provides a way for the tracing parent to distinguish diff --git a/init/Kconfig b/init/Kconfig index 5ad8b775f2ac..fe25ffbe818b 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -355,7 +355,7 @@ config AUDIT config AUDITSYSCALL bool "Enable system-call auditing support" - depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH) + depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH || ARM) default y if SECURITY_SELINUX help Enable low-overhead system-call auditing infrastructure that From 02d86a568c6d2d335256864451ac8ce781bc5652 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:08 -0500 Subject: [PATCH 547/820] audit: allow interfield comparison in audit rules We wish to be able to audit when a uid=500 task accesses a file which is uid=0. Or vice versa. This patch introduces a new audit filter type AUDIT_FIELD_COMPARE which takes as an 'enum' which indicates which fields should be compared. At this point we only define the task->uid vs inode->uid, but other comparisons can be added. Signed-off-by: Eric Paris --- include/linux/audit.h | 4 ++++ kernel/auditfilter.c | 5 ++++- kernel/auditsc.c | 30 +++++++++++++++++++++++++++++- 3 files changed, 37 insertions(+), 2 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 7cbd6fe41573..838e05fc0582 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -182,7 +182,10 @@ * AUDIT_UNUSED_BITS is updated if need be. */ #define AUDIT_UNUSED_BITS 0x07FFFC00 +/* AUDIT_FIELD_COMPARE rule list */ +#define AUDIT_COMPARE_UID_TO_OBJ_UID 1 +#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_UID_TO_OBJ_UID /* Rule fields */ /* These are useful when checking the * task structure at task creation time @@ -225,6 +228,7 @@ #define AUDIT_FILETYPE 108 #define AUDIT_OBJ_UID 109 #define AUDIT_OBJ_GID 110 +#define AUDIT_FIELD_COMPARE 111 #define AUDIT_ARG0 200 #define AUDIT_ARG1 (AUDIT_ARG0+1) diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index f10605c787e6..a6c3f1abd206 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -526,7 +526,6 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, goto exit_free; break; case AUDIT_FILTERKEY: - err = -EINVAL; if (entry->rule.filterkey || f->val > AUDIT_MAX_KEY_LEN) goto exit_free; str = audit_unpack_string(&bufp, &remain, f->val); @@ -543,6 +542,10 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, if (f->val & ~S_IFMT) goto exit_free; break; + case AUDIT_FIELD_COMPARE: + if (f->val > AUDIT_MAX_FIELD_COMPARE) + goto exit_free; + break; default: goto exit_free; } diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 9161e70a4379..8fb2c8e6d624 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -463,6 +463,32 @@ static int match_tree_refs(struct audit_context *ctx, struct audit_tree *tree) return 0; } +static int audit_field_compare(struct task_struct *tsk, + const struct cred *cred, + struct audit_field *f, + struct audit_context *ctx, + struct audit_names *name) +{ + struct audit_names *n; + + switch (f->val) { + case AUDIT_COMPARE_UID_TO_OBJ_UID: + if (name) { + return audit_comparator(cred->uid, f->op, name->uid); + } else if (ctx) { + list_for_each_entry(n, &ctx->names_list, list) { + if (audit_comparator(cred->uid, f->op, n->uid)) + return 1; + } + } + break; + default: + WARN(1, "Missing AUDIT_COMPARE define. Report as a bug\n"); + return 0; + } + return 0; +} + /* Determine if any context name data matches a rule's watch data */ /* Compare a task_struct with an audit_rule. Return 1 on match, 0 * otherwise. @@ -693,8 +719,10 @@ static int audit_filter_rules(struct task_struct *tsk, case AUDIT_FILETYPE: result = audit_match_filetype(ctx, f->val); break; + case AUDIT_FIELD_COMPARE: + result = audit_field_compare(tsk, cred, f, ctx, name); + break; } - if (!result) return 0; } From b34b039324bf081554ee8678f9b8c5d937e5206c Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:08 -0500 Subject: [PATCH 548/820] audit: complex interfield comparison helper Rather than code the same loop over and over implement a helper function which uses some pointer magic to make it generic enough to be used numerous places as we implement more audit interfield comparisons Signed-off-by: Eric Paris --- kernel/auditsc.c | 50 +++++++++++++++++++++++++++++++++++++----------- 1 file changed, 39 insertions(+), 11 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 8fb2c8e6d624..b12cc32fe377 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -463,25 +463,53 @@ static int match_tree_refs(struct audit_context *ctx, struct audit_tree *tree) return 0; } +static int audit_compare_id(uid_t uid1, + struct audit_names *name, + unsigned long name_offset, + struct audit_field *f, + struct audit_context *ctx) +{ + struct audit_names *n; + unsigned long addr; + uid_t uid2; + int rc; + + if (name) { + addr = (unsigned long)name; + addr += name_offset; + + uid2 = *(uid_t *)addr; + rc = audit_comparator(uid1, f->op, uid2); + if (rc) + return rc; + } + + if (ctx) { + list_for_each_entry(n, &ctx->names_list, list) { + addr = (unsigned long)n; + addr += name_offset; + + uid2 = *(uid_t *)addr; + + rc = audit_comparator(uid1, f->op, uid2); + if (rc) + return rc; + } + } + return 0; +} + static int audit_field_compare(struct task_struct *tsk, const struct cred *cred, struct audit_field *f, struct audit_context *ctx, struct audit_names *name) { - struct audit_names *n; - switch (f->val) { case AUDIT_COMPARE_UID_TO_OBJ_UID: - if (name) { - return audit_comparator(cred->uid, f->op, name->uid); - } else if (ctx) { - list_for_each_entry(n, &ctx->names_list, list) { - if (audit_comparator(cred->uid, f->op, n->uid)) - return 1; - } - } - break; + return audit_compare_id(cred->uid, + name, offsetof(struct audit_names, uid), + f, ctx); default: WARN(1, "Missing AUDIT_COMPARE define. Report as a bug\n"); return 0; From c9fe685f7a17a0ee8bf3fbe51e40b1c8b8e65896 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 3 Jan 2012 14:23:08 -0500 Subject: [PATCH 549/820] audit: allow interfield comparison between gid and ogid Allow audit rules to compare the gid of the running task to the gid of the inode in question. Signed-off-by: Eric Paris --- include/linux/audit.h | 3 ++- kernel/auditsc.c | 6 ++++++ 2 files changed, 8 insertions(+), 1 deletion(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 838e05fc0582..fffbc2176ee1 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -184,8 +184,9 @@ /* AUDIT_FIELD_COMPARE rule list */ #define AUDIT_COMPARE_UID_TO_OBJ_UID 1 +#define AUDIT_COMPARE_GID_TO_OBJ_GID 2 -#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_UID_TO_OBJ_UID +#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_GID_TO_OBJ_GID /* Rule fields */ /* These are useful when checking the * task structure at task creation time diff --git a/kernel/auditsc.c b/kernel/auditsc.c index b12cc32fe377..861c7b9c565a 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -474,6 +474,8 @@ static int audit_compare_id(uid_t uid1, uid_t uid2; int rc; + BUILD_BUG_ON(sizeof(uid_t) != sizeof(gid_t)); + if (name) { addr = (unsigned long)name; addr += name_offset; @@ -510,6 +512,10 @@ static int audit_field_compare(struct task_struct *tsk, return audit_compare_id(cred->uid, name, offsetof(struct audit_names, uid), f, ctx); + case AUDIT_COMPARE_GID_TO_OBJ_GID: + return audit_compare_id(cred->gid, + name, offsetof(struct audit_names, gid), + f, ctx); default: WARN(1, "Missing AUDIT_COMPARE define. Report as a bug\n"); return 0; From 4a6633ed08af5ba67790b4d1adcdeb8ceb55677e Mon Sep 17 00:00:00 2001 From: Peter Moody Date: Tue, 13 Dec 2011 16:17:51 -0800 Subject: [PATCH 550/820] audit: implement all object interfield comparisons This completes the matrix of interfield comparisons between uid/gid information for the current task and the uid/gid information for inodes. aka I can audit based on differences between the euid of the process and the uid of fs objects. Signed-off-by: Peter Moody Signed-off-by: Eric Paris --- include/linux/audit.h | 10 +++++++++- kernel/auditsc.c | 29 +++++++++++++++++++++++++++++ 2 files changed, 38 insertions(+), 1 deletion(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index fffbc2176ee1..67113cb4bc15 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -185,8 +185,16 @@ /* AUDIT_FIELD_COMPARE rule list */ #define AUDIT_COMPARE_UID_TO_OBJ_UID 1 #define AUDIT_COMPARE_GID_TO_OBJ_GID 2 +#define AUDIT_COMPARE_EUID_TO_OBJ_UID 3 +#define AUDIT_COMPARE_EGID_TO_OBJ_GID 4 +#define AUDIT_COMPARE_AUID_TO_OBJ_UID 5 +#define AUDIT_COMPARE_SUID_TO_OBJ_UID 6 +#define AUDIT_COMPARE_SGID_TO_OBJ_GID 7 +#define AUDIT_COMPARE_FSUID_TO_OBJ_UID 8 +#define AUDIT_COMPARE_FSGID_TO_OBJ_GID 9 + +#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_FSGID_TO_OBJ_GID -#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_GID_TO_OBJ_GID /* Rule fields */ /* These are useful when checking the * task structure at task creation time diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 861c7b9c565a..b8cee462b99e 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -508,6 +508,7 @@ static int audit_field_compare(struct task_struct *tsk, struct audit_names *name) { switch (f->val) { + /* process to file object comparisons */ case AUDIT_COMPARE_UID_TO_OBJ_UID: return audit_compare_id(cred->uid, name, offsetof(struct audit_names, uid), @@ -516,6 +517,34 @@ static int audit_field_compare(struct task_struct *tsk, return audit_compare_id(cred->gid, name, offsetof(struct audit_names, gid), f, ctx); + case AUDIT_COMPARE_EUID_TO_OBJ_UID: + return audit_compare_id(cred->euid, + name, offsetof(struct audit_names, uid), + f, ctx); + case AUDIT_COMPARE_EGID_TO_OBJ_GID: + return audit_compare_id(cred->egid, + name, offsetof(struct audit_names, gid), + f, ctx); + case AUDIT_COMPARE_AUID_TO_OBJ_UID: + return audit_compare_id(tsk->loginuid, + name, offsetof(struct audit_names, uid), + f, ctx); + case AUDIT_COMPARE_SUID_TO_OBJ_UID: + return audit_compare_id(cred->suid, + name, offsetof(struct audit_names, uid), + f, ctx); + case AUDIT_COMPARE_SGID_TO_OBJ_GID: + return audit_compare_id(cred->sgid, + name, offsetof(struct audit_names, gid), + f, ctx); + case AUDIT_COMPARE_FSUID_TO_OBJ_UID: + return audit_compare_id(cred->fsuid, + name, offsetof(struct audit_names, uid), + f, ctx); + case AUDIT_COMPARE_FSGID_TO_OBJ_GID: + return audit_compare_id(cred->fsgid, + name, offsetof(struct audit_names, gid), + f, ctx); default: WARN(1, "Missing AUDIT_COMPARE define. Report as a bug\n"); return 0; From 10d68360871657204885371cdf2594412675d2f9 Mon Sep 17 00:00:00 2001 From: Peter Moody Date: Wed, 4 Jan 2012 15:24:31 -0500 Subject: [PATCH 551/820] audit: comparison on interprocess fields This allows audit to specify rules in which we compare two fields of a process. Such as is the running process uid != to the running process euid? Signed-off-by: Peter Moody Signed-off-by: Eric Paris --- include/linux/audit.h | 24 +++++++++++++++++++++++- kernel/auditsc.c | 39 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 62 insertions(+), 1 deletion(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 67113cb4bc15..9ff7a2c48b50 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -193,7 +193,29 @@ #define AUDIT_COMPARE_FSUID_TO_OBJ_UID 8 #define AUDIT_COMPARE_FSGID_TO_OBJ_GID 9 -#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_FSGID_TO_OBJ_GID +#define AUDIT_COMPARE_UID_TO_AUID 10 +#define AUDIT_COMPARE_UID_TO_EUID 11 +#define AUDIT_COMPARE_UID_TO_FSUID 12 +#define AUDIT_COMPARE_UID_TO_SUID 13 + +#define AUDIT_COMPARE_AUID_TO_FSUID 14 +#define AUDIT_COMPARE_AUID_TO_SUID 15 +#define AUDIT_COMPARE_AUID_TO_EUID 16 + +#define AUDIT_COMPARE_EUID_TO_SUID 17 +#define AUDIT_COMPARE_EUID_TO_FSUID 18 + +#define AUDIT_COMPARE_SUID_TO_FSUID 19 + +#define AUDIT_COMPARE_GID_TO_EGID 20 +#define AUDIT_COMPARE_GID_TO_FSGID 21 +#define AUDIT_COMPARE_GID_TO_SGID 22 + +#define AUDIT_COMPARE_EGID_TO_FSGID 23 +#define AUDIT_COMPARE_EGID_TO_SGID 24 +#define AUDIT_COMPARE_SGID_TO_FSGID 25 + +#define AUDIT_MAX_FIELD_COMPARE AUDIT_COMPARE_SGID_TO_FSGID /* Rule fields */ /* These are useful when checking the diff --git a/kernel/auditsc.c b/kernel/auditsc.c index b8cee462b99e..593237e3654d 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -545,6 +545,45 @@ static int audit_field_compare(struct task_struct *tsk, return audit_compare_id(cred->fsgid, name, offsetof(struct audit_names, gid), f, ctx); + /* uid comparisons */ + case AUDIT_COMPARE_UID_TO_AUID: + return audit_comparator(cred->uid, f->op, tsk->loginuid); + case AUDIT_COMPARE_UID_TO_EUID: + return audit_comparator(cred->uid, f->op, cred->euid); + case AUDIT_COMPARE_UID_TO_SUID: + return audit_comparator(cred->uid, f->op, cred->suid); + case AUDIT_COMPARE_UID_TO_FSUID: + return audit_comparator(cred->uid, f->op, cred->fsuid); + /* auid comparisons */ + case AUDIT_COMPARE_AUID_TO_EUID: + return audit_comparator(tsk->loginuid, f->op, cred->euid); + case AUDIT_COMPARE_AUID_TO_SUID: + return audit_comparator(tsk->loginuid, f->op, cred->suid); + case AUDIT_COMPARE_AUID_TO_FSUID: + return audit_comparator(tsk->loginuid, f->op, cred->fsuid); + /* euid comparisons */ + case AUDIT_COMPARE_EUID_TO_SUID: + return audit_comparator(cred->euid, f->op, cred->suid); + case AUDIT_COMPARE_EUID_TO_FSUID: + return audit_comparator(cred->euid, f->op, cred->fsuid); + /* suid comparisons */ + case AUDIT_COMPARE_SUID_TO_FSUID: + return audit_comparator(cred->suid, f->op, cred->fsuid); + /* gid comparisons */ + case AUDIT_COMPARE_GID_TO_EGID: + return audit_comparator(cred->gid, f->op, cred->egid); + case AUDIT_COMPARE_GID_TO_SGID: + return audit_comparator(cred->gid, f->op, cred->sgid); + case AUDIT_COMPARE_GID_TO_FSGID: + return audit_comparator(cred->gid, f->op, cred->fsgid); + /* egid comparisons */ + case AUDIT_COMPARE_EGID_TO_SGID: + return audit_comparator(cred->egid, f->op, cred->sgid); + case AUDIT_COMPARE_EGID_TO_FSGID: + return audit_comparator(cred->egid, f->op, cred->fsgid); + /* sgid comparison */ + case AUDIT_COMPARE_SGID_TO_FSGID: + return audit_comparator(cred->sgid, f->op, cred->fsgid); default: WARN(1, "Missing AUDIT_COMPARE define. Report as a bug\n"); return 0; From 5afb8a3f96573f7ea018abb768f5b6ebe1a6c1a4 Mon Sep 17 00:00:00 2001 From: Xi Wang Date: Tue, 20 Dec 2011 18:39:41 -0500 Subject: [PATCH 552/820] audit: fix signedness bug in audit_log_execve_info() In the loop, a size_t "len" is used to hold the return value of audit_log_single_execve_arg(), which returns -1 on error. In that case the error handling (len <= 0) will be bypassed since "len" is unsigned, and the loop continues with (p += len) being wrapped. Change the type of "len" to signed int to fix the error handling. size_t len; ... for (...) { len = audit_log_single_execve_arg(...); if (len <= 0) break; p += len; } Signed-off-by: Xi Wang Signed-off-by: Eric Paris --- kernel/auditsc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 593237e3654d..86584ecb1039 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1362,8 +1362,8 @@ static void audit_log_execve_info(struct audit_context *context, struct audit_buffer **ab, struct audit_aux_data_execve *axi) { - int i; - size_t len, len_sent = 0; + int i, len; + size_t len_sent = 0; const char __user *p; char *buf; From 41fdc3054e23e3229edea27053522fe052d02ec2 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Sat, 7 Jan 2012 10:41:04 -0800 Subject: [PATCH 553/820] audit: treat s_id as an untrusted string The use of s_id should go through the untrusted string path, just to be extra careful. Signed-off-by: Kees Cook Acked-by: Mimi Zohar Signed-off-by: Eric Paris --- security/integrity/ima/ima_audit.c | 8 +++++--- security/lsm_audit.c | 23 +++++++++++++---------- 2 files changed, 18 insertions(+), 13 deletions(-) diff --git a/security/integrity/ima/ima_audit.c b/security/integrity/ima/ima_audit.c index c5c5a72c30be..2ad942fb1e23 100644 --- a/security/integrity/ima/ima_audit.c +++ b/security/integrity/ima/ima_audit.c @@ -56,9 +56,11 @@ void integrity_audit_msg(int audit_msgno, struct inode *inode, audit_log_format(ab, " name="); audit_log_untrustedstring(ab, fname); } - if (inode) - audit_log_format(ab, " dev=%s ino=%lu", - inode->i_sb->s_id, inode->i_ino); + if (inode) { + audit_log_format(ab, " dev="); + audit_log_untrustedstring(ab, inode->i_sb->s_id); + audit_log_format(ab, " ino=%lu", inode->i_ino); + } audit_log_format(ab, " res=%d", !result ? 0 : 1); audit_log_end(ab); } diff --git a/security/lsm_audit.c b/security/lsm_audit.c index 7bd6f138236b..fc41b7cccb53 100644 --- a/security/lsm_audit.c +++ b/security/lsm_audit.c @@ -235,10 +235,11 @@ static void dump_common_audit_data(struct audit_buffer *ab, audit_log_d_path(ab, "path=", &a->u.path); inode = a->u.path.dentry->d_inode; - if (inode) - audit_log_format(ab, " dev=%s ino=%lu", - inode->i_sb->s_id, - inode->i_ino); + if (inode) { + audit_log_format(ab, " dev="); + audit_log_untrustedstring(ab, inode->i_sb->s_id); + audit_log_format(ab, " ino=%lu", inode->i_ino); + } break; } case LSM_AUDIT_DATA_DENTRY: { @@ -248,10 +249,11 @@ static void dump_common_audit_data(struct audit_buffer *ab, audit_log_untrustedstring(ab, a->u.dentry->d_name.name); inode = a->u.dentry->d_inode; - if (inode) - audit_log_format(ab, " dev=%s ino=%lu", - inode->i_sb->s_id, - inode->i_ino); + if (inode) { + audit_log_format(ab, " dev="); + audit_log_untrustedstring(ab, inode->i_sb->s_id); + audit_log_format(ab, " ino=%lu", inode->i_ino); + } break; } case LSM_AUDIT_DATA_INODE: { @@ -266,8 +268,9 @@ static void dump_common_audit_data(struct audit_buffer *ab, dentry->d_name.name); dput(dentry); } - audit_log_format(ab, " dev=%s ino=%lu", inode->i_sb->s_id, - inode->i_ino); + audit_log_format(ab, " dev="); + audit_log_untrustedstring(ab, inode->i_sb->s_id); + audit_log_format(ab, " ino=%lu", inode->i_ino); break; } case LSM_AUDIT_DATA_TASK: From c158a35c8a681cf68d36f22f058f9f5466386c71 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 6 Jan 2012 14:07:10 -0800 Subject: [PATCH 554/820] audit: no leading space in audit_log_d_path prefix audit_log_d_path() injects an additional space before the prefix, which serves no purpose and doesn't mix well with other audit_log*() functions that do not sneak extra characters into the log. Signed-off-by: Kees Cook Signed-off-by: Eric Paris --- kernel/audit.c | 2 +- kernel/auditsc.c | 6 +++--- security/lsm_audit.c | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) diff --git a/kernel/audit.c b/kernel/audit.c index 00efe4758c86..705c25a70bff 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -1423,7 +1423,7 @@ void audit_log_d_path(struct audit_buffer *ab, const char *prefix, char *p, *pathname; if (prefix) - audit_log_format(ab, " %s", prefix); + audit_log_format(ab, "%s", prefix); /* We will allow 11 spaces for ' (deleted)' to be appended */ pathname = kmalloc(PATH_MAX+11, ab->gfp_mask); diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 86584ecb1039..caaea6e944f8 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1171,7 +1171,7 @@ static void audit_log_task_info(struct audit_buffer *ab, struct task_struct *tsk while (vma) { if ((vma->vm_flags & VM_EXECUTABLE) && vma->vm_file) { - audit_log_d_path(ab, "exe=", + audit_log_d_path(ab, " exe=", &vma->vm_file->f_path); break; } @@ -1540,7 +1540,7 @@ static void audit_log_name(struct audit_context *context, struct audit_names *n, case 0: /* name was specified as a relative path and the * directory component is the cwd */ - audit_log_d_path(ab, "name=", &context->pwd); + audit_log_d_path(ab, " name=", &context->pwd); break; default: /* log the name's directory component */ @@ -1725,7 +1725,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts if (context->pwd.dentry && context->pwd.mnt) { ab = audit_log_start(context, GFP_KERNEL, AUDIT_CWD); if (ab) { - audit_log_d_path(ab, "cwd=", &context->pwd); + audit_log_d_path(ab, " cwd=", &context->pwd); audit_log_end(ab); } } diff --git a/security/lsm_audit.c b/security/lsm_audit.c index fc41b7cccb53..293b8c45b1d1 100644 --- a/security/lsm_audit.c +++ b/security/lsm_audit.c @@ -232,7 +232,7 @@ static void dump_common_audit_data(struct audit_buffer *ab, case LSM_AUDIT_DATA_PATH: { struct inode *inode; - audit_log_d_path(ab, "path=", &a->u.path); + audit_log_d_path(ab, " path=", &a->u.path); inode = a->u.path.dentry->d_inode; if (inode) { @@ -318,7 +318,7 @@ static void dump_common_audit_data(struct audit_buffer *ab, .dentry = u->dentry, .mnt = u->mnt }; - audit_log_d_path(ab, "path=", &path); + audit_log_d_path(ab, " path=", &path); break; } if (!u->addr) From e1616300a20c80396109c1cf013ba9a36055a3da Mon Sep 17 00:00:00 2001 From: Kazuya Mio Date: Thu, 1 Dec 2011 16:51:07 +0900 Subject: [PATCH 555/820] wake up s_wait_unfrozen when ->freeze_fs fails dd slept infinitely when fsfeeze failed because of EIO. To fix this problem, if ->freeze_fs fails, freeze_super() wakes up the tasks waiting for the filesystem to become unfrozen. When s_frozen isn't SB_UNFROZEN in __generic_file_aio_write(), the function sleeps until FITHAW ioctl wakes up s_wait_unfrozen. However, if ->freeze_fs fails, s_frozen is set to SB_UNFROZEN and then freeze_super() returns an error number. In this case, FITHAW ioctl returns EINVAL because s_frozen is already SB_UNFROZEN. There is no way to wake up s_wait_unfrozen, so __generic_file_aio_write() sleeps infinitely. Signed-off-by: Kazuya Mio Signed-off-by: Al Viro --- fs/super.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/super.c b/fs/super.c index de41e1e46f09..6015c02296b7 100644 --- a/fs/super.c +++ b/fs/super.c @@ -1186,6 +1186,8 @@ int freeze_super(struct super_block *sb) printk(KERN_ERR "VFS:Filesystem freeze failed\n"); sb->s_frozen = SB_UNFROZEN; + smp_wmb(); + wake_up(&sb->s_wait_unfrozen); deactivate_locked_super(sb); return ret; } From 424a5334a5235c2fbb80090b18a065eeceb51d64 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Thu, 12 Jan 2012 12:41:36 +0100 Subject: [PATCH 556/820] vfs: remove printk from set_nlink() Don't log a message for set_nlink(0). Signed-off-by: Miklos Szeredi Signed-off-by: Al Viro --- fs/inode.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/fs/inode.c b/fs/inode.c index 4fa4f0916af9..fb10d86ffad7 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -322,9 +322,6 @@ EXPORT_SYMBOL(clear_nlink); void set_nlink(struct inode *inode, unsigned int nlink) { if (!nlink) { - printk_ratelimited(KERN_INFO - "set_nlink() clearing i_nlink on %s inode %li\n", - inode->i_sb->s_type->name, inode->i_ino); clear_nlink(inode); } else { /* Yes, some filesystems do change nlink from zero to one */ From 2727b1753934e154931d6b3bdf20c9b2398457a2 Mon Sep 17 00:00:00 2001 From: Alexander Aring Date: Thu, 8 Dec 2011 15:43:53 +0100 Subject: [PATCH 557/820] I2C: OMAP: correct SYSC register offset for OMAP4 Correct OMAP_I2C_SYSC_REG offset in omap4 register map. Offset 0x20 is reserved and OMAP_I2C_SYSC_REG has 0x10 as offset. Signed-off-by: Alexander Aring [khilman@ti.com: minor changelog edits] Cc: stable@vger.kernel.org Signed-off-by: Kevin Hilman --- drivers/i2c/busses/i2c-omap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c index fa23faa20f0e..257c1a52c8c4 100644 --- a/drivers/i2c/busses/i2c-omap.c +++ b/drivers/i2c/busses/i2c-omap.c @@ -235,7 +235,7 @@ static const u8 reg_map_ip_v2[] = { [OMAP_I2C_BUF_REG] = 0x94, [OMAP_I2C_CNT_REG] = 0x98, [OMAP_I2C_DATA_REG] = 0x9c, - [OMAP_I2C_SYSC_REG] = 0x20, + [OMAP_I2C_SYSC_REG] = 0x10, [OMAP_I2C_CON_REG] = 0xa4, [OMAP_I2C_OA_REG] = 0xa8, [OMAP_I2C_SA_REG] = 0xac, From 78e1cf42ee3efd2c3e8af0b97cf0a0a570cc0c1f Mon Sep 17 00:00:00 2001 From: Jan Weitzel Date: Wed, 7 Dec 2011 11:50:16 -0800 Subject: [PATCH 558/820] I2C: OMAP: NACK without STP On OMAP4 OMAP_I2C_STAT_NACK is causing a timeout on the next access. The isr cleans all flags in OMAP_I2C_CON_REG by setting OMAP_I2C_CON_STP OMAP_I2C_CON_STP is also set in omap_i2c_xfer_msg on the last message. According to the TI TSR the sequence for OMAP_I2C_STAT_NACK and OMAP_I2C_STAT_AL are nearly the same. Removing the OMAP_I2C_CON_STP part in the isr fix the problem. Tested on OMAP4430 and OMAP3530 (here NACK was not a problem) Fixes also booting on 2430sdp. Signed-off-by: Jan Weitzel Acked-by: Tony Lindgren Signed-off-by: Kevin Hilman --- drivers/i2c/busses/i2c-omap.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c index 257c1a52c8c4..e0733b776079 100644 --- a/drivers/i2c/busses/i2c-omap.c +++ b/drivers/i2c/busses/i2c-omap.c @@ -830,11 +830,9 @@ complete: ~(OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR | OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR)); - if (stat & OMAP_I2C_STAT_NACK) { + if (stat & OMAP_I2C_STAT_NACK) err |= OMAP_I2C_STAT_NACK; - omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, - OMAP_I2C_CON_STP); - } + if (stat & OMAP_I2C_STAT_AL) { dev_err(dev->dev, "Arbitration lost\n"); err |= OMAP_I2C_STAT_AL; From 6145197be6cc0583fa1a2f4ec1079d366137061e Mon Sep 17 00:00:00 2001 From: Benoit Cousson Date: Thu, 22 Dec 2011 15:56:36 +0100 Subject: [PATCH 559/820] i2c: OMAP: Add DT support for i2c controller Add initial DT support to retrieve the frequency using a DT attribute instead of the pdata pointer if of_node exist. Add documentation for omap i2c controller binding. Based on original patches from Manju and Grant. Signed-off-by: Benoit Cousson Cc: Ben Dooks Reviewed-by: Rob Herring Acked-by: Grant Likely Signed-off-by: Kevin Hilman --- .../devicetree/bindings/i2c/omap-i2c.txt | 30 ++++++ drivers/i2c/busses/i2c-omap.c | 102 ++++++++++++------ 2 files changed, 97 insertions(+), 35 deletions(-) create mode 100644 Documentation/devicetree/bindings/i2c/omap-i2c.txt diff --git a/Documentation/devicetree/bindings/i2c/omap-i2c.txt b/Documentation/devicetree/bindings/i2c/omap-i2c.txt new file mode 100644 index 000000000000..56564aa4b444 --- /dev/null +++ b/Documentation/devicetree/bindings/i2c/omap-i2c.txt @@ -0,0 +1,30 @@ +I2C for OMAP platforms + +Required properties : +- compatible : Must be "ti,omap3-i2c" or "ti,omap4-i2c" +- ti,hwmods : Must be "i2c", n being the instance number (1-based) +- #address-cells = <1>; +- #size-cells = <0>; + +Recommended properties : +- clock-frequency : Desired I2C bus clock frequency in Hz. Otherwise + the default 100 kHz frequency will be used. + +Optional properties: +- Child nodes conforming to i2c bus binding + +Note: Current implementation will fetch base address, irq and dma +from omap hwmod data base during device registration. +Future plan is to migrate hwmod data base contents into device tree +blob so that, all the required data will be used from device tree dts +file. + +Examples : + +i2c1: i2c@0 { + compatible = "ti,omap3-i2c"; + #address-cells = <1>; + #size-cells = <0>; + ti,hwmods = "i2c1"; + clock-frequency = <400000>; +}; diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c index e0733b776079..f713eac55047 100644 --- a/drivers/i2c/busses/i2c-omap.c +++ b/drivers/i2c/busses/i2c-omap.c @@ -37,6 +37,9 @@ #include #include #include +#include +#include +#include #include #include #include @@ -182,7 +185,9 @@ struct omap_i2c_dev { u32 latency; /* maximum mpu wkup latency */ void (*set_mpu_wkup_lat)(struct device *dev, long latency); - u32 speed; /* Speed of bus in Khz */ + u32 speed; /* Speed of bus in kHz */ + u32 dtrev; /* extra revision from DT */ + u32 flags; u16 cmd_err; u8 *buf; u8 *regs; @@ -266,11 +271,7 @@ static inline u16 omap_i2c_read_reg(struct omap_i2c_dev *i2c_dev, int reg) static void omap_i2c_unidle(struct omap_i2c_dev *dev) { - struct omap_i2c_bus_platform_data *pdata; - - pdata = dev->dev->platform_data; - - if (pdata->flags & OMAP_I2C_FLAG_RESET_REGS_POSTIDLE) { + if (dev->flags & OMAP_I2C_FLAG_RESET_REGS_POSTIDLE) { omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, 0); omap_i2c_write_reg(dev, OMAP_I2C_PSC_REG, dev->pscstate); omap_i2c_write_reg(dev, OMAP_I2C_SCLL_REG, dev->scllstate); @@ -291,13 +292,10 @@ static void omap_i2c_unidle(struct omap_i2c_dev *dev) static void omap_i2c_idle(struct omap_i2c_dev *dev) { - struct omap_i2c_bus_platform_data *pdata; u16 iv; - pdata = dev->dev->platform_data; - dev->iestate = omap_i2c_read_reg(dev, OMAP_I2C_IE_REG); - if (pdata->rev == OMAP_I2C_IP_VERSION_2) + if (dev->dtrev == OMAP_I2C_IP_VERSION_2) omap_i2c_write_reg(dev, OMAP_I2C_IP_V2_IRQENABLE_CLR, 1); else omap_i2c_write_reg(dev, OMAP_I2C_IE_REG, 0); @@ -320,9 +318,6 @@ static int omap_i2c_init(struct omap_i2c_dev *dev) unsigned long timeout; unsigned long internal_clk = 0; struct clk *fclk; - struct omap_i2c_bus_platform_data *pdata; - - pdata = dev->dev->platform_data; if (dev->rev >= OMAP_I2C_OMAP1_REV_2) { /* Disable I2C controller before soft reset */ @@ -373,7 +368,7 @@ static int omap_i2c_init(struct omap_i2c_dev *dev) } omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, 0); - if (pdata->flags & OMAP_I2C_FLAG_ALWAYS_ARMXOR_CLK) { + if (dev->flags & OMAP_I2C_FLAG_ALWAYS_ARMXOR_CLK) { /* * The I2C functional clock is the armxor_ck, so there's * no need to get "armxor_ck" separately. Now, if OMAP2420 @@ -397,7 +392,7 @@ static int omap_i2c_init(struct omap_i2c_dev *dev) psc = fclk_rate / 12000000; } - if (!(pdata->flags & OMAP_I2C_FLAG_SIMPLE_CLOCK)) { + if (!(dev->flags & OMAP_I2C_FLAG_SIMPLE_CLOCK)) { /* * HSI2C controller internal clk rate should be 19.2 Mhz for @@ -406,7 +401,7 @@ static int omap_i2c_init(struct omap_i2c_dev *dev) * The filter is iclk (fclk for HS) period. */ if (dev->speed > 400 || - pdata->flags & OMAP_I2C_FLAG_FORCE_19200_INT_CLK) + dev->flags & OMAP_I2C_FLAG_FORCE_19200_INT_CLK) internal_clk = 19200; else if (dev->speed > 100) internal_clk = 9600; @@ -475,7 +470,7 @@ static int omap_i2c_init(struct omap_i2c_dev *dev) dev->errata = 0; - if (pdata->flags & OMAP_I2C_FLAG_APPLY_ERRATA_I207) + if (dev->flags & OMAP_I2C_FLAG_APPLY_ERRATA_I207) dev->errata |= I2C_OMAP_ERRATA_I207; /* Enable interrupts */ @@ -484,7 +479,7 @@ static int omap_i2c_init(struct omap_i2c_dev *dev) OMAP_I2C_IE_AL) | ((dev->fifo_size) ? (OMAP_I2C_IE_RDR | OMAP_I2C_IE_XDR) : 0); omap_i2c_write_reg(dev, OMAP_I2C_IE_REG, dev->iestate); - if (pdata->flags & OMAP_I2C_FLAG_RESET_REGS_POSTIDLE) { + if (dev->flags & OMAP_I2C_FLAG_RESET_REGS_POSTIDLE) { dev->pscstate = psc; dev->scllstate = scll; dev->sclhstate = sclh; @@ -804,9 +799,6 @@ omap_i2c_isr(int this_irq, void *dev_id) u16 bits; u16 stat, w; int err, count = 0; - struct omap_i2c_bus_platform_data *pdata; - - pdata = dev->dev->platform_data; if (pm_runtime_suspended(dev->dev)) return IRQ_NONE; @@ -873,7 +865,7 @@ complete: * Data reg in 2430, omap3 and * omap4 is 8 bit wide */ - if (pdata->flags & + if (dev->flags & OMAP_I2C_FLAG_16BIT_DATA_REG) { if (dev->buf_len) { *dev->buf++ = w >> 8; @@ -916,7 +908,7 @@ complete: * Data reg in 2430, omap3 and * omap4 is 8 bit wide */ - if (pdata->flags & + if (dev->flags & OMAP_I2C_FLAG_16BIT_DATA_REG) { if (dev->buf_len) { w |= *dev->buf++ << 8; @@ -963,6 +955,32 @@ static const struct i2c_algorithm omap_i2c_algo = { .functionality = omap_i2c_func, }; +#ifdef CONFIG_OF +static struct omap_i2c_bus_platform_data omap3_pdata = { + .rev = OMAP_I2C_IP_VERSION_1, + .flags = OMAP_I2C_FLAG_APPLY_ERRATA_I207 | + OMAP_I2C_FLAG_RESET_REGS_POSTIDLE | + OMAP_I2C_FLAG_BUS_SHIFT_2, +}; + +static struct omap_i2c_bus_platform_data omap4_pdata = { + .rev = OMAP_I2C_IP_VERSION_2, +}; + +static const struct of_device_id omap_i2c_of_match[] = { + { + .compatible = "ti,omap4-i2c", + .data = &omap4_pdata, + }, + { + .compatible = "ti,omap3-i2c", + .data = &omap3_pdata, + }, + { }, +}; +MODULE_DEVICE_TABLE(of, omap_i2c_of_match); +#endif + static int __devinit omap_i2c_probe(struct platform_device *pdev) { @@ -970,9 +988,10 @@ omap_i2c_probe(struct platform_device *pdev) struct i2c_adapter *adap; struct resource *mem, *irq, *ioarea; struct omap_i2c_bus_platform_data *pdata = pdev->dev.platform_data; + struct device_node *node = pdev->dev.of_node; + const struct of_device_id *match; irq_handler_t isr; int r; - u32 speed = 0; /* NOTE: driver uses the static register mapping */ mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); @@ -999,15 +1018,24 @@ omap_i2c_probe(struct platform_device *pdev) goto err_release_region; } - if (pdata != NULL) { - speed = pdata->clkrate; + match = of_match_device(omap_i2c_of_match, &pdev->dev); + if (match) { + u32 freq = 100000; /* default to 100000 Hz */ + + pdata = match->data; + dev->dtrev = pdata->rev; + dev->flags = pdata->flags; + + of_property_read_u32(node, "clock-frequency", &freq); + /* convert DT freq value in Hz into kHz for speed */ + dev->speed = freq / 1000; + } else if (pdata != NULL) { + dev->speed = pdata->clkrate; + dev->flags = pdata->flags; dev->set_mpu_wkup_lat = pdata->set_mpu_wkup_lat; - } else { - speed = 100; /* Default speed */ - dev->set_mpu_wkup_lat = NULL; + dev->dtrev = pdata->rev; } - dev->speed = speed; dev->dev = &pdev->dev; dev->irq = irq->start; dev->base = ioremap(mem->start, resource_size(mem)); @@ -1018,9 +1046,9 @@ omap_i2c_probe(struct platform_device *pdev) platform_set_drvdata(pdev, dev); - dev->reg_shift = (pdata->flags >> OMAP_I2C_FLAG_BUS_SHIFT__SHIFT) & 3; + dev->reg_shift = (dev->flags >> OMAP_I2C_FLAG_BUS_SHIFT__SHIFT) & 3; - if (pdata->rev == OMAP_I2C_IP_VERSION_2) + if (dev->dtrev == OMAP_I2C_IP_VERSION_2) dev->regs = (u8 *)reg_map_ip_v2; else dev->regs = (u8 *)reg_map_ip_v1; @@ -1033,7 +1061,7 @@ omap_i2c_probe(struct platform_device *pdev) if (dev->rev <= OMAP_I2C_REV_ON_3430) dev->errata |= I2C_OMAP3_1P153; - if (!(pdata->flags & OMAP_I2C_FLAG_NO_FIFO)) { + if (!(dev->flags & OMAP_I2C_FLAG_NO_FIFO)) { u16 s; /* Set up the fifo size - Get total size */ @@ -1056,7 +1084,7 @@ omap_i2c_probe(struct platform_device *pdev) /* calculate wakeup latency constraint for MPU */ if (dev->set_mpu_wkup_lat != NULL) dev->latency = (1000000 * dev->fifo_size) / - (1000 * speed / 8); + (1000 * dev->speed / 8); } /* reset ASAP, clearing any IRQs */ @@ -1072,7 +1100,7 @@ omap_i2c_probe(struct platform_device *pdev) } dev_info(dev->dev, "bus %d rev%d.%d.%d at %d kHz\n", pdev->id, - pdata->rev, dev->rev >> 4, dev->rev & 0xf, dev->speed); + dev->dtrev, dev->rev >> 4, dev->rev & 0xf, dev->speed); pm_runtime_put(dev->dev); @@ -1083,6 +1111,7 @@ omap_i2c_probe(struct platform_device *pdev) strlcpy(adap->name, "OMAP I2C adapter", sizeof(adap->name)); adap->algo = &omap_i2c_algo; adap->dev.parent = &pdev->dev; + adap->dev.of_node = pdev->dev.of_node; /* i2c device drivers may be active on return from add_adapter() */ adap->nr = pdev->id; @@ -1092,6 +1121,8 @@ omap_i2c_probe(struct platform_device *pdev) goto err_free_irq; } + of_i2c_register_devices(adap); + return 0; err_free_irq: @@ -1164,6 +1195,7 @@ static struct platform_driver omap_i2c_driver = { .name = "omap_i2c", .owner = THIS_MODULE, .pm = OMAP_I2C_PM_OPS, + .of_match_table = of_match_ptr(omap_i2c_of_match), }, }; From e268337dfe26dfc7efd422a804dbb27977a3cccc Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 17 Jan 2012 15:21:19 -0800 Subject: [PATCH 560/820] proc: clean up and fix /proc//mem handling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Jüri Aedla reported that the /proc//mem handling really isn't very robust, and it also doesn't match the permission checking of any of the other related files. This changes it to do the permission checks at open time, and instead of tracking the process, it tracks the VM at the time of the open. That simplifies the code a lot, but does mean that if you hold the file descriptor open over an execve(), you'll continue to read from the _old_ VM. That is different from our previous behavior, but much simpler. If somebody actually finds a load where this matters, we'll need to revert this commit. I suspect that nobody will ever notice - because the process mapping addresses will also have changed as part of the execve. So you cannot actually usefully access the fd across a VM change simply because all the offsets for IO would have changed too. Reported-by: Jüri Aedla Cc: Al Viro Signed-off-by: Linus Torvalds --- fs/proc/base.c | 145 +++++++++++++------------------------------------ 1 file changed, 39 insertions(+), 106 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index 5485a5388ecb..662ddf2ec4f1 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -198,65 +198,7 @@ static int proc_root_link(struct dentry *dentry, struct path *path) return result; } -static struct mm_struct *__check_mem_permission(struct task_struct *task) -{ - struct mm_struct *mm; - - mm = get_task_mm(task); - if (!mm) - return ERR_PTR(-EINVAL); - - /* - * A task can always look at itself, in case it chooses - * to use system calls instead of load instructions. - */ - if (task == current) - return mm; - - /* - * If current is actively ptrace'ing, and would also be - * permitted to freshly attach with ptrace now, permit it. - */ - if (task_is_stopped_or_traced(task)) { - int match; - rcu_read_lock(); - match = (ptrace_parent(task) == current); - rcu_read_unlock(); - if (match && ptrace_may_access(task, PTRACE_MODE_ATTACH)) - return mm; - } - - /* - * No one else is allowed. - */ - mmput(mm); - return ERR_PTR(-EPERM); -} - -/* - * If current may access user memory in @task return a reference to the - * corresponding mm, otherwise ERR_PTR. - */ -static struct mm_struct *check_mem_permission(struct task_struct *task) -{ - struct mm_struct *mm; - int err; - - /* - * Avoid racing if task exec's as we might get a new mm but validate - * against old credentials. - */ - err = mutex_lock_killable(&task->signal->cred_guard_mutex); - if (err) - return ERR_PTR(err); - - mm = __check_mem_permission(task); - mutex_unlock(&task->signal->cred_guard_mutex); - - return mm; -} - -struct mm_struct *mm_for_maps(struct task_struct *task) +static struct mm_struct *mm_access(struct task_struct *task, unsigned int mode) { struct mm_struct *mm; int err; @@ -267,7 +209,7 @@ struct mm_struct *mm_for_maps(struct task_struct *task) mm = get_task_mm(task); if (mm && mm != current->mm && - !ptrace_may_access(task, PTRACE_MODE_READ)) { + !ptrace_may_access(task, mode)) { mmput(mm); mm = ERR_PTR(-EACCES); } @@ -276,6 +218,11 @@ struct mm_struct *mm_for_maps(struct task_struct *task) return mm; } +struct mm_struct *mm_for_maps(struct task_struct *task) +{ + return mm_access(task, PTRACE_MODE_READ); +} + static int proc_pid_cmdline(struct task_struct *task, char * buffer) { int res = 0; @@ -752,38 +699,39 @@ static const struct file_operations proc_single_file_operations = { static int mem_open(struct inode* inode, struct file* file) { - file->private_data = (void*)((long)current->self_exec_id); + struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); + struct mm_struct *mm; + + if (!task) + return -ESRCH; + + mm = mm_access(task, PTRACE_MODE_ATTACH); + put_task_struct(task); + + if (IS_ERR(mm)) + return PTR_ERR(mm); + /* OK to pass negative loff_t, we can catch out-of-range */ file->f_mode |= FMODE_UNSIGNED_OFFSET; + file->private_data = mm; + return 0; } static ssize_t mem_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); + int ret; char *page; unsigned long src = *ppos; - int ret = -ESRCH; - struct mm_struct *mm; + struct mm_struct *mm = file->private_data; - if (!task) - goto out_no_task; + if (!mm) + return 0; - ret = -ENOMEM; page = (char *)__get_free_page(GFP_TEMPORARY); if (!page) - goto out; - - mm = check_mem_permission(task); - ret = PTR_ERR(mm); - if (IS_ERR(mm)) - goto out_free; - - ret = -EIO; - - if (file->private_data != (void*)((long)current->self_exec_id)) - goto out_put; + return -ENOMEM; ret = 0; @@ -810,13 +758,7 @@ static ssize_t mem_read(struct file * file, char __user * buf, } *ppos = src; -out_put: - mmput(mm); -out_free: free_page((unsigned long) page); -out: - put_task_struct(task); -out_no_task: return ret; } @@ -825,27 +767,15 @@ static ssize_t mem_write(struct file * file, const char __user *buf, { int copied; char *page; - struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); unsigned long dst = *ppos; - struct mm_struct *mm; + struct mm_struct *mm = file->private_data; - copied = -ESRCH; - if (!task) - goto out_no_task; + if (!mm) + return 0; - copied = -ENOMEM; page = (char *)__get_free_page(GFP_TEMPORARY); if (!page) - goto out_task; - - mm = check_mem_permission(task); - copied = PTR_ERR(mm); - if (IS_ERR(mm)) - goto out_free; - - copied = -EIO; - if (file->private_data != (void *)((long)current->self_exec_id)) - goto out_mm; + return -ENOMEM; copied = 0; while (count > 0) { @@ -869,13 +799,7 @@ static ssize_t mem_write(struct file * file, const char __user *buf, } *ppos = dst; -out_mm: - mmput(mm); -out_free: free_page((unsigned long) page); -out_task: - put_task_struct(task); -out_no_task: return copied; } @@ -895,11 +819,20 @@ loff_t mem_lseek(struct file *file, loff_t offset, int orig) return file->f_pos; } +static int mem_release(struct inode *inode, struct file *file) +{ + struct mm_struct *mm = file->private_data; + + mmput(mm); + return 0; +} + static const struct file_operations proc_mem_operations = { .llseek = mem_lseek, .read = mem_read, .write = mem_write, .open = mem_open, + .release = mem_release, }; static ssize_t environ_read(struct file *file, char __user *buf, From 07e8a51ff68353e01d795cceafbac9f54c49132b Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Thu, 12 Jan 2012 15:38:02 +0800 Subject: [PATCH 561/820] i2c-eg20t: use i2c_add_numbered_adapter to get a fixed bus number For EG20T and ML7213 IOH, the i2c controller numbers are fixed, using fixed bus number will make it much easier for platform code to use i2c_register_board_info() to register i2c devices. Signed-off-by: Feng Tang Signed-off-by: Ben Dooks --- drivers/i2c/busses/i2c-eg20t.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/i2c/busses/i2c-eg20t.c b/drivers/i2c/busses/i2c-eg20t.c index 3ef3557b6e32..8dc3f75d2faf 100644 --- a/drivers/i2c/busses/i2c-eg20t.c +++ b/drivers/i2c/busses/i2c-eg20t.c @@ -918,7 +918,9 @@ static int __devinit pch_i2c_probe(struct pci_dev *pdev, pch_adap->dev.parent = &pdev->dev; pch_i2c_init(&adap_info->pch_data[i]); - ret = i2c_add_adapter(pch_adap); + + pch_adap->nr = i; + ret = i2c_add_numbered_adapter(pch_adap); if (ret) { pch_pci_err(pdev, "i2c_add_adapter[ch:%d] FAILED\n", i); goto err_add_adapter; From ff35e8b18984ad2a82cbd259fc07f0be4b34b1aa Mon Sep 17 00:00:00 2001 From: Toshiharu Okada Date: Mon, 26 Sep 2011 16:16:23 +0900 Subject: [PATCH 562/820] i2c-eg20t: modified the setting of transfer rate. This patch modified the setting value of I2C Bus Transfer Rate Setting Counter regisrer. Signed-off-by: Toshiharu Okada Signed-off-by: Ben Dooks --- drivers/i2c/busses/i2c-eg20t.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/i2c/busses/i2c-eg20t.c b/drivers/i2c/busses/i2c-eg20t.c index 8dc3f75d2faf..357a66c66064 100644 --- a/drivers/i2c/busses/i2c-eg20t.c +++ b/drivers/i2c/busses/i2c-eg20t.c @@ -243,7 +243,7 @@ static void pch_i2c_init(struct i2c_algo_pch_data *adap) if (pch_clk > PCH_MAX_CLK) pch_clk = 62500; - pch_i2cbc = (pch_clk + (pch_i2c_speed * 4)) / pch_i2c_speed * 8; + pch_i2cbc = (pch_clk + (pch_i2c_speed * 4)) / (pch_i2c_speed * 8); /* Set transfer speed in I2CBC */ iowrite32(pch_i2cbc, p + PCH_I2CBC); From c3f4661f55f86b541d23c66dbb53ce7b0a34e777 Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Fri, 28 Oct 2011 09:40:10 +0900 Subject: [PATCH 563/820] i2c-eg20t: Support new device LAPIS Semiconductor ML7831 IOH ML7831 is companion chip for Intel Atom E6xx series. Signed-off-by: Tomoya MORINAGA Signed-off-by: Ben Dooks --- drivers/i2c/busses/Kconfig | 14 +++++++------- drivers/i2c/busses/i2c-eg20t.c | 3 +++ 2 files changed, 10 insertions(+), 7 deletions(-) diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig index cbe7a2fb779f..3101dd59e379 100644 --- a/drivers/i2c/busses/Kconfig +++ b/drivers/i2c/busses/Kconfig @@ -682,19 +682,19 @@ config I2C_XILINX will be called xilinx_i2c. config I2C_EG20T - tristate "Intel EG20T PCH / OKI SEMICONDUCTOR IOH(ML7213/ML7223)" + tristate "Intel EG20T PCH/LAPIS Semicon IOH(ML7213/ML7223/ML7831) I2C" depends on PCI help This driver is for PCH(Platform controller Hub) I2C of EG20T which is an IOH(Input/Output Hub) for x86 embedded processor. This driver can access PCH I2C bus device. - This driver also can be used for OKI SEMICONDUCTOR IOH(Input/ - Output Hub), ML7213 and ML7223. - ML7213 IOH is for IVI(In-Vehicle Infotainment) use and ML7223 IOH is - for MP(Media Phone) use. - ML7213/ML7223 is companion chip for Intel Atom E6xx series. - ML7213/ML7223 is completely compatible for Intel EG20T PCH. + This driver also can be used for LAPIS Semiconductor IOH(Input/ + Output Hub), ML7213, ML7223 and ML7831. + ML7213 IOH is for IVI(In-Vehicle Infotainment) use, ML7223 IOH is + for MP(Media Phone) use and ML7831 IOH is for general purpose use. + ML7213/ML7223/ML7831 is companion chip for Intel Atom E6xx series. + ML7213/ML7223/ML7831 is completely compatible for Intel EG20T PCH. comment "External I2C/SMBus adapter drivers" diff --git a/drivers/i2c/busses/i2c-eg20t.c b/drivers/i2c/busses/i2c-eg20t.c index 357a66c66064..c72d7f998487 100644 --- a/drivers/i2c/busses/i2c-eg20t.c +++ b/drivers/i2c/busses/i2c-eg20t.c @@ -137,6 +137,7 @@ Set the number of I2C instance max Intel EG20T PCH : 1ch OKI SEMICONDUCTOR ML7213 IOH : 2ch +OKI SEMICONDUCTOR ML7831 IOH : 1ch */ #define PCH_I2C_MAX_DEV 2 @@ -184,11 +185,13 @@ static DEFINE_MUTEX(pch_mutex); #define PCI_VENDOR_ID_ROHM 0x10DB #define PCI_DEVICE_ID_ML7213_I2C 0x802D #define PCI_DEVICE_ID_ML7223_I2C 0x8010 +#define PCI_DEVICE_ID_ML7831_I2C 0x8817 static DEFINE_PCI_DEVICE_TABLE(pch_pcidev_id) = { { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_PCH_I2C), 1, }, { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_I2C), 2, }, { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_I2C), 1, }, + { PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_I2C), 1, }, {0,} }; From 8956dc102ca26357850830f1d26132719c1ce6ee Mon Sep 17 00:00:00 2001 From: Tomoya MORINAGA Date: Fri, 28 Oct 2011 09:40:11 +0900 Subject: [PATCH 564/820] i2c-eg20t: Change-company-name-OKI-SEMICONDUCTOR to LAPIS Semiconductor On October 1 in 2011, OKI SEMICONDUCTOR Co., Ltd. changed the company name in to LAPIS Semiconductor Co., Ltd. Signed-off-by: Tomoya MORINAGA Signed-off-by: Ben Dooks --- drivers/i2c/busses/i2c-eg20t.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/i2c/busses/i2c-eg20t.c b/drivers/i2c/busses/i2c-eg20t.c index c72d7f998487..ca8877641040 100644 --- a/drivers/i2c/busses/i2c-eg20t.c +++ b/drivers/i2c/busses/i2c-eg20t.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2010 OKI SEMICONDUCTOR CO., LTD. + * Copyright (C) 2011 LAPIS Semiconductor Co., Ltd. * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -136,8 +136,8 @@ /* Set the number of I2C instance max Intel EG20T PCH : 1ch -OKI SEMICONDUCTOR ML7213 IOH : 2ch -OKI SEMICONDUCTOR ML7831 IOH : 1ch +LAPIS Semiconductor ML7213 IOH : 2ch +LAPIS Semiconductor ML7831 IOH : 1ch */ #define PCH_I2C_MAX_DEV 2 @@ -181,7 +181,7 @@ static int pch_clk = 50000; /* specifies I2C clock speed in KHz */ static wait_queue_head_t pch_event; static DEFINE_MUTEX(pch_mutex); -/* Definition for ML7213 by OKI SEMICONDUCTOR */ +/* Definition for ML7213 by LAPIS Semiconductor */ #define PCI_VENDOR_ID_ROHM 0x10DB #define PCI_DEVICE_ID_ML7213_I2C 0x802D #define PCI_DEVICE_ID_ML7223_I2C 0x8010 @@ -1063,8 +1063,8 @@ static void __exit pch_pci_exit(void) } module_exit(pch_pci_exit); -MODULE_DESCRIPTION("Intel EG20T PCH/OKI SEMICONDUCTOR ML7213 IOH I2C Driver"); +MODULE_DESCRIPTION("Intel EG20T PCH/LAPIS Semico ML7213/ML7223/ML7831 IOH I2C"); MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Tomoya MORINAGA. "); +MODULE_AUTHOR("Tomoya MORINAGA. "); module_param(pch_i2c_speed, int, (S_IRUSR | S_IWUSR)); module_param(pch_clk, int, (S_IRUSR | S_IWUSR)); From 89879a7eb81f69e6f63bdb2a442fb765c46482c0 Mon Sep 17 00:00:00 2001 From: James Morris Date: Wed, 18 Jan 2012 10:40:44 +1100 Subject: [PATCH 565/820] security: update MAINTAINERS file with new git repo Update MAINTAINERS file with new git repo: git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git Signed-off-by: James Morris --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 2a90101309d1..ece8935025e3 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5846,7 +5846,7 @@ F: drivers/mmc/host/sdhci-spear.c SECURITY SUBSYSTEM M: James Morris L: linux-security-module@vger.kernel.org (suggested Cc:) -T: git git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6.git +T: git git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security.git W: http://security.wiki.kernel.org/ S: Supported F: security/ From 67175b855bfd6ed95ffeff95532173c07de6432d Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Tue, 17 Jan 2012 21:14:05 +0000 Subject: [PATCH 566/820] Fix compile breakage with kref.h This set of build failures just started appearing on parisc: In file included from drivers/input/serio/serio_raw.c:12: include/linux/kref.h: In function 'kref_get': include/linux/kref.h:40: error: 'TAINT_WARN' undeclared (first use in this function) include/linux/kref.h:40: error: (Each undeclared identifier is reported only once include/linux/kref.h:40: error: for each function it appears in.) include/linux/kref.h: In function 'kref_sub': include/linux/kref.h:65: error: 'TAINT_WARN' undeclared (first use in this function) It happens because TAINT_WARN is defined in kernel.h and this particular compile doesn't seem to include it (no idea why it's just manifesting .. probably some #include file untangling exposed it). Fix by adding #include to linux/kref.h Signed-off-by: James Bottomley Acked-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- include/linux/kref.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/linux/kref.h b/include/linux/kref.h index abc0120b09b7..9c07dcebded7 100644 --- a/include/linux/kref.h +++ b/include/linux/kref.h @@ -17,6 +17,7 @@ #include #include +#include struct kref { atomic_t refcount; From 25add8cf99c9ec8b8dc0acd8b9241e963fc0d29c Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Sun, 15 Jan 2012 11:05:59 +0900 Subject: [PATCH 567/820] TOMOYO: Accept \000 as a valid character. TOMOYO 2.5 in Linux 3.2 and later handles Unix domain socket's address. Thus, tomoyo_correct_word2() needs to accept \000 as a valid character, or TOMOYO 2.5 cannot handle Unix domain's abstract socket address. Reported-by: Steven Allen Signed-off-by: Tetsuo Handa CC: stable@vger.kernel.org [3.2+] Signed-off-by: James Morris --- security/tomoyo/util.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/security/tomoyo/util.c b/security/tomoyo/util.c index 4a9b4b2eb755..867558c98334 100644 --- a/security/tomoyo/util.c +++ b/security/tomoyo/util.c @@ -492,13 +492,13 @@ static bool tomoyo_correct_word2(const char *string, size_t len) if (d < '0' || d > '7' || e < '0' || e > '7') break; c = tomoyo_make_byte(c, d, e); - if (tomoyo_invalid(c)) - continue; /* pattern is not \000 */ + if (c <= ' ' || c >= 127) + continue; } goto out; } else if (in_repetition && c == '/') { goto out; - } else if (tomoyo_invalid(c)) { + } else if (c <= ' ' || c >= 127) { goto out; } } From efde8b6e16f11e7d1681c68d86c7fd51053cada7 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 17 Jan 2012 20:39:40 +0000 Subject: [PATCH 568/820] KEYS: Add missing smp_rmb() primitives to the keyring search code Add missing smp_rmb() primitives to the keyring search code. When keyring payloads are appended to without replacement (thus using up spare slots in the key pointer array), an smp_wmb() is issued between the pointer assignment and the increment of the key count (nkeys). There should be corresponding read barriers between the read of nkeys and dereferences of keys[n] when n is dependent on the value of nkeys. Signed-off-by: David Howells Reviewed-by: Paul E. McKenney Signed-off-by: James Morris --- security/keys/gc.c | 4 +++- security/keys/keyring.c | 22 +++++++++++++++------- 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/security/keys/gc.c b/security/keys/gc.c index bf4d8da5a795..a42b45531aac 100644 --- a/security/keys/gc.c +++ b/security/keys/gc.c @@ -145,7 +145,9 @@ static void key_gc_keyring(struct key *keyring, time_t limit) if (!klist) goto unlock_dont_gc; - for (loop = klist->nkeys - 1; loop >= 0; loop--) { + loop = klist->nkeys; + smp_rmb(); + for (loop--; loop >= 0; loop--) { key = klist->keys[loop]; if (test_bit(KEY_FLAG_DEAD, &key->flags) || (key->expiry > 0 && key->expiry <= limit)) diff --git a/security/keys/keyring.c b/security/keys/keyring.c index 37a7f3b28852..d605f75292e4 100644 --- a/security/keys/keyring.c +++ b/security/keys/keyring.c @@ -319,7 +319,7 @@ key_ref_t keyring_search_aux(key_ref_t keyring_ref, struct key *keyring, *key; key_ref_t key_ref; long err; - int sp, kix; + int sp, nkeys, kix; keyring = key_ref_to_ptr(keyring_ref); possessed = is_key_possessed(keyring_ref); @@ -380,7 +380,9 @@ descend: goto not_this_keyring; /* iterate through the keys in this keyring first */ - for (kix = 0; kix < keylist->nkeys; kix++) { + nkeys = keylist->nkeys; + smp_rmb(); + for (kix = 0; kix < nkeys; kix++) { key = keylist->keys[kix]; kflags = key->flags; @@ -421,7 +423,9 @@ descend: /* search through the keyrings nested in this one */ kix = 0; ascend: - for (; kix < keylist->nkeys; kix++) { + nkeys = keylist->nkeys; + smp_rmb(); + for (; kix < nkeys; kix++) { key = keylist->keys[kix]; if (key->type != &key_type_keyring) continue; @@ -515,7 +519,7 @@ key_ref_t __keyring_search_one(key_ref_t keyring_ref, struct keyring_list *klist; unsigned long possessed; struct key *keyring, *key; - int loop; + int nkeys, loop; keyring = key_ref_to_ptr(keyring_ref); possessed = is_key_possessed(keyring_ref); @@ -524,7 +528,9 @@ key_ref_t __keyring_search_one(key_ref_t keyring_ref, klist = rcu_dereference(keyring->payload.subscriptions); if (klist) { - for (loop = 0; loop < klist->nkeys; loop++) { + nkeys = klist->nkeys; + smp_rmb(); + for (loop = 0; loop < nkeys ; loop++) { key = klist->keys[loop]; if (key->type == ktype && @@ -622,7 +628,7 @@ static int keyring_detect_cycle(struct key *A, struct key *B) struct keyring_list *keylist; struct key *subtree, *key; - int sp, kix, ret; + int sp, nkeys, kix, ret; rcu_read_lock(); @@ -645,7 +651,9 @@ descend: ascend: /* iterate through the remaining keys in this keyring */ - for (; kix < keylist->nkeys; kix++) { + nkeys = keylist->nkeys; + smp_rmb(); + for (; kix < nkeys; kix++) { key = keylist->keys[kix]; if (key == A) From ee0b31a25a010116f44fca6c96f4516d417793dd Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Tue, 17 Jan 2012 20:39:51 +0000 Subject: [PATCH 569/820] keys: fix trusted/encrypted keys sparse rcu_assign_pointer messages Define rcu_assign_keypointer(), which uses the key payload.rcudata instead of payload.data, to resolve the CONFIG_SPARSE_RCU_POINTER message: "incompatible types in comparison expression (different address spaces)" Replace the rcu_assign_pointer() calls in encrypted/trusted keys with rcu_assign_keypointer(). Signed-off-by: Mimi Zohar Signed-off-by: David Howells Signed-off-by: James Morris --- include/linux/key.h | 3 +++ security/keys/encrypted-keys/encrypted.c | 4 ++-- security/keys/encrypted-keys/masterkey_trusted.c | 2 ++ security/keys/trusted.c | 4 ++-- 4 files changed, 9 insertions(+), 4 deletions(-) diff --git a/include/linux/key.h b/include/linux/key.h index 183a6af7715d..bfc014c57351 100644 --- a/include/linux/key.h +++ b/include/linux/key.h @@ -293,6 +293,9 @@ static inline bool key_is_instantiated(const struct key *key) (rcu_dereference_protected((KEY)->payload.rcudata, \ rwsem_is_locked(&((struct key *)(KEY))->sem))) +#define rcu_assign_keypointer(KEY, PAYLOAD) \ + (rcu_assign_pointer((KEY)->payload.rcudata, PAYLOAD)) + #ifdef CONFIG_SYSCTL extern ctl_table key_sysctls[]; #endif diff --git a/security/keys/encrypted-keys/encrypted.c b/security/keys/encrypted-keys/encrypted.c index 41144f71d615..d91efb6901e9 100644 --- a/security/keys/encrypted-keys/encrypted.c +++ b/security/keys/encrypted-keys/encrypted.c @@ -810,7 +810,7 @@ static int encrypted_instantiate(struct key *key, const void *data, goto out; } - rcu_assign_pointer(key->payload.data, epayload); + rcu_assign_keypointer(key, epayload); out: kfree(datablob); return ret; @@ -874,7 +874,7 @@ static int encrypted_update(struct key *key, const void *data, size_t datalen) memcpy(new_epayload->payload_data, epayload->payload_data, epayload->payload_datalen); - rcu_assign_pointer(key->payload.data, new_epayload); + rcu_assign_keypointer(key, new_epayload); call_rcu(&epayload->rcu, encrypted_rcu_free); out: kfree(buf); diff --git a/security/keys/encrypted-keys/masterkey_trusted.c b/security/keys/encrypted-keys/masterkey_trusted.c index df87272e3f51..8c16c3e472e7 100644 --- a/security/keys/encrypted-keys/masterkey_trusted.c +++ b/security/keys/encrypted-keys/masterkey_trusted.c @@ -18,6 +18,8 @@ #include #include #include +#include +#include "encrypted.h" /* * request_trusted_key - request the trusted key diff --git a/security/keys/trusted.c b/security/keys/trusted.c index 0ed5fdf238a2..2d5d041f2049 100644 --- a/security/keys/trusted.c +++ b/security/keys/trusted.c @@ -993,7 +993,7 @@ out: kfree(datablob); kfree(options); if (!ret) - rcu_assign_pointer(key->payload.data, payload); + rcu_assign_keypointer(key, payload); else kfree(payload); return ret; @@ -1067,7 +1067,7 @@ static int trusted_update(struct key *key, const void *data, size_t datalen) goto out; } } - rcu_assign_pointer(key->payload.data, new_p); + rcu_assign_keypointer(key, new_p); call_rcu(&p->rcu, trusted_rcu_free); out: kfree(datablob); From 6ac6172a935d1faf7ef259802267657bc0007a62 Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Tue, 17 Jan 2012 20:40:02 +0000 Subject: [PATCH 570/820] encrypted-keys: fix rcu and sparse messages Enabling CONFIG_PROVE_RCU and CONFIG_SPARSE_RCU_POINTER resulted in "suspicious rcu_dereference_check() usage!" and "incompatible types in comparison expression (different address spaces)" messages. Access the masterkey directly when holding the rwsem. Changelog v1: - Use either rcu_read_lock()/rcu_derefence_key()/rcu_read_unlock() or remove the unnecessary rcu_derefence() - David Howells Reported-by: Dmitry Kasatkin Signed-off-by: Mimi Zohar Signed-off-by: David Howells Signed-off-by: James Morris --- security/keys/encrypted-keys/encrypted.c | 2 +- security/keys/encrypted-keys/masterkey_trusted.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/security/keys/encrypted-keys/encrypted.c b/security/keys/encrypted-keys/encrypted.c index d91efb6901e9..2d1bb8af7696 100644 --- a/security/keys/encrypted-keys/encrypted.c +++ b/security/keys/encrypted-keys/encrypted.c @@ -314,7 +314,7 @@ static struct key *request_user_key(const char *master_desc, u8 **master_key, goto error; down_read(&ukey->sem); - upayload = rcu_dereference(ukey->payload.data); + upayload = ukey->payload.data; *master_key = upayload->data; *master_keylen = upayload->datalen; error: diff --git a/security/keys/encrypted-keys/masterkey_trusted.c b/security/keys/encrypted-keys/masterkey_trusted.c index 8c16c3e472e7..013f7e5d3a2f 100644 --- a/security/keys/encrypted-keys/masterkey_trusted.c +++ b/security/keys/encrypted-keys/masterkey_trusted.c @@ -39,7 +39,7 @@ struct key *request_trusted_key(const char *trusted_desc, goto error; down_read(&tkey->sem); - tpayload = rcu_dereference(tkey->payload.data); + tpayload = tkey->payload.data; *master_key = tpayload->key; *master_keylen = tpayload->key_len; error: From 5e8898e97a5db4125d944070922164d1d09a2689 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Tue, 17 Jan 2012 17:12:03 +0200 Subject: [PATCH 571/820] lib: digital signature config option name change It was reported that DIGSIG is confusing name for digital signature module. It was suggested to rename DIGSIG to SIGNATURE. Requested-by: Linus Torvalds Suggested-by: Pavel Machek Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- include/linux/digsig.h | 4 ++-- lib/Kconfig | 2 +- lib/Makefile | 2 +- security/integrity/Kconfig | 2 +- 4 files changed, 5 insertions(+), 5 deletions(-) diff --git a/include/linux/digsig.h b/include/linux/digsig.h index efae755017d7..b01558b15814 100644 --- a/include/linux/digsig.h +++ b/include/linux/digsig.h @@ -46,7 +46,7 @@ struct signature_hdr { char mpi[0]; } __packed; -#if defined(CONFIG_DIGSIG) || defined(CONFIG_DIGSIG_MODULE) +#if defined(CONFIG_SIGNATURE) || defined(CONFIG_SIGNATURE_MODULE) int digsig_verify(struct key *keyring, const char *sig, int siglen, const char *digest, int digestlen); @@ -59,6 +59,6 @@ static inline int digsig_verify(struct key *keyring, const char *sig, return -EOPNOTSUPP; } -#endif /* CONFIG_DIGSIG */ +#endif /* CONFIG_SIGNATURE */ #endif /* _DIGSIG_H */ diff --git a/lib/Kconfig b/lib/Kconfig index 201e1b33d721..854735d96dc3 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -302,7 +302,7 @@ config MPILIB_EXTRA This code in unnecessary for RSA digital signature verification, and can be compiled if needed. -config DIGSIG +config SIGNATURE tristate "In-kernel signature checker" depends on KEYS select MPILIB diff --git a/lib/Makefile b/lib/Makefile index dace162c7e1c..d71aae1b01b3 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -119,7 +119,7 @@ obj-$(CONFIG_CORDIC) += cordic.o obj-$(CONFIG_DQL) += dynamic_queue_limits.o obj-$(CONFIG_MPILIB) += mpi/ -obj-$(CONFIG_DIGSIG) += digsig.o +obj-$(CONFIG_SIGNATURE) += digsig.o hostprogs-y := gen_crc32table clean-files := crc32table.h diff --git a/security/integrity/Kconfig b/security/integrity/Kconfig index d384ea921482..ff60bf72881f 100644 --- a/security/integrity/Kconfig +++ b/security/integrity/Kconfig @@ -7,7 +7,7 @@ config INTEGRITY_DIGSIG boolean "Digital signature verification using multiple keyrings" depends on INTEGRITY && KEYS default n - select DIGSIG + select SIGNATURE help This option enables digital signature verification support using multiple keyrings. It defines separate keyrings for each From be440ec77325d02ad5e43676eb8d868426d35cd5 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Tue, 17 Jan 2012 17:12:04 +0200 Subject: [PATCH 572/820] lib: digital signature dependency fix Randy Dunlap reported build break: ERROR: "crypto_alloc_shash" [lib/digsig.ko] undefined! ERROR: "crypto_shash_final" [lib/digsig.ko] undefined! ERROR: "crypto_shash_update" [lib/digsig.ko] undefined! ERROR: "crypto_destroy_tfm" [lib/digsig.ko] undefined! Added CRYPTO dependency and selected SHA1 algorithm. Reported-by: Randy Dunlap Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- lib/Kconfig | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/lib/Kconfig b/lib/Kconfig index 854735d96dc3..dbaf19ea84e8 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -304,7 +304,8 @@ config MPILIB_EXTRA config SIGNATURE tristate "In-kernel signature checker" - depends on KEYS + depends on KEYS && CRYPTO + select CRYPTO_SHA1 select MPILIB help Digital signature verification. Currently only RSA is supported. From 68adcad570a2ce0496918ddd9f8c83138d9bcc93 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Tue, 17 Jan 2012 17:12:05 +0200 Subject: [PATCH 573/820] lib: MPILIB Kconfig description update It was reported that description of the MPILIB_EXTRA is confusing. Indeed it was copy-paste typo. It is fixed here. Reported-by: Valdis Kletnieks Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- lib/Kconfig | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/lib/Kconfig b/lib/Kconfig index dbaf19ea84e8..1cbc1fe13619 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -296,11 +296,9 @@ config MPILIB_EXTRA bool "Multiprecision maths library - additional sources" depends on MPILIB help - Multiprecision maths library from GnuPG. - It is used to implement RSA digital signature verification, - which is used by IMA/EVM digital signature extension. - This code in unnecessary for RSA digital signature verification, - and can be compiled if needed. + Additional sources of multiprecision maths library from GnuPG. + This code is unnecessary for RSA digital signature verification, + but can be compiled if needed. config SIGNATURE tristate "In-kernel signature checker" From 2e5f094b9dbf9463ab93f86351cd1a8dc88942cc Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Tue, 17 Jan 2012 17:12:06 +0200 Subject: [PATCH 574/820] lib: Removed MPILIB, MPILIB_EXTRA, and SIGNATURE prompts As modules are expected to select MPILIB, MPILIB_EXTRA, and SIGNATURE, removed Kconfig prompts. Requested-by: Linus Torvalds Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- lib/Kconfig | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/lib/Kconfig b/lib/Kconfig index 1cbc1fe13619..169eb7c598e5 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -286,14 +286,14 @@ config CORDIC calculations are in fixed point. Module will be called cordic. config MPILIB - tristate "Multiprecision maths library" + tristate help Multiprecision maths library from GnuPG. It is used to implement RSA digital signature verification, which is used by IMA/EVM digital signature extension. config MPILIB_EXTRA - bool "Multiprecision maths library - additional sources" + bool depends on MPILIB help Additional sources of multiprecision maths library from GnuPG. @@ -301,7 +301,7 @@ config MPILIB_EXTRA but can be compiled if needed. config SIGNATURE - tristate "In-kernel signature checker" + tristate depends on KEYS && CRYPTO select CRYPTO_SHA1 select MPILIB From f1be242c95257b199d8b679bc952ca33487c9af6 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Tue, 17 Jan 2012 17:12:07 +0200 Subject: [PATCH 575/820] integrity: digital signature config option name change Similar to SIGNATURE, rename INTEGRITY_DIGSIG to INTEGRITY_SIGNATURE. Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- security/integrity/Kconfig | 2 +- security/integrity/Makefile | 2 +- security/integrity/integrity.h | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/security/integrity/Kconfig b/security/integrity/Kconfig index ff60bf72881f..5bd1cc1b4a54 100644 --- a/security/integrity/Kconfig +++ b/security/integrity/Kconfig @@ -3,7 +3,7 @@ config INTEGRITY def_bool y depends on IMA || EVM -config INTEGRITY_DIGSIG +config INTEGRITY_SIGNATURE boolean "Digital signature verification using multiple keyrings" depends on INTEGRITY && KEYS default n diff --git a/security/integrity/Makefile b/security/integrity/Makefile index bece0563ee5e..d43799cc14f6 100644 --- a/security/integrity/Makefile +++ b/security/integrity/Makefile @@ -3,7 +3,7 @@ # obj-$(CONFIG_INTEGRITY) += integrity.o -obj-$(CONFIG_INTEGRITY_DIGSIG) += digsig.o +obj-$(CONFIG_INTEGRITY_SIGNATURE) += digsig.o integrity-y := iint.o diff --git a/security/integrity/integrity.h b/security/integrity/integrity.h index 4da6ba81d153..7a25ecec5aaa 100644 --- a/security/integrity/integrity.h +++ b/security/integrity/integrity.h @@ -51,7 +51,7 @@ struct integrity_iint_cache *integrity_iint_find(struct inode *inode); #define INTEGRITY_KEYRING_IMA 2 #define INTEGRITY_KEYRING_MAX 3 -#ifdef CONFIG_INTEGRITY_DIGSIG +#ifdef CONFIG_INTEGRITY_SIGNATURE int integrity_digsig_verify(const unsigned int id, const char *sig, int siglen, const char *digest, int digestlen); @@ -65,7 +65,7 @@ static inline int integrity_digsig_verify(const unsigned int id, return -EOPNOTSUPP; } -#endif /* CONFIG_INTEGRITY_DIGSIG */ +#endif /* CONFIG_INTEGRITY_SIGNATURE */ /* set during initialization */ extern int iint_initialized; From 68f30fbee19cc67849b9fa8e153ede70758afe81 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 17 Jan 2012 15:35:37 -0800 Subject: [PATCH 576/820] x86, tsc: Fix SMI induced variation in quick_pit_calibrate() pit_expect_msb() returns success wrongly in the below SMI scenario: a. pit_verify_msb() has not yet seen the MSB transition. b. we are close to the MSB transition though and got a SMI immediately after returning from pit_verify_msb() which didn't see the MSB transition. PIT MSB transition has happened somewhere during SMI execution. c. returned from SMI and we noted down the 'tsc', saw the pit MSB change now and exited the loop to calculate 'deltatsc'. Instead of noting the TSC at the MSB transition, we are way off because of the SMI. And as the SMI happened between the pit_verify_msb() and before the 'tsc' is recorded in the for loop, 'delattsc' (d1/d2 in quick_pit_calibrate()) will be small and quick_pit_calibrate() will not notice this error. Depending on whether SMI disturbance happens while computing d1 or d2, we will see the TSC calibrated value smaller or bigger than the expected value. As a result, in a cluster we were seeing a variation of approximately +/- 20MHz in the calibrated values, resulting in NTP failures. [ As far as the SMI source is concerned, this is a periodic SMI that gets disabled after ACPI is enabled by the OS. But the TSC calibration happens before the ACPI is enabled. ] To address this, change pit_expect_msb() so that - the 'tsc' is the TSC in between the two reads that read the MSB change from the PIT (same as before) - the 'delta' is the difference in TSC from *before* the MSB changed to *after* the MSB changed. Now the delta is twice as big as before (it covers four PIT accesses, roughly 4us) and quick_pit_calibrate() will loop a bit longer to get the calibrated value with in the 500ppm precision. As the delta (d1/d2) covers four PIT accesses, actual calibrated result might be closer to 250ppm precision. As the loop now takes longer to stabilize, double MAX_QUICK_PIT_MS to 50. SMI disturbance will showup as much larger delta's and the loop will take longer than usual for the result to be with in the accepted precision. Or will fallback to slow PIT calibration if it takes more than 50msec. Also while we are at this, remove the calibration correction that aims to get the result to the middle of the error bars. We really don't know which direction to correct into, so remove it. Reported-and-tested-by: Suresh Siddha Signed-off-by: Linus Torvalds Signed-off-by: Suresh Siddha Link: http://lkml.kernel.org/r/1326843337.5291.4.camel@sbsiddha-mobl2 Signed-off-by: H. Peter Anvin --- arch/x86/kernel/tsc.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 2c9cf0fd78f5..f54694611172 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -290,14 +290,15 @@ static inline int pit_verify_msb(unsigned char val) static inline int pit_expect_msb(unsigned char val, u64 *tscp, unsigned long *deltap) { int count; - u64 tsc = 0; + u64 tsc = 0, prev_tsc = 0; for (count = 0; count < 50000; count++) { if (!pit_verify_msb(val)) break; + prev_tsc = tsc; tsc = get_cycles(); } - *deltap = get_cycles() - tsc; + *deltap = get_cycles() - prev_tsc; *tscp = tsc; /* @@ -311,9 +312,9 @@ static inline int pit_expect_msb(unsigned char val, u64 *tscp, unsigned long *de * How many MSB values do we want to see? We aim for * a maximum error rate of 500ppm (in practice the * real error is much smaller), but refuse to spend - * more than 25ms on it. + * more than 50ms on it. */ -#define MAX_QUICK_PIT_MS 25 +#define MAX_QUICK_PIT_MS 50 #define MAX_QUICK_PIT_ITERATIONS (MAX_QUICK_PIT_MS * PIT_TICK_RATE / 1000 / 256) static unsigned long quick_pit_calibrate(void) @@ -383,15 +384,12 @@ success: * * As a result, we can depend on there not being * any odd delays anywhere, and the TSC reads are - * reliable (within the error). We also adjust the - * delta to the middle of the error bars, just - * because it looks nicer. + * reliable (within the error). * * kHz = ticks / time-in-seconds / 1000; * kHz = (t2 - t1) / (I * 256 / PIT_TICK_RATE) / 1000 * kHz = ((t2 - t1) * PIT_TICK_RATE) / (I * 256 * 1000) */ - delta += (long)(d2 - d1)/2; delta *= PIT_TICK_RATE; do_div(delta, i*256*1000); printk("Fast TSC calibration using PIT\n"); From 265fe02b6fade1704f01c4e96f99ac051b7fab9d Mon Sep 17 00:00:00 2001 From: Grant Likely Date: Wed, 11 Jan 2012 11:03:41 -0700 Subject: [PATCH 577/820] gpio/it8761e: Restrict it8761e gpio driver to x86. This driver does an unconditional read of io space during module init which causes a bad dereference on ARM. It looks to me like this is an x86 only drivers, so restrict it to only compile on x86. Signed-off-by: Grant Likely Cc: Denis Turischev --- drivers/gpio/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index 37c4bd1cacd5..d0c41188d4e5 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -87,6 +87,7 @@ config GPIO_GENERIC_PLATFORM config GPIO_IT8761E tristate "IT8761E GPIO support" + depends on X86 # unconditional access to IO space. help Say yes here to support GPIO functionality of IT8761E super I/O chip. From 53e5ab946a1773f589b97df6ae1cc72b52656a7d Mon Sep 17 00:00:00 2001 From: Sachin Kamat Date: Tue, 10 Jan 2012 05:46:57 -0300 Subject: [PATCH 578/820] [media] s5p-fimc: Fix incorrect control ID assignment This patch fixes the mismatch between control IDs (CID) and controls for hflip, vflip and rotate. Signed-off-by: Sachin Kamat Acked-by: Kyungmin Park Signed-off-by: Marek Szyprowski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-fimc/fimc-core.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/media/video/s5p-fimc/fimc-core.c b/drivers/media/video/s5p-fimc/fimc-core.c index f5cbb8a4c540..81bcbb9492ea 100644 --- a/drivers/media/video/s5p-fimc/fimc-core.c +++ b/drivers/media/video/s5p-fimc/fimc-core.c @@ -848,11 +848,11 @@ int fimc_ctrls_create(struct fimc_ctx *ctx) v4l2_ctrl_handler_init(&ctx->ctrl_handler, 4); ctx->ctrl_rotate = v4l2_ctrl_new_std(&ctx->ctrl_handler, &fimc_ctrl_ops, - V4L2_CID_HFLIP, 0, 1, 1, 0); + V4L2_CID_ROTATE, 0, 270, 90, 0); ctx->ctrl_hflip = v4l2_ctrl_new_std(&ctx->ctrl_handler, &fimc_ctrl_ops, - V4L2_CID_VFLIP, 0, 1, 1, 0); + V4L2_CID_HFLIP, 0, 1, 1, 0); ctx->ctrl_vflip = v4l2_ctrl_new_std(&ctx->ctrl_handler, &fimc_ctrl_ops, - V4L2_CID_ROTATE, 0, 270, 90, 0); + V4L2_CID_VFLIP, 0, 1, 1, 0); if (variant->has_alpha) ctx->ctrl_alpha = v4l2_ctrl_new_std(&ctx->ctrl_handler, &fimc_ctrl_ops, V4L2_CID_ALPHA_COMPONENT, From b1aa60892ecc1ba03185e7dfed99cc9202ef03b7 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Thu, 12 Jan 2012 17:49:28 -0300 Subject: [PATCH 579/820] [media] drivers/media/video/s5p-fimc/fimc-capture.c: adjust double test Rewrite a duplicated test to test the correct value The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @@ expression E; @@ ( * E || ... || E | * E && ... && E ) // Signed-off-by: Julia Lawall Acked-by: Kyungmin Park Signed-off-by: Marek Szyprowski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-fimc/fimc-capture.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/media/video/s5p-fimc/fimc-capture.c b/drivers/media/video/s5p-fimc/fimc-capture.c index 510cfab477ff..a9e9653beeb4 100644 --- a/drivers/media/video/s5p-fimc/fimc-capture.c +++ b/drivers/media/video/s5p-fimc/fimc-capture.c @@ -693,7 +693,7 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx, mf->code = 0; continue; } - if (mf->width != tfmt->width || mf->width != tfmt->width) { + if (mf->width != tfmt->width || mf->height != tfmt->height) { u32 fcc = ffmt->fourcc; tfmt->width = mf->width; tfmt->height = mf->height; @@ -702,7 +702,8 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx, NULL, &fcc, FIMC_SD_PAD_SOURCE); if (ffmt && ffmt->mbus_code) mf->code = ffmt->mbus_code; - if (mf->width != tfmt->width || mf->width != tfmt->width) + if (mf->width != tfmt->width || + mf->height != tfmt->height) continue; tfmt->code = mf->code; } @@ -710,7 +711,7 @@ static int fimc_pipeline_try_format(struct fimc_ctx *ctx, ret = v4l2_subdev_call(csis, pad, set_fmt, NULL, &sfmt); if (mf->code == tfmt->code && - mf->width == tfmt->width && mf->width == tfmt->width) + mf->width == tfmt->width && mf->height == tfmt->height) break; } From bb869368f36abecc427ffa94193439a05c0aac8d Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Thu, 12 Jan 2012 17:49:30 -0300 Subject: [PATCH 580/820] [media] drivers/media/video/s5p-mfc/s5p_mfc.c: adjust double test Rewrite a duplicated test to test the correct value The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @@ expression E; @@ ( * E || ... || E | * E && ... && E ) // Signed-off-by: Julia Lawall Acked-by: Kyungmin Park Signed-off-by: Marek Szyprowski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-mfc/s5p_mfc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/video/s5p-mfc/s5p_mfc.c b/drivers/media/video/s5p-mfc/s5p_mfc.c index e43e128baf5f..90cc05855ecb 100644 --- a/drivers/media/video/s5p-mfc/s5p_mfc.c +++ b/drivers/media/video/s5p-mfc/s5p_mfc.c @@ -475,7 +475,7 @@ static void s5p_mfc_handle_seq_done(struct s5p_mfc_ctx *ctx, ctx->mv_size = 0; } ctx->dpb_count = s5p_mfc_get_dpb_count(); - if (ctx->img_width == 0 || ctx->img_width == 0) + if (ctx->img_width == 0 || ctx->img_height == 0) ctx->state = MFCINST_ERROR; else ctx->state = MFCINST_HEAD_PARSED; From 0da0a7833824b51488b81db002bb47ad8688f9bb Mon Sep 17 00:00:00 2001 From: Kamil Debski Date: Tue, 27 Dec 2011 10:07:24 -0300 Subject: [PATCH 581/820] [media] s5p-mfc: Fix volatile controls setup Signed-off-by: Kamil Debski Signed-off-by: Kyungmin Park Signed-off-by: Marek Szyprowski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-mfc/s5p_mfc_dec.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/video/s5p-mfc/s5p_mfc_dec.c b/drivers/media/video/s5p-mfc/s5p_mfc_dec.c index 844a4d7797bc..c25ec022d267 100644 --- a/drivers/media/video/s5p-mfc/s5p_mfc_dec.c +++ b/drivers/media/video/s5p-mfc/s5p_mfc_dec.c @@ -165,7 +165,7 @@ static struct mfc_control controls[] = { .maximum = 32, .step = 1, .default_value = 1, - .flags = V4L2_CTRL_FLAG_VOLATILE, + .is_volatile = 1, }, }; From 7f6cce69e2667fdf0ee978641ee4325ea05ce3f8 Mon Sep 17 00:00:00 2001 From: Kamil Debski Date: Mon, 2 Jan 2012 09:19:25 -0300 Subject: [PATCH 582/820] [media] s5p-g2d: fixed a bug in controls setting function Signed-off-by: Kamil Debski Signed-off-by: Kyungmin Park Signed-off-by: Marek Szyprowski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-g2d/g2d.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/video/s5p-g2d/g2d.c b/drivers/media/video/s5p-g2d/g2d.c index c40b0dde1883..febaa673d363 100644 --- a/drivers/media/video/s5p-g2d/g2d.c +++ b/drivers/media/video/s5p-g2d/g2d.c @@ -184,6 +184,7 @@ static int g2d_s_ctrl(struct v4l2_ctrl *ctrl) ctx->rop = ROP4_INVERT; else ctx->rop = ROP4_COPY; + break; default: v4l2_err(&ctx->dev->v4l2_dev, "unknown control\n"); return -EINVAL; From 719c174ee9005e44b61bc8ab295f05851a989e21 Mon Sep 17 00:00:00 2001 From: Marek Szyprowski Date: Fri, 13 Jan 2012 05:12:38 -0300 Subject: [PATCH 583/820] [media] s5p-jpeg: adapt to recent videobuf2 changes queue_setup callback has been extended with struct v4l2_format *fmt parameter in 2d86401c2c commit. This patch adds this parameter to s5p-jpeg driver. Signed-off-by: Marek Szyprowski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-jpeg/jpeg-core.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/media/video/s5p-jpeg/jpeg-core.c b/drivers/media/video/s5p-jpeg/jpeg-core.c index f841a3e9845c..1105a8749c8b 100644 --- a/drivers/media/video/s5p-jpeg/jpeg-core.c +++ b/drivers/media/video/s5p-jpeg/jpeg-core.c @@ -989,9 +989,10 @@ static struct v4l2_m2m_ops s5p_jpeg_m2m_ops = { * ============================================================================ */ -static int s5p_jpeg_queue_setup(struct vb2_queue *vq, unsigned int *nbuffers, - unsigned int *nplanes, unsigned int sizes[], - void *alloc_ctxs[]) +static int s5p_jpeg_queue_setup(struct vb2_queue *vq, + const struct v4l2_format *fmt, + unsigned int *nbuffers, unsigned int *nplanes, + unsigned int sizes[], void *alloc_ctxs[]) { struct s5p_jpeg_ctx *ctx = vb2_get_drv_priv(vq); struct s5p_jpeg_q_data *q_data = NULL; From bad321f1bb8f45adc348a3913632387891dd6ac6 Mon Sep 17 00:00:00 2001 From: Patrick Boettcher Date: Tue, 17 Jan 2012 05:49:14 -0300 Subject: [PATCH 584/820] [media] DVB-CORE: remove superfluous DTV_CMDs This small patch removes superfluous DTV_CMDs from dvb_frontend.c which were added in the initially when ISBD-T support was added. They were there unnoticed even though compilers should have warning about those duplicates. Finally they did and now we can remove them. Thanks to Dan Carpenter for pointing that out. Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-core/dvb_frontend.c | 19 ------------------- 1 file changed, 19 deletions(-) diff --git a/drivers/media/dvb/dvb-core/dvb_frontend.c b/drivers/media/dvb/dvb-core/dvb_frontend.c index 2998f38d64a2..fbbe545a74cb 100644 --- a/drivers/media/dvb/dvb-core/dvb_frontend.c +++ b/drivers/media/dvb/dvb-core/dvb_frontend.c @@ -1012,25 +1012,6 @@ static struct dtv_cmds_h dtv_cmds[DTV_MAX_COMMAND + 1] = { _DTV_CMD(DTV_ISDBT_LAYERC_SEGMENT_COUNT, 1, 0), _DTV_CMD(DTV_ISDBT_LAYERC_TIME_INTERLEAVING, 1, 0), - _DTV_CMD(DTV_ISDBT_PARTIAL_RECEPTION, 0, 0), - _DTV_CMD(DTV_ISDBT_SOUND_BROADCASTING, 0, 0), - _DTV_CMD(DTV_ISDBT_SB_SUBCHANNEL_ID, 0, 0), - _DTV_CMD(DTV_ISDBT_SB_SEGMENT_IDX, 0, 0), - _DTV_CMD(DTV_ISDBT_SB_SEGMENT_COUNT, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYER_ENABLED, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERA_FEC, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERA_MODULATION, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERA_SEGMENT_COUNT, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERA_TIME_INTERLEAVING, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERB_FEC, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERB_MODULATION, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERB_SEGMENT_COUNT, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERB_TIME_INTERLEAVING, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERC_FEC, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERC_MODULATION, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERC_SEGMENT_COUNT, 0, 0), - _DTV_CMD(DTV_ISDBT_LAYERC_TIME_INTERLEAVING, 0, 0), - _DTV_CMD(DTV_ISDBS_TS_ID, 1, 0), _DTV_CMD(DTV_DVBT2_PLP_ID, 1, 0), From 68dc8bc5806a583b547c9543c4a864ec93b423dc Mon Sep 17 00:00:00 2001 From: Patrick Boettcher Date: Tue, 17 Jan 2012 08:41:20 -0300 Subject: [PATCH 585/820] [media] : add MODULE_FIRMWARE to dib0700 The dib0700 needs a binary firmware file. This patch added the MODULE_FIRMWARE-macro. Signed-off-by: Christoph Anton Mitterer Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/dib0700_core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/dvb/dvb-usb/dib0700_core.c b/drivers/media/dvb/dvb-usb/dib0700_core.c index 156cbfc9c79d..2d259ca34b05 100644 --- a/drivers/media/dvb/dvb-usb/dib0700_core.c +++ b/drivers/media/dvb/dvb-usb/dib0700_core.c @@ -854,6 +854,7 @@ static void __exit dib0700_module_exit(void) module_init (dib0700_module_init); module_exit (dib0700_module_exit); +MODULE_FIRMWARE("dvb-usb-dib0700-1.20.fw"); MODULE_AUTHOR("Patrick Boettcher "); MODULE_DESCRIPTION("Driver for devices based on DiBcom DiB0700 - USB bridge"); MODULE_VERSION("1.0"); From 9b6ba57bd45da30173f37d5099805d687440e5c0 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Tue, 10 Jan 2012 14:11:22 -0300 Subject: [PATCH 586/820] [media] DVB: dib0700, move Nova-TD Stick to a separate set To properly support the three LEDs which are on the stick, we need a special handling in the ->frontend_attach function. Thus let's have a separate ->frontend_attach instead of ifs in the common one. The hadnling itself will be added in further patches. Signed-off-by: Jiri Slaby Acked-by: Michael Krufky Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/dib0700_devices.c | 57 +++++++++++++++++++-- 1 file changed, 52 insertions(+), 5 deletions(-) diff --git a/drivers/media/dvb/dvb-usb/dib0700_devices.c b/drivers/media/dvb/dvb-usb/dib0700_devices.c index 81ef4b46f790..3c6ee54d8fc3 100644 --- a/drivers/media/dvb/dvb-usb/dib0700_devices.c +++ b/drivers/media/dvb/dvb-usb/dib0700_devices.c @@ -3892,7 +3892,58 @@ struct dvb_usb_device_properties dib0700_devices[] = { } }, - .num_device_descs = 6, + .num_device_descs = 1, + .devices = { + { "Hauppauge Nova-TD Stick (52009)", + { &dib0700_usb_id_table[35], NULL }, + { NULL }, + }, + }, + + .rc.core = { + .rc_interval = DEFAULT_RC_INTERVAL, + .rc_codes = RC_MAP_DIB0700_RC5_TABLE, + .module_name = "dib0700", + .rc_query = dib0700_rc_query_old_firmware, + .allowed_protos = RC_TYPE_RC5 | + RC_TYPE_RC6 | + RC_TYPE_NEC, + .change_protocol = dib0700_change_protocol, + }, + }, { DIB0700_DEFAULT_DEVICE_PROPERTIES, + + .num_adapters = 2, + .adapter = { + { + .num_frontends = 1, + .fe = {{ + .caps = DVB_USB_ADAP_HAS_PID_FILTER | DVB_USB_ADAP_PID_FILTER_CAN_BE_TURNED_OFF, + .pid_filter_count = 32, + .pid_filter = stk70x0p_pid_filter, + .pid_filter_ctrl = stk70x0p_pid_filter_ctrl, + .frontend_attach = stk7070pd_frontend_attach0, + .tuner_attach = dib7070p_tuner_attach, + + DIB0700_DEFAULT_STREAMING_CONFIG(0x02), + }}, + .size_of_priv = sizeof(struct dib0700_adapter_state), + }, { + .num_frontends = 1, + .fe = {{ + .caps = DVB_USB_ADAP_HAS_PID_FILTER | DVB_USB_ADAP_PID_FILTER_CAN_BE_TURNED_OFF, + .pid_filter_count = 32, + .pid_filter = stk70x0p_pid_filter, + .pid_filter_ctrl = stk70x0p_pid_filter_ctrl, + .frontend_attach = stk7070pd_frontend_attach1, + .tuner_attach = dib7070p_tuner_attach, + + DIB0700_DEFAULT_STREAMING_CONFIG(0x03), + }}, + .size_of_priv = sizeof(struct dib0700_adapter_state), + } + }, + + .num_device_descs = 5, .devices = { { "DiBcom STK7070PD reference design", { &dib0700_usb_id_table[17], NULL }, @@ -3902,10 +3953,6 @@ struct dvb_usb_device_properties dib0700_devices[] = { { &dib0700_usb_id_table[18], NULL }, { NULL }, }, - { "Hauppauge Nova-TD Stick (52009)", - { &dib0700_usb_id_table[35], NULL }, - { NULL }, - }, { "Hauppauge Nova-TD-500 (84xxx)", { &dib0700_usb_id_table[36], NULL }, { NULL }, From 52fd5b2ea72863c53c8ab2e839983004ce33e559 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Tue, 10 Jan 2012 14:11:23 -0300 Subject: [PATCH 587/820] [media] DVB: dib0700, separate stk7070pd initialization The start is common for both stk7070pd and novatd specific routine. This is just a preparation for the next patch. Signed-off-by: Jiri Slaby Acked-by: Michael Krufky Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/dib0700_devices.c | 26 +++++++++++++-------- 1 file changed, 16 insertions(+), 10 deletions(-) diff --git a/drivers/media/dvb/dvb-usb/dib0700_devices.c b/drivers/media/dvb/dvb-usb/dib0700_devices.c index 3c6ee54d8fc3..e5c2bd2b2be8 100644 --- a/drivers/media/dvb/dvb-usb/dib0700_devices.c +++ b/drivers/media/dvb/dvb-usb/dib0700_devices.c @@ -3066,19 +3066,25 @@ static struct dib7000p_config stk7070pd_dib7000p_config[2] = { } }; +static void stk7070pd_init(struct dvb_usb_device *dev) +{ + dib0700_set_gpio(dev, GPIO6, GPIO_OUT, 1); + msleep(10); + dib0700_set_gpio(dev, GPIO9, GPIO_OUT, 1); + dib0700_set_gpio(dev, GPIO4, GPIO_OUT, 1); + dib0700_set_gpio(dev, GPIO7, GPIO_OUT, 1); + dib0700_set_gpio(dev, GPIO10, GPIO_OUT, 0); + + dib0700_ctrl_clock(dev, 72, 1); + + msleep(10); + dib0700_set_gpio(dev, GPIO10, GPIO_OUT, 1); +} + static int stk7070pd_frontend_attach0(struct dvb_usb_adapter *adap) { - dib0700_set_gpio(adap->dev, GPIO6, GPIO_OUT, 1); - msleep(10); - dib0700_set_gpio(adap->dev, GPIO9, GPIO_OUT, 1); - dib0700_set_gpio(adap->dev, GPIO4, GPIO_OUT, 1); - dib0700_set_gpio(adap->dev, GPIO7, GPIO_OUT, 1); - dib0700_set_gpio(adap->dev, GPIO10, GPIO_OUT, 0); + stk7070pd_init(adap->dev); - dib0700_ctrl_clock(adap->dev, 72, 1); - - msleep(10); - dib0700_set_gpio(adap->dev, GPIO10, GPIO_OUT, 1); msleep(10); dib0700_set_gpio(adap->dev, GPIO0, GPIO_OUT, 1); From 2b05b881ec9b04cc308f0880ea604843bcb2265a Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Tue, 10 Jan 2012 14:11:24 -0300 Subject: [PATCH 588/820] [media] DVB: dib0700, add corrected Nova-TD frontend_attach This means cut & paste from the former f. attach. But while at it write to the right GPIO to turn on the right LED. Also turn the other two off jsut for sure. Signed-off-by: Jiri Slaby Acked-by: Michael Krufky Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/dib0700_devices.c | 36 +++++++++++++++++++-- 1 file changed, 34 insertions(+), 2 deletions(-) diff --git a/drivers/media/dvb/dvb-usb/dib0700_devices.c b/drivers/media/dvb/dvb-usb/dib0700_devices.c index e5c2bd2b2be8..3ab45aedbdfb 100644 --- a/drivers/media/dvb/dvb-usb/dib0700_devices.c +++ b/drivers/media/dvb/dvb-usb/dib0700_devices.c @@ -3105,6 +3105,38 @@ static int stk7070pd_frontend_attach1(struct dvb_usb_adapter *adap) return adap->fe_adap[0].fe == NULL ? -ENODEV : 0; } +/** + * novatd_frontend_attach - Nova-TD specific attach + * + * Nova-TD has GPIO0, 1 and 2 for LEDs. So do not fiddle with them except for + * information purposes. + */ +static int novatd_frontend_attach(struct dvb_usb_adapter *adap) +{ + struct dvb_usb_device *dev = adap->dev; + + if (adap->id == 0) { + stk7070pd_init(dev); + + /* turn the power LED on, the other two off (just in case) */ + dib0700_set_gpio(dev, GPIO0, GPIO_OUT, 0); + dib0700_set_gpio(dev, GPIO1, GPIO_OUT, 0); + dib0700_set_gpio(dev, GPIO2, GPIO_OUT, 1); + + if (dib7000p_i2c_enumeration(&dev->i2c_adap, 2, 18, + stk7070pd_dib7000p_config) != 0) { + err("%s: dib7000p_i2c_enumeration failed. Cannot continue\n", + __func__); + return -ENODEV; + } + } + + adap->fe_adap[0].fe = dvb_attach(dib7000p_attach, &dev->i2c_adap, + adap->id == 0 ? 0x80 : 0x82, + &stk7070pd_dib7000p_config[adap->id]); + return adap->fe_adap[0].fe == NULL ? -ENODEV : 0; +} + /* S5H1411 */ static struct s5h1411_config pinnacle_801e_config = { .output_mode = S5H1411_PARALLEL_OUTPUT, @@ -3876,7 +3908,7 @@ struct dvb_usb_device_properties dib0700_devices[] = { .pid_filter_count = 32, .pid_filter = stk70x0p_pid_filter, .pid_filter_ctrl = stk70x0p_pid_filter_ctrl, - .frontend_attach = stk7070pd_frontend_attach0, + .frontend_attach = novatd_frontend_attach, .tuner_attach = dib7070p_tuner_attach, DIB0700_DEFAULT_STREAMING_CONFIG(0x02), @@ -3889,7 +3921,7 @@ struct dvb_usb_device_properties dib0700_devices[] = { .pid_filter_count = 32, .pid_filter = stk70x0p_pid_filter, .pid_filter_ctrl = stk70x0p_pid_filter_ctrl, - .frontend_attach = stk7070pd_frontend_attach1, + .frontend_attach = novatd_frontend_attach, .tuner_attach = dib7070p_tuner_attach, DIB0700_DEFAULT_STREAMING_CONFIG(0x03), From d43272a4e898a1e43e5ac351ab625b7a40b39e88 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Tue, 10 Jan 2012 14:11:25 -0300 Subject: [PATCH 589/820] [media] DVB: dib0700, add support for Nova-TD LEDs Add an override of read_status to intercept lock status. This allows us to switch LEDs appropriately on and off with signal un/locked. The second phase is to override sleep to properly turn off both. This is a hackish way to achieve that. Thanks to Mike Krufky for his help. Signed-off-by: Jiri Slaby Acked-by: Michael Krufky Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/dib0700.h | 2 + drivers/media/dvb/dvb-usb/dib0700_devices.c | 41 ++++++++++++++++++++- 2 files changed, 42 insertions(+), 1 deletion(-) diff --git a/drivers/media/dvb/dvb-usb/dib0700.h b/drivers/media/dvb/dvb-usb/dib0700.h index 9bd6d51b3b93..7de125c0b36f 100644 --- a/drivers/media/dvb/dvb-usb/dib0700.h +++ b/drivers/media/dvb/dvb-usb/dib0700.h @@ -48,6 +48,8 @@ struct dib0700_state { u8 disable_streaming_master_mode; u32 fw_version; u32 nb_packet_buffer_size; + int (*read_status)(struct dvb_frontend *, fe_status_t *); + int (*sleep)(struct dvb_frontend* fe); u8 buf[255]; }; diff --git a/drivers/media/dvb/dvb-usb/dib0700_devices.c b/drivers/media/dvb/dvb-usb/dib0700_devices.c index 3ab45aedbdfb..f9e966aa26e7 100644 --- a/drivers/media/dvb/dvb-usb/dib0700_devices.c +++ b/drivers/media/dvb/dvb-usb/dib0700_devices.c @@ -3105,6 +3105,35 @@ static int stk7070pd_frontend_attach1(struct dvb_usb_adapter *adap) return adap->fe_adap[0].fe == NULL ? -ENODEV : 0; } +static int novatd_read_status_override(struct dvb_frontend *fe, + fe_status_t *stat) +{ + struct dvb_usb_adapter *adap = fe->dvb->priv; + struct dvb_usb_device *dev = adap->dev; + struct dib0700_state *state = dev->priv; + int ret; + + ret = state->read_status(fe, stat); + + if (!ret) + dib0700_set_gpio(dev, adap->id == 0 ? GPIO1 : GPIO0, GPIO_OUT, + !!(*stat & FE_HAS_LOCK)); + + return ret; +} + +static int novatd_sleep_override(struct dvb_frontend* fe) +{ + struct dvb_usb_adapter *adap = fe->dvb->priv; + struct dvb_usb_device *dev = adap->dev; + struct dib0700_state *state = dev->priv; + + /* turn off LED */ + dib0700_set_gpio(dev, adap->id == 0 ? GPIO1 : GPIO0, GPIO_OUT, 0); + + return state->sleep(fe); +} + /** * novatd_frontend_attach - Nova-TD specific attach * @@ -3114,6 +3143,7 @@ static int stk7070pd_frontend_attach1(struct dvb_usb_adapter *adap) static int novatd_frontend_attach(struct dvb_usb_adapter *adap) { struct dvb_usb_device *dev = adap->dev; + struct dib0700_state *st = dev->priv; if (adap->id == 0) { stk7070pd_init(dev); @@ -3134,7 +3164,16 @@ static int novatd_frontend_attach(struct dvb_usb_adapter *adap) adap->fe_adap[0].fe = dvb_attach(dib7000p_attach, &dev->i2c_adap, adap->id == 0 ? 0x80 : 0x82, &stk7070pd_dib7000p_config[adap->id]); - return adap->fe_adap[0].fe == NULL ? -ENODEV : 0; + + if (adap->fe_adap[0].fe == NULL) + return -ENODEV; + + st->read_status = adap->fe_adap[0].fe->ops.read_status; + adap->fe_adap[0].fe->ops.read_status = novatd_read_status_override; + st->sleep = adap->fe_adap[0].fe->ops.sleep; + adap->fe_adap[0].fe->ops.sleep = novatd_sleep_override; + + return 0; } /* S5H1411 */ From f08aacf84f0a2c36cfd20d19459b036b20d2e1fb Mon Sep 17 00:00:00 2001 From: Hans Verkuil Date: Mon, 16 Jan 2012 12:27:15 -0300 Subject: [PATCH 590/820] [media] v4l2-ctrls: make control names consistent Several control names used inconsistent capitalization or were inconsistent in other ways. I also corrected a spelling mistake and fixed four strings that were too long (>31 characters). Harmless, but the string is cut off when it is returned with QUERYCTRL. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/v4l2-ctrls.c | 54 ++++++++++++++++---------------- 1 file changed, 27 insertions(+), 27 deletions(-) diff --git a/drivers/media/video/v4l2-ctrls.c b/drivers/media/video/v4l2-ctrls.c index da1f4c2d2d4b..cccd42be718a 100644 --- a/drivers/media/video/v4l2-ctrls.c +++ b/drivers/media/video/v4l2-ctrls.c @@ -465,8 +465,8 @@ const char *v4l2_ctrl_get_name(u32 id) case V4L2_CID_CHROMA_GAIN: return "Chroma Gain"; case V4L2_CID_ILLUMINATORS_1: return "Illuminator 1"; case V4L2_CID_ILLUMINATORS_2: return "Illuminator 2"; - case V4L2_CID_MIN_BUFFERS_FOR_CAPTURE: return "Minimum Number of Capture Buffers"; - case V4L2_CID_MIN_BUFFERS_FOR_OUTPUT: return "Minimum Number of Output Buffers"; + case V4L2_CID_MIN_BUFFERS_FOR_CAPTURE: return "Min Number of Capture Buffers"; + case V4L2_CID_MIN_BUFFERS_FOR_OUTPUT: return "Min Number of Output Buffers"; case V4L2_CID_ALPHA_COMPONENT: return "Alpha Component"; /* MPEG controls */ @@ -506,25 +506,25 @@ const char *v4l2_ctrl_get_name(u32 id) case V4L2_CID_MPEG_VIDEO_MUTE_YUV: return "Video Mute YUV"; case V4L2_CID_MPEG_VIDEO_DECODER_SLICE_INTERFACE: return "Decoder Slice Interface"; case V4L2_CID_MPEG_VIDEO_DECODER_MPEG4_DEBLOCK_FILTER: return "MPEG4 Loop Filter Enable"; - case V4L2_CID_MPEG_VIDEO_CYCLIC_INTRA_REFRESH_MB: return "The Number of Intra Refresh MBs"; + case V4L2_CID_MPEG_VIDEO_CYCLIC_INTRA_REFRESH_MB: return "Number of Intra Refresh MBs"; case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE: return "Frame Level Rate Control Enable"; case V4L2_CID_MPEG_VIDEO_MB_RC_ENABLE: return "H264 MB Level Rate Control"; case V4L2_CID_MPEG_VIDEO_HEADER_MODE: return "Sequence Header Mode"; - case V4L2_CID_MPEG_VIDEO_MAX_REF_PIC: return "The Max Number of Reference Picture"; + case V4L2_CID_MPEG_VIDEO_MAX_REF_PIC: return "Max Number of Reference Pics"; case V4L2_CID_MPEG_VIDEO_H263_I_FRAME_QP: return "H263 I-Frame QP Value"; - case V4L2_CID_MPEG_VIDEO_H263_P_FRAME_QP: return "H263 P frame QP Value"; - case V4L2_CID_MPEG_VIDEO_H263_B_FRAME_QP: return "H263 B frame QP Value"; + case V4L2_CID_MPEG_VIDEO_H263_P_FRAME_QP: return "H263 P-Frame QP Value"; + case V4L2_CID_MPEG_VIDEO_H263_B_FRAME_QP: return "H263 B-Frame QP Value"; case V4L2_CID_MPEG_VIDEO_H263_MIN_QP: return "H263 Minimum QP Value"; case V4L2_CID_MPEG_VIDEO_H263_MAX_QP: return "H263 Maximum QP Value"; case V4L2_CID_MPEG_VIDEO_H264_I_FRAME_QP: return "H264 I-Frame QP Value"; - case V4L2_CID_MPEG_VIDEO_H264_P_FRAME_QP: return "H264 P frame QP Value"; - case V4L2_CID_MPEG_VIDEO_H264_B_FRAME_QP: return "H264 B frame QP Value"; + case V4L2_CID_MPEG_VIDEO_H264_P_FRAME_QP: return "H264 P-Frame QP Value"; + case V4L2_CID_MPEG_VIDEO_H264_B_FRAME_QP: return "H264 B-Frame QP Value"; case V4L2_CID_MPEG_VIDEO_H264_MAX_QP: return "H264 Maximum QP Value"; case V4L2_CID_MPEG_VIDEO_H264_MIN_QP: return "H264 Minimum QP Value"; case V4L2_CID_MPEG_VIDEO_H264_8X8_TRANSFORM: return "H264 8x8 Transform Enable"; case V4L2_CID_MPEG_VIDEO_H264_CPB_SIZE: return "H264 CPB Buffer Size"; - case V4L2_CID_MPEG_VIDEO_H264_ENTROPY_MODE: return "H264 Entorpy Mode"; - case V4L2_CID_MPEG_VIDEO_H264_I_PERIOD: return "H264 I Period"; + case V4L2_CID_MPEG_VIDEO_H264_ENTROPY_MODE: return "H264 Entropy Mode"; + case V4L2_CID_MPEG_VIDEO_H264_I_PERIOD: return "H264 I-Frame Period"; case V4L2_CID_MPEG_VIDEO_H264_LEVEL: return "H264 Level"; case V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_ALPHA: return "H264 Loop Filter Alpha Offset"; case V4L2_CID_MPEG_VIDEO_H264_LOOP_FILTER_BETA: return "H264 Loop Filter Beta Offset"; @@ -535,16 +535,16 @@ const char *v4l2_ctrl_get_name(u32 id) case V4L2_CID_MPEG_VIDEO_H264_VUI_SAR_ENABLE: return "Aspect Ratio VUI Enable"; case V4L2_CID_MPEG_VIDEO_H264_VUI_SAR_IDC: return "VUI Aspect Ratio IDC"; case V4L2_CID_MPEG_VIDEO_MPEG4_I_FRAME_QP: return "MPEG4 I-Frame QP Value"; - case V4L2_CID_MPEG_VIDEO_MPEG4_P_FRAME_QP: return "MPEG4 P frame QP Value"; - case V4L2_CID_MPEG_VIDEO_MPEG4_B_FRAME_QP: return "MPEG4 B frame QP Value"; + case V4L2_CID_MPEG_VIDEO_MPEG4_P_FRAME_QP: return "MPEG4 P-Frame QP Value"; + case V4L2_CID_MPEG_VIDEO_MPEG4_B_FRAME_QP: return "MPEG4 B-Frame QP Value"; case V4L2_CID_MPEG_VIDEO_MPEG4_MIN_QP: return "MPEG4 Minimum QP Value"; case V4L2_CID_MPEG_VIDEO_MPEG4_MAX_QP: return "MPEG4 Maximum QP Value"; case V4L2_CID_MPEG_VIDEO_MPEG4_LEVEL: return "MPEG4 Level"; case V4L2_CID_MPEG_VIDEO_MPEG4_PROFILE: return "MPEG4 Profile"; case V4L2_CID_MPEG_VIDEO_MPEG4_QPEL: return "Quarter Pixel Search Enable"; - case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_BYTES: return "The Maximum Bytes Per Slice"; - case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_MB: return "The Number of MB in a Slice"; - case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MODE: return "The Slice Partitioning Method"; + case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_BYTES: return "Maximum Bytes in a Slice"; + case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MAX_MB: return "Number of MBs in a Slice"; + case V4L2_CID_MPEG_VIDEO_MULTI_SLICE_MODE: return "Slice Partitioning Method"; case V4L2_CID_MPEG_VIDEO_VBV_SIZE: return "VBV Buffer Size"; /* CAMERA controls */ @@ -580,7 +580,7 @@ const char *v4l2_ctrl_get_name(u32 id) case V4L2_CID_AUDIO_LIMITER_ENABLED: return "Audio Limiter Feature Enabled"; case V4L2_CID_AUDIO_LIMITER_RELEASE_TIME: return "Audio Limiter Release Time"; case V4L2_CID_AUDIO_LIMITER_DEVIATION: return "Audio Limiter Deviation"; - case V4L2_CID_AUDIO_COMPRESSION_ENABLED: return "Audio Compression Feature Enabled"; + case V4L2_CID_AUDIO_COMPRESSION_ENABLED: return "Audio Compression Enabled"; case V4L2_CID_AUDIO_COMPRESSION_GAIN: return "Audio Compression Gain"; case V4L2_CID_AUDIO_COMPRESSION_THRESHOLD: return "Audio Compression Threshold"; case V4L2_CID_AUDIO_COMPRESSION_ATTACK_TIME: return "Audio Compression Attack Time"; @@ -588,24 +588,24 @@ const char *v4l2_ctrl_get_name(u32 id) case V4L2_CID_PILOT_TONE_ENABLED: return "Pilot Tone Feature Enabled"; case V4L2_CID_PILOT_TONE_DEVIATION: return "Pilot Tone Deviation"; case V4L2_CID_PILOT_TONE_FREQUENCY: return "Pilot Tone Frequency"; - case V4L2_CID_TUNE_PREEMPHASIS: return "Pre-emphasis settings"; + case V4L2_CID_TUNE_PREEMPHASIS: return "Pre-Emphasis"; case V4L2_CID_TUNE_POWER_LEVEL: return "Tune Power Level"; case V4L2_CID_TUNE_ANTENNA_CAPACITOR: return "Tune Antenna Capacitor"; /* Flash controls */ - case V4L2_CID_FLASH_CLASS: return "Flash controls"; - case V4L2_CID_FLASH_LED_MODE: return "LED mode"; - case V4L2_CID_FLASH_STROBE_SOURCE: return "Strobe source"; + case V4L2_CID_FLASH_CLASS: return "Flash Controls"; + case V4L2_CID_FLASH_LED_MODE: return "LED Mode"; + case V4L2_CID_FLASH_STROBE_SOURCE: return "Strobe Source"; case V4L2_CID_FLASH_STROBE: return "Strobe"; - case V4L2_CID_FLASH_STROBE_STOP: return "Stop strobe"; - case V4L2_CID_FLASH_STROBE_STATUS: return "Strobe status"; - case V4L2_CID_FLASH_TIMEOUT: return "Strobe timeout"; - case V4L2_CID_FLASH_INTENSITY: return "Intensity, flash mode"; - case V4L2_CID_FLASH_TORCH_INTENSITY: return "Intensity, torch mode"; - case V4L2_CID_FLASH_INDICATOR_INTENSITY: return "Intensity, indicator"; + case V4L2_CID_FLASH_STROBE_STOP: return "Stop Strobe"; + case V4L2_CID_FLASH_STROBE_STATUS: return "Strobe Status"; + case V4L2_CID_FLASH_TIMEOUT: return "Strobe Timeout"; + case V4L2_CID_FLASH_INTENSITY: return "Intensity, Flash Mode"; + case V4L2_CID_FLASH_TORCH_INTENSITY: return "Intensity, Torch Mode"; + case V4L2_CID_FLASH_INDICATOR_INTENSITY: return "Intensity, Indicator"; case V4L2_CID_FLASH_FAULT: return "Faults"; case V4L2_CID_FLASH_CHARGE: return "Charge"; - case V4L2_CID_FLASH_READY: return "Ready to strobe"; + case V4L2_CID_FLASH_READY: return "Ready to Strobe"; default: return NULL; From 3a9888f913595515893fa46aca70422aba8a3799 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 17 Jan 2012 03:28:51 -0300 Subject: [PATCH 591/820] [media] ds3000: using logical && instead of bitwise & The intent here was to test if the FE_HAS_LOCK was set. The current test is equivalent to "if (status) { ..." Signed-off-by: Dan Carpenter Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/ds3000.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/dvb/frontends/ds3000.c b/drivers/media/dvb/frontends/ds3000.c index 938777065de6..af65d013db11 100644 --- a/drivers/media/dvb/frontends/ds3000.c +++ b/drivers/media/dvb/frontends/ds3000.c @@ -1195,7 +1195,7 @@ static int ds3000_set_frontend(struct dvb_frontend *fe) for (i = 0; i < 30 ; i++) { ds3000_read_status(fe, &status); - if (status && FE_HAS_LOCK) + if (status & FE_HAS_LOCK) break; msleep(10); From 2bd43341217b6e8b75e382243328f458ac67fcbe Mon Sep 17 00:00:00 2001 From: Sachin Kamat Date: Tue, 17 Jan 2012 04:52:41 -0300 Subject: [PATCH 592/820] [media] s5p-mfc: Remove linux/version.h include from s5p_mfc.c Signed-off-by: Sachin Kamat Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-mfc/s5p_mfc.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/media/video/s5p-mfc/s5p_mfc.c b/drivers/media/video/s5p-mfc/s5p_mfc.c index 90cc05855ecb..83fe461af263 100644 --- a/drivers/media/video/s5p-mfc/s5p_mfc.c +++ b/drivers/media/video/s5p-mfc/s5p_mfc.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include #include From a2e2baad60065614c48fee6f55479d6809b0ce83 Mon Sep 17 00:00:00 2001 From: Sachin Kamat Date: Tue, 17 Jan 2012 04:55:25 -0300 Subject: [PATCH 593/820] [media] s5p-fimc: Remove linux/version.h include from fimc-mdevice.c Signed-off-by: Sachin Kamat Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/s5p-fimc/fimc-mdevice.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/media/video/s5p-fimc/fimc-mdevice.c b/drivers/media/video/s5p-fimc/fimc-mdevice.c index 615c862f0360..8ea4ee116e46 100644 --- a/drivers/media/video/s5p-fimc/fimc-mdevice.c +++ b/drivers/media/video/s5p-fimc/fimc-mdevice.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include From 36be126cb0ebe3000a65c1049f339a3e882a9a47 Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Sun, 8 Jan 2012 20:18:19 -0300 Subject: [PATCH 594/820] [media] as3645a: Fix compilation by including slab.h The as3645a driver calls kzalloc(). Include slab.h. Reported-by: Fredrik Lingvall Signed-off-by: Laurent Pinchart Acked-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/as3645a.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/video/as3645a.c b/drivers/media/video/as3645a.c index ec859a580651..f241702a0f36 100644 --- a/drivers/media/video/as3645a.c +++ b/drivers/media/video/as3645a.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include From 966f1212e1ac5fe3ddf04479d21488ddb36a2608 Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Mon, 16 Jan 2012 13:23:23 +0800 Subject: [PATCH 595/820] [libata] enable ata port async suspend This saves devices suspend/resume time. Tested system suspend/resume with SATA IDE/AHCI mode 3 times. Below is the time took for devices suspend/resume. SATA mode vanilla-kernel patched-kernel --------- --------------------- --------------------- IDE suspend: 0.744 suspend: 0.432 (0.716, 0.768, 0.748) (0.440, 0.428, 0.428) resume: 5.084 resume: 2.209 (5.100, 5.064, 5.088) (2.168, 2.232, 2.228) AHCI: suspend: 0.725 suspend: 0.449 (0.740, 0.708, 0.728) (0.456, 0.448, 0.444) resume: 2.556 resume: 1.896 (2.604, 2.492, 2.572) (1.932, 1.872, 1.884) Signed-off-by: Lin Ming Signed-off-by: Jeff Garzik --- drivers/ata/libata-transport.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/ata/libata-transport.c b/drivers/ata/libata-transport.c index 9a7f0ea565df..74aaee30e264 100644 --- a/drivers/ata/libata-transport.c +++ b/drivers/ata/libata-transport.c @@ -291,6 +291,7 @@ int ata_tport_add(struct device *parent, goto tport_err; } + device_enable_async_suspend(dev); pm_runtime_set_active(dev); pm_runtime_enable(dev); From 93272b132a72450dfc16f13d32223fe47aaf5061 Mon Sep 17 00:00:00 2001 From: Jerry Huang Date: Tue, 20 Dec 2011 14:50:27 +0800 Subject: [PATCH 596/820] [libata] sata_fsl: fix the controller operating mode Configure the FSL SATA controller to the preferred, enterprise mode. Signed-off-by: Yutaka Ando Signed-off-by: Jerry Huang CC: Jeff Garzik Signed-off-by: Jeff Garzik --- drivers/ata/sata_fsl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/ata/sata_fsl.c b/drivers/ata/sata_fsl.c index 5a2c95ba050a..0120b0d1e9a5 100644 --- a/drivers/ata/sata_fsl.c +++ b/drivers/ata/sata_fsl.c @@ -140,6 +140,7 @@ enum { */ HCONTROL_ONLINE_PHY_RST = (1 << 31), HCONTROL_FORCE_OFFLINE = (1 << 30), + HCONTROL_LEGACY = (1 << 28), HCONTROL_PARITY_PROT_MOD = (1 << 14), HCONTROL_DPATH_PARITY = (1 << 12), HCONTROL_SNOOP_ENABLE = (1 << 10), @@ -1223,6 +1224,10 @@ static int sata_fsl_init_controller(struct ata_host *host) * part of the port_start() callback */ + /* sata controller to operate in enterprise mode */ + temp = ioread32(hcr_base + HCONTROL); + iowrite32(temp & ~HCONTROL_LEGACY, hcr_base + HCONTROL); + /* ack. any pending IRQs for this controller/port */ temp = ioread32(hcr_base + HSTATUS); if (temp & 0x3F) @@ -1421,6 +1426,12 @@ static int sata_fsl_resume(struct platform_device *op) /* Recovery the CHBA register in host controller cmd register set */ iowrite32(pp->cmdslot_paddr & 0xffffffff, hcr_base + CHBA); + iowrite32((ioread32(hcr_base + HCONTROL) + | HCONTROL_ONLINE_PHY_RST + | HCONTROL_SNOOP_ENABLE + | HCONTROL_PMP_ATTACHED), + hcr_base + HCONTROL); + ata_host_resume(host); return 0; } From 81b0287d341535ac722de891b19f7c49212ac91c Mon Sep 17 00:00:00 2001 From: Sonic Zhang Date: Wed, 4 Jan 2012 14:06:51 +0800 Subject: [PATCH 597/820] [libata] pata_bf54x: Support sg list in bmdma transfer. BF54x on-chip ATAPI controller allows maximum 0x1fffe bytes to be transfered in one ATAPI transfer. So, set the max sg_tablesize to 4. Signed-off-by: Sonic Zhang Signed-off-by: Jeff Garzik --- drivers/ata/pata_bf54x.c | 175 ++++++++++++++++++++------------------- 1 file changed, 92 insertions(+), 83 deletions(-) diff --git a/drivers/ata/pata_bf54x.c b/drivers/ata/pata_bf54x.c index d6a4677fdf71..1e65842e2ca7 100644 --- a/drivers/ata/pata_bf54x.c +++ b/drivers/ata/pata_bf54x.c @@ -251,6 +251,8 @@ static const u32 udma_tenvmin = 20; static const u32 udma_tackmin = 20; static const u32 udma_tssmin = 50; +#define BFIN_MAX_SG_SEGMENTS 4 + /** * * Function: num_clocks_min @@ -829,30 +831,74 @@ static void bfin_set_devctl(struct ata_port *ap, u8 ctl) static void bfin_bmdma_setup(struct ata_queued_cmd *qc) { - unsigned short config = WDSIZE_16; + struct ata_port *ap = qc->ap; + struct dma_desc_array *dma_desc_cpu = (struct dma_desc_array *)ap->bmdma_prd; + void __iomem *base = (void __iomem *)ap->ioaddr.ctl_addr; + unsigned short config = DMAFLOW_ARRAY | NDSIZE_5 | RESTART | WDSIZE_16 | DMAEN; struct scatterlist *sg; unsigned int si; + unsigned int channel; + unsigned int dir; + unsigned int size = 0; dev_dbg(qc->ap->dev, "in atapi dma setup\n"); /* Program the ATA_CTRL register with dir */ if (qc->tf.flags & ATA_TFLAG_WRITE) { - /* fill the ATAPI DMA controller */ - set_dma_config(CH_ATAPI_TX, config); - set_dma_x_modify(CH_ATAPI_TX, 2); - for_each_sg(qc->sg, sg, qc->n_elem, si) { - set_dma_start_addr(CH_ATAPI_TX, sg_dma_address(sg)); - set_dma_x_count(CH_ATAPI_TX, sg_dma_len(sg) >> 1); - } + channel = CH_ATAPI_TX; + dir = DMA_TO_DEVICE; } else { + channel = CH_ATAPI_RX; + dir = DMA_FROM_DEVICE; config |= WNR; - /* fill the ATAPI DMA controller */ - set_dma_config(CH_ATAPI_RX, config); - set_dma_x_modify(CH_ATAPI_RX, 2); - for_each_sg(qc->sg, sg, qc->n_elem, si) { - set_dma_start_addr(CH_ATAPI_RX, sg_dma_address(sg)); - set_dma_x_count(CH_ATAPI_RX, sg_dma_len(sg) >> 1); - } } + + dma_map_sg(ap->dev, qc->sg, qc->n_elem, dir); + + /* fill the ATAPI DMA controller */ + for_each_sg(qc->sg, sg, qc->n_elem, si) { + dma_desc_cpu[si].start_addr = sg_dma_address(sg); + dma_desc_cpu[si].cfg = config; + dma_desc_cpu[si].x_count = sg_dma_len(sg) >> 1; + dma_desc_cpu[si].x_modify = 2; + size += sg_dma_len(sg); + } + + /* Set the last descriptor to stop mode */ + dma_desc_cpu[qc->n_elem - 1].cfg &= ~(DMAFLOW | NDSIZE); + + flush_dcache_range((unsigned int)dma_desc_cpu, + (unsigned int)dma_desc_cpu + + qc->n_elem * sizeof(struct dma_desc_array)); + + /* Enable ATA DMA operation*/ + set_dma_curr_desc_addr(channel, (unsigned long *)ap->bmdma_prd_dma); + set_dma_x_count(channel, 0); + set_dma_x_modify(channel, 0); + set_dma_config(channel, config); + + SSYNC(); + + /* Send ATA DMA command */ + bfin_exec_command(ap, &qc->tf); + + if (qc->tf.flags & ATA_TFLAG_WRITE) { + /* set ATA DMA write direction */ + ATAPI_SET_CONTROL(base, (ATAPI_GET_CONTROL(base) + | XFER_DIR)); + } else { + /* set ATA DMA read direction */ + ATAPI_SET_CONTROL(base, (ATAPI_GET_CONTROL(base) + & ~XFER_DIR)); + } + + /* Reset all transfer count */ + ATAPI_SET_CONTROL(base, ATAPI_GET_CONTROL(base) | TFRCNT_RST); + + /* Set ATAPI state machine contorl in terminate sequence */ + ATAPI_SET_CONTROL(base, ATAPI_GET_CONTROL(base) | END_ON_TERM); + + /* Set transfer length to the total size of sg buffers */ + ATAPI_SET_XFER_LEN(base, size >> 1); } /** @@ -866,59 +912,13 @@ static void bfin_bmdma_start(struct ata_queued_cmd *qc) { struct ata_port *ap = qc->ap; void __iomem *base = (void __iomem *)ap->ioaddr.ctl_addr; - struct scatterlist *sg; - unsigned int si; dev_dbg(qc->ap->dev, "in atapi dma start\n"); + if (!(ap->udma_mask || ap->mwdma_mask)) return; - /* start ATAPI DMA controller*/ - if (qc->tf.flags & ATA_TFLAG_WRITE) { - /* - * On blackfin arch, uncacheable memory is not - * allocated with flag GFP_DMA. DMA buffer from - * common kenel code should be flushed if WB - * data cache is enabled. Otherwise, this loop - * is an empty loop and optimized out. - */ - for_each_sg(qc->sg, sg, qc->n_elem, si) { - flush_dcache_range(sg_dma_address(sg), - sg_dma_address(sg) + sg_dma_len(sg)); - } - enable_dma(CH_ATAPI_TX); - dev_dbg(qc->ap->dev, "enable udma write\n"); - - /* Send ATA DMA write command */ - bfin_exec_command(ap, &qc->tf); - - /* set ATA DMA write direction */ - ATAPI_SET_CONTROL(base, (ATAPI_GET_CONTROL(base) - | XFER_DIR)); - } else { - enable_dma(CH_ATAPI_RX); - dev_dbg(qc->ap->dev, "enable udma read\n"); - - /* Send ATA DMA read command */ - bfin_exec_command(ap, &qc->tf); - - /* set ATA DMA read direction */ - ATAPI_SET_CONTROL(base, (ATAPI_GET_CONTROL(base) - & ~XFER_DIR)); - } - - /* Reset all transfer count */ - ATAPI_SET_CONTROL(base, ATAPI_GET_CONTROL(base) | TFRCNT_RST); - - /* Set ATAPI state machine contorl in terminate sequence */ - ATAPI_SET_CONTROL(base, ATAPI_GET_CONTROL(base) | END_ON_TERM); - - /* Set transfer length to buffer len */ - for_each_sg(qc->sg, sg, qc->n_elem, si) { - ATAPI_SET_XFER_LEN(base, (sg_dma_len(sg) >> 1)); - } - - /* Enable ATA DMA operation*/ + /* start ATAPI transfer*/ if (ap->udma_mask) ATAPI_SET_CONTROL(base, ATAPI_GET_CONTROL(base) | ULTRA_START); @@ -935,34 +935,23 @@ static void bfin_bmdma_start(struct ata_queued_cmd *qc) static void bfin_bmdma_stop(struct ata_queued_cmd *qc) { struct ata_port *ap = qc->ap; - struct scatterlist *sg; - unsigned int si; + unsigned int dir; dev_dbg(qc->ap->dev, "in atapi dma stop\n"); + if (!(ap->udma_mask || ap->mwdma_mask)) return; /* stop ATAPI DMA controller*/ - if (qc->tf.flags & ATA_TFLAG_WRITE) + if (qc->tf.flags & ATA_TFLAG_WRITE) { + dir = DMA_TO_DEVICE; disable_dma(CH_ATAPI_TX); - else { + } else { + dir = DMA_FROM_DEVICE; disable_dma(CH_ATAPI_RX); - if (ap->hsm_task_state & HSM_ST_LAST) { - /* - * On blackfin arch, uncacheable memory is not - * allocated with flag GFP_DMA. DMA buffer from - * common kenel code should be invalidated if - * data cache is enabled. Otherwise, this loop - * is an empty loop and optimized out. - */ - for_each_sg(qc->sg, sg, qc->n_elem, si) { - invalidate_dcache_range( - sg_dma_address(sg), - sg_dma_address(sg) - + sg_dma_len(sg)); - } - } } + + dma_unmap_sg(ap->dev, qc->sg, qc->n_elem, dir); } /** @@ -1260,6 +1249,11 @@ static void bfin_port_stop(struct ata_port *ap) { dev_dbg(ap->dev, "in atapi port stop\n"); if (ap->udma_mask != 0 || ap->mwdma_mask != 0) { + dma_free_coherent(ap->dev, + BFIN_MAX_SG_SEGMENTS * sizeof(struct dma_desc_array), + ap->bmdma_prd, + ap->bmdma_prd_dma); + free_dma(CH_ATAPI_RX); free_dma(CH_ATAPI_TX); } @@ -1271,14 +1265,29 @@ static int bfin_port_start(struct ata_port *ap) if (!(ap->udma_mask || ap->mwdma_mask)) return 0; + ap->bmdma_prd = dma_alloc_coherent(ap->dev, + BFIN_MAX_SG_SEGMENTS * sizeof(struct dma_desc_array), + &ap->bmdma_prd_dma, + GFP_KERNEL); + + if (ap->bmdma_prd == NULL) { + dev_info(ap->dev, "Unable to allocate DMA descriptor array.\n"); + goto out; + } + if (request_dma(CH_ATAPI_RX, "BFIN ATAPI RX DMA") >= 0) { if (request_dma(CH_ATAPI_TX, "BFIN ATAPI TX DMA") >= 0) return 0; free_dma(CH_ATAPI_RX); + dma_free_coherent(ap->dev, + BFIN_MAX_SG_SEGMENTS * sizeof(struct dma_desc_array), + ap->bmdma_prd, + ap->bmdma_prd_dma); } +out: ap->udma_mask = 0; ap->mwdma_mask = 0; dev_err(ap->dev, "Unable to request ATAPI DMA!" @@ -1400,7 +1409,7 @@ static irqreturn_t bfin_ata_interrupt(int irq, void *dev_instance) static struct scsi_host_template bfin_sht = { ATA_BASE_SHT(DRV_NAME), - .sg_tablesize = SG_NONE, + .sg_tablesize = BFIN_MAX_SG_SEGMENTS, .dma_boundary = ATA_DMA_BOUNDARY, }; From 3a25179e068b2fae8df7a7ab63a2a5ed71a2e927 Mon Sep 17 00:00:00 2001 From: Vladimir LAVALLADE Date: Sun, 8 Jan 2012 13:50:13 +0100 Subject: [PATCH 598/820] [libata] add DVRTD08A and DVR-215 to NOSETXFER device quirk list Signed-off-by: Jeff Garzik --- drivers/ata/libata-core.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 11c9aea4f4f7..c06e0ec11556 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -4125,6 +4125,8 @@ static const struct ata_blacklist_entry ata_device_blacklist [] = { * device and controller are SATA. */ { "PIONEER DVD-RW DVRTD08", NULL, ATA_HORKAGE_NOSETXFER }, + { "PIONEER DVD-RW DVRTD08A", NULL, ATA_HORKAGE_NOSETXFER }, + { "PIONEER DVD-RW DVR-215", NULL, ATA_HORKAGE_NOSETXFER }, { "PIONEER DVD-RW DVR-212D", NULL, ATA_HORKAGE_NOSETXFER }, { "PIONEER DVD-RW DVR-216D", NULL, ATA_HORKAGE_NOSETXFER }, From b73fa4630e63e4d23407181c2bc0a54777bd0ce8 Mon Sep 17 00:00:00 2001 From: Benjamin Larsson Date: Sun, 8 Jan 2012 00:39:10 +0100 Subject: [PATCH 599/820] [libata] ata_piix: Add Toshiba Satellite Pro A120 to the quirks list due to broken suspend functionality. Signed-off-by: Benjamin Larsson Signed-off-by: Jeff Garzik --- drivers/ata/ata_piix.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/ata/ata_piix.c b/drivers/ata/ata_piix.c index 69ac373c72ab..fdf27b9fce43 100644 --- a/drivers/ata/ata_piix.c +++ b/drivers/ata/ata_piix.c @@ -1116,6 +1116,13 @@ static int piix_broken_suspend(void) DMI_MATCH(DMI_PRODUCT_NAME, "SATELLITE U205"), }, }, + { + .ident = "Satellite Pro A120", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "TOSHIBA"), + DMI_MATCH(DMI_PRODUCT_NAME, "Satellite Pro A120"), + }, + }, { .ident = "Portege M500", .matches = { From 6015ff103133c7e50a753c198c69bcabc3a5e3b0 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 18 Jan 2012 01:51:22 +0000 Subject: [PATCH 600/820] x86-32: Fix build failure with AUDIT=y, AUDITSYSCALL=n JONGMAN HEO reports: With current linus git (commit a25a2b84), I got following build error, arch/x86/kernel/vm86_32.c: In function 'do_sys_vm86': arch/x86/kernel/vm86_32.c:340: error: implicit declaration of function '__audit_syscall_exit' make[3]: *** [arch/x86/kernel/vm86_32.o] Error 1 OK, I can reproduce it (32bit allmodconfig with AUDIT=y, AUDITSYSCALL=n) It's due to commit d7e7528bcd45: "Audit: push audit success and retcode into arch ptrace.h". Reported-by: JONGMAN HEO Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86/kernel/vm86_32.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index af17e1c966dc..b466cab5ba15 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -336,8 +336,10 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk mark_screen_rdonly(tsk->mm); /*call __audit_syscall_exit since we do not exit via the normal paths */ +#ifdef CONFIG_AUDITSYSCALL if (unlikely(current->audit_context)) __audit_syscall_exit(1, 0); +#endif __asm__ __volatile__( "movl %0,%%esp\n\t" From f5fffcee27c09143ba80e5257dbd1f381d86342f Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 17 Jan 2012 13:49:17 -0500 Subject: [PATCH 601/820] cifs: better instrumentation for coalesce_t2 When coalesce_t2 returns an error, have it throw a cFYI message that explains the reason. Also rename some variables to clarify what they represent. Reported-and-Tested-by: Konstantinos Skarlatos Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 90 ++++++++++++++++++++++++++++------------------- 1 file changed, 53 insertions(+), 37 deletions(-) diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 4666780f315d..5cc15856e4ad 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -225,74 +225,90 @@ static int check2ndT2(struct smb_hdr *pSMB) static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) { - struct smb_t2_rsp *pSMB2 = (struct smb_t2_rsp *)psecond; + struct smb_t2_rsp *pSMBs = (struct smb_t2_rsp *)psecond; struct smb_t2_rsp *pSMBt = (struct smb_t2_rsp *)pTargetSMB; - char *data_area_of_target; - char *data_area_of_buf2; + char *data_area_of_tgt; + char *data_area_of_src; int remaining; - unsigned int byte_count, total_in_buf; - __u16 total_data_size, total_in_buf2; + unsigned int byte_count, total_in_tgt; + __u16 tgt_total_cnt, src_total_cnt, total_in_src; - total_data_size = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); + src_total_cnt = get_unaligned_le16(&pSMBs->t2_rsp.TotalDataCount); + tgt_total_cnt = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); - if (total_data_size != - get_unaligned_le16(&pSMB2->t2_rsp.TotalDataCount)) - cFYI(1, "total data size of primary and secondary t2 differ"); + if (tgt_total_cnt != src_total_cnt) + cFYI(1, "total data count of primary and secondary t2 differ " + "source=%hu target=%hu", src_total_cnt, tgt_total_cnt); - total_in_buf = get_unaligned_le16(&pSMBt->t2_rsp.DataCount); + total_in_tgt = get_unaligned_le16(&pSMBt->t2_rsp.DataCount); - remaining = total_data_size - total_in_buf; + remaining = tgt_total_cnt - total_in_tgt; - if (remaining < 0) + if (remaining < 0) { + cFYI(1, "Server sent too much data. tgt_total_cnt=%hu " + "total_in_tgt=%hu", tgt_total_cnt, total_in_tgt); return -EPROTO; - - if (remaining == 0) /* nothing to do, ignore */ - return 0; - - total_in_buf2 = get_unaligned_le16(&pSMB2->t2_rsp.DataCount); - if (remaining < total_in_buf2) { - cFYI(1, "transact2 2nd response contains too much data"); } + if (remaining == 0) { + /* nothing to do, ignore */ + cFYI(1, "no more data remains"); + return 0; + } + + total_in_src = get_unaligned_le16(&pSMBs->t2_rsp.DataCount); + if (remaining < total_in_src) + cFYI(1, "transact2 2nd response contains too much data"); + /* find end of first SMB data area */ - data_area_of_target = (char *)&pSMBt->hdr.Protocol + + data_area_of_tgt = (char *)&pSMBt->hdr.Protocol + get_unaligned_le16(&pSMBt->t2_rsp.DataOffset); + /* validate target area */ + data_area_of_src = (char *)&pSMBs->hdr.Protocol + + get_unaligned_le16(&pSMBs->t2_rsp.DataOffset); - data_area_of_buf2 = (char *)&pSMB2->hdr.Protocol + - get_unaligned_le16(&pSMB2->t2_rsp.DataOffset); + data_area_of_tgt += total_in_tgt; - data_area_of_target += total_in_buf; - - /* copy second buffer into end of first buffer */ - total_in_buf += total_in_buf2; + total_in_tgt += total_in_src; /* is the result too big for the field? */ - if (total_in_buf > USHRT_MAX) + if (total_in_tgt > USHRT_MAX) { + cFYI(1, "coalesced DataCount too large (%u)", total_in_tgt); return -EPROTO; - put_unaligned_le16(total_in_buf, &pSMBt->t2_rsp.DataCount); + } + put_unaligned_le16(total_in_tgt, &pSMBt->t2_rsp.DataCount); /* fix up the BCC */ byte_count = get_bcc(pTargetSMB); - byte_count += total_in_buf2; + byte_count += total_in_src; /* is the result too big for the field? */ - if (byte_count > USHRT_MAX) + if (byte_count > USHRT_MAX) { + cFYI(1, "coalesced BCC too large (%u)", byte_count); return -EPROTO; + } put_bcc(byte_count, pTargetSMB); byte_count = be32_to_cpu(pTargetSMB->smb_buf_length); - byte_count += total_in_buf2; + byte_count += total_in_src; /* don't allow buffer to overflow */ - if (byte_count > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) + if (byte_count > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { + cFYI(1, "coalesced BCC exceeds buffer size (%u)", byte_count); return -ENOBUFS; + } pTargetSMB->smb_buf_length = cpu_to_be32(byte_count); - memcpy(data_area_of_target, data_area_of_buf2, total_in_buf2); + /* copy second buffer into end of first buffer */ + memcpy(data_area_of_tgt, data_area_of_src, total_in_src); - if (remaining == total_in_buf2) { - cFYI(1, "found the last secondary response"); - return 0; /* we are done */ - } else /* more responses to go */ + if (remaining != total_in_src) { + /* more responses to go */ + cFYI(1, "waiting for more secondary responses"); return 1; + } + + /* we are done */ + cFYI(1, "found the last secondary response"); + return 0; } static void From ce91acb3acae26f4163c5a6f1f695d1a1e8d9009 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 17 Jan 2012 16:08:51 -0500 Subject: [PATCH 602/820] cifs: lower default wsize when unix extensions are not used We've had some reports of servers (namely, the Solaris in-kernel CIFS server) that don't deal properly with writes that are "too large" even though they set CAP_LARGE_WRITE_ANDX. Change the default to better mirror what windows clients do. Cc: stable@vger.kernel.org Cc: Pavel Shilovsky Reported-by: Nick Davis Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 23 +++++++++++++++++++---- 1 file changed, 19 insertions(+), 4 deletions(-) diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 5cc15856e4ad..a66dcb52988c 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2930,18 +2930,33 @@ void cifs_setup_cifs_sb(struct smb_vol *pvolume_info, #define CIFS_DEFAULT_IOSIZE (1024 * 1024) /* - * Windows only supports a max of 60k reads. Default to that when posix - * extensions aren't in force. + * Windows only supports a max of 60kb reads and 65535 byte writes. Default to + * those values when posix extensions aren't in force. In actuality here, we + * use 65536 to allow for a write that is a multiple of 4k. Most servers seem + * to be ok with the extra byte even though Windows doesn't send writes that + * are that large. + * + * Citation: + * + * http://blogs.msdn.com/b/openspecification/archive/2009/04/10/smb-maximum-transmit-buffer-size-and-performance-tuning.aspx */ #define CIFS_DEFAULT_NON_POSIX_RSIZE (60 * 1024) +#define CIFS_DEFAULT_NON_POSIX_WSIZE (65536) static unsigned int cifs_negotiate_wsize(struct cifs_tcon *tcon, struct smb_vol *pvolume_info) { __u64 unix_cap = le64_to_cpu(tcon->fsUnixInfo.Capability); struct TCP_Server_Info *server = tcon->ses->server; - unsigned int wsize = pvolume_info->wsize ? pvolume_info->wsize : - CIFS_DEFAULT_IOSIZE; + unsigned int wsize; + + /* start with specified wsize, or default */ + if (pvolume_info->wsize) + wsize = pvolume_info->wsize; + else if (tcon->unix_ext && (unix_cap & CIFS_UNIX_LARGE_WRITE_CAP)) + wsize = CIFS_DEFAULT_IOSIZE; + else + wsize = CIFS_DEFAULT_NON_POSIX_WSIZE; /* can server support 24-bit write sizes? (via UNIX extensions) */ if (!tcon->unix_ext || !(unix_cap & CIFS_UNIX_LARGE_WRITE_CAP)) From 9f6ed2ca257fa8650b876377833e6f14e272848b Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 17 Jan 2012 16:09:11 -0500 Subject: [PATCH 603/820] keys: add a "logon" key type For CIFS, we want to be able to store NTLM credentials (aka username and password) in the keyring. We do not, however want to allow users to fetch those keys back out of the keyring since that would be a security risk. Unfortunately, due to the nuances of key permission bits, it's not possible to do this. We need to grant search permissions so the kernel can find these keys, but that also implies permissions to read the payload. Resolve this by adding a new key_type. This key type is essentially the same as key_type_user, but does not define a .read op. This prevents the payload from ever being visible from userspace. This key type also vets the description to ensure that it's "qualified" by checking to ensure that it has a ':' in it that is preceded by other characters. Acked-by: David Howells Signed-off-by: Jeff Layton Signed-off-by: Steve French --- include/keys/user-type.h | 3 ++- security/keys/internal.h | 1 + security/keys/key.c | 1 + security/keys/user_defined.c | 37 ++++++++++++++++++++++++++++++++++++ 4 files changed, 41 insertions(+), 1 deletion(-) diff --git a/include/keys/user-type.h b/include/keys/user-type.h index c37c34275a44..bc9ec1d7698c 100644 --- a/include/keys/user-type.h +++ b/include/keys/user-type.h @@ -17,7 +17,7 @@ /*****************************************************************************/ /* - * the payload for a key of type "user" + * the payload for a key of type "user" or "logon" * - once filled in and attached to a key: * - the payload struct is invariant may not be changed, only replaced * - the payload must be read with RCU procedures or with the key semaphore @@ -33,6 +33,7 @@ struct user_key_payload { }; extern struct key_type key_type_user; +extern struct key_type key_type_logon; extern int user_instantiate(struct key *key, const void *data, size_t datalen); extern int user_update(struct key *key, const void *data, size_t datalen); diff --git a/security/keys/internal.h b/security/keys/internal.h index c7a7caec4830..65647f825584 100644 --- a/security/keys/internal.h +++ b/security/keys/internal.h @@ -33,6 +33,7 @@ extern struct key_type key_type_dead; extern struct key_type key_type_user; +extern struct key_type key_type_logon; /*****************************************************************************/ /* diff --git a/security/keys/key.c b/security/keys/key.c index 4f64c7267afb..7ada8019be1f 100644 --- a/security/keys/key.c +++ b/security/keys/key.c @@ -999,6 +999,7 @@ void __init key_init(void) list_add_tail(&key_type_keyring.link, &key_types_list); list_add_tail(&key_type_dead.link, &key_types_list); list_add_tail(&key_type_user.link, &key_types_list); + list_add_tail(&key_type_logon.link, &key_types_list); /* record the root user tracking */ rb_link_node(&root_key_user.node, diff --git a/security/keys/user_defined.c b/security/keys/user_defined.c index 69ff52c08e97..6e1a6276649f 100644 --- a/security/keys/user_defined.c +++ b/security/keys/user_defined.c @@ -18,6 +18,8 @@ #include #include "internal.h" +static int logon_vet_description(const char *desc); + /* * user defined keys take an arbitrary string as the description and an * arbitrary blob of data as the payload @@ -35,6 +37,24 @@ struct key_type key_type_user = { EXPORT_SYMBOL_GPL(key_type_user); +/* + * This key type is essentially the same as key_type_user, but it does + * not define a .read op. This is suitable for storing username and + * password pairs in the keyring that you do not want to be readable + * from userspace. + */ +struct key_type key_type_logon = { + .name = "logon", + .instantiate = user_instantiate, + .update = user_update, + .match = user_match, + .revoke = user_revoke, + .destroy = user_destroy, + .describe = user_describe, + .vet_description = logon_vet_description, +}; +EXPORT_SYMBOL_GPL(key_type_logon); + /* * instantiate a user defined key */ @@ -189,3 +209,20 @@ long user_read(const struct key *key, char __user *buffer, size_t buflen) } EXPORT_SYMBOL_GPL(user_read); + +/* Vet the description for a "logon" key */ +static int logon_vet_description(const char *desc) +{ + char *p; + + /* require a "qualified" description string */ + p = strchr(desc, ':'); + if (!p) + return -EINVAL; + + /* also reject description with ':' as first char */ + if (p == desc) + return -EINVAL; + + return 0; +} From 04febabcf55beeffb8794a0d8c539e571bd2ae29 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 17 Jan 2012 16:09:15 -0500 Subject: [PATCH 604/820] cifs: sanitize username handling Currently, it's not very clear whether you're allowed to have a NULL vol->username or ses->user_name. Some places check for it and some don't. Make it clear that a NULL pointer is OK in these fields, and ensure that all the callers check for that. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifs_spnego.c | 10 +++++++--- fs/cifs/cifsencrypt.c | 11 ++++++++--- fs/cifs/connect.c | 19 ++++++++++++------- 3 files changed, 27 insertions(+), 13 deletions(-) diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index 2272fd5fe5b7..e622863b292f 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -113,9 +113,11 @@ cifs_get_spnego_key(struct cifs_ses *sesInfo) MAX_MECH_STR_LEN + UID_KEY_LEN + (sizeof(uid_t) * 2) + CREDUID_KEY_LEN + (sizeof(uid_t) * 2) + - USER_KEY_LEN + strlen(sesInfo->user_name) + PID_KEY_LEN + (sizeof(pid_t) * 2) + 1; + if (sesInfo->user_name) + desc_len += USER_KEY_LEN + strlen(sesInfo->user_name); + spnego_key = ERR_PTR(-ENOMEM); description = kzalloc(desc_len, GFP_KERNEL); if (description == NULL) @@ -152,8 +154,10 @@ cifs_get_spnego_key(struct cifs_ses *sesInfo) dp = description + strlen(description); sprintf(dp, ";creduid=0x%x", sesInfo->cred_uid); - dp = description + strlen(description); - sprintf(dp, ";user=%s", sesInfo->user_name); + if (sesInfo->user_name) { + dp = description + strlen(description); + sprintf(dp, ";user=%s", sesInfo->user_name); + } dp = description + strlen(description); sprintf(dp, ";pid=0x%x", current->pid); diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index 5d9b9acc5fce..bce99e6a4950 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -420,15 +420,20 @@ static int calc_ntlmv2_hash(struct cifs_ses *ses, char *ntlmv2_hash, } /* convert ses->user_name to unicode and uppercase */ - len = strlen(ses->user_name); + len = ses->user_name ? strlen(ses->user_name) : 0; user = kmalloc(2 + (len * 2), GFP_KERNEL); if (user == NULL) { cERROR(1, "calc_ntlmv2_hash: user mem alloc failure\n"); rc = -ENOMEM; return rc; } - len = cifs_strtoUCS((__le16 *)user, ses->user_name, len, nls_cp); - UniStrupr(user); + + if (len) { + len = cifs_strtoUCS((__le16 *)user, ses->user_name, len, nls_cp); + UniStrupr(user); + } else { + memset(user, '\0', 2); + } rc = crypto_shash_update(&ses->server->secmech.sdeschmacmd5->shash, (char *)user, 2 * len); diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index a66dcb52988c..b952a21e917b 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1997,10 +1997,16 @@ static int match_session(struct cifs_ses *ses, struct smb_vol *vol) return 0; break; default: + /* NULL username means anonymous session */ + if (ses->user_name == NULL) { + if (!vol->nullauth) + return 0; + break; + } + /* anything else takes username/password */ - if (ses->user_name == NULL) - return 0; - if (strncmp(ses->user_name, vol->username, + if (strncmp(ses->user_name, + vol->username ? vol->username : "", MAX_USERNAME_SIZE)) return 0; if (strlen(vol->username) != 0 && @@ -3167,10 +3173,9 @@ cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data, return -EINVAL; if (volume_info->nullauth) { - cFYI(1, "null user"); - volume_info->username = kzalloc(1, GFP_KERNEL); - if (volume_info->username == NULL) - return -ENOMEM; + cFYI(1, "Anonymous login"); + kfree(volume_info->username); + volume_info->username = NULL; } else if (volume_info->username) { /* BB fixme parse for domain name here */ cFYI(1, "Username: %s", volume_info->username); From 8a8798a5ff90977d6459ce1d657cf8fe13a51e97 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 17 Jan 2012 16:09:15 -0500 Subject: [PATCH 605/820] cifs: fetch credentials out of keyring for non-krb5 auth multiuser mounts Fix up multiuser mounts to set the secType and set the username and password from the key payload in the vol info for non-krb5 auth types. Look for a key of type "secret" with a description of "cifs:a:" or "cifs:d:". If that's found, then scrape the username and password out of the key payload and use that to create a new user session. Finally, don't have the code enforce krb5 auth on multiuser mounts, but do require a kernel with keys support. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 175 +++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 165 insertions(+), 10 deletions(-) diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index b952a21e917b..28f23c03da53 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include "cifspdu.h" #include "cifsglob.h" @@ -1594,11 +1595,14 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, } } - if (vol->multiuser && !(vol->secFlg & CIFSSEC_MAY_KRB5)) { - cERROR(1, "Multiuser mounts currently require krb5 " - "authentication!"); +#ifndef CONFIG_KEYS + /* Muliuser mounts require CONFIG_KEYS support */ + if (vol->multiuser) { + cERROR(1, "Multiuser mounts require kernels with " + "CONFIG_KEYS enabled."); goto cifs_parse_mount_err; } +#endif if (vol->UNCip == NULL) vol->UNCip = &vol->UNC[2]; @@ -2061,6 +2065,132 @@ cifs_put_smb_ses(struct cifs_ses *ses) cifs_put_tcp_session(server); } +#ifdef CONFIG_KEYS + +/* strlen("cifs:a:") + INET6_ADDRSTRLEN + 1 */ +#define CIFSCREDS_DESC_SIZE (7 + INET6_ADDRSTRLEN + 1) + +/* Populate username and pw fields from keyring if possible */ +static int +cifs_set_cifscreds(struct smb_vol *vol, struct cifs_ses *ses) +{ + int rc = 0; + char *desc, *delim, *payload; + ssize_t len; + struct key *key; + struct TCP_Server_Info *server = ses->server; + struct sockaddr_in *sa; + struct sockaddr_in6 *sa6; + struct user_key_payload *upayload; + + desc = kmalloc(CIFSCREDS_DESC_SIZE, GFP_KERNEL); + if (!desc) + return -ENOMEM; + + /* try to find an address key first */ + switch (server->dstaddr.ss_family) { + case AF_INET: + sa = (struct sockaddr_in *)&server->dstaddr; + sprintf(desc, "cifs:a:%pI4", &sa->sin_addr.s_addr); + break; + case AF_INET6: + sa6 = (struct sockaddr_in6 *)&server->dstaddr; + sprintf(desc, "cifs:a:%pI6c", &sa6->sin6_addr.s6_addr); + break; + default: + cFYI(1, "Bad ss_family (%hu)", server->dstaddr.ss_family); + rc = -EINVAL; + goto out_err; + } + + cFYI(1, "%s: desc=%s", __func__, desc); + key = request_key(&key_type_logon, desc, ""); + if (IS_ERR(key)) { + if (!ses->domainName) { + cFYI(1, "domainName is NULL"); + rc = PTR_ERR(key); + goto out_err; + } + + /* didn't work, try to find a domain key */ + sprintf(desc, "cifs:d:%s", ses->domainName); + cFYI(1, "%s: desc=%s", __func__, desc); + key = request_key(&key_type_logon, desc, ""); + if (IS_ERR(key)) { + rc = PTR_ERR(key); + goto out_err; + } + } + + down_read(&key->sem); + upayload = key->payload.data; + if (IS_ERR_OR_NULL(upayload)) { + rc = PTR_ERR(key); + goto out_key_put; + } + + /* find first : in payload */ + payload = (char *)upayload->data; + delim = strnchr(payload, upayload->datalen, ':'); + cFYI(1, "payload=%s", payload); + if (!delim) { + cFYI(1, "Unable to find ':' in payload (datalen=%d)", + upayload->datalen); + rc = -EINVAL; + goto out_key_put; + } + + len = delim - payload; + if (len > MAX_USERNAME_SIZE || len <= 0) { + cFYI(1, "Bad value from username search (len=%ld)", len); + rc = -EINVAL; + goto out_key_put; + } + + vol->username = kstrndup(payload, len, GFP_KERNEL); + if (!vol->username) { + cFYI(1, "Unable to allocate %ld bytes for username", len); + rc = -ENOMEM; + goto out_key_put; + } + cFYI(1, "%s: username=%s", __func__, vol->username); + + len = key->datalen - (len + 1); + if (len > MAX_PASSWORD_SIZE || len <= 0) { + cFYI(1, "Bad len for password search (len=%ld)", len); + rc = -EINVAL; + kfree(vol->username); + vol->username = NULL; + goto out_key_put; + } + + ++delim; + vol->password = kstrndup(delim, len, GFP_KERNEL); + if (!vol->password) { + cFYI(1, "Unable to allocate %ld bytes for password", len); + rc = -ENOMEM; + kfree(vol->username); + vol->username = NULL; + goto out_key_put; + } + +out_key_put: + up_read(&key->sem); + key_put(key); +out_err: + kfree(desc); + cFYI(1, "%s: returning %d", __func__, rc); + return rc; +} +#else /* ! CONFIG_KEYS */ +static inline int +cifs_set_cifscreds(struct smb_vol *vol __attribute__((unused)), + struct cifs_ses *ses __attribute__((unused))) +{ + return -ENOSYS; +} +#endif /* CONFIG_KEYS */ + static bool warned_on_ntlm; /* globals init to false automatically */ static struct cifs_ses * @@ -3693,16 +3823,38 @@ int cifs_setup_session(unsigned int xid, struct cifs_ses *ses, return rc; } +static int +cifs_set_vol_auth(struct smb_vol *vol, struct cifs_ses *ses) +{ + switch (ses->server->secType) { + case Kerberos: + vol->secFlg = CIFSSEC_MUST_KRB5; + return 0; + case NTLMv2: + vol->secFlg = CIFSSEC_MUST_NTLMV2; + break; + case NTLM: + vol->secFlg = CIFSSEC_MUST_NTLM; + break; + case RawNTLMSSP: + vol->secFlg = CIFSSEC_MUST_NTLMSSP; + break; + case LANMAN: + vol->secFlg = CIFSSEC_MUST_LANMAN; + break; + } + + return cifs_set_cifscreds(vol, ses); +} + static struct cifs_tcon * cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) { + int rc; struct cifs_tcon *master_tcon = cifs_sb_master_tcon(cifs_sb); struct cifs_ses *ses; struct cifs_tcon *tcon = NULL; struct smb_vol *vol_info; - char username[28]; /* big enough for "krb50x" + hex of ULONG_MAX 6+16 */ - /* We used to have this as MAX_USERNAME which is */ - /* way too big now (256 instead of 32) */ vol_info = kzalloc(sizeof(*vol_info), GFP_KERNEL); if (vol_info == NULL) { @@ -3710,8 +3862,6 @@ cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) goto out; } - snprintf(username, sizeof(username), "krb50x%x", fsuid); - vol_info->username = username; vol_info->local_nls = cifs_sb->local_nls; vol_info->linux_uid = fsuid; vol_info->cred_uid = fsuid; @@ -3721,8 +3871,11 @@ cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) vol_info->local_lease = master_tcon->local_lease; vol_info->no_linux_ext = !master_tcon->unix_ext; - /* FIXME: allow for other secFlg settings */ - vol_info->secFlg = CIFSSEC_MUST_KRB5; + rc = cifs_set_vol_auth(vol_info, master_tcon->ses); + if (rc) { + tcon = ERR_PTR(rc); + goto out; + } /* get a reference for the same TCP session */ spin_lock(&cifs_tcp_ses_lock); @@ -3745,6 +3898,8 @@ cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) if (ses->capabilities & CAP_UNIX) reset_cifs_unix_caps(0, tcon, NULL, vol_info); out: + kfree(vol_info->username); + kfree(vol_info->password); kfree(vol_info); return tcon; From 789b4588da40cf572ef982bdc5d590ec1b0386fe Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 17 Jan 2012 16:09:15 -0500 Subject: [PATCH 606/820] cifs: warn about impending deprecation of legacy MultiuserMount code We'll allow a grace period of 2 releases (3.3 and 3.4) and then remove the legacy code in 3.5. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifs_debug.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index 84e8c0724704..24b3dfc05282 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -676,14 +676,23 @@ static ssize_t cifs_multiuser_mount_proc_write(struct file *file, { char c; int rc; + static bool warned; rc = get_user(c, buffer); if (rc) return rc; if (c == '0' || c == 'n' || c == 'N') multiuser_mount = 0; - else if (c == '1' || c == 'y' || c == 'Y') + else if (c == '1' || c == 'y' || c == 'Y') { multiuser_mount = 1; + if (!warned) { + warned = true; + printk(KERN_WARNING "CIFS VFS: The legacy multiuser " + "mount code is scheduled to be deprecated in " + "3.5. Please switch to using the multiuser " + "mount option."); + } + } return count; } From 766e9f1be128bcdc15aa7d07084d0d51e873b5ed Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 17 Jan 2012 12:45:36 +0000 Subject: [PATCH 607/820] ipv6: fix per device IP snmp counters In commit 4ce3c183fca (snmp: 64bit ipstats_mib for all arches), I forgot to change the /proc/net/dev_snmp6/xxx output for IP counters. percpu array is 64bit per counter but the folding still used the 'long' variant, and output garbage on 32bit arches. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv6/proc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv6/proc.c b/net/ipv6/proc.c index fdeb6d03da81..da2e92d05c15 100644 --- a/net/ipv6/proc.c +++ b/net/ipv6/proc.c @@ -237,8 +237,8 @@ static int snmp6_dev_seq_show(struct seq_file *seq, void *v) struct inet6_dev *idev = (struct inet6_dev *)seq->private; seq_printf(seq, "%-32s\t%u\n", "ifIndex", idev->dev->ifindex); - snmp6_seq_show_item(seq, (void __percpu **)idev->stats.ipv6, NULL, - snmp6_ipstats_list); + snmp6_seq_show_item64(seq, (void __percpu **)idev->stats.ipv6, + snmp6_ipstats_list, offsetof(struct ipstats_mib, syncp)); snmp6_seq_show_item(seq, NULL, idev->stats.icmpv6dev->mibs, snmp6_icmp6_list); snmp6_seq_show_icmpv6msg(seq, idev->stats.icmpv6msgdev->mibs); From 77676fdbd54f0c6fdb71d55d9758bebc69a00fc4 Mon Sep 17 00:00:00 2001 From: Ben Pfaff Date: Tue, 17 Jan 2012 13:33:39 +0000 Subject: [PATCH 608/820] openvswitch: Fix multipart datapath dumps. The logic to split up the list of datapaths into multiple Netlink messages was simply wrong, causing the list to be terminated after the first part. Only about the first 50 datapaths would be dumped. This fixes the problem. Reported-by: Paul Ingram Signed-off-by: Ben Pfaff Signed-off-by: Jesse Gross Signed-off-by: David S. Miller --- net/openvswitch/datapath.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c index d9d01cbf0fcb..ce64c18b8c79 100644 --- a/net/openvswitch/datapath.c +++ b/net/openvswitch/datapath.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2007-2011 Nicira Networks. + * Copyright (c) 2007-2012 Nicira Networks. * * This program is free software; you can redistribute it and/or * modify it under the terms of version 2 of the GNU General Public @@ -1396,9 +1396,8 @@ static int ovs_dp_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb) int i = 0; list_for_each_entry(dp, &dps, list_node) { - if (i < skip) - continue; - if (ovs_dp_cmd_fill_info(dp, skb, NETLINK_CB(cb->skb).pid, + if (i >= skip && + ovs_dp_cmd_fill_info(dp, skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, NLM_F_MULTI, OVS_DP_CMD_NEW) < 0) break; From c3b5003b628d8e373262bee42c7260d6a799c73e Mon Sep 17 00:00:00 2001 From: Matt Carlson Date: Tue, 17 Jan 2012 15:27:23 +0000 Subject: [PATCH 609/820] tg3: Fix single-vector MSI-X code Kdump kernels leave MSI-X interrupts (as setup by the crashed kernel) enabled. However, kdump only enables one CPU in the new environment, thus causing tg3 to abort MSI-X setup. When the driver attempts to enable INTA or MSI interrupt modes on a kdump kernel, interrupt delivery fails. This patch attempts to workaround the problem by forcing the driver to enable a single MSI-X interrupt. In such a configuration, the device's multivector interrupt mode must be disabled. Signed-off-by: Matt Carlson Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/tg3.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-) diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index 076e02a415a0..d529af99157d 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -8846,9 +8846,11 @@ static int tg3_reset_hw(struct tg3 *tp, int reset_phy) tw32_f(GRC_LOCAL_CTRL, tp->grc_local_ctrl); udelay(100); - if (tg3_flag(tp, USING_MSIX) && tp->irq_cnt > 1) { + if (tg3_flag(tp, USING_MSIX)) { val = tr32(MSGINT_MODE); - val |= MSGINT_MODE_MULTIVEC_EN | MSGINT_MODE_ENABLE; + val |= MSGINT_MODE_ENABLE; + if (tp->irq_cnt > 1) + val |= MSGINT_MODE_MULTIVEC_EN; if (!tg3_flag(tp, 1SHOT_MSI)) val |= MSGINT_MODE_ONE_SHOT_DISABLE; tw32(MSGINT_MODE, val); @@ -9548,19 +9550,18 @@ static int tg3_request_firmware(struct tg3 *tp) static bool tg3_enable_msix(struct tg3 *tp) { - int i, rc, cpus = num_online_cpus(); + int i, rc; struct msix_entry msix_ent[tp->irq_max]; - if (cpus == 1) - /* Just fallback to the simpler MSI mode. */ - return false; - - /* - * We want as many rx rings enabled as there are cpus. - * The first MSIX vector only deals with link interrupts, etc, - * so we add one to the number of vectors we are requesting. - */ - tp->irq_cnt = min_t(unsigned, cpus + 1, tp->irq_max); + tp->irq_cnt = num_online_cpus(); + if (tp->irq_cnt > 1) { + /* We want as many rx rings enabled as there are cpus. + * In multiqueue MSI-X mode, the first MSI-X vector + * only deals with link interrupts, etc, so we add + * one to the number of vectors we are requesting. + */ + tp->irq_cnt = min_t(unsigned, tp->irq_cnt + 1, tp->irq_max); + } for (i = 0; i < tp->irq_max; i++) { msix_ent[i].entry = i; From 3bd81a8710710f8bf5d1a5ebac315c842c20bdd3 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Tue, 17 Jan 2012 22:40:07 +0100 Subject: [PATCH 610/820] ACPI processor: Remove unneeded variable passed by acpi_processor_hotadd_init V2 V2: Fix typo: pr->handle -> pr, here: acpi_processor_hotadd_init(pr) This is a very small part taken from patches which afaik are coming from Yunhong Jiang (for a Xen not a Linus repo?). Cleanup only: no functional change. Advantage (beside cleanup) is that other data of the pr (acpi_processor) struct in the acpi_processor_hotadd_init() is needed later, for example a newly introduced flag: pr->flags.need_hotplug_init Signed-off-by: Thomas Renninger CC: Bjorn Helgaas CC: Jiang, Yunhong Signed-off-by: Len Brown --- drivers/acpi/processor_driver.c | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-) diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index 3616ce5509c8..b9cbd9b09450 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -82,7 +82,7 @@ MODULE_LICENSE("GPL"); static int acpi_processor_add(struct acpi_device *device); static int acpi_processor_remove(struct acpi_device *device, int type); static void acpi_processor_notify(struct acpi_device *device, u32 event); -static acpi_status acpi_processor_hotadd_init(acpi_handle handle, int *p_cpu); +static acpi_status acpi_processor_hotadd_init(struct acpi_processor *pr); static int acpi_processor_handle_eject(struct acpi_processor *pr); @@ -324,10 +324,8 @@ static int acpi_processor_get_info(struct acpi_device *device) * they are physically not present. */ if (pr->id == -1) { - if (ACPI_FAILURE - (acpi_processor_hotadd_init(pr->handle, &pr->id))) { + if (ACPI_FAILURE(acpi_processor_hotadd_init(pr))) return -ENODEV; - } } /* * On some boxes several processors use the same processor bus id. @@ -721,18 +719,19 @@ processor_walk_namespace_cb(acpi_handle handle, return (AE_OK); } -static acpi_status acpi_processor_hotadd_init(acpi_handle handle, int *p_cpu) +static acpi_status acpi_processor_hotadd_init(struct acpi_processor *pr) { + acpi_handle handle = pr->handle; if (!is_processor_present(handle)) { return AE_ERROR; } - if (acpi_map_lsapic(handle, p_cpu)) + if (acpi_map_lsapic(handle, &pr->id)) return AE_ERROR; - if (arch_register_cpu(*p_cpu)) { - acpi_unmap_lsapic(*p_cpu); + if (arch_register_cpu(pr->id)) { + acpi_unmap_lsapic(pr->id); return AE_ERROR; } @@ -749,7 +748,7 @@ static int acpi_processor_handle_eject(struct acpi_processor *pr) return (0); } #else -static acpi_status acpi_processor_hotadd_init(acpi_handle handle, int *p_cpu) +static acpi_status acpi_processor_hotadd_init(struct acpi_processor *pr) { return AE_ERROR; } From 65b7f839ceecc0a36c7969c0c9151d5748cd4242 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Tue, 17 Jan 2012 22:40:08 +0100 Subject: [PATCH 611/820] intel_idle: Split up and provide per CPU initialization func Function split up, should have no functional change. Provides entry point for physically hotplugged CPUs to initialize and activate cpuidle. Signed-off-by: Thomas Renninger CC: Deepthi Dharwar CC: Shaohua Li CC: Andrew Morton Signed-off-by: Len Brown --- drivers/idle/intel_idle.c | 82 ++++++++++++++++++++------------------- include/linux/cpuidle.h | 7 ++++ 2 files changed, 49 insertions(+), 40 deletions(-) diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 5d2f8e13cf0e..ef0c04d8dc22 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -478,64 +478,60 @@ static int intel_idle_cpuidle_driver_init(void) /* - * intel_idle_cpuidle_devices_init() + * intel_idle_cpu_init() * allocate, initialize, register cpuidle_devices + * @cpu: cpu/core to initialize */ -static int intel_idle_cpuidle_devices_init(void) +int intel_idle_cpu_init(int cpu) { - int i, cstate; + int cstate; struct cpuidle_device *dev; - intel_idle_cpuidle_devices = alloc_percpu(struct cpuidle_device); - if (intel_idle_cpuidle_devices == NULL) - return -ENOMEM; + dev = per_cpu_ptr(intel_idle_cpuidle_devices, cpu); - for_each_online_cpu(i) { - dev = per_cpu_ptr(intel_idle_cpuidle_devices, i); + dev->state_count = 1; - dev->state_count = 1; + for (cstate = 1; cstate < MWAIT_MAX_NUM_CSTATES; ++cstate) { + int num_substates; - for (cstate = 1; cstate < MWAIT_MAX_NUM_CSTATES; ++cstate) { - int num_substates; + if (cstate > max_cstate) { + printk(PREFIX "max_cstate %d reached\n", + max_cstate); + break; + } - if (cstate > max_cstate) { - printk(PREFIX "max_cstate %d reached\n", - max_cstate); - break; - } + /* does the state exist in CPUID.MWAIT? */ + num_substates = (mwait_substates >> ((cstate) * 4)) + & MWAIT_SUBSTATE_MASK; + if (num_substates == 0) + continue; + /* is the state not enabled? */ + if (cpuidle_state_table[cstate].enter == NULL) + continue; - /* does the state exist in CPUID.MWAIT? */ - num_substates = (mwait_substates >> ((cstate) * 4)) - & MWAIT_SUBSTATE_MASK; - if (num_substates == 0) - continue; - /* is the state not enabled? */ - if (cpuidle_state_table[cstate].enter == NULL) { - continue; - } - - dev->states_usage[dev->state_count].driver_data = - (void *)get_driver_data(cstate); + dev->states_usage[dev->state_count].driver_data = + (void *)get_driver_data(cstate); dev->state_count += 1; } + dev->cpu = cpu; - dev->cpu = i; - if (cpuidle_register_device(dev)) { - pr_debug(PREFIX "cpuidle_register_device %d failed!\n", - i); - intel_idle_cpuidle_devices_uninit(); - return -EIO; - } + if (cpuidle_register_device(dev)) { + pr_debug(PREFIX "cpuidle_register_device %d failed!\n", cpu); + intel_idle_cpuidle_devices_uninit(); + return -EIO; } + if (auto_demotion_disable_flags) + smp_call_function_single(cpu, auto_demotion_disable, NULL, 1); + return 0; } static int __init intel_idle_init(void) { - int retval; + int retval, i; /* Do not load intel_idle at all for now if idle= is passed */ if (boot_option_idle_override != IDLE_NO_OVERRIDE) @@ -553,10 +549,16 @@ static int __init intel_idle_init(void) return retval; } - retval = intel_idle_cpuidle_devices_init(); - if (retval) { - cpuidle_unregister_driver(&intel_idle_driver); - return retval; + intel_idle_cpuidle_devices = alloc_percpu(struct cpuidle_device); + if (intel_idle_cpuidle_devices == NULL) + return -ENOMEM; + + for_each_online_cpu(i) { + retval = intel_idle_cpu_init(i); + if (retval) { + cpuidle_unregister_driver(&intel_idle_driver); + return retval; + } } return 0; diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index 7408af843b8a..93df66ea794a 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -188,7 +188,14 @@ struct cpuidle_governor { extern int cpuidle_register_governor(struct cpuidle_governor *gov); extern void cpuidle_unregister_governor(struct cpuidle_governor *gov); +#ifdef CONFIG_INTEL_IDLE +extern int intel_idle_cpu_init(int cpu); #else +static inline int intel_idle_cpu_init(int cpu) { return -1; } +#endif + +#else +static inline int intel_idle_cpu_init(int cpu) { return -1; } static inline int cpuidle_register_governor(struct cpuidle_governor *gov) {return 0;} From c130bd6f82e5dda28b1a19741c4c2fe269713199 Mon Sep 17 00:00:00 2001 From: Tony Luck Date: Tue, 17 Jan 2012 12:10:16 -0800 Subject: [PATCH 612/820] acpi/apei/einj: Add extensions to EINJ from rev 5.0 of acpi spec ACPI 5.0 provides extensions to the EINJ mechanism to specify the target for the error injection - by APICID for cpu related errors, by address for memory related errors, and by segment/bus/device/function for PCIe related errors. Also extensions for vendor specific error injections. Tested-by: Chen Gong Signed-off-by: Tony Luck Signed-off-by: Len Brown --- Documentation/acpi/apei/einj.txt | 55 ++++++-- drivers/acpi/apei/einj.c | 224 ++++++++++++++++++++++++++----- include/acpi/actbl1.h | 3 +- 3 files changed, 234 insertions(+), 48 deletions(-) diff --git a/Documentation/acpi/apei/einj.txt b/Documentation/acpi/apei/einj.txt index 5cc699ba5453..e7cc36397217 100644 --- a/Documentation/acpi/apei/einj.txt +++ b/Documentation/acpi/apei/einj.txt @@ -47,20 +47,53 @@ directory apei/einj. The following files are provided. - param1 This file is used to set the first error parameter value. Effect of - parameter depends on error_type specified. For memory error, this is - physical memory address. Only available if param_extension module - parameter is specified. + parameter depends on error_type specified. - param2 This file is used to set the second error parameter value. Effect of - parameter depends on error_type specified. For memory error, this is - physical memory address mask. Only available if param_extension - module parameter is specified. + parameter depends on error_type specified. + +BIOS versions based in the ACPI 4.0 specification have limited options +to control where the errors are injected. Your BIOS may support an +extension (enabled with the param_extension=1 module parameter, or +boot command line einj.param_extension=1). This allows the address +and mask for memory injections to be specified by the param1 and +param2 files in apei/einj. + +BIOS versions using the ACPI 5.0 specification have more control over +the target of the injection. For processor related errors (type 0x1, +0x2 and 0x4) the APICID of the target should be provided using the +param1 file in apei/einj. For memory errors (type 0x8, 0x10 and 0x20) +the address is set using param1 with a mask in param2 (0x0 is equivalent +to all ones). For PCI express errors (type 0x40, 0x80 and 0x100) the +segment, bus, device and function are specified using param1: + + 31 24 23 16 15 11 10 8 7 0 + +-------------------------------------------------+ + | segment | bus | device | function | reserved | + +-------------------------------------------------+ + +An ACPI 5.0 BIOS may also allow vendor specific errors to be injected. +In this case a file named vendor will contain identifying information +from the BIOS that hopefully will allow an application wishing to use +the vendor specific extension to tell that they are running on a BIOS +that supports it. All vendor extensions have the 0x80000000 bit set in +error_type. A file vendor_flags controls the interpretation of param1 +and param2 (1 = PROCESSOR, 2 = MEMORY, 4 = PCI). See your BIOS vendor +documentation for details (and expect changes to this API if vendors +creativity in using this feature expands beyond our expectations). + +Example: +# cd /sys/kernel/debug/apei/einj +# cat available_error_type # See which errors can be injected +0x00000002 Processor Uncorrectable non-fatal +0x00000008 Memory Correctable +0x00000010 Memory Uncorrectable non-fatal +# echo 0x12345000 > param1 # Set memory address for injection +# echo 0xfffffffffffff000 > param2 # Mask - anywhere in this page +# echo 0x8 > error_type # Choose correctable memory error +# echo 1 > error_inject # Inject now -Injecting parameter support is a BIOS version specific extension, that -is, it only works on some BIOS version. If you want to use it, please -make sure your BIOS version has the proper support and specify -"param_extension=y" in module parameter. For more information about EINJ, please refer to ACPI specification -version 4.0, section 17.5. +version 4.0, section 17.5 and ACPI 5.0, section 18.6. diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index 589b96c38704..31546fd21029 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -42,6 +42,42 @@ /* Firmware should respond within 1 milliseconds */ #define FIRMWARE_TIMEOUT (1 * NSEC_PER_MSEC) +/* + * ACPI version 5 provides a SET_ERROR_TYPE_WITH_ADDRESS action. + */ +static int acpi5; + +struct set_error_type_with_address { + u32 type; + u32 vendor_extension; + u32 flags; + u32 apicid; + u64 memory_address; + u64 memory_address_range; + u32 pcie_sbdf; +}; +enum { + SETWA_FLAGS_APICID = 1, + SETWA_FLAGS_MEM = 2, + SETWA_FLAGS_PCIE_SBDF = 4, +}; + +/* + * Vendor extensions for platform specific operations + */ +struct vendor_error_type_extension { + u32 length; + u32 pcie_sbdf; + u16 vendor_id; + u16 device_id; + u8 rev_id; + u8 reserved[3]; +}; + +static u32 vendor_flags; +static struct debugfs_blob_wrapper vendor_blob; +static char vendor_dev[64]; + /* * Some BIOSes allow parameters to the SET_ERROR_TYPE entries in the * EINJ table through an unpublished extension. Use with caution as @@ -103,7 +139,14 @@ static struct apei_exec_ins_type einj_ins_type[] = { */ static DEFINE_MUTEX(einj_mutex); -static struct einj_parameter *einj_param; +static void *einj_param; + +#ifndef readq +static inline __u64 readq(volatile void __iomem *addr) +{ + return ((__u64)readl(addr+4) << 32) + readl(addr); +} +#endif #ifndef writeq static inline void writeq(__u64 val, volatile void __iomem *addr) @@ -158,10 +201,31 @@ static int einj_timedout(u64 *t) return 0; } -static u64 einj_get_parameter_address(void) +static void check_vendor_extension(u64 paddr, + struct set_error_type_with_address *v5param) +{ + int offset = readl(&v5param->vendor_extension); + struct vendor_error_type_extension *v; + u32 sbdf; + + if (!offset) + return; + v = ioremap(paddr + offset, sizeof(*v)); + if (!v) + return; + sbdf = readl(&v->pcie_sbdf); + sprintf(vendor_dev, "%x:%x:%x.%x vendor_id=%x device_id=%x rev_id=%x\n", + sbdf >> 24, (sbdf >> 16) & 0xff, + (sbdf >> 11) & 0x1f, (sbdf >> 8) & 0x7, + readw(&v->vendor_id), readw(&v->device_id), + readb(&v->rev_id)); + iounmap(v); +} + +static void *einj_get_parameter_address(void) { int i; - u64 paddr = 0; + u64 paddrv4 = 0, paddrv5 = 0; struct acpi_whea_header *entry; entry = EINJ_TAB_ENTRY(einj_tab); @@ -170,12 +234,40 @@ static u64 einj_get_parameter_address(void) entry->instruction == ACPI_EINJ_WRITE_REGISTER && entry->register_region.space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) - memcpy(&paddr, &entry->register_region.address, - sizeof(paddr)); + memcpy(&paddrv4, &entry->register_region.address, + sizeof(paddrv4)); + if (entry->action == ACPI_EINJ_SET_ERROR_TYPE_WITH_ADDRESS && + entry->instruction == ACPI_EINJ_WRITE_REGISTER && + entry->register_region.space_id == + ACPI_ADR_SPACE_SYSTEM_MEMORY) + memcpy(&paddrv5, &entry->register_region.address, + sizeof(paddrv5)); entry++; } + if (paddrv5) { + struct set_error_type_with_address *v5param; - return paddr; + v5param = ioremap(paddrv5, sizeof(*v5param)); + if (v5param) { + acpi5 = 1; + check_vendor_extension(paddrv5, v5param); + return v5param; + } + } + if (paddrv4) { + struct einj_parameter *v4param; + + v4param = ioremap(paddrv4, sizeof(*v4param)); + if (!v4param) + return 0; + if (readq(&v4param->reserved1) || readq(&v4param->reserved2)) { + iounmap(v4param); + return 0; + } + return v4param; + } + + return 0; } /* do sanity check to trigger table */ @@ -293,12 +385,56 @@ static int __einj_error_inject(u32 type, u64 param1, u64 param2) if (rc) return rc; apei_exec_ctx_set_input(&ctx, type); - rc = apei_exec_run(&ctx, ACPI_EINJ_SET_ERROR_TYPE); - if (rc) - return rc; - if (einj_param) { - writeq(param1, &einj_param->param1); - writeq(param2, &einj_param->param2); + if (acpi5) { + struct set_error_type_with_address *v5param = einj_param; + + writel(type, &v5param->type); + if (type & 0x80000000) { + switch (vendor_flags) { + case SETWA_FLAGS_APICID: + writel(param1, &v5param->apicid); + break; + case SETWA_FLAGS_MEM: + writeq(param1, &v5param->memory_address); + writeq(param2, &v5param->memory_address_range); + break; + case SETWA_FLAGS_PCIE_SBDF: + writel(param1, &v5param->pcie_sbdf); + break; + } + writel(vendor_flags, &v5param->flags); + } else { + switch (type) { + case ACPI_EINJ_PROCESSOR_CORRECTABLE: + case ACPI_EINJ_PROCESSOR_UNCORRECTABLE: + case ACPI_EINJ_PROCESSOR_FATAL: + writel(param1, &v5param->apicid); + writel(SETWA_FLAGS_APICID, &v5param->flags); + break; + case ACPI_EINJ_MEMORY_CORRECTABLE: + case ACPI_EINJ_MEMORY_UNCORRECTABLE: + case ACPI_EINJ_MEMORY_FATAL: + writeq(param1, &v5param->memory_address); + writeq(param2, &v5param->memory_address_range); + writel(SETWA_FLAGS_MEM, &v5param->flags); + break; + case ACPI_EINJ_PCIX_CORRECTABLE: + case ACPI_EINJ_PCIX_UNCORRECTABLE: + case ACPI_EINJ_PCIX_FATAL: + writel(param1, &v5param->pcie_sbdf); + writel(SETWA_FLAGS_PCIE_SBDF, &v5param->flags); + break; + } + } + } else { + rc = apei_exec_run(&ctx, ACPI_EINJ_SET_ERROR_TYPE); + if (rc) + return rc; + if (einj_param) { + struct einj_parameter *v4param = einj_param; + writeq(param1, &v4param->param1); + writeq(param2, &v4param->param2); + } } rc = apei_exec_run(&ctx, ACPI_EINJ_EXECUTE_OPERATION); if (rc) @@ -408,15 +544,25 @@ static int error_type_set(void *data, u64 val) { int rc; u32 available_error_type = 0; + u32 tval, vendor; + + /* + * Vendor defined types have 0x80000000 bit set, and + * are not enumerated by ACPI_EINJ_GET_ERROR_TYPE + */ + vendor = val & 0x80000000; + tval = val & 0x7fffffff; /* Only one error type can be specified */ - if (val & (val - 1)) - return -EINVAL; - rc = einj_get_available_error_type(&available_error_type); - if (rc) - return rc; - if (!(val & available_error_type)) + if (tval & (tval - 1)) return -EINVAL; + if (!vendor) { + rc = einj_get_available_error_type(&available_error_type); + if (rc) + return rc; + if (!(val & available_error_type)) + return -EINVAL; + } error_type = val; return 0; @@ -455,7 +601,6 @@ static int einj_check_table(struct acpi_table_einj *einj_tab) static int __init einj_init(void) { int rc; - u64 param_paddr; acpi_status status; struct dentry *fentry; struct apei_exec_context ctx; @@ -509,23 +654,30 @@ static int __init einj_init(void) rc = apei_exec_pre_map_gars(&ctx); if (rc) goto err_release; - if (param_extension) { - param_paddr = einj_get_parameter_address(); - if (param_paddr) { - einj_param = ioremap(param_paddr, sizeof(*einj_param)); - rc = -ENOMEM; - if (!einj_param) - goto err_unmap; - fentry = debugfs_create_x64("param1", S_IRUSR | S_IWUSR, - einj_debug_dir, &error_param1); - if (!fentry) - goto err_unmap; - fentry = debugfs_create_x64("param2", S_IRUSR | S_IWUSR, - einj_debug_dir, &error_param2); - if (!fentry) - goto err_unmap; - } else - pr_warn(EINJ_PFX "Parameter extension is not supported.\n"); + + einj_param = einj_get_parameter_address(); + if ((param_extension || acpi5) && einj_param) { + fentry = debugfs_create_x64("param1", S_IRUSR | S_IWUSR, + einj_debug_dir, &error_param1); + if (!fentry) + goto err_unmap; + fentry = debugfs_create_x64("param2", S_IRUSR | S_IWUSR, + einj_debug_dir, &error_param2); + if (!fentry) + goto err_unmap; + } + + if (vendor_dev[0]) { + vendor_blob.data = vendor_dev; + vendor_blob.size = strlen(vendor_dev); + fentry = debugfs_create_blob("vendor", S_IRUSR, + einj_debug_dir, &vendor_blob); + if (!fentry) + goto err_unmap; + fentry = debugfs_create_x32("vendor_flags", S_IRUSR | S_IWUSR, + einj_debug_dir, &vendor_flags); + if (!fentry) + goto err_unmap; } pr_info(EINJ_PFX "Error INJection is initialized.\n"); diff --git a/include/acpi/actbl1.h b/include/acpi/actbl1.h index 7504bc99b29b..f25d7efee630 100644 --- a/include/acpi/actbl1.h +++ b/include/acpi/actbl1.h @@ -228,7 +228,8 @@ enum acpi_einj_actions { ACPI_EINJ_EXECUTE_OPERATION = 5, ACPI_EINJ_CHECK_BUSY_STATUS = 6, ACPI_EINJ_GET_COMMAND_STATUS = 7, - ACPI_EINJ_ACTION_RESERVED = 8, /* 8 and greater are reserved */ + ACPI_EINJ_SET_ERROR_TYPE_WITH_ADDRESS = 8, + ACPI_EINJ_ACTION_RESERVED = 9, /* 9 and greater are reserved */ ACPI_EINJ_TRIGGER_ERROR = 0xFF /* Except for this value */ }; From f8b5a31877b93f7136ce8c22ce44930e39b41204 Mon Sep 17 00:00:00 2001 From: Jerry Huang Date: Thu, 5 Jan 2012 09:40:56 +0800 Subject: [PATCH 613/820] powerpc/85xx: Fix cmd12 bug and add the chip compatible for eSDHC According to latest kernel, the auto-cmd12 property should be "sdhci,auto-cmd12", and according to the SDHC binding and the workaround for the special chip, add the chip compatible for eSDHC: "fsl,p1022-esdhc", "fsl,mpc8536-esdhc", "fsl,p1020-esdhc", "fsl,p2020-esdhc" and "fsl,p1010-esdhc". Signed-off-by: Jerry Huang Signed-off-by: Kumar Gala --- arch/powerpc/boot/dts/fsl/mpc8536si-post.dtsi | 4 ++++ arch/powerpc/boot/dts/fsl/p1010si-post.dtsi | 3 ++- arch/powerpc/boot/dts/fsl/p1020si-post.dtsi | 4 ++++ arch/powerpc/boot/dts/fsl/p1022si-post.dtsi | 3 ++- arch/powerpc/boot/dts/fsl/p2020si-post.dtsi | 4 ++++ 5 files changed, 16 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/boot/dts/fsl/mpc8536si-post.dtsi b/arch/powerpc/boot/dts/fsl/mpc8536si-post.dtsi index 89af62637707..b37da56018b6 100644 --- a/arch/powerpc/boot/dts/fsl/mpc8536si-post.dtsi +++ b/arch/powerpc/boot/dts/fsl/mpc8536si-post.dtsi @@ -236,6 +236,10 @@ }; /include/ "pq3-esdhc-0.dtsi" + sdhc@2e000 { + compatible = "fsl,mpc8536-esdhc", "fsl,esdhc"; + }; + /include/ "pq3-sec3.0-0.dtsi" /include/ "pq3-mpic.dtsi" /include/ "pq3-mpic-timer-B.dtsi" diff --git a/arch/powerpc/boot/dts/fsl/p1010si-post.dtsi b/arch/powerpc/boot/dts/fsl/p1010si-post.dtsi index bd9e163c764b..a97d1263372c 100644 --- a/arch/powerpc/boot/dts/fsl/p1010si-post.dtsi +++ b/arch/powerpc/boot/dts/fsl/p1010si-post.dtsi @@ -158,7 +158,8 @@ /include/ "pq3-usb2-dr-0.dtsi" /include/ "pq3-esdhc-0.dtsi" sdhc@2e000 { - fsl,sdhci-auto-cmd12; + compatible = "fsl,p1010-esdhc", "fsl,esdhc"; + sdhci,auto-cmd12; }; /include/ "pq3-sec4.4-0.dtsi" diff --git a/arch/powerpc/boot/dts/fsl/p1020si-post.dtsi b/arch/powerpc/boot/dts/fsl/p1020si-post.dtsi index fc924c5ffebe..5de5fc351314 100644 --- a/arch/powerpc/boot/dts/fsl/p1020si-post.dtsi +++ b/arch/powerpc/boot/dts/fsl/p1020si-post.dtsi @@ -145,6 +145,10 @@ /include/ "pq3-usb2-dr-1.dtsi" /include/ "pq3-esdhc-0.dtsi" + sdhc@2e000 { + compatible = "fsl,p1020-esdhc", "fsl,esdhc"; + sdhci,auto-cmd12; + }; /include/ "pq3-sec3.3-0.dtsi" /include/ "pq3-mpic.dtsi" diff --git a/arch/powerpc/boot/dts/fsl/p1022si-post.dtsi b/arch/powerpc/boot/dts/fsl/p1022si-post.dtsi index 16239b199d0a..ff9ed1d87929 100644 --- a/arch/powerpc/boot/dts/fsl/p1022si-post.dtsi +++ b/arch/powerpc/boot/dts/fsl/p1022si-post.dtsi @@ -203,7 +203,8 @@ /include/ "pq3-esdhc-0.dtsi" sdhc@2e000 { - fsl,sdhci-auto-cmd12; + compatible = "fsl,p1022-esdhc", "fsl,esdhc"; + sdhci,auto-cmd12; }; /include/ "pq3-sec3.3-0.dtsi" diff --git a/arch/powerpc/boot/dts/fsl/p2020si-post.dtsi b/arch/powerpc/boot/dts/fsl/p2020si-post.dtsi index c041050561a7..332e9e75e6c2 100644 --- a/arch/powerpc/boot/dts/fsl/p2020si-post.dtsi +++ b/arch/powerpc/boot/dts/fsl/p2020si-post.dtsi @@ -182,6 +182,10 @@ /include/ "pq3-etsec1-1.dtsi" /include/ "pq3-etsec1-2.dtsi" /include/ "pq3-esdhc-0.dtsi" + sdhc@2e000 { + compatible = "fsl,p2020-esdhc", "fsl,esdhc"; + }; + /include/ "pq3-sec3.1-0.dtsi" /include/ "pq3-mpic.dtsi" /include/ "pq3-mpic-timer-B.dtsi" From 70a0ac66869907ce3914f26fe1eba984c9eb438e Mon Sep 17 00:00:00 2001 From: Ramneek Mehresh Date: Wed, 18 Jan 2012 11:20:39 +0530 Subject: [PATCH 614/820] powerpc/85xx: Enable USB2 controller node for P1020RDB Enable USB2 controller node for P1020RDB. USB2 controller is used only when board boots from SPI or SD as it is muxed with eLBC Signed-off-by: Ramneek Mehresh Signed-off-by: Kumar Gala --- arch/powerpc/boot/dts/p1020rdb.dtsi | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/arch/powerpc/boot/dts/p1020rdb.dtsi b/arch/powerpc/boot/dts/p1020rdb.dtsi index b5bd86f4baf2..17559c4dc5d8 100644 --- a/arch/powerpc/boot/dts/p1020rdb.dtsi +++ b/arch/powerpc/boot/dts/p1020rdb.dtsi @@ -1,7 +1,7 @@ /* * P1020 RDB Device Tree Source stub (no addresses or top-level ranges) * - * Copyright 2011 Freescale Semiconductor Inc. + * Copyright 2011-2012 Freescale Semiconductor Inc. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions are met: @@ -192,15 +192,12 @@ phy_type = "ulpi"; }; - /* USB2 is shared with localbus, so it must be disabled - by default. We can't put 'status = "disabled";' here - since U-Boot doesn't clear the status property when - it enables USB2. OTOH, U-Boot does create a new node - when there isn't any. So, just comment it out. + /* USB2 is shared with localbus. It is used + only in case of SPI and SD boot after + appropriate device-tree fixup done by uboot */ usb@23000 { phy_type = "ulpi"; }; - */ mdio@24000 { phy0: ethernet-phy@0 { From 621c4b999e3e1dc6f72b0c1680029981edc03caa Mon Sep 17 00:00:00 2001 From: Ramneek Mehresh Date: Wed, 18 Jan 2012 12:18:46 +0530 Subject: [PATCH 615/820] powerpc/85xx: Add dr_mode property in USB nodes Add usb2 controller node for P1020RDB, P2020RDB, P2020DS, P1021MDS Signed-off-by: Ramneek Mehresh Signed-off-by: Kumar Gala --- arch/powerpc/boot/dts/p1020rdb.dtsi | 2 ++ arch/powerpc/boot/dts/p1021mds.dts | 3 ++- arch/powerpc/boot/dts/p2020ds.dtsi | 3 ++- arch/powerpc/boot/dts/p2020rdb.dts | 3 ++- 4 files changed, 8 insertions(+), 3 deletions(-) diff --git a/arch/powerpc/boot/dts/p1020rdb.dtsi b/arch/powerpc/boot/dts/p1020rdb.dtsi index 17559c4dc5d8..1fb7e0e0940f 100644 --- a/arch/powerpc/boot/dts/p1020rdb.dtsi +++ b/arch/powerpc/boot/dts/p1020rdb.dtsi @@ -190,6 +190,7 @@ usb@22000 { phy_type = "ulpi"; + dr_mode = "host"; }; /* USB2 is shared with localbus. It is used @@ -197,6 +198,7 @@ appropriate device-tree fixup done by uboot */ usb@23000 { phy_type = "ulpi"; + dr_mode = "host"; }; mdio@24000 { diff --git a/arch/powerpc/boot/dts/p1021mds.dts b/arch/powerpc/boot/dts/p1021mds.dts index d9540791e434..97116f198a37 100644 --- a/arch/powerpc/boot/dts/p1021mds.dts +++ b/arch/powerpc/boot/dts/p1021mds.dts @@ -1,7 +1,7 @@ /* * P1021 MDS Device Tree Source * - * Copyright 2010 Freescale Semiconductor Inc. + * Copyright 2010,2012 Freescale Semiconductor Inc. * * This program is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the @@ -151,6 +151,7 @@ usb@22000 { phy_type = "ulpi"; + dr_mode = "host"; }; mdio@24000 { diff --git a/arch/powerpc/boot/dts/p2020ds.dtsi b/arch/powerpc/boot/dts/p2020ds.dtsi index c1cf6cef4dd6..d3b939c573b0 100644 --- a/arch/powerpc/boot/dts/p2020ds.dtsi +++ b/arch/powerpc/boot/dts/p2020ds.dtsi @@ -1,7 +1,7 @@ /* * P2020DS Device Tree Source stub (no addresses or top-level ranges) * - * Copyright 2011 Freescale Semiconductor Inc. + * Copyright 2011-2012 Freescale Semiconductor Inc. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions are met: @@ -134,6 +134,7 @@ &board_soc { usb@22000 { phy_type = "ulpi"; + dr_mode = "host"; }; mdio@24520 { diff --git a/arch/powerpc/boot/dts/p2020rdb.dts b/arch/powerpc/boot/dts/p2020rdb.dts index 26759a591712..eb8a6aa2bda5 100644 --- a/arch/powerpc/boot/dts/p2020rdb.dts +++ b/arch/powerpc/boot/dts/p2020rdb.dts @@ -1,7 +1,7 @@ /* * P2020 RDB Device Tree Source * - * Copyright 2009-2011 Freescale Semiconductor Inc. + * Copyright 2009-2012 Freescale Semiconductor Inc. * * This program is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the @@ -197,6 +197,7 @@ usb@22000 { phy_type = "ulpi"; + dr_mode = "host"; }; mdio@24520 { From 9446f3efc53512e5ad9e0966539021a2a41fe5a0 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 17 Jan 2012 10:32:01 +0100 Subject: [PATCH 616/820] mac80211: fix debugfs key->station symlink Since stations moved into a virtual interface subdirectory, this link has been broken. Fix it. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/debugfs_key.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/mac80211/debugfs_key.c b/net/mac80211/debugfs_key.c index 38e6101190d9..59edcd95a58d 100644 --- a/net/mac80211/debugfs_key.c +++ b/net/mac80211/debugfs_key.c @@ -225,9 +225,9 @@ KEY_OPS(key); key, &key_##name##_ops); void ieee80211_debugfs_key_add(struct ieee80211_key *key) - { +{ static int keycount; - char buf[50]; + char buf[100]; struct sta_info *sta; if (!key->local->debugfs.keys) @@ -244,7 +244,8 @@ void ieee80211_debugfs_key_add(struct ieee80211_key *key) sta = key->sta; if (sta) { - sprintf(buf, "../../stations/%pM", sta->sta.addr); + sprintf(buf, "../../netdev:%s/stations/%pM", + sta->sdata->name, sta->sta.addr); key->debugfs.stalink = debugfs_create_symlink("station", key->debugfs.dir, buf); } From f96b08a7e6f69c0f0a576554df3df5b1b519c479 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Tue, 17 Jan 2012 12:38:50 +0100 Subject: [PATCH 617/820] brcmsmac: fix tx queue flush infinite loop This patch workaround live deadlock problem caused by infinite loop in brcms_c_wait_for_tx_completion(). I do not consider the patch as the proper fix, which should fix the real reason of tx queue flush failure, but patch helps with system lockup. Reference: https://bugzilla.kernel.org/show_bug.cgi?id=42576 Reported-and-tested-by: Patrick Cc: stable@vger.kernel.org # 3.2+ Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville --- drivers/net/wireless/brcm80211/brcmsmac/main.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/brcm80211/brcmsmac/main.c b/drivers/net/wireless/brcm80211/brcmsmac/main.c index f7ed34034f88..f6affc6fd12a 100644 --- a/drivers/net/wireless/brcm80211/brcmsmac/main.c +++ b/drivers/net/wireless/brcm80211/brcmsmac/main.c @@ -7981,13 +7981,21 @@ int brcms_c_get_curband(struct brcms_c_info *wlc) void brcms_c_wait_for_tx_completion(struct brcms_c_info *wlc, bool drop) { + int timeout = 20; + /* flush packet queue when requested */ if (drop) brcmu_pktq_flush(&wlc->pkt_queue->q, false, NULL, NULL); /* wait for queue and DMA fifos to run dry */ - while (!pktq_empty(&wlc->pkt_queue->q) || brcms_txpktpendtot(wlc) > 0) + while (!pktq_empty(&wlc->pkt_queue->q) || brcms_txpktpendtot(wlc) > 0) { brcms_msleep(wlc->wl, 1); + + if (--timeout == 0) + break; + } + + WARN_ON_ONCE(timeout == 0); } void brcms_c_set_beacon_listen_interval(struct brcms_c_info *wlc, u8 interval) From 65e8b0ccb6cf176f8eddb1b05534be46580da9dd Mon Sep 17 00:00:00 2001 From: Javier Cardona Date: Tue, 17 Jan 2012 18:17:46 -0800 Subject: [PATCH 618/820] mac80211: Use the right headroom size for mesh mgmt frames Use local->tx_headroom instad of local->hw.extra_tx_headroom. local->tx_headroom is the max of hw.extra_tx_headroom required by the driver and the headroom required by mac80211 for status reporting. On drivers where hw.extra_tx_headroom is smaller than what mac80211 requires (e.g. ath5k), we would not reserve sufficient buffer space to report tx status. Also, don't reserve local->tx_headroom + local->hw.extra_tx_headroom. Reported-by: Simon Morgenthaler Reported-by: Kai Scharwies Signed-off-by: Javier Cardona Signed-off-by: John W. Linville --- net/mac80211/mesh_hwmp.c | 8 ++++---- net/mac80211/mesh_plink.c | 4 ++-- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/net/mac80211/mesh_hwmp.c b/net/mac80211/mesh_hwmp.c index 73abb7524b2c..54df1b2bafd2 100644 --- a/net/mac80211/mesh_hwmp.c +++ b/net/mac80211/mesh_hwmp.c @@ -119,12 +119,12 @@ static int mesh_path_sel_frame_tx(enum mpath_frame_type action, u8 flags, int hdr_len = offsetof(struct ieee80211_mgmt, u.action.u.mesh_action) + sizeof(mgmt->u.action.u.mesh_action); - skb = dev_alloc_skb(local->hw.extra_tx_headroom + + skb = dev_alloc_skb(local->tx_headroom + hdr_len + 2 + 37); /* max HWMP IE */ if (!skb) return -1; - skb_reserve(skb, local->hw.extra_tx_headroom); + skb_reserve(skb, local->tx_headroom); mgmt = (struct ieee80211_mgmt *) skb_put(skb, hdr_len); memset(mgmt, 0, hdr_len); mgmt->frame_control = cpu_to_le16(IEEE80211_FTYPE_MGMT | @@ -250,12 +250,12 @@ int mesh_path_error_tx(u8 ttl, u8 *target, __le32 target_sn, if (time_before(jiffies, ifmsh->next_perr)) return -EAGAIN; - skb = dev_alloc_skb(local->hw.extra_tx_headroom + + skb = dev_alloc_skb(local->tx_headroom + hdr_len + 2 + 15 /* PERR IE */); if (!skb) return -1; - skb_reserve(skb, local->tx_headroom + local->hw.extra_tx_headroom); + skb_reserve(skb, local->tx_headroom); mgmt = (struct ieee80211_mgmt *) skb_put(skb, hdr_len); memset(mgmt, 0, hdr_len); mgmt->frame_control = cpu_to_le16(IEEE80211_FTYPE_MGMT | diff --git a/net/mac80211/mesh_plink.c b/net/mac80211/mesh_plink.c index 41ef1b476442..a17251730b9e 100644 --- a/net/mac80211/mesh_plink.c +++ b/net/mac80211/mesh_plink.c @@ -172,7 +172,7 @@ static int mesh_plink_frame_tx(struct ieee80211_sub_if_data *sdata, int hdr_len = offsetof(struct ieee80211_mgmt, u.action.u.self_prot) + sizeof(mgmt->u.action.u.self_prot); - skb = dev_alloc_skb(local->hw.extra_tx_headroom + + skb = dev_alloc_skb(local->tx_headroom + hdr_len + 2 + /* capability info */ 2 + /* AID */ @@ -186,7 +186,7 @@ static int mesh_plink_frame_tx(struct ieee80211_sub_if_data *sdata, sdata->u.mesh.ie_len); if (!skb) return -1; - skb_reserve(skb, local->hw.extra_tx_headroom); + skb_reserve(skb, local->tx_headroom); mgmt = (struct ieee80211_mgmt *) skb_put(skb, hdr_len); memset(mgmt, 0, hdr_len); mgmt->frame_control = cpu_to_le16(IEEE80211_FTYPE_MGMT | From bc4934bc61d0a11fd62c5187ff83645628f8be8b Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 18 Jan 2012 14:10:25 +0100 Subject: [PATCH 619/820] mac80211: fix work removal on deauth request When deauth is requested while an auth or assoc work item is in progress, we currently delete it without regard for any state it might need to clean up. Fix it by cleaning up for those items. In the case Pontus found, the problem manifested itself as such: authenticate with 00:23:69:aa:dd:7b (try 1) authenticated failed to insert Dummy STA entry for the AP (error -17) deauthenticating from 00:23:69:aa:dd:7b by local choice (reason=2) It could also happen differently if the driver uses the tx_sync callback. We can't just call the ->done() method of the work items because that will lock up due to the locking in cfg80211. This fix isn't very clean, but that seems acceptable since I have patches pending to remove this code completely. Cc: stable@vger.kernel.org Reported-by: Pontus Fuchs Tested-by: Pontus Fuchs Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/mlme.c | 38 +++++++++++++++++++++++++++----------- 1 file changed, 27 insertions(+), 11 deletions(-) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index ecb4c84c1bb3..295be92f7c77 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -2750,7 +2750,6 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, { struct ieee80211_local *local = sdata->local; struct ieee80211_if_managed *ifmgd = &sdata->u.mgd; - struct ieee80211_work *wk; u8 bssid[ETH_ALEN]; bool assoc_bss = false; @@ -2763,30 +2762,47 @@ int ieee80211_mgd_deauth(struct ieee80211_sub_if_data *sdata, assoc_bss = true; } else { bool not_auth_yet = false; + struct ieee80211_work *tmp, *wk = NULL; mutex_unlock(&ifmgd->mtx); mutex_lock(&local->mtx); - list_for_each_entry(wk, &local->work_list, list) { - if (wk->sdata != sdata) + list_for_each_entry(tmp, &local->work_list, list) { + if (tmp->sdata != sdata) continue; - if (wk->type != IEEE80211_WORK_DIRECT_PROBE && - wk->type != IEEE80211_WORK_AUTH && - wk->type != IEEE80211_WORK_ASSOC && - wk->type != IEEE80211_WORK_ASSOC_BEACON_WAIT) + if (tmp->type != IEEE80211_WORK_DIRECT_PROBE && + tmp->type != IEEE80211_WORK_AUTH && + tmp->type != IEEE80211_WORK_ASSOC && + tmp->type != IEEE80211_WORK_ASSOC_BEACON_WAIT) continue; - if (memcmp(req->bss->bssid, wk->filter_ta, ETH_ALEN)) + if (memcmp(req->bss->bssid, tmp->filter_ta, ETH_ALEN)) continue; - not_auth_yet = wk->type == IEEE80211_WORK_DIRECT_PROBE; - list_del_rcu(&wk->list); - free_work(wk); + not_auth_yet = tmp->type == IEEE80211_WORK_DIRECT_PROBE; + list_del_rcu(&tmp->list); + synchronize_rcu(); + wk = tmp; break; } mutex_unlock(&local->mtx); + if (wk && wk->type == IEEE80211_WORK_ASSOC) { + /* clean up dummy sta & TX sync */ + sta_info_destroy_addr(wk->sdata, wk->filter_ta); + if (wk->assoc.synced) + drv_finish_tx_sync(local, wk->sdata, + wk->filter_ta, + IEEE80211_TX_SYNC_ASSOC); + } else if (wk && wk->type == IEEE80211_WORK_AUTH) { + if (wk->probe_auth.synced) + drv_finish_tx_sync(local, wk->sdata, + wk->filter_ta, + IEEE80211_TX_SYNC_AUTH); + } + kfree(wk); + /* * If somebody requests authentication and we haven't * sent out an auth frame yet there's no need to send From 4f3d09de38d234ce7ffba5ec5a7e6704f983d375 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Wed, 11 Jan 2012 15:50:15 -0500 Subject: [PATCH 620/820] b43: add option to avoid duplicating device support with brcmsmac Signed-off-by: John W. Linville --- drivers/net/wireless/b43/Kconfig | 6 ++++++ drivers/net/wireless/b43/main.c | 2 ++ 2 files changed, 8 insertions(+) diff --git a/drivers/net/wireless/b43/Kconfig b/drivers/net/wireless/b43/Kconfig index b97a40ed5fff..3876c7ea54f4 100644 --- a/drivers/net/wireless/b43/Kconfig +++ b/drivers/net/wireless/b43/Kconfig @@ -31,6 +31,12 @@ config B43_BCMA depends on B43 && BCMA default y +config B43_BCMA_EXTRA + bool "Hardware support that overlaps with the brcmsmac driver" + depends on B43_BCMA + default n if BRCMSMAC || BRCMSMAC_MODULE + default y + config B43_SSB bool depends on B43 && SSB diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index b91f28ef1032..23ffb1b9a86f 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -116,8 +116,10 @@ MODULE_PARM_DESC(pio, "Use PIO accesses by default: 0=DMA, 1=PIO"); #ifdef CONFIG_B43_BCMA static const struct bcma_device_id b43_bcma_tbl[] = { BCMA_CORE(BCMA_MANUF_BCM, BCMA_CORE_80211, 0x11, BCMA_ANY_CLASS), +#ifdef CONFIG_B43_BCMA_EXTRA BCMA_CORE(BCMA_MANUF_BCM, BCMA_CORE_80211, 0x17, BCMA_ANY_CLASS), BCMA_CORE(BCMA_MANUF_BCM, BCMA_CORE_80211, 0x18, BCMA_ANY_CLASS), +#endif BCMA_CORE(BCMA_MANUF_BCM, BCMA_CORE_80211, 0x1D, BCMA_ANY_CLASS), BCMA_CORETABLE_END }; From 94bd2442d25454a874e070d871f50f4ce9d57101 Mon Sep 17 00:00:00 2001 From: Laxman Dewangan Date: Wed, 18 Jan 2012 20:07:35 +0530 Subject: [PATCH 621/820] gpio: tps65910: Use correct offset for gpio initialization Using the correct gpio offset for setting the initial value of gpio when setting output direction. Signed-off-by: Laxman Dewangan Signed-off-by: Grant Likely --- drivers/gpio/gpio-tps65910.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-tps65910.c b/drivers/gpio/gpio-tps65910.c index b9c1c297669e..91f45b965d1e 100644 --- a/drivers/gpio/gpio-tps65910.c +++ b/drivers/gpio/gpio-tps65910.c @@ -52,7 +52,7 @@ static int tps65910_gpio_output(struct gpio_chip *gc, unsigned offset, struct tps65910 *tps65910 = container_of(gc, struct tps65910, gpio); /* Set the initial value */ - tps65910_gpio_set(gc, 0, value); + tps65910_gpio_set(gc, offset, value); return tps65910_set_bits(tps65910, TPS65910_GPIO0 + offset, GPIO_CFG_MASK); From d00a9dd21bdf7908b70866794c8313ee8a5abd5c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 18 Jan 2012 07:21:42 +0000 Subject: [PATCH 622/820] net: bpf_jit: fix divide by 0 generation Several problems fixed in this patch : 1) Target of the conditional jump in case a divide by 0 is performed by a bpf is wrong. 2) Must 'generate' the full function prologue/epilogue at pass=0, or else we can stop too early in pass=1 if the proglen doesnt change. (if the increase of prologue/epilogue equals decrease of all instructions length because some jumps are converted to near jumps) 3) Change the wrong length detection at the end of code generation to issue a more explicit message, no need for a full stack trace. Reported-by: Phil Oester Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- arch/x86/net/bpf_jit_comp.c | 36 ++++++++++++++++++++++-------------- 1 file changed, 22 insertions(+), 14 deletions(-) diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index 7b65f752c5f8..7c1b765ecc59 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -151,17 +151,18 @@ void bpf_jit_compile(struct sk_filter *fp) cleanup_addr = proglen; /* epilogue address */ for (pass = 0; pass < 10; pass++) { + u8 seen_or_pass0 = (pass == 0) ? (SEEN_XREG | SEEN_DATAREF | SEEN_MEM) : seen; /* no prologue/epilogue for trivial filters (RET something) */ proglen = 0; prog = temp; - if (seen) { + if (seen_or_pass0) { EMIT4(0x55, 0x48, 0x89, 0xe5); /* push %rbp; mov %rsp,%rbp */ EMIT4(0x48, 0x83, 0xec, 96); /* subq $96,%rsp */ /* note : must save %rbx in case bpf_error is hit */ - if (seen & (SEEN_XREG | SEEN_DATAREF)) + if (seen_or_pass0 & (SEEN_XREG | SEEN_DATAREF)) EMIT4(0x48, 0x89, 0x5d, 0xf8); /* mov %rbx, -8(%rbp) */ - if (seen & SEEN_XREG) + if (seen_or_pass0 & SEEN_XREG) CLEAR_X(); /* make sure we dont leek kernel memory */ /* @@ -170,7 +171,7 @@ void bpf_jit_compile(struct sk_filter *fp) * r9 = skb->len - skb->data_len * r8 = skb->data */ - if (seen & SEEN_DATAREF) { + if (seen_or_pass0 & SEEN_DATAREF) { if (offsetof(struct sk_buff, len) <= 127) /* mov off8(%rdi),%r9d */ EMIT4(0x44, 0x8b, 0x4f, offsetof(struct sk_buff, len)); @@ -260,9 +261,14 @@ void bpf_jit_compile(struct sk_filter *fp) case BPF_S_ALU_DIV_X: /* A /= X; */ seen |= SEEN_XREG; EMIT2(0x85, 0xdb); /* test %ebx,%ebx */ - if (pc_ret0 != -1) - EMIT_COND_JMP(X86_JE, addrs[pc_ret0] - (addrs[i] - 4)); - else { + if (pc_ret0 > 0) { + /* addrs[pc_ret0 - 1] is start address of target + * (addrs[i] - 4) is the address following this jmp + * ("xor %edx,%edx; div %ebx" being 4 bytes long) + */ + EMIT_COND_JMP(X86_JE, addrs[pc_ret0 - 1] - + (addrs[i] - 4)); + } else { EMIT_COND_JMP(X86_JNE, 2 + 5); CLEAR_A(); EMIT1_off32(0xe9, cleanup_addr - (addrs[i] - 4)); /* jmp .+off32 */ @@ -335,12 +341,12 @@ void bpf_jit_compile(struct sk_filter *fp) } /* fallinto */ case BPF_S_RET_A: - if (seen) { + if (seen_or_pass0) { if (i != flen - 1) { EMIT_JMP(cleanup_addr - addrs[i]); break; } - if (seen & SEEN_XREG) + if (seen_or_pass0 & SEEN_XREG) EMIT4(0x48, 0x8b, 0x5d, 0xf8); /* mov -8(%rbp),%rbx */ EMIT1(0xc9); /* leaveq */ } @@ -483,8 +489,9 @@ common_load: seen |= SEEN_DATAREF; goto common_load; case BPF_S_LDX_B_MSH: if ((int)K < 0) { - if (pc_ret0 != -1) { - EMIT_JMP(addrs[pc_ret0] - addrs[i]); + if (pc_ret0 > 0) { + /* addrs[pc_ret0 - 1] is the start address */ + EMIT_JMP(addrs[pc_ret0 - 1] - addrs[i]); break; } CLEAR_A(); @@ -599,13 +606,14 @@ cond_branch: f_offset = addrs[i + filter[i].jf] - addrs[i]; * use it to give the cleanup instruction(s) addr */ cleanup_addr = proglen - 1; /* ret */ - if (seen) + if (seen_or_pass0) cleanup_addr -= 1; /* leaveq */ - if (seen & SEEN_XREG) + if (seen_or_pass0 & SEEN_XREG) cleanup_addr -= 4; /* mov -8(%rbp),%rbx */ if (image) { - WARN_ON(proglen != oldproglen); + if (proglen != oldproglen) + pr_err("bpb_jit_compile proglen=%u != oldproglen=%u\n", proglen, oldproglen); break; } if (proglen == oldproglen) { From 93d3e3678f23109363cd6f99f2944d2cda616b23 Mon Sep 17 00:00:00 2001 From: Yevgeny Petrilin Date: Tue, 17 Jan 2012 22:54:55 +0000 Subject: [PATCH 623/820] mlx4_en: set number of rx rings used by RSS using ethtool Value must be a power of 2 due to HW limitation. Driver supports only 'equal' mode in ethtool and can't be set by using weights. Signed-off-by: Amir Vadai Reviewed-by: Yevgeny Petrilin Signed-off-by: David S. Miller --- .../net/ethernet/mellanox/mlx4/en_ethtool.c | 93 +++++++++++++++++++ drivers/net/ethernet/mellanox/mlx4/en_main.c | 6 +- drivers/net/ethernet/mellanox/mlx4/en_rx.c | 8 +- drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 2 +- 4 files changed, 102 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c index 7dbc6a230779..53c66869aecd 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c @@ -479,6 +479,95 @@ static void mlx4_en_get_ringparam(struct net_device *dev, param->tx_pending = priv->tx_ring[0].size; } +static u32 mlx4_en_get_rxfh_indir_size(struct net_device *dev) +{ + struct mlx4_en_priv *priv = netdev_priv(dev); + + return priv->rx_ring_num; +} + +static int mlx4_en_get_rxfh_indir(struct net_device *dev, u32 *ring_index) +{ + struct mlx4_en_priv *priv = netdev_priv(dev); + struct mlx4_en_rss_map *rss_map = &priv->rss_map; + int rss_rings; + size_t n = priv->rx_ring_num; + int err = 0; + + rss_rings = priv->prof->rss_rings ?: priv->rx_ring_num; + + while (n--) { + ring_index[n] = rss_map->qps[n % rss_rings].qpn - + rss_map->base_qpn; + } + + return err; +} + +static int mlx4_en_set_rxfh_indir(struct net_device *dev, + const u32 *ring_index) +{ + struct mlx4_en_priv *priv = netdev_priv(dev); + struct mlx4_en_dev *mdev = priv->mdev; + int port_up = 0; + int err = 0; + int i; + int rss_rings = 0; + + /* Calculate RSS table size and make sure flows are spread evenly + * between rings + */ + for (i = 0; i < priv->rx_ring_num; i++) { + if (i > 0 && !ring_index[i] && !rss_rings) + rss_rings = i; + + if (ring_index[i] != (i % (rss_rings ?: priv->rx_ring_num))) + return -EINVAL; + } + + if (!rss_rings) + rss_rings = priv->rx_ring_num; + + /* RSS table size must be an order of 2 */ + if (!is_power_of_2(rss_rings)) + return -EINVAL; + + mutex_lock(&mdev->state_lock); + if (priv->port_up) { + port_up = 1; + mlx4_en_stop_port(dev); + } + + priv->prof->rss_rings = rss_rings; + + if (port_up) { + err = mlx4_en_start_port(dev); + if (err) + en_err(priv, "Failed starting port\n"); + } + + mutex_unlock(&mdev->state_lock); + return err; +} + +static int mlx4_en_get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *cmd, + u32 *rule_locs) +{ + struct mlx4_en_priv *priv = netdev_priv(dev); + int err = 0; + + switch (cmd->cmd) { + case ETHTOOL_GRXRINGS: + cmd->data = priv->rx_ring_num; + break; + default: + err = -EOPNOTSUPP; + break; + } + + return err; +} + const struct ethtool_ops mlx4_en_ethtool_ops = { .get_drvinfo = mlx4_en_get_drvinfo, .get_settings = mlx4_en_get_settings, @@ -498,6 +587,10 @@ const struct ethtool_ops mlx4_en_ethtool_ops = { .set_pauseparam = mlx4_en_set_pauseparam, .get_ringparam = mlx4_en_get_ringparam, .set_ringparam = mlx4_en_set_ringparam, + .get_rxnfc = mlx4_en_get_rxnfc, + .get_rxfh_indir_size = mlx4_en_get_rxfh_indir_size, + .get_rxfh_indir = mlx4_en_get_rxfh_indir, + .set_rxfh_indir = mlx4_en_set_rxfh_indir, }; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_main.c b/drivers/net/ethernet/mellanox/mlx4/en_main.c index a06096fcc0b8..2097a7d3c5b8 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_main.c @@ -62,10 +62,6 @@ static const char mlx4_en_version[] = * Device scope module parameters */ - -/* Enable RSS TCP traffic */ -MLX4_EN_PARM_INT(tcp_rss, 1, - "Enable RSS for incomming TCP traffic or disabled (0)"); /* Enable RSS UDP traffic */ MLX4_EN_PARM_INT(udp_rss, 1, "Enable RSS for incomming UDP traffic or disabled (0)"); @@ -104,7 +100,6 @@ static int mlx4_en_get_profile(struct mlx4_en_dev *mdev) struct mlx4_en_profile *params = &mdev->profile; int i; - params->tcp_rss = tcp_rss; params->udp_rss = udp_rss; if (params->udp_rss && !(mdev->dev->caps.flags & MLX4_DEV_CAP_FLAG_UDP_RSS)) { @@ -120,6 +115,7 @@ static int mlx4_en_get_profile(struct mlx4_en_dev *mdev) params->prof[i].rx_ring_size = MLX4_EN_DEF_RX_RING_SIZE; params->prof[i].tx_ring_num = MLX4_EN_NUM_TX_RINGS + (!!pfcrx) * MLX4_EN_NUM_PPP_RINGS; + params->prof[i].rss_rings = 0; } return 0; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c index e8d6ad2dce0a..971d4b6b8dfe 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c @@ -853,6 +853,7 @@ int mlx4_en_config_rss_steer(struct mlx4_en_priv *priv) struct mlx4_en_rss_map *rss_map = &priv->rss_map; struct mlx4_qp_context context; struct mlx4_rss_context *rss_context; + int rss_rings; void *ptr; u8 rss_mask = (MLX4_RSS_IPV4 | MLX4_RSS_TCP_IPV4 | MLX4_RSS_IPV6 | MLX4_RSS_TCP_IPV6); @@ -893,10 +894,15 @@ int mlx4_en_config_rss_steer(struct mlx4_en_priv *priv) mlx4_en_fill_qp_context(priv, 0, 0, 0, 1, priv->base_qpn, priv->rx_ring[0].cqn, &context); + if (!priv->prof->rss_rings || priv->prof->rss_rings > priv->rx_ring_num) + rss_rings = priv->rx_ring_num; + else + rss_rings = priv->prof->rss_rings; + ptr = ((void *) &context) + offsetof(struct mlx4_qp_context, pri_path) + MLX4_RSS_OFFSET_IN_QPC_PRI_PATH; rss_context = ptr; - rss_context->base_qpn = cpu_to_be32(ilog2(priv->rx_ring_num) << 24 | + rss_context->base_qpn = cpu_to_be32(ilog2(rss_rings) << 24 | (rss_map->base_qpn)); rss_context->default_qpn = cpu_to_be32(rss_map->base_qpn); if (priv->mdev->profile.udp_rss) { diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index f2a8e65f5f88..f4d189a1290e 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -325,11 +325,11 @@ struct mlx4_en_port_profile { u8 rx_ppp; u8 tx_pause; u8 tx_ppp; + int rss_rings; }; struct mlx4_en_profile { int rss_xor; - int tcp_rss; int udp_rss; u8 rss_mask; u32 active_ports; From 3a4adef5c0adbbd0d898578e13ba1adbbdecbbd5 Mon Sep 17 00:00:00 2001 From: Roopa Prabhu Date: Wed, 18 Jan 2012 04:23:55 +0000 Subject: [PATCH 624/820] enic: This patch adds pci id 0x71 for SRIOV VF's This patch adds pci id 0x71 for sriov VF's. Signed-off-by: Roopa Prabhu Signed-off-by: Christian Benvenuti Signed-off-by: Sujith Sankar Signed-off-by: David S. Miller --- drivers/net/ethernet/cisco/enic/enic.h | 2 +- drivers/net/ethernet/cisco/enic/enic_main.c | 7 +++++++ 2 files changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/cisco/enic/enic.h b/drivers/net/ethernet/cisco/enic/enic.h index fe0c29acdbe6..ee93a2087fe6 100644 --- a/drivers/net/ethernet/cisco/enic/enic.h +++ b/drivers/net/ethernet/cisco/enic/enic.h @@ -32,7 +32,7 @@ #define DRV_NAME "enic" #define DRV_DESCRIPTION "Cisco VIC Ethernet NIC Driver" -#define DRV_VERSION "2.1.1.28" +#define DRV_VERSION "2.1.1.31" #define DRV_COPYRIGHT "Copyright 2008-2011 Cisco Systems, Inc" #define ENIC_BARS_MAX 6 diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index 2fd9db4b1be5..a634f58976af 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -57,11 +57,13 @@ #define PCI_DEVICE_ID_CISCO_VIC_ENET 0x0043 /* ethernet vnic */ #define PCI_DEVICE_ID_CISCO_VIC_ENET_DYN 0x0044 /* enet dynamic vnic */ +#define PCI_DEVICE_ID_CISCO_VIC_ENET_VF 0x0071 /* enet SRIOV VF */ /* Supported devices */ static DEFINE_PCI_DEVICE_TABLE(enic_id_table) = { { PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET) }, { PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET_DYN) }, + { PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET_VF) }, { 0, } /* end of table */ }; @@ -132,6 +134,11 @@ int enic_sriov_enabled(struct enic *enic) return (enic->priv_flags & ENIC_SRIOV_ENABLED) ? 1 : 0; } +static int enic_is_sriov_vf(struct enic *enic) +{ + return enic->pdev->device == PCI_DEVICE_ID_CISCO_VIC_ENET_VF; +} + int enic_is_valid_vf(struct enic *enic, int vf) { #ifdef CONFIG_PCI_IOV From 7335903cba523fc2fd801a178a0e4fcf8ccae756 Mon Sep 17 00:00:00 2001 From: Roopa Prabhu Date: Wed, 18 Jan 2012 04:24:02 +0000 Subject: [PATCH 625/820] enic: Add sriov vf device id checks in port profile code This patch adds checks for sriov vf's in enic port profile handling code. sriov vf's are same as dynamic vnics but with a different device id. Signed-off-by: Roopa Prabhu Signed-off-by: Christian Benvenuti Signed-off-by: Sujith Sankar Signed-off-by: David S. Miller --- drivers/net/ethernet/cisco/enic/enic_main.c | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index a634f58976af..9f37d92fc76d 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -444,7 +444,7 @@ static void enic_mtu_check(struct enic *enic) if (mtu && mtu != enic->port_mtu) { enic->port_mtu = mtu; - if (enic_is_dynamic(enic)) { + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) { mtu = max_t(int, ENIC_MIN_MTU, min_t(int, ENIC_MAX_MTU, mtu)); if (mtu != netdev->mtu) @@ -856,7 +856,7 @@ static int enic_set_mac_addr(struct net_device *netdev, char *addr) { struct enic *enic = netdev_priv(netdev); - if (enic_is_dynamic(enic)) { + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) { if (!is_valid_ether_addr(addr) && !is_zero_ether_addr(addr)) return -EADDRNOTAVAIL; } else { @@ -1615,7 +1615,7 @@ static int enic_open(struct net_device *netdev) for (i = 0; i < enic->rq_count; i++) vnic_rq_enable(&enic->rq[i]); - if (!enic_is_dynamic(enic)) + if (!enic_is_dynamic(enic) && !enic_is_sriov_vf(enic)) enic_dev_add_station_addr(enic); enic_set_rx_mode(netdev); @@ -1666,7 +1666,7 @@ static int enic_stop(struct net_device *netdev) netif_carrier_off(netdev); netif_tx_disable(netdev); - if (!enic_is_dynamic(enic)) + if (!enic_is_dynamic(enic) && !enic_is_sriov_vf(enic)) enic_dev_del_station_addr(enic); for (i = 0; i < enic->wq_count; i++) { @@ -1703,7 +1703,7 @@ static int enic_change_mtu(struct net_device *netdev, int new_mtu) if (new_mtu < ENIC_MIN_MTU || new_mtu > ENIC_MAX_MTU) return -EINVAL; - if (enic_is_dynamic(enic)) + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) return -EOPNOTSUPP; if (running) @@ -2433,7 +2433,7 @@ static int __devinit enic_probe(struct pci_dev *pdev, * called later by an upper layer. */ - if (!enic_is_dynamic(enic)) { + if (!enic_is_dynamic(enic) && !enic_is_sriov_vf(enic)) { err = vnic_dev_init(enic->vdev, 0); if (err) { dev_err(dev, "vNIC dev init failed, aborting\n"); @@ -2467,8 +2467,7 @@ static int __devinit enic_probe(struct pci_dev *pdev, (void)enic_change_mtu(netdev, enic->port_mtu); #ifdef CONFIG_PCI_IOV - if (enic_is_dynamic(enic) && pdev->is_virtfn && - is_zero_ether_addr(enic->mac_addr)) + if (enic_is_sriov_vf(enic) && is_zero_ether_addr(enic->mac_addr)) random_ether_addr(enic->mac_addr); #endif @@ -2481,7 +2480,7 @@ static int __devinit enic_probe(struct pci_dev *pdev, enic->tx_coalesce_usecs = enic->config.intr_timer_usec; enic->rx_coalesce_usecs = enic->tx_coalesce_usecs; - if (enic_is_dynamic(enic)) + if (enic_is_dynamic(enic) || enic_is_sriov_vf(enic)) netdev->netdev_ops = &enic_netdev_dynamic_ops; else netdev->netdev_ops = &enic_netdev_ops; From ca2b721de25b9893310fd695a38b495d14c70ff3 Mon Sep 17 00:00:00 2001 From: Roopa Prabhu Date: Wed, 18 Jan 2012 04:24:07 +0000 Subject: [PATCH 626/820] enic: rearrange some of the port profile code This patch rearranges some of the port profile code in enic_probe. It moves out some lines of port profile related code currently inside CONFIG_PCI_IOV. This is only done to move all port profile related code together so that it can help isolate the port profile handling code under a separate #ifdef in our internal build scripts. Signed-off-by: Roopa Prabhu Signed-off-by: Christian Benvenuti Signed-off-by: Sujith Sankar Signed-off-by: David S. Miller --- drivers/net/ethernet/cisco/enic/enic_main.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index 9f37d92fc76d..03b0577bd619 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -2270,10 +2270,10 @@ static int __devinit enic_probe(struct pci_dev *pdev, int using_dac = 0; unsigned int i; int err; - int num_pps = 1; #ifdef CONFIG_PCI_IOV int pos = 0; #endif + int num_pps; /* Allocate net device structure and initialize. Private * instance data is initialized to zero. @@ -2380,17 +2380,17 @@ static int __devinit enic_probe(struct pci_dev *pdev, goto err_out_vnic_unregister; } enic->priv_flags |= ENIC_SRIOV_ENABLED; - num_pps = enic->num_vfs; } } - #endif + + num_pps = enic->num_vfs ? enic->num_vfs : 1; /* Allocate structure for port profiles */ enic->pp = kcalloc(num_pps, sizeof(*enic->pp), GFP_KERNEL); if (!enic->pp) { pr_err("port profile alloc failed, aborting\n"); err = -ENOMEM; - goto err_out_disable_sriov; + goto err_out_disable_sriov_pp; } /* Issue device open to get device in known state @@ -2399,7 +2399,7 @@ static int __devinit enic_probe(struct pci_dev *pdev, err = enic_dev_open(enic); if (err) { dev_err(dev, "vNIC dev open failed, aborting\n"); - goto err_out_free_pp; + goto err_out_disable_sriov; } /* Setup devcmd lock @@ -2522,9 +2522,9 @@ err_out_dev_deinit: enic_dev_deinit(enic); err_out_dev_close: vnic_dev_close(enic->vdev); -err_out_free_pp: - kfree(enic->pp); err_out_disable_sriov: + kfree(enic->pp); +err_out_disable_sriov_pp: #ifdef CONFIG_PCI_IOV if (enic_sriov_enabled(enic)) { pci_disable_sriov(pdev); From 35d87e3325e91ec6bc446bf71f50eccc9a7b3b8f Mon Sep 17 00:00:00 2001 From: Roopa Prabhu Date: Wed, 18 Jan 2012 04:24:12 +0000 Subject: [PATCH 627/820] enic: fix location of vnic dev unregister in enic_probe cleanup code The vnic_dev_unregister is erroneously under CONFIG_PCI_IOV. This patch moves it out of CONFIG_PCI_IOV Signed-off-by: Roopa Prabhu Signed-off-by: Christian Benvenuti Signed-off-by: Sujith Sankar Signed-off-by: David S. Miller --- drivers/net/ethernet/cisco/enic/enic_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index 03b0577bd619..045c468ac8ef 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -2531,8 +2531,8 @@ err_out_disable_sriov_pp: enic->priv_flags &= ~ENIC_SRIOV_ENABLED; } err_out_vnic_unregister: - vnic_dev_unregister(enic->vdev); #endif + vnic_dev_unregister(enic->vdev); err_out_iounmap: enic_iounmap(enic); err_out_release_regions: From 013d97e9da1877f1334aa8ff3a19921ebbfe99b5 Mon Sep 17 00:00:00 2001 From: Francesco Ruggeri Date: Mon, 16 Jan 2012 10:40:10 +0000 Subject: [PATCH 628/820] net: race condition in ipv6 forwarding and disable_ipv6 parameters There is a race condition in addrconf_sysctl_forward() and addrconf_sysctl_disable(). These functions change idev->cnf.forwarding (resp. idev->cnf.disable_ipv6) and then try to grab the rtnl lock before performing any actions. If that fails they restore the original value and restart the syscall. This creates race conditions if ipv6 code tries to access these parameters, or if multiple instances try to do the same operation. As an example of the former, if __ipv6_ifa_notify() finds a 0 in idev->cnf.forwarding when invoked by addrconf_ifdown() it may not free anycast addresses, ultimately resulting in the net_device not being freed. This patch reads the user parameters into a temporary location and only writes the actual parameters when the rtnl lock is acquired. Tested in 2.6.38.8. Signed-off-by: Francesco Ruggeri Signed-off-by: David S. Miller --- net/ipv6/addrconf.c | 61 +++++++++++++++++++++++++++++---------------- 1 file changed, 40 insertions(+), 21 deletions(-) diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index a225d5ee3c2f..c02280a4d126 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -502,29 +502,31 @@ static void addrconf_forward_change(struct net *net, __s32 newf) rcu_read_unlock(); } -static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int old) +static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int newf) { struct net *net; + int old; + + if (!rtnl_trylock()) + return restart_syscall(); net = (struct net *)table->extra2; - if (p == &net->ipv6.devconf_dflt->forwarding) - return 0; + old = *p; + *p = newf; - if (!rtnl_trylock()) { - /* Restore the original values before restarting */ - *p = old; - return restart_syscall(); + if (p == &net->ipv6.devconf_dflt->forwarding) { + rtnl_unlock(); + return 0; } if (p == &net->ipv6.devconf_all->forwarding) { - __s32 newf = net->ipv6.devconf_all->forwarding; net->ipv6.devconf_dflt->forwarding = newf; addrconf_forward_change(net, newf); - } else if ((!*p) ^ (!old)) + } else if ((!newf) ^ (!old)) dev_forward_change((struct inet6_dev *)table->extra1); rtnl_unlock(); - if (*p) + if (newf) rt6_purge_dflt_routers(net); return 1; } @@ -4260,9 +4262,17 @@ int addrconf_sysctl_forward(ctl_table *ctl, int write, int *valp = ctl->data; int val = *valp; loff_t pos = *ppos; + ctl_table lctl; int ret; - ret = proc_dointvec(ctl, write, buffer, lenp, ppos); + /* + * ctl->data points to idev->cnf.forwarding, we should + * not modify it until we get the rtnl lock. + */ + lctl = *ctl; + lctl.data = &val; + + ret = proc_dointvec(&lctl, write, buffer, lenp, ppos); if (write) ret = addrconf_fixup_forwarding(ctl, valp, val); @@ -4300,26 +4310,27 @@ static void addrconf_disable_change(struct net *net, __s32 newf) rcu_read_unlock(); } -static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int old) +static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf) { struct net *net; + int old; + + if (!rtnl_trylock()) + return restart_syscall(); net = (struct net *)table->extra2; + old = *p; + *p = newf; - if (p == &net->ipv6.devconf_dflt->disable_ipv6) + if (p == &net->ipv6.devconf_dflt->disable_ipv6) { + rtnl_unlock(); return 0; - - if (!rtnl_trylock()) { - /* Restore the original values before restarting */ - *p = old; - return restart_syscall(); } if (p == &net->ipv6.devconf_all->disable_ipv6) { - __s32 newf = net->ipv6.devconf_all->disable_ipv6; net->ipv6.devconf_dflt->disable_ipv6 = newf; addrconf_disable_change(net, newf); - } else if ((!*p) ^ (!old)) + } else if ((!newf) ^ (!old)) dev_disable_change((struct inet6_dev *)table->extra1); rtnl_unlock(); @@ -4333,9 +4344,17 @@ int addrconf_sysctl_disable(ctl_table *ctl, int write, int *valp = ctl->data; int val = *valp; loff_t pos = *ppos; + ctl_table lctl; int ret; - ret = proc_dointvec(ctl, write, buffer, lenp, ppos); + /* + * ctl->data points to idev->cnf.disable_ipv6, we should + * not modify it until we get the rtnl lock. + */ + lctl = *ctl; + lctl.data = &val; + + ret = proc_dointvec(&lctl, write, buffer, lenp, ppos); if (write) ret = addrconf_disable_ipv6(ctl, valp, val); From 2b53d07891630dead46d65c8f896955fd3ae0302 Mon Sep 17 00:00:00 2001 From: Darren Hart Date: Mon, 16 Jan 2012 09:50:19 +0000 Subject: [PATCH 629/820] pch_gbe: Do not abort probe on bad MAC If the MAC is invalid or not implemented, do not abort the probe. Issue a warning and prevent bringing the interface up until a MAC is set manually (via ifconfig $IFACE hw ether $MAC). Tested on two platforms, one with a valid MAC, the other without a MAC. The real MAC is used if present, the interface fails to come up until the MAC is set on the other. They successfully get an IP over DHCP and pass a simple ping and login over ssh test. This is meant to allow the Inforce SYS940X development board: http://www.inforcecomputing.com/SYS940X_ECX.html (and others suffering from a missing MAC) to work with the mainline kernel. Without this patch, the probe will fail and the interface will not be created, preventing the user from configuring the MAC manually. This does not make any attempt to address a missing or invalid MAC for the pch_phub driver. Signed-off-by: Darren Hart CC: Arjan van de Ven CC: Alan Cox CC: Tomoya MORINAGA CC: Jeff Kirsher CC: "David S. Miller" CC: Paul Gortmaker CC: Jon Mason CC: netdev@vger.kernel.org CC: Mark Brown CC: David Laight CC: Joe Perches Signed-off-by: David S. Miller --- .../ethernet/oki-semi/pch_gbe/pch_gbe_main.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c index 964e9c0948bc..3ead111111e1 100644 --- a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c +++ b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c @@ -1745,6 +1745,12 @@ int pch_gbe_up(struct pch_gbe_adapter *adapter) struct pch_gbe_rx_ring *rx_ring = adapter->rx_ring; int err; + /* Ensure we have a valid MAC */ + if (!is_valid_ether_addr(adapter->hw.mac.addr)) { + pr_err("Error: Invalid MAC address\n"); + return -EINVAL; + } + /* hardware has been reset, we need to reload some things */ pch_gbe_set_multi(netdev); @@ -2468,9 +2474,14 @@ static int pch_gbe_probe(struct pci_dev *pdev, memcpy(netdev->dev_addr, adapter->hw.mac.addr, netdev->addr_len); if (!is_valid_ether_addr(netdev->dev_addr)) { - dev_err(&pdev->dev, "Invalid MAC Address\n"); - ret = -EIO; - goto err_free_adapter; + /* + * If the MAC is invalid (or just missing), display a warning + * but do not abort setting up the device. pch_gbe_up will + * prevent the interface from being brought up until a valid MAC + * is set. + */ + dev_err(&pdev->dev, "Invalid MAC address, " + "interface disabled.\n"); } setup_timer(&adapter->watchdog_timer, pch_gbe_watchdog, (unsigned long)adapter); From 88a4412b798236bfdd9284d5c251d76679f944e1 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 18 Jan 2012 17:13:47 -0600 Subject: [PATCH 630/820] [CIFS] Fix build break with multiuser patch when LANMAN disabled CC: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsglob.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index ba53c1c6c6cc..76e7d8b6da17 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -879,6 +879,8 @@ require use of the stronger protocol */ #define CIFSSEC_MASK 0xB70B7 /* current flags supported if weak */ #endif /* UPCALL */ #else /* do not allow weak pw hash */ +#define CIFSSEC_MUST_LANMAN 0 +#define CIFSSEC_MUST_PLNTXT 0 #ifdef CONFIG_CIFS_UPCALL #define CIFSSEC_MASK 0x8F08F /* flags supported if no weak allowed */ #else From 9ef9b20bd0eef609f07960a997c13cab8fe15d2e Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Thu, 19 Jan 2012 10:24:31 +1100 Subject: [PATCH 631/820] xen: using EXPORT_SYMBOL requires including export.h Fix these warnings: drivers/xen/biomerge.c:14:1: warning: data definition has no type or storage class [enabled by default] drivers/xen/biomerge.c:14:1: warning: type defaults to 'int' in declaration of 'EXPORT_SYMBOL' [-Wimplicit-int] drivers/xen/biomerge.c:14:1: warning: parameter names (without types) in function declaration [enabled by default] And this build error: ERROR: "xen_biovec_phys_mergeable" [drivers/block/nvme.ko] undefined! Signed-off-by: Stephen Rothwell Signed-off-by: Linus Torvalds --- drivers/xen/biomerge.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/xen/biomerge.c b/drivers/xen/biomerge.c index 18c1bb6ffce3..0edb91c0de6b 100644 --- a/drivers/xen/biomerge.c +++ b/drivers/xen/biomerge.c @@ -1,5 +1,6 @@ #include #include +#include #include bool xen_biovec_phys_mergeable(const struct bio_vec *vec1, From 93c3d65b28bab6da520c2add9cb387a0303f8b2d Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 18 Jan 2012 15:41:27 -0800 Subject: [PATCH 632/820] nvme: fix merge error due to change of 'make_request_fn' fn type The type of 'make_request_fn' changed in 5a7bbad27a4 ("block: remove support for bio remapping from ->make_request"), but the merge of the nvme driver didn't take that into account, and as a result the driver would compile with a warning: drivers/block/nvme.c: In function 'nvme_alloc_ns': drivers/block/nvme.c:1336:2: warning: passing argument 2 of 'blk_queue_make_request' from incompatible pointer type [enabled by default] include/linux/blkdev.h:830:13: note: expected 'void (*)(struct request_queue *, struct bio *)' but argument is of type 'int (*)(struct request_queue *, struct bio *)' It's benign, but the warning is annoying. Reported-by: Stephen Rothwell Cc: Matthew Wilcox Signed-off-by: Linus Torvalds --- drivers/block/nvme.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/drivers/block/nvme.c b/drivers/block/nvme.c index f4996b0e4b1a..c1dc4d86c221 100644 --- a/drivers/block/nvme.c +++ b/drivers/block/nvme.c @@ -613,11 +613,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, return result; } -/* - * NB: return value of non-zero would mean that we were a stacking driver. - * make_request must always succeed. - */ -static int nvme_make_request(struct request_queue *q, struct bio *bio) +static void nvme_make_request(struct request_queue *q, struct bio *bio) { struct nvme_ns *ns = q->queuedata; struct nvme_queue *nvmeq = get_nvmeq(ns->dev); @@ -634,8 +630,6 @@ static int nvme_make_request(struct request_queue *q, struct bio *bio) spin_unlock_irq(&nvmeq->q_lock); put_nvmeq(nvmeq); - - return 0; } static irqreturn_t nvme_process_cq(struct nvme_queue *nvmeq) From be405411f712489f2f780ab085e1069e8fb85f19 Mon Sep 17 00:00:00 2001 From: Stefan Berger Date: Tue, 17 Jan 2012 22:07:30 -0500 Subject: [PATCH 633/820] tpm: fix (ACPI S3) suspend regression This patch fixes an (ACPI S3) suspend regression introduced in commit 68d6e6713fcb ("tpm: Introduce function to poll for result of self test") and occurring with an Infineon TPM and tpm_tis and tpm_infineon drivers active. The suspend problem occurred if the TPM was disabled and/or deactivated and therefore the TPM_PCRRead checking the result of the (asynchronous) self test returned an error code which then caused the tpm_tis driver to become inactive and this then seemed to have negatively influenced the suspend support by the tpm_infineon driver... Besides that the tpm_tis drive may stay active even if the TPM is disabled and/or deactivated. Signed-off-by: Stefan Berger Tested-by: Rafael J. Wysocki Acked-by: Rajiv Andrade Signed-off-by: Linus Torvalds --- drivers/char/tpm/tpm.c | 9 +++++++++ drivers/char/tpm/tpm.h | 3 +++ 2 files changed, 12 insertions(+) diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c index 6a8771f47a55..32362cf35b8d 100644 --- a/drivers/char/tpm/tpm.c +++ b/drivers/char/tpm/tpm.c @@ -846,6 +846,15 @@ int tpm_do_selftest(struct tpm_chip *chip) do { rc = __tpm_pcr_read(chip, 0, digest); + if (rc == TPM_ERR_DISABLED || rc == TPM_ERR_DEACTIVATED) { + dev_info(chip->dev, + "TPM is disabled/deactivated (0x%X)\n", rc); + /* TPM is disabled and/or deactivated; driver can + * proceed and TPM does handle commands for + * suspend/resume correctly + */ + return 0; + } if (rc != TPM_WARN_DOING_SELFTEST) return rc; msleep(delay_msec); diff --git a/drivers/char/tpm/tpm.h b/drivers/char/tpm/tpm.h index 8c1df302fbb6..010547138281 100644 --- a/drivers/char/tpm/tpm.h +++ b/drivers/char/tpm/tpm.h @@ -39,6 +39,9 @@ enum tpm_addr { }; #define TPM_WARN_DOING_SELFTEST 0x802 +#define TPM_ERR_DEACTIVATED 0x6 +#define TPM_ERR_DISABLED 0x7 + #define TPM_HEADER_SIZE 10 extern ssize_t tpm_show_pubek(struct device *, struct device_attribute *attr, char *); From c56001879bc091eee0c7a8e6e94ea0bea63c3012 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 18 Jan 2012 17:19:11 -0600 Subject: [PATCH 634/820] [CIFS] ACL and FSCACHE support no longer EXPERIMENTAL CIFS ACL support and FSCACHE support have been in long enough to be no longer considered experimental. Remove obsolete Kconfig dependency. Signed-off-by: Steve French Acked-by: Jeff Layton --- fs/cifs/Kconfig | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index f66cc1625150..0554b00a7b33 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -140,7 +140,6 @@ config CIFS_DFS_UPCALL config CIFS_FSCACHE bool "Provide CIFS client caching support (EXPERIMENTAL)" - depends on EXPERIMENTAL depends on CIFS=m && FSCACHE || CIFS=y && FSCACHE=y help Makes CIFS FS-Cache capable. Say Y here if you want your CIFS data @@ -149,7 +148,7 @@ config CIFS_FSCACHE config CIFS_ACL bool "Provide CIFS ACL support (EXPERIMENTAL)" - depends on EXPERIMENTAL && CIFS_XATTR && KEYS + depends on CIFS_XATTR && KEYS help Allows to fetch CIFS/NTFS ACL from the server. The DACL blob is handed over to the application/caller. From b924551bed09f61b64f21bffe241afc5526b091a Mon Sep 17 00:00:00 2001 From: Jiri Bohac Date: Wed, 18 Jan 2012 12:24:54 +0000 Subject: [PATCH 635/820] bonding: fix enslaving in alb mode when link down bond_alb_init_slave() is called from bond_enslave() and sets the slave's MAC address. This is done differently for TLB and ALB modes. bond->alb_info.rlb_enabled is used to discriminate between the two modes but this flag may be uninitialized if the slave is being enslaved prior to calling bond_open() -> bond_alb_initialize() on the master. It turns out all the callers of alb_set_slave_mac_addr() pass bond->alb_info.rlb_enabled as the hw parameter. This patch cleans up the unnecessary parameter of alb_set_slave_mac_addr() and makes the function decide based on the bonding mode instead, which fixes the above problem. Reported-by: Narendra K Signed-off-by: Jiri Bohac Signed-off-by: Jay Vosburgh Signed-off-by: David S. Miller --- drivers/net/bonding/bond_alb.c | 27 +++++++++------------------ 1 file changed, 9 insertions(+), 18 deletions(-) diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index 342626f4bc46..f820b26b9db3 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -909,16 +909,12 @@ static void alb_send_learning_packets(struct slave *slave, u8 mac_addr[]) } } -/* hw is a boolean parameter that determines whether we should try and - * set the hw address of the device as well as the hw address of the - * net_device - */ -static int alb_set_slave_mac_addr(struct slave *slave, u8 addr[], int hw) +static int alb_set_slave_mac_addr(struct slave *slave, u8 addr[]) { struct net_device *dev = slave->dev; struct sockaddr s_addr; - if (!hw) { + if (slave->bond->params.mode == BOND_MODE_TLB) { memcpy(dev->dev_addr, addr, dev->addr_len); return 0; } @@ -948,8 +944,8 @@ static void alb_swap_mac_addr(struct bonding *bond, struct slave *slave1, struct u8 tmp_mac_addr[ETH_ALEN]; memcpy(tmp_mac_addr, slave1->dev->dev_addr, ETH_ALEN); - alb_set_slave_mac_addr(slave1, slave2->dev->dev_addr, bond->alb_info.rlb_enabled); - alb_set_slave_mac_addr(slave2, tmp_mac_addr, bond->alb_info.rlb_enabled); + alb_set_slave_mac_addr(slave1, slave2->dev->dev_addr); + alb_set_slave_mac_addr(slave2, tmp_mac_addr); } @@ -1096,8 +1092,7 @@ static int alb_handle_addr_collision_on_attach(struct bonding *bond, struct slav /* Try setting slave mac to bond address and fall-through to code handling that situation below... */ - alb_set_slave_mac_addr(slave, bond->dev->dev_addr, - bond->alb_info.rlb_enabled); + alb_set_slave_mac_addr(slave, bond->dev->dev_addr); } /* The slave's address is equal to the address of the bond. @@ -1133,8 +1128,7 @@ static int alb_handle_addr_collision_on_attach(struct bonding *bond, struct slav } if (free_mac_slave) { - alb_set_slave_mac_addr(slave, free_mac_slave->perm_hwaddr, - bond->alb_info.rlb_enabled); + alb_set_slave_mac_addr(slave, free_mac_slave->perm_hwaddr); pr_warning("%s: Warning: the hw address of slave %s is in use by the bond; giving it the hw address of %s\n", bond->dev->name, slave->dev->name, @@ -1491,8 +1485,7 @@ int bond_alb_init_slave(struct bonding *bond, struct slave *slave) { int res; - res = alb_set_slave_mac_addr(slave, slave->perm_hwaddr, - bond->alb_info.rlb_enabled); + res = alb_set_slave_mac_addr(slave, slave->perm_hwaddr); if (res) { return res; } @@ -1643,8 +1636,7 @@ void bond_alb_handle_active_change(struct bonding *bond, struct slave *new_slave alb_swap_mac_addr(bond, swap_slave, new_slave); } else { /* set the new_slave to the bond mac address */ - alb_set_slave_mac_addr(new_slave, bond->dev->dev_addr, - bond->alb_info.rlb_enabled); + alb_set_slave_mac_addr(new_slave, bond->dev->dev_addr); } if (swap_slave) { @@ -1704,8 +1696,7 @@ int bond_alb_set_mac_address(struct net_device *bond_dev, void *addr) alb_swap_mac_addr(bond, swap_slave, bond->curr_active_slave); alb_fasten_mac_swap(bond, swap_slave, bond->curr_active_slave); } else { - alb_set_slave_mac_addr(bond->curr_active_slave, bond_dev->dev_addr, - bond->alb_info.rlb_enabled); + alb_set_slave_mac_addr(bond->curr_active_slave, bond_dev->dev_addr); read_lock(&bond->lock); alb_send_learning_packets(bond->curr_active_slave, bond_dev->dev_addr); From 17f1bbca1baf2f652fa9c102ec239d6abded94c1 Mon Sep 17 00:00:00 2001 From: Thomas Faber Date: Wed, 18 Jan 2012 13:45:44 +0000 Subject: [PATCH 636/820] net: ftgmac100/ftmac100: add missing interrupt.h include Fixes compilation failure of these modules due to missing irqreturn_t type for the ft(g)mac100_interrupt definition. Signed-off-by: Thomas Faber Signed-off-by: David S. Miller --- drivers/net/ethernet/faraday/ftgmac100.c | 1 + drivers/net/ethernet/faraday/ftmac100.c | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/net/ethernet/faraday/ftgmac100.c b/drivers/net/ethernet/faraday/ftgmac100.c index fb5579a3b19d..47f85c337cf7 100644 --- a/drivers/net/ethernet/faraday/ftgmac100.c +++ b/drivers/net/ethernet/faraday/ftgmac100.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/net/ethernet/faraday/ftmac100.c b/drivers/net/ethernet/faraday/ftmac100.c index a127cb2476c7..bb336a0959c9 100644 --- a/drivers/net/ethernet/faraday/ftmac100.c +++ b/drivers/net/ethernet/faraday/ftmac100.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include From 4bf1924c008dffdc154f82507b4052e49263a6f4 Mon Sep 17 00:00:00 2001 From: David Howells Date: Wed, 18 Jan 2012 10:03:54 +0000 Subject: [PATCH 637/820] MPILIB: Add a missing ENOMEM check Add a missing ENOMEM check. Signed-off-by: David Howells Acked-by: Mimi Zohar Signed-off-by: James Morris --- lib/mpi/mpicoder.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/lib/mpi/mpicoder.c b/lib/mpi/mpicoder.c index fe84bb978e3b..716802b774ea 100644 --- a/lib/mpi/mpicoder.c +++ b/lib/mpi/mpicoder.c @@ -255,6 +255,8 @@ void *mpi_get_buffer(MPI a, unsigned *nbytes, int *sign) if (!n) n++; /* avoid zero length allocation */ p = buffer = kmalloc(n, GFP_KERNEL); + if (!p) + return NULL; for (i = a->nlimbs - 1; i >= 0; i--) { alimb = a->d[i]; From 90a4c0f51e8e44111a926be6f4c87af3938a79c3 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 18 Jan 2012 19:26:11 -0800 Subject: [PATCH 638/820] uml: fix compile for x86-64 Randy Dunlap reports that we get arch/x86/um/shared/sysdep/ptrace.h:7:20: error: redefinition of 'regs_return_value' arch/x86/um/shared/sysdep/ptrace.h:7:20: note: previous definition of 'regs_return_value' was here when compiling UML for x86-64. Stephen Rothwell root-caused it and says: "Caused by commit d7e7528bcd45 ("Audit: push audit success and retcode into arch ptrace.h") (another patch that was never in linux-next :-(). This file now needs protection against double inclusion." so let's do as the man says. Reported-by: Randy Dunlap Analyzed-by: Stephen Rothwell Signed-off-by: Linus Torvalds --- arch/x86/um/shared/sysdep/ptrace.h | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/x86/um/shared/sysdep/ptrace.h b/arch/x86/um/shared/sysdep/ptrace.h index 5ef9344a8b24..2bbe1ec2d96a 100644 --- a/arch/x86/um/shared/sysdep/ptrace.h +++ b/arch/x86/um/shared/sysdep/ptrace.h @@ -1,3 +1,6 @@ +#ifndef __SYSDEP_X86_PTRACE_H +#define __SYSDEP_X86_PTRACE_H + #ifdef __i386__ #include "ptrace_32.h" #else @@ -8,3 +11,5 @@ static inline long regs_return_value(struct uml_pt_regs *regs) { return UPT_SYSCALL_RET(regs); } + +#endif /* __SYSDEP_X86_PTRACE_H */ From acbbb76a26648dfae6fed0989879e40d75692bfc Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 18 Jan 2012 22:32:33 -0600 Subject: [PATCH 639/820] CIFS: Rename *UCS* functions to *UTF16* to reflect the unicode encoding used by CIFS protocol. Signed-off-by: Pavel Shilovsky Acked-by: Jeff Layton Reviewed-by: Shirish Pargaonkar --- fs/cifs/cifs_unicode.c | 41 ++++++----- fs/cifs/cifs_unicode.h | 20 ++--- fs/cifs/cifsencrypt.c | 12 +-- fs/cifs/cifssmb.c | 162 +++++++++++++++++++++-------------------- fs/cifs/connect.c | 4 +- fs/cifs/readdir.c | 9 ++- fs/cifs/sess.c | 34 ++++----- fs/cifs/smbencrypt.c | 2 +- 8 files changed, 146 insertions(+), 138 deletions(-) diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index 1b2e180b018d..fbb9da951843 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -27,17 +27,17 @@ #include "cifs_debug.h" /* - * cifs_ucs2_bytes - how long will a string be after conversion? - * @ucs - pointer to input string + * cifs_utf16_bytes - how long will a string be after conversion? + * @utf16 - pointer to input string * @maxbytes - don't go past this many bytes of input string * @codepage - destination codepage * - * Walk a ucs2le string and return the number of bytes that the string will + * Walk a utf16le string and return the number of bytes that the string will * be after being converted to the given charset, not including any null * termination required. Don't walk past maxbytes in the source buffer. */ int -cifs_ucs2_bytes(const __le16 *from, int maxbytes, +cifs_utf16_bytes(const __le16 *from, int maxbytes, const struct nls_table *codepage) { int i; @@ -122,7 +122,7 @@ cp_convert: } /* - * cifs_from_ucs2 - convert utf16le string to local charset + * cifs_from_utf16 - convert utf16le string to local charset * @to - destination buffer * @from - source buffer * @tolen - destination buffer size (in bytes) @@ -130,7 +130,7 @@ cp_convert: * @codepage - codepage to which characters should be converted * @mapchar - should characters be remapped according to the mapchars option? * - * Convert a little-endian ucs2le string (as sent by the server) to a string + * Convert a little-endian utf16le string (as sent by the server) to a string * in the provided codepage. The tolen and fromlen parameters are to ensure * that the code doesn't walk off of the end of the buffer (which is always * a danger if the alignment of the source buffer is off). The destination @@ -139,12 +139,12 @@ cp_convert: * null terminator). * * Note that some windows versions actually send multiword UTF-16 characters - * instead of straight UCS-2. The linux nls routines however aren't able to + * instead of straight UTF16-2. The linux nls routines however aren't able to * deal with those characters properly. In the event that we get some of * those characters, they won't be translated properly. */ int -cifs_from_ucs2(char *to, const __le16 *from, int tolen, int fromlen, +cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen, const struct nls_table *codepage, bool mapchar) { int i, charlen, safelen; @@ -190,13 +190,13 @@ cifs_from_ucs2(char *to, const __le16 *from, int tolen, int fromlen, } /* - * NAME: cifs_strtoUCS() + * NAME: cifs_strtoUTF16() * * FUNCTION: Convert character string to unicode string * */ int -cifs_strtoUCS(__le16 *to, const char *from, int len, +cifs_strtoUTF16(__le16 *to, const char *from, int len, const struct nls_table *codepage) { int charlen; @@ -206,7 +206,7 @@ cifs_strtoUCS(__le16 *to, const char *from, int len, for (i = 0; len && *from; i++, from += charlen, len -= charlen) { charlen = codepage->char2uni(from, len, &wchar_to); if (charlen < 1) { - cERROR(1, "strtoUCS: char2uni of 0x%x returned %d", + cERROR(1, "strtoUTF16: char2uni of 0x%x returned %d", *from, charlen); /* A question mark */ wchar_to = 0x003f; @@ -220,7 +220,8 @@ cifs_strtoUCS(__le16 *to, const char *from, int len, } /* - * cifs_strndup_from_ucs - copy a string from wire format to the local codepage + * cifs_strndup_from_utf16 - copy a string from wire format to the local + * codepage * @src - source string * @maxlen - don't walk past this many bytes in the source string * @is_unicode - is this a unicode string? @@ -231,19 +232,19 @@ cifs_strtoUCS(__le16 *to, const char *from, int len, * error. */ char * -cifs_strndup_from_ucs(const char *src, const int maxlen, const bool is_unicode, - const struct nls_table *codepage) +cifs_strndup_from_utf16(const char *src, const int maxlen, + const bool is_unicode, const struct nls_table *codepage) { int len; char *dst; if (is_unicode) { - len = cifs_ucs2_bytes((__le16 *) src, maxlen, codepage); + len = cifs_utf16_bytes((__le16 *) src, maxlen, codepage); len += nls_nullsize(codepage); dst = kmalloc(len, GFP_KERNEL); if (!dst) return NULL; - cifs_from_ucs2(dst, (__le16 *) src, len, maxlen, codepage, + cifs_from_utf16(dst, (__le16 *) src, len, maxlen, codepage, false); } else { len = strnlen(src, maxlen); @@ -264,7 +265,7 @@ cifs_strndup_from_ucs(const char *src, const int maxlen, const bool is_unicode, * names are little endian 16 bit Unicode on the wire */ int -cifsConvertToUCS(__le16 *target, const char *source, int srclen, +cifsConvertToUTF16(__le16 *target, const char *source, int srclen, const struct nls_table *cp, int mapChars) { int i, j, charlen; @@ -273,7 +274,7 @@ cifsConvertToUCS(__le16 *target, const char *source, int srclen, wchar_t tmp; if (!mapChars) - return cifs_strtoUCS(target, source, PATH_MAX, cp); + return cifs_strtoUTF16(target, source, PATH_MAX, cp); for (i = 0, j = 0; i < srclen; j++) { src_char = source[i]; @@ -281,7 +282,7 @@ cifsConvertToUCS(__le16 *target, const char *source, int srclen, switch (src_char) { case 0: put_unaligned(0, &target[j]); - goto ctoUCS_out; + goto ctoUTF16_out; case ':': dst_char = cpu_to_le16(UNI_COLON); break; @@ -326,7 +327,7 @@ cifsConvertToUCS(__le16 *target, const char *source, int srclen, put_unaligned(dst_char, &target[j]); } -ctoUCS_out: +ctoUTF16_out: return i; } diff --git a/fs/cifs/cifs_unicode.h b/fs/cifs/cifs_unicode.h index 6d02fd560566..a513a546700b 100644 --- a/fs/cifs/cifs_unicode.h +++ b/fs/cifs/cifs_unicode.h @@ -74,16 +74,16 @@ extern const struct UniCaseRange CifsUniLowerRange[]; #endif /* UNIUPR_NOLOWER */ #ifdef __KERNEL__ -int cifs_from_ucs2(char *to, const __le16 *from, int tolen, int fromlen, - const struct nls_table *codepage, bool mapchar); -int cifs_ucs2_bytes(const __le16 *from, int maxbytes, - const struct nls_table *codepage); -int cifs_strtoUCS(__le16 *, const char *, int, const struct nls_table *); -char *cifs_strndup_from_ucs(const char *src, const int maxlen, - const bool is_unicode, - const struct nls_table *codepage); -extern int cifsConvertToUCS(__le16 *target, const char *source, int maxlen, - const struct nls_table *cp, int mapChars); +int cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen, + const struct nls_table *codepage, bool mapchar); +int cifs_utf16_bytes(const __le16 *from, int maxbytes, + const struct nls_table *codepage); +int cifs_strtoUTF16(__le16 *, const char *, int, const struct nls_table *); +char *cifs_strndup_from_utf16(const char *src, const int maxlen, + const bool is_unicode, + const struct nls_table *codepage); +extern int cifsConvertToUTF16(__le16 *target, const char *source, int maxlen, + const struct nls_table *cp, int mapChars); #endif diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index bce99e6a4950..63c460e503b6 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -327,7 +327,7 @@ build_avpair_blob(struct cifs_ses *ses, const struct nls_table *nls_cp) attrptr->type = cpu_to_le16(NTLMSSP_AV_NB_DOMAIN_NAME); attrptr->length = cpu_to_le16(2 * dlen); blobptr = (unsigned char *)attrptr + sizeof(struct ntlmssp2_name); - cifs_strtoUCS((__le16 *)blobptr, ses->domainName, dlen, nls_cp); + cifs_strtoUTF16((__le16 *)blobptr, ses->domainName, dlen, nls_cp); return 0; } @@ -376,7 +376,7 @@ find_domain_name(struct cifs_ses *ses, const struct nls_table *nls_cp) kmalloc(attrsize + 1, GFP_KERNEL); if (!ses->domainName) return -ENOMEM; - cifs_from_ucs2(ses->domainName, + cifs_from_utf16(ses->domainName, (__le16 *)blobptr, attrsize, attrsize, nls_cp, false); break; @@ -429,7 +429,7 @@ static int calc_ntlmv2_hash(struct cifs_ses *ses, char *ntlmv2_hash, } if (len) { - len = cifs_strtoUCS((__le16 *)user, ses->user_name, len, nls_cp); + len = cifs_strtoUTF16((__le16 *)user, ses->user_name, len, nls_cp); UniStrupr(user); } else { memset(user, '\0', 2); @@ -453,8 +453,8 @@ static int calc_ntlmv2_hash(struct cifs_ses *ses, char *ntlmv2_hash, rc = -ENOMEM; return rc; } - len = cifs_strtoUCS((__le16 *)domain, ses->domainName, len, - nls_cp); + len = cifs_strtoUTF16((__le16 *)domain, ses->domainName, len, + nls_cp); rc = crypto_shash_update(&ses->server->secmech.sdeschmacmd5->shash, (char *)domain, 2 * len); @@ -473,7 +473,7 @@ static int calc_ntlmv2_hash(struct cifs_ses *ses, char *ntlmv2_hash, rc = -ENOMEM; return rc; } - len = cifs_strtoUCS((__le16 *)server, ses->serverName, len, + len = cifs_strtoUTF16((__le16 *)server, ses->serverName, len, nls_cp); rc = crypto_shash_update(&ses->server->secmech.sdeschmacmd5->shash, diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 6600aa2d2ef3..8b7794c31591 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -821,8 +821,8 @@ PsxDelete: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB add path length overrun check */ @@ -893,8 +893,8 @@ DelFileRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->fileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->fileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve check for buffer overruns BB */ @@ -938,8 +938,8 @@ RmDirRetry: return rc; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { - name_len = cifsConvertToUCS((__le16 *) pSMB->DirName, dirName, - PATH_MAX, nls_codepage, remap); + name_len = cifsConvertToUTF16((__le16 *) pSMB->DirName, dirName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve check for buffer overruns BB */ @@ -981,8 +981,8 @@ MkDirRetry: return rc; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { - name_len = cifsConvertToUCS((__le16 *) pSMB->DirName, name, - PATH_MAX, nls_codepage, remap); + name_len = cifsConvertToUTF16((__le16 *) pSMB->DirName, name, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve check for buffer overruns BB */ @@ -1030,8 +1030,8 @@ PsxCreat: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, name, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, name, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -1197,8 +1197,8 @@ OldOpenRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { count = 1; /* account for one byte pad to word boundary */ name_len = - cifsConvertToUCS((__le16 *) (pSMB->fileName + 1), - fileName, PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) (pSMB->fileName + 1), + fileName, PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve check for buffer overruns BB */ @@ -1304,8 +1304,8 @@ openRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { count = 1; /* account for one byte pad to word boundary */ name_len = - cifsConvertToUCS((__le16 *) (pSMB->fileName + 1), - fileName, PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) (pSMB->fileName + 1), + fileName, PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; pSMB->NameLength = cpu_to_le16(name_len); @@ -2649,16 +2649,16 @@ renameRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->OldFileName, fromName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->OldFileName, fromName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; pSMB->OldFileName[name_len] = 0x04; /* pad */ /* protocol requires ASCII signature byte on Unicode string */ pSMB->OldFileName[name_len + 1] = 0x00; name_len2 = - cifsConvertToUCS((__le16 *)&pSMB->OldFileName[name_len + 2], - toName, PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *)&pSMB->OldFileName[name_len+2], + toName, PATH_MAX, nls_codepage, remap); name_len2 += 1 /* trailing null */ + 1 /* Signature word */ ; name_len2 *= 2; /* convert to bytes */ } else { /* BB improve the check for buffer overruns BB */ @@ -2738,10 +2738,12 @@ int CIFSSMBRenameOpenFile(const int xid, struct cifs_tcon *pTcon, /* unicode only call */ if (target_name == NULL) { sprintf(dummy_string, "cifs%x", pSMB->hdr.Mid); - len_of_str = cifsConvertToUCS((__le16 *)rename_info->target_name, + len_of_str = + cifsConvertToUTF16((__le16 *)rename_info->target_name, dummy_string, 24, nls_codepage, remap); } else { - len_of_str = cifsConvertToUCS((__le16 *)rename_info->target_name, + len_of_str = + cifsConvertToUTF16((__le16 *)rename_info->target_name, target_name, PATH_MAX, nls_codepage, remap); } @@ -2795,17 +2797,17 @@ copyRetry: pSMB->Flags = cpu_to_le16(flags & COPY_TREE); if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { - name_len = cifsConvertToUCS((__le16 *) pSMB->OldFileName, - fromName, PATH_MAX, nls_codepage, - remap); + name_len = cifsConvertToUTF16((__le16 *) pSMB->OldFileName, + fromName, PATH_MAX, nls_codepage, + remap); name_len++; /* trailing null */ name_len *= 2; pSMB->OldFileName[name_len] = 0x04; /* pad */ /* protocol requires ASCII signature byte on Unicode string */ pSMB->OldFileName[name_len + 1] = 0x00; name_len2 = - cifsConvertToUCS((__le16 *)&pSMB->OldFileName[name_len + 2], - toName, PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *)&pSMB->OldFileName[name_len+2], + toName, PATH_MAX, nls_codepage, remap); name_len2 += 1 /* trailing null */ + 1 /* Signature word */ ; name_len2 *= 2; /* convert to bytes */ } else { /* BB improve the check for buffer overruns BB */ @@ -2861,9 +2863,9 @@ createSymLinkRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifs_strtoUCS((__le16 *) pSMB->FileName, fromName, PATH_MAX - /* find define for this maxpathcomponent */ - , nls_codepage); + cifs_strtoUTF16((__le16 *) pSMB->FileName, fromName, + /* find define for this maxpathcomponent */ + PATH_MAX, nls_codepage); name_len++; /* trailing null */ name_len *= 2; @@ -2885,9 +2887,9 @@ createSymLinkRetry: data_offset = (char *) (&pSMB->hdr.Protocol) + offset; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len_target = - cifs_strtoUCS((__le16 *) data_offset, toName, PATH_MAX - /* find define for this maxpathcomponent */ - , nls_codepage); + cifs_strtoUTF16((__le16 *) data_offset, toName, PATH_MAX + /* find define for this maxpathcomponent */ + , nls_codepage); name_len_target++; /* trailing null */ name_len_target *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -2949,8 +2951,8 @@ createHardLinkRetry: return rc; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { - name_len = cifsConvertToUCS((__le16 *) pSMB->FileName, toName, - PATH_MAX, nls_codepage, remap); + name_len = cifsConvertToUTF16((__le16 *) pSMB->FileName, toName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; @@ -2972,8 +2974,8 @@ createHardLinkRetry: data_offset = (char *) (&pSMB->hdr.Protocol) + offset; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len_target = - cifsConvertToUCS((__le16 *) data_offset, fromName, PATH_MAX, - nls_codepage, remap); + cifsConvertToUTF16((__le16 *) data_offset, fromName, + PATH_MAX, nls_codepage, remap); name_len_target++; /* trailing null */ name_len_target *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -3042,8 +3044,8 @@ winCreateHardLinkRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->OldFileName, fromName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->OldFileName, fromName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; @@ -3051,8 +3053,8 @@ winCreateHardLinkRetry: pSMB->OldFileName[name_len] = 0x04; pSMB->OldFileName[name_len + 1] = 0x00; /* pad */ name_len2 = - cifsConvertToUCS((__le16 *)&pSMB->OldFileName[name_len + 2], - toName, PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *)&pSMB->OldFileName[name_len+2], + toName, PATH_MAX, nls_codepage, remap); name_len2 += 1 /* trailing null */ + 1 /* Signature word */ ; name_len2 *= 2; /* convert to bytes */ } else { /* BB improve the check for buffer overruns BB */ @@ -3108,8 +3110,8 @@ querySymLinkRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifs_strtoUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage); + cifs_strtoUTF16((__le16 *) pSMB->FileName, searchName, + PATH_MAX, nls_codepage); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -3166,8 +3168,8 @@ querySymLinkRetry: is_unicode = false; /* BB FIXME investigate remapping reserved chars here */ - *symlinkinfo = cifs_strndup_from_ucs(data_start, count, - is_unicode, nls_codepage); + *symlinkinfo = cifs_strndup_from_utf16(data_start, + count, is_unicode, nls_codepage); if (!*symlinkinfo) rc = -ENOMEM; } @@ -3450,8 +3452,9 @@ queryAclRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, + searchName, PATH_MAX, nls_codepage, + remap); name_len++; /* trailing null */ name_len *= 2; pSMB->FileName[name_len] = 0; @@ -3537,8 +3540,8 @@ setAclRetry: return rc; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -3948,8 +3951,9 @@ QInfRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, + searchName, PATH_MAX, nls_codepage, + remap); name_len++; /* trailing null */ name_len *= 2; } else { @@ -4086,8 +4090,8 @@ QPathInfoRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, searchName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -4255,8 +4259,8 @@ UnixQPathInfoRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, searchName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -4344,8 +4348,8 @@ findFirstRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, searchName, + PATH_MAX, nls_codepage, remap); /* We can not add the asterik earlier in case it got remapped to 0xF03A as if it were part of the directory name instead of a wildcard */ @@ -4656,8 +4660,9 @@ GetInodeNumberRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, + searchName, PATH_MAX, nls_codepage, + remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -4794,9 +4799,9 @@ parse_DFS_referrals(TRANSACTION2_GET_DFS_REFER_RSP *pSMBr, rc = -ENOMEM; goto parse_DFS_referrals_exit; } - cifsConvertToUCS((__le16 *) tmp, searchName, - PATH_MAX, nls_codepage, remap); - node->path_consumed = cifs_ucs2_bytes(tmp, + cifsConvertToUTF16((__le16 *) tmp, searchName, + PATH_MAX, nls_codepage, remap); + node->path_consumed = cifs_utf16_bytes(tmp, le16_to_cpu(pSMBr->PathConsumed), nls_codepage); kfree(tmp); @@ -4809,8 +4814,8 @@ parse_DFS_referrals(TRANSACTION2_GET_DFS_REFER_RSP *pSMBr, /* copy DfsPath */ temp = (char *)ref + le16_to_cpu(ref->DfsPathOffset); max_len = data_end - temp; - node->path_name = cifs_strndup_from_ucs(temp, max_len, - is_unicode, nls_codepage); + node->path_name = cifs_strndup_from_utf16(temp, max_len, + is_unicode, nls_codepage); if (!node->path_name) { rc = -ENOMEM; goto parse_DFS_referrals_exit; @@ -4819,8 +4824,8 @@ parse_DFS_referrals(TRANSACTION2_GET_DFS_REFER_RSP *pSMBr, /* copy link target UNC */ temp = (char *)ref + le16_to_cpu(ref->NetworkAddressOffset); max_len = data_end - temp; - node->node_name = cifs_strndup_from_ucs(temp, max_len, - is_unicode, nls_codepage); + node->node_name = cifs_strndup_from_utf16(temp, max_len, + is_unicode, nls_codepage); if (!node->node_name) rc = -ENOMEM; } @@ -4873,8 +4878,9 @@ getDFSRetry: if (ses->capabilities & CAP_UNICODE) { pSMB->hdr.Flags2 |= SMBFLG2_UNICODE; name_len = - cifsConvertToUCS((__le16 *) pSMB->RequestFileName, - searchName, PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->RequestFileName, + searchName, PATH_MAX, nls_codepage, + remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -5506,8 +5512,8 @@ SetEOFRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -5796,8 +5802,8 @@ SetTimesRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -5877,8 +5883,8 @@ SetAttrLgcyRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - ConvertToUCS((__le16 *) pSMB->fileName, fileName, - PATH_MAX, nls_codepage); + ConvertToUTF16((__le16 *) pSMB->fileName, fileName, + PATH_MAX, nls_codepage); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -6030,8 +6036,8 @@ setPermsRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -6123,8 +6129,8 @@ QAllEAsRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { list_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, searchName, + PATH_MAX, nls_codepage, remap); list_len++; /* trailing null */ list_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -6301,8 +6307,8 @@ SetEARetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifsConvertToUCS((__le16 *) pSMB->FileName, fileName, - PATH_MAX, nls_codepage, remap); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fileName, + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 28f23c03da53..986709a8d903 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -3644,7 +3644,7 @@ CIFSTCon(unsigned int xid, struct cifs_ses *ses, if (ses->capabilities & CAP_UNICODE) { smb_buffer->Flags2 |= SMBFLG2_UNICODE; length = - cifs_strtoUCS((__le16 *) bcc_ptr, tree, + cifs_strtoUTF16((__le16 *) bcc_ptr, tree, 6 /* max utf8 char length in bytes */ * (/* server len*/ + 256 /* share len */), nls_codepage); bcc_ptr += 2 * length; /* convert num 16 bit words to bytes */ @@ -3699,7 +3699,7 @@ CIFSTCon(unsigned int xid, struct cifs_ses *ses, /* mostly informational -- no need to fail on error here */ kfree(tcon->nativeFileSystem); - tcon->nativeFileSystem = cifs_strndup_from_ucs(bcc_ptr, + tcon->nativeFileSystem = cifs_strndup_from_utf16(bcc_ptr, bytes_left, is_unicode, nls_codepage); diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index a090bbe6ee29..e2bbc683e018 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -647,10 +647,11 @@ static int cifs_filldir(char *find_entry, struct file *file, filldir_t filldir, name.name = scratch_buf; name.len = - cifs_from_ucs2((char *)name.name, (__le16 *)de.name, - UNICODE_NAME_MAX, - min(de.namelen, (size_t)max_len), nlt, - cifs_sb->mnt_cifs_flags & + cifs_from_utf16((char *)name.name, (__le16 *)de.name, + UNICODE_NAME_MAX, + min_t(size_t, de.namelen, + (size_t)max_len), nlt, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); name.len -= nls_nullsize(nlt); } else { diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 4ec3ee9d72cc..d85efad5765f 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -167,16 +167,16 @@ unicode_oslm_strings(char **pbcc_area, const struct nls_table *nls_cp) int bytes_ret = 0; /* Copy OS version */ - bytes_ret = cifs_strtoUCS((__le16 *)bcc_ptr, "Linux version ", 32, - nls_cp); + bytes_ret = cifs_strtoUTF16((__le16 *)bcc_ptr, "Linux version ", 32, + nls_cp); bcc_ptr += 2 * bytes_ret; - bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, init_utsname()->release, - 32, nls_cp); + bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, init_utsname()->release, + 32, nls_cp); bcc_ptr += 2 * bytes_ret; bcc_ptr += 2; /* trailing null */ - bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, CIFS_NETWORK_OPSYS, - 32, nls_cp); + bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, CIFS_NETWORK_OPSYS, + 32, nls_cp); bcc_ptr += 2 * bytes_ret; bcc_ptr += 2; /* trailing null */ @@ -197,8 +197,8 @@ static void unicode_domain_string(char **pbcc_area, struct cifs_ses *ses, *(bcc_ptr+1) = 0; bytes_ret = 0; } else - bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->domainName, - 256, nls_cp); + bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, ses->domainName, + 256, nls_cp); bcc_ptr += 2 * bytes_ret; bcc_ptr += 2; /* account for null terminator */ @@ -226,8 +226,8 @@ static void unicode_ssetup_strings(char **pbcc_area, struct cifs_ses *ses, *bcc_ptr = 0; *(bcc_ptr+1) = 0; } else { - bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->user_name, - MAX_USERNAME_SIZE, nls_cp); + bytes_ret = cifs_strtoUTF16((__le16 *) bcc_ptr, ses->user_name, + MAX_USERNAME_SIZE, nls_cp); } bcc_ptr += 2 * bytes_ret; bcc_ptr += 2; /* account for null termination */ @@ -287,7 +287,7 @@ decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifs_ses *ses, cFYI(1, "bleft %d", bleft); kfree(ses->serverOS); - ses->serverOS = cifs_strndup_from_ucs(data, bleft, true, nls_cp); + ses->serverOS = cifs_strndup_from_utf16(data, bleft, true, nls_cp); cFYI(1, "serverOS=%s", ses->serverOS); len = (UniStrnlen((wchar_t *) data, bleft / 2) * 2) + 2; data += len; @@ -296,7 +296,7 @@ decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifs_ses *ses, return; kfree(ses->serverNOS); - ses->serverNOS = cifs_strndup_from_ucs(data, bleft, true, nls_cp); + ses->serverNOS = cifs_strndup_from_utf16(data, bleft, true, nls_cp); cFYI(1, "serverNOS=%s", ses->serverNOS); len = (UniStrnlen((wchar_t *) data, bleft / 2) * 2) + 2; data += len; @@ -305,7 +305,7 @@ decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifs_ses *ses, return; kfree(ses->serverDomain); - ses->serverDomain = cifs_strndup_from_ucs(data, bleft, true, nls_cp); + ses->serverDomain = cifs_strndup_from_utf16(data, bleft, true, nls_cp); cFYI(1, "serverDomain=%s", ses->serverDomain); return; @@ -502,8 +502,8 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, tmp += 2; } else { int len; - len = cifs_strtoUCS((__le16 *)tmp, ses->domainName, - MAX_USERNAME_SIZE, nls_cp); + len = cifs_strtoUTF16((__le16 *)tmp, ses->domainName, + MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ sec_blob->DomainName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->DomainName.Length = cpu_to_le16(len); @@ -518,8 +518,8 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, tmp += 2; } else { int len; - len = cifs_strtoUCS((__le16 *)tmp, ses->user_name, - MAX_USERNAME_SIZE, nls_cp); + len = cifs_strtoUTF16((__le16 *)tmp, ses->user_name, + MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->UserName.Length = cpu_to_le16(len); diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c index 80d850881938..d5cd9aa7eacc 100644 --- a/fs/cifs/smbencrypt.c +++ b/fs/cifs/smbencrypt.c @@ -213,7 +213,7 @@ E_md4hash(const unsigned char *passwd, unsigned char *p16, /* Password cannot be longer than 128 characters */ if (passwd) /* Password must be converted to NT unicode */ - len = cifs_strtoUCS(wpwd, passwd, 128, codepage); + len = cifs_strtoUTF16(wpwd, passwd, 128, codepage); else { len = 0; *wpwd = 0; /* Ensure string is null terminated */ From 3db59dd93309710c40aaf1571c607cb0feef3ecb Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Tue, 17 Jan 2012 22:11:28 -0500 Subject: [PATCH 640/820] ima: fix cred sparse warning Fix ima_policy.c sparse "warning: dereference of noderef expression" message, by accessing cred->uid using current_cred(). Changelog v1: - Change __cred to just cred (based on David Howell's comment) Signed-off-by: Mimi Zohar Signed-off-by: James Morris --- security/integrity/ima/ima_policy.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/security/integrity/ima/ima_policy.c b/security/integrity/ima/ima_policy.c index d661afbe474c..d45061d02fee 100644 --- a/security/integrity/ima/ima_policy.c +++ b/security/integrity/ima/ima_policy.c @@ -99,6 +99,7 @@ static bool ima_match_rules(struct ima_measure_rule_entry *rule, struct inode *inode, enum ima_hooks func, int mask) { struct task_struct *tsk = current; + const struct cred *cred = current_cred(); int i; if ((rule->flags & IMA_FUNC) && rule->func != func) @@ -108,7 +109,7 @@ static bool ima_match_rules(struct ima_measure_rule_entry *rule, if ((rule->flags & IMA_FSMAGIC) && rule->fsmagic != inode->i_sb->s_magic) return false; - if ((rule->flags & IMA_UID) && rule->uid != tsk->cred->uid) + if ((rule->flags & IMA_UID) && rule->uid != cred->uid) return false; for (i = 0; i < MAX_LSM_RULES; i++) { int rc = 0; From f6b24579d099ebb67f39cd7924a72a7eec0ce6ae Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Wed, 18 Jan 2012 10:03:14 +0000 Subject: [PATCH 641/820] keys: fix user_defined key sparse messages Replace the rcu_assign_pointer() calls with rcu_assign_keypointer(). Signed-off-by: Mimi Zohar Signed-off-by: David Howells Signed-off-by: James Morris --- security/keys/user_defined.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/security/keys/user_defined.c b/security/keys/user_defined.c index 69ff52c08e97..2aee3c5a3b99 100644 --- a/security/keys/user_defined.c +++ b/security/keys/user_defined.c @@ -59,7 +59,7 @@ int user_instantiate(struct key *key, const void *data, size_t datalen) /* attach the data */ upayload->datalen = datalen; memcpy(upayload->data, data, datalen); - rcu_assign_pointer(key->payload.data, upayload); + rcu_assign_keypointer(key, upayload); ret = 0; error: @@ -98,7 +98,7 @@ int user_update(struct key *key, const void *data, size_t datalen) if (ret == 0) { /* attach the new data, displacing the old */ zap = key->payload.data; - rcu_assign_pointer(key->payload.data, upayload); + rcu_assign_keypointer(key, upayload); key->expiry = 0; } @@ -133,7 +133,7 @@ void user_revoke(struct key *key) key_payload_reserve(key, 0); if (upayload) { - rcu_assign_pointer(key->payload.data, NULL); + rcu_assign_keypointer(key, NULL); kfree_rcu(upayload, rcu); } } From 456a8167e94b66f406c27400a46a707b870452b0 Mon Sep 17 00:00:00 2001 From: David Howells Date: Wed, 18 Jan 2012 10:04:29 +0000 Subject: [PATCH 642/820] KEYS: Permit key_serial() to be called with a const key pointer Permit key_serial() to be called with a const key pointer. Signed-off-by: David Howells Signed-off-by: James Morris --- include/linux/key.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/key.h b/include/linux/key.h index bfc014c57351..5253471cd2ea 100644 --- a/include/linux/key.h +++ b/include/linux/key.h @@ -271,7 +271,7 @@ extern int keyring_add_key(struct key *keyring, extern struct key *key_lookup(key_serial_t id); -static inline key_serial_t key_serial(struct key *key) +static inline key_serial_t key_serial(const struct key *key) { return key ? key->serial : 0; } From b90bf1de7cb65e7f61798fcfbcf74ae72207b0dc Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 19 Jan 2012 11:42:55 +0100 Subject: [PATCH 643/820] ALSA: hda/realtek - Avoid multi-ios conflicting with multi-speakers When a machine has multiple speakers, we don't need to create the controls for multi-ios. Check the number of primary outputs beforehand. Note that this workaround might not work always with new codecs in future; this assumes that both speakers and multi-io jacks share the same mixers/DACs. If they are routed with different mixers, the individual mixer controls should be needed. But, so far, this doesn't happen with the existing ALC codecs. Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 5e82acf77c5a..61ccbe832b75 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -3233,7 +3233,7 @@ static int alc_auto_create_multi_out_ctls(struct hda_codec *codec, int i, err, noutputs; noutputs = cfg->line_outs; - if (spec->multi_ios > 0) + if (spec->multi_ios > 0 && cfg->line_outs < 3) noutputs += spec->multi_ios; for (i = 0; i < noutputs; i++) { From f21d78e2698b6380a5387461e3b126bb2dee23aa Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 19 Jan 2012 12:10:29 +0100 Subject: [PATCH 644/820] ALSA: hda/realtek - Avoid conflict of unsol-events with static quirks The recently added jack-kctl support sets the unsol event tags dynamically, while static quirks usually set the fixed tags in the init_verbs array. Due to this conflict, the own unsol event handler can't retrieve the tag and handle it properly any more. For fixing this, avoid calling snd_hda_jack_add_kctls() for static quirks, and always let them use own handlers instead of the standard one for the auto-pareser. Reported-by: David Henningsson Signed-off-by: Takashi Iwai --- sound/pci/hda/alc880_quirks.c | 17 +++++++++---- sound/pci/hda/alc882_quirks.c | 15 ++++++++---- sound/pci/hda/patch_realtek.c | 45 ++++++++++++++++++++++++----------- 3 files changed, 53 insertions(+), 24 deletions(-) diff --git a/sound/pci/hda/alc880_quirks.c b/sound/pci/hda/alc880_quirks.c index 5b68435d195b..501501ef36a9 100644 --- a/sound/pci/hda/alc880_quirks.c +++ b/sound/pci/hda/alc880_quirks.c @@ -762,16 +762,22 @@ static void alc880_uniwill_unsol_event(struct hda_codec *codec, /* Looks like the unsol event is incompatible with the standard * definition. 4bit tag is placed at 28 bit! */ - switch (res >> 28) { + res >>= 28; + switch (res) { case ALC_MIC_EVENT: alc88x_simple_mic_automute(codec); break; default: - alc_sku_unsol_event(codec, res); + alc_exec_unsol_event(codec, res); break; } } +static void alc880_unsol_event(struct hda_codec *codec, unsigned int res) +{ + alc_exec_unsol_event(codec, res >> 28); +} + static void alc880_uniwill_p53_setup(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; @@ -800,10 +806,11 @@ static void alc880_uniwill_p53_unsol_event(struct hda_codec *codec, /* Looks like the unsol event is incompatible with the standard * definition. 4bit tag is placed at 28 bit! */ - if ((res >> 28) == ALC_DCVOL_EVENT) + res >>= 28; + if (res == ALC_DCVOL_EVENT) alc880_uniwill_p53_dcvol_automute(codec); else - alc_sku_unsol_event(codec, res); + alc_exec_unsol_event(codec, res); } /* @@ -1677,7 +1684,7 @@ static const struct alc_config_preset alc880_presets[] = { .channel_mode = alc880_lg_ch_modes, .need_dac_fix = 1, .input_mux = &alc880_lg_capture_source, - .unsol_event = alc_sku_unsol_event, + .unsol_event = alc880_unsol_event, .setup = alc880_lg_setup, .init_hook = alc_hp_automute, #ifdef CONFIG_SND_HDA_POWER_SAVE diff --git a/sound/pci/hda/alc882_quirks.c b/sound/pci/hda/alc882_quirks.c index bdf0ed4ab3e2..bb364a53f546 100644 --- a/sound/pci/hda/alc882_quirks.c +++ b/sound/pci/hda/alc882_quirks.c @@ -730,6 +730,11 @@ static void alc889A_mb31_unsol_event(struct hda_codec *codec, unsigned int res) alc889A_mb31_automute(codec); } +static void alc882_unsol_event(struct hda_codec *codec, unsigned int res) +{ + alc_exec_unsol_event(codec, res >> 26); +} + /* * configuration and preset */ @@ -775,7 +780,7 @@ static const struct alc_config_preset alc882_presets[] = { .channel_mode = alc885_mba21_ch_modes, .num_channel_mode = ARRAY_SIZE(alc885_mba21_ch_modes), .input_mux = &alc882_capture_source, - .unsol_event = alc_sku_unsol_event, + .unsol_event = alc882_unsol_event, .setup = alc885_mba21_setup, .init_hook = alc_hp_automute, }, @@ -791,7 +796,7 @@ static const struct alc_config_preset alc882_presets[] = { .input_mux = &alc882_capture_source, .dig_out_nid = ALC882_DIGOUT_NID, .dig_in_nid = ALC882_DIGIN_NID, - .unsol_event = alc_sku_unsol_event, + .unsol_event = alc882_unsol_event, .setup = alc885_mbp3_setup, .init_hook = alc_hp_automute, }, @@ -806,7 +811,7 @@ static const struct alc_config_preset alc882_presets[] = { .input_mux = &mb5_capture_source, .dig_out_nid = ALC882_DIGOUT_NID, .dig_in_nid = ALC882_DIGIN_NID, - .unsol_event = alc_sku_unsol_event, + .unsol_event = alc882_unsol_event, .setup = alc885_mb5_setup, .init_hook = alc_hp_automute, }, @@ -821,7 +826,7 @@ static const struct alc_config_preset alc882_presets[] = { .input_mux = &macmini3_capture_source, .dig_out_nid = ALC882_DIGOUT_NID, .dig_in_nid = ALC882_DIGIN_NID, - .unsol_event = alc_sku_unsol_event, + .unsol_event = alc882_unsol_event, .setup = alc885_macmini3_setup, .init_hook = alc_hp_automute, }, @@ -836,7 +841,7 @@ static const struct alc_config_preset alc882_presets[] = { .input_mux = &alc889A_imac91_capture_source, .dig_out_nid = ALC882_DIGOUT_NID, .dig_in_nid = ALC882_DIGIN_NID, - .unsol_event = alc_sku_unsol_event, + .unsol_event = alc882_unsol_event, .setup = alc885_imac91_setup, .init_hook = alc_hp_automute, }, diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 61ccbe832b75..2326bf379525 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -621,17 +621,10 @@ static void alc_mic_automute(struct hda_codec *codec) alc_mux_select(codec, 0, spec->int_mic_idx, false); } -/* unsolicited event for HP jack sensing */ -static void alc_sku_unsol_event(struct hda_codec *codec, unsigned int res) +/* handle the specified unsol action (ALC_XXX_EVENT) */ +static void alc_exec_unsol_event(struct hda_codec *codec, int action) { - struct alc_spec *spec = codec->spec; - if (codec->vendor_id == 0x10ec0880) - res >>= 28; - else - res >>= 26; - if (spec->use_jack_tbl) - res = snd_hda_jack_get_action(codec, res); - switch (res) { + switch (action) { case ALC_HP_EVENT: alc_hp_automute(codec); break; @@ -645,6 +638,19 @@ static void alc_sku_unsol_event(struct hda_codec *codec, unsigned int res) snd_hda_jack_report_sync(codec); } +/* unsolicited event for HP jack sensing */ +static void alc_sku_unsol_event(struct hda_codec *codec, unsigned int res) +{ + struct alc_spec *spec = codec->spec; + if (codec->vendor_id == 0x10ec0880) + res >>= 28; + else + res >>= 26; + if (spec->use_jack_tbl) + res = snd_hda_jack_get_action(codec, res); + alc_exec_unsol_event(codec, res); +} + /* call init functions of standard auto-mute helpers */ static void alc_inithook(struct hda_codec *codec) { @@ -1883,7 +1889,7 @@ static const struct snd_kcontrol_new alc_beep_mixer[] = { }; #endif -static int alc_build_controls(struct hda_codec *codec) +static int __alc_build_controls(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; struct snd_kcontrol *kctl = NULL; @@ -2029,11 +2035,16 @@ static int alc_build_controls(struct hda_codec *codec) alc_free_kctls(codec); /* no longer needed */ - err = snd_hda_jack_add_kctls(codec, &spec->autocfg); + return 0; +} + +static int alc_build_controls(struct hda_codec *codec) +{ + struct alc_spec *spec = codec->spec; + int err = __alc_build_controls(codec); if (err < 0) return err; - - return 0; + return snd_hda_jack_add_kctls(codec, &spec->autocfg); } @@ -4168,6 +4179,8 @@ static int patch_alc880(struct hda_codec *codec) codec->patch_ops = alc_patch_ops; if (board_config == ALC_MODEL_AUTO) spec->init_hook = alc_auto_init_std; + else + codec->patch_ops.build_controls = __alc_build_controls; #ifdef CONFIG_SND_HDA_POWER_SAVE if (!spec->loopback.amplist) spec->loopback.amplist = alc880_loopbacks; @@ -4297,6 +4310,8 @@ static int patch_alc260(struct hda_codec *codec) codec->patch_ops = alc_patch_ops; if (board_config == ALC_MODEL_AUTO) spec->init_hook = alc_auto_init_std; + else + codec->patch_ops.build_controls = __alc_build_controls; spec->shutup = alc_eapd_shutup; #ifdef CONFIG_SND_HDA_POWER_SAVE if (!spec->loopback.amplist) @@ -4691,6 +4706,8 @@ static int patch_alc882(struct hda_codec *codec) codec->patch_ops = alc_patch_ops; if (board_config == ALC_MODEL_AUTO) spec->init_hook = alc_auto_init_std; + else + codec->patch_ops.build_controls = __alc_build_controls; #ifdef CONFIG_SND_HDA_POWER_SAVE if (!spec->loopback.amplist) From a7309792c4e313d4e4c30084dc0ecbc834082433 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 19 Jan 2012 15:03:48 +0100 Subject: [PATCH 645/820] ALSA: hda/realtek - Remove use_jack_tbl field Now that all quirks have the own unsol handlers, we don't need to check use_jack_tbl flag any more. Let's kill it. Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 2326bf379525..ddbed9705ef4 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -185,7 +185,6 @@ struct alc_spec { unsigned int vol_in_capsrc:1; /* use capsrc volume (ADC has no vol) */ unsigned int parse_flags; /* passed to snd_hda_parse_pin_defcfg() */ unsigned int shared_mic_hp:1; /* HP/Mic-in sharing */ - unsigned int use_jack_tbl:1; /* 1 for model=auto */ /* auto-mute control */ int automute_mode; @@ -646,8 +645,7 @@ static void alc_sku_unsol_event(struct hda_codec *codec, unsigned int res) res >>= 28; else res >>= 26; - if (spec->use_jack_tbl) - res = snd_hda_jack_get_action(codec, res); + res = snd_hda_jack_get_action(codec, res); alc_exec_unsol_event(codec, res); } @@ -3915,7 +3913,6 @@ static void set_capture_mixer(struct hda_codec *codec) static void alc_auto_init_std(struct hda_codec *codec) { struct alc_spec *spec = codec->spec; - spec->use_jack_tbl = 1; alc_auto_init_multi_out(codec); alc_auto_init_extra_out(codec); alc_auto_init_analog_input(codec); From c50b52a0c7ec8522983f5021c0b0952b4d678adf Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 18 Jan 2012 22:13:26 +0000 Subject: [PATCH 646/820] igb: make local functions static Sparse caught two functions that were only being used in one file. Signed-off-by: Stephen Hemminger Tested-by: Aaron Brown Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/igb/e1000_mac.c | 2 +- drivers/net/ethernet/intel/igb/igb_main.c | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/intel/igb/e1000_mac.c b/drivers/net/ethernet/intel/igb/e1000_mac.c index 73aac082c44d..36450d35f621 100644 --- a/drivers/net/ethernet/intel/igb/e1000_mac.c +++ b/drivers/net/ethernet/intel/igb/e1000_mac.c @@ -151,7 +151,7 @@ void igb_clear_vfta_i350(struct e1000_hw *hw) * Writes value at the given offset in the register array which stores * the VLAN filter table. **/ -void igb_write_vfta_i350(struct e1000_hw *hw, u32 offset, u32 value) +static void igb_write_vfta_i350(struct e1000_hw *hw, u32 offset, u32 value) { int i; diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index 01e5e89ef959..92fd642b2dfe 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -4003,8 +4003,8 @@ set_itr_now: } } -void igb_tx_ctxtdesc(struct igb_ring *tx_ring, u32 vlan_macip_lens, - u32 type_tucmd, u32 mss_l4len_idx) +static void igb_tx_ctxtdesc(struct igb_ring *tx_ring, u32 vlan_macip_lens, + u32 type_tucmd, u32 mss_l4len_idx) { struct e1000_adv_tx_context_desc *context_desc; u16 i = tx_ring->next_to_use; @@ -5623,7 +5623,7 @@ static irqreturn_t igb_intr(int irq, void *data) return IRQ_HANDLED; } -void igb_ring_irq_enable(struct igb_q_vector *q_vector) +static void igb_ring_irq_enable(struct igb_q_vector *q_vector) { struct igb_adapter *adapter = q_vector->adapter; struct e1000_hw *hw = &adapter->hw; From 6e861326b1d78bb439c0724864a6ca83ec23d289 Mon Sep 17 00:00:00 2001 From: Carolyn Wyborny Date: Wed, 18 Jan 2012 22:13:27 +0000 Subject: [PATCH 647/820] igb: Update Copyright on all Intel copyrighted files. Signed-off-by: Carolyn Wyborny Tested-by: Aaron Brown Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/igb/Makefile | 2 +- drivers/net/ethernet/intel/igb/e1000_82575.c | 2 +- drivers/net/ethernet/intel/igb/e1000_82575.h | 2 +- drivers/net/ethernet/intel/igb/e1000_defines.h | 2 +- drivers/net/ethernet/intel/igb/e1000_hw.h | 2 +- drivers/net/ethernet/intel/igb/e1000_mac.c | 2 +- drivers/net/ethernet/intel/igb/e1000_mac.h | 2 +- drivers/net/ethernet/intel/igb/e1000_mbx.c | 2 +- drivers/net/ethernet/intel/igb/e1000_mbx.h | 2 +- drivers/net/ethernet/intel/igb/e1000_nvm.c | 2 +- drivers/net/ethernet/intel/igb/e1000_nvm.h | 2 +- drivers/net/ethernet/intel/igb/e1000_phy.c | 2 +- drivers/net/ethernet/intel/igb/e1000_phy.h | 2 +- drivers/net/ethernet/intel/igb/e1000_regs.h | 2 +- drivers/net/ethernet/intel/igb/igb.h | 2 +- drivers/net/ethernet/intel/igb/igb_ethtool.c | 2 +- drivers/net/ethernet/intel/igb/igb_main.c | 4 ++-- 17 files changed, 18 insertions(+), 18 deletions(-) diff --git a/drivers/net/ethernet/intel/igb/Makefile b/drivers/net/ethernet/intel/igb/Makefile index c6e4621b6262..6565c463185c 100644 --- a/drivers/net/ethernet/intel/igb/Makefile +++ b/drivers/net/ethernet/intel/igb/Makefile @@ -1,7 +1,7 @@ ################################################################################ # # Intel 82575 PCI-Express Ethernet Linux driver -# Copyright(c) 1999 - 2011 Intel Corporation. +# Copyright(c) 1999 - 2012 Intel Corporation. # # This program is free software; you can redistribute it and/or modify it # under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_82575.c b/drivers/net/ethernet/intel/igb/e1000_82575.c index b8e20f037d0a..08bdc33715ee 100644 --- a/drivers/net/ethernet/intel/igb/e1000_82575.c +++ b/drivers/net/ethernet/intel/igb/e1000_82575.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_82575.h b/drivers/net/ethernet/intel/igb/e1000_82575.h index 08a757eb6608..b927d79ab536 100644 --- a/drivers/net/ethernet/intel/igb/e1000_82575.h +++ b/drivers/net/ethernet/intel/igb/e1000_82575.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_defines.h b/drivers/net/ethernet/intel/igb/e1000_defines.h index f5fc5725ea94..aed217449f0d 100644 --- a/drivers/net/ethernet/intel/igb/e1000_defines.h +++ b/drivers/net/ethernet/intel/igb/e1000_defines.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_hw.h b/drivers/net/ethernet/intel/igb/e1000_hw.h index 4519a1367170..f67cbd3fa307 100644 --- a/drivers/net/ethernet/intel/igb/e1000_hw.h +++ b/drivers/net/ethernet/intel/igb/e1000_hw.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_mac.c b/drivers/net/ethernet/intel/igb/e1000_mac.c index 36450d35f621..f57338afd71f 100644 --- a/drivers/net/ethernet/intel/igb/e1000_mac.c +++ b/drivers/net/ethernet/intel/igb/e1000_mac.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_mac.h b/drivers/net/ethernet/intel/igb/e1000_mac.h index e45996b4ea34..cbddc4e51e30 100644 --- a/drivers/net/ethernet/intel/igb/e1000_mac.h +++ b/drivers/net/ethernet/intel/igb/e1000_mac.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_mbx.c b/drivers/net/ethernet/intel/igb/e1000_mbx.c index 469d95eaa154..5988b8958baf 100644 --- a/drivers/net/ethernet/intel/igb/e1000_mbx.c +++ b/drivers/net/ethernet/intel/igb/e1000_mbx.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_mbx.h b/drivers/net/ethernet/intel/igb/e1000_mbx.h index eddb0f83dcea..dbcfa3d5caec 100644 --- a/drivers/net/ethernet/intel/igb/e1000_mbx.h +++ b/drivers/net/ethernet/intel/igb/e1000_mbx.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_nvm.c b/drivers/net/ethernet/intel/igb/e1000_nvm.c index 40407124e722..fa2c6ba62139 100644 --- a/drivers/net/ethernet/intel/igb/e1000_nvm.c +++ b/drivers/net/ethernet/intel/igb/e1000_nvm.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_nvm.h b/drivers/net/ethernet/intel/igb/e1000_nvm.h index a2a7ca9fa733..825b0228cac0 100644 --- a/drivers/net/ethernet/intel/igb/e1000_nvm.h +++ b/drivers/net/ethernet/intel/igb/e1000_nvm.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2011 Intel Corporation. + Copyright(c) 2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_phy.c b/drivers/net/ethernet/intel/igb/e1000_phy.c index b17d7c20f817..789de5b83aad 100644 --- a/drivers/net/ethernet/intel/igb/e1000_phy.c +++ b/drivers/net/ethernet/intel/igb/e1000_phy.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_phy.h b/drivers/net/ethernet/intel/igb/e1000_phy.h index 8510797b9d81..4c32ac66ff39 100644 --- a/drivers/net/ethernet/intel/igb/e1000_phy.h +++ b/drivers/net/ethernet/intel/igb/e1000_phy.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/e1000_regs.h b/drivers/net/ethernet/intel/igb/e1000_regs.h index 0a860bc1198e..ccdf36d503fd 100644 --- a/drivers/net/ethernet/intel/igb/e1000_regs.h +++ b/drivers/net/ethernet/intel/igb/e1000_regs.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/igb.h b/drivers/net/ethernet/intel/igb/igb.h index 3d12e67eebb4..8e33bdd33eea 100644 --- a/drivers/net/ethernet/intel/igb/igb.h +++ b/drivers/net/ethernet/intel/igb/igb.h @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/igb_ethtool.c b/drivers/net/ethernet/intel/igb/igb_ethtool.c index 7998bf4d5946..aa399a8a8f0d 100644 --- a/drivers/net/ethernet/intel/igb/igb_ethtool.c +++ b/drivers/net/ethernet/intel/igb/igb_ethtool.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index 92fd642b2dfe..e91d73c8aa4e 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -1,7 +1,7 @@ /******************************************************************************* Intel(R) Gigabit Ethernet Linux driver - Copyright(c) 2007-2011 Intel Corporation. + Copyright(c) 2007-2012 Intel Corporation. This program is free software; you can redistribute it and/or modify it under the terms and conditions of the GNU General Public License, @@ -68,7 +68,7 @@ char igb_driver_name[] = "igb"; char igb_driver_version[] = DRV_VERSION; static const char igb_driver_string[] = "Intel(R) Gigabit Ethernet Network Driver"; -static const char igb_copyright[] = "Copyright (c) 2007-2011 Intel Corporation."; +static const char igb_copyright[] = "Copyright (c) 2007-2012 Intel Corporation."; static const struct e1000_info *igb_info_tbl[] = { [board_82575] = &e1000_82575_info, From b2f0f6bb874df645766930e0f023a208d93c12cd Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 18 Jan 2012 22:13:28 +0000 Subject: [PATCH 648/820] igbvf: remove unneeded cast The cast and comment are unnecessary in the current upstream kernel. Signed-off-by: Stephen Hemminger Tested-by: Garrett, RobertX E Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/igbvf/ethtool.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/ethernet/intel/igbvf/ethtool.c b/drivers/net/ethernet/intel/igbvf/ethtool.c index 7b600a1f6366..2dba53446064 100644 --- a/drivers/net/ethernet/intel/igbvf/ethtool.c +++ b/drivers/net/ethernet/intel/igbvf/ethtool.c @@ -468,6 +468,5 @@ static const struct ethtool_ops igbvf_ethtool_ops = { void igbvf_set_ethtool_ops(struct net_device *netdev) { - /* have to "undeclare" const on this struct to remove warnings */ - SET_ETHTOOL_OPS(netdev, (struct ethtool_ops *)&igbvf_ethtool_ops); + SET_ETHTOOL_OPS(netdev, &igbvf_ethtool_ops); } From 56b68960ce996db7947440179cc153652a504e9a Mon Sep 17 00:00:00 2001 From: Mitch A Williams Date: Wed, 18 Jan 2012 22:13:29 +0000 Subject: [PATCH 649/820] igbvf: Remove unnecessary irq disable/enable This irq disable/enable pair used to wrap access to the driver's vlgrp struct, which is no longer present. So, then, this could also so no longer be present. Signed-off-by: Mitch Williams Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/igbvf/netdev.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/drivers/net/ethernet/intel/igbvf/netdev.c b/drivers/net/ethernet/intel/igbvf/netdev.c index fd3da3076c2f..a4b20c865759 100644 --- a/drivers/net/ethernet/intel/igbvf/netdev.c +++ b/drivers/net/ethernet/intel/igbvf/netdev.c @@ -1194,11 +1194,6 @@ static int igbvf_vlan_rx_kill_vid(struct net_device *netdev, u16 vid) struct igbvf_adapter *adapter = netdev_priv(netdev); struct e1000_hw *hw = &adapter->hw; - igbvf_irq_disable(adapter); - - if (!test_bit(__IGBVF_DOWN, &adapter->state)) - igbvf_irq_enable(adapter); - if (hw->mac.ops.set_vfta(hw, vid, false)) { dev_err(&adapter->pdev->dev, "Failed to remove vlan id %d\n", vid); From 795be954dcc4a72329c8591aafe5eadb03b264af Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Wed, 18 Jan 2012 22:13:30 +0000 Subject: [PATCH 650/820] ixgbe: Fix register defines to correctly handle complex expressions This patch is meant to address possible issues with the IXGBE register defines generating incorrect values when given a complex expression for the register offset. Signed-off-by: Alexander Duyck Tested-by: Phil Schmitt Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbe/ixgbe_type.h | 32 +++++++++---------- 1 file changed, 16 insertions(+), 16 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h index 802bfa0f62cc..775602ef90e5 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_type.h @@ -161,19 +161,19 @@ /* Receive DMA Registers */ #define IXGBE_RDBAL(_i) (((_i) < 64) ? (0x01000 + ((_i) * 0x40)) : \ - (0x0D000 + ((_i - 64) * 0x40))) + (0x0D000 + (((_i) - 64) * 0x40))) #define IXGBE_RDBAH(_i) (((_i) < 64) ? (0x01004 + ((_i) * 0x40)) : \ - (0x0D004 + ((_i - 64) * 0x40))) + (0x0D004 + (((_i) - 64) * 0x40))) #define IXGBE_RDLEN(_i) (((_i) < 64) ? (0x01008 + ((_i) * 0x40)) : \ - (0x0D008 + ((_i - 64) * 0x40))) + (0x0D008 + (((_i) - 64) * 0x40))) #define IXGBE_RDH(_i) (((_i) < 64) ? (0x01010 + ((_i) * 0x40)) : \ - (0x0D010 + ((_i - 64) * 0x40))) + (0x0D010 + (((_i) - 64) * 0x40))) #define IXGBE_RDT(_i) (((_i) < 64) ? (0x01018 + ((_i) * 0x40)) : \ - (0x0D018 + ((_i - 64) * 0x40))) + (0x0D018 + (((_i) - 64) * 0x40))) #define IXGBE_RXDCTL(_i) (((_i) < 64) ? (0x01028 + ((_i) * 0x40)) : \ - (0x0D028 + ((_i - 64) * 0x40))) + (0x0D028 + (((_i) - 64) * 0x40))) #define IXGBE_RSCCTL(_i) (((_i) < 64) ? (0x0102C + ((_i) * 0x40)) : \ - (0x0D02C + ((_i - 64) * 0x40))) + (0x0D02C + (((_i) - 64) * 0x40))) #define IXGBE_RSCDBU 0x03028 #define IXGBE_RDDCC 0x02F20 #define IXGBE_RXMEMWRAP 0x03190 @@ -186,7 +186,7 @@ */ #define IXGBE_SRRCTL(_i) (((_i) <= 15) ? (0x02100 + ((_i) * 4)) : \ (((_i) < 64) ? (0x01014 + ((_i) * 0x40)) : \ - (0x0D014 + ((_i - 64) * 0x40)))) + (0x0D014 + (((_i) - 64) * 0x40)))) /* * Rx DCA Control Register: * 00-15 : 0x02200 + n*4 @@ -195,7 +195,7 @@ */ #define IXGBE_DCA_RXCTRL(_i) (((_i) <= 15) ? (0x02200 + ((_i) * 4)) : \ (((_i) < 64) ? (0x0100C + ((_i) * 0x40)) : \ - (0x0D00C + ((_i - 64) * 0x40)))) + (0x0D00C + (((_i) - 64) * 0x40)))) #define IXGBE_RDRXCTL 0x02F00 #define IXGBE_RXPBSIZE(_i) (0x03C00 + ((_i) * 4)) /* 8 of these 0x03C00 - 0x03C1C */ @@ -344,9 +344,9 @@ #define IXGBE_WUPL 0x05900 #define IXGBE_WUPM 0x05A00 /* wake up pkt memory 0x5A00-0x5A7C */ -#define IXGBE_FHFT(_n) (0x09000 + (_n * 0x100)) /* Flex host filter table */ -#define IXGBE_FHFT_EXT(_n) (0x09800 + (_n * 0x100)) /* Ext Flexible Host - * Filter Table */ +#define IXGBE_FHFT(_n) (0x09000 + ((_n) * 0x100)) /* Flex host filter table */ +#define IXGBE_FHFT_EXT(_n) (0x09800 + ((_n) * 0x100)) /* Ext Flexible Host + * Filter Table */ #define IXGBE_FLEXIBLE_FILTER_COUNT_MAX 4 #define IXGBE_EXT_FLEXIBLE_FILTER_COUNT_MAX 2 @@ -1485,7 +1485,7 @@ enum { #define IXGBE_LED_BLINK_BASE 0x00000080 #define IXGBE_LED_MODE_MASK_BASE 0x0000000F #define IXGBE_LED_OFFSET(_base, _i) (_base << (8 * (_i))) -#define IXGBE_LED_MODE_SHIFT(_i) (8*(_i)) +#define IXGBE_LED_MODE_SHIFT(_i) (8 * (_i)) #define IXGBE_LED_IVRT(_i) IXGBE_LED_OFFSET(IXGBE_LED_IVRT_BASE, _i) #define IXGBE_LED_BLINK(_i) IXGBE_LED_OFFSET(IXGBE_LED_BLINK_BASE, _i) #define IXGBE_LED_MODE_MASK(_i) IXGBE_LED_OFFSET(IXGBE_LED_MODE_MASK_BASE, _i) @@ -2068,9 +2068,9 @@ enum { /* SR-IOV specific macros */ #define IXGBE_MBVFICR_INDEX(vf_number) (vf_number >> 4) -#define IXGBE_MBVFICR(_i) (0x00710 + (_i * 4)) -#define IXGBE_VFLRE(_i) (((_i & 1) ? 0x001C0 : 0x00600)) -#define IXGBE_VFLREC(_i) (0x00700 + (_i * 4)) +#define IXGBE_MBVFICR(_i) (0x00710 + ((_i) * 4)) +#define IXGBE_VFLRE(_i) ((((_i) & 1) ? 0x001C0 : 0x00600)) +#define IXGBE_VFLREC(_i) (0x00700 + ((_i) * 4)) enum ixgbe_fdir_pballoc_type { IXGBE_FDIR_PBALLOC_NONE = 0, From 375b27cf5db963e2bc2a34dc5643d0d7ceca1ee2 Mon Sep 17 00:00:00 2001 From: Greg Rose Date: Wed, 18 Jan 2012 22:13:31 +0000 Subject: [PATCH 651/820] ixgbevf: Prevent possible race condition by checking for message The mailbox interrupt routine might cause a race condition sometimes and cause a message to be missed. Signed-off-by: Greg Rose Tested-by: Sibai Li Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- .../net/ethernet/intel/ixgbevf/ixgbevf_main.c | 39 ++++++++++--------- 1 file changed, 21 insertions(+), 18 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index 891162d1610c..1cad3b6d2c30 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -917,31 +917,34 @@ static irqreturn_t ixgbevf_msix_mbx(int irq, void *data) struct ixgbe_hw *hw = &adapter->hw; u32 eicr; u32 msg; + bool got_ack = false; eicr = IXGBE_READ_REG(hw, IXGBE_VTEICS); IXGBE_WRITE_REG(hw, IXGBE_VTEICR, eicr); - if (!hw->mbx.ops.check_for_ack(hw)) { - /* - * checking for the ack clears the PFACK bit. Place - * it back in the v2p_mailbox cache so that anyone - * polling for an ack will not miss it. Also - * avoid the read below because the code to read - * the mailbox will also clear the ack bit. This was - * causing lost acks. Just cache the bit and exit - * the IRQ handler. - */ - hw->mbx.v2p_mailbox |= IXGBE_VFMAILBOX_PFACK; + if (!hw->mbx.ops.check_for_ack(hw)) + got_ack = true; + + if (!hw->mbx.ops.check_for_msg(hw)) { + hw->mbx.ops.read(hw, &msg, 1); + + if ((msg & IXGBE_MBVFICR_VFREQ_MASK) == IXGBE_PF_CONTROL_MSG) + mod_timer(&adapter->watchdog_timer, + round_jiffies(jiffies + 1)); + + if (msg & IXGBE_VT_MSGTYPE_NACK) + pr_warn("Last Request of type %2.2x to PF Nacked\n", + msg & 0xFF); goto out; } - /* Not an ack interrupt, go ahead and read the message */ - hw->mbx.ops.read(hw, &msg, 1); - - if ((msg & IXGBE_MBVFICR_VFREQ_MASK) == IXGBE_PF_CONTROL_MSG) - mod_timer(&adapter->watchdog_timer, - round_jiffies(jiffies + 1)); - + /* + * checking for the ack clears the PFACK bit. Place + * it back in the v2p_mailbox cache so that anyone + * polling for an ack will not miss it + */ + if (got_ack) + hw->mbx.v2p_mailbox |= IXGBE_VFMAILBOX_PFACK; out: return IRQ_HANDLED; } From b47aca135d6df5f676c768368baf3c099f054fcd Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 18 Jan 2012 22:13:32 +0000 Subject: [PATCH 652/820] ixgbevf: make ethtool ops and strings const Signed-off-by: Stephen Hemminger Tested-by: Sibai Li Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbevf/ethtool.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ethtool.c b/drivers/net/ethernet/intel/ixgbevf/ethtool.c index dc8e6511c640..c85700318147 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ethtool.c +++ b/drivers/net/ethernet/intel/ixgbevf/ethtool.c @@ -56,7 +56,8 @@ struct ixgbe_stats { offsetof(struct ixgbevf_adapter, m), \ offsetof(struct ixgbevf_adapter, b), \ offsetof(struct ixgbevf_adapter, r) -static struct ixgbe_stats ixgbe_gstrings_stats[] = { + +static const struct ixgbe_stats ixgbe_gstrings_stats[] = { {"rx_packets", IXGBEVF_STAT(stats.vfgprc, stats.base_vfgprc, stats.saved_reset_vfgprc)}, {"tx_packets", IXGBEVF_STAT(stats.vfgptc, stats.base_vfgptc, @@ -671,7 +672,7 @@ static int ixgbevf_nway_reset(struct net_device *netdev) return 0; } -static struct ethtool_ops ixgbevf_ethtool_ops = { +static const struct ethtool_ops ixgbevf_ethtool_ops = { .get_settings = ixgbevf_get_settings, .get_drvinfo = ixgbevf_get_drvinfo, .get_regs_len = ixgbevf_get_regs_len, From b5417bf8e8952401bca69ded67c30ead484af823 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 18 Jan 2012 22:13:33 +0000 Subject: [PATCH 653/820] ixgbevf: fix sparse warnings Fixes sparse warnings: drivers/net/ethernet/intel/ixgbevf/vf.c:418:21: warning: symbol 'ixgbevf_82599_vf_info' was not declared. Should it be static? drivers/net/ethernet/intel/ixgbevf/vf.c:423:21: warning: symbol 'ixgbevf_X540_vf_info' was not declared. Should it be static? drivers/net/ethernet/intel/ixgbevf/mbx.c:331:29: warning: symbol 'ixgbevf_mbx_ops' was not declared. Should it be static? Signed-off-by: Stephen Hemminger Tested-by: Sibai Li Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbevf/ixgbevf.h | 2 +- drivers/net/ethernet/intel/ixgbevf/mbx.c | 3 ++- drivers/net/ethernet/intel/ixgbevf/vf.c | 1 + 3 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h index e6c9d1a927a9..74b22ee497e7 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h @@ -281,7 +281,7 @@ enum ixgbevf_boards { extern struct ixgbevf_info ixgbevf_82599_vf_info; extern struct ixgbevf_info ixgbevf_X540_vf_info; -extern struct ixgbe_mbx_operations ixgbevf_mbx_ops; +extern const struct ixgbe_mbx_operations ixgbevf_mbx_ops; /* needed by ethtool.c */ extern char ixgbevf_driver_name[]; diff --git a/drivers/net/ethernet/intel/ixgbevf/mbx.c b/drivers/net/ethernet/intel/ixgbevf/mbx.c index 930fa83f2568..13532d9ba72d 100644 --- a/drivers/net/ethernet/intel/ixgbevf/mbx.c +++ b/drivers/net/ethernet/intel/ixgbevf/mbx.c @@ -26,6 +26,7 @@ *******************************************************************************/ #include "mbx.h" +#include "ixgbevf.h" /** * ixgbevf_poll_for_msg - Wait for message notification @@ -328,7 +329,7 @@ static s32 ixgbevf_init_mbx_params_vf(struct ixgbe_hw *hw) return 0; } -struct ixgbe_mbx_operations ixgbevf_mbx_ops = { +const struct ixgbe_mbx_operations ixgbevf_mbx_ops = { .init_params = ixgbevf_init_mbx_params_vf, .read = ixgbevf_read_mbx_vf, .write = ixgbevf_write_mbx_vf, diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.c b/drivers/net/ethernet/intel/ixgbevf/vf.c index 21533e300367..5c8c23f8346f 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.c +++ b/drivers/net/ethernet/intel/ixgbevf/vf.c @@ -26,6 +26,7 @@ *******************************************************************************/ #include "vf.h" +#include "ixgbevf.h" /** * ixgbevf_start_hw_vf - Prepare hardware for Tx/Rx From 3d8fe98f8d5cb303d907d8f94ea6dc1f9a1d8b7a Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 18 Jan 2012 22:13:34 +0000 Subject: [PATCH 654/820] ixgbevf: make operations tables const The arrays of function pointers should be const to make life harder for rootkits. Signed-off-by: Stephen Hemminger Tested-by: Sibai Li Signed-off-by: Jeff Kirsher Signed-off-by: David S. Miller --- drivers/net/ethernet/intel/ixgbevf/ixgbevf.h | 6 +++--- drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c | 2 +- drivers/net/ethernet/intel/ixgbevf/vf.c | 6 +++--- drivers/net/ethernet/intel/ixgbevf/vf.h | 2 +- 4 files changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h index 74b22ee497e7..9075c1d61039 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf.h @@ -279,12 +279,12 @@ enum ixgbevf_boards { board_X540_vf, }; -extern struct ixgbevf_info ixgbevf_82599_vf_info; -extern struct ixgbevf_info ixgbevf_X540_vf_info; +extern const struct ixgbevf_info ixgbevf_82599_vf_info; +extern const struct ixgbevf_info ixgbevf_X540_vf_info; extern const struct ixgbe_mbx_operations ixgbevf_mbx_ops; /* needed by ethtool.c */ -extern char ixgbevf_driver_name[]; +extern const char ixgbevf_driver_name[]; extern const char ixgbevf_driver_version[]; extern int ixgbevf_up(struct ixgbevf_adapter *adapter); diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index 1cad3b6d2c30..bed411bada21 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -53,7 +53,7 @@ #include "ixgbevf.h" -char ixgbevf_driver_name[] = "ixgbevf"; +const char ixgbevf_driver_name[] = "ixgbevf"; static const char ixgbevf_driver_string[] = "Intel(R) 10 Gigabit PCI Express Virtual Function Network Driver"; diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.c b/drivers/net/ethernet/intel/ixgbevf/vf.c index 5c8c23f8346f..d0138d7a31a1 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.c +++ b/drivers/net/ethernet/intel/ixgbevf/vf.c @@ -402,7 +402,7 @@ static s32 ixgbevf_check_mac_link_vf(struct ixgbe_hw *hw, return 0; } -static struct ixgbe_mac_operations ixgbevf_mac_ops = { +static const struct ixgbe_mac_operations ixgbevf_mac_ops = { .init_hw = ixgbevf_init_hw_vf, .reset_hw = ixgbevf_reset_hw_vf, .start_hw = ixgbevf_start_hw_vf, @@ -416,12 +416,12 @@ static struct ixgbe_mac_operations ixgbevf_mac_ops = { .set_vfta = ixgbevf_set_vfta_vf, }; -struct ixgbevf_info ixgbevf_82599_vf_info = { +const struct ixgbevf_info ixgbevf_82599_vf_info = { .mac = ixgbe_mac_82599_vf, .mac_ops = &ixgbevf_mac_ops, }; -struct ixgbevf_info ixgbevf_X540_vf_info = { +const struct ixgbevf_info ixgbevf_X540_vf_info = { .mac = ixgbe_mac_X540_vf, .mac_ops = &ixgbevf_mac_ops, }; diff --git a/drivers/net/ethernet/intel/ixgbevf/vf.h b/drivers/net/ethernet/intel/ixgbevf/vf.h index 10306b492ee6..d556619a9212 100644 --- a/drivers/net/ethernet/intel/ixgbevf/vf.h +++ b/drivers/net/ethernet/intel/ixgbevf/vf.h @@ -167,7 +167,7 @@ struct ixgbevf_hw_stats { struct ixgbevf_info { enum ixgbe_mac_type mac; - struct ixgbe_mac_operations *mac_ops; + const struct ixgbe_mac_operations *mac_ops; }; #endif /* __IXGBE_VF_H__ */ From 1aab323ea5cd67d2d2572a1f2794978583ff8545 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 19 Jan 2012 13:19:42 -0500 Subject: [PATCH 655/820] qnx4: di_fname is an array, for crying out loud... (struct qnx4_inode_entry *)(bh->b_data + some_offset)->di_fname is not going to be NULL, TYVM... Signed-off-by: Al Viro --- fs/qnx4/inode.c | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 2bfd987f4853..63e0f7471fb5 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -194,20 +194,18 @@ static const char *qnx4_checkroot(struct super_block *sb) } for (i = 0; i < QNX4_INODES_PER_BLOCK; i++) { rootdir = (struct qnx4_inode_entry *) (bh->b_data + i * QNX4_DIR_ENTRY_SIZE); - if (rootdir->di_fname != NULL) { - QNX4DEBUG((KERN_INFO "rootdir entry found : [%s]\n", rootdir->di_fname)); - if (!strcmp(rootdir->di_fname, - QNX4_BMNAME)) { - found = 1; - qnx4_sb(sb)->BitMap = kmemdup(rootdir, - sizeof(struct qnx4_inode_entry), - GFP_KERNEL); - if (!qnx4_sb(sb)->BitMap) { - brelse (bh); - return "not enough memory for bitmap inode"; - }/* keep bitmap inode known */ - break; - } + QNX4DEBUG((KERN_INFO "rootdir entry found : [%s]\n", rootdir->di_fname)); + if (!strcmp(rootdir->di_fname, + QNX4_BMNAME)) { + found = 1; + qnx4_sb(sb)->BitMap = kmemdup(rootdir, + sizeof(struct qnx4_inode_entry), + GFP_KERNEL); + if (!qnx4_sb(sb)->BitMap) { + brelse (bh); + return "not enough memory for bitmap inode"; + }/* keep bitmap inode known */ + break; } } brelse(bh); From 4134bf81ffd962f4de9bbeca55130d2238bd3698 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 19 Jan 2012 13:40:57 -0500 Subject: [PATCH 656/820] qnx4: reduce the insane nesting in qnx4_checkroot() Signed-off-by: Al Viro --- fs/qnx4/inode.c | 56 +++++++++++++++++++------------------------------ 1 file changed, 22 insertions(+), 34 deletions(-) diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 63e0f7471fb5..3fd121c7c30a 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -179,45 +179,33 @@ static const char *qnx4_checkroot(struct super_block *sb) struct qnx4_inode_entry *rootdir; int rd, rl; int i, j; - int found = 0; - if (*(qnx4_sb(sb)->sb->RootDir.di_fname) != '/') { + if (*(qnx4_sb(sb)->sb->RootDir.di_fname) != '/') return "no qnx4 filesystem (no root dir)."; - } else { - QNX4DEBUG((KERN_NOTICE "QNX4 filesystem found on dev %s.\n", sb->s_id)); - rd = le32_to_cpu(qnx4_sb(sb)->sb->RootDir.di_first_xtnt.xtnt_blk) - 1; - rl = le32_to_cpu(qnx4_sb(sb)->sb->RootDir.di_first_xtnt.xtnt_size); - for (j = 0; j < rl; j++) { - bh = sb_bread(sb, rd + j); /* root dir, first block */ - if (bh == NULL) { - return "unable to read root entry."; - } - for (i = 0; i < QNX4_INODES_PER_BLOCK; i++) { - rootdir = (struct qnx4_inode_entry *) (bh->b_data + i * QNX4_DIR_ENTRY_SIZE); - QNX4DEBUG((KERN_INFO "rootdir entry found : [%s]\n", rootdir->di_fname)); - if (!strcmp(rootdir->di_fname, - QNX4_BMNAME)) { - found = 1; - qnx4_sb(sb)->BitMap = kmemdup(rootdir, - sizeof(struct qnx4_inode_entry), - GFP_KERNEL); - if (!qnx4_sb(sb)->BitMap) { - brelse (bh); - return "not enough memory for bitmap inode"; - }/* keep bitmap inode known */ - break; - } - } + QNX4DEBUG((KERN_NOTICE "QNX4 filesystem found on dev %s.\n", sb->s_id)); + rd = le32_to_cpu(qnx4_sb(sb)->sb->RootDir.di_first_xtnt.xtnt_blk) - 1; + rl = le32_to_cpu(qnx4_sb(sb)->sb->RootDir.di_first_xtnt.xtnt_size); + for (j = 0; j < rl; j++) { + bh = sb_bread(sb, rd + j); /* root dir, first block */ + if (bh == NULL) + return "unable to read root entry."; + rootdir = (struct qnx4_inode_entry *) bh->b_data; + for (i = 0; i < QNX4_INODES_PER_BLOCK; i++, rootdir++) { + QNX4DEBUG((KERN_INFO "rootdir entry found : [%s]\n", rootdir->di_fname)); + if (strcmp(rootdir->di_fname, QNX4_BMNAME) != 0) + continue; + qnx4_sb(sb)->BitMap = kmemdup(rootdir, + sizeof(struct qnx4_inode_entry), + GFP_KERNEL); brelse(bh); - if (found != 0) { - break; - } - } - if (found == 0) { - return "bitmap file not found."; + if (!qnx4_sb(sb)->BitMap) + return "not enough memory for bitmap inode"; + /* keep bitmap inode known */ + return NULL; } + brelse(bh); } - return NULL; + return "bitmap file not found."; } static int qnx4_fill_super(struct super_block *s, void *data, int silent) From 8bc5191b261c4fd9a5e9052cebe04ce2ef05f2e7 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 19 Jan 2012 13:54:36 -0500 Subject: [PATCH 657/820] qnx4: don't leak ->BitMap on late failure exits Signed-off-by: Al Viro --- fs/qnx4/inode.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 3fd121c7c30a..6b009548d2e0 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -256,7 +256,7 @@ static int qnx4_fill_super(struct super_block *s, void *data, int silent) if (IS_ERR(root)) { printk(KERN_ERR "qnx4: get inode failed\n"); ret = PTR_ERR(root); - goto out; + goto outb; } ret = -ENOMEM; @@ -269,6 +269,8 @@ static int qnx4_fill_super(struct super_block *s, void *data, int silent) outi: iput(root); + outb: + kfree(qs->BitMap); out: brelse(bh); outnobh: From 4f2f81a5621de47d42476d0b929be2e0d565df84 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Thu, 19 Jan 2012 12:41:25 -0800 Subject: [PATCH 658/820] x86, syscall: Need __ARCH_WANT_SYS_IPC for 32 bits In checkin 303395ac3bf3 x86: Generate system call tables and unistd_*.h from tables the feature macros in were unified between 32 and 64 bits. Unfortunately 32 bits requires __ARCH_WANT_SYS_IPC and this was inadvertently dropped. Reported-by: Dmitry Kasatkin Cc: Linus Torvalds Signed-off-by: H. Peter Anvin Link: http://lkml.kernel.org/r/CALLzPKbeXN5gdngo8uYYU8mAow=XhrwBFBhKfG811f37BubQOg@mail.gmail.com --- arch/x86/include/asm/unistd.h | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/include/asm/unistd.h b/arch/x86/include/asm/unistd.h index b4a3db7ce140..21f77b89e47a 100644 --- a/arch/x86/include/asm/unistd.h +++ b/arch/x86/include/asm/unistd.h @@ -7,6 +7,7 @@ # include # define __ARCH_WANT_IPC_PARSE_VERSION # define __ARCH_WANT_STAT64 +# define __ARCH_WANT_SYS_IPC # define __ARCH_WANT_SYS_OLD_MMAP # define __ARCH_WANT_SYS_OLD_SELECT From f581b63aa1049ac030d6eb6c24e1be1ce2072ae7 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 19 Jan 2012 23:22:38 +0100 Subject: [PATCH 659/820] PM / Documentation: Fix spelling mistake in basic-pm-debugging.txt Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- Documentation/power/basic-pm-debugging.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/power/basic-pm-debugging.txt b/Documentation/power/basic-pm-debugging.txt index 40a4c65f380a..262acf56fa79 100644 --- a/Documentation/power/basic-pm-debugging.txt +++ b/Documentation/power/basic-pm-debugging.txt @@ -15,7 +15,7 @@ test at least a couple of times in a row for confidence. [This is necessary, because some problems only show up on a second attempt at suspending and resuming the system.] Moreover, hibernating in the "reboot" and "shutdown" modes causes the PM core to skip some platform-related callbacks which on ACPI -systems might be necessary to make hibernation work. Thus, if you machine fails +systems might be necessary to make hibernation work. Thus, if your machine fails to hibernate or resume in the "reboot" mode, you should try the "platform" mode: # echo platform > /sys/power/disk From 5eb6f9ad96967be4e0da55521a253e11b534bd3f Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 19 Jan 2012 23:22:49 +0100 Subject: [PATCH 660/820] PM / Documentation: Fix minor issue in freezing_of_tasks.txt In a paragraph, "kernel thread" is mistakenly written as "kernel". Fix this by adding thread after word "kernel". Changes are shown in multiple lines, as they are realigned to 80 col width. Signed-off-by: Viresh Kumar Acked-by: Pavel Machek Signed-off-by: Rafael J. Wysocki --- Documentation/power/freezing-of-tasks.txt | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/Documentation/power/freezing-of-tasks.txt b/Documentation/power/freezing-of-tasks.txt index 6ccb68f68da6..ebd7490ef1df 100644 --- a/Documentation/power/freezing-of-tasks.txt +++ b/Documentation/power/freezing-of-tasks.txt @@ -120,10 +120,10 @@ So in practice, the 'at all' may become a 'why freeze kernel threads?' and freezing user threads I don't find really objectionable." Still, there are kernel threads that may want to be freezable. For example, if -a kernel that belongs to a device driver accesses the device directly, it in -principle needs to know when the device is suspended, so that it doesn't try to -access it at that time. However, if the kernel thread is freezable, it will be -frozen before the driver's .suspend() callback is executed and it will be +a kernel thread that belongs to a device driver accesses the device directly, it +in principle needs to know when the device is suspended, so that it doesn't try +to access it at that time. However, if the kernel thread is freezable, it will +be frozen before the driver's .suspend() callback is executed and it will be thawed after the driver's .resume() callback has run, so it won't be accessing the device while it's suspended. From 160cb5a97daef0cb894685d84c9d4700bb7cccb4 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Thu, 19 Jan 2012 23:23:10 +0100 Subject: [PATCH 661/820] PM / Hibernate: Correct additional pages number calculation The struct bm_block is allocated by chain_alloc(), so it'd better counting it in LINKED_PAGE_DATA_SIZE. Signed-off-by: Namhyung Kim Signed-off-by: Rafael J. Wysocki --- kernel/power/snapshot.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index 1cf88900ec4f..6a768e537001 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -812,7 +812,8 @@ unsigned int snapshot_additional_pages(struct zone *zone) unsigned int res; res = DIV_ROUND_UP(zone->spanned_pages, BM_BITS_PER_BLOCK); - res += DIV_ROUND_UP(res * sizeof(struct bm_block), PAGE_SIZE); + res += DIV_ROUND_UP(res * sizeof(struct bm_block), + LINKED_PAGE_DATA_SIZE); return 2 * res; } From 72081624d5ad3cf56deb6e727b78c4e7a55e4eec Mon Sep 17 00:00:00 2001 From: "Srivatsa S. Bhat" Date: Thu, 19 Jan 2012 23:25:33 +0100 Subject: [PATCH 662/820] PM / Hibernate: Rewrite unlock_system_sleep() to fix s2disk regression Commit 33e638b, "PM / Sleep: Use the freezer_count() functions in [un]lock_system_sleep() APIs" introduced an undesirable change in the behaviour of unlock_system_sleep() since freezer_count() internally calls try_to_freeze() - which we don't need in unlock_system_sleep(). And commit bcda53f, "PM / Sleep: Replace mutex_[un]lock(&pm_mutex) with [un]lock_system_sleep()" made these APIs wide-spread. This caused a regression in suspend-to-disk where snapshot_read() and snapshot_write() were getting frozen due to the try_to_freeze embedded in unlock_system_sleep(), since these functions were invoked when the freezing condition was still in effect. Fix this by rewriting unlock_system_sleep() by open-coding freezer_count() and dropping the try_to_freeze() part. Not only will this fix the regression but this will also ensure that the API only does what it is intended to do, and nothing more, under the hood. While at it, make the code more correct and robust by ensuring that the PF_FREEZER_SKIP flag gets cleared with pm_mutex held, to avoid a race with the freezer. Also, to be on the safer side, open-code freezer_do_not_count() as well (inside lock_system_sleep()), to ensure that any unrelated modification to freezer[_do_not]_count() does not break things again! Reported-and-tested-by: Rafael J. Wysocki Signed-off-by: Srivatsa S. Bhat Acked-by: Tejun Heo Signed-off-by: Rafael J. Wysocki --- include/linux/suspend.h | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 95040cc33107..91784a4f8608 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -357,14 +357,29 @@ extern bool pm_save_wakeup_count(unsigned int count); static inline void lock_system_sleep(void) { - freezer_do_not_count(); + current->flags |= PF_FREEZER_SKIP; mutex_lock(&pm_mutex); } static inline void unlock_system_sleep(void) { + /* + * Don't use freezer_count() because we don't want the call to + * try_to_freeze() here. + * + * Reason: + * Fundamentally, we just don't need it, because freezing condition + * doesn't come into effect until we release the pm_mutex lock, + * since the freezer always works with pm_mutex held. + * + * More importantly, in the case of hibernation, + * unlock_system_sleep() gets called in snapshot_read() and + * snapshot_write() when the freezing condition is still in effect. + * Which means, if we use try_to_freeze() here, it would make them + * enter the refrigerator, thus causing hibernation to lockup. + */ + current->flags &= ~PF_FREEZER_SKIP; mutex_unlock(&pm_mutex); - freezer_count(); } #else /* !CONFIG_PM_SLEEP */ From dcd6c92267155e70a94b3927bce681ce74b80d1f Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 19 Jan 2012 15:04:48 -0800 Subject: [PATCH 663/820] Linux 3.3-rc1 --- Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Makefile b/Makefile index 156ac69c961e..71e6ed21dd15 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 3 -PATCHLEVEL = 2 +PATCHLEVEL = 3 SUBLEVEL = 0 -EXTRAVERSION = +EXTRAVERSION = -rc1 NAME = Saber-toothed Squirrel # *DOCUMENTATION* From 54d5dcc45af7adbb907072d042bbece4c2b4de6e Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Thu, 19 Jan 2012 18:18:42 +0100 Subject: [PATCH 664/820] ACPI processor hotplug: Split up acpi_processor_add No functional change. This is needed because: When a CPU gets hotplugged, it's totally uninitialized and offline. cpuinfo_x86 struct (cpu_data(cpu)) is mostly zero (CPU feature flags, model, family,..). When a CPU gets hotplugged, struct processor is alloc'd, some sysfs files are set up but acpi_processor_add() must not try to access a MSR on this CPU or try to read out CPU feature,family, etc. This must be done in acpi_processor_start(). The next patch will delay the call of acpi_processor_start() for physically hotpluggedcores, to the time when they are onlined the first time. There it is safe then to access cpu_data(cpu) cpuinfo_x86 struct or access MSRs which is needed to set up cpuidle, throttling and other features. Tested and Signed-off-by: Thomas Renninger Signed-off-by: Len Brown --- drivers/acpi/processor_driver.c | 92 +++++++++++++++++++-------------- 1 file changed, 54 insertions(+), 38 deletions(-) diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index 0034ede38710..bec55937cf10 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -440,6 +440,58 @@ static struct notifier_block acpi_cpu_notifier = .notifier_call = acpi_cpu_soft_notify, }; +static int __cpuinit acpi_processor_start(struct acpi_processor *pr) +{ + struct acpi_device *device = per_cpu(processor_device_array, pr->id); + int result = 0; + +#ifdef CONFIG_CPU_FREQ + acpi_processor_ppc_has_changed(pr, 0); +#endif + acpi_processor_get_throttling_info(pr); + acpi_processor_get_limit_info(pr); + + if (!cpuidle_get_driver() || cpuidle_get_driver() == &acpi_idle_driver) + acpi_processor_power_init(pr, device); + + pr->cdev = thermal_cooling_device_register("Processor", device, + &processor_cooling_ops); + if (IS_ERR(pr->cdev)) { + result = PTR_ERR(pr->cdev); + goto err_power_exit; + } + + dev_dbg(&device->dev, "registered as cooling_device%d\n", + pr->cdev->id); + + result = sysfs_create_link(&device->dev.kobj, + &pr->cdev->device.kobj, + "thermal_cooling"); + if (result) { + printk(KERN_ERR PREFIX "Create sysfs link\n"); + goto err_thermal_unregister; + } + result = sysfs_create_link(&pr->cdev->device.kobj, + &device->dev.kobj, + "device"); + if (result) { + printk(KERN_ERR PREFIX "Create sysfs link\n"); + goto err_remove_sysfs_thermal; + } + + return 0; + +err_remove_sysfs_thermal: + sysfs_remove_link(&device->dev.kobj, "thermal_cooling"); +err_thermal_unregister: + thermal_cooling_device_unregister(pr->cdev); +err_power_exit: + acpi_processor_power_exit(pr, device); + + return result; +} + + static int __cpuinit acpi_processor_add(struct acpi_device *device) { struct acpi_processor *pr = NULL; @@ -494,49 +546,13 @@ static int __cpuinit acpi_processor_add(struct acpi_device *device) result = -EFAULT; goto err_free_cpumask; } - -#ifdef CONFIG_CPU_FREQ - acpi_processor_ppc_has_changed(pr, 0); -#endif - acpi_processor_get_throttling_info(pr); - acpi_processor_get_limit_info(pr); - - if (!cpuidle_get_driver() || cpuidle_get_driver() == &acpi_idle_driver) - acpi_processor_power_init(pr, device); - - pr->cdev = thermal_cooling_device_register("Processor", device, - &processor_cooling_ops); - if (IS_ERR(pr->cdev)) { - result = PTR_ERR(pr->cdev); - goto err_power_exit; - } - - dev_dbg(&device->dev, "registered as cooling_device%d\n", - pr->cdev->id); - - result = sysfs_create_link(&device->dev.kobj, - &pr->cdev->device.kobj, - "thermal_cooling"); - if (result) { - printk(KERN_ERR PREFIX "Create sysfs link\n"); - goto err_thermal_unregister; - } - result = sysfs_create_link(&pr->cdev->device.kobj, - &device->dev.kobj, - "device"); - if (result) { - printk(KERN_ERR PREFIX "Create sysfs link\n"); + result = acpi_processor_start(pr); + if (result) goto err_remove_sysfs; - } return 0; err_remove_sysfs: - sysfs_remove_link(&device->dev.kobj, "thermal_cooling"); -err_thermal_unregister: - thermal_cooling_device_unregister(pr->cdev); -err_power_exit: - acpi_processor_power_exit(pr, device); sysfs_remove_link(&device->dev.kobj, "sysdev"); err_free_cpumask: free_cpumask_var(pr->throttling.shared_cpu_map); From 99b725084450bbc6f8e1ab20a0df4cc291c342b5 Mon Sep 17 00:00:00 2001 From: Thomas Renninger Date: Thu, 19 Jan 2012 18:18:43 +0100 Subject: [PATCH 665/820] ACPI processor hotplug: Delay acpi_processor_start() call for hotplugged cores Delay the setting up of features (cpuidle, throttling by calling acpi_processor_start()) to the time when the hotplugged core got onlined the first time and got fully initialized. Signed-off-by: Thomas Renninger Signed-off-by: Len Brown --- drivers/acpi/processor_driver.c | 72 +++++++++++++++++++++++++++++---- drivers/idle/intel_idle.c | 2 +- include/acpi/processor.h | 1 + 3 files changed, 67 insertions(+), 8 deletions(-) diff --git a/drivers/acpi/processor_driver.c b/drivers/acpi/processor_driver.c index bec55937cf10..2b805d7ef317 100644 --- a/drivers/acpi/processor_driver.c +++ b/drivers/acpi/processor_driver.c @@ -84,7 +84,7 @@ static int acpi_processor_remove(struct acpi_device *device, int type); static void acpi_processor_notify(struct acpi_device *device, u32 event); static acpi_status acpi_processor_hotadd_init(struct acpi_processor *pr); static int acpi_processor_handle_eject(struct acpi_processor *pr); - +static int acpi_processor_start(struct acpi_processor *pr); static const struct acpi_device_id processor_device_ids[] = { {ACPI_PROCESSOR_OBJECT_HID, 0}, @@ -423,10 +423,29 @@ static int acpi_cpu_soft_notify(struct notifier_block *nfb, struct acpi_processor *pr = per_cpu(processors, cpu); if (action == CPU_ONLINE && pr) { - acpi_processor_ppc_has_changed(pr, 0); - acpi_processor_hotplug(pr); - acpi_processor_reevaluate_tstate(pr, action); - acpi_processor_tstate_has_changed(pr); + /* CPU got physically hotplugged and onlined the first time: + * Initialize missing things + */ + if (pr->flags.need_hotplug_init) { + struct cpuidle_driver *idle_driver = + cpuidle_get_driver(); + + printk(KERN_INFO "Will online and init hotplugged " + "CPU: %d\n", pr->id); + WARN(acpi_processor_start(pr), "Failed to start CPU:" + " %d\n", pr->id); + pr->flags.need_hotplug_init = 0; + if (idle_driver && !strcmp(idle_driver->name, + "intel_idle")) { + intel_idle_cpu_init(pr->id); + } + /* Normal CPU soft online event */ + } else { + acpi_processor_ppc_has_changed(pr, 0); + acpi_processor_cst_has_changed(pr); + acpi_processor_reevaluate_tstate(pr, action); + acpi_processor_tstate_has_changed(pr); + } } if (action == CPU_DEAD && pr) { /* invalidate the flag.throttling after one CPU is offline */ @@ -440,7 +459,15 @@ static struct notifier_block acpi_cpu_notifier = .notifier_call = acpi_cpu_soft_notify, }; -static int __cpuinit acpi_processor_start(struct acpi_processor *pr) +/* + * acpi_processor_start() is called by the cpu_hotplug_notifier func: + * acpi_cpu_soft_notify(). Getting it __cpuinit{data} is difficult, the + * root cause seem to be that acpi_processor_uninstall_hotplug_notify() + * is in the module_exit (__exit) func. Allowing acpi_processor_start() + * to not be in __cpuinit section, but being called from __cpuinit funcs + * via __ref looks like the right thing to do here. + */ +static __ref int acpi_processor_start(struct acpi_processor *pr) { struct acpi_device *device = per_cpu(processor_device_array, pr->id); int result = 0; @@ -491,7 +518,12 @@ err_power_exit: return result; } - +/* + * Do not put anything in here which needs the core to be online. + * For example MSR access or setting up things which check for cpuinfo_x86 + * (cpu_data(cpu)) values, like CPU feature flags, family, model, etc. + * Such things have to be put in and set up above in acpi_processor_start() + */ static int __cpuinit acpi_processor_add(struct acpi_device *device) { struct acpi_processor *pr = NULL; @@ -546,6 +578,21 @@ static int __cpuinit acpi_processor_add(struct acpi_device *device) result = -EFAULT; goto err_free_cpumask; } + + /* + * Do not start hotplugged CPUs now, but when they + * are onlined the first time + */ + if (pr->flags.need_hotplug_init) + return 0; + + /* + * Do not start hotplugged CPUs now, but when they + * are onlined the first time + */ + if (pr->flags.need_hotplug_init) + return 0; + result = acpi_processor_start(pr); if (result) goto err_remove_sysfs; @@ -751,6 +798,17 @@ static acpi_status acpi_processor_hotadd_init(struct acpi_processor *pr) return AE_ERROR; } + /* CPU got hot-plugged, but cpu_data is not initialized yet + * Set flag to delay cpu_idle/throttling initialization + * in: + * acpi_processor_add() + * acpi_processor_get_info() + * and do it when the CPU gets online the first time + * TBD: Cleanup above functions and try to do this more elegant. + */ + printk(KERN_INFO "CPU %d got hotplugged\n", pr->id); + pr->flags.need_hotplug_init = 1; + return AE_OK; } diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c index 20bce51c2e82..54ab97bae042 100644 --- a/drivers/idle/intel_idle.c +++ b/drivers/idle/intel_idle.c @@ -527,7 +527,7 @@ int intel_idle_cpu_init(int cpu) return 0; } - +EXPORT_SYMBOL_GPL(intel_idle_cpu_init); static int __init intel_idle_init(void) { diff --git a/include/acpi/processor.h b/include/acpi/processor.h index 610f6fb1bbc2..8cf7e98a2c7b 100644 --- a/include/acpi/processor.h +++ b/include/acpi/processor.h @@ -195,6 +195,7 @@ struct acpi_processor_flags { u8 has_cst:1; u8 power_setup_done:1; u8 bm_rld_set:1; + u8 need_hotplug_init:1; }; struct acpi_processor { From b9ecc4ee28a5ff5b3997da247cd9df1320c602a9 Mon Sep 17 00:00:00 2001 From: Albert Pool Date: Thu, 19 Jan 2012 22:08:50 +0100 Subject: [PATCH 666/820] snd-hda-intel: better Alienware M17x R3 quirk I have been told that this way the rear headphone connector is working as well; with model=alienware only laptop speakers work. The subsystem of both controller and codec is 1028:0490. Signed-off-by: Albert Pool Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_sigmatel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 3556408d6ece..1a26dbca9483 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -1608,7 +1608,7 @@ static const struct snd_pci_quirk stac92hd73xx_codec_id_cfg_tbl[] = { SND_PCI_QUIRK(PCI_VENDOR_ID_DELL, 0x043a, "Alienware M17x", STAC_ALIENWARE_M17X), SND_PCI_QUIRK(PCI_VENDOR_ID_DELL, 0x0490, - "Alienware M17x", STAC_ALIENWARE_M17X), + "Alienware M17x R3", STAC_DELL_EQ), {} /* terminator */ }; From cb0cdebbf8b834110ef67ed9335d5bafed7835df Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 20 Jan 2012 12:14:12 +0100 Subject: [PATCH 667/820] ALSA: hda - Fix a unused variable warning MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Just overlooked. sound/pci/hda/patch_realtek.c: In function ‘alc_sku_unsol_event’: sound/pci/hda/patch_realtek.c:643:19: warning: unused variable ‘spec’ [-Wunused-variable] Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 1 - 1 file changed, 1 deletion(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index ddbed9705ef4..c95c8bde12d0 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -640,7 +640,6 @@ static void alc_exec_unsol_event(struct hda_codec *codec, int action) /* unsolicited event for HP jack sensing */ static void alc_sku_unsol_event(struct hda_codec *codec, unsigned int res) { - struct alc_spec *spec = codec->spec; if (codec->vendor_id == 0x10ec0880) res >>= 28; else From 421e8d2de3bd8b089dc6322d8589b7eb38437a23 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Fri, 20 Jan 2012 13:39:37 +0000 Subject: [PATCH 668/820] regmap: Reset cache status when reinitialsing the cache When we reinitialise the cache make sure that we reset the cache access flags, ensuring that the reinitialised cache is in the default state which is what callers would and do expect given the function name. This is particularly likely to cause issues in systems where there was no cache previously as those systems have cache bypass enabled, as for the wm8994 driver where this was noticed. Signed-off-by: Mark Brown --- drivers/base/regmap/regmap.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/base/regmap/regmap.c b/drivers/base/regmap/regmap.c index be10a4ff6609..65558034318f 100644 --- a/drivers/base/regmap/regmap.c +++ b/drivers/base/regmap/regmap.c @@ -284,6 +284,9 @@ int regmap_reinit_cache(struct regmap *map, const struct regmap_config *config) map->precious_reg = config->precious_reg; map->cache_type = config->cache_type; + map->cache_bypass = false; + map->cache_only = false; + ret = regcache_init(map, config); mutex_unlock(&map->lock); From bb362e2e4f4874f3fd4cbc2497385b9bceb3a08a Mon Sep 17 00:00:00 2001 From: Zeng Zhaoming Date: Wed, 18 Jan 2012 13:58:07 +0800 Subject: [PATCH 669/820] ASoC: sgtl5000: Fix wrong register name in restore Correct SGTL5000_CHIP_CLK_CTRL to SGTL5000_CHIP_REF_CTRL in sgtl5000_restore_regs(), and add comment to explain the restore order. Reported-by: Julia Lawall Signed-off-by: Zeng Zhaoming Acked-by: Dong Aisheng Signed-off-by: Mark Brown --- sound/soc/codecs/sgtl5000.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/sgtl5000.c b/sound/soc/codecs/sgtl5000.c index f8863ebb4304..7f4ba819a9f6 100644 --- a/sound/soc/codecs/sgtl5000.c +++ b/sound/soc/codecs/sgtl5000.c @@ -987,12 +987,12 @@ static int sgtl5000_restore_regs(struct snd_soc_codec *codec) /* restore regular registers */ for (reg = 0; reg <= SGTL5000_CHIP_SHORT_CTRL; reg += 2) { - /* this regs depends on the others */ + /* These regs should restore in particular order */ if (reg == SGTL5000_CHIP_ANA_POWER || reg == SGTL5000_CHIP_CLK_CTRL || reg == SGTL5000_CHIP_LINREG_CTRL || reg == SGTL5000_CHIP_LINE_OUT_CTRL || - reg == SGTL5000_CHIP_CLK_CTRL) + reg == SGTL5000_CHIP_REF_CTRL) continue; snd_soc_write(codec, reg, cache[reg]); @@ -1003,8 +1003,17 @@ static int sgtl5000_restore_regs(struct snd_soc_codec *codec) snd_soc_write(codec, reg, cache[reg]); /* - * restore power and other regs according - * to set_power() and set_clock() + * restore these regs according to the power setting sequence in + * sgtl5000_set_power_regs() and clock setting sequence in + * sgtl5000_set_clock(). + * + * The order of restore is: + * 1. SGTL5000_CHIP_CLK_CTRL MCLK_FREQ bits (1:0) should be restore after + * SGTL5000_CHIP_ANA_POWER PLL bits set + * 2. SGTL5000_CHIP_LINREG_CTRL should be set before + * SGTL5000_CHIP_ANA_POWER LINREG_D restored + * 3. SGTL5000_CHIP_REF_CTRL controls Analog Ground Voltage, + * prefer to resotre it after SGTL5000_CHIP_ANA_POWER restored */ snd_soc_write(codec, SGTL5000_CHIP_LINREG_CTRL, cache[SGTL5000_CHIP_LINREG_CTRL]); From 01b37e94c04bc6dae2c4837a2eb6fff6819ea82a Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Wed, 18 Jan 2012 11:48:58 +0100 Subject: [PATCH 670/820] ASoC: tlv320aic32x4: always enable dividers Dividers (such as MDAC) are always needed, independent of the codec being I2S master or slave. Needed on a custom board where the codec has to be slave. Signed-off-by: Wolfram Sang Acked-by: Javier Martin Signed-off-by: Mark Brown --- sound/soc/codecs/tlv320aic32x4.c | 82 ++++++++++++++------------------ 1 file changed, 36 insertions(+), 46 deletions(-) diff --git a/sound/soc/codecs/tlv320aic32x4.c b/sound/soc/codecs/tlv320aic32x4.c index eb401ef021fb..3806cb6d9d4d 100644 --- a/sound/soc/codecs/tlv320aic32x4.c +++ b/sound/soc/codecs/tlv320aic32x4.c @@ -60,7 +60,6 @@ struct aic32x4_rate_divs { struct aic32x4_priv { u32 sysclk; - s32 master; u8 page_no; void *control_data; u32 power_cfg; @@ -369,7 +368,6 @@ static int aic32x4_set_dai_sysclk(struct snd_soc_dai *codec_dai, static int aic32x4_set_dai_fmt(struct snd_soc_dai *codec_dai, unsigned int fmt) { struct snd_soc_codec *codec = codec_dai->codec; - struct aic32x4_priv *aic32x4 = snd_soc_codec_get_drvdata(codec); u8 iface_reg_1; u8 iface_reg_2; u8 iface_reg_3; @@ -384,11 +382,9 @@ static int aic32x4_set_dai_fmt(struct snd_soc_dai *codec_dai, unsigned int fmt) /* set master/slave audio interface */ switch (fmt & SND_SOC_DAIFMT_MASTER_MASK) { case SND_SOC_DAIFMT_CBM_CFM: - aic32x4->master = 1; iface_reg_1 |= AIC32X4_BCLKMASTER | AIC32X4_WCLKMASTER; break; case SND_SOC_DAIFMT_CBS_CFS: - aic32x4->master = 0; break; default: printk(KERN_ERR "aic32x4: invalid DAI master/slave interface\n"); @@ -526,64 +522,58 @@ static int aic32x4_mute(struct snd_soc_dai *dai, int mute) static int aic32x4_set_bias_level(struct snd_soc_codec *codec, enum snd_soc_bias_level level) { - struct aic32x4_priv *aic32x4 = snd_soc_codec_get_drvdata(codec); - switch (level) { case SND_SOC_BIAS_ON: - if (aic32x4->master) { - /* Switch on PLL */ - snd_soc_update_bits(codec, AIC32X4_PLLPR, - AIC32X4_PLLEN, AIC32X4_PLLEN); + /* Switch on PLL */ + snd_soc_update_bits(codec, AIC32X4_PLLPR, + AIC32X4_PLLEN, AIC32X4_PLLEN); - /* Switch on NDAC Divider */ - snd_soc_update_bits(codec, AIC32X4_NDAC, - AIC32X4_NDACEN, AIC32X4_NDACEN); + /* Switch on NDAC Divider */ + snd_soc_update_bits(codec, AIC32X4_NDAC, + AIC32X4_NDACEN, AIC32X4_NDACEN); - /* Switch on MDAC Divider */ - snd_soc_update_bits(codec, AIC32X4_MDAC, - AIC32X4_MDACEN, AIC32X4_MDACEN); + /* Switch on MDAC Divider */ + snd_soc_update_bits(codec, AIC32X4_MDAC, + AIC32X4_MDACEN, AIC32X4_MDACEN); - /* Switch on NADC Divider */ - snd_soc_update_bits(codec, AIC32X4_NADC, - AIC32X4_NADCEN, AIC32X4_NADCEN); + /* Switch on NADC Divider */ + snd_soc_update_bits(codec, AIC32X4_NADC, + AIC32X4_NADCEN, AIC32X4_NADCEN); - /* Switch on MADC Divider */ - snd_soc_update_bits(codec, AIC32X4_MADC, - AIC32X4_MADCEN, AIC32X4_MADCEN); + /* Switch on MADC Divider */ + snd_soc_update_bits(codec, AIC32X4_MADC, + AIC32X4_MADCEN, AIC32X4_MADCEN); - /* Switch on BCLK_N Divider */ - snd_soc_update_bits(codec, AIC32X4_BCLKN, - AIC32X4_BCLKEN, AIC32X4_BCLKEN); - } + /* Switch on BCLK_N Divider */ + snd_soc_update_bits(codec, AIC32X4_BCLKN, + AIC32X4_BCLKEN, AIC32X4_BCLKEN); break; case SND_SOC_BIAS_PREPARE: break; case SND_SOC_BIAS_STANDBY: - if (aic32x4->master) { - /* Switch off PLL */ - snd_soc_update_bits(codec, AIC32X4_PLLPR, - AIC32X4_PLLEN, 0); + /* Switch off PLL */ + snd_soc_update_bits(codec, AIC32X4_PLLPR, + AIC32X4_PLLEN, 0); - /* Switch off NDAC Divider */ - snd_soc_update_bits(codec, AIC32X4_NDAC, - AIC32X4_NDACEN, 0); + /* Switch off NDAC Divider */ + snd_soc_update_bits(codec, AIC32X4_NDAC, + AIC32X4_NDACEN, 0); - /* Switch off MDAC Divider */ - snd_soc_update_bits(codec, AIC32X4_MDAC, - AIC32X4_MDACEN, 0); + /* Switch off MDAC Divider */ + snd_soc_update_bits(codec, AIC32X4_MDAC, + AIC32X4_MDACEN, 0); - /* Switch off NADC Divider */ - snd_soc_update_bits(codec, AIC32X4_NADC, - AIC32X4_NADCEN, 0); + /* Switch off NADC Divider */ + snd_soc_update_bits(codec, AIC32X4_NADC, + AIC32X4_NADCEN, 0); - /* Switch off MADC Divider */ - snd_soc_update_bits(codec, AIC32X4_MADC, - AIC32X4_MADCEN, 0); + /* Switch off MADC Divider */ + snd_soc_update_bits(codec, AIC32X4_MADC, + AIC32X4_MADCEN, 0); - /* Switch off BCLK_N Divider */ - snd_soc_update_bits(codec, AIC32X4_BCLKN, - AIC32X4_BCLKEN, 0); - } + /* Switch off BCLK_N Divider */ + snd_soc_update_bits(codec, AIC32X4_BCLKN, + AIC32X4_BCLKEN, 0); break; case SND_SOC_BIAS_OFF: break; From 0c93a167a6b3fa510c74e88477852c41defda075 Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Wed, 18 Jan 2012 11:48:59 +0100 Subject: [PATCH 671/820] ASoC: tlv320aic32x4: always enable analouge block Register LDOCTLEN must always be initialized to clear the analog power control bit, otherwise the analog block will stay deactivated. Signed-off-by: Wolfram Sang Acked-by: Javier Martin Signed-off-by: Mark Brown --- sound/soc/codecs/tlv320aic32x4.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/tlv320aic32x4.c b/sound/soc/codecs/tlv320aic32x4.c index 3806cb6d9d4d..372b0b83bd9f 100644 --- a/sound/soc/codecs/tlv320aic32x4.c +++ b/sound/soc/codecs/tlv320aic32x4.c @@ -641,9 +641,11 @@ static int aic32x4_probe(struct snd_soc_codec *codec) if (aic32x4->power_cfg & AIC32X4_PWR_AVDD_DVDD_WEAK_DISABLE) { snd_soc_write(codec, AIC32X4_PWRCFG, AIC32X4_AVDDWEAKDISABLE); } - if (aic32x4->power_cfg & AIC32X4_PWR_AIC32X4_LDO_ENABLE) { - snd_soc_write(codec, AIC32X4_LDOCTL, AIC32X4_LDOCTLEN); - } + + tmp_reg = (aic32x4->power_cfg & AIC32X4_PWR_AIC32X4_LDO_ENABLE) ? + AIC32X4_LDOCTLEN : 0; + snd_soc_write(codec, AIC32X4_LDOCTL, tmp_reg); + tmp_reg = snd_soc_read(codec, AIC32X4_CMMODE); if (aic32x4->power_cfg & AIC32X4_PWR_CMMODE_LDOIN_RANGE_18_36) { tmp_reg |= AIC32X4_LDOIN_18_36; From e53e417331c57b9b97e3f8be870214a02c99265c Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 18 Jan 2012 20:02:38 +0000 Subject: [PATCH 672/820] ASoC: Mark WM5100 register map cache only when going into BIAS_OFF Writing to the registers won't work if we do actually manage to hit a fully powered off state. Signed-off-by: Mark Brown Cc: stable@kernel.org --- sound/soc/codecs/wm5100.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/codecs/wm5100.c b/sound/soc/codecs/wm5100.c index 8b24323d6b2c..3fd9cfe6dcd7 100644 --- a/sound/soc/codecs/wm5100.c +++ b/sound/soc/codecs/wm5100.c @@ -1402,6 +1402,7 @@ static int wm5100_set_bias_level(struct snd_soc_codec *codec, break; case SND_SOC_BIAS_OFF: + regcache_cache_only(wm5100->regmap, true); if (wm5100->pdata.ldo_ena) gpio_set_value_cansleep(wm5100->pdata.ldo_ena, 0); regulator_bulk_disable(ARRAY_SIZE(wm5100->core_supplies), From 495174a8ffbaa0d15153d855cf206cdc46d51cf4 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Thu, 19 Jan 2012 11:16:37 +0000 Subject: [PATCH 673/820] ASoC: Don't go through cache when applying WM5100 rev A updates These are all to either uncached registers or fixes to register defaults, in the former case the cache won't do anything and in the latter case we're fixing things so the cache sync will do the right thing. Signed-off-by: Mark Brown Cc: stable@kernel.org --- sound/soc/codecs/wm5100.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/soc/codecs/wm5100.c b/sound/soc/codecs/wm5100.c index 3fd9cfe6dcd7..66f0611e68b6 100644 --- a/sound/soc/codecs/wm5100.c +++ b/sound/soc/codecs/wm5100.c @@ -1377,6 +1377,7 @@ static int wm5100_set_bias_level(struct snd_soc_codec *codec, switch (wm5100->rev) { case 0: + regcache_cache_bypass(wm5100->regmap, true); snd_soc_write(codec, 0x11, 0x3); snd_soc_write(codec, 0x203, 0xc); snd_soc_write(codec, 0x206, 0); @@ -1392,6 +1393,7 @@ static int wm5100_set_bias_level(struct snd_soc_codec *codec, snd_soc_write(codec, wm5100_reva_patches[i].reg, wm5100_reva_patches[i].val); + regcache_cache_bypass(wm5100->regmap, false); break; default: break; From fed22007113cb857e917913ce016d9b539dc3a80 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Wed, 18 Jan 2012 19:17:06 +0000 Subject: [PATCH 674/820] ASoC: Disable register synchronisation for low frequency WM8996 SYSCLK With a low frequency SYSCLK and a fast I2C clock register synchronisation may occasionally take too long to take effect, causing I/O issues. Disable synchronisation in order to avoid any issues. Signed-off-by: Mark Brown Cc: stable@kernel.org --- sound/soc/codecs/wm8996.c | 4 ++++ sound/soc/codecs/wm8996.h | 4 ++++ 2 files changed, 8 insertions(+) diff --git a/sound/soc/codecs/wm8996.c b/sound/soc/codecs/wm8996.c index d8da10fe5b52..86f5b6bd7af2 100644 --- a/sound/soc/codecs/wm8996.c +++ b/sound/soc/codecs/wm8996.c @@ -2007,6 +2007,7 @@ static int wm8996_set_sysclk(struct snd_soc_dai *dai, struct wm8996_priv *wm8996 = snd_soc_codec_get_drvdata(codec); int lfclk = 0; int ratediv = 0; + int sync = WM8996_REG_SYNC; int src; int old; @@ -2051,6 +2052,7 @@ static int wm8996_set_sysclk(struct snd_soc_dai *dai, case 32000: case 32768: lfclk = WM8996_LFCLK_ENA; + sync = 0; break; default: dev_warn(codec->dev, "Unsupported clock rate %dHz\n", @@ -2064,6 +2066,8 @@ static int wm8996_set_sysclk(struct snd_soc_dai *dai, WM8996_SYSCLK_SRC_MASK | WM8996_SYSCLK_DIV_MASK, src << WM8996_SYSCLK_SRC_SHIFT | ratediv); snd_soc_update_bits(codec, WM8996_CLOCKING_1, WM8996_LFCLK_ENA, lfclk); + snd_soc_update_bits(codec, WM8996_CONTROL_INTERFACE_1, + WM8996_REG_SYNC, sync); snd_soc_update_bits(codec, WM8996_AIF_CLOCKING_1, WM8996_SYSCLK_ENA, old); diff --git a/sound/soc/codecs/wm8996.h b/sound/soc/codecs/wm8996.h index 0fde643194ce..de9ac3e44aec 100644 --- a/sound/soc/codecs/wm8996.h +++ b/sound/soc/codecs/wm8996.h @@ -1567,6 +1567,10 @@ int wm8996_detect(struct snd_soc_codec *codec, struct snd_soc_jack *jack, /* * R257 (0x101) - Control Interface (1) */ +#define WM8996_REG_SYNC 0x8000 /* REG_SYNC */ +#define WM8996_REG_SYNC_MASK 0x8000 /* REG_SYNC */ +#define WM8996_REG_SYNC_SHIFT 15 /* REG_SYNC */ +#define WM8996_REG_SYNC_WIDTH 1 /* REG_SYNC */ #define WM8996_AUTO_INC 0x0004 /* AUTO_INC */ #define WM8996_AUTO_INC_MASK 0x0004 /* AUTO_INC */ #define WM8996_AUTO_INC_SHIFT 2 /* AUTO_INC */ From 53b6b3e00b8454aeab497119443d4f365ee4e46e Mon Sep 17 00:00:00 2001 From: Sumit Semwal Date: Fri, 20 Jan 2012 15:04:25 +0530 Subject: [PATCH 675/820] MAINTAINERS: Add dma-buf sharing framework maintainer Adding maintainer info for dma-buf buffer sharing framework; some mailing lists interested in this work are also added. Signed-off-by: Sumit Semwal Signed-off-by: Sumit Semwal Acked-by: Arnd Bergmann Acked-by: Daniel Vetter Acked-by: Dave Airlie --- MAINTAINERS | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 89b70df91f4f..93c68d5f1cf4 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2246,6 +2246,17 @@ T: git git://git.kernel.org/pub/scm/linux/kernel/git/teigland/dlm.git S: Supported F: fs/dlm/ +DMA BUFFER SHARING FRAMEWORK +M: Sumit Semwal +S: Maintained +L: linux-media@vger.kernel.org +L: dri-devel@lists.freedesktop.org +L: linaro-mm-sig@lists.linaro.org +F: drivers/base/dma-buf* +F: include/linux/dma-buf* +F: Documentation/dma-buf-sharing.txt +T: git git://git.linaro.org/people/sumitsemwal/linux-dma-buf.git + DMA GENERIC OFFLOAD ENGINE SUBSYSTEM M: Vinod Koul M: Dan Williams From 6b35f924b80a0e6d71711e66f5b3c16f427f3d2a Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Thu, 19 Jan 2012 10:23:22 -0200 Subject: [PATCH 676/820] ASoC: mxs: Fix mxs-saif timeout On a mx28evk board the following errors happens on mxs-sgtl5000 probe: [ 0.660000] saif0_clk_set_rate: divider writing timeout [ 0.670000] mxs-sgtl5000: probe of mxs-sgtl5000.0 failed with error -110 [ 0.670000] ALSA device list: [ 0.680000] No soundcards found. This timeout happens because clk_set_rate will result in writing to the DIV bits of register HW_CLKCTRL_SAIF0 with the saif clock gated (CLKGATE bit set to one). MX28 Reference states the following about CLKGATE: "The DIV field can change ONLY when this clock gate bit field is low." So call clk_prepare_enable prior to clk_set_rate to fix this problem. After this change the mxs-saif driver can be correctly probed and audio is functional. Signed-off-by: Fabio Estevam Signed-off-by: Mark Brown --- sound/soc/mxs/mxs-saif.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/sound/soc/mxs/mxs-saif.c b/sound/soc/mxs/mxs-saif.c index dccfb37a9626..f204dbac11d4 100644 --- a/sound/soc/mxs/mxs-saif.c +++ b/sound/soc/mxs/mxs-saif.c @@ -124,6 +124,8 @@ static int mxs_saif_set_clk(struct mxs_saif *saif, * * If MCLK is not used, we just set saif clk to 512*fs. */ + clk_prepare_enable(master_saif->clk); + if (master_saif->mclk_in_use) { if (mclk % 32 == 0) { scr &= ~BM_SAIF_CTRL_BITCLK_BASE_RATE; @@ -133,6 +135,7 @@ static int mxs_saif_set_clk(struct mxs_saif *saif, ret = clk_set_rate(master_saif->clk, 384 * rate); } else { /* SAIF MCLK should be either 32x or 48x */ + clk_disable_unprepare(master_saif->clk); return -EINVAL; } } else { @@ -140,6 +143,8 @@ static int mxs_saif_set_clk(struct mxs_saif *saif, scr &= ~BM_SAIF_CTRL_BITCLK_BASE_RATE; } + clk_disable_unprepare(master_saif->clk); + if (ret) return ret; From b67f231ded332461dd31123c4f659c4681223fb1 Mon Sep 17 00:00:00 2001 From: Roopa Prabhu Date: Thu, 19 Jan 2012 22:25:36 +0000 Subject: [PATCH 677/820] enic: fix compile when CONFIG_PCI_IOV is not enabled reverting back change that access enic->num_vfs outside CONFIG_PCI_IOV Reported-by: Randy Dunlap Signed-off-by: Roopa Prabhu Signed-off-by: David S. Miller --- drivers/net/ethernet/cisco/enic/enic_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index 045c468ac8ef..ab3f67f980d8 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -2273,7 +2273,7 @@ static int __devinit enic_probe(struct pci_dev *pdev, #ifdef CONFIG_PCI_IOV int pos = 0; #endif - int num_pps; + int num_pps = 1; /* Allocate net device structure and initialize. Private * instance data is initialized to zero. @@ -2380,11 +2380,11 @@ static int __devinit enic_probe(struct pci_dev *pdev, goto err_out_vnic_unregister; } enic->priv_flags |= ENIC_SRIOV_ENABLED; + num_pps = enic->num_vfs; } } #endif - num_pps = enic->num_vfs ? enic->num_vfs : 1; /* Allocate structure for port profiles */ enic->pp = kcalloc(num_pps, sizeof(*enic->pp), GFP_KERNEL); if (!enic->pp) { From fc16dcd8c2e1e9bc91ed765957e1f2bbf334253e Mon Sep 17 00:00:00 2001 From: Neal Cardwell Date: Wed, 18 Jan 2012 17:47:58 +0000 Subject: [PATCH 678/820] tcp: fix undo after RTO for BIC This patch fixes BIC so that cwnd reductions made during RTOs can be undone (just as they already can be undone when using the default/Reno behavior). When undoing cwnd reductions, BIC-derived congestion control modules were restoring the cwnd from last_max_cwnd. There were two problems with using last_max_cwnd to restore a cwnd during undo: (a) last_max_cwnd was set to 0 on state transitions into TCP_CA_Loss (by calling the module's reset() functions), so cwnd reductions from RTOs could not be undone. (b) when fast_covergence is enabled (which it is by default) last_max_cwnd does not actually hold the value of snd_cwnd before the loss; instead, it holds a scaled-down version of snd_cwnd. This patch makes the following changes: (1) upon undo, revert snd_cwnd to ca->loss_cwnd, which is already, as the existing comment notes, the "congestion window at last loss" (2) stop forgetting ca->loss_cwnd on TCP_CA_Loss events (3) use ca->last_max_cwnd to check if we're in slow start Signed-off-by: Neal Cardwell Signed-off-by: David S. Miller --- net/ipv4/tcp_bic.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/net/ipv4/tcp_bic.c b/net/ipv4/tcp_bic.c index 6187eb4d1dcf..f45e1c242440 100644 --- a/net/ipv4/tcp_bic.c +++ b/net/ipv4/tcp_bic.c @@ -63,7 +63,6 @@ static inline void bictcp_reset(struct bictcp *ca) { ca->cnt = 0; ca->last_max_cwnd = 0; - ca->loss_cwnd = 0; ca->last_cwnd = 0; ca->last_time = 0; ca->epoch_start = 0; @@ -72,7 +71,11 @@ static inline void bictcp_reset(struct bictcp *ca) static void bictcp_init(struct sock *sk) { - bictcp_reset(inet_csk_ca(sk)); + struct bictcp *ca = inet_csk_ca(sk); + + bictcp_reset(ca); + ca->loss_cwnd = 0; + if (initial_ssthresh) tcp_sk(sk)->snd_ssthresh = initial_ssthresh; } @@ -127,7 +130,7 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd) } /* if in slow start or link utilization is very low */ - if (ca->loss_cwnd == 0) { + if (ca->last_max_cwnd == 0) { if (ca->cnt > 20) /* increase cwnd 5% per RTT */ ca->cnt = 20; } @@ -185,7 +188,7 @@ static u32 bictcp_undo_cwnd(struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); const struct bictcp *ca = inet_csk_ca(sk); - return max(tp->snd_cwnd, ca->last_max_cwnd); + return max(tp->snd_cwnd, ca->loss_cwnd); } static void bictcp_state(struct sock *sk, u8 new_state) From 5a45f0086a2dcf50db7e6a0bf5be933880f85127 Mon Sep 17 00:00:00 2001 From: Neal Cardwell Date: Wed, 18 Jan 2012 17:47:59 +0000 Subject: [PATCH 679/820] tcp: fix undo after RTO for CUBIC This patch fixes CUBIC so that cwnd reductions made during RTOs can be undone (just as they already can be undone when using the default/Reno behavior). When undoing cwnd reductions, BIC-derived congestion control modules were restoring the cwnd from last_max_cwnd. There were two problems with using last_max_cwnd to restore a cwnd during undo: (a) last_max_cwnd was set to 0 on state transitions into TCP_CA_Loss (by calling the module's reset() functions), so cwnd reductions from RTOs could not be undone. (b) when fast_covergence is enabled (which it is by default) last_max_cwnd does not actually hold the value of snd_cwnd before the loss; instead, it holds a scaled-down version of snd_cwnd. This patch makes the following changes: (1) upon undo, revert snd_cwnd to ca->loss_cwnd, which is already, as the existing comment notes, the "congestion window at last loss" (2) stop forgetting ca->loss_cwnd on TCP_CA_Loss events (3) use ca->last_max_cwnd to check if we're in slow start Signed-off-by: Neal Cardwell Acked-by: Stephen Hemminger Acked-by: Sangtae Ha Signed-off-by: David S. Miller --- net/ipv4/tcp_cubic.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c index f376b05cca81..a9077f441cb2 100644 --- a/net/ipv4/tcp_cubic.c +++ b/net/ipv4/tcp_cubic.c @@ -107,7 +107,6 @@ static inline void bictcp_reset(struct bictcp *ca) { ca->cnt = 0; ca->last_max_cwnd = 0; - ca->loss_cwnd = 0; ca->last_cwnd = 0; ca->last_time = 0; ca->bic_origin_point = 0; @@ -142,7 +141,10 @@ static inline void bictcp_hystart_reset(struct sock *sk) static void bictcp_init(struct sock *sk) { - bictcp_reset(inet_csk_ca(sk)); + struct bictcp *ca = inet_csk_ca(sk); + + bictcp_reset(ca); + ca->loss_cwnd = 0; if (hystart) bictcp_hystart_reset(sk); @@ -275,7 +277,7 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd) * The initial growth of cubic function may be too conservative * when the available bandwidth is still unknown. */ - if (ca->loss_cwnd == 0 && ca->cnt > 20) + if (ca->last_max_cwnd == 0 && ca->cnt > 20) ca->cnt = 20; /* increase cwnd 5% per RTT */ /* TCP Friendly */ @@ -342,7 +344,7 @@ static u32 bictcp_undo_cwnd(struct sock *sk) { struct bictcp *ca = inet_csk_ca(sk); - return max(tcp_sk(sk)->snd_cwnd, ca->last_max_cwnd); + return max(tcp_sk(sk)->snd_cwnd, ca->loss_cwnd); } static void bictcp_state(struct sock *sk, u8 new_state) From e615bf5b5519862ab66172f4dec7455d6543a578 Mon Sep 17 00:00:00 2001 From: Myron Stowe Date: Fri, 20 Jan 2012 19:13:24 -0700 Subject: [PATCH 680/820] ACPI, APEI: Add 64-bit read/write support for APEI on i386 Base ACPI (CA) currently does not support atomic 64-bit reads and writes (acpi_read() and acpi_write() split 64-bit loads/stores into two 32-bit transfers) yet APEI expects 64-bit transfer capability, even when running on 32-bit systems. This patch implements 64-bit read and write routines for APEI usage. This patch re-factors similar functionality introduced in commit 04c25997c97, bringing it into the ACPI subsystem in preparation for removing ./drivers/acpi/atomicio.[ch]. In the implementation I have replicated acpi_os_read_memory() and acpi_os_write_memory(), creating 64-bit versions for APEI to utilize, as opposed to something more elegant. My thinking is that we should attempt to see if we can get ACPI's CA/OSL changed so that the existing acpi_read() and acpi_write() interfaces are natively 64-bit capable and then subsequently remove the replication. Signed-off-by: Myron Stowe Signed-off-by: Len Brown --- drivers/acpi/apei/apei-base.c | 35 ++-------- drivers/acpi/osl.c | 116 ++++++++++++++++++++++++++++++++++ include/acpi/acpiosxf.h | 4 ++ 3 files changed, 124 insertions(+), 31 deletions(-) diff --git a/drivers/acpi/apei/apei-base.c b/drivers/acpi/apei/apei-base.c index e45350cb6ac8..e5d53b7ddc7e 100644 --- a/drivers/acpi/apei/apei-base.c +++ b/drivers/acpi/apei/apei-base.c @@ -596,33 +596,19 @@ int apei_read(u64 *val, struct acpi_generic_address *reg) { int rc; u64 address; - u32 tmp, width = reg->bit_width; acpi_status status; rc = apei_check_gar(reg, &address); if (rc) return rc; - if (width == 64) - width = 32; /* Break into two 32-bit transfers */ - *val = 0; switch(reg->space_id) { case ACPI_ADR_SPACE_SYSTEM_MEMORY: - status = acpi_os_read_memory((acpi_physical_address) - address, &tmp, width); + status = acpi_os_read_memory64((acpi_physical_address) + address, val, reg->bit_width); if (ACPI_FAILURE(status)) return -EIO; - *val = tmp; - - if (reg->bit_width == 64) { - /* Read the top 32 bits */ - status = acpi_os_read_memory((acpi_physical_address) - (address + 4), &tmp, 32); - if (ACPI_FAILURE(status)) - return -EIO; - *val |= ((u64)tmp << 32); - } break; case ACPI_ADR_SPACE_SYSTEM_IO: status = acpi_os_read_port(address, (u32 *)val, reg->bit_width); @@ -642,31 +628,18 @@ int apei_write(u64 val, struct acpi_generic_address *reg) { int rc; u64 address; - u32 width = reg->bit_width; acpi_status status; rc = apei_check_gar(reg, &address); if (rc) return rc; - if (width == 64) - width = 32; /* Break into two 32-bit transfers */ - switch (reg->space_id) { case ACPI_ADR_SPACE_SYSTEM_MEMORY: - status = acpi_os_write_memory((acpi_physical_address) - address, ACPI_LODWORD(val), - width); + status = acpi_os_write_memory64((acpi_physical_address) + address, val, reg->bit_width); if (ACPI_FAILURE(status)) return -EIO; - - if (reg->bit_width == 64) { - status = acpi_os_write_memory((acpi_physical_address) - (address + 4), - ACPI_HIDWORD(val), 32); - if (ACPI_FAILURE(status)) - return -EIO; - } break; case ACPI_ADR_SPACE_SYSTEM_IO: status = acpi_os_write_port(address, val, reg->bit_width); diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index fcc12d842bcc..5498a6d88ba2 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -710,6 +710,67 @@ acpi_os_read_memory(acpi_physical_address phys_addr, u32 * value, u32 width) return AE_OK; } +#ifdef readq +static inline u64 read64(const volatile void __iomem *addr) +{ + return readq(addr); +} +#else +static inline u64 read64(const volatile void __iomem *addr) +{ + u64 l, h; + l = readl(addr); + h = readl(addr+4); + return l | (h << 32); +} +#endif + +acpi_status +acpi_os_read_memory64(acpi_physical_address phys_addr, u64 *value, u32 width) +{ + void __iomem *virt_addr; + unsigned int size = width / 8; + bool unmap = false; + u64 dummy; + + rcu_read_lock(); + virt_addr = acpi_map_vaddr_lookup(phys_addr, size); + if (!virt_addr) { + rcu_read_unlock(); + virt_addr = acpi_os_ioremap(phys_addr, size); + if (!virt_addr) + return AE_BAD_ADDRESS; + unmap = true; + } + + if (!value) + value = &dummy; + + switch (width) { + case 8: + *(u8 *) value = readb(virt_addr); + break; + case 16: + *(u16 *) value = readw(virt_addr); + break; + case 32: + *(u32 *) value = readl(virt_addr); + break; + case 64: + *(u64 *) value = read64(virt_addr); + break; + default: + BUG(); + } + + if (unmap) + iounmap(virt_addr); + else + rcu_read_unlock(); + + return AE_OK; +} + acpi_status acpi_os_write_memory(acpi_physical_address phys_addr, u32 value, u32 width) { @@ -749,6 +810,61 @@ acpi_os_write_memory(acpi_physical_address phys_addr, u32 value, u32 width) return AE_OK; } +#ifdef writeq +static inline void write64(u64 val, volatile void __iomem *addr) +{ + writeq(val, addr); +} +#else +static inline void write64(u64 val, volatile void __iomem *addr) +{ + writel(val, addr); + writel(val>>32, addr+4); +} +#endif + +acpi_status +acpi_os_write_memory64(acpi_physical_address phys_addr, u64 value, u32 width) +{ + void __iomem *virt_addr; + unsigned int size = width / 8; + bool unmap = false; + + rcu_read_lock(); + virt_addr = acpi_map_vaddr_lookup(phys_addr, size); + if (!virt_addr) { + rcu_read_unlock(); + virt_addr = acpi_os_ioremap(phys_addr, size); + if (!virt_addr) + return AE_BAD_ADDRESS; + unmap = true; + } + + switch (width) { + case 8: + writeb(value, virt_addr); + break; + case 16: + writew(value, virt_addr); + break; + case 32: + writel(value, virt_addr); + break; + case 64: + write64(value, virt_addr); + break; + default: + BUG(); + } + + if (unmap) + iounmap(virt_addr); + else + rcu_read_unlock(); + + return AE_OK; +} + acpi_status acpi_os_read_pci_configuration(struct acpi_pci_id * pci_id, u32 reg, u64 *value, u32 width) diff --git a/include/acpi/acpiosxf.h b/include/acpi/acpiosxf.h index 2fe8639b3ae7..7c9aebe8a7aa 100644 --- a/include/acpi/acpiosxf.h +++ b/include/acpi/acpiosxf.h @@ -218,9 +218,13 @@ acpi_status acpi_os_write_port(acpi_io_address address, u32 value, u32 width); */ acpi_status acpi_os_read_memory(acpi_physical_address address, u32 * value, u32 width); +acpi_status +acpi_os_read_memory64(acpi_physical_address address, u64 *value, u32 width); acpi_status acpi_os_write_memory(acpi_physical_address address, u32 value, u32 width); +acpi_status +acpi_os_write_memory64(acpi_physical_address address, u64 value, u32 width); /* * Platform and hardware-independent PCI configuration space access From ba242d5b1a84bc6611732296517ee40d5a80a4d9 Mon Sep 17 00:00:00 2001 From: Myron Stowe Date: Fri, 20 Jan 2012 19:13:30 -0700 Subject: [PATCH 681/820] ACPI, APEI: Add RAM mapping support to ACPI This patch adds support for RAM to ACPI's mapping capabilities in order to support APEI error injection (EINJ) actions. This patch re-factors similar functionality introduced in commit 76da3fb3575, bringing it into osl.c in preparation for removing ./drivers/acpi/atomicio.[ch]. Signed-off-by: Myron Stowe Signed-off-by: Len Brown --- drivers/acpi/osl.c | 36 ++++++++++++++++++++++++++++++++++-- 1 file changed, 34 insertions(+), 2 deletions(-) diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 5498a6d88ba2..412a1e04a922 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include @@ -321,6 +322,37 @@ acpi_map_lookup_virt(void __iomem *virt, acpi_size size) return NULL; } +#ifndef CONFIG_IA64 +#define should_use_kmap(pfn) page_is_ram(pfn) +#else +/* ioremap will take care of cache attributes */ +#define should_use_kmap(pfn) 0 +#endif + +static void __iomem *acpi_map(acpi_physical_address pg_off, unsigned long pg_sz) +{ + unsigned long pfn; + + pfn = pg_off >> PAGE_SHIFT; + if (should_use_kmap(pfn)) { + if (pg_sz > PAGE_SIZE) + return NULL; + return (void __iomem __force *)kmap(pfn_to_page(pfn)); + } else + return acpi_os_ioremap(pg_off, pg_sz); +} + +static void acpi_unmap(acpi_physical_address pg_off, void __iomem *vaddr) +{ + unsigned long pfn; + + pfn = pg_off >> PAGE_SHIFT; + if (page_is_ram(pfn)) + kunmap(pfn_to_page(pfn)); + else + iounmap(vaddr); +} + void __iomem *__init_refok acpi_os_map_memory(acpi_physical_address phys, acpi_size size) { @@ -353,7 +385,7 @@ acpi_os_map_memory(acpi_physical_address phys, acpi_size size) pg_off = round_down(phys, PAGE_SIZE); pg_sz = round_up(phys + size, PAGE_SIZE) - pg_off; - virt = acpi_os_ioremap(pg_off, pg_sz); + virt = acpi_map(pg_off, pg_sz); if (!virt) { mutex_unlock(&acpi_ioremap_lock); kfree(map); @@ -384,7 +416,7 @@ static void acpi_os_map_cleanup(struct acpi_ioremap *map) { if (!map->refcount) { synchronize_rcu(); - iounmap(map->virt); + acpi_unmap(map->phys, map->virt); kfree(map); } } From 1001a3a307ca2425c028fd23b8ad05ad8bb048dd Mon Sep 17 00:00:00 2001 From: Myron Stowe Date: Fri, 20 Jan 2012 19:13:37 -0700 Subject: [PATCH 682/820] ACPI: Remove ./drivers/acpi/atomicio.[ch] With the conversion of atomicio's routines in place (see commits 6f68c91c55e and 700130b41f4), atomicio.[ch] can be removed, replacing the APEI specific pre-mapping capabilities with the more generalized versions that drivers/acpi/osl.c provides. Signed-off-by: Myron Stowe Signed-off-by: Len Brown --- drivers/acpi/Makefile | 1 - drivers/acpi/atomicio.c | 422 ---------------------------------------- include/acpi/atomicio.h | 10 - 3 files changed, 433 deletions(-) delete mode 100644 drivers/acpi/atomicio.c delete mode 100644 include/acpi/atomicio.h diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile index c07f44f05f9d..1567028d2038 100644 --- a/drivers/acpi/Makefile +++ b/drivers/acpi/Makefile @@ -19,7 +19,6 @@ obj-y += acpi.o \ # All the builtin files are in the "acpi." module_param namespace. acpi-y += osl.o utils.o reboot.o -acpi-y += atomicio.o acpi-y += nvs.o # sleep related files diff --git a/drivers/acpi/atomicio.c b/drivers/acpi/atomicio.c deleted file mode 100644 index d4a5b3d3657b..000000000000 --- a/drivers/acpi/atomicio.c +++ /dev/null @@ -1,422 +0,0 @@ -/* - * atomicio.c - ACPI IO memory pre-mapping/post-unmapping, then - * accessing in atomic context. - * - * This is used for NMI handler to access IO memory area, because - * ioremap/iounmap can not be used in NMI handler. The IO memory area - * is pre-mapped in process context and accessed in NMI handler. - * - * Copyright (C) 2009-2010, Intel Corp. - * Author: Huang Ying - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License version - * 2 as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#define ACPI_PFX "ACPI: " - -static LIST_HEAD(acpi_iomaps); -/* - * Used for mutual exclusion between writers of acpi_iomaps list, for - * synchronization between readers and writer, RCU is used. - */ -static DEFINE_SPINLOCK(acpi_iomaps_lock); - -struct acpi_iomap { - struct list_head list; - void __iomem *vaddr; - unsigned long size; - phys_addr_t paddr; - struct kref ref; -}; - -/* acpi_iomaps_lock or RCU read lock must be held before calling */ -static struct acpi_iomap *__acpi_find_iomap(phys_addr_t paddr, - unsigned long size) -{ - struct acpi_iomap *map; - - list_for_each_entry_rcu(map, &acpi_iomaps, list) { - if (map->paddr + map->size >= paddr + size && - map->paddr <= paddr) - return map; - } - return NULL; -} - -/* - * Atomic "ioremap" used by NMI handler, if the specified IO memory - * area is not pre-mapped, NULL will be returned. - * - * acpi_iomaps_lock or RCU read lock must be held before calling - */ -static void __iomem *__acpi_ioremap_fast(phys_addr_t paddr, - unsigned long size) -{ - struct acpi_iomap *map; - - map = __acpi_find_iomap(paddr, size/8); - if (map) - return map->vaddr + (paddr - map->paddr); - else - return NULL; -} - -/* acpi_iomaps_lock must be held before calling */ -static void __iomem *__acpi_try_ioremap(phys_addr_t paddr, - unsigned long size) -{ - struct acpi_iomap *map; - - map = __acpi_find_iomap(paddr, size); - if (map) { - kref_get(&map->ref); - return map->vaddr + (paddr - map->paddr); - } else - return NULL; -} - -#ifndef CONFIG_IA64 -#define should_use_kmap(pfn) page_is_ram(pfn) -#else -/* ioremap will take care of cache attributes */ -#define should_use_kmap(pfn) 0 -#endif - -static void __iomem *acpi_map(phys_addr_t pg_off, unsigned long pg_sz) -{ - unsigned long pfn; - - pfn = pg_off >> PAGE_SHIFT; - if (should_use_kmap(pfn)) { - if (pg_sz > PAGE_SIZE) - return NULL; - return (void __iomem __force *)kmap(pfn_to_page(pfn)); - } else - return ioremap(pg_off, pg_sz); -} - -static void acpi_unmap(phys_addr_t pg_off, void __iomem *vaddr) -{ - unsigned long pfn; - - pfn = pg_off >> PAGE_SHIFT; - if (page_is_ram(pfn)) - kunmap(pfn_to_page(pfn)); - else - iounmap(vaddr); -} - -/* - * Used to pre-map the specified IO memory area. First try to find - * whether the area is already pre-mapped, if it is, increase the - * reference count (in __acpi_try_ioremap) and return; otherwise, do - * the real ioremap, and add the mapping into acpi_iomaps list. - */ -static void __iomem *acpi_pre_map(phys_addr_t paddr, - unsigned long size) -{ - void __iomem *vaddr; - struct acpi_iomap *map; - unsigned long pg_sz, flags; - phys_addr_t pg_off; - - spin_lock_irqsave(&acpi_iomaps_lock, flags); - vaddr = __acpi_try_ioremap(paddr, size); - spin_unlock_irqrestore(&acpi_iomaps_lock, flags); - if (vaddr) - return vaddr; - - pg_off = paddr & PAGE_MASK; - pg_sz = ((paddr + size + PAGE_SIZE - 1) & PAGE_MASK) - pg_off; - vaddr = acpi_map(pg_off, pg_sz); - if (!vaddr) - return NULL; - map = kmalloc(sizeof(*map), GFP_KERNEL); - if (!map) - goto err_unmap; - INIT_LIST_HEAD(&map->list); - map->paddr = pg_off; - map->size = pg_sz; - map->vaddr = vaddr; - kref_init(&map->ref); - - spin_lock_irqsave(&acpi_iomaps_lock, flags); - vaddr = __acpi_try_ioremap(paddr, size); - if (vaddr) { - spin_unlock_irqrestore(&acpi_iomaps_lock, flags); - acpi_unmap(pg_off, map->vaddr); - kfree(map); - return vaddr; - } - list_add_tail_rcu(&map->list, &acpi_iomaps); - spin_unlock_irqrestore(&acpi_iomaps_lock, flags); - - return map->vaddr + (paddr - map->paddr); -err_unmap: - acpi_unmap(pg_off, vaddr); - return NULL; -} - -/* acpi_iomaps_lock must be held before calling */ -static void __acpi_kref_del_iomap(struct kref *ref) -{ - struct acpi_iomap *map; - - map = container_of(ref, struct acpi_iomap, ref); - list_del_rcu(&map->list); -} - -/* - * Used to post-unmap the specified IO memory area. The iounmap is - * done only if the reference count goes zero. - */ -static void acpi_post_unmap(phys_addr_t paddr, unsigned long size) -{ - struct acpi_iomap *map; - unsigned long flags; - int del; - - spin_lock_irqsave(&acpi_iomaps_lock, flags); - map = __acpi_find_iomap(paddr, size); - BUG_ON(!map); - del = kref_put(&map->ref, __acpi_kref_del_iomap); - spin_unlock_irqrestore(&acpi_iomaps_lock, flags); - - if (!del) - return; - - synchronize_rcu(); - acpi_unmap(map->paddr, map->vaddr); - kfree(map); -} - -/* In NMI handler, should set silent = 1 */ -static int acpi_check_gar(struct acpi_generic_address *reg, - u64 *paddr, int silent) -{ - u32 width, space_id; - - width = reg->bit_width; - space_id = reg->space_id; - /* Handle possible alignment issues */ - memcpy(paddr, ®->address, sizeof(*paddr)); - if (!*paddr) { - if (!silent) - pr_warning(FW_BUG ACPI_PFX - "Invalid physical address in GAR [0x%llx/%u/%u]\n", - *paddr, width, space_id); - return -EINVAL; - } - - if ((width != 8) && (width != 16) && (width != 32) && (width != 64)) { - if (!silent) - pr_warning(FW_BUG ACPI_PFX - "Invalid bit width in GAR [0x%llx/%u/%u]\n", - *paddr, width, space_id); - return -EINVAL; - } - - if (space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY && - space_id != ACPI_ADR_SPACE_SYSTEM_IO) { - if (!silent) - pr_warning(FW_BUG ACPI_PFX - "Invalid address space type in GAR [0x%llx/%u/%u]\n", - *paddr, width, space_id); - return -EINVAL; - } - - return 0; -} - -/* Pre-map, working on GAR */ -int acpi_pre_map_gar(struct acpi_generic_address *reg) -{ - u64 paddr; - void __iomem *vaddr; - int rc; - - if (reg->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) - return 0; - - rc = acpi_check_gar(reg, &paddr, 0); - if (rc) - return rc; - - vaddr = acpi_pre_map(paddr, reg->bit_width / 8); - if (!vaddr) - return -EIO; - - return 0; -} -EXPORT_SYMBOL_GPL(acpi_pre_map_gar); - -/* Post-unmap, working on GAR */ -int acpi_post_unmap_gar(struct acpi_generic_address *reg) -{ - u64 paddr; - int rc; - - if (reg->space_id != ACPI_ADR_SPACE_SYSTEM_MEMORY) - return 0; - - rc = acpi_check_gar(reg, &paddr, 0); - if (rc) - return rc; - - acpi_post_unmap(paddr, reg->bit_width / 8); - - return 0; -} -EXPORT_SYMBOL_GPL(acpi_post_unmap_gar); - -#ifdef readq -static inline u64 read64(const volatile void __iomem *addr) -{ - return readq(addr); -} -#else -static inline u64 read64(const volatile void __iomem *addr) -{ - u64 l, h; - l = readl(addr); - h = readl(addr+4); - return l | (h << 32); -} -#endif - -/* - * Can be used in atomic (including NMI) or process context. RCU read - * lock can only be released after the IO memory area accessing. - */ -static int acpi_atomic_read_mem(u64 paddr, u64 *val, u32 width) -{ - void __iomem *addr; - - rcu_read_lock(); - addr = __acpi_ioremap_fast(paddr, width); - switch (width) { - case 8: - *val = readb(addr); - break; - case 16: - *val = readw(addr); - break; - case 32: - *val = readl(addr); - break; - case 64: - *val = read64(addr); - break; - default: - return -EINVAL; - } - rcu_read_unlock(); - - return 0; -} - -#ifdef writeq -static inline void write64(u64 val, volatile void __iomem *addr) -{ - writeq(val, addr); -} -#else -static inline void write64(u64 val, volatile void __iomem *addr) -{ - writel(val, addr); - writel(val>>32, addr+4); -} -#endif - -static int acpi_atomic_write_mem(u64 paddr, u64 val, u32 width) -{ - void __iomem *addr; - - rcu_read_lock(); - addr = __acpi_ioremap_fast(paddr, width); - switch (width) { - case 8: - writeb(val, addr); - break; - case 16: - writew(val, addr); - break; - case 32: - writel(val, addr); - break; - case 64: - write64(val, addr); - break; - default: - return -EINVAL; - } - rcu_read_unlock(); - - return 0; -} - -/* GAR accessing in atomic (including NMI) or process context */ -int acpi_atomic_read(u64 *val, struct acpi_generic_address *reg) -{ - u64 paddr; - int rc; - - rc = acpi_check_gar(reg, &paddr, 1); - if (rc) - return rc; - - *val = 0; - switch (reg->space_id) { - case ACPI_ADR_SPACE_SYSTEM_MEMORY: - return acpi_atomic_read_mem(paddr, val, reg->bit_width); - case ACPI_ADR_SPACE_SYSTEM_IO: - return acpi_os_read_port(paddr, (u32 *)val, reg->bit_width); - default: - return -EINVAL; - } -} -EXPORT_SYMBOL_GPL(acpi_atomic_read); - -int acpi_atomic_write(u64 val, struct acpi_generic_address *reg) -{ - u64 paddr; - int rc; - - rc = acpi_check_gar(reg, &paddr, 1); - if (rc) - return rc; - - switch (reg->space_id) { - case ACPI_ADR_SPACE_SYSTEM_MEMORY: - return acpi_atomic_write_mem(paddr, val, reg->bit_width); - case ACPI_ADR_SPACE_SYSTEM_IO: - return acpi_os_write_port(paddr, val, reg->bit_width); - default: - return -EINVAL; - } -} -EXPORT_SYMBOL_GPL(acpi_atomic_write); diff --git a/include/acpi/atomicio.h b/include/acpi/atomicio.h deleted file mode 100644 index 8b9fb4b0b9ce..000000000000 --- a/include/acpi/atomicio.h +++ /dev/null @@ -1,10 +0,0 @@ -#ifndef ACPI_ATOMIC_IO_H -#define ACPI_ATOMIC_IO_H - -int acpi_pre_map_gar(struct acpi_generic_address *reg); -int acpi_post_unmap_gar(struct acpi_generic_address *reg); - -int acpi_atomic_read(u64 *val, struct acpi_generic_address *reg); -int acpi_atomic_write(u64 val, struct acpi_generic_address *reg); - -#endif From 93f770846e8dedc5d9117bd4ad9d7efd18420627 Mon Sep 17 00:00:00 2001 From: Lan Tianyu Date: Sat, 21 Jan 2012 09:23:56 +0800 Subject: [PATCH 683/820] ACPI / PM: Add Sony Vaio VPCCW29FX to nonvs blacklist. Sony Vaio VPCCW29FX does not resume correctly without acpi_sleep=nonvs, so add it to the ACPI sleep blacklist. https://bugzilla.kernel.org/show_bug.cgi?id=34722 Signed-off-by: Lan Tianyu Signed-off-by: Len Brown --- drivers/acpi/sleep.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index 0a7ed69546ba..ca191ff97844 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -438,6 +438,14 @@ static struct dmi_system_id __initdata acpisleep_dmi_table[] = { }, { .callback = init_nvs_nosave, + .ident = "Sony Vaio VPCCW29FX", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Sony Corporation"), + DMI_MATCH(DMI_PRODUCT_NAME, "VPCCW29FX"), + }, + }, + { + .callback = init_nvs_nosave, .ident = "Averatec AV1020-ED2", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "AVERATEC"), From fd45c15f13e754f3c106427e857310f3e0813951 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Fri, 20 Jan 2012 10:12:45 +0900 Subject: [PATCH 684/820] perf: Don't call release_callchain_buffers() if allocation fails When alloc_callchain_buffers() fails, it frees all of entries before return. In addition, calling the release_callchain_buffers() will cause a NULL pointer dereference since callchain_cpu_entries is not set. Signed-off-by: Namhyung Kim Acked-by: Frederic Weisbecker Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Link: http://lkml.kernel.org/r/1327021966-27688-1-git-send-email-namhyung.kim@lge.com Signed-off-by: Ingo Molnar --- kernel/events/callchain.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c index 057e24b665cf..6581a040f399 100644 --- a/kernel/events/callchain.c +++ b/kernel/events/callchain.c @@ -115,8 +115,6 @@ int get_callchain_buffers(void) } err = alloc_callchain_buffers(); - if (err) - release_callchain_buffers(); exit: mutex_unlock(&callchain_mutex); From 46cd6a7f680d14f6f80ede9f04aeb70fa83bd266 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Fri, 20 Jan 2012 10:12:46 +0900 Subject: [PATCH 685/820] perf: Call perf_cgroup_event_time() directly The perf_event_time() will call perf_cgroup_event_time() if @event is a cgroup event. Just do it directly and avoid the extra check.. Signed-off-by: Namhyung Kim Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Link: http://lkml.kernel.org/r/1327021966-27688-2-git-send-email-namhyung.kim@lge.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index a8f4ac001a00..32b48c889711 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -815,7 +815,7 @@ static void update_event_times(struct perf_event *event) * here. */ if (is_cgroup_event(event)) - run_end = perf_event_time(event); + run_end = perf_cgroup_event_time(event); else if (ctx->is_active) run_end = ctx->time; else From aa557875cc6ed78a8c6035dffa354a09d48b16f6 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Wed, 13 Jul 2011 22:33:13 +0200 Subject: [PATCH 686/820] m68k/irq: Remove obsolete IRQ_FLG_* definitions The m68k core irq code stopped honoring these flags during the irq restructuring in 2006. Signed-off-by: Geert Uytterhoeven --- arch/m68k/include/asm/irq.h | 13 ------------- 1 file changed, 13 deletions(-) diff --git a/arch/m68k/include/asm/irq.h b/arch/m68k/include/asm/irq.h index 6198df5ff245..d3a8acd4f1fe 100644 --- a/arch/m68k/include/asm/irq.h +++ b/arch/m68k/include/asm/irq.h @@ -49,19 +49,6 @@ #define IRQ_USER 8 -/* - * various flags for request_irq() - the Amiga now uses the standard - * mechanism like all other architectures - IRQF_DISABLED and - * IRQF_SHARED are your friends. - */ -#ifndef MACH_AMIGA_ONLY -#define IRQ_FLG_LOCK (0x0001) /* handler is not replaceable */ -#define IRQ_FLG_REPLACE (0x0002) /* replace existing handler */ -#define IRQ_FLG_FAST (0x0004) -#define IRQ_FLG_SLOW (0x0008) -#define IRQ_FLG_STD (0x8000) /* internally used */ -#endif - struct irq_data; struct irq_chip; struct irq_desc; From d9070fc4997e255532f0519709c9326d043501b2 Mon Sep 17 00:00:00 2001 From: Finn Thain Date: Mon, 24 Oct 2011 01:11:10 +1100 Subject: [PATCH 687/820] macfb: fix black and white modes macfb won't init in black & white modes since fb_alloc_cmap() no longer works for zero cmap length. Fix this and also clean up a few printk's and some stylistic inconsistencies. Signed-off-by: Finn Thain Signed-off-by: Geert Uytterhoeven --- drivers/video/macfb.c | 60 ++++++++++++++++++++----------------------- 1 file changed, 28 insertions(+), 32 deletions(-) diff --git a/drivers/video/macfb.c b/drivers/video/macfb.c index 43207cc6cc19..fe01add3700e 100644 --- a/drivers/video/macfb.c +++ b/drivers/video/macfb.c @@ -592,12 +592,12 @@ static int __init macfb_init(void) if (!fb_info.screen_base) return -ENODEV; - printk("macfb: framebuffer at 0x%08lx, mapped to 0x%p, size %dk\n", - macfb_fix.smem_start, fb_info.screen_base, - macfb_fix.smem_len / 1024); - printk("macfb: mode is %dx%dx%d, linelength=%d\n", - macfb_defined.xres, macfb_defined.yres, - macfb_defined.bits_per_pixel, macfb_fix.line_length); + pr_info("macfb: framebuffer at 0x%08lx, mapped to 0x%p, size %dk\n", + macfb_fix.smem_start, fb_info.screen_base, + macfb_fix.smem_len / 1024); + pr_info("macfb: mode is %dx%dx%d, linelength=%d\n", + macfb_defined.xres, macfb_defined.yres, + macfb_defined.bits_per_pixel, macfb_fix.line_length); /* Fill in the available video resolution */ macfb_defined.xres_virtual = macfb_defined.xres; @@ -613,14 +613,10 @@ static int __init macfb_init(void) switch (macfb_defined.bits_per_pixel) { case 1: - /* - * XXX: I think this will catch any program that tries - * to do FBIO_PUTCMAP when the visual is monochrome. - */ macfb_defined.red.length = macfb_defined.bits_per_pixel; macfb_defined.green.length = macfb_defined.bits_per_pixel; macfb_defined.blue.length = macfb_defined.bits_per_pixel; - video_cmap_len = 0; + video_cmap_len = 2; macfb_fix.visual = FB_VISUAL_MONO01; break; case 2: @@ -660,11 +656,10 @@ static int __init macfb_init(void) macfb_fix.visual = FB_VISUAL_TRUECOLOR; break; default: - video_cmap_len = 0; - macfb_fix.visual = FB_VISUAL_MONO01; - printk("macfb: unknown or unsupported bit depth: %d\n", + pr_err("macfb: unknown or unsupported bit depth: %d\n", macfb_defined.bits_per_pixel); - break; + err = -EINVAL; + goto fail_unmap; } /* @@ -734,8 +729,8 @@ static int __init macfb_init(void) case MAC_MODEL_Q950: strcpy(macfb_fix.id, "DAFB"); macfb_setpalette = dafb_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; dafb_cmap_regs = ioremap(DAFB_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; /* @@ -744,8 +739,8 @@ static int __init macfb_init(void) case MAC_MODEL_LCII: strcpy(macfb_fix.id, "V8"); macfb_setpalette = v8_brazil_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; v8_brazil_cmap_regs = ioremap(DAC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; /* @@ -758,8 +753,8 @@ static int __init macfb_init(void) case MAC_MODEL_P600: strcpy(macfb_fix.id, "Brazil"); macfb_setpalette = v8_brazil_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; v8_brazil_cmap_regs = ioremap(DAC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; /* @@ -773,10 +768,10 @@ static int __init macfb_init(void) case MAC_MODEL_P520: case MAC_MODEL_P550: case MAC_MODEL_P460: - macfb_setpalette = v8_brazil_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; strcpy(macfb_fix.id, "Sonora"); + macfb_setpalette = v8_brazil_setpalette; v8_brazil_cmap_regs = ioremap(DAC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; /* @@ -786,10 +781,10 @@ static int __init macfb_init(void) */ case MAC_MODEL_IICI: case MAC_MODEL_IISI: - macfb_setpalette = rbv_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; strcpy(macfb_fix.id, "RBV"); + macfb_setpalette = rbv_setpalette; rbv_cmap_regs = ioremap(DAC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; /* @@ -797,10 +792,10 @@ static int __init macfb_init(void) */ case MAC_MODEL_Q840: case MAC_MODEL_C660: - macfb_setpalette = civic_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; strcpy(macfb_fix.id, "Civic"); + macfb_setpalette = civic_setpalette; civic_cmap_regs = ioremap(CIVIC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; @@ -809,26 +804,26 @@ static int __init macfb_init(void) * We think this may be like the LC II */ case MAC_MODEL_LC: + strcpy(macfb_fix.id, "LC"); if (vidtest) { macfb_setpalette = v8_brazil_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; v8_brazil_cmap_regs = ioremap(DAC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; } - strcpy(macfb_fix.id, "LC"); break; /* * We think this may be like the LC II */ case MAC_MODEL_CCL: + strcpy(macfb_fix.id, "Color Classic"); if (vidtest) { macfb_setpalette = v8_brazil_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; v8_brazil_cmap_regs = ioremap(DAC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; } - strcpy(macfb_fix.id, "Color Classic"); break; /* @@ -893,10 +888,10 @@ static int __init macfb_init(void) case MAC_MODEL_PB270C: case MAC_MODEL_PB280: case MAC_MODEL_PB280C: - macfb_setpalette = csc_setpalette; - macfb_defined.activate = FB_ACTIVATE_NOW; strcpy(macfb_fix.id, "CSC"); + macfb_setpalette = csc_setpalette; csc_cmap_regs = ioremap(CSC_BASE, 0x1000); + macfb_defined.activate = FB_ACTIVATE_NOW; break; default: @@ -918,8 +913,9 @@ static int __init macfb_init(void) if (err) goto fail_dealloc; - printk("fb%d: %s frame buffer device\n", - fb_info.node, fb_info.fix.id); + pr_info("fb%d: %s frame buffer device\n", + fb_info.node, fb_info.fix.id); + return 0; fail_dealloc: From 37be2c86f0fbd89b66792008767e688c819b7c32 Mon Sep 17 00:00:00 2001 From: Finn Thain Date: Mon, 24 Oct 2011 01:11:23 +1100 Subject: [PATCH 688/820] mac_scsi: dont enable mac_scsi irq before requesting it Don't enable the SCSI irq when initialising the chip -- the irq has no handler yet. Signed-off-by: Finn Thain Signed-off-by: Geert Uytterhoeven --- drivers/scsi/mac_scsi.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/drivers/scsi/mac_scsi.c b/drivers/scsi/mac_scsi.c index af3a6af97cc7..737d526c0813 100644 --- a/drivers/scsi/mac_scsi.c +++ b/drivers/scsi/mac_scsi.c @@ -340,9 +340,6 @@ static void mac_scsi_reset_boot(struct Scsi_Host *instance) printk(KERN_INFO "Macintosh SCSI: resetting the SCSI bus..." ); - /* switch off SCSI IRQ - catch an interrupt without IRQ bit set else */ - disable_irq(IRQ_MAC_SCSI); - /* get in phase */ NCR5380_write( TARGET_COMMAND_REG, PHASE_SR_TO_TCR( NCR5380_read(STATUS_REG) )); @@ -358,9 +355,6 @@ static void mac_scsi_reset_boot(struct Scsi_Host *instance) for( end = jiffies + AFTER_RESET_DELAY; time_before(jiffies, end); ) barrier(); - /* switch on SCSI IRQ again */ - enable_irq(IRQ_MAC_SCSI); - printk(KERN_INFO " done\n" ); } #endif From c808d3d839ab70c87a6c9356c50569c87661378e Mon Sep 17 00:00:00 2001 From: Finn Thain Date: Mon, 24 Oct 2011 01:11:24 +1100 Subject: [PATCH 689/820] mac_esp: rename irq Rename the "Mac ESP" irq as "ESP" to be consistent with all the other Mac drivers and ESP drivers. Signed-off-by: Finn Thain Signed-off-by: Geert Uytterhoeven --- drivers/scsi/mac_esp.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/scsi/mac_esp.c b/drivers/scsi/mac_esp.c index 4ceeace80453..70eb1f79b1ba 100644 --- a/drivers/scsi/mac_esp.c +++ b/drivers/scsi/mac_esp.c @@ -565,8 +565,7 @@ static int __devinit esp_mac_probe(struct platform_device *dev) esp_chips[dev->id] = esp; mb(); if (esp_chips[!dev->id] == NULL) { - err = request_irq(host->irq, mac_scsi_esp_intr, 0, - "Mac ESP", NULL); + err = request_irq(host->irq, mac_scsi_esp_intr, 0, "ESP", NULL); if (err < 0) { esp_chips[dev->id] = NULL; goto fail_free_priv; From 2a3535069e33d8b416f406c159ce924427315303 Mon Sep 17 00:00:00 2001 From: Andreas Schwab Date: Mon, 9 Jan 2012 15:10:15 +0100 Subject: [PATCH 690/820] m68k: Fix assembler constraint to prevent overeager gcc optimisation Passing the address of a variable as an operand to an asm statement doesn't mark the value of this variable as used, so gcc may optimize its initialisation away. Fix this by using the "m" constraint instead. Signed-off-by: Andreas Schwab Signed-off-by: Geert Uytterhoeven Cc: stable@vger.kernel.org --- arch/m68k/atari/config.c | 8 ++++---- arch/m68k/kernel/process_mm.c | 4 ++-- arch/m68k/kernel/process_no.c | 4 ++-- arch/m68k/kernel/traps.c | 36 +++++++++++++++++------------------ arch/m68k/mm/cache.c | 6 +++--- 5 files changed, 28 insertions(+), 30 deletions(-) diff --git a/arch/m68k/atari/config.c b/arch/m68k/atari/config.c index 4203d101363c..c4ac15c4f065 100644 --- a/arch/m68k/atari/config.c +++ b/arch/m68k/atari/config.c @@ -414,9 +414,9 @@ void __init config_atari(void) * FDC val = 4 -> Supervisor only */ asm volatile ("\n" " .chip 68030\n" - " pmove %0@,%/tt1\n" + " pmove %0,%/tt1\n" " .chip 68k" - : : "a" (&tt1_val)); + : : "m" (tt1_val)); } else { asm volatile ("\n" " .chip 68040\n" @@ -569,10 +569,10 @@ static void atari_reset(void) : "d0"); } else asm volatile ("\n" - " pmove %0@,%%tc\n" + " pmove %0,%%tc\n" " jmp %1@" : /* no outputs */ - : "a" (&tc_val), "a" (reset_addr)); + : "m" (tc_val), "a" (reset_addr)); } diff --git a/arch/m68k/kernel/process_mm.c b/arch/m68k/kernel/process_mm.c index 1bc223aa07ec..aa4ffb882366 100644 --- a/arch/m68k/kernel/process_mm.c +++ b/arch/m68k/kernel/process_mm.c @@ -189,8 +189,8 @@ void flush_thread(void) current->thread.fs = __USER_DS; if (!FPU_IS_EMU) asm volatile (".chip 68k/68881\n\t" - "frestore %0@\n\t" - ".chip 68k" : : "a" (&zero)); + "frestore %0\n\t" + ".chip 68k" : : "m" (zero)); } /* diff --git a/arch/m68k/kernel/process_no.c b/arch/m68k/kernel/process_no.c index 69c1803fcf1b..5e1078cabe0e 100644 --- a/arch/m68k/kernel/process_no.c +++ b/arch/m68k/kernel/process_no.c @@ -163,8 +163,8 @@ void flush_thread(void) #ifdef CONFIG_FPU if (!FPU_IS_EMU) asm volatile (".chip 68k/68881\n\t" - "frestore %0@\n\t" - ".chip 68k" : : "a" (&zero)); + "frestore %0\n\t" + ".chip 68k" : : "m" (zero)); #endif } diff --git a/arch/m68k/kernel/traps.c b/arch/m68k/kernel/traps.c index 89362f2bb56a..eb6746978083 100644 --- a/arch/m68k/kernel/traps.c +++ b/arch/m68k/kernel/traps.c @@ -552,13 +552,13 @@ static inline void bus_error030 (struct frame *fp) #ifdef DEBUG asm volatile ("ptestr %3,%2@,#7,%0\n\t" - "pmove %%psr,%1@" - : "=a&" (desc) - : "a" (&temp), "a" (addr), "d" (ssw)); + "pmove %%psr,%1" + : "=a&" (desc), "=m" (temp) + : "a" (addr), "d" (ssw)); #else asm volatile ("ptestr %2,%1@,#7\n\t" - "pmove %%psr,%0@" - : : "a" (&temp), "a" (addr), "d" (ssw)); + "pmove %%psr,%0" + : "=m" (temp) : "a" (addr), "d" (ssw)); #endif mmusr = temp; @@ -605,20 +605,18 @@ static inline void bus_error030 (struct frame *fp) !(ssw & RW) ? "write" : "read", addr, fp->ptregs.pc, ssw); asm volatile ("ptestr #1,%1@,#0\n\t" - "pmove %%psr,%0@" - : /* no outputs */ - : "a" (&temp), "a" (addr)); + "pmove %%psr,%0" + : "=m" (temp) + : "a" (addr)); mmusr = temp; printk ("level 0 mmusr is %#x\n", mmusr); #if 0 - asm volatile ("pmove %%tt0,%0@" - : /* no outputs */ - : "a" (&tlong)); + asm volatile ("pmove %%tt0,%0" + : "=m" (tlong)); printk("tt0 is %#lx, ", tlong); - asm volatile ("pmove %%tt1,%0@" - : /* no outputs */ - : "a" (&tlong)); + asm volatile ("pmove %%tt1,%0" + : "=m" (tlong)); printk("tt1 is %#lx\n", tlong); #endif #ifdef DEBUG @@ -668,13 +666,13 @@ static inline void bus_error030 (struct frame *fp) #ifdef DEBUG asm volatile ("ptestr #1,%2@,#7,%0\n\t" - "pmove %%psr,%1@" - : "=a&" (desc) - : "a" (&temp), "a" (addr)); + "pmove %%psr,%1" + : "=a&" (desc), "=m" (temp) + : "a" (addr)); #else asm volatile ("ptestr #1,%1@,#7\n\t" - "pmove %%psr,%0@" - : : "a" (&temp), "a" (addr)); + "pmove %%psr,%0" + : "=m" (temp) : "a" (addr)); #endif mmusr = temp; diff --git a/arch/m68k/mm/cache.c b/arch/m68k/mm/cache.c index 5437fff5fe07..5550aa4fd811 100644 --- a/arch/m68k/mm/cache.c +++ b/arch/m68k/mm/cache.c @@ -52,9 +52,9 @@ static unsigned long virt_to_phys_slow(unsigned long vaddr) unsigned long *descaddr; asm volatile ("ptestr %3,%2@,#7,%0\n\t" - "pmove %%psr,%1@" - : "=a&" (descaddr) - : "a" (&mmusr), "a" (vaddr), "d" (get_fs().seg)); + "pmove %%psr,%1" + : "=a&" (descaddr), "=m" (mmusr) + : "a" (vaddr), "d" (get_fs().seg)); if (mmusr & (MMU_I|MMU_B|MMU_L)) return 0; descaddr = phys_to_virt((unsigned long)descaddr); From a14304edcd5e8323205db34b08f709feb5357e64 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Sat, 21 Jan 2012 21:48:53 +0000 Subject: [PATCH 691/820] ASoC: wm8996: Call _POST_PMU callback for CPVDD We should be allowing a 5ms delay after the charge pump is started in order to ensure it has finished ramping. Signed-off-by: Mark Brown Cc: stable@kernel.org --- sound/soc/codecs/wm8996.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/sound/soc/codecs/wm8996.c b/sound/soc/codecs/wm8996.c index 86f5b6bd7af2..13aa2bdaa7d7 100644 --- a/sound/soc/codecs/wm8996.c +++ b/sound/soc/codecs/wm8996.c @@ -1120,7 +1120,8 @@ SND_SOC_DAPM_SUPPLY_S("SYSCLK", 1, WM8996_AIF_CLOCKING_1, 0, 0, NULL, 0), SND_SOC_DAPM_SUPPLY_S("SYSDSPCLK", 2, WM8996_CLOCKING_1, 1, 0, NULL, 0), SND_SOC_DAPM_SUPPLY_S("AIFCLK", 2, WM8996_CLOCKING_1, 2, 0, NULL, 0), SND_SOC_DAPM_SUPPLY_S("Charge Pump", 2, WM8996_CHARGE_PUMP_1, 15, 0, cp_event, - SND_SOC_DAPM_PRE_PMU | SND_SOC_DAPM_POST_PMD), + SND_SOC_DAPM_PRE_PMU | SND_SOC_DAPM_POST_PMU | + SND_SOC_DAPM_POST_PMD), SND_SOC_DAPM_SUPPLY("Bandgap", SND_SOC_NOPM, 0, 0, bg_event, SND_SOC_DAPM_PRE_PMU | SND_SOC_DAPM_POST_PMD), SND_SOC_DAPM_SUPPLY("LDO2", WM8996_POWER_MANAGEMENT_2, 1, 0, NULL, 0), From b477ba628a283ba93e631d66907f91df80e82267 Mon Sep 17 00:00:00 2001 From: Eugenia Emantayev Date: Thu, 19 Jan 2012 09:42:37 +0000 Subject: [PATCH 692/820] mlx4_en: clear all eth statistics when port goes up Bug fix: Not all stats fields were cleared. Signed-off-by: Eugenia Emantayev Reviewed-by: Yevgeny Petriln Signed-off-by: David S. Miller --- .../net/ethernet/mellanox/mlx4/en_netdev.c | 44 ++++++++++++------- 1 file changed, 28 insertions(+), 16 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index 72fa807b69ce..be3f4156aaab 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -807,12 +807,37 @@ static void mlx4_en_restart(struct work_struct *work) mutex_unlock(&mdev->state_lock); } +static void mlx4_en_clear_stats(struct net_device *dev) +{ + struct mlx4_en_priv *priv = netdev_priv(dev); + struct mlx4_en_dev *mdev = priv->mdev; + int i; + + if (mlx4_en_DUMP_ETH_STATS(mdev, priv->port, 1)) + en_dbg(HW, priv, "Failed dumping statistics\n"); + + memset(&priv->stats, 0, sizeof(priv->stats)); + memset(&priv->pstats, 0, sizeof(priv->pstats)); + memset(&priv->pkstats, 0, sizeof(priv->pkstats)); + memset(&priv->port_stats, 0, sizeof(priv->port_stats)); + + for (i = 0; i < priv->tx_ring_num; i++) { + priv->tx_ring[i].bytes = 0; + priv->tx_ring[i].packets = 0; + priv->tx_ring[i].tx_csum = 0; + } + for (i = 0; i < priv->rx_ring_num; i++) { + priv->rx_ring[i].bytes = 0; + priv->rx_ring[i].packets = 0; + priv->rx_ring[i].csum_ok = 0; + priv->rx_ring[i].csum_none = 0; + } +} static int mlx4_en_open(struct net_device *dev) { struct mlx4_en_priv *priv = netdev_priv(dev); struct mlx4_en_dev *mdev = priv->mdev; - int i; int err = 0; mutex_lock(&mdev->state_lock); @@ -823,21 +848,8 @@ static int mlx4_en_open(struct net_device *dev) goto out; } - /* Reset HW statistics and performance counters */ - if (mlx4_en_DUMP_ETH_STATS(mdev, priv->port, 1)) - en_dbg(HW, priv, "Failed dumping statistics\n"); - - memset(&priv->stats, 0, sizeof(priv->stats)); - memset(&priv->pstats, 0, sizeof(priv->pstats)); - - for (i = 0; i < priv->tx_ring_num; i++) { - priv->tx_ring[i].bytes = 0; - priv->tx_ring[i].packets = 0; - } - for (i = 0; i < priv->rx_ring_num; i++) { - priv->rx_ring[i].bytes = 0; - priv->rx_ring[i].packets = 0; - } + /* Reset HW statistics and SW counters */ + mlx4_en_clear_stats(dev); err = mlx4_en_start_port(dev); if (err) From 35fb9afbdeef9d5859d9a878d0372907baf119e1 Mon Sep 17 00:00:00 2001 From: Eugenia Emantayev Date: Thu, 19 Jan 2012 09:44:37 +0000 Subject: [PATCH 693/820] mlx4: VF is not allowed to perform dump stats In multifunction mode - DUMP_STATS command is not executed for VFs. Signed-off-by: Eugenia Emantayev Reviewed-by: Yevgeny Petrilin Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/port.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx4/port.c b/drivers/net/ethernet/mellanox/mlx4/port.c index 88b52e547524..1a551d69ddcb 100644 --- a/drivers/net/ethernet/mellanox/mlx4/port.c +++ b/drivers/net/ethernet/mellanox/mlx4/port.c @@ -898,6 +898,8 @@ int mlx4_DUMP_ETH_STATS_wrapper(struct mlx4_dev *dev, int slave, struct mlx4_cmd_mailbox *outbox, struct mlx4_cmd_info *cmd) { + if (slave != dev->caps.function) + return 0; return mlx4_common_dump_eth_stats(dev, slave, vhcr->in_modifier, outbox); } From 93ece0c1a7ace88f10411dbb5643d2aa2fe00ebf Mon Sep 17 00:00:00 2001 From: Eugenia Emantayev Date: Thu, 19 Jan 2012 09:45:05 +0000 Subject: [PATCH 694/820] mlx4_en: eth statistics modification In native mode display all available staticstics. In SRIOV mode on VF display only SW counters statistics, in SRIOV mode on hypervisor display SW counters and errors (got from FW) statistics. Signed-off-by: Eugenia Emantayev Reviewed-by: Yevgeny Petrilin Signed-off-by: David S. Miller --- .../net/ethernet/mellanox/mlx4/en_ethtool.c | 66 ++++++++++++++----- .../net/ethernet/mellanox/mlx4/en_netdev.c | 2 + drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 1 + drivers/net/ethernet/mellanox/mlx4/port.c | 21 ++++++ include/linux/mlx4/device.h | 1 + 5 files changed, 75 insertions(+), 16 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c index 53c66869aecd..70346fd7f9c4 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c @@ -183,10 +183,11 @@ static int mlx4_en_set_wol(struct net_device *netdev, static int mlx4_en_get_sset_count(struct net_device *dev, int sset) { struct mlx4_en_priv *priv = netdev_priv(dev); + int bit_count = hweight64(priv->stats_bitmap); switch (sset) { case ETH_SS_STATS: - return NUM_ALL_STATS + + return (priv->stats_bitmap ? bit_count : NUM_ALL_STATS) + (priv->tx_ring_num + priv->rx_ring_num) * 2; case ETH_SS_TEST: return MLX4_EN_NUM_SELF_TEST - !(priv->mdev->dev->caps.flags @@ -201,14 +202,34 @@ static void mlx4_en_get_ethtool_stats(struct net_device *dev, { struct mlx4_en_priv *priv = netdev_priv(dev); int index = 0; - int i; + int i, j = 0; spin_lock_bh(&priv->stats_lock); - for (i = 0; i < NUM_MAIN_STATS; i++) - data[index++] = ((unsigned long *) &priv->stats)[i]; - for (i = 0; i < NUM_PORT_STATS; i++) - data[index++] = ((unsigned long *) &priv->port_stats)[i]; + if (!(priv->stats_bitmap)) { + for (i = 0; i < NUM_MAIN_STATS; i++) + data[index++] = + ((unsigned long *) &priv->stats)[i]; + for (i = 0; i < NUM_PORT_STATS; i++) + data[index++] = + ((unsigned long *) &priv->port_stats)[i]; + for (i = 0; i < NUM_PKT_STATS; i++) + data[index++] = + ((unsigned long *) &priv->pkstats)[i]; + } else { + for (i = 0; i < NUM_MAIN_STATS; i++) { + if ((priv->stats_bitmap >> j) & 1) + data[index++] = + ((unsigned long *) &priv->stats)[i]; + j++; + } + for (i = 0; i < NUM_PORT_STATS; i++) { + if ((priv->stats_bitmap >> j) & 1) + data[index++] = + ((unsigned long *) &priv->port_stats)[i]; + j++; + } + } for (i = 0; i < priv->tx_ring_num; i++) { data[index++] = priv->tx_ring[i].packets; data[index++] = priv->tx_ring[i].bytes; @@ -217,8 +238,6 @@ static void mlx4_en_get_ethtool_stats(struct net_device *dev, data[index++] = priv->rx_ring[i].packets; data[index++] = priv->rx_ring[i].bytes; } - for (i = 0; i < NUM_PKT_STATS; i++) - data[index++] = ((unsigned long *) &priv->pkstats)[i]; spin_unlock_bh(&priv->stats_lock); } @@ -247,11 +266,29 @@ static void mlx4_en_get_strings(struct net_device *dev, case ETH_SS_STATS: /* Add main counters */ - for (i = 0; i < NUM_MAIN_STATS; i++) - strcpy(data + (index++) * ETH_GSTRING_LEN, main_strings[i]); - for (i = 0; i< NUM_PORT_STATS; i++) - strcpy(data + (index++) * ETH_GSTRING_LEN, - main_strings[i + NUM_MAIN_STATS]); + if (!priv->stats_bitmap) { + for (i = 0; i < NUM_MAIN_STATS; i++) + strcpy(data + (index++) * ETH_GSTRING_LEN, + main_strings[i]); + for (i = 0; i < NUM_PORT_STATS; i++) + strcpy(data + (index++) * ETH_GSTRING_LEN, + main_strings[i + + NUM_MAIN_STATS]); + for (i = 0; i < NUM_PKT_STATS; i++) + strcpy(data + (index++) * ETH_GSTRING_LEN, + main_strings[i + + NUM_MAIN_STATS + + NUM_PORT_STATS]); + } else + for (i = 0; i < NUM_MAIN_STATS + NUM_PORT_STATS; i++) { + if ((priv->stats_bitmap >> i) & 1) { + strcpy(data + + (index++) * ETH_GSTRING_LEN, + main_strings[i]); + } + if (!(priv->stats_bitmap >> i)) + break; + } for (i = 0; i < priv->tx_ring_num; i++) { sprintf(data + (index++) * ETH_GSTRING_LEN, "tx%d_packets", i); @@ -264,9 +301,6 @@ static void mlx4_en_get_strings(struct net_device *dev, sprintf(data + (index++) * ETH_GSTRING_LEN, "rx%d_bytes", i); } - for (i = 0; i< NUM_PKT_STATS; i++) - strcpy(data + (index++) * ETH_GSTRING_LEN, - main_strings[i + NUM_MAIN_STATS + NUM_PORT_STATS]); break; } } diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index be3f4156aaab..467ae5824875 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -702,6 +702,8 @@ int mlx4_en_start_port(struct net_device *dev) /* Schedule multicast task to populate multicast list */ queue_work(mdev->workqueue, &priv->mcast_task); + mlx4_set_stats_bitmap(mdev->dev, &priv->stats_bitmap); + priv->port_up = true; netif_tx_start_all_queues(dev); return 0; diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index f4d189a1290e..35f08840813c 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -476,6 +476,7 @@ struct mlx4_en_priv { struct mlx4_en_perf_stats pstats; struct mlx4_en_pkt_stats pkstats; struct mlx4_en_port_stats port_stats; + u64 stats_bitmap; char *mc_addrs; int mc_addrs_cnt; struct mlx4_en_stat_out_mbox hw_stats; diff --git a/drivers/net/ethernet/mellanox/mlx4/port.c b/drivers/net/ethernet/mellanox/mlx4/port.c index 1a551d69ddcb..f44ae555bf43 100644 --- a/drivers/net/ethernet/mellanox/mlx4/port.c +++ b/drivers/net/ethernet/mellanox/mlx4/port.c @@ -44,6 +44,11 @@ #define MLX4_VLAN_VALID (1u << 31) #define MLX4_VLAN_MASK 0xfff +#define MLX4_STATS_TRAFFIC_COUNTERS_MASK 0xfULL +#define MLX4_STATS_TRAFFIC_DROPS_MASK 0xc0ULL +#define MLX4_STATS_ERROR_COUNTERS_MASK 0x1ffc30ULL +#define MLX4_STATS_PORT_COUNTERS_MASK 0x1fe00000ULL + void mlx4_init_mac_table(struct mlx4_dev *dev, struct mlx4_mac_table *table) { int i; @@ -903,3 +908,19 @@ int mlx4_DUMP_ETH_STATS_wrapper(struct mlx4_dev *dev, int slave, return mlx4_common_dump_eth_stats(dev, slave, vhcr->in_modifier, outbox); } + +void mlx4_set_stats_bitmap(struct mlx4_dev *dev, u64 *stats_bitmap) +{ + if (!mlx4_is_mfunc(dev)) { + *stats_bitmap = 0; + return; + } + + *stats_bitmap = (MLX4_STATS_TRAFFIC_COUNTERS_MASK | + MLX4_STATS_TRAFFIC_DROPS_MASK | + MLX4_STATS_PORT_COUNTERS_MASK); + + if (mlx4_is_master(dev)) + *stats_bitmap |= MLX4_STATS_ERROR_COUNTERS_MASK; +} +EXPORT_SYMBOL(mlx4_set_stats_bitmap); diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h index 5c4fe8e5bfe5..aea61905499b 100644 --- a/include/linux/mlx4/device.h +++ b/include/linux/mlx4/device.h @@ -621,6 +621,7 @@ void mlx4_unregister_mac(struct mlx4_dev *dev, u8 port, u64 mac); int mlx4_replace_mac(struct mlx4_dev *dev, u8 port, int qpn, u64 new_mac); int mlx4_get_eth_qp(struct mlx4_dev *dev, u8 port, u64 mac, int *qpn); void mlx4_put_eth_qp(struct mlx4_dev *dev, u8 port, u64 mac, int qpn); +void mlx4_set_stats_bitmap(struct mlx4_dev *dev, u64 *stats_bitmap); int mlx4_find_cached_vlan(struct mlx4_dev *dev, u8 port, u16 vid, int *idx); int mlx4_register_vlan(struct mlx4_dev *dev, u8 port, u16 vlan, int *index); From eb41049f2f5eee1525fe1c699e6b1f03c8c51532 Mon Sep 17 00:00:00 2001 From: Marcel Apfelbaum Date: Thu, 19 Jan 2012 09:45:19 +0000 Subject: [PATCH 695/820] mlx4_core: removed function index from vf. The Virtual Functions should not be aware their function number. Signed-off-by: Marcel Apfelbaum Reviewed-by: Jack Morgenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/cq.c | 4 +-- drivers/net/ethernet/mellanox/mlx4/eq.c | 4 +-- drivers/net/ethernet/mellanox/mlx4/fw.c | 7 ---- drivers/net/ethernet/mellanox/mlx4/fw.h | 1 - drivers/net/ethernet/mellanox/mlx4/main.c | 1 - drivers/net/ethernet/mellanox/mlx4/mr.c | 2 +- drivers/net/ethernet/mellanox/mlx4/pd.c | 3 +- drivers/net/ethernet/mellanox/mlx4/qp.c | 2 +- .../ethernet/mellanox/mlx4/resource_tracker.c | 35 ------------------- drivers/net/ethernet/mellanox/mlx4/srq.c | 2 +- 10 files changed, 8 insertions(+), 53 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/cq.c b/drivers/net/ethernet/mellanox/mlx4/cq.c index 475f9d6af955..7e64033d7de3 100644 --- a/drivers/net/ethernet/mellanox/mlx4/cq.c +++ b/drivers/net/ethernet/mellanox/mlx4/cq.c @@ -96,7 +96,7 @@ void mlx4_cq_event(struct mlx4_dev *dev, u32 cqn, int event_type) static int mlx4_SW2HW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int cq_num) { - return mlx4_cmd(dev, mailbox->dma | dev->caps.function, cq_num, 0, + return mlx4_cmd(dev, mailbox->dma, cq_num, 0, MLX4_CMD_SW2HW_CQ, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_WRAPPED); } @@ -111,7 +111,7 @@ static int mlx4_MODIFY_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox static int mlx4_HW2SW_CQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int cq_num) { - return mlx4_cmd_box(dev, dev->caps.function, mailbox ? mailbox->dma : 0, + return mlx4_cmd_box(dev, 0, mailbox ? mailbox->dma : 0, cq_num, mailbox ? 0 : 1, MLX4_CMD_HW2SW_CQ, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_WRAPPED); } diff --git a/drivers/net/ethernet/mellanox/mlx4/eq.c b/drivers/net/ethernet/mellanox/mlx4/eq.c index 1e9b55eb7217..e163de62604b 100644 --- a/drivers/net/ethernet/mellanox/mlx4/eq.c +++ b/drivers/net/ethernet/mellanox/mlx4/eq.c @@ -546,7 +546,7 @@ static int mlx4_MAP_EQ(struct mlx4_dev *dev, u64 event_mask, int unmap, static int mlx4_SW2HW_EQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int eq_num) { - return mlx4_cmd(dev, mailbox->dma | dev->caps.function, eq_num, 0, + return mlx4_cmd(dev, mailbox->dma, eq_num, 0, MLX4_CMD_SW2HW_EQ, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_WRAPPED); } @@ -554,7 +554,7 @@ static int mlx4_SW2HW_EQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, static int mlx4_HW2SW_EQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int eq_num) { - return mlx4_cmd_box(dev, dev->caps.function, mailbox->dma, eq_num, + return mlx4_cmd_box(dev, 0, mailbox->dma, eq_num, 0, MLX4_CMD_HW2SW_EQ, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_WRAPPED); } diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c index a424a19280cc..8a21e10952ea 100644 --- a/drivers/net/ethernet/mellanox/mlx4/fw.c +++ b/drivers/net/ethernet/mellanox/mlx4/fw.c @@ -158,7 +158,6 @@ int mlx4_QUERY_FUNC_CAP_wrapper(struct mlx4_dev *dev, int slave, #define QUERY_FUNC_CAP_FLAGS_OFFSET 0x0 #define QUERY_FUNC_CAP_NUM_PORTS_OFFSET 0x1 -#define QUERY_FUNC_CAP_FUNCTION_OFFSET 0x3 #define QUERY_FUNC_CAP_PF_BHVR_OFFSET 0x4 #define QUERY_FUNC_CAP_QP_QUOTA_OFFSET 0x10 #define QUERY_FUNC_CAP_CQ_QUOTA_OFFSET 0x14 @@ -182,9 +181,6 @@ int mlx4_QUERY_FUNC_CAP_wrapper(struct mlx4_dev *dev, int slave, field = 1 << 7; /* enable only ethernet interface */ MLX4_PUT(outbox->buf, field, QUERY_FUNC_CAP_FLAGS_OFFSET); - field = slave; - MLX4_PUT(outbox->buf, field, QUERY_FUNC_CAP_FUNCTION_OFFSET); - field = dev->caps.num_ports; MLX4_PUT(outbox->buf, field, QUERY_FUNC_CAP_NUM_PORTS_OFFSET); @@ -249,9 +245,6 @@ int mlx4_QUERY_FUNC_CAP(struct mlx4_dev *dev, struct mlx4_func_cap *func_cap) goto out; } - MLX4_GET(field, outbox, QUERY_FUNC_CAP_FUNCTION_OFFSET); - func_cap->function = field; - MLX4_GET(field, outbox, QUERY_FUNC_CAP_NUM_PORTS_OFFSET); func_cap->num_ports = field; diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.h b/drivers/net/ethernet/mellanox/mlx4/fw.h index 119e0cc9fab3..e1a5fa56bcbc 100644 --- a/drivers/net/ethernet/mellanox/mlx4/fw.h +++ b/drivers/net/ethernet/mellanox/mlx4/fw.h @@ -119,7 +119,6 @@ struct mlx4_dev_cap { }; struct mlx4_func_cap { - u8 function; u8 num_ports; u8 flags; u32 pf_context_behaviour; diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c index 6bb62c580e2d..91c8bc953f3f 100644 --- a/drivers/net/ethernet/mellanox/mlx4/main.c +++ b/drivers/net/ethernet/mellanox/mlx4/main.c @@ -471,7 +471,6 @@ static int mlx4_slave_cap(struct mlx4_dev *dev) return -ENOSYS; } - dev->caps.function = func_cap.function; dev->caps.num_ports = func_cap.num_ports; dev->caps.num_qps = func_cap.qp_quota; dev->caps.num_srqs = func_cap.srq_quota; diff --git a/drivers/net/ethernet/mellanox/mlx4/mr.c b/drivers/net/ethernet/mellanox/mlx4/mr.c index 01df5567e16e..8deeef98280c 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mr.c +++ b/drivers/net/ethernet/mellanox/mlx4/mr.c @@ -291,7 +291,7 @@ static u32 key_to_hw_index(u32 key) static int mlx4_SW2HW_MPT(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int mpt_index) { - return mlx4_cmd(dev, mailbox->dma | dev->caps.function , mpt_index, + return mlx4_cmd(dev, mailbox->dma, mpt_index, 0, MLX4_CMD_SW2HW_MPT, MLX4_CMD_TIME_CLASS_B, MLX4_CMD_WRAPPED); } diff --git a/drivers/net/ethernet/mellanox/mlx4/pd.c b/drivers/net/ethernet/mellanox/mlx4/pd.c index 5c9a54df17ab..db4746d0dca7 100644 --- a/drivers/net/ethernet/mellanox/mlx4/pd.c +++ b/drivers/net/ethernet/mellanox/mlx4/pd.c @@ -52,8 +52,7 @@ int mlx4_pd_alloc(struct mlx4_dev *dev, u32 *pdn) *pdn = mlx4_bitmap_alloc(&priv->pd_bitmap); if (*pdn == -1) return -ENOMEM; - if (mlx4_is_mfunc(dev)) - *pdn |= (dev->caps.function + 1) << NOT_MASKED_PD_BITS; + return 0; } EXPORT_SYMBOL_GPL(mlx4_pd_alloc); diff --git a/drivers/net/ethernet/mellanox/mlx4/qp.c b/drivers/net/ethernet/mellanox/mlx4/qp.c index 6b03ac8b9002..738f950a1ce5 100644 --- a/drivers/net/ethernet/mellanox/mlx4/qp.c +++ b/drivers/net/ethernet/mellanox/mlx4/qp.c @@ -162,7 +162,7 @@ static int __mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt, ((struct mlx4_qp_context *) (mailbox->buf + 8))->local_qpn = cpu_to_be32(qp->qpn); - ret = mlx4_cmd(dev, mailbox->dma | dev->caps.function, + ret = mlx4_cmd(dev, mailbox->dma, qp->qpn | (!!sqd_event << 31), new_state == MLX4_QP_STATE_RST ? 2 : 0, op[cur_state][new_state], MLX4_CMD_TIME_CLASS_C, native); diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c index ed20751a057d..a30cf197a358 100644 --- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c +++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c @@ -1561,11 +1561,6 @@ static int mr_get_mtt_size(struct mlx4_mpt_entry *mpt) return be32_to_cpu(mpt->mtt_sz); } -static int mr_get_pdn(struct mlx4_mpt_entry *mpt) -{ - return be32_to_cpu(mpt->pd_flags) & 0xffffff; -} - static int qp_get_mtt_addr(struct mlx4_qp_context *qpc) { return be32_to_cpu(qpc->mtt_base_addr_l) & 0xfffffff8; @@ -1602,16 +1597,6 @@ static int qp_get_mtt_size(struct mlx4_qp_context *qpc) return total_pages; } -static int qp_get_pdn(struct mlx4_qp_context *qpc) -{ - return be32_to_cpu(qpc->pd) & 0xffffff; -} - -static int pdn2slave(int pdn) -{ - return (pdn >> NOT_MASKED_PD_BITS) - 1; -} - static int check_mtt_range(struct mlx4_dev *dev, int slave, int start, int size, struct res_mtt *mtt) { @@ -1656,11 +1641,6 @@ int mlx4_SW2HW_MPT_wrapper(struct mlx4_dev *dev, int slave, mpt->mtt = mtt; } - if (pdn2slave(mr_get_pdn(inbox->buf)) != slave) { - err = -EPERM; - goto ex_put; - } - err = mlx4_DMA_wrapper(dev, slave, vhcr, inbox, outbox, cmd); if (err) goto ex_put; @@ -1792,11 +1772,6 @@ int mlx4_RST2INIT_QP_wrapper(struct mlx4_dev *dev, int slave, if (err) goto ex_put_mtt; - if (pdn2slave(qp_get_pdn(qpc)) != slave) { - err = -EPERM; - goto ex_put_mtt; - } - err = get_res(dev, slave, rcqn, RES_CQ, &rcq); if (err) goto ex_put_mtt; @@ -2289,11 +2264,6 @@ ex_put: return err; } -static int srq_get_pdn(struct mlx4_srq_context *srqc) -{ - return be32_to_cpu(srqc->pd) & 0xffffff; -} - static int srq_get_mtt_size(struct mlx4_srq_context *srqc) { int log_srq_size = (be32_to_cpu(srqc->state_logsize_srqn) >> 24) & 0xf; @@ -2333,11 +2303,6 @@ int mlx4_SW2HW_SRQ_wrapper(struct mlx4_dev *dev, int slave, if (err) goto ex_put_mtt; - if (pdn2slave(srq_get_pdn(srqc)) != slave) { - err = -EPERM; - goto ex_put_mtt; - } - err = mlx4_DMA_wrapper(dev, slave, vhcr, inbox, outbox, cmd); if (err) goto ex_put_mtt; diff --git a/drivers/net/ethernet/mellanox/mlx4/srq.c b/drivers/net/ethernet/mellanox/mlx4/srq.c index 2823fffc6383..feda6c00829f 100644 --- a/drivers/net/ethernet/mellanox/mlx4/srq.c +++ b/drivers/net/ethernet/mellanox/mlx4/srq.c @@ -67,7 +67,7 @@ void mlx4_srq_event(struct mlx4_dev *dev, u32 srqn, int event_type) static int mlx4_SW2HW_SRQ(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox, int srq_num) { - return mlx4_cmd(dev, mailbox->dma | dev->caps.function, srq_num, 0, + return mlx4_cmd(dev, mailbox->dma, srq_num, 0, MLX4_CMD_SW2HW_SRQ, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_WRAPPED); } From 9fd7a1e147bd41fb999a58da8a8826b9f3415384 Mon Sep 17 00:00:00 2001 From: Marcel Apfelbaum Date: Thu, 19 Jan 2012 09:45:31 +0000 Subject: [PATCH 696/820] mlx4_core: Fix mtt profile issue Num mtts from profile is really the number of mtt segments. Thus, in make profile, to get the proper number of MTT entries, must multiply num_mtts by mtts per segment. Signed-off-by: Marcel Apfelbaum Reviewed-by: Jack Morgenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/main.c | 2 +- drivers/net/ethernet/mellanox/mlx4/profile.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c index 91c8bc953f3f..678558b502fc 100644 --- a/drivers/net/ethernet/mellanox/mlx4/main.c +++ b/drivers/net/ethernet/mellanox/mlx4/main.c @@ -108,7 +108,7 @@ static struct mlx4_profile default_profile = { .num_cq = 1 << 16, .num_mcg = 1 << 13, .num_mpt = 1 << 19, - .num_mtt = 1 << 20, + .num_mtt = 1 << 20, /* It is really num mtt segements */ }; static int log_num_mac = 7; diff --git a/drivers/net/ethernet/mellanox/mlx4/profile.c b/drivers/net/ethernet/mellanox/mlx4/profile.c index 66f91ca7a7c6..1129677daa62 100644 --- a/drivers/net/ethernet/mellanox/mlx4/profile.c +++ b/drivers/net/ethernet/mellanox/mlx4/profile.c @@ -110,7 +110,7 @@ u64 mlx4_make_profile(struct mlx4_dev *dev, profile[MLX4_RES_EQ].num = min_t(unsigned, dev_cap->max_eqs, MAX_MSIX); profile[MLX4_RES_DMPT].num = request->num_mpt; profile[MLX4_RES_CMPT].num = MLX4_NUM_CMPTS; - profile[MLX4_RES_MTT].num = request->num_mtt; + profile[MLX4_RES_MTT].num = request->num_mtt * (1 << log_mtts_per_seg); profile[MLX4_RES_MCG].num = request->num_mcg; for (i = 0; i < MLX4_RES_NUM; ++i) { From 803143fbda719106017309d551d082e4b4e7c8c3 Mon Sep 17 00:00:00 2001 From: Marcel Apfelbaum Date: Thu, 19 Jan 2012 09:45:46 +0000 Subject: [PATCH 697/820] mlx4_core: map async events to arbitrary slave eqs Slave async events were mapped to single eq. This patch fixes this issue, so the slaves can map the async events to any eq. Signed-off-by: Marcel Apfelbaum Reviewed-by: Jack Morgenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/cmd.c | 9 ++++++++- drivers/net/ethernet/mellanox/mlx4/eq.c | 17 +++++++---------- drivers/net/ethernet/mellanox/mlx4/mlx4.h | 8 +++++--- .../ethernet/mellanox/mlx4/resource_tracker.c | 4 ++-- 4 files changed, 22 insertions(+), 16 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/cmd.c b/drivers/net/ethernet/mellanox/mlx4/cmd.c index 978f593094c0..405e6ac3faf6 100644 --- a/drivers/net/ethernet/mellanox/mlx4/cmd.c +++ b/drivers/net/ethernet/mellanox/mlx4/cmd.c @@ -1247,6 +1247,7 @@ static void mlx4_master_do_cmd(struct mlx4_dev *dev, int slave, u8 cmd, u32 reply; u32 slave_status = 0; u8 is_going_down = 0; + int i; slave_state[slave].comm_toggle ^= 1; reply = (u32) slave_state[slave].comm_toggle << 31; @@ -1258,6 +1259,10 @@ static void mlx4_master_do_cmd(struct mlx4_dev *dev, int slave, u8 cmd, if (cmd == MLX4_COMM_CMD_RESET) { mlx4_warn(dev, "Received reset from slave:%d\n", slave); slave_state[slave].active = false; + for (i = 0; i < MLX4_EVENT_TYPES_NUM; ++i) { + slave_state[slave].event_eq[i].eqn = -1; + slave_state[slave].event_eq[i].token = 0; + } /*check if we are in the middle of FLR process, if so return "retry" status to the slave*/ if (MLX4_COMM_CMD_FLR == slave_state[slave].last_cmd) { @@ -1452,7 +1457,7 @@ int mlx4_multi_func_init(struct mlx4_dev *dev) { struct mlx4_priv *priv = mlx4_priv(dev); struct mlx4_slave_state *s_state; - int i, err, port; + int i, j, err, port; priv->mfunc.vhcr = dma_alloc_coherent(&(dev->pdev->dev), PAGE_SIZE, &priv->mfunc.vhcr_dma, @@ -1485,6 +1490,8 @@ int mlx4_multi_func_init(struct mlx4_dev *dev) for (i = 0; i < dev->num_slaves; ++i) { s_state = &priv->mfunc.master.slave_state[i]; s_state->last_cmd = MLX4_COMM_CMD_RESET; + for (j = 0; j < MLX4_EVENT_TYPES_NUM; ++j) + s_state->event_eq[j].eqn = -1; __raw_writel((__force u32) 0, &priv->mfunc.comm[i].slave_write); __raw_writel((__force u32) 0, diff --git a/drivers/net/ethernet/mellanox/mlx4/eq.c b/drivers/net/ethernet/mellanox/mlx4/eq.c index e163de62604b..55d7bd4e210a 100644 --- a/drivers/net/ethernet/mellanox/mlx4/eq.c +++ b/drivers/net/ethernet/mellanox/mlx4/eq.c @@ -513,25 +513,22 @@ int mlx4_MAP_EQ_wrapper(struct mlx4_dev *dev, int slave, { struct mlx4_priv *priv = mlx4_priv(dev); struct mlx4_slave_event_eq_info *event_eq = - &priv->mfunc.master.slave_state[slave].event_eq; + priv->mfunc.master.slave_state[slave].event_eq; u32 in_modifier = vhcr->in_modifier; u32 eqn = in_modifier & 0x1FF; u64 in_param = vhcr->in_param; int err = 0; + int i; if (slave == dev->caps.function) err = mlx4_cmd(dev, in_param, (in_modifier & 0x80000000) | eqn, 0, MLX4_CMD_MAP_EQ, MLX4_CMD_TIME_CLASS_B, MLX4_CMD_NATIVE); - if (!err) { - if (in_modifier >> 31) { - /* unmap */ - event_eq->event_type &= ~in_param; - } else { - event_eq->eqn = eqn; - event_eq->event_type = in_param; - } - } + if (!err) + for (i = 0; i < MLX4_EVENT_TYPES_NUM; ++i) + if (in_param & (1LL << i)) + event_eq[i].eqn = in_modifier >> 31 ? -1 : eqn; + return err; } diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4.h b/drivers/net/ethernet/mellanox/mlx4/mlx4.h index a80121a2b519..c92269f8c057 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4.h @@ -388,9 +388,8 @@ struct mlx4_slave_eqe { }; struct mlx4_slave_event_eq_info { - u32 eqn; + int eqn; u16 token; - u64 event_type; }; struct mlx4_profile { @@ -449,6 +448,8 @@ struct mlx4_steer_index { struct list_head duplicates; }; +#define MLX4_EVENT_TYPES_NUM 64 + struct mlx4_slave_state { u8 comm_toggle; u8 last_cmd; @@ -461,7 +462,8 @@ struct mlx4_slave_state { struct mlx4_slave_eqe eq[MLX4_MFUNC_MAX_EQES]; struct list_head mcast_filters[MLX4_MAX_PORTS + 1]; struct mlx4_vlan_fltr *vlan_filter[MLX4_MAX_PORTS + 1]; - struct mlx4_slave_event_eq_info event_eq; + /* event type to eq number lookup */ + struct mlx4_slave_event_eq_info event_eq[MLX4_EVENT_TYPES_NUM]; u16 eq_pi; u16 eq_ci; spinlock_t lock; diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c index a30cf197a358..dcd819bfb2f0 100644 --- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c +++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c @@ -2023,10 +2023,10 @@ int mlx4_GEN_EQE(struct mlx4_dev *dev, int slave, struct mlx4_eqe *eqe) if (!priv->mfunc.master.slave_state) return -EINVAL; - event_eq = &priv->mfunc.master.slave_state[slave].event_eq; + event_eq = &priv->mfunc.master.slave_state[slave].event_eq[eqe->type]; /* Create the event only if the slave is registered */ - if ((event_eq->event_type & (1 << eqe->type)) == 0) + if (event_eq->eqn < 0) return 0; mutex_lock(&priv->mfunc.master.gen_eqe_mutex[slave]); From a42b4799c683723e8c464de4026af085b2ebd5fa Mon Sep 17 00:00:00 2001 From: Vijay Subramanian Date: Thu, 19 Jan 2012 10:20:59 +0000 Subject: [PATCH 698/820] netem: Fix off-by-one bug in reordering With netem reordering, a gap of N is supposed to reorder every Nth packet with given reorder probability. However, the code currently skips N packets and reorders every (N+1)th packet. Signed-off-by: Vijay Subramanian Signed-off-by: David S. Miller --- net/sched/sch_netem.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index e7e1d0b57b3d..2776012132ea 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -419,7 +419,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch) cb = netem_skb_cb(skb); if (q->gap == 0 || /* not doing reordering */ - q->counter < q->gap || /* inside last reordering gap */ + q->counter < q->gap - 1 || /* inside last reordering gap */ q->reorder < get_crandom(&q->reorder_cor)) { psched_time_t now; psched_tdiff_t delay; From 19f9ad789396fe5b4fe0302344b3acfa07cb6772 Mon Sep 17 00:00:00 2001 From: stephen hemminger Date: Thu, 19 Jan 2012 14:35:25 +0000 Subject: [PATCH 699/820] skge: don't assert carrier until link is up Skge device would assert carrier (link up) as soon as network device open was called, rather than waiting until PHY has detected link. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/skge.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c index 18a87a57fc0a..f580f0535bb4 100644 --- a/drivers/net/ethernet/marvell/skge.c +++ b/drivers/net/ethernet/marvell/skge.c @@ -2576,6 +2576,7 @@ static int skge_up(struct net_device *dev) } /* Initialize MAC */ + netif_carrier_off(dev); spin_lock_bh(&hw->phy_lock); if (is_genesis(hw)) genesis_mac_init(hw, port); From d0249e44432aa0ffcf710b64449b8eaa3722547e Mon Sep 17 00:00:00 2001 From: stephen hemminger Date: Thu, 19 Jan 2012 14:37:18 +0000 Subject: [PATCH 700/820] skge: check for PCI dma mapping errors Driver should check for mapping errors. Machines with limited DMA maps may return an error when a PCI map is requested (not an issue on standard x86). Also use upper/lower 32 bits macros for clarity. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/skge.c | 71 +++++++++++++++++++++++------ 1 file changed, 58 insertions(+), 13 deletions(-) diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c index f580f0535bb4..299c33bd5345 100644 --- a/drivers/net/ethernet/marvell/skge.c +++ b/drivers/net/ethernet/marvell/skge.c @@ -931,17 +931,20 @@ static int skge_ring_alloc(struct skge_ring *ring, void *vaddr, u32 base) } /* Allocate and setup a new buffer for receiving */ -static void skge_rx_setup(struct skge_port *skge, struct skge_element *e, - struct sk_buff *skb, unsigned int bufsize) +static int skge_rx_setup(struct pci_dev *pdev, + struct skge_element *e, + struct sk_buff *skb, unsigned int bufsize) { struct skge_rx_desc *rd = e->desc; - u64 map; + dma_addr_t map; - map = pci_map_single(skge->hw->pdev, skb->data, bufsize, + map = pci_map_single(pdev, skb->data, bufsize, PCI_DMA_FROMDEVICE); + if (pci_dma_mapping_error(pdev, map)) + goto mapping_error; - rd->dma_lo = map; - rd->dma_hi = map >> 32; + rd->dma_lo = lower_32_bits(map); + rd->dma_hi = upper_32_bits(map); e->skb = skb; rd->csum1_start = ETH_HLEN; rd->csum2_start = ETH_HLEN; @@ -953,6 +956,13 @@ static void skge_rx_setup(struct skge_port *skge, struct skge_element *e, rd->control = BMU_OWN | BMU_STF | BMU_IRQ_EOF | BMU_TCP_CHECK | bufsize; dma_unmap_addr_set(e, mapaddr, map); dma_unmap_len_set(e, maplen, bufsize); + return 0; + +mapping_error: + if (net_ratelimit()) + dev_warn(&pdev->dev, "%s: rx mapping error\n", + skb->dev->name); + return -EIO; } /* Resume receiving using existing skb, @@ -1014,7 +1024,11 @@ static int skge_rx_fill(struct net_device *dev) return -ENOMEM; skb_reserve(skb, NET_IP_ALIGN); - skge_rx_setup(skge, e, skb, skge->rx_buf_size); + if (skge_rx_setup(skge->hw->pdev, e, skb, skge->rx_buf_size)) { + kfree_skb(skb); + return -ENOMEM; + } + } while ((e = e->next) != ring->start); ring->to_clean = ring->start; @@ -2729,7 +2743,7 @@ static netdev_tx_t skge_xmit_frame(struct sk_buff *skb, struct skge_tx_desc *td; int i; u32 control, len; - u64 map; + dma_addr_t map; if (skb_padto(skb, ETH_ZLEN)) return NETDEV_TX_OK; @@ -2743,11 +2757,14 @@ static netdev_tx_t skge_xmit_frame(struct sk_buff *skb, e->skb = skb; len = skb_headlen(skb); map = pci_map_single(hw->pdev, skb->data, len, PCI_DMA_TODEVICE); + if (pci_dma_mapping_error(hw->pdev, map)) + goto mapping_error; + dma_unmap_addr_set(e, mapaddr, map); dma_unmap_len_set(e, maplen, len); - td->dma_lo = map; - td->dma_hi = map >> 32; + td->dma_lo = lower_32_bits(map); + td->dma_hi = upper_32_bits(map); if (skb->ip_summed == CHECKSUM_PARTIAL) { const int offset = skb_checksum_start_offset(skb); @@ -2778,14 +2795,16 @@ static netdev_tx_t skge_xmit_frame(struct sk_buff *skb, map = skb_frag_dma_map(&hw->pdev->dev, frag, 0, skb_frag_size(frag), DMA_TO_DEVICE); + if (dma_mapping_error(&hw->pdev->dev, map)) + goto mapping_unwind; e = e->next; e->skb = skb; tf = e->desc; BUG_ON(tf->control & BMU_OWN); - tf->dma_lo = map; - tf->dma_hi = (u64) map >> 32; + tf->dma_lo = lower_32_bits(map); + tf->dma_hi = upper_32_bits(map); dma_unmap_addr_set(e, mapaddr, map); dma_unmap_len_set(e, maplen, skb_frag_size(frag)); @@ -2813,6 +2832,28 @@ static netdev_tx_t skge_xmit_frame(struct sk_buff *skb, } return NETDEV_TX_OK; + +mapping_unwind: + /* unroll any pages that were already mapped. */ + if (e != skge->tx_ring.to_use) { + struct skge_element *u; + + for (u = skge->tx_ring.to_use->next; u != e; u = u->next) + pci_unmap_page(hw->pdev, dma_unmap_addr(u, mapaddr), + dma_unmap_len(u, maplen), + PCI_DMA_TODEVICE); + e = skge->tx_ring.to_use; + } + /* undo the mapping for the skb header */ + pci_unmap_single(hw->pdev, dma_unmap_addr(e, mapaddr), + dma_unmap_len(e, maplen), + PCI_DMA_TODEVICE); +mapping_error: + /* mapping error causes error message and packet to be discarded. */ + if (net_ratelimit()) + dev_warn(&hw->pdev->dev, "%s: tx mapping error\n", dev->name); + dev_kfree_skb(skb); + return NETDEV_TX_OK; } @@ -3060,13 +3101,17 @@ static struct sk_buff *skge_rx_get(struct net_device *dev, if (!nskb) goto resubmit; + if (unlikely(skge_rx_setup(skge->hw->pdev, e, nskb, skge->rx_buf_size))) { + dev_kfree_skb(nskb); + goto resubmit; + } + pci_unmap_single(skge->hw->pdev, dma_unmap_addr(e, mapaddr), dma_unmap_len(e, maplen), PCI_DMA_FROMDEVICE); skb = e->skb; prefetch(skb->data); - skge_rx_setup(skge, e, nskb, skge->rx_buf_size); } skb_put(skb, len); From 974c12360dfe6ab01201fe9e708e7755c413f8b6 Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Thu, 19 Jan 2012 14:42:21 +0000 Subject: [PATCH 701/820] tcp: detect loss above high_seq in recovery Correctly implement a loss detection heuristic: New sequences (above high_seq) sent during the fast recovery are deemed lost when higher sequences are SACKed. Current code does not catch these losses, because tcp_mark_head_lost() does not check packets beyond high_seq. The fix is straight-forward by checking packets until the highest sacked packet. In addition, all the FLAG_DATA_LOST logic are in-effective and redundant and can be removed. Update the loss heuristic comments. The algorithm above is documented as heuristic B, but it is redundant too because heuristic A already covers B. Note that this change only marks some forward-retransmitted packets LOST. It does NOT forbid TCP performing further CWR on new losses. A potential follow-up patch under preparation is to perform another CWR on "new" losses such as 1) sequence above high_seq is lost (by resetting high_seq to snd_nxt) 2) retransmission is lost. Signed-off-by: Yuchung Cheng Signed-off-by: David S. Miller --- include/linux/snmp.h | 1 - net/ipv4/proc.c | 1 - net/ipv4/tcp_input.c | 41 +++++++++++++++-------------------------- 3 files changed, 15 insertions(+), 28 deletions(-) diff --git a/include/linux/snmp.h b/include/linux/snmp.h index e16557a357e5..c1241c428179 100644 --- a/include/linux/snmp.h +++ b/include/linux/snmp.h @@ -192,7 +192,6 @@ enum LINUX_MIB_TCPPARTIALUNDO, /* TCPPartialUndo */ LINUX_MIB_TCPDSACKUNDO, /* TCPDSACKUndo */ LINUX_MIB_TCPLOSSUNDO, /* TCPLossUndo */ - LINUX_MIB_TCPLOSS, /* TCPLoss */ LINUX_MIB_TCPLOSTRETRANSMIT, /* TCPLostRetransmit */ LINUX_MIB_TCPRENOFAILURES, /* TCPRenoFailures */ LINUX_MIB_TCPSACKFAILURES, /* TCPSackFailures */ diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c index 3569d8ecaeac..6afc807ee2ad 100644 --- a/net/ipv4/proc.c +++ b/net/ipv4/proc.c @@ -216,7 +216,6 @@ static const struct snmp_mib snmp4_net_list[] = { SNMP_MIB_ITEM("TCPPartialUndo", LINUX_MIB_TCPPARTIALUNDO), SNMP_MIB_ITEM("TCPDSACKUndo", LINUX_MIB_TCPDSACKUNDO), SNMP_MIB_ITEM("TCPLossUndo", LINUX_MIB_TCPLOSSUNDO), - SNMP_MIB_ITEM("TCPLoss", LINUX_MIB_TCPLOSS), SNMP_MIB_ITEM("TCPLostRetransmit", LINUX_MIB_TCPLOSTRETRANSMIT), SNMP_MIB_ITEM("TCPRenoFailures", LINUX_MIB_TCPRENOFAILURES), SNMP_MIB_ITEM("TCPSackFailures", LINUX_MIB_TCPSACKFAILURES), diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 2877c3e09587..976034f82320 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -105,7 +105,6 @@ int sysctl_tcp_abc __read_mostly; #define FLAG_SYN_ACKED 0x10 /* This ACK acknowledged SYN. */ #define FLAG_DATA_SACKED 0x20 /* New SACK. */ #define FLAG_ECE 0x40 /* ECE in this ACK */ -#define FLAG_DATA_LOST 0x80 /* SACK detected data lossage. */ #define FLAG_SLOWPATH 0x100 /* Do not skip RFC checks for window update.*/ #define FLAG_ONLY_ORIG_SACKED 0x200 /* SACKs only non-rexmit sent before RTO */ #define FLAG_SND_UNA_ADVANCED 0x400 /* Snd_una was changed (!= FLAG_DATA_ACKED) */ @@ -1040,13 +1039,11 @@ static void tcp_skb_mark_lost_uncond_verify(struct tcp_sock *tp, * These 6 states form finite state machine, controlled by the following events: * 1. New ACK (+SACK) arrives. (tcp_sacktag_write_queue()) * 2. Retransmission. (tcp_retransmit_skb(), tcp_xmit_retransmit_queue()) - * 3. Loss detection event of one of three flavors: + * 3. Loss detection event of two flavors: * A. Scoreboard estimator decided the packet is lost. * A'. Reno "three dupacks" marks head of queue lost. - * A''. Its FACK modfication, head until snd.fack is lost. - * B. SACK arrives sacking data transmitted after never retransmitted - * hole was sent out. - * C. SACK arrives sacking SND.NXT at the moment, when the + * A''. Its FACK modification, head until snd.fack is lost. + * B. SACK arrives sacking SND.NXT at the moment, when the * segment was retransmitted. * 4. D-SACK added new rule: D-SACK changes any tag to S. * @@ -1153,7 +1150,7 @@ static int tcp_is_sackblock_valid(struct tcp_sock *tp, int is_dsack, } /* Check for lost retransmit. This superb idea is borrowed from "ratehalving". - * Event "C". Later note: FACK people cheated me again 8), we have to account + * Event "B". Later note: FACK people cheated me again 8), we have to account * for reordering! Ugly, but should help. * * Search retransmitted skbs from write_queue that were sent when snd_nxt was @@ -1844,10 +1841,6 @@ tcp_sacktag_write_queue(struct sock *sk, const struct sk_buff *ack_skb, if (found_dup_sack && ((i + 1) == first_sack_index)) next_dup = &sp[i + 1]; - /* Event "B" in the comment above. */ - if (after(end_seq, tp->high_seq)) - state.flag |= FLAG_DATA_LOST; - /* Skip too early cached blocks */ while (tcp_sack_cache_ok(tp, cache) && !before(start_seq, cache->end_seq)) @@ -2515,8 +2508,11 @@ static void tcp_timeout_skbs(struct sock *sk) tcp_verify_left_out(tp); } -/* Mark head of queue up as lost. With RFC3517 SACK, the packets is - * is against sacked "cnt", otherwise it's against facked "cnt" +/* Detect loss in event "A" above by marking head of queue up as lost. + * For FACK or non-SACK(Reno) senders, the first "packets" number of segments + * are considered lost. For RFC3517 SACK, a segment is considered lost if it + * has at least tp->reordering SACKed seqments above it; "packets" refers to + * the maximum SACKed segments to pass before reaching this limit. */ static void tcp_mark_head_lost(struct sock *sk, int packets, int mark_head) { @@ -2525,6 +2521,8 @@ static void tcp_mark_head_lost(struct sock *sk, int packets, int mark_head) int cnt, oldcnt; int err; unsigned int mss; + /* Use SACK to deduce losses of new sequences sent during recovery */ + const u32 loss_high = tcp_is_sack(tp) ? tp->snd_nxt : tp->high_seq; WARN_ON(packets > tp->packets_out); if (tp->lost_skb_hint) { @@ -2546,7 +2544,7 @@ static void tcp_mark_head_lost(struct sock *sk, int packets, int mark_head) tp->lost_skb_hint = skb; tp->lost_cnt_hint = cnt; - if (after(TCP_SKB_CB(skb)->end_seq, tp->high_seq)) + if (after(TCP_SKB_CB(skb)->end_seq, loss_high)) break; oldcnt = cnt; @@ -3033,19 +3031,10 @@ static void tcp_fastretrans_alert(struct sock *sk, int pkts_acked, if (tcp_check_sack_reneging(sk, flag)) return; - /* C. Process data loss notification, provided it is valid. */ - if (tcp_is_fack(tp) && (flag & FLAG_DATA_LOST) && - before(tp->snd_una, tp->high_seq) && - icsk->icsk_ca_state != TCP_CA_Open && - tp->fackets_out > tp->reordering) { - tcp_mark_head_lost(sk, tp->fackets_out - tp->reordering, 0); - NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPLOSS); - } - - /* D. Check consistency of the current state. */ + /* C. Check consistency of the current state. */ tcp_verify_left_out(tp); - /* E. Check state exit conditions. State can be terminated + /* D. Check state exit conditions. State can be terminated * when high_seq is ACKed. */ if (icsk->icsk_ca_state == TCP_CA_Open) { WARN_ON(tp->retrans_out != 0); @@ -3077,7 +3066,7 @@ static void tcp_fastretrans_alert(struct sock *sk, int pkts_acked, } } - /* F. Process state. */ + /* E. Process state. */ switch (icsk->icsk_ca_state) { case TCP_CA_Recovery: if (!(flag & FLAG_SND_UNA_ADVANCED)) { From bf0813bd282f6ded0a5efca3db238287f7a3dbe8 Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Thu, 19 Jan 2012 15:40:06 +0000 Subject: [PATCH 702/820] pktgen: Fix unsigned function that is returning negative vals Every call to num_args() immediately checks the return value for less than zero, as it will return -EFAULT for a failed get_user() call. So it makes no sense for the function to be declared as an unsigned long. Signed-off-by: Paul Gortmaker Signed-off-by: David S. Miller --- net/core/pktgen.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/core/pktgen.c b/net/core/pktgen.c index 65f80c7b1656..4d8ce93cd503 100644 --- a/net/core/pktgen.c +++ b/net/core/pktgen.c @@ -767,8 +767,8 @@ done: return i; } -static unsigned long num_arg(const char __user * user_buffer, - unsigned long maxlen, unsigned long *num) +static long num_arg(const char __user *user_buffer, unsigned long maxlen, + unsigned long *num) { int i; *num = 0; From df505eb804d5221c3164ebecd1286cb7fc7f49ba Mon Sep 17 00:00:00 2001 From: Sathya Perla Date: Thu, 19 Jan 2012 20:34:04 +0000 Subject: [PATCH 703/820] be2net: create RSS rings even in multi-channel configs Currently RSS rings are not created in a multi-channel config. RSS rings can be created on one (out of four) interfaces per port in a multi-channel config. Doing this insulates the driver from a FW bug wherin multi-channel config is wrongly reported even when not enabled. This also helps performance in a multi-channel config, as one interface per port gets RSS rings. Signed-off-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/ethernet/emulex/benet/be_main.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c index a6bcdb5cd2be..e703d64434f8 100644 --- a/drivers/net/ethernet/emulex/benet/be_main.c +++ b/drivers/net/ethernet/emulex/benet/be_main.c @@ -1786,8 +1786,7 @@ static void be_rx_queues_destroy(struct be_adapter *adapter) static u32 be_num_rxqs_want(struct be_adapter *adapter) { if ((adapter->function_caps & BE_FUNCTION_CAPS_RSS) && - !sriov_enabled(adapter) && be_physfn(adapter) && - !be_is_mc(adapter)) { + !sriov_enabled(adapter) && be_physfn(adapter)) { return 1 + MAX_RSS_QS; /* one default non-RSS queue */ } else { dev_warn(&adapter->pdev->dev, From 8a622e71f58ec9f092fc99eacae0e6cf14f6e742 Mon Sep 17 00:00:00 2001 From: shawnlu Date: Fri, 20 Jan 2012 12:22:04 +0000 Subject: [PATCH 704/820] tcp: md5: using remote adress for md5 lookup in rst packet md5 key is added in socket through remote address. remote address should be used in finding md5 key when sending out reset packet. Signed-off-by: shawnlu Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 2 +- net/ipv6/tcp_ipv6.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 1eb4ad57670e..337ba4cca052 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -631,7 +631,7 @@ static void tcp_v4_send_reset(struct sock *sk, struct sk_buff *skb) arg.iov[0].iov_len = sizeof(rep.th); #ifdef CONFIG_TCP_MD5SIG - key = sk ? tcp_v4_md5_do_lookup(sk, ip_hdr(skb)->daddr) : NULL; + key = sk ? tcp_v4_md5_do_lookup(sk, ip_hdr(skb)->saddr) : NULL; if (key) { rep.opt[0] = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) | diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 906c7ca43542..3edd05ae4388 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1083,7 +1083,7 @@ static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb) #ifdef CONFIG_TCP_MD5SIG if (sk) - key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr); + key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->saddr); #endif if (th->ack) From af58f1d62853cd883c1fa26556fb470f05878f21 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 09:03:04 +0000 Subject: [PATCH 705/820] kernel-doc: fix new warning in net/phy/mdio_bus.c Fix new kernel-doc warning: Warning(drivers/net/phy/mdio_bus.c:49): No description found for parameter 'size' Signed-off-by: Randy Dunlap Signed-off-by: David S. Miller --- drivers/net/phy/mdio_bus.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/phy/mdio_bus.c b/drivers/net/phy/mdio_bus.c index 88cc5db9affd..8985cc62cf41 100644 --- a/drivers/net/phy/mdio_bus.c +++ b/drivers/net/phy/mdio_bus.c @@ -38,12 +38,11 @@ /** * mdiobus_alloc_size - allocate a mii_bus structure + * @size: extra amount of memory to allocate for private storage. + * If non-zero, then bus->priv is points to that memory. * * Description: called by a bus driver to allocate an mii_bus * structure to fill in. - * - * 'size' is an an extra amount of memory to allocate for private storage. - * If non-zero, then bus->priv is points to that memory. */ struct mii_bus *mdiobus_alloc_size(size_t size) { From 1a3bc369ba547c11ca8b3ed079d7584f27499e70 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 09:03:10 +0000 Subject: [PATCH 706/820] kernel-doc: fix new warning in net/sock.h Fix new kernel-doc warning: Warning(include/net/sock.h:372): No description found for parameter 'sk_cgrp_prioidx' Signed-off-by: Randy Dunlap Signed-off-by: David S. Miller --- include/net/sock.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/net/sock.h b/include/net/sock.h index bb972d254dff..97fc0ad47da0 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -226,6 +226,7 @@ struct cg_proto; * @sk_ack_backlog: current listen backlog * @sk_max_ack_backlog: listen backlog set in listen() * @sk_priority: %SO_PRIORITY setting + * @sk_cgrp_prioidx: socket group's priority map index * @sk_type: socket type (%SOCK_STREAM, etc) * @sk_protocol: which protocol this socket belongs in this network family * @sk_peer_pid: &struct pid for this socket's peer From 376be5ff8a6a36efadd131860cf26841f366d44c Mon Sep 17 00:00:00 2001 From: Glauber Costa Date: Fri, 20 Jan 2012 04:57:14 +0000 Subject: [PATCH 707/820] net: fix socket memcg build with !CONFIG_NET There is still a build bug with the sock memcg code, that triggers with !CONFIG_NET, that survived my series of randconfig builds. Signed-off-by: Glauber Costa Reported-by: Randy Dunlap CC: Hiroyouki Kamezawa Signed-off-by: David S. Miller --- include/net/sock.h | 2 +- mm/memcontrol.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 97fc0ad47da0..0e7a9b05f92b 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -922,7 +922,7 @@ inline void sk_refcnt_debug_release(const struct sock *sk) #define sk_refcnt_debug_release(sk) do { } while (0) #endif /* SOCK_REFCNT_DEBUG */ -#ifdef CONFIG_CGROUP_MEM_RES_CTLR_KMEM +#if defined(CONFIG_CGROUP_MEM_RES_CTLR_KMEM) && defined(CONFIG_NET) extern struct jump_label_key memcg_socket_limit_enabled; static inline struct cg_proto *parent_cg_proto(struct proto *proto, struct cg_proto *cg_proto) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 3dbff4dcde35..c3688dfd9a5f 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -379,7 +379,7 @@ static void mem_cgroup_put(struct mem_cgroup *memcg); static bool mem_cgroup_is_root(struct mem_cgroup *memcg); void sock_update_memcg(struct sock *sk) { - if (static_branch(&memcg_socket_limit_enabled)) { + if (mem_cgroup_sockets_enabled) { struct mem_cgroup *memcg; BUG_ON(!sk->sk_prot->proto_cgroup); @@ -411,7 +411,7 @@ EXPORT_SYMBOL(sock_update_memcg); void sock_release_memcg(struct sock *sk) { - if (static_branch(&memcg_socket_limit_enabled) && sk->sk_cgrp) { + if (mem_cgroup_sockets_enabled && sk->sk_cgrp) { struct mem_cgroup *memcg; WARN_ON(!sk->sk_cgrp->memcg); memcg = sk->sk_cgrp->memcg; From 8cfd14ad1eb52e44cb1fe7b47a68126e45e04026 Mon Sep 17 00:00:00 2001 From: Glauber Costa Date: Fri, 20 Jan 2012 04:57:15 +0000 Subject: [PATCH 708/820] cgroup: make sure memcg margin is 0 when over limit For the memcg sock code, we'll need to register allocations that are temporarily over limit. Let's make sure that margin is 0 in this case. I am keeping this as a separate patch, so that if any weirdness interaction appears in the future, we can now exactly what caused it. Suggested by Johannes Weiner Signed-off-by: Glauber Costa CC: KAMEZAWA Hiroyuki CC: Johannes Weiner CC: Michal Hocko CC: Tejun Heo CC: Li Zefan Acked-by: Tejun Heo Signed-off-by: David S. Miller --- include/linux/res_counter.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/res_counter.h b/include/linux/res_counter.h index c9d625ca659e..d06d014afda6 100644 --- a/include/linux/res_counter.h +++ b/include/linux/res_counter.h @@ -142,7 +142,10 @@ static inline unsigned long long res_counter_margin(struct res_counter *cnt) unsigned long flags; spin_lock_irqsave(&cnt->lock, flags); - margin = cnt->limit - cnt->usage; + if (cnt->limit > cnt->usage) + margin = cnt->limit - cnt->usage; + else + margin = 0; spin_unlock_irqrestore(&cnt->lock, flags); return margin; } From 0e90b31f4ba77027a7c21cbfc66404df0851ca21 Mon Sep 17 00:00:00 2001 From: Glauber Costa Date: Fri, 20 Jan 2012 04:57:16 +0000 Subject: [PATCH 709/820] net: introduce res_counter_charge_nofail() for socket allocations There is a case in __sk_mem_schedule(), where an allocation is beyond the maximum, but yet we are allowed to proceed. It happens under the following condition: sk->sk_wmem_queued + size >= sk->sk_sndbuf The network code won't revert the allocation in this case, meaning that at some point later it'll try to do it. Since this is never communicated to the underlying res_counter code, there is an inbalance in res_counter uncharge operation. I see two ways of fixing this: 1) storing the information about those allocations somewhere in memcg, and then deducting from that first, before we start draining the res_counter, 2) providing a slightly different allocation function for the res_counter, that matches the original behavior of the network code more closely. I decided to go for #2 here, believing it to be more elegant, since #1 would require us to do basically that, but in a more obscure way. Signed-off-by: Glauber Costa Cc: KAMEZAWA Hiroyuki Cc: Johannes Weiner Cc: Michal Hocko CC: Tejun Heo CC: Li Zefan CC: Laurent Chavey Acked-by: Tejun Heo Signed-off-by: David S. Miller --- include/linux/res_counter.h | 6 ++++++ include/net/sock.h | 10 ++++------ kernel/res_counter.c | 25 +++++++++++++++++++++++++ net/core/sock.c | 4 ++-- 4 files changed, 37 insertions(+), 8 deletions(-) diff --git a/include/linux/res_counter.h b/include/linux/res_counter.h index d06d014afda6..da81af086eaf 100644 --- a/include/linux/res_counter.h +++ b/include/linux/res_counter.h @@ -109,12 +109,18 @@ void res_counter_init(struct res_counter *counter, struct res_counter *parent); * * returns 0 on success and <0 if the counter->usage will exceed the * counter->limit _locked call expects the counter->lock to be taken + * + * charge_nofail works the same, except that it charges the resource + * counter unconditionally, and returns < 0 if the after the current + * charge we are over limit. */ int __must_check res_counter_charge_locked(struct res_counter *counter, unsigned long val); int __must_check res_counter_charge(struct res_counter *counter, unsigned long val, struct res_counter **limit_fail_at); +int __must_check res_counter_charge_nofail(struct res_counter *counter, + unsigned long val, struct res_counter **limit_fail_at); /* * uncharge - tell that some portion of the resource is released diff --git a/include/net/sock.h b/include/net/sock.h index 0e7a9b05f92b..4c69ac165e6b 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1008,9 +1008,8 @@ static inline void memcg_memory_allocated_add(struct cg_proto *prot, struct res_counter *fail; int ret; - ret = res_counter_charge(prot->memory_allocated, - amt << PAGE_SHIFT, &fail); - + ret = res_counter_charge_nofail(prot->memory_allocated, + amt << PAGE_SHIFT, &fail); if (ret < 0) *parent_status = OVER_LIMIT; } @@ -1054,12 +1053,11 @@ sk_memory_allocated_add(struct sock *sk, int amt, int *parent_status) } static inline void -sk_memory_allocated_sub(struct sock *sk, int amt, int parent_status) +sk_memory_allocated_sub(struct sock *sk, int amt) { struct proto *prot = sk->sk_prot; - if (mem_cgroup_sockets_enabled && sk->sk_cgrp && - parent_status != OVER_LIMIT) /* Otherwise was uncharged already */ + if (mem_cgroup_sockets_enabled && sk->sk_cgrp) memcg_memory_allocated_sub(sk->sk_cgrp, amt); atomic_long_sub(amt, prot->memory_allocated); diff --git a/kernel/res_counter.c b/kernel/res_counter.c index 6d269cce7aa1..d508363858b3 100644 --- a/kernel/res_counter.c +++ b/kernel/res_counter.c @@ -66,6 +66,31 @@ done: return ret; } +int res_counter_charge_nofail(struct res_counter *counter, unsigned long val, + struct res_counter **limit_fail_at) +{ + int ret, r; + unsigned long flags; + struct res_counter *c; + + r = ret = 0; + *limit_fail_at = NULL; + local_irq_save(flags); + for (c = counter; c != NULL; c = c->parent) { + spin_lock(&c->lock); + r = res_counter_charge_locked(c, val); + if (r) + c->usage += val; + spin_unlock(&c->lock); + if (r < 0 && ret == 0) { + *limit_fail_at = c; + ret = r; + } + } + local_irq_restore(flags); + + return ret; +} void res_counter_uncharge_locked(struct res_counter *counter, unsigned long val) { if (WARN_ON(counter->usage < val)) diff --git a/net/core/sock.c b/net/core/sock.c index 5c5af9988f94..3e81fd2e3c75 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1827,7 +1827,7 @@ suppress_allocation: /* Alas. Undo changes. */ sk->sk_forward_alloc -= amt * SK_MEM_QUANTUM; - sk_memory_allocated_sub(sk, amt, parent_status); + sk_memory_allocated_sub(sk, amt); return 0; } @@ -1840,7 +1840,7 @@ EXPORT_SYMBOL(__sk_mem_schedule); void __sk_mem_reclaim(struct sock *sk) { sk_memory_allocated_sub(sk, - sk->sk_forward_alloc >> SK_MEM_QUANTUM_SHIFT, 0); + sk->sk_forward_alloc >> SK_MEM_QUANTUM_SHIFT); sk->sk_forward_alloc &= SK_MEM_QUANTUM - 1; if (sk_under_memory_pressure(sk) && From b1cc16b8e643096adb92bbcb76c6c4c564141c40 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Sun, 22 Jan 2012 14:45:14 -0500 Subject: [PATCH 710/820] bluetooth: hci: Fix type of "enable_hs" to bool. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fixes: net/bluetooth/hci_core.c: In function ‘__check_enable_hs’: net/bluetooth/hci_core.c:2587:1: warning: return from incompatible pointer type [enabled by default] Signed-off-by: David S. Miller --- include/net/bluetooth/hci.h | 2 +- net/bluetooth/hci_core.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/include/net/bluetooth/hci.h b/include/net/bluetooth/hci.h index 5b2fed5eebf2..00596e816b4d 100644 --- a/include/net/bluetooth/hci.h +++ b/include/net/bluetooth/hci.h @@ -1388,6 +1388,6 @@ struct hci_inquiry_req { }; #define IREQ_CACHE_FLUSH 0x0001 -extern int enable_hs; +extern bool enable_hs; #endif /* __HCI_H */ diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 845da3ee56a0..9de93714213a 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -55,7 +55,7 @@ #define AUTO_OFF_TIMEOUT 2000 -int enable_hs; +bool enable_hs; static void hci_rx_work(struct work_struct *work); static void hci_cmd_work(struct work_struct *work); From f80400a26a2e8bff541de12834a1134358bb6642 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20Miros=C5=82aw?= Date: Sun, 22 Jan 2012 00:20:40 +0000 Subject: [PATCH 711/820] ethtool: allow ETHTOOL_GSSET_INFO for users MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Allow ETHTOOL_GSSET_INFO ethtool ioctl() for unprivileged users. ETHTOOL_GSTRINGS is already allowed, but is unusable without this one. Signed-off-by: Michał Mirosław Acked-by: Ben Hutchings Signed-off-by: David S. Miller --- net/core/ethtool.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/ethtool.c b/net/core/ethtool.c index 921aa2b4b415..369b41894527 100644 --- a/net/core/ethtool.c +++ b/net/core/ethtool.c @@ -1311,6 +1311,7 @@ int dev_ethtool(struct net *net, struct ifreq *ifr) case ETHTOOL_GRXCSUM: case ETHTOOL_GTXCSUM: case ETHTOOL_GSG: + case ETHTOOL_GSSET_INFO: case ETHTOOL_GSTRINGS: case ETHTOOL_GTSO: case ETHTOOL_GPERMADDR: From 2d58d7ea9164da59d0ea82fdf80e3ababe52d58c Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 4 Nov 2011 10:31:04 +0100 Subject: [PATCH 712/820] thermal: Rename generate_netlink_event It doesn't seem right for the thermal subsystem to export a symbol named generate_netlink_event. This function is thermal-specific and its name should reflect that fact. Rename it to thermal_generate_netlink_event. Signed-off-by: Jean Delvare Acked-by: Rafael J. Wysocki Acked-by: R.Durgadoss Signed-off-by: Len Brown --- Documentation/thermal/sysfs-api.txt | 2 +- drivers/thermal/thermal_sys.c | 4 ++-- include/linux/thermal.h | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/Documentation/thermal/sysfs-api.txt b/Documentation/thermal/sysfs-api.txt index b61e46f449aa..1733ab947a95 100644 --- a/Documentation/thermal/sysfs-api.txt +++ b/Documentation/thermal/sysfs-api.txt @@ -284,7 +284,7 @@ method, the sys I/F structure will be built like this: The framework includes a simple notification mechanism, in the form of a netlink event. Netlink socket initialization is done during the _init_ of the framework. Drivers which intend to use the notification mechanism -just need to call generate_netlink_event() with two arguments viz +just need to call thermal_generate_netlink_event() with two arguments viz (originator, event). Typically the originator will be an integer assigned to a thermal_zone_device when it registers itself with the framework. The event will be one of:{THERMAL_AUX0, THERMAL_AUX1, THERMAL_CRITICAL, diff --git a/drivers/thermal/thermal_sys.c b/drivers/thermal/thermal_sys.c index dd9a5743fa99..220ce7e31cf5 100644 --- a/drivers/thermal/thermal_sys.c +++ b/drivers/thermal/thermal_sys.c @@ -1304,7 +1304,7 @@ static struct genl_multicast_group thermal_event_mcgrp = { .name = THERMAL_GENL_MCAST_GROUP_NAME, }; -int generate_netlink_event(u32 orig, enum events event) +int thermal_generate_netlink_event(u32 orig, enum events event) { struct sk_buff *skb; struct nlattr *attr; @@ -1363,7 +1363,7 @@ int generate_netlink_event(u32 orig, enum events event) return result; } -EXPORT_SYMBOL(generate_netlink_event); +EXPORT_SYMBOL(thermal_generate_netlink_event); static int genetlink_init(void) { diff --git a/include/linux/thermal.h b/include/linux/thermal.h index 47b4a27e6e97..796f1ff0388c 100644 --- a/include/linux/thermal.h +++ b/include/linux/thermal.h @@ -152,9 +152,9 @@ struct thermal_cooling_device *thermal_cooling_device_register(char *, void *, void thermal_cooling_device_unregister(struct thermal_cooling_device *); #ifdef CONFIG_NET -extern int generate_netlink_event(u32 orig, enum events event); +extern int thermal_generate_netlink_event(u32 orig, enum events event); #else -static inline int generate_netlink_event(u32 orig, enum events event) +static inline int thermal_generate_netlink_event(u32 orig, enum events event) { return 0; } From 4e087a7a1f3884750790bda580e22e9eccd5f4fa Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Sat, 14 Jan 2012 21:20:37 +0100 Subject: [PATCH 713/820] gpu, drm, sis: Don't return uninitialized variable from sis_driver_load() In sis_driver_load(), the only use of 'ret' is as the return value from the function, unfortunately it is never initialized, so the function just returns garbage when it succeeds. To fix that, remove the variable and just return 0 directly on success. Signed-off-by: Jesper Juhl Reviewed-by: Daniel Vetter Signed-off-by: Dave Airlie --- drivers/gpu/drm/sis/sis_drv.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/gpu/drm/sis/sis_drv.c b/drivers/gpu/drm/sis/sis_drv.c index 06da063ece2e..573220cc5269 100644 --- a/drivers/gpu/drm/sis/sis_drv.c +++ b/drivers/gpu/drm/sis/sis_drv.c @@ -40,7 +40,6 @@ static struct pci_device_id pciidlist[] = { static int sis_driver_load(struct drm_device *dev, unsigned long chipset) { drm_sis_private_t *dev_priv; - int ret; dev_priv = kzalloc(sizeof(drm_sis_private_t), GFP_KERNEL); if (dev_priv == NULL) @@ -50,7 +49,7 @@ static int sis_driver_load(struct drm_device *dev, unsigned long chipset) dev_priv->chipset = chipset; idr_init(&dev->object_name_idr); - return ret; + return 0; } static int sis_driver_unload(struct drm_device *dev) From 44517c44496062180a6376cc704b33129441ce60 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Sun, 15 Jan 2012 08:51:12 -0500 Subject: [PATCH 714/820] drm/radeon/kms: Add an MSI quirk for Dell RS690 Interrupts only work with MSIs. https://bugs.freedesktop.org/show_bug.cgi?id=37679 Reported-by: Dmitry Podgorny Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon_irq_kms.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c index be38921bf761..66d5fe1c8174 100644 --- a/drivers/gpu/drm/radeon/radeon_irq_kms.c +++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c @@ -135,6 +135,12 @@ static bool radeon_msi_ok(struct radeon_device *rdev) (rdev->pdev->subsystem_device == 0x30c2)) return true; + /* Dell RS690 only seems to work with MSIs. */ + if ((rdev->pdev->device == 0x791f) && + (rdev->pdev->subsystem_vendor == 0x1028) && + (rdev->pdev->subsystem_device == 0x01fc)) + return true; + /* Dell RS690 only seems to work with MSIs. */ if ((rdev->pdev->device == 0x791f) && (rdev->pdev->subsystem_vendor == 0x1028) && From 11ef3f1f8780b64425a4cadbf42a46aa2e36895f Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 20 Jan 2012 14:47:43 -0500 Subject: [PATCH 715/820] drm/radeon/kms: add some missing semaphore init MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Alex Deucher Reviewed-by: Michel Dänzer Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/evergreen.c | 1 + drivers/gpu/drm/radeon/evergreend.h | 1 + drivers/gpu/drm/radeon/ni.c | 1 + drivers/gpu/drm/radeon/nid.h | 1 + 4 files changed, 4 insertions(+) diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 636660fca8c2..ae09fe82afbc 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1455,6 +1455,7 @@ int evergreen_cp_resume(struct radeon_device *rdev) #endif WREG32(CP_RB_CNTL, tmp); WREG32(CP_SEM_WAIT_TIMER, 0x0); + WREG32(CP_SEM_INCOMPLETE_TIMER_CNTL, 0x0); /* Set the write pointer delay */ WREG32(CP_RB_WPTR_DELAY, 0); diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h index b502216d42af..74713d42df29 100644 --- a/drivers/gpu/drm/radeon/evergreend.h +++ b/drivers/gpu/drm/radeon/evergreend.h @@ -108,6 +108,7 @@ #define CP_RB_WPTR_ADDR_HI 0xC11C #define CP_RB_WPTR_DELAY 0x8704 #define CP_SEM_WAIT_TIMER 0x85BC +#define CP_SEM_INCOMPLETE_TIMER_CNTL 0x85C8 #define CP_DEBUG 0xC1FC diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c index 321137295400..db09065e68fd 100644 --- a/drivers/gpu/drm/radeon/ni.c +++ b/drivers/gpu/drm/radeon/ni.c @@ -1219,6 +1219,7 @@ int cayman_cp_resume(struct radeon_device *rdev) RREG32(GRBM_SOFT_RESET); WREG32(CP_SEM_WAIT_TIMER, 0x0); + WREG32(CP_SEM_INCOMPLETE_TIMER_CNTL, 0x0); /* Set the write pointer delay */ WREG32(CP_RB_WPTR_DELAY, 0); diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h index f9df2a645e79..9a7f3b6e02de 100644 --- a/drivers/gpu/drm/radeon/nid.h +++ b/drivers/gpu/drm/radeon/nid.h @@ -222,6 +222,7 @@ #define SCRATCH_UMSK 0x8540 #define SCRATCH_ADDR 0x8544 #define CP_SEM_WAIT_TIMER 0x85BC +#define CP_SEM_INCOMPLETE_TIMER_CNTL 0x85C8 #define CP_COHER_CNTL2 0x85E8 #define CP_ME_CNTL 0x86D8 #define CP_ME_HALT (1 << 28) From 52409aa6a0e96337da137c069856298f4dd825a0 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 23 Jan 2012 17:10:24 +0100 Subject: [PATCH 716/820] ALSA: hda - Fix buffer-alignment regression with Nvidia HDMI The commit 2ae66c26550cd94b0e2606a9275eb0ab7070ad0e ALSA: hda: option to enable arbitrary buffer/period sizes introduced a regression on machines with Intel controller and Nvidia HDMI. The reason is that the driver modifies the global variable align_buffer_size when an Intel controller is found, and the Nvidia HDMI controller is probed after Intel although Nvidia chips require the aligned buffers. This patch fixes the problem by moving the flag into the local struct so that it's not affected by other controllers. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=42567 Cc: Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_intel.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index fb35474c1203..95dfb6874941 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -469,6 +469,7 @@ struct azx { unsigned int irq_pending_warned :1; unsigned int probing :1; /* codec probing phase */ unsigned int snoop:1; + unsigned int align_buffer_size:1; /* for debugging */ unsigned int last_cmd[AZX_MAX_CODECS]; @@ -1690,7 +1691,7 @@ static int azx_pcm_open(struct snd_pcm_substream *substream) runtime->hw.rates = hinfo->rates; snd_pcm_limit_hw_rates(runtime); snd_pcm_hw_constraint_integer(runtime, SNDRV_PCM_HW_PARAM_PERIODS); - if (align_buffer_size) + if (chip->align_buffer_size) /* constrain buffer sizes to be multiple of 128 bytes. This is more efficient in terms of memory access but isn't required by the HDA spec and @@ -2773,8 +2774,9 @@ static int __devinit azx_create(struct snd_card *card, struct pci_dev *pci, } /* disable buffer size rounding to 128-byte multiples if supported */ + chip->align_buffer_size = align_buffer_size; if (chip->driver_caps & AZX_DCAPS_BUFSIZE) - align_buffer_size = 0; + chip->align_buffer_size = 0; /* allow 64bit DMA address if supported by H/W */ if ((gcap & ICH6_GCAP_64OK) && !pci_set_dma_mask(pci, DMA_BIT_MASK(64))) From 29c5fbbcfefba5225a6783683c46c39e10877703 Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Mon, 23 Jan 2012 16:39:55 +0100 Subject: [PATCH 717/820] ALSA: HDA: Use model=auto for Thinkpad T510 The user reports that model=auto works fine for him. Using model=auto bring in new features such as jack detection notification to userspace. Alsa info is available at http://paste.ubuntu.com/805351/ Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_conexant.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c index 8a32a69c83c3..a7a5733aa4d2 100644 --- a/sound/pci/hda/patch_conexant.c +++ b/sound/pci/hda/patch_conexant.c @@ -3027,7 +3027,7 @@ static const struct snd_pci_quirk cxt5066_cfg_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x20f2, "Lenovo T400s", CXT5066_THINKPAD), SND_PCI_QUIRK(0x17aa, 0x21c5, "Thinkpad Edge 13", CXT5066_THINKPAD), SND_PCI_QUIRK(0x17aa, 0x21c6, "Thinkpad Edge 13", CXT5066_ASUS), - SND_PCI_QUIRK(0x17aa, 0x215e, "Lenovo Thinkpad", CXT5066_THINKPAD), + SND_PCI_QUIRK(0x17aa, 0x215e, "Lenovo T510", CXT5066_AUTO), SND_PCI_QUIRK(0x17aa, 0x21cf, "Lenovo T520 & W520", CXT5066_AUTO), SND_PCI_QUIRK(0x17aa, 0x21da, "Lenovo X220", CXT5066_THINKPAD), SND_PCI_QUIRK(0x17aa, 0x21db, "Lenovo X220-tablet", CXT5066_THINKPAD), From 6536e3123e5d3371a6f52e32a3d0694bcc987702 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Fri, 20 Jan 2012 14:33:53 -0800 Subject: [PATCH 718/820] mm: fix warnings regarding enum migrate_mode sparc64 allmodconfig: In file included from include/linux/compat.h:15, from /usr/src/25/arch/sparc/include/asm/siginfo.h:19, from include/linux/signal.h:5, from include/linux/sched.h:73, from arch/sparc/kernel/asm-offsets.c:13: include/linux/fs.h:618: warning: parameter has incomplete type It seems that my sparc64 compiler (gcc-3.4.5) doesn't like the forward declaration of enums. Fix this by moving the "enum migrate_mode" definition into its own header file. Acked-by: Mel Gorman Cc: Rik van Riel Cc: Andrea Arcangeli Cc: Minchan Kim Cc: Dave Jones Cc: Jan Kara Cc: Andy Isaacson Cc: Nai Xia Cc: Johannes Weiner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/fs.h | 2 +- include/linux/migrate.h | 14 +------------- include/linux/migrate_mode.h | 16 ++++++++++++++++ 3 files changed, 18 insertions(+), 14 deletions(-) create mode 100644 include/linux/migrate_mode.h diff --git a/include/linux/fs.h b/include/linux/fs.h index 0244082d42c5..4b3a41fe22bf 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -10,6 +10,7 @@ #include #include #include +#include /* * It's silly to have NR_OPEN bigger than NR_FILE, but you can change @@ -526,7 +527,6 @@ enum positive_aop_returns { struct page; struct address_space; struct writeback_control; -enum migrate_mode; struct iov_iter { const struct iovec *iov; diff --git a/include/linux/migrate.h b/include/linux/migrate.h index eaf867412f7a..05ed2828a553 100644 --- a/include/linux/migrate.h +++ b/include/linux/migrate.h @@ -3,22 +3,10 @@ #include #include +#include typedef struct page *new_page_t(struct page *, unsigned long private, int **); -/* - * MIGRATE_ASYNC means never block - * MIGRATE_SYNC_LIGHT in the current implementation means to allow blocking - * on most operations but not ->writepage as the potential stall time - * is too significant - * MIGRATE_SYNC will block when migrating pages - */ -enum migrate_mode { - MIGRATE_ASYNC, - MIGRATE_SYNC_LIGHT, - MIGRATE_SYNC, -}; - #ifdef CONFIG_MIGRATION #define PAGE_MIGRATION 1 diff --git a/include/linux/migrate_mode.h b/include/linux/migrate_mode.h new file mode 100644 index 000000000000..ebf3d89a3919 --- /dev/null +++ b/include/linux/migrate_mode.h @@ -0,0 +1,16 @@ +#ifndef MIGRATE_MODE_H_INCLUDED +#define MIGRATE_MODE_H_INCLUDED +/* + * MIGRATE_ASYNC means never block + * MIGRATE_SYNC_LIGHT in the current implementation means to allow blocking + * on most operations but not ->writepage as the potential stall time + * is too significant + * MIGRATE_SYNC will block when migrating pages + */ +enum migrate_mode { + MIGRATE_ASYNC, + MIGRATE_SYNC_LIGHT, + MIGRATE_SYNC, +}; + +#endif /* MIGRATE_MODE_H_INCLUDED */ From 687875fb7de4a95223af20ee024282fa9099f860 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Fri, 20 Jan 2012 14:33:55 -0800 Subject: [PATCH 719/820] mm: fix NULL ptr dereference in __count_immobile_pages Fix the following NULL ptr dereference caused by cat /sys/devices/system/memory/memory0/removable Pid: 13979, comm: sed Not tainted 3.0.13-0.5-default #1 IBM BladeCenter LS21 -[7971PAM]-/Server Blade RIP: __count_immobile_pages+0x4/0x100 Process sed (pid: 13979, threadinfo ffff880221c36000, task ffff88022e788480) Call Trace: is_pageblock_removable_nolock+0x34/0x40 is_mem_section_removable+0x74/0xf0 show_mem_removable+0x41/0x70 sysfs_read_file+0xfe/0x1c0 vfs_read+0xc7/0x130 sys_read+0x53/0xa0 system_call_fastpath+0x16/0x1b We are crashing because we are trying to dereference NULL zone which came from pfn=0 (struct page ffffea0000000000). According to the boot log this page is marked reserved: e820 update range: 0000000000000000 - 0000000000010000 (usable) ==> (reserved) and early_node_map confirms that: early_node_map[3] active PFN ranges 1: 0x00000010 -> 0x0000009c 1: 0x00000100 -> 0x000bffa3 1: 0x00100000 -> 0x00240000 The problem is that memory_present works in PAGE_SECTION_MASK aligned blocks so the reserved range sneaks into the the section as well. This also means that free_area_init_node will not take care of those reserved pages and they stay uninitialized. When we try to read the removable status we walk through all available sections and hope that the zone is valid for all pages in the section. But this is not true in this case as the zone and nid are not initialized. We have only one node in this particular case and it is marked as node=1 (rather than 0) and that made the problem visible because page_to_nid will return 0 and there are no zones on the node. Let's check that the zone is valid and that the given pfn falls into its boundaries and mark the section not removable. This might cause some false positives, probably, but we do not have any sane way to find out whether the page is reserved by the platform or it is just not used for whatever other reasons. Signed-off-by: Michal Hocko Acked-by: Mel Gorman Cc: KAMEZAWA Hiroyuki Cc: Andrea Arcangeli Cc: David Rientjes Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 0027d8f4a1bb..21c272d517bc 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -5414,6 +5414,17 @@ __count_immobile_pages(struct zone *zone, struct page *page, int count) bool is_pageblock_removable_nolock(struct page *page) { struct zone *zone = page_zone(page); + unsigned long pfn = page_to_pfn(page); + + /* + * We have to be careful here because we are iterating over memory + * sections which are not zone aware so we might end up outside of + * the zone but still within the section. + */ + if (!zone || zone->zone_start_pfn > pfn || + zone->zone_start_pfn + zone->spanned_pages <= pfn) + return false; + return __count_immobile_pages(zone, page, 0); } From 656a070629adfe23c12768e35ddf91feab469ff7 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Fri, 20 Jan 2012 14:33:58 -0800 Subject: [PATCH 720/820] mm: __count_immobile_pages(): make sure the node is online page_zone() requires an online node otherwise we are accessing NULL NODE_DATA. This is not an issue at the moment because node_zones are located at the structure beginning but this might change in the future so better be careful about that. Signed-off-by: Michal Hocko Signed-off-by: KAMEZAWA Hiroyuki Acked-by: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 21c272d517bc..d2186ecb36f7 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -5413,15 +5413,22 @@ __count_immobile_pages(struct zone *zone, struct page *page, int count) bool is_pageblock_removable_nolock(struct page *page) { - struct zone *zone = page_zone(page); - unsigned long pfn = page_to_pfn(page); + struct zone *zone; + unsigned long pfn; /* * We have to be careful here because we are iterating over memory * sections which are not zone aware so we might end up outside of * the zone but still within the section. + * We have to take care about the node as well. If the node is offline + * its NODE_DATA will be NULL - see page_zone. */ - if (!zone || zone->zone_start_pfn > pfn || + if (!node_online(page_to_nid(page))) + return false; + + zone = page_zone(page); + pfn = page_to_pfn(page); + if (zone->zone_start_pfn > pfn || zone->zone_start_pfn + zone->spanned_pages <= pfn) return false; From e9a4593cc5e36c6d47c87b439cb41c2568e7395f Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Fri, 20 Jan 2012 14:33:59 -0800 Subject: [PATCH 721/820] leds: add led driver for Bachmann's ot200 Add support for leds on Bachmann's ot200 visualisation device. The device has three leds on the back panel (led_err, led_init and led_run) and can handle up to seven leds on the front panel. The driver was written by Linutronix on behalf of Bachmann electronic GmbH. It incorporates feedback from Lars-Peter Clausen [akpm@linux-foundation.org: add dependency on HAS_IOMEM] Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Christian Gmeiner Cc: Lars-Peter Clausen Cc: Richard Purdie Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/leds/Kconfig | 7 ++ drivers/leds/Makefile | 1 + drivers/leds/leds-ot200.c | 171 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 179 insertions(+) create mode 100644 drivers/leds/leds-ot200.c diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig index c957c344233f..9ca28fced2b9 100644 --- a/drivers/leds/Kconfig +++ b/drivers/leds/Kconfig @@ -403,6 +403,13 @@ config LEDS_MAX8997 This option enables support for on-chip LED drivers on MAXIM MAX8997 PMIC. +config LEDS_OT200 + tristate "LED support for the Bachmann OT200" + depends on LEDS_CLASS && HAS_IOMEM + help + This option enables support for the LEDs on the Bachmann OT200. + Say Y to enable LEDs on the Bachmann OT200. + config LEDS_TRIGGERS bool "LED Trigger support" depends on LEDS_CLASS diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile index b8a9723477f0..1fc6875a8b20 100644 --- a/drivers/leds/Makefile +++ b/drivers/leds/Makefile @@ -28,6 +28,7 @@ obj-$(CONFIG_LEDS_LP5523) += leds-lp5523.o obj-$(CONFIG_LEDS_TCA6507) += leds-tca6507.o obj-$(CONFIG_LEDS_CLEVO_MAIL) += leds-clevo-mail.o obj-$(CONFIG_LEDS_HP6XX) += leds-hp6xx.o +obj-$(CONFIG_LEDS_OT200) += leds-ot200.o obj-$(CONFIG_LEDS_FSG) += leds-fsg.o obj-$(CONFIG_LEDS_PCA955X) += leds-pca955x.o obj-$(CONFIG_LEDS_DA903X) += leds-da903x.o diff --git a/drivers/leds/leds-ot200.c b/drivers/leds/leds-ot200.c new file mode 100644 index 000000000000..c4646825a620 --- /dev/null +++ b/drivers/leds/leds-ot200.c @@ -0,0 +1,171 @@ +/* + * Bachmann ot200 leds driver. + * + * Author: Sebastian Andrzej Siewior + * Christian Gmeiner + * + * License: GPL as published by the FSF. + */ + +#include +#include +#include +#include +#include +#include +#include + + +struct ot200_led { + struct led_classdev cdev; + const char *name; + unsigned long port; + u8 mask; +}; + +/* + * The device has three leds on the back panel (led_err, led_init and led_run) + * and can handle up to seven leds on the front panel. + */ + +static struct ot200_led leds[] = { + { + .name = "led_run", + .port = 0x5a, + .mask = BIT(0), + }, + { + .name = "led_init", + .port = 0x5a, + .mask = BIT(1), + }, + { + .name = "led_err", + .port = 0x5a, + .mask = BIT(2), + }, + { + .name = "led_1", + .port = 0x49, + .mask = BIT(7), + }, + { + .name = "led_2", + .port = 0x49, + .mask = BIT(6), + }, + { + .name = "led_3", + .port = 0x49, + .mask = BIT(5), + }, + { + .name = "led_4", + .port = 0x49, + .mask = BIT(4), + }, + { + .name = "led_5", + .port = 0x49, + .mask = BIT(3), + }, + { + .name = "led_6", + .port = 0x49, + .mask = BIT(2), + }, + { + .name = "led_7", + .port = 0x49, + .mask = BIT(1), + } +}; + +static DEFINE_SPINLOCK(value_lock); + +/* + * we need to store the current led states, as it is not + * possible to read the current led state via inb(). + */ +static u8 leds_back; +static u8 leds_front; + +static void ot200_led_brightness_set(struct led_classdev *led_cdev, + enum led_brightness value) +{ + struct ot200_led *led = container_of(led_cdev, struct ot200_led, cdev); + u8 *val; + unsigned long flags; + + spin_lock_irqsave(&value_lock, flags); + + if (led->port == 0x49) + val = &leds_front; + else if (led->port == 0x5a) + val = &leds_back; + else + BUG(); + + if (value == LED_OFF) + *val &= ~led->mask; + else + *val |= led->mask; + + outb(*val, led->port); + spin_unlock_irqrestore(&value_lock, flags); +} + +static int __devinit ot200_led_probe(struct platform_device *pdev) +{ + int i; + int ret; + + for (i = 0; i < ARRAY_SIZE(leds); i++) { + + leds[i].cdev.name = leds[i].name; + leds[i].cdev.brightness_set = ot200_led_brightness_set; + + ret = led_classdev_register(&pdev->dev, &leds[i].cdev); + if (ret < 0) + goto err; + } + + leds_front = 0; /* turn off all front leds */ + leds_back = BIT(1); /* turn on init led */ + outb(leds_front, 0x49); + outb(leds_back, 0x5a); + + return 0; + +err: + for (i = i - 1; i >= 0; i--) + led_classdev_unregister(&leds[i].cdev); + + return ret; +} + +static int __devexit ot200_led_remove(struct platform_device *pdev) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(leds); i++) + led_classdev_unregister(&leds[i].cdev); + + return 0; +} + +static struct platform_driver ot200_led_driver = { + .probe = ot200_led_probe, + .remove = __devexit_p(ot200_led_remove), + .driver = { + .name = "leds-ot200", + .owner = THIS_MODULE, + }, +}; + +module_platform_driver(ot200_led_driver); + +MODULE_AUTHOR("Sebastian A. Siewior "); +MODULE_DESCRIPTION("ot200 LED driver"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS("platform:leds-ot200"); From 2a4e64b8f6bcbf23ddd375b78342051ae8862284 Mon Sep 17 00:00:00 2001 From: Davidlohr Bueso Date: Fri, 20 Jan 2012 14:34:01 -0800 Subject: [PATCH 722/820] ipc/mqueue: simplify reading msgqueue limit Because the current task is being used to get the limit, we can simply use rlimit() instead of task_rlimit(). Signed-off-by: Davidlohr Bueso Acked-by: KOSAKI Motohiro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- ipc/mqueue.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/ipc/mqueue.c b/ipc/mqueue.c index 9b7c8ab7d75c..86ee272de210 100644 --- a/ipc/mqueue.c +++ b/ipc/mqueue.c @@ -128,7 +128,6 @@ static struct inode *mqueue_get_inode(struct super_block *sb, if (S_ISREG(mode)) { struct mqueue_inode_info *info; - struct task_struct *p = current; unsigned long mq_bytes, mq_msg_tblsz; inode->i_fop = &mqueue_file_operations; @@ -159,7 +158,7 @@ static struct inode *mqueue_get_inode(struct super_block *sb, spin_lock(&mq_lock); if (u->mq_bytes + mq_bytes < u->mq_bytes || - u->mq_bytes + mq_bytes > task_rlimit(p, RLIMIT_MSGQUEUE)) { + u->mq_bytes + mq_bytes > rlimit(RLIMIT_MSGQUEUE)) { spin_unlock(&mq_lock); /* mqueue_evict_inode() releases info->messages */ ret = -EMFILE; From d496aab567e7e52b3e974c9192a5de6e77dce32c Mon Sep 17 00:00:00 2001 From: Ananth N Mavinakayanahalli Date: Fri, 20 Jan 2012 14:34:04 -0800 Subject: [PATCH 723/820] kprobes: initialize before using a hlist Commit ef53d9c5e ("kprobes: improve kretprobe scalability with hashed locking") introduced a bug where we can potentially leak kretprobe_instances since we initialize a hlist head after having used it. Initialize the hlist head before using it. Reported by: Jim Keniston Acked-by: Jim Keniston Signed-off-by: Ananth N Mavinakayanahalli Acked-by: Masami Hiramatsu Cc: Srinivasa D S Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/kprobes.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 95dd7212e610..29f5b65bee29 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -1077,6 +1077,7 @@ void __kprobes kprobe_flush_task(struct task_struct *tk) /* Early boot. kretprobe_table_locks not yet initialized. */ return; + INIT_HLIST_HEAD(&empty_rp); hash = hash_ptr(tk, KPROBE_HASH_BITS); head = &kretprobe_inst_table[hash]; kretprobe_table_lock(hash, &flags); @@ -1085,7 +1086,6 @@ void __kprobes kprobe_flush_task(struct task_struct *tk) recycle_rp_inst(ri, &empty_rp); } kretprobe_table_unlock(hash, &flags); - INIT_HLIST_HEAD(&empty_rp); hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) { hlist_del(&ri->hlist); kfree(ri); From 36c3e75907c8cb515fad260190ca1beb7e53df96 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 20 Jan 2012 14:34:05 -0800 Subject: [PATCH 724/820] drivers/video/backlight/adp88x0_bl.c: fix bit testing logic We need to write new value if the bit mask fields of new value is not equal to old value. It does not make sense to write new value only when all the bit_mask bits are zero. Signed-off-by: Axel Lin Cc: Michael Hennerich Cc: Richard Purdie Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/video/backlight/adp8860_bl.c | 2 +- drivers/video/backlight/adp8870_bl.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/video/backlight/adp8860_bl.c b/drivers/video/backlight/adp8860_bl.c index 66bc74d9ce2a..378276c9d3cf 100644 --- a/drivers/video/backlight/adp8860_bl.c +++ b/drivers/video/backlight/adp8860_bl.c @@ -146,7 +146,7 @@ static int adp8860_set_bits(struct i2c_client *client, int reg, uint8_t bit_mask ret = adp8860_read(client, reg, ®_val); - if (!ret && ((reg_val & bit_mask) == 0)) { + if (!ret && ((reg_val & bit_mask) != bit_mask)) { reg_val |= bit_mask; ret = adp8860_write(client, reg, reg_val); } diff --git a/drivers/video/backlight/adp8870_bl.c b/drivers/video/backlight/adp8870_bl.c index 6c68a6899e87..6735059376d6 100644 --- a/drivers/video/backlight/adp8870_bl.c +++ b/drivers/video/backlight/adp8870_bl.c @@ -160,7 +160,7 @@ static int adp8870_set_bits(struct i2c_client *client, int reg, uint8_t bit_mask ret = adp8870_read(client, reg, ®_val); - if (!ret && ((reg_val & bit_mask) == 0)) { + if (!ret && ((reg_val & bit_mask) != bit_mask)) { reg_val |= bit_mask; ret = adp8870_write(client, reg, reg_val); } From d59d9ebaacba32b63f24d53b1463519b445b4683 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 20 Jan 2012 14:34:06 -0800 Subject: [PATCH 725/820] drivers/video/backlight/l4f00242t03.c: return proper error in l4f00242t03_probe if regulator_get() fails Signed-off-by: Axel Lin Acked-by: Alberto Panizzo Cc: Richard Purdie Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/video/backlight/l4f00242t03.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/video/backlight/l4f00242t03.c b/drivers/video/backlight/l4f00242t03.c index 4f5d1c4cb6ab..27d1d7a29c77 100644 --- a/drivers/video/backlight/l4f00242t03.c +++ b/drivers/video/backlight/l4f00242t03.c @@ -190,6 +190,7 @@ static int __devinit l4f00242t03_probe(struct spi_device *spi) priv->io_reg = regulator_get(&spi->dev, "vdd"); if (IS_ERR(priv->io_reg)) { + ret = PTR_ERR(priv->io_reg); dev_err(&spi->dev, "%s: Unable to get the IO regulator\n", __func__); goto err3; @@ -197,6 +198,7 @@ static int __devinit l4f00242t03_probe(struct spi_device *spi) priv->core_reg = regulator_get(&spi->dev, "vcore"); if (IS_ERR(priv->core_reg)) { + ret = PTR_ERR(priv->core_reg); dev_err(&spi->dev, "%s: Unable to get the core regulator\n", __func__); goto err4; From 85e72aa5384b1a614563ad63257ded0e91d1a620 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 20 Jan 2012 14:34:09 -0800 Subject: [PATCH 726/820] proc: clear_refs: do not clear reserved pages /proc/pid/clear_refs is used to clear the Referenced and YOUNG bits for pages and corresponding page table entries of the task with PID pid, which includes any special mappings inserted into the page tables in order to provide things like vDSOs and user helper functions. On ARM this causes a problem because the vectors page is mapped as a global mapping and since ec706dab ("ARM: add a vma entry for the user accessible vector page"), a VMA is also inserted into each task for this page to aid unwinding through signals and syscall restarts. Since the vectors page is required for handling faults, clearing the YOUNG bit (and subsequently writing a faulting pte) means that we lose the vectors page *globally* and cannot fault it back in. This results in a system deadlock on the next exception. To see this problem in action, just run: $ echo 1 > /proc/self/clear_refs on an ARM platform (as any user) and watch your system hang. I think this has been the case since 2.6.37 This patch avoids clearing the aforementioned bits for reserved pages, therefore leaving the vectors page intact on ARM. Since reserved pages are not candidates for swap, this change should not have any impact on the usefulness of clear_refs. Signed-off-by: Will Deacon Reported-by: Moussa Ba Acked-by: Hugh Dickins Cc: David Rientjes Cc: Russell King Acked-by: Nicolas Pitre Cc: Matt Mackall Cc: [2.6.37+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/task_mmu.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index e418c5abdb0e..7dcd2a250495 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -518,6 +518,9 @@ static int clear_refs_pte_range(pmd_t *pmd, unsigned long addr, if (!page) continue; + if (PageReserved(page)) + continue; + /* Clear accessed and referenced bits. */ ptep_test_and_clear_young(vma, addr, pte); ClearPageReferenced(page); From 6568d4a9c9ff16d6c4f0b14dfea567806ce579e4 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Fri, 20 Jan 2012 14:34:12 -0800 Subject: [PATCH 727/820] mm: memcg: update the correct soft limit tree during migration end_migration() passes the old page instead of the new page to commit the charge. This page descriptor is not used for committing itself, though, since we also pass the (correct) page_cgroup descriptor. But it's used to find the soft limit tree through the page's zone, so the soft limit tree of the old page's zone is updated instead of that of the new page's, which might get slightly out of date until the next charge reaches the ratelimit point. This glitch has been present since 5564e88 ("memcg: condense page_cgroup-to-page lookup points"). This fixes a bug that I introduced in 2.6.38. It's benign enough (to my knowledge) that we probably don't want this for stable. Reported-by: Hugh Dickins Signed-off-by: Johannes Weiner Acked-by: KAMEZAWA Hiroyuki Acked-by: Michal Hocko Acked-by: Kirill A. Shutemov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memcontrol.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 3dbff4dcde35..4baddbae94cc 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -3247,7 +3247,7 @@ int mem_cgroup_prepare_migration(struct page *page, ctype = MEM_CGROUP_CHARGE_TYPE_CACHE; else ctype = MEM_CGROUP_CHARGE_TYPE_SHMEM; - __mem_cgroup_commit_charge(memcg, page, 1, pc, ctype); + __mem_cgroup_commit_charge(memcg, newpage, 1, pc, ctype); return ret; } From 409eb8c2611b4310947a150af988111f7f52ab15 Mon Sep 17 00:00:00 2001 From: Hillf Danton Date: Fri, 20 Jan 2012 14:34:13 -0800 Subject: [PATCH 728/820] mm/hugetlb.c: undo change to page mapcount in fault handler Page mapcount should be updated only if we are sure that the page ends up in the page table otherwise we would leak if we couldn't COW due to reservations or if idx is out of bounds. Signed-off-by: Hillf Danton Reviewed-by: Michal Hocko Acked-by: KAMEZAWA Hiroyuki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/hugetlb.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index ea8c3a4cd2ae..5f34bd8dda34 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -2508,6 +2508,7 @@ static int hugetlb_no_page(struct mm_struct *mm, struct vm_area_struct *vma, { struct hstate *h = hstate_vma(vma); int ret = VM_FAULT_SIGBUS; + int anon_rmap = 0; pgoff_t idx; unsigned long size; struct page *page; @@ -2562,14 +2563,13 @@ retry: spin_lock(&inode->i_lock); inode->i_blocks += blocks_per_huge_page(h); spin_unlock(&inode->i_lock); - page_dup_rmap(page); } else { lock_page(page); if (unlikely(anon_vma_prepare(vma))) { ret = VM_FAULT_OOM; goto backout_unlocked; } - hugepage_add_new_anon_rmap(page, vma, address); + anon_rmap = 1; } } else { /* @@ -2582,7 +2582,6 @@ retry: VM_FAULT_SET_HINDEX(h - hstates); goto backout_unlocked; } - page_dup_rmap(page); } /* @@ -2606,6 +2605,10 @@ retry: if (!huge_pte_none(huge_ptep_get(ptep))) goto backout; + if (anon_rmap) + hugepage_add_new_anon_rmap(page, vma, address); + else + page_dup_rmap(page); new_pte = make_huge_pte(vma, page, ((vma->vm_flags & VM_WRITE) && (vma->vm_flags & VM_SHARED))); set_huge_pte_at(mm, address, ptep, new_pte); From cb78edfdcef5259ac9e9088bd63810d21299928d Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Fri, 20 Jan 2012 14:34:16 -0800 Subject: [PATCH 729/820] kdump: define KEXEC_NOTE_BYTES arch specific for s390x kdump only allocates memory for the prstatus ELF note. For s390x, besides of prstatus multiple ELF notes for various different register types are stored. Therefore the currently allocated memory is not sufficient. With this patch the KEXEC_NOTE_BYTES macro can be defined by architecture code and for s390x it is set to the correct size now. Signed-off-by: Michael Holzheu Cc: "Eric W. Biederman" Cc: Vivek Goyal Cc: Martin Schwidefsky Cc: Heiko Carstens Reviewed-by: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/include/asm/kexec.h | 18 ++++++++++++++++++ include/linux/kexec.h | 2 ++ 2 files changed, 20 insertions(+) diff --git a/arch/s390/include/asm/kexec.h b/arch/s390/include/asm/kexec.h index cf4e47b0948c..3f30dac804ea 100644 --- a/arch/s390/include/asm/kexec.h +++ b/arch/s390/include/asm/kexec.h @@ -42,6 +42,24 @@ /* The native architecture */ #define KEXEC_ARCH KEXEC_ARCH_S390 +/* + * Size for s390x ELF notes per CPU + * + * Seven notes plus zero note at the end: prstatus, fpregset, timer, + * tod_cmp, tod_reg, control regs, and prefix + */ +#define KEXEC_NOTE_BYTES \ + (ALIGN(sizeof(struct elf_note), 4) * 8 + \ + ALIGN(sizeof("CORE"), 4) * 7 + \ + ALIGN(sizeof(struct elf_prstatus), 4) + \ + ALIGN(sizeof(elf_fpregset_t), 4) + \ + ALIGN(sizeof(u64), 4) + \ + ALIGN(sizeof(u64), 4) + \ + ALIGN(sizeof(u32), 4) + \ + ALIGN(sizeof(u64) * 16, 4) + \ + ALIGN(sizeof(u32), 4) \ + ) + /* Provide a dummy definition to avoid build failures. */ static inline void crash_setup_regs(struct pt_regs *newregs, struct pt_regs *oldregs) { } diff --git a/include/linux/kexec.h b/include/linux/kexec.h index 2fa0901219d4..0d7d6a1b172f 100644 --- a/include/linux/kexec.h +++ b/include/linux/kexec.h @@ -50,9 +50,11 @@ * note header. For kdump, the code in vmcore.c runs in the context * of the second kernel to combine them into one note. */ +#ifndef KEXEC_NOTE_BYTES #define KEXEC_NOTE_BYTES ( (KEXEC_NOTE_HEAD_BYTES * 2) + \ KEXEC_CORE_NOTE_NAME_BYTES + \ KEXEC_CORE_NOTE_DESC_BYTES ) +#endif /* * This structure is used to hold the arguments that are used when loading From 85046579bde15e532983438f86b36856e358f417 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Fri, 20 Jan 2012 14:34:19 -0800 Subject: [PATCH 730/820] SHM_UNLOCK: fix long unpreemptible section scan_mapping_unevictable_pages() is used to make SysV SHM_LOCKed pages evictable again once the shared memory is unlocked. It does this with pagevec_lookup()s across the whole object (which might occupy most of memory), and takes 300ms to unlock 7GB here. A cond_resched() every PAGEVEC_SIZE pages would be good. However, KOSAKI-san points out that this is called under shmem.c's info->lock, and it's also under shm.c's shm_lock(), both spinlocks. There is no strong reason for that: we need to take these pages off the unevictable list soonish, but those locks are not required for it. So move the call to scan_mapping_unevictable_pages() from shmem.c's unlock handling up to shm.c's unlock handling. Remove the recently added barrier, not needed now we have spin_unlock() before the scan. Use get_file(), with subsequent fput(), to make sure we have a reference to mapping throughout scan_mapping_unevictable_pages(): that's something that was previously guaranteed by the shm_lock(). Remove shmctl's lru_add_drain_all(): we don't fault in pages at SHM_LOCK time, and we lazily discover them to be Unevictable later, so it serves no purpose for SHM_LOCK; and serves no purpose for SHM_UNLOCK, since pages still on pagevec are not marked Unevictable. The original code avoided redundant rescans by checking VM_LOCKED flag at its level: now avoid them by checking shp's SHM_LOCKED. The original code called scan_mapping_unevictable_pages() on a locked area at shm_destroy() time: perhaps we once had accounting cross-checks which required that, but not now, so skip the overhead and just let inode eviction deal with them. Put check_move_unevictable_page() and scan_mapping_unevictable_pages() under CONFIG_SHMEM (with stub for the TINY case when ramfs is used), more as comment than to save space; comment them used for SHM_UNLOCK. Signed-off-by: Hugh Dickins Reviewed-by: KOSAKI Motohiro Cc: Minchan Kim Cc: Rik van Riel Cc: Shaohua Li Cc: Eric Dumazet Cc: Johannes Weiner Cc: Michel Lespinasse Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- ipc/shm.c | 37 ++++++++++++++++++++++--------------- mm/shmem.c | 7 ------- mm/vmscan.c | 12 +++++++++++- 3 files changed, 33 insertions(+), 23 deletions(-) diff --git a/ipc/shm.c b/ipc/shm.c index 02ecf2c078fc..854ab58e5f6e 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -870,9 +870,7 @@ SYSCALL_DEFINE3(shmctl, int, shmid, int, cmd, struct shmid_ds __user *, buf) case SHM_LOCK: case SHM_UNLOCK: { - struct file *uninitialized_var(shm_file); - - lru_add_drain_all(); /* drain pagevecs to lru lists */ + struct file *shm_file; shp = shm_lock_check(ns, shmid); if (IS_ERR(shp)) { @@ -895,22 +893,31 @@ SYSCALL_DEFINE3(shmctl, int, shmid, int, cmd, struct shmid_ds __user *, buf) err = security_shm_shmctl(shp, cmd); if (err) goto out_unlock; - - if(cmd==SHM_LOCK) { + + shm_file = shp->shm_file; + if (is_file_hugepages(shm_file)) + goto out_unlock; + + if (cmd == SHM_LOCK) { struct user_struct *user = current_user(); - if (!is_file_hugepages(shp->shm_file)) { - err = shmem_lock(shp->shm_file, 1, user); - if (!err && !(shp->shm_perm.mode & SHM_LOCKED)){ - shp->shm_perm.mode |= SHM_LOCKED; - shp->mlock_user = user; - } + err = shmem_lock(shm_file, 1, user); + if (!err && !(shp->shm_perm.mode & SHM_LOCKED)) { + shp->shm_perm.mode |= SHM_LOCKED; + shp->mlock_user = user; } - } else if (!is_file_hugepages(shp->shm_file)) { - shmem_lock(shp->shm_file, 0, shp->mlock_user); - shp->shm_perm.mode &= ~SHM_LOCKED; - shp->mlock_user = NULL; + goto out_unlock; } + + /* SHM_UNLOCK */ + if (!(shp->shm_perm.mode & SHM_LOCKED)) + goto out_unlock; + shmem_lock(shm_file, 0, shp->mlock_user); + shp->shm_perm.mode &= ~SHM_LOCKED; + shp->mlock_user = NULL; + get_file(shm_file); shm_unlock(shp); + scan_mapping_unevictable_pages(shm_file->f_mapping); + fput(shm_file); goto out; } case IPC_RMID: diff --git a/mm/shmem.c b/mm/shmem.c index feead1943d92..4aaa53abe302 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -1068,13 +1068,6 @@ int shmem_lock(struct file *file, int lock, struct user_struct *user) user_shm_unlock(inode->i_size, user); info->flags &= ~VM_LOCKED; mapping_clear_unevictable(file->f_mapping); - /* - * Ensure that a racing putback_lru_page() can see - * the pages of this mapping are evictable when we - * skip them due to !PageLRU during the scan. - */ - smp_mb__after_clear_bit(); - scan_mapping_unevictable_pages(file->f_mapping); } retval = 0; diff --git a/mm/vmscan.c b/mm/vmscan.c index 2880396f7953..e097c1026b58 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -3499,6 +3499,7 @@ int page_evictable(struct page *page, struct vm_area_struct *vma) return 1; } +#ifdef CONFIG_SHMEM /** * check_move_unevictable_page - check page for evictability and move to appropriate zone lru list * @page: page to check evictability and move to appropriate lru list @@ -3509,6 +3510,8 @@ int page_evictable(struct page *page, struct vm_area_struct *vma) * * Restrictions: zone->lru_lock must be held, page must be on LRU and must * have PageUnevictable set. + * + * This function is only used for SysV IPC SHM_UNLOCK. */ static void check_move_unevictable_page(struct page *page, struct zone *zone) { @@ -3545,6 +3548,8 @@ retry: * * Scan all pages in mapping. Check unevictable pages for * evictability and move them to the appropriate zone lru list. + * + * This function is only used for SysV IPC SHM_UNLOCK. */ void scan_mapping_unevictable_pages(struct address_space *mapping) { @@ -3590,9 +3595,14 @@ void scan_mapping_unevictable_pages(struct address_space *mapping) pagevec_release(&pvec); count_vm_events(UNEVICTABLE_PGSCANNED, pg_scanned); + cond_resched(); } - } +#else +void scan_mapping_unevictable_pages(struct address_space *mapping) +{ +} +#endif /* CONFIG_SHMEM */ static void warn_scan_unevictable_pages(void) { From 245132643e1cfcd145bbc86a716c1818371fcb93 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Fri, 20 Jan 2012 14:34:21 -0800 Subject: [PATCH 731/820] SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins Reviewed-by: KOSAKI Motohiro Cc: Minchan Kim Cc: Rik van Riel Cc: Shaohua Li Cc: Eric Dumazet Cc: Johannes Weiner Cc: Michel Lespinasse Cc: [back to 3.1 but will need respins] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/shmem_fs.h | 1 + include/linux/swap.h | 2 +- ipc/shm.c | 2 +- mm/shmem.c | 46 ++++++++++++-- mm/vmscan.c | 132 ++++++++++++--------------------------- 5 files changed, 85 insertions(+), 98 deletions(-) diff --git a/include/linux/shmem_fs.h b/include/linux/shmem_fs.h index e4c711c6f321..79ab2555b3b0 100644 --- a/include/linux/shmem_fs.h +++ b/include/linux/shmem_fs.h @@ -48,6 +48,7 @@ extern struct file *shmem_file_setup(const char *name, loff_t size, unsigned long flags); extern int shmem_zero_setup(struct vm_area_struct *); extern int shmem_lock(struct file *file, int lock, struct user_struct *user); +extern void shmem_unlock_mapping(struct address_space *mapping); extern struct page *shmem_read_mapping_page_gfp(struct address_space *mapping, pgoff_t index, gfp_t gfp_mask); extern void shmem_truncate_range(struct inode *inode, loff_t start, loff_t end); diff --git a/include/linux/swap.h b/include/linux/swap.h index 06061a7f8e69..3e60228e7299 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -273,7 +273,7 @@ static inline int zone_reclaim(struct zone *z, gfp_t mask, unsigned int order) #endif extern int page_evictable(struct page *page, struct vm_area_struct *vma); -extern void scan_mapping_unevictable_pages(struct address_space *); +extern void check_move_unevictable_pages(struct page **, int nr_pages); extern unsigned long scan_unevictable_pages; extern int scan_unevictable_handler(struct ctl_table *, int, diff --git a/ipc/shm.c b/ipc/shm.c index 854ab58e5f6e..b76be5bda6c2 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -916,7 +916,7 @@ SYSCALL_DEFINE3(shmctl, int, shmid, int, cmd, struct shmid_ds __user *, buf) shp->mlock_user = NULL; get_file(shm_file); shm_unlock(shp); - scan_mapping_unevictable_pages(shm_file->f_mapping); + shmem_unlock_mapping(shm_file->f_mapping); fput(shm_file); goto out; } diff --git a/mm/shmem.c b/mm/shmem.c index 4aaa53abe302..269d049294ab 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -379,7 +379,7 @@ static int shmem_free_swap(struct address_space *mapping, /* * Pagevec may contain swap entries, so shuffle up pages before releasing. */ -static void shmem_pagevec_release(struct pagevec *pvec) +static void shmem_deswap_pagevec(struct pagevec *pvec) { int i, j; @@ -389,7 +389,36 @@ static void shmem_pagevec_release(struct pagevec *pvec) pvec->pages[j++] = page; } pvec->nr = j; - pagevec_release(pvec); +} + +/* + * SysV IPC SHM_UNLOCK restore Unevictable pages to their evictable lists. + */ +void shmem_unlock_mapping(struct address_space *mapping) +{ + struct pagevec pvec; + pgoff_t indices[PAGEVEC_SIZE]; + pgoff_t index = 0; + + pagevec_init(&pvec, 0); + /* + * Minor point, but we might as well stop if someone else SHM_LOCKs it. + */ + while (!mapping_unevictable(mapping)) { + /* + * Avoid pagevec_lookup(): find_get_pages() returns 0 as if it + * has finished, if it hits a row of PAGEVEC_SIZE swap entries. + */ + pvec.nr = shmem_find_get_pages_and_swap(mapping, index, + PAGEVEC_SIZE, pvec.pages, indices); + if (!pvec.nr) + break; + index = indices[pvec.nr - 1] + 1; + shmem_deswap_pagevec(&pvec); + check_move_unevictable_pages(pvec.pages, pvec.nr); + pagevec_release(&pvec); + cond_resched(); + } } /* @@ -440,7 +469,8 @@ void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) } unlock_page(page); } - shmem_pagevec_release(&pvec); + shmem_deswap_pagevec(&pvec); + pagevec_release(&pvec); mem_cgroup_uncharge_end(); cond_resched(); index++; @@ -470,7 +500,8 @@ void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) continue; } if (index == start && indices[0] > end) { - shmem_pagevec_release(&pvec); + shmem_deswap_pagevec(&pvec); + pagevec_release(&pvec); break; } mem_cgroup_uncharge_start(); @@ -494,7 +525,8 @@ void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) } unlock_page(page); } - shmem_pagevec_release(&pvec); + shmem_deswap_pagevec(&pvec); + pagevec_release(&pvec); mem_cgroup_uncharge_end(); index++; } @@ -2438,6 +2470,10 @@ int shmem_lock(struct file *file, int lock, struct user_struct *user) return 0; } +void shmem_unlock_mapping(struct address_space *mapping) +{ +} + void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) { truncate_inode_pages_range(inode->i_mapping, lstart, lend); diff --git a/mm/vmscan.c b/mm/vmscan.c index e097c1026b58..c52b23552659 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -26,7 +26,6 @@ #include /* for try_to_release_page(), buffer_heads_over_limit */ #include -#include #include #include #include @@ -661,7 +660,7 @@ redo: * When racing with an mlock or AS_UNEVICTABLE clearing * (page is unlocked) make sure that if the other thread * does not observe our setting of PG_lru and fails - * isolation/check_move_unevictable_page, + * isolation/check_move_unevictable_pages, * we see PG_mlocked/AS_UNEVICTABLE cleared below and move * the page back to the evictable list. * @@ -3501,106 +3500,57 @@ int page_evictable(struct page *page, struct vm_area_struct *vma) #ifdef CONFIG_SHMEM /** - * check_move_unevictable_page - check page for evictability and move to appropriate zone lru list - * @page: page to check evictability and move to appropriate lru list - * @zone: zone page is in + * check_move_unevictable_pages - check pages for evictability and move to appropriate zone lru list + * @pages: array of pages to check + * @nr_pages: number of pages to check * - * Checks a page for evictability and moves the page to the appropriate - * zone lru list. - * - * Restrictions: zone->lru_lock must be held, page must be on LRU and must - * have PageUnevictable set. + * Checks pages for evictability and moves them to the appropriate lru list. * * This function is only used for SysV IPC SHM_UNLOCK. */ -static void check_move_unevictable_page(struct page *page, struct zone *zone) +void check_move_unevictable_pages(struct page **pages, int nr_pages) { struct lruvec *lruvec; + struct zone *zone = NULL; + int pgscanned = 0; + int pgrescued = 0; + int i; - VM_BUG_ON(PageActive(page)); -retry: - ClearPageUnevictable(page); - if (page_evictable(page, NULL)) { - enum lru_list l = page_lru_base_type(page); + for (i = 0; i < nr_pages; i++) { + struct page *page = pages[i]; + struct zone *pagezone; - __dec_zone_state(zone, NR_UNEVICTABLE); - lruvec = mem_cgroup_lru_move_lists(zone, page, - LRU_UNEVICTABLE, l); - list_move(&page->lru, &lruvec->lists[l]); - __inc_zone_state(zone, NR_INACTIVE_ANON + l); - __count_vm_event(UNEVICTABLE_PGRESCUED); - } else { - /* - * rotate unevictable list - */ - SetPageUnevictable(page); - lruvec = mem_cgroup_lru_move_lists(zone, page, LRU_UNEVICTABLE, - LRU_UNEVICTABLE); - list_move(&page->lru, &lruvec->lists[LRU_UNEVICTABLE]); - if (page_evictable(page, NULL)) - goto retry; - } -} - -/** - * scan_mapping_unevictable_pages - scan an address space for evictable pages - * @mapping: struct address_space to scan for evictable pages - * - * Scan all pages in mapping. Check unevictable pages for - * evictability and move them to the appropriate zone lru list. - * - * This function is only used for SysV IPC SHM_UNLOCK. - */ -void scan_mapping_unevictable_pages(struct address_space *mapping) -{ - pgoff_t next = 0; - pgoff_t end = (i_size_read(mapping->host) + PAGE_CACHE_SIZE - 1) >> - PAGE_CACHE_SHIFT; - struct zone *zone; - struct pagevec pvec; - - if (mapping->nrpages == 0) - return; - - pagevec_init(&pvec, 0); - while (next < end && - pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE)) { - int i; - int pg_scanned = 0; - - zone = NULL; - - for (i = 0; i < pagevec_count(&pvec); i++) { - struct page *page = pvec.pages[i]; - pgoff_t page_index = page->index; - struct zone *pagezone = page_zone(page); - - pg_scanned++; - if (page_index > next) - next = page_index; - next++; - - if (pagezone != zone) { - if (zone) - spin_unlock_irq(&zone->lru_lock); - zone = pagezone; - spin_lock_irq(&zone->lru_lock); - } - - if (PageLRU(page) && PageUnevictable(page)) - check_move_unevictable_page(page, zone); + pgscanned++; + pagezone = page_zone(page); + if (pagezone != zone) { + if (zone) + spin_unlock_irq(&zone->lru_lock); + zone = pagezone; + spin_lock_irq(&zone->lru_lock); } - if (zone) - spin_unlock_irq(&zone->lru_lock); - pagevec_release(&pvec); - count_vm_events(UNEVICTABLE_PGSCANNED, pg_scanned); - cond_resched(); + if (!PageLRU(page) || !PageUnevictable(page)) + continue; + + if (page_evictable(page, NULL)) { + enum lru_list lru = page_lru_base_type(page); + + VM_BUG_ON(PageActive(page)); + ClearPageUnevictable(page); + __dec_zone_state(zone, NR_UNEVICTABLE); + lruvec = mem_cgroup_lru_move_lists(zone, page, + LRU_UNEVICTABLE, lru); + list_move(&page->lru, &lruvec->lists[lru]); + __inc_zone_state(zone, NR_INACTIVE_ANON + lru); + pgrescued++; + } + } + + if (zone) { + __count_vm_events(UNEVICTABLE_PGRESCUED, pgrescued); + __count_vm_events(UNEVICTABLE_PGSCANNED, pgscanned); + spin_unlock_irq(&zone->lru_lock); } -} -#else -void scan_mapping_unevictable_pages(struct address_space *mapping) -{ } #endif /* CONFIG_SHMEM */ From 9f9f1acd713d69fae2af286fbeedc6c8963411c6 Mon Sep 17 00:00:00 2001 From: Konstantin Khlebnikov Date: Fri, 20 Jan 2012 14:34:24 -0800 Subject: [PATCH 732/820] mm: fix rss count leakage during migration Memory migration fills a pte with a migration entry and it doesn't update the rss counters. Then it replaces the migration entry with the new page (or the old one if migration failed). But between these two passes this pte can be unmaped, or a task can fork a child and it will get a copy of this migration entry. Nobody accounts for this in the rss counters. This patch properly adjust rss counters for migration entries in zap_pte_range() and copy_one_pte(). Thus we avoid extra atomic operations on the migration fast-path. Signed-off-by: Konstantin Khlebnikov Cc: Hugh Dickins Cc: KAMEZAWA Hiroyuki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memory.c | 37 ++++++++++++++++++++++++++++--------- 1 file changed, 28 insertions(+), 9 deletions(-) diff --git a/mm/memory.c b/mm/memory.c index 5e30583c2605..fa2f04e0337c 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -878,15 +878,24 @@ copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, } if (likely(!non_swap_entry(entry))) rss[MM_SWAPENTS]++; - else if (is_write_migration_entry(entry) && - is_cow_mapping(vm_flags)) { - /* - * COW mappings require pages in both parent - * and child to be set to read. - */ - make_migration_entry_read(&entry); - pte = swp_entry_to_pte(entry); - set_pte_at(src_mm, addr, src_pte, pte); + else if (is_migration_entry(entry)) { + page = migration_entry_to_page(entry); + + if (PageAnon(page)) + rss[MM_ANONPAGES]++; + else + rss[MM_FILEPAGES]++; + + if (is_write_migration_entry(entry) && + is_cow_mapping(vm_flags)) { + /* + * COW mappings require pages in both + * parent and child to be set to read. + */ + make_migration_entry_read(&entry); + pte = swp_entry_to_pte(entry); + set_pte_at(src_mm, addr, src_pte, pte); + } } } goto out_set_pte; @@ -1191,6 +1200,16 @@ again: if (!non_swap_entry(entry)) rss[MM_SWAPENTS]--; + else if (is_migration_entry(entry)) { + struct page *page; + + page = migration_entry_to_page(entry); + + if (PageAnon(page)) + rss[MM_ANONPAGES]--; + else + rss[MM_FILEPAGES]--; + } if (unlikely(!free_swap_and_cache(entry))) print_bad_pte(vma, addr, ptent, NULL); } From c25a785d6647984505fa165b5cd84cfc9a95970b Mon Sep 17 00:00:00 2001 From: Dan Rosenberg Date: Fri, 20 Jan 2012 14:34:27 -0800 Subject: [PATCH 733/820] score: fix off-by-one index into syscall table If the provided system call number is equal to __NR_syscalls, the current check will pass and a function pointer just after the system call table may be called, since sys_call_table is an array with total size __NR_syscalls. Whether or not this is a security bug depends on what the compiler puts immediately after the system call table. It's likely that this won't do anything bad because there is an additional NULL check on the syscall entry, but if there happens to be a non-NULL value immediately after the system call table, this may result in local privilege escalation. Signed-off-by: Dan Rosenberg Cc: Cc: Chen Liqin Cc: Lennox Wu Cc: Eugene Teo Cc: Arnd Bergmann Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/score/kernel/entry.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/score/kernel/entry.S b/arch/score/kernel/entry.S index 577abba3fac6..83bb96079c43 100644 --- a/arch/score/kernel/entry.S +++ b/arch/score/kernel/entry.S @@ -408,7 +408,7 @@ ENTRY(handle_sys) sw r9, [r0, PT_EPC] cmpi.c r27, __NR_syscalls # check syscall number - bgtu illegal_syscall + bgeu illegal_syscall slli r8, r27, 2 # get syscall routine la r11, sys_call_table From 70c95b00b87f85ce5a15b7899aaa862eacd2364a Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 10:31:54 -0800 Subject: [PATCH 734/820] scripts/kernel-doc: fix fatal error caused by cfg80211.h include/net/cfg80211.h uses __must_check in functions that have kernel-doc notation. This was confusing scripts/kernel-doc, so have scripts/kernel-doc ignore "__must_check". Error(include/net/cfg80211.h:2702): cannot understand prototype: 'struct cfg80211_bss * __must_check cfg80211_inform_bss(...) Signed-off-by: Randy Dunlap Cc: Johannes Berg Signed-off-by: Linus Torvalds --- scripts/kernel-doc | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/scripts/kernel-doc b/scripts/kernel-doc index d793001929cf..9b0c0b8b4ab4 100755 --- a/scripts/kernel-doc +++ b/scripts/kernel-doc @@ -5,7 +5,7 @@ use strict; ## Copyright (c) 1998 Michael Zucchi, All Rights Reserved ## ## Copyright (C) 2000, 1 Tim Waugh ## ## Copyright (C) 2001 Simon Huggins ## -## Copyright (C) 2005-2010 Randy Dunlap ## +## Copyright (C) 2005-2012 Randy Dunlap ## ## ## ## #define enhancements by Armin Kuster ## ## Copyright (c) 2000 MontaVista Software, Inc. ## @@ -1785,6 +1785,7 @@ sub dump_function($$) { $prototype =~ s/__devinit +//; $prototype =~ s/__init +//; $prototype =~ s/__init_or_module +//; + $prototype =~ s/__must_check +//; $prototype =~ s/^#\s*define\s+//; #ak added $prototype =~ s/__attribute__\s*\(\([a-z,]*\)\)//; From 42ae610c1a820ddecb80943d4ccfc936f7772535 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:24 -0800 Subject: [PATCH 735/820] kernel-doc: fix new warnings in auditsc.c Fix new kernel-doc warnings in auditsc.c: Warning(kernel/auditsc.c:1875): No description found for parameter 'success' Warning(kernel/auditsc.c:1875): No description found for parameter 'return_code' Warning(kernel/auditsc.c:1875): Excess function parameter 'pt_regs' description in '__audit_syscall_exit' Signed-off-by: Randy Dunlap Cc: Al Viro Cc: Eric Paris Signed-off-by: Linus Torvalds --- kernel/auditsc.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index caaea6e944f8..af1de0f34eae 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -1863,11 +1863,12 @@ void __audit_syscall_entry(int arch, int major, /** * audit_syscall_exit - deallocate audit context after a system call - * @pt_regs: syscall registers + * @success: success value of the syscall + * @return_code: return value of the syscall * * Tear down after system call. If the audit context has been marked as * auditable (either because of the AUDIT_RECORD_CONTEXT state from - * filtering, or because some other part of the kernel write an audit + * filtering, or because some other part of the kernel wrote an audit * message), then write out the syscall information. In call cases, * free the names stored from getname(). */ From 78d79559f2af1e77034436326aa20f2654074e4c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:28 -0800 Subject: [PATCH 736/820] kernel-doc: fix new warnings in driver-core Fix new kernel-doc warnings: Warning(drivers/base/bus.c:925): No description found for parameter 'key' Warning(drivers/base/bus.c:1241): No description found for parameter 'subsys' Warning(drivers/base/bus.c:1241): No description found for parameter 'groups' Signed-off-by: Randy Dunlap Cc: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/base/bus.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/base/bus.c b/drivers/base/bus.c index 99dc5921e1dd..40fb12288ce2 100644 --- a/drivers/base/bus.c +++ b/drivers/base/bus.c @@ -915,9 +915,10 @@ static BUS_ATTR(uevent, S_IWUSR, NULL, bus_uevent_store); /** * __bus_register - register a driver-core subsystem - * @bus: bus. + * @bus: bus to register + * @key: lockdep class key * - * Once we have that, we registered the bus with the kobject + * Once we have that, we register the bus with the kobject * infrastructure, then register the children subsystems it has: * the devices and drivers that belong to the subsystem. */ @@ -1220,8 +1221,8 @@ static void system_root_device_release(struct device *dev) } /** * subsys_system_register - register a subsystem at /sys/devices/system/ - * @subsys - system subsystem - * @groups - default attributes for the root device + * @subsys: system subsystem + * @groups: default attributes for the root device * * All 'system' subsystems have a /sys/devices/system/ root device * with the name of the subsystem. The root device can carry subsystem- From 6e9292c588894bd39eb2d093013f0aee558ddf0e Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:35 -0800 Subject: [PATCH 737/820] kernel-doc: fix new warnings in pci Fix new kernel-doc warnings: Warning(drivers/pci/pci.c:2811): No description found for parameter 'dev' Warning(drivers/pci/pci.c:2811): Excess function parameter 'pdev' description in 'pci_intx_mask_supported' Warning(drivers/pci/pci.c:2894): No description found for parameter 'dev' Warning(drivers/pci/pci.c:2894): Excess function parameter 'pdev' description in 'pci_check_and_mask_intx' Warning(drivers/pci/pci.c:2908): No description found for parameter 'dev' Warning(drivers/pci/pci.c:2908): Excess function parameter 'pdev' description in 'pci_check_and_unmask_intx' Signed-off-by: Randy Dunlap Cc: Jesse Barnes Signed-off-by: Linus Torvalds --- drivers/pci/pci.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 97fff785e97e..af295bb21d62 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -2802,7 +2802,7 @@ pci_intx(struct pci_dev *pdev, int enable) /** * pci_intx_mask_supported - probe for INTx masking support - * @pdev: the PCI device to operate on + * @dev: the PCI device to operate on * * Check if the device dev support INTx masking via the config space * command word. @@ -2884,7 +2884,7 @@ done: /** * pci_check_and_mask_intx - mask INTx on pending interrupt - * @pdev: the PCI device to operate on + * @dev: the PCI device to operate on * * Check if the device dev has its INTx line asserted, mask it and * return true in that case. False is returned if not interrupt was @@ -2898,7 +2898,7 @@ EXPORT_SYMBOL_GPL(pci_check_and_mask_intx); /** * pci_check_and_mask_intx - unmask INTx of no interrupt is pending - * @pdev: the PCI device to operate on + * @dev: the PCI device to operate on * * Check if the device dev has its INTx line asserted, unmask it if not * and return true. False is returned and the mask remains active if From 5bc75a886353fa5f386c5ce49a93da1756006d8f Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:38 -0800 Subject: [PATCH 738/820] kernel-doc: fix new warning in regulator core Fix new kernel-doc warning: Warning(drivers/regulator/core.c:2741): No description found for parameter 'of_node' Signed-off-by: Randy Dunlap Cc: Liam Girdwood Cc: Mark Brown Signed-off-by: Linus Torvalds --- drivers/regulator/core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index ca86f39a0fdc..b1fa25d6583a 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -2731,6 +2731,7 @@ static void rdev_init_debugfs(struct regulator_dev *rdev) * @dev: struct device for the regulator * @init_data: platform provided init data, passed through by driver * @driver_data: private regulator data + * @of_node: target open firmware device structure (may be NULL) * * Called by regulator drivers to register a regulator. * Returns 0 on success. From b5763accd3b5fc131ee06e26ce56e63ae0322c9b Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:42 -0800 Subject: [PATCH 739/820] kernel-doc: fix new warnings in debugfs Fix new kernel-doc warnings: Warning(fs/debugfs/file.c:556): No description found for parameter 'nregs' Warning(fs/debugfs/file.c:556): Excess function parameter 'mregs' description in 'debugfs_print_regs32' Signed-off-by: Randy Dunlap Cc: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- fs/debugfs/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index f65d4455c5e5..ef023eef0464 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -540,7 +540,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_blob); * debugfs_print_regs32 - use seq_print to describe a set of registers * @s: the seq_file structure being used to generate output * @regs: an array if struct debugfs_reg32 structures - * @mregs: the length of the above array + * @nregs: the length of the above array * @base: the base address to be used in reading the registers * @prefix: a string to be prefixed to every output line * From 2eda013f4894bc200124f791a56c4defb613a0cc Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:51 -0800 Subject: [PATCH 740/820] kernel-doc: fix new warnings in device.h Fix new kernel-doc warnings: Warning(include/linux/device.h:299): No description found for parameter 'name' Warning(include/linux/device.h:299): No description found for parameter 'subsys' Warning(include/linux/device.h:299): No description found for parameter 'node' Warning(include/linux/device.h:299): No description found for parameter 'add_dev' Warning(include/linux/device.h:299): No description found for parameter 'remove_dev' Warning(include/linux/device.h:685): No description found for parameter 'id' Warning(include/linux/device.h:1009): No description found for parameter '__driver' Warning(include/linux/device.h:1009): No description found for parameter '__register' Warning(include/linux/device.h:1009): No description found for parameter '__unregister' Signed-off-by: Randy Dunlap Cc: Lars-Peter Clausen Cc: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- include/linux/device.h | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git a/include/linux/device.h b/include/linux/device.h index 5b3adb8f9588..b63fb393aa58 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -279,11 +279,11 @@ struct device *driver_find_device(struct device_driver *drv, /** * struct subsys_interface - interfaces to device functions - * @name name of the device function - * @subsystem subsytem of the devices to attach to - * @node the list of functions registered at the subsystem - * @add device hookup to device function handler - * @remove device hookup to device function handler + * @name: name of the device function + * @subsys: subsytem of the devices to attach to + * @node: the list of functions registered at the subsystem + * @add_dev: device hookup to device function handler + * @remove_dev: device hookup to device function handler * * Simple interfaces attached to a subsystem. Multiple interfaces can * attach to a subsystem and its devices. Unlike drivers, they do not @@ -612,6 +612,7 @@ struct device_dma_parameters { * @archdata: For arch-specific additions. * @of_node: Associated device tree node. * @devt: For creating the sysfs "dev". + * @id: device instance * @devres_lock: Spinlock to protect the resource of the device. * @devres_head: The resources list of the device. * @knode_class: The node used to add the device to the class list. @@ -1003,6 +1004,10 @@ extern long sysfs_deprecated; * Each module may only use this macro once, and calling it replaces * module_init() and module_exit(). * + * @__driver: driver name + * @__register: register function for this driver type + * @__unregister: unregister function for this driver type + * * Use this macro to construct bus specific macros for registering * drivers, and do not use it on its own. */ From 4d922612df8bd1202a1f51d95b78aca3d67302cd Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:02:56 -0800 Subject: [PATCH 741/820] kernel-doc: fix new warning in usb.h Fix new kernel-doc warning: Warning(include/linux/usb.h:1251): No description found for parameter 'num_mapped_sgs' Signed-off-by: Randy Dunlap Cc: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- include/linux/usb.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/linux/usb.h b/include/linux/usb.h index 27a4e16d2bf1..69d845739bc2 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -1073,6 +1073,7 @@ typedef void (*usb_complete_t)(struct urb *); * which the host controller driver should use in preference to the * transfer_buffer. * @sg: scatter gather buffer list + * @num_mapped_sgs: (internal) number of mapped sg entries * @num_sgs: number of entries in the sg list * @transfer_buffer_length: How big is transfer_buffer. The transfer may * be broken up into chunks according to the current maximum packet From 2f6c76aa5f0fb53059730c628ac59ef62a14ed7a Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:03:00 -0800 Subject: [PATCH 742/820] kernel-doc: fix new warnings in cfg80211.h Fix new kernel-doc warnings: Warning(include/net/cfg80211.h:1165): No description found for parameter 'channel_type' Warning(include/net/cfg80211.h:2090): No description found for parameter 'probe_resp_offload' Signed-off-by: Randy Dunlap Cc: Johannes Berg Cc: linux-wireless@vger.kernel.org Signed-off-by: Linus Torvalds --- include/net/cfg80211.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 15f4be7d768e..a067d30ce73e 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1140,6 +1140,7 @@ struct cfg80211_disassoc_request { * @bssid: Fixed BSSID requested, maybe be %NULL, if set do not * search for IBSSs with a different BSSID. * @channel: The channel to use if no IBSS can be found to join. + * @channel_type: channel type (HT mode) * @channel_fixed: The channel should be fixed -- do not search for * IBSSs to join on other channels. * @ie: information element(s) to include in the beacon @@ -1978,6 +1979,11 @@ struct wiphy_wowlan_support { * configured as RX antennas. Antenna configuration commands will be * rejected unless this or @available_antennas_tx is set. * + * @probe_resp_offload: + * Bitmap of supported protocols for probe response offloading. + * See &enum nl80211_probe_resp_offload_support_attr. Only valid + * when the wiphy flag @WIPHY_FLAG_AP_PROBE_RESP_OFFLOAD is set. + * * @max_remain_on_channel_duration: Maximum time a remain-on-channel operation * may request, if implemented. * From fa757281a08799fd6c0f7ec6f111d1cd66afc97b Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:03:13 -0800 Subject: [PATCH 743/820] kernel-doc: fix kernel-doc warnings in sched Fix new kernel-doc notation warnings: Warning(include/linux/sched.h:2094): No description found for parameter 'p' Warning(include/linux/sched.h:2094): Excess function parameter 'tsk' description in 'is_idle_task' Warning(kernel/sched/cpupri.c:139): No description found for parameter 'newpri' Warning(kernel/sched/cpupri.c:139): Excess function parameter 'pri' description in 'cpupri_set' Warning(kernel/sched/cpupri.c:208): Excess function parameter 'bootmem' description in 'cpupri_init' Signed-off-by: Randy Dunlap Cc: Ingo Molnar Cc: Peter Zijlstra Signed-off-by: Linus Torvalds --- include/linux/sched.h | 2 +- kernel/sched/cpupri.c | 3 +-- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 4032ec1cf836..513f52459872 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2088,7 +2088,7 @@ extern int sched_setscheduler_nocheck(struct task_struct *, int, extern struct task_struct *idle_task(int cpu); /** * is_idle_task - is the specified task an idle task? - * @tsk: the task in question. + * @p: the task in question. */ static inline bool is_idle_task(struct task_struct *p) { diff --git a/kernel/sched/cpupri.c b/kernel/sched/cpupri.c index b0d798eaf130..d72586fdf660 100644 --- a/kernel/sched/cpupri.c +++ b/kernel/sched/cpupri.c @@ -129,7 +129,7 @@ int cpupri_find(struct cpupri *cp, struct task_struct *p, * cpupri_set - update the cpu priority setting * @cp: The cpupri context * @cpu: The target cpu - * @pri: The priority (INVALID-RT99) to assign to this CPU + * @newpri: The priority (INVALID-RT99) to assign to this CPU * * Note: Assumes cpu_rq(cpu)->lock is locked * @@ -200,7 +200,6 @@ void cpupri_set(struct cpupri *cp, int cpu, int newpri) /** * cpupri_init - initialize the cpupri structure * @cp: The cpupri context - * @bootmem: true if allocations need to use bootmem * * Returns: -ENOMEM if memory fails. */ From 73e4a987390aacce102d7d36095a2dd63665f032 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:03:16 -0800 Subject: [PATCH 744/820] docbook: don't use serial_core.h in device-drivers book Fix new kernel-doc warning. This file no longer contains kernel-doc comments. Warning(include/linux/serial_core.h): no structured comments found Signed-off-by: Randy Dunlap Cc: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- Documentation/DocBook/device-drivers.tmpl | 1 - 1 file changed, 1 deletion(-) diff --git a/Documentation/DocBook/device-drivers.tmpl b/Documentation/DocBook/device-drivers.tmpl index b638e50cf8f6..b330b327032a 100644 --- a/Documentation/DocBook/device-drivers.tmpl +++ b/Documentation/DocBook/device-drivers.tmpl @@ -216,7 +216,6 @@ X!Isound/sound_firmware.c 16x50 UART Driver -!Iinclude/linux/serial_core.h !Edrivers/tty/serial/serial_core.c !Edrivers/tty/serial/8250.c From 4e35d28751752859b9985074e02320393b6f1c33 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:03:20 -0800 Subject: [PATCH 745/820] docbook: change iomap source filename in deviceiobook Fix warning since kernel-doc comments were moved to another source file. Warning(lib/iomap.c): no structured comments found Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds --- Documentation/DocBook/deviceiobook.tmpl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/DocBook/deviceiobook.tmpl b/Documentation/DocBook/deviceiobook.tmpl index c1ed6a49e598..54199a0dcf9a 100644 --- a/Documentation/DocBook/deviceiobook.tmpl +++ b/Documentation/DocBook/deviceiobook.tmpl @@ -317,7 +317,7 @@ CPU B: spin_unlock_irqrestore(&dev_lock, flags) Public Functions Provided !Iarch/x86/include/asm/io.h -!Elib/iomap.c +!Elib/pci_iomap.c From b4d20859362fde976bb2fa53eb51b798cdba1afc Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 21 Jan 2012 11:03:24 -0800 Subject: [PATCH 746/820] docbook: fix sched source file names in device-drivers book Fix warning since kernel scheduler functions and kernel-doc notation were moved to other files. docproc: lnx-33-rc1/kernel/sched.c: No such file or directory Signed-off-by: Randy Dunlap Cc: Ingo Molnar Cc: Peter Zijlstra Signed-off-by: Linus Torvalds --- Documentation/DocBook/device-drivers.tmpl | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/Documentation/DocBook/device-drivers.tmpl b/Documentation/DocBook/device-drivers.tmpl index b330b327032a..2f7fd4360848 100644 --- a/Documentation/DocBook/device-drivers.tmpl +++ b/Documentation/DocBook/device-drivers.tmpl @@ -50,7 +50,9 @@ Delaying, scheduling, and timer routines !Iinclude/linux/sched.h -!Ekernel/sched.c +!Ekernel/sched/core.c +!Ikernel/sched/cpupri.c +!Ikernel/sched/fair.c !Iinclude/linux/completion.h !Ekernel/timer.c From b4ead019afc201f71c39cd0dfcaafed4a97b3dd2 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 23 Jan 2012 18:23:36 +0100 Subject: [PATCH 747/820] ALSA: hda - Fix silent outputs from docking-station jacks of Dell laptops The recent change of the power-widget handling for IDT codecs caused the silent output from the docking-station line-out jack. This was partially fixed by the commit f2cbba7602383cd9cdd21f0a5d0b8bd1aad47b33 "ALSA: hda - Fix the lost power-setup of seconary pins after PM resume". But the line-out on the docking-station is still silent when booted with the jack plugged even by this fix. The remainig bug is that the power-widget is set off in stac92xx_init() because the pins in cfg->line_out_pins[] aren't checked there properly but only hp_pins[] are checked in is_nid_hp_pin(). This patch fixes the problem by checking both HP and line-out pins and leaving the power-map correctly. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=42637 Cc: Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_sigmatel.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 1a26dbca9483..336cfcd324f9 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -4163,13 +4163,15 @@ static int enable_pin_detect(struct hda_codec *codec, hda_nid_t nid, return 1; } -static int is_nid_hp_pin(struct auto_pin_cfg *cfg, hda_nid_t nid) +static int is_nid_out_jack_pin(struct auto_pin_cfg *cfg, hda_nid_t nid) { int i; for (i = 0; i < cfg->hp_outs; i++) if (cfg->hp_pins[i] == nid) return 1; /* nid is a HP-Out */ - + for (i = 0; i < cfg->line_outs; i++) + if (cfg->line_out_pins[i] == nid) + return 1; /* nid is a line-Out */ return 0; /* nid is not a HP-Out */ }; @@ -4375,7 +4377,7 @@ static int stac92xx_init(struct hda_codec *codec) continue; } - if (is_nid_hp_pin(cfg, nid)) + if (is_nid_out_jack_pin(cfg, nid)) continue; /* already has an unsol event */ pinctl = snd_hda_codec_read(codec, nid, 0, From 4ec7ac1203bcf21f5e3d977c9818b1a56c9ef40d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 23 Jan 2012 05:38:59 +0000 Subject: [PATCH 748/820] macvlan: fix a possible use after free Commit bc416d9768 (macvlan: handle fragmented multicast frames) added a possible use after free in macvlan_handle_frame(), since ip_check_defrag() uses pskb_may_pull() : skb header can be reallocated. Signed-off-by: Eric Dumazet Cc: Ben Greear Signed-off-by: David S. Miller --- drivers/net/macvlan.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index f2f820c4b40a..9ea99217f116 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -173,6 +173,7 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb) skb = ip_check_defrag(skb, IP_DEFRAG_MACVLAN); if (!skb) return RX_HANDLER_CONSUMED; + eth = eth_hdr(skb); src = macvlan_hash_lookup(port, eth->h_source); if (!src) /* frame comes from an external address */ From 460a25cdaef1a2b6b8e14e371d868aa91b0e72e8 Mon Sep 17 00:00:00 2001 From: Yuval Mintz Date: Mon, 23 Jan 2012 07:31:51 +0000 Subject: [PATCH 749/820] bnx2x: credit-leakage fixup on vlan_mac_del_all Upon insertion of elements into the execution queue, it is validated that there are enough credits to support additional vlan-macs, and the credits are consumed. However, when removing a pending command in `bnx2x_vland_mac_del_all' the consumed credits are not released, which might cause leakage and eventually the inability to add new vlan-macs in certain scenarios. Signed-off-by: Yuval Mintz Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- .../net/ethernet/broadcom/bnx2x/bnx2x_sp.c | 44 ++++++++++++++++++- .../net/ethernet/broadcom/bnx2x/bnx2x_sp.h | 11 +++++ 2 files changed, 54 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.c index 5ac616093f9f..cb6339c35571 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.c @@ -50,6 +50,7 @@ static inline void bnx2x_exe_queue_init(struct bnx2x *bp, int exe_len, union bnx2x_qable_obj *owner, exe_q_validate validate, + exe_q_remove remove, exe_q_optimize optimize, exe_q_execute exec, exe_q_get get) @@ -66,6 +67,7 @@ static inline void bnx2x_exe_queue_init(struct bnx2x *bp, /* Owner specific callbacks */ o->validate = validate; + o->remove = remove; o->optimize = optimize; o->execute = exec; o->get = get; @@ -1340,6 +1342,35 @@ static int bnx2x_validate_vlan_mac(struct bnx2x *bp, } } +static int bnx2x_remove_vlan_mac(struct bnx2x *bp, + union bnx2x_qable_obj *qo, + struct bnx2x_exeq_elem *elem) +{ + int rc = 0; + + /* If consumption wasn't required, nothing to do */ + if (test_bit(BNX2X_DONT_CONSUME_CAM_CREDIT, + &elem->cmd_data.vlan_mac.vlan_mac_flags)) + return 0; + + switch (elem->cmd_data.vlan_mac.cmd) { + case BNX2X_VLAN_MAC_ADD: + case BNX2X_VLAN_MAC_MOVE: + rc = qo->vlan_mac.put_credit(&qo->vlan_mac); + break; + case BNX2X_VLAN_MAC_DEL: + rc = qo->vlan_mac.get_credit(&qo->vlan_mac); + break; + default: + return -EINVAL; + } + + if (rc != true) + return -EINVAL; + + return 0; +} + /** * bnx2x_wait_vlan_mac - passivly wait for 5 seconds until all work completes. * @@ -1801,8 +1832,14 @@ static int bnx2x_vlan_mac_del_all(struct bnx2x *bp, list_for_each_entry_safe(exeq_pos, exeq_pos_n, &exeq->exe_queue, link) { if (exeq_pos->cmd_data.vlan_mac.vlan_mac_flags == - *vlan_mac_flags) + *vlan_mac_flags) { + rc = exeq->remove(bp, exeq->owner, exeq_pos); + if (rc) { + BNX2X_ERR("Failed to remove command\n"); + return rc; + } list_del(&exeq_pos->link); + } } spin_unlock_bh(&exeq->lock); @@ -1908,6 +1945,7 @@ void bnx2x_init_mac_obj(struct bnx2x *bp, bnx2x_exe_queue_init(bp, &mac_obj->exe_queue, 1, qable_obj, bnx2x_validate_vlan_mac, + bnx2x_remove_vlan_mac, bnx2x_optimize_vlan_mac, bnx2x_execute_vlan_mac, bnx2x_exeq_get_mac); @@ -1924,6 +1962,7 @@ void bnx2x_init_mac_obj(struct bnx2x *bp, bnx2x_exe_queue_init(bp, &mac_obj->exe_queue, CLASSIFY_RULES_COUNT, qable_obj, bnx2x_validate_vlan_mac, + bnx2x_remove_vlan_mac, bnx2x_optimize_vlan_mac, bnx2x_execute_vlan_mac, bnx2x_exeq_get_mac); @@ -1963,6 +2002,7 @@ void bnx2x_init_vlan_obj(struct bnx2x *bp, bnx2x_exe_queue_init(bp, &vlan_obj->exe_queue, CLASSIFY_RULES_COUNT, qable_obj, bnx2x_validate_vlan_mac, + bnx2x_remove_vlan_mac, bnx2x_optimize_vlan_mac, bnx2x_execute_vlan_mac, bnx2x_exeq_get_vlan); @@ -2009,6 +2049,7 @@ void bnx2x_init_vlan_mac_obj(struct bnx2x *bp, bnx2x_exe_queue_init(bp, &vlan_mac_obj->exe_queue, 1, qable_obj, bnx2x_validate_vlan_mac, + bnx2x_remove_vlan_mac, bnx2x_optimize_vlan_mac, bnx2x_execute_vlan_mac, bnx2x_exeq_get_vlan_mac); @@ -2025,6 +2066,7 @@ void bnx2x_init_vlan_mac_obj(struct bnx2x *bp, &vlan_mac_obj->exe_queue, CLASSIFY_RULES_COUNT, qable_obj, bnx2x_validate_vlan_mac, + bnx2x_remove_vlan_mac, bnx2x_optimize_vlan_mac, bnx2x_execute_vlan_mac, bnx2x_exeq_get_vlan_mac); diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.h index 992308ff82e8..66da39f0c84a 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.h +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.h @@ -161,6 +161,10 @@ typedef int (*exe_q_validate)(struct bnx2x *bp, union bnx2x_qable_obj *o, struct bnx2x_exeq_elem *elem); +typedef int (*exe_q_remove)(struct bnx2x *bp, + union bnx2x_qable_obj *o, + struct bnx2x_exeq_elem *elem); + /** * @return positive is entry was optimized, 0 - if not, negative * in case of an error. @@ -203,11 +207,18 @@ struct bnx2x_exe_queue_obj { */ exe_q_validate validate; + /** + * Called before removing pending commands, cleaning allocated + * resources (e.g., credits from validate) + */ + exe_q_remove remove; /** * This will try to cancel the current pending commands list * considering the new command. * + * Returns the number of optimized commands or a negative error code + * * Must run under exe_queue->lock */ exe_q_optimize optimize; From d5e836329bd836d24b168004827532426cad2f39 Mon Sep 17 00:00:00 2001 From: Yuval Mintz Date: Mon, 23 Jan 2012 07:31:52 +0000 Subject: [PATCH 750/820] bnx2x: fixed ethtool statistics for MF modes Previosuly, in MF modes `ethtool -S' lacked some of the statistics which appeared in non-MF modes. This has been fixed. Signed-off-by: Yuval Mintz Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- .../ethernet/broadcom/bnx2x/bnx2x_ethtool.c | 108 +++++++----------- 1 file changed, 42 insertions(+), 66 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c index f99c6e312a5d..7d32e0059966 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c @@ -2121,18 +2121,16 @@ static int bnx2x_get_sset_count(struct net_device *dev, int stringset) case ETH_SS_STATS: if (is_multi(bp)) { num_stats = bnx2x_num_stat_queues(bp) * - BNX2X_NUM_Q_STATS; - if (!IS_MF_MODE_STAT(bp)) - num_stats += BNX2X_NUM_STATS; - } else { - if (IS_MF_MODE_STAT(bp)) { - num_stats = 0; - for (i = 0; i < BNX2X_NUM_STATS; i++) - if (IS_FUNC_STAT(i)) - num_stats++; - } else - num_stats = BNX2X_NUM_STATS; - } + BNX2X_NUM_Q_STATS; + } else + num_stats = 0; + if (IS_MF_MODE_STAT(bp)) { + for (i = 0; i < BNX2X_NUM_STATS; i++) + if (IS_FUNC_STAT(i)) + num_stats++; + } else + num_stats += BNX2X_NUM_STATS; + return num_stats; case ETH_SS_TEST: @@ -2151,8 +2149,8 @@ static void bnx2x_get_strings(struct net_device *dev, u32 stringset, u8 *buf) switch (stringset) { case ETH_SS_STATS: + k = 0; if (is_multi(bp)) { - k = 0; for_each_eth_queue(bp, i) { memset(queue_name, 0, sizeof(queue_name)); sprintf(queue_name, "%d", i); @@ -2163,20 +2161,17 @@ static void bnx2x_get_strings(struct net_device *dev, u32 stringset, u8 *buf) queue_name); k += BNX2X_NUM_Q_STATS; } - if (IS_MF_MODE_STAT(bp)) - break; - for (j = 0; j < BNX2X_NUM_STATS; j++) - strcpy(buf + (k + j)*ETH_GSTRING_LEN, - bnx2x_stats_arr[j].string); - } else { - for (i = 0, j = 0; i < BNX2X_NUM_STATS; i++) { - if (IS_MF_MODE_STAT(bp) && IS_PORT_STAT(i)) - continue; - strcpy(buf + j*ETH_GSTRING_LEN, - bnx2x_stats_arr[i].string); - j++; - } } + + + for (i = 0, j = 0; i < BNX2X_NUM_STATS; i++) { + if (IS_MF_MODE_STAT(bp) && IS_PORT_STAT(i)) + continue; + strcpy(buf + (k + j)*ETH_GSTRING_LEN, + bnx2x_stats_arr[i].string); + j++; + } + break; case ETH_SS_TEST: @@ -2190,10 +2185,9 @@ static void bnx2x_get_ethtool_stats(struct net_device *dev, { struct bnx2x *bp = netdev_priv(dev); u32 *hw_stats, *offset; - int i, j, k; + int i, j, k = 0; if (is_multi(bp)) { - k = 0; for_each_eth_queue(bp, i) { hw_stats = (u32 *)&bp->fp[i].eth_q_stats; for (j = 0; j < BNX2X_NUM_Q_STATS; j++) { @@ -2214,46 +2208,28 @@ static void bnx2x_get_ethtool_stats(struct net_device *dev, } k += BNX2X_NUM_Q_STATS; } - if (IS_MF_MODE_STAT(bp)) - return; - hw_stats = (u32 *)&bp->eth_stats; - for (j = 0; j < BNX2X_NUM_STATS; j++) { - if (bnx2x_stats_arr[j].size == 0) { - /* skip this counter */ - buf[k + j] = 0; - continue; - } - offset = (hw_stats + bnx2x_stats_arr[j].offset); - if (bnx2x_stats_arr[j].size == 4) { - /* 4-byte counter */ - buf[k + j] = (u64) *offset; - continue; - } - /* 8-byte counter */ - buf[k + j] = HILO_U64(*offset, *(offset + 1)); - } - } else { - hw_stats = (u32 *)&bp->eth_stats; - for (i = 0, j = 0; i < BNX2X_NUM_STATS; i++) { - if (IS_MF_MODE_STAT(bp) && IS_PORT_STAT(i)) - continue; - if (bnx2x_stats_arr[i].size == 0) { - /* skip this counter */ - buf[j] = 0; - j++; - continue; - } - offset = (hw_stats + bnx2x_stats_arr[i].offset); - if (bnx2x_stats_arr[i].size == 4) { - /* 4-byte counter */ - buf[j] = (u64) *offset; - j++; - continue; - } - /* 8-byte counter */ - buf[j] = HILO_U64(*offset, *(offset + 1)); + } + + hw_stats = (u32 *)&bp->eth_stats; + for (i = 0, j = 0; i < BNX2X_NUM_STATS; i++) { + if (IS_MF_MODE_STAT(bp) && IS_PORT_STAT(i)) + continue; + if (bnx2x_stats_arr[i].size == 0) { + /* skip this counter */ + buf[k + j] = 0; j++; + continue; } + offset = (hw_stats + bnx2x_stats_arr[i].offset); + if (bnx2x_stats_arr[i].size == 4) { + /* 4-byte counter */ + buf[k + j] = (u64) *offset; + j++; + continue; + } + /* 8-byte counter */ + buf[k + j] = HILO_U64(*offset, *(offset + 1)); + j++; } } From b0700b1e6b9556aa99bb9bf6ad6d830ae38344c7 Mon Sep 17 00:00:00 2001 From: Dmitry Kravkov Date: Mon, 23 Jan 2012 07:31:53 +0000 Subject: [PATCH 751/820] bnx2x: fix Big-Endianess in ethtool -t Signed-off-by: Dmitry Kravkov Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c index 7d32e0059966..31a8b38ab15e 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c @@ -1738,7 +1738,7 @@ static int bnx2x_run_loopback(struct bnx2x *bp, int loopback_mode) struct bnx2x_fp_txdata *txdata = &fp_tx->txdata[0]; u16 tx_start_idx, tx_idx; u16 rx_start_idx, rx_idx; - u16 pkt_prod, bd_prod, rx_comp_cons; + u16 pkt_prod, bd_prod; struct sw_tx_bd *tx_buf; struct eth_tx_start_bd *tx_start_bd; struct eth_tx_parse_bd_e1x *pbd_e1x = NULL; @@ -1873,8 +1873,7 @@ static int bnx2x_run_loopback(struct bnx2x *bp, int loopback_mode) if (rx_idx != rx_start_idx + num_pkts) goto test_loopback_exit; - rx_comp_cons = le16_to_cpu(fp_rx->rx_comp_cons); - cqe = &fp_rx->rx_comp_ring[RCQ_BD(rx_comp_cons)]; + cqe = &fp_rx->rx_comp_ring[RCQ_BD(fp_rx->rx_comp_cons)]; cqe_fp_flags = cqe->fast_path_cqe.type_error_flags; cqe_fp_type = cqe_fp_flags & ETH_FAST_PATH_RX_CQE_TYPE; if (!CQE_TYPE_FAST(cqe_fp_type) || (cqe_fp_flags & ETH_RX_ERROR_FALGS)) From 1fdf155158886514c82e5401ab7b1264beb375bf Mon Sep 17 00:00:00 2001 From: Dmitry Kravkov Date: Mon, 23 Jan 2012 07:31:54 +0000 Subject: [PATCH 752/820] bnx2x: allow user to change ring size in ISCSI SD mode Signed-off-by: Dmitry Kravkov Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c index 2b731b253598..03f3935fd8c2 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c @@ -3117,7 +3117,7 @@ static int bnx2x_alloc_fp_mem_at(struct bnx2x *bp, int index) int rx_ring_size = 0; #ifdef BCM_CNIC - if (IS_MF_ISCSI_SD(bp)) { + if (!bp->rx_ring_size && IS_MF_ISCSI_SD(bp)) { rx_ring_size = MIN_RX_SIZE_NONTPA; bp->rx_ring_size = rx_ring_size; } else From 65087cfee50595185f6bbf3d78272eeb34186d2b Mon Sep 17 00:00:00 2001 From: Ariel Elior Date: Mon, 23 Jan 2012 07:31:55 +0000 Subject: [PATCH 753/820] bnx2x: handle CHIP_REVISION during init_one The macro `CHIP_IS_E1x' requires `bp' to be initialized. As `bp' is not yet initialized during this phase of `bnx2x_init_dev', it accessed uninitialized fields in the struct. Signed-off-by: Ariel Elior Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c index ffeaaa95ed96..f4c2fe52ab1c 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c @@ -10536,6 +10536,9 @@ static int __devinit bnx2x_init_dev(struct pci_dev *pdev, { struct bnx2x *bp; int rc; + bool chip_is_e1x = (board_type == BCM57710 || + board_type == BCM57711 || + board_type == BCM57711E); SET_NETDEV_DEV(dev, &pdev->dev); bp = netdev_priv(dev); @@ -10624,7 +10627,7 @@ static int __devinit bnx2x_init_dev(struct pci_dev *pdev, REG_WR(bp, PXP2_REG_PGL_ADDR_90_F0, 0); REG_WR(bp, PXP2_REG_PGL_ADDR_94_F0, 0); - if (CHIP_IS_E1x(bp)) { + if (chip_is_e1x) { REG_WR(bp, PXP2_REG_PGL_ADDR_88_F1, 0); REG_WR(bp, PXP2_REG_PGL_ADDR_8C_F1, 0); REG_WR(bp, PXP2_REG_PGL_ADDR_90_F1, 0); @@ -10635,9 +10638,7 @@ static int __devinit bnx2x_init_dev(struct pci_dev *pdev, * Enable internal target-read (in case we are probed after PF FLR). * Must be done prior to any BAR read access. Only for 57712 and up */ - if (board_type != BCM57710 && - board_type != BCM57711 && - board_type != BCM57711E) + if (!chip_is_e1x) REG_WR(bp, PGLUE_B_REG_INTERNAL_PFID_ENABLE_TARGET_READ, 1); /* Reset the load counter */ From 44151acb9f13563e40d40d14c3e5c11ce21b59e1 Mon Sep 17 00:00:00 2001 From: Yuval Mintz Date: Mon, 23 Jan 2012 07:31:56 +0000 Subject: [PATCH 754/820] bnx2x: fix compilation error with SOE in fw_dump Signed-off-by: Yuval Mintz Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c index f4c2fe52ab1c..1e3f978ee6da 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c @@ -941,7 +941,7 @@ void bnx2x_panic_dump(struct bnx2x *bp) struct sw_rx_bd *sw_bd = &fp->rx_buf_ring[j]; BNX2X_ERR("fp%d: rx_bd[%x]=[%x:%x] sw_bd=[%p]\n", - i, j, rx_bd[1], rx_bd[0], sw_bd->skb); + i, j, rx_bd[1], rx_bd[0], sw_bd->data); } start = RX_SGE(fp->rx_sge_prod); From 302476c99863fe6d08eed6145e37322892ab7f55 Mon Sep 17 00:00:00 2001 From: Paulius Zaleckas Date: Mon, 23 Jan 2012 01:16:35 +0000 Subject: [PATCH 755/820] mv643xx_eth: Add Rx Discard and Rx Overrun statistics These statistics helped me a lot while searching who is losing packets in my setup. I added these stats to MIB group since they are very similar, but just in other registers. I have tested this patch on 88F6281 SoC. Signed-off-by: Paulius Zaleckas Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/mv643xx_eth.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/drivers/net/ethernet/marvell/mv643xx_eth.c b/drivers/net/ethernet/marvell/mv643xx_eth.c index 9c049d2cb97d..9edecfa1f0f4 100644 --- a/drivers/net/ethernet/marvell/mv643xx_eth.c +++ b/drivers/net/ethernet/marvell/mv643xx_eth.c @@ -136,6 +136,8 @@ static char mv643xx_eth_driver_version[] = "1.4"; #define INT_MASK 0x0068 #define INT_MASK_EXT 0x006c #define TX_FIFO_URGENT_THRESHOLD 0x0074 +#define RX_DISCARD_FRAME_CNT 0x0084 +#define RX_OVERRUN_FRAME_CNT 0x0088 #define TXQ_FIX_PRIO_CONF_MOVED 0x00dc #define TX_BW_RATE_MOVED 0x00e0 #define TX_BW_MTU_MOVED 0x00e8 @@ -334,6 +336,9 @@ struct mib_counters { u32 bad_crc_event; u32 collision; u32 late_collision; + /* Non MIB hardware counters */ + u32 rx_discard; + u32 rx_overrun; }; struct lro_counters { @@ -1225,6 +1230,10 @@ static void mib_counters_clear(struct mv643xx_eth_private *mp) for (i = 0; i < 0x80; i += 4) mib_read(mp, i); + + /* Clear non MIB hw counters also */ + rdlp(mp, RX_DISCARD_FRAME_CNT); + rdlp(mp, RX_OVERRUN_FRAME_CNT); } static void mib_counters_update(struct mv643xx_eth_private *mp) @@ -1262,6 +1271,9 @@ static void mib_counters_update(struct mv643xx_eth_private *mp) p->bad_crc_event += mib_read(mp, 0x74); p->collision += mib_read(mp, 0x78); p->late_collision += mib_read(mp, 0x7c); + /* Non MIB hardware counters */ + p->rx_discard += rdlp(mp, RX_DISCARD_FRAME_CNT); + p->rx_overrun += rdlp(mp, RX_OVERRUN_FRAME_CNT); spin_unlock_bh(&mp->mib_counters_lock); mod_timer(&mp->mib_counters_timer, jiffies + 30 * HZ); @@ -1413,6 +1425,8 @@ static const struct mv643xx_eth_stats mv643xx_eth_stats[] = { MIBSTAT(bad_crc_event), MIBSTAT(collision), MIBSTAT(late_collision), + MIBSTAT(rx_discard), + MIBSTAT(rx_overrun), LROSTAT(lro_aggregated), LROSTAT(lro_flushed), LROSTAT(lro_no_desc), From da057fb7d272c7e7609465a54bcac8ec8072ead5 Mon Sep 17 00:00:00 2001 From: stephen hemminger Date: Sun, 22 Jan 2012 09:40:40 +0000 Subject: [PATCH 756/820] skge: add byte queue limit support This also changes the cleanup logic slightly to aggregate completed notifications for multiple packets. Signed-off-by: Stephen Hemminger Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- drivers/net/ethernet/marvell/skge.c | 37 +++++++++++++++++++---------- 1 file changed, 24 insertions(+), 13 deletions(-) diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c index 299c33bd5345..edb9bda55d55 100644 --- a/drivers/net/ethernet/marvell/skge.c +++ b/drivers/net/ethernet/marvell/skge.c @@ -2817,6 +2817,8 @@ static netdev_tx_t skge_xmit_frame(struct sk_buff *skb, td->control = BMU_OWN | BMU_SW | BMU_STF | control | len; wmb(); + netdev_sent_queue(dev, skb->len); + skge_write8(hw, Q_ADDR(txqaddr[skge->port], Q_CSR), CSR_START); netif_printk(skge, tx_queued, KERN_DEBUG, skge->netdev, @@ -2858,11 +2860,9 @@ mapping_error: /* Free resources associated with this reing element */ -static void skge_tx_free(struct skge_port *skge, struct skge_element *e, - u32 control) +static inline void skge_tx_unmap(struct pci_dev *pdev, struct skge_element *e, + u32 control) { - struct pci_dev *pdev = skge->hw->pdev; - /* skb header vs. fragment */ if (control & BMU_STF) pci_unmap_single(pdev, dma_unmap_addr(e, mapaddr), @@ -2872,13 +2872,6 @@ static void skge_tx_free(struct skge_port *skge, struct skge_element *e, pci_unmap_page(pdev, dma_unmap_addr(e, mapaddr), dma_unmap_len(e, maplen), PCI_DMA_TODEVICE); - - if (control & BMU_EOF) { - netif_printk(skge, tx_done, KERN_DEBUG, skge->netdev, - "tx done slot %td\n", e - skge->tx_ring.start); - - dev_kfree_skb(e->skb); - } } /* Free all buffers in transmit ring */ @@ -2889,10 +2882,15 @@ static void skge_tx_clean(struct net_device *dev) for (e = skge->tx_ring.to_clean; e != skge->tx_ring.to_use; e = e->next) { struct skge_tx_desc *td = e->desc; - skge_tx_free(skge, e, td->control); + + skge_tx_unmap(skge->hw->pdev, e, td->control); + + if (td->control & BMU_EOF) + dev_kfree_skb(e->skb); td->control = 0; } + netdev_reset_queue(dev); skge->tx_ring.to_clean = e; } @@ -3157,6 +3155,7 @@ static void skge_tx_done(struct net_device *dev) struct skge_port *skge = netdev_priv(dev); struct skge_ring *ring = &skge->tx_ring; struct skge_element *e; + unsigned int bytes_compl = 0, pkts_compl = 0; skge_write8(skge->hw, Q_ADDR(txqaddr[skge->port], Q_CSR), CSR_IRQ_CL_F); @@ -3166,8 +3165,20 @@ static void skge_tx_done(struct net_device *dev) if (control & BMU_OWN) break; - skge_tx_free(skge, e, control); + skge_tx_unmap(skge->hw->pdev, e, control); + + if (control & BMU_EOF) { + netif_printk(skge, tx_done, KERN_DEBUG, skge->netdev, + "tx done slot %td\n", + e - skge->tx_ring.start); + + pkts_compl++; + bytes_compl += e->skb->len; + + dev_kfree_skb(e->skb); + } } + netdev_completed_queue(dev, pkts_compl, bytes_compl); skge->tx_ring.to_clean = e; /* Can run lockless until we need to synchronize to restart queue. */ From a5a1195559f2e20bd975f58e50f53ebe84d5cca6 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 23 Jan 2012 01:22:09 +0000 Subject: [PATCH 757/820] tg3: fix ipv6 header length computation tg3_start_xmit() makes the wrong assumption for TSOV6 that skb->head doesnt include any payload data. if (skb_is_gso_v6(skb)) hdr_len = skb_headlen(skb) - ETH_HLEN; This is not true anymore after commit f07d960df3 (tcp: avoid frag allocation for small frames) We should instead use : skb_transport_offset(skb) + tcp_hdrlen(skb) Its also true for IPv4 Signed-off-by: Eric Dumazet CC: Matt Carlson CC: Michael Chan Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/tg3.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index d529af99157d..a1f2e0fed78b 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -6667,14 +6667,9 @@ static netdev_tx_t tg3_start_xmit(struct sk_buff *skb, struct net_device *dev) iph = ip_hdr(skb); tcp_opt_len = tcp_optlen(skb); - if (skb_is_gso_v6(skb)) { - hdr_len = skb_headlen(skb) - ETH_HLEN; - } else { - u32 ip_tcp_len; - - ip_tcp_len = ip_hdrlen(skb) + sizeof(struct tcphdr); - hdr_len = ip_tcp_len + tcp_opt_len; + hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb) - ETH_HLEN; + if (!skb_is_gso_v6(skb)) { iph->check = 0; iph->tot_len = htons(mss + hdr_len); } From edd664bbba53f771d4a6d4559ed6e1ff48b46406 Mon Sep 17 00:00:00 2001 From: Chris Healy Date: Sun, 22 Jan 2012 21:20:54 +0000 Subject: [PATCH 758/820] dsa: Add reporting of silicon revision for Marvell 88E6123/88E6161/88E6165 switches. Add reporting of silicon revision during the probe function for Marvell 88E6123/88E6161/88E6165 switches. Signed-off-by: Chris Healy Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6123_61_65.c | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/drivers/net/dsa/mv88e6123_61_65.c b/drivers/net/dsa/mv88e6123_61_65.c index c0a458fc698f..6f23c9521f0e 100644 --- a/drivers/net/dsa/mv88e6123_61_65.c +++ b/drivers/net/dsa/mv88e6123_61_65.c @@ -20,12 +20,25 @@ static char *mv88e6123_61_65_probe(struct mii_bus *bus, int sw_addr) ret = __mv88e6xxx_reg_read(bus, sw_addr, REG_PORT(0), 0x03); if (ret >= 0) { - ret &= 0xfff0; - if (ret == 0x1210) + if (ret == 0x1212) + return "Marvell 88E6123 (A1)"; + if (ret == 0x1213) + return "Marvell 88E6123 (A2)"; + if ((ret & 0xfff0) == 0x1210) return "Marvell 88E6123"; - if (ret == 0x1610) + + if (ret == 0x1612) + return "Marvell 88E6161 (A1)"; + if (ret == 0x1613) + return "Marvell 88E6161 (A2)"; + if ((ret & 0xfff0) == 0x1610) return "Marvell 88E6161"; - if (ret == 0x1650) + + if (ret == 0x1652) + return "Marvell 88E6165 (A1)"; + if (ret == 0x1653) + return "Marvell 88e6165 (A2)"; + if ((ret & 0xfff0) == 0x1650) return "Marvell 88E6165"; } From 9bf31efa84c898a0cf294bacdfe8edcac24e6318 Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Wed, 18 Jan 2012 13:57:33 -0300 Subject: [PATCH 759/820] [media] cxd2820r: fix dvb_frontend_ops Fix bug introduced by multi-frontend to single-frontend change. * Add missing DVB-C caps * Change frontend name as single frontend does all the standards Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/cxd2820r_core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/media/dvb/frontends/cxd2820r_core.c b/drivers/media/dvb/frontends/cxd2820r_core.c index caae7f79c837..5fe591d226f5 100644 --- a/drivers/media/dvb/frontends/cxd2820r_core.c +++ b/drivers/media/dvb/frontends/cxd2820r_core.c @@ -562,7 +562,7 @@ static const struct dvb_frontend_ops cxd2820r_ops = { .delsys = { SYS_DVBT, SYS_DVBT2, SYS_DVBC_ANNEX_A }, /* default: DVB-T/T2 */ .info = { - .name = "Sony CXD2820R (DVB-T/T2)", + .name = "Sony CXD2820R", .caps = FE_CAN_FEC_1_2 | FE_CAN_FEC_2_3 | @@ -572,7 +572,9 @@ static const struct dvb_frontend_ops cxd2820r_ops = { FE_CAN_FEC_AUTO | FE_CAN_QPSK | FE_CAN_QAM_16 | + FE_CAN_QAM_32 | FE_CAN_QAM_64 | + FE_CAN_QAM_128 | FE_CAN_QAM_256 | FE_CAN_QAM_AUTO | FE_CAN_TRANSMISSION_MODE_AUTO | From c2bbbe7b5e79974c5ed1c828690731f6f5106bee Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Thu, 19 Jan 2012 14:46:43 -0300 Subject: [PATCH 760/820] [media] cxd2820r: remove unused parameter from cxd2820r_attach Fix bug introduced by multi-frontend to single-frontend change. This parameter is no longer used after multi-frontend to single-frontend change. Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/anysee.c | 3 +-- drivers/media/dvb/frontends/cxd2820r.h | 6 ++---- drivers/media/dvb/frontends/cxd2820r_core.c | 3 +-- drivers/media/video/em28xx/em28xx-dvb.c | 3 +-- 4 files changed, 5 insertions(+), 10 deletions(-) diff --git a/drivers/media/dvb/dvb-usb/anysee.c b/drivers/media/dvb/dvb-usb/anysee.c index ecc3addc77ec..7b77c7b29f65 100644 --- a/drivers/media/dvb/dvb-usb/anysee.c +++ b/drivers/media/dvb/dvb-usb/anysee.c @@ -887,8 +887,7 @@ static int anysee_frontend_attach(struct dvb_usb_adapter *adap) /* attach demod */ adap->fe_adap[state->fe_id].fe = dvb_attach(cxd2820r_attach, - &anysee_cxd2820r_config, &adap->dev->i2c_adap, - NULL); + &anysee_cxd2820r_config, &adap->dev->i2c_adap); state->has_ci = true; diff --git a/drivers/media/dvb/frontends/cxd2820r.h b/drivers/media/dvb/frontends/cxd2820r.h index cf0f546aa1d1..5aa306ebb7ef 100644 --- a/drivers/media/dvb/frontends/cxd2820r.h +++ b/drivers/media/dvb/frontends/cxd2820r.h @@ -77,14 +77,12 @@ struct cxd2820r_config { (defined(CONFIG_DVB_CXD2820R_MODULE) && defined(MODULE)) extern struct dvb_frontend *cxd2820r_attach( const struct cxd2820r_config *config, - struct i2c_adapter *i2c, - struct dvb_frontend *fe + struct i2c_adapter *i2c ); #else static inline struct dvb_frontend *cxd2820r_attach( const struct cxd2820r_config *config, - struct i2c_adapter *i2c, - struct dvb_frontend *fe + struct i2c_adapter *i2c ) { printk(KERN_WARNING "%s: driver disabled by Kconfig\n", __func__); diff --git a/drivers/media/dvb/frontends/cxd2820r_core.c b/drivers/media/dvb/frontends/cxd2820r_core.c index 5fe591d226f5..bdfa207a883e 100644 --- a/drivers/media/dvb/frontends/cxd2820r_core.c +++ b/drivers/media/dvb/frontends/cxd2820r_core.c @@ -604,8 +604,7 @@ static const struct dvb_frontend_ops cxd2820r_ops = { }; struct dvb_frontend *cxd2820r_attach(const struct cxd2820r_config *cfg, - struct i2c_adapter *i2c, - struct dvb_frontend *fe) + struct i2c_adapter *i2c) { struct cxd2820r_priv *priv = NULL; int ret; diff --git a/drivers/media/video/em28xx/em28xx-dvb.c b/drivers/media/video/em28xx/em28xx-dvb.c index 9449423098e0..aabbf4854f66 100644 --- a/drivers/media/video/em28xx/em28xx-dvb.c +++ b/drivers/media/video/em28xx/em28xx-dvb.c @@ -853,8 +853,7 @@ static int em28xx_dvb_init(struct em28xx *dev) case EM28174_BOARD_PCTV_290E: dvb->fe[0] = dvb_attach(cxd2820r_attach, &em28xx_cxd2820r_config, - &dev->i2c_adap, - NULL); + &dev->i2c_adap); if (dvb->fe[0]) { /* FE 0 attach tuner */ if (!dvb_attach(tda18271_attach, From 46de20a78ae4b122b79fc02633e9a6c3d539ecad Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Fri, 20 Jan 2012 17:39:17 -0300 Subject: [PATCH 761/820] [media] anysee: fix CI init No more error that error seen when device is plugged: dvb_ca adapter 0: Invalid PC card inserted :( Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/anysee.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/media/dvb/dvb-usb/anysee.c b/drivers/media/dvb/dvb-usb/anysee.c index 7b77c7b29f65..fdee856acab0 100644 --- a/drivers/media/dvb/dvb-usb/anysee.c +++ b/drivers/media/dvb/dvb-usb/anysee.c @@ -1188,6 +1188,14 @@ static int anysee_ci_init(struct dvb_usb_device *d) if (ret) return ret; + ret = anysee_wr_reg_mask(d, REG_IOD, (0 << 2)|(0 << 1)|(0 << 0), 0x07); + if (ret) + return ret; + + ret = anysee_wr_reg_mask(d, REG_IOD, (1 << 2)|(1 << 1)|(1 << 0), 0x07); + if (ret) + return ret; + ret = dvb_ca_en50221_init(&d->adapter[0].dvb_adap, &state->ci, 0, 1); if (ret) return ret; From 72565224609a23a60d10fcdf42f87a2fa8f7b16d Mon Sep 17 00:00:00 2001 From: Antti Palosaari Date: Fri, 20 Jan 2012 19:48:28 -0300 Subject: [PATCH 762/820] [media] cxd2820r: sleep on DVB-T/T2 delivery system switch Fix bug introduced by multi-frontend to single-frontend change. It is safer to put DVB-T parts sleeping when auto-switching to DVB-T2 and vice versa. That was original behaviour. Signed-off-by: Antti Palosaari Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/frontends/cxd2820r_core.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/drivers/media/dvb/frontends/cxd2820r_core.c b/drivers/media/dvb/frontends/cxd2820r_core.c index bdfa207a883e..5c7c2aaf9bf5 100644 --- a/drivers/media/dvb/frontends/cxd2820r_core.c +++ b/drivers/media/dvb/frontends/cxd2820r_core.c @@ -482,10 +482,19 @@ static enum dvbfe_search cxd2820r_search(struct dvb_frontend *fe) /* switch between DVB-T and DVB-T2 when tune fails */ if (priv->last_tune_failed) { - if (priv->delivery_system == SYS_DVBT) + if (priv->delivery_system == SYS_DVBT) { + ret = cxd2820r_sleep_t(fe); + if (ret) + goto error; + c->delivery_system = SYS_DVBT2; - else if (priv->delivery_system == SYS_DVBT2) + } else if (priv->delivery_system == SYS_DVBT2) { + ret = cxd2820r_sleep_t2(fe); + if (ret) + goto error; + c->delivery_system = SYS_DVBT; + } } /* set frontend */ From 03652e0ad4b140523ec5ef7fec8d2b3c7218447b Mon Sep 17 00:00:00 2001 From: Josh Wu Date: Wed, 11 Jan 2012 00:58:29 -0300 Subject: [PATCH 763/820] [media] V4L: atmel-isi: add clk_prepare()/clk_unprepare() functions Signed-off-by: Josh Wu Acked-by: Nicolas Ferre Signed-off-by: Guennadi Liakhovetski Signed-off-by: Mauro Carvalho Chehab --- drivers/media/video/atmel-isi.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/drivers/media/video/atmel-isi.c b/drivers/media/video/atmel-isi.c index 9fe4519176a4..ec3f6a06f9c3 100644 --- a/drivers/media/video/atmel-isi.c +++ b/drivers/media/video/atmel-isi.c @@ -922,7 +922,9 @@ static int __devexit atmel_isi_remove(struct platform_device *pdev) isi->fb_descriptors_phys); iounmap(isi->regs); + clk_unprepare(isi->mck); clk_put(isi->mck); + clk_unprepare(isi->pclk); clk_put(isi->pclk); kfree(isi); @@ -955,6 +957,10 @@ static int __devinit atmel_isi_probe(struct platform_device *pdev) if (IS_ERR(pclk)) return PTR_ERR(pclk); + ret = clk_prepare(pclk); + if (ret) + goto err_clk_prepare_pclk; + isi = kzalloc(sizeof(struct atmel_isi), GFP_KERNEL); if (!isi) { ret = -ENOMEM; @@ -978,6 +984,10 @@ static int __devinit atmel_isi_probe(struct platform_device *pdev) goto err_clk_get; } + ret = clk_prepare(isi->mck); + if (ret) + goto err_clk_prepare_mck; + /* Set ISI_MCK's frequency, it should be faster than pixel clock */ ret = clk_set_rate(isi->mck, pdata->mck_hz); if (ret < 0) @@ -1059,10 +1069,14 @@ err_alloc_ctx: isi->fb_descriptors_phys); err_alloc_descriptors: err_set_mck_rate: + clk_unprepare(isi->mck); +err_clk_prepare_mck: clk_put(isi->mck); err_clk_get: kfree(isi); err_alloc_isi: + clk_unprepare(pclk); +err_clk_prepare_pclk: clk_put(pclk); return ret; From c79eba92406acc4898adcd1689fc21a6aa91ed0b Mon Sep 17 00:00:00 2001 From: Mauro Carvalho Chehab Date: Mon, 23 Jan 2012 13:15:22 -0200 Subject: [PATCH 764/820] [media] cinergyT2-fe: Fix bandwdith settings Changeset 7830bbaff9f mangled the bandwidth field for CinergyT2. Properly fill it. Signed-off-by: Mauro Carvalho Chehab --- drivers/media/dvb/dvb-usb/cinergyT2-fe.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/media/dvb/dvb-usb/cinergyT2-fe.c b/drivers/media/dvb/dvb-usb/cinergyT2-fe.c index 8a57ed8272de..1efc028a76c9 100644 --- a/drivers/media/dvb/dvb-usb/cinergyT2-fe.c +++ b/drivers/media/dvb/dvb-usb/cinergyT2-fe.c @@ -276,14 +276,15 @@ static int cinergyt2_fe_set_frontend(struct dvb_frontend *fe) param.flags = 0; switch (fep->bandwidth_hz) { + default: case 8000000: - param.bandwidth = 0; + param.bandwidth = 8; break; case 7000000: - param.bandwidth = 1; + param.bandwidth = 7; break; case 6000000: - param.bandwidth = 2; + param.bandwidth = 6; break; } From e4c89a508f4385a0cd8681c2749a2cd2fa476e40 Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Mon, 23 Jan 2012 21:59:08 +0100 Subject: [PATCH 765/820] PM / Sleep: Fix read_unlock_usermodehelper() call. Commit b298d289 "PM / Sleep: Fix freezer failures due to racy usermodehelper_is_disabled()" added read_unlock_usermodehelper() but read_unlock_usermodehelper() is called without read_lock_usermodehelper() when kmalloc() failed. Signed-off-by: Tetsuo Handa Acked-by: Srivatsa S. Bhat Signed-off-by: Rafael J. Wysocki --- drivers/base/firmware_class.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c index 26ab358dac62..6c9387d646ec 100644 --- a/drivers/base/firmware_class.c +++ b/drivers/base/firmware_class.c @@ -525,8 +525,7 @@ static int _request_firmware(const struct firmware **firmware_p, if (!firmware) { dev_err(device, "%s: kmalloc(struct firmware) failed\n", __func__); - retval = -ENOMEM; - goto out; + return -ENOMEM; } if (fw_get_builtin_firmware(firmware, name)) { From 7edf1a4f27f44588d69cbde955651990090eb25d Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Mon, 23 Jan 2012 21:15:48 +0100 Subject: [PATCH 766/820] ASoC: wm8958: Use correct format string in dev_err() call To print a value of type size_t one should use %zd, not %d. Signed-off-by: Jesper Juhl Signed-off-by: Mark Brown --- sound/soc/codecs/wm8958-dsp2.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/wm8958-dsp2.c b/sound/soc/codecs/wm8958-dsp2.c index 8d4ea43d40a3..40ac888faf3d 100644 --- a/sound/soc/codecs/wm8958-dsp2.c +++ b/sound/soc/codecs/wm8958-dsp2.c @@ -55,7 +55,7 @@ static int wm8958_dsp2_fw(struct snd_soc_codec *codec, const char *name, return 0; if (fw->size < 32) { - dev_err(codec->dev, "%s: firmware too short (%d bytes)\n", + dev_err(codec->dev, "%s: firmware too short (%zd bytes)\n", name, fw->size); goto err; } From 4c40aed869a200a621b53bcb491c5ee8a34ef5f1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Niklas=20S=C3=B6derlund?= Date: Mon, 9 Jan 2012 16:49:44 +0100 Subject: [PATCH 767/820] ACPI, APEI, EINJ Allow empty Trigger Error Action Table MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit According to the ACPI spec [1] section 18.6.4 the TRIGGER_ERROR action table can consists of zero elements. [1] Advanced Configuration and Power Interface Specification Revision 5.0, December 6, 2011 http://www.acpi.info/DOWNLOADS/ACPIspec50.pdf Signed-off-by: Niklas Söderlund Acked-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/einj.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index 5b898d4dda99..bb20b03ddc39 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -276,7 +276,7 @@ static int einj_check_trigger_header(struct acpi_einj_trigger *trigger_tab) if (trigger_tab->header_size != sizeof(struct acpi_einj_trigger)) return -EINVAL; if (trigger_tab->table_size > PAGE_SIZE || - trigger_tab->table_size <= trigger_tab->header_size) + trigger_tab->table_size < trigger_tab->header_size) return -EINVAL; if (trigger_tab->entry_count != (trigger_tab->table_size - trigger_tab->header_size) / @@ -340,6 +340,11 @@ static int __einj_error_trigger(u64 trigger_paddr, u32 type, "The trigger error action table is invalid\n"); goto out_rel_header; } + + /* No action structures in the TRIGGER_ERROR table, nothing to do */ + if (!trigger_tab->entry_count) + goto out_rel_header; + rc = -EIO; table_size = trigger_tab->table_size; r = request_mem_region(trigger_paddr + sizeof(*trigger_tab), From 29924b9f8f39e37275cff93740835d28b9e6fb36 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 20 Jan 2012 10:57:14 +0300 Subject: [PATCH 768/820] ACPI, APEI, EINJ, cleanup 0 vs NULL confusion This function is returning pointers. Sparse complains here: drivers/acpi/apei/einj.c:262:32: warning: Using plain integer as NULL pointer Signed-off-by: Dan Carpenter Acked-by: Huang Ying Signed-off-by: Len Brown --- drivers/acpi/apei/einj.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index bb20b03ddc39..c89b0e5a2293 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -259,15 +259,15 @@ static void *einj_get_parameter_address(void) v4param = ioremap(paddrv4, sizeof(*v4param)); if (!v4param) - return 0; + return NULL; if (readq(&v4param->reserved1) || readq(&v4param->reserved2)) { iounmap(v4param); - return 0; + return NULL; } return v4param; } - return 0; + return NULL; } /* do sanity check to trigger table */ From 459413db33d6c99fc13d60f88899fb72d2530ed3 Mon Sep 17 00:00:00 2001 From: "Luck, Tony" Date: Mon, 23 Jan 2012 15:27:56 -0800 Subject: [PATCH 769/820] Use acpi_os_map_memory() instead of ioremap() in einj driver ioremap() has become more picky and is now spitting out console messages like: ioremap error for 0xbddbd000-0xbddbe000, requested 0x10, got 0x0 when loading the einj driver. What we are trying to so here is map a couple of data structures that the EINJ table points to. Perhaps acpi_os_map_memory() is a better tool for this? Most importantly it works, but as a side benefit it maps the structures into kernel virtual space so we can access them with normal C memory dereferences, so instead of using: writel(param1, &v5param->apicid); we can use the more natural: v5param->apicid = param1; Signed-off-by: Tony Luck Signed-off-by: Len Brown --- drivers/acpi/apei/einj.c | 82 +++++++++++++++++++--------------------- 1 file changed, 38 insertions(+), 44 deletions(-) diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c index c89b0e5a2293..4ca087dd5f4f 100644 --- a/drivers/acpi/apei/einj.c +++ b/drivers/acpi/apei/einj.c @@ -141,21 +141,6 @@ static DEFINE_MUTEX(einj_mutex); static void *einj_param; -#ifndef readq -static inline __u64 readq(volatile void __iomem *addr) -{ - return ((__u64)readl(addr+4) << 32) + readl(addr); -} -#endif - -#ifndef writeq -static inline void writeq(__u64 val, volatile void __iomem *addr) -{ - writel(val, addr); - writel(val >> 32, addr+4); -} -#endif - static void einj_exec_ctx_init(struct apei_exec_context *ctx) { apei_exec_ctx_init(ctx, einj_ins_type, ARRAY_SIZE(einj_ins_type), @@ -204,22 +189,21 @@ static int einj_timedout(u64 *t) static void check_vendor_extension(u64 paddr, struct set_error_type_with_address *v5param) { - int offset = readl(&v5param->vendor_extension); + int offset = v5param->vendor_extension; struct vendor_error_type_extension *v; u32 sbdf; if (!offset) return; - v = ioremap(paddr + offset, sizeof(*v)); + v = acpi_os_map_memory(paddr + offset, sizeof(*v)); if (!v) return; - sbdf = readl(&v->pcie_sbdf); + sbdf = v->pcie_sbdf; sprintf(vendor_dev, "%x:%x:%x.%x vendor_id=%x device_id=%x rev_id=%x\n", sbdf >> 24, (sbdf >> 16) & 0xff, (sbdf >> 11) & 0x1f, (sbdf >> 8) & 0x7, - readw(&v->vendor_id), readw(&v->device_id), - readb(&v->rev_id)); - iounmap(v); + v->vendor_id, v->device_id, v->rev_id); + acpi_os_unmap_memory(v, sizeof(*v)); } static void *einj_get_parameter_address(void) @@ -247,7 +231,7 @@ static void *einj_get_parameter_address(void) if (paddrv5) { struct set_error_type_with_address *v5param; - v5param = ioremap(paddrv5, sizeof(*v5param)); + v5param = acpi_os_map_memory(paddrv5, sizeof(*v5param)); if (v5param) { acpi5 = 1; check_vendor_extension(paddrv5, v5param); @@ -257,11 +241,11 @@ static void *einj_get_parameter_address(void) if (paddrv4) { struct einj_parameter *v4param; - v4param = ioremap(paddrv4, sizeof(*v4param)); + v4param = acpi_os_map_memory(paddrv4, sizeof(*v4param)); if (!v4param) return NULL; - if (readq(&v4param->reserved1) || readq(&v4param->reserved2)) { - iounmap(v4param); + if (v4param->reserved1 || v4param->reserved2) { + acpi_os_unmap_memory(v4param, sizeof(*v4param)); return NULL; } return v4param; @@ -440,41 +424,41 @@ static int __einj_error_inject(u32 type, u64 param1, u64 param2) if (acpi5) { struct set_error_type_with_address *v5param = einj_param; - writel(type, &v5param->type); + v5param->type = type; if (type & 0x80000000) { switch (vendor_flags) { case SETWA_FLAGS_APICID: - writel(param1, &v5param->apicid); + v5param->apicid = param1; break; case SETWA_FLAGS_MEM: - writeq(param1, &v5param->memory_address); - writeq(param2, &v5param->memory_address_range); + v5param->memory_address = param1; + v5param->memory_address_range = param2; break; case SETWA_FLAGS_PCIE_SBDF: - writel(param1, &v5param->pcie_sbdf); + v5param->pcie_sbdf = param1; break; } - writel(vendor_flags, &v5param->flags); + v5param->flags = vendor_flags; } else { switch (type) { case ACPI_EINJ_PROCESSOR_CORRECTABLE: case ACPI_EINJ_PROCESSOR_UNCORRECTABLE: case ACPI_EINJ_PROCESSOR_FATAL: - writel(param1, &v5param->apicid); - writel(SETWA_FLAGS_APICID, &v5param->flags); + v5param->apicid = param1; + v5param->flags = SETWA_FLAGS_APICID; break; case ACPI_EINJ_MEMORY_CORRECTABLE: case ACPI_EINJ_MEMORY_UNCORRECTABLE: case ACPI_EINJ_MEMORY_FATAL: - writeq(param1, &v5param->memory_address); - writeq(param2, &v5param->memory_address_range); - writel(SETWA_FLAGS_MEM, &v5param->flags); + v5param->memory_address = param1; + v5param->memory_address_range = param2; + v5param->flags = SETWA_FLAGS_MEM; break; case ACPI_EINJ_PCIX_CORRECTABLE: case ACPI_EINJ_PCIX_UNCORRECTABLE: case ACPI_EINJ_PCIX_FATAL: - writel(param1, &v5param->pcie_sbdf); - writel(SETWA_FLAGS_PCIE_SBDF, &v5param->flags); + v5param->pcie_sbdf = param1; + v5param->flags = SETWA_FLAGS_PCIE_SBDF; break; } } @@ -484,8 +468,8 @@ static int __einj_error_inject(u32 type, u64 param1, u64 param2) return rc; if (einj_param) { struct einj_parameter *v4param = einj_param; - writeq(param1, &v4param->param1); - writeq(param2, &v4param->param2); + v4param->param1 = param1; + v4param->param2 = param2; } } rc = apei_exec_run(&ctx, ACPI_EINJ_EXECUTE_OPERATION); @@ -736,8 +720,13 @@ static int __init einj_init(void) return 0; err_unmap: - if (einj_param) - iounmap(einj_param); + if (einj_param) { + acpi_size size = (acpi5) ? + sizeof(struct set_error_type_with_address) : + sizeof(struct einj_parameter); + + acpi_os_unmap_memory(einj_param, size); + } apei_exec_post_unmap_gars(&ctx); err_release: apei_resources_release(&einj_resources); @@ -753,8 +742,13 @@ static void __exit einj_exit(void) { struct apei_exec_context ctx; - if (einj_param) - iounmap(einj_param); + if (einj_param) { + acpi_size size = (acpi5) ? + sizeof(struct set_error_type_with_address) : + sizeof(struct einj_parameter); + + acpi_os_unmap_memory(einj_param, size); + } einj_exec_ctx_init(&ctx); apei_exec_post_unmap_gars(&ctx); apei_resources_release(&einj_resources); From c1aab02dac690af7ff634d8e1cb3be6a04387eef Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Tue, 24 Jan 2012 11:41:32 +1100 Subject: [PATCH 770/820] migrate_mode.h is not exported to user mode so move its include into fs.h inside the __KERNEL__ protection. Signed-off-by: Stephen Rothwell Signed-off-by: Linus Torvalds --- include/linux/fs.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/fs.h b/include/linux/fs.h index 4b3a41fe22bf..386da09f229d 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -10,7 +10,6 @@ #include #include #include -#include /* * It's silly to have NR_OPEN bigger than NR_FILE, but you can change @@ -397,6 +396,7 @@ struct inodes_stat_t { #include #include #include +#include #include From c83f1d7e71625801c72f4013291194e09b6f0a6e Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Mon, 23 Jan 2012 22:28:44 +0100 Subject: [PATCH 771/820] ASoC: wm2000: Fix use-after-free - don't release_firmware() twice on error In wm2000_i2c_probe(), if we take the true branch in " ret = snd_soc_register_codec(&i2c->dev, &soc_codec_dev_wm2000, NULL, 0); if (ret != 0) goto err_fw; " then we'll release_firmware(fw) at the 'err_fw' label. But we've already done that just a few lines above. That's a use-after-free bug. This patch restructures the code so that we always call release_firmware(fw) before leaving the function, but only ever call it once. This means that we have to initialize 'fw' to NULL since some paths may now end up calling it without having called request_firmware(), but since request_firmware() deals gracefully with NULL pointers, we are fine if we just NULL initialize it. Signed-off-by: Jesper Juhl Signed-off-by: Mark Brown --- sound/soc/codecs/wm2000.c | 31 +++++++++++++------------------ 1 file changed, 13 insertions(+), 18 deletions(-) diff --git a/sound/soc/codecs/wm2000.c b/sound/soc/codecs/wm2000.c index c2880907fced..a75c3766aede 100644 --- a/sound/soc/codecs/wm2000.c +++ b/sound/soc/codecs/wm2000.c @@ -733,8 +733,9 @@ static int __devinit wm2000_i2c_probe(struct i2c_client *i2c, struct wm2000_priv *wm2000; struct wm2000_platform_data *pdata; const char *filename; - const struct firmware *fw; - int reg, ret; + const struct firmware *fw = NULL; + int ret; + int reg; u16 id; wm2000 = devm_kzalloc(&i2c->dev, sizeof(struct wm2000_priv), @@ -751,7 +752,7 @@ static int __devinit wm2000_i2c_probe(struct i2c_client *i2c, ret = PTR_ERR(wm2000->regmap); dev_err(&i2c->dev, "Failed to allocate register map: %d\n", ret); - goto err; + goto out; } /* Verify that this is a WM2000 */ @@ -763,7 +764,7 @@ static int __devinit wm2000_i2c_probe(struct i2c_client *i2c, if (id != 0x2000) { dev_err(&i2c->dev, "Device is not a WM2000 - ID %x\n", id); ret = -ENODEV; - goto err_regmap; + goto out_regmap_exit; } reg = wm2000_read(i2c, WM2000_REG_REVISON); @@ -782,7 +783,7 @@ static int __devinit wm2000_i2c_probe(struct i2c_client *i2c, ret = request_firmware(&fw, filename, &i2c->dev); if (ret != 0) { dev_err(&i2c->dev, "Failed to acquire ANC data: %d\n", ret); - goto err_regmap; + goto out_regmap_exit; } /* Pre-cook the concatenation of the register address onto the image */ @@ -793,15 +794,13 @@ static int __devinit wm2000_i2c_probe(struct i2c_client *i2c, if (wm2000->anc_download == NULL) { dev_err(&i2c->dev, "Out of memory\n"); ret = -ENOMEM; - goto err_fw; + goto out_regmap_exit; } wm2000->anc_download[0] = 0x80; wm2000->anc_download[1] = 0x00; memcpy(wm2000->anc_download + 2, fw->data, fw->size); - release_firmware(fw); - wm2000->anc_eng_ena = 1; wm2000->anc_active = 1; wm2000->spk_ena = 1; @@ -809,18 +808,14 @@ static int __devinit wm2000_i2c_probe(struct i2c_client *i2c, wm2000_reset(wm2000); - ret = snd_soc_register_codec(&i2c->dev, &soc_codec_dev_wm2000, - NULL, 0); - if (ret != 0) - goto err_fw; + ret = snd_soc_register_codec(&i2c->dev, &soc_codec_dev_wm2000, NULL, 0); + if (!ret) + goto out; - return 0; - -err_fw: - release_firmware(fw); -err_regmap: +out_regmap_exit: regmap_exit(wm2000->regmap); -err: +out: + release_firmware(fw); return ret; } From a3f83ab1a717c0e6c2f59a4cfdaa10707cc35c55 Mon Sep 17 00:00:00 2001 From: Igor Murzov Date: Sun, 22 Jan 2012 18:43:25 +0400 Subject: [PATCH 772/820] drm/radeon: fix invalid memory access in radeon_atrm_get_bios() At a boot time I observed following bug: BUG: unable to handle kernel paging request at ffff8800a4244000 IP: [] memcpy+0xb/0x120 PGD 1816063 PUD 1fe7d067 PMD 1ff9f067 PTE 80000000a4244160 Oops: 0000 [#1] SMP DEBUG_PAGEALLOC CPU 0 Modules linked in: btusb bluetooth brcmsmac brcmutil crc8 cordic b43 radeon(+) mac80211 cfg80211 ttm ohci_hcd drm_kms_helper rfkill drm ssb agpgart mmc_core sp5100_tco video battery ac thermal processor rtc_cmos thermal_sys snd_hda_codec_hdmi joydev snd_hda_codec_conexant button bcma pcmcia snd_hda_intel snd_hda_codec snd_hwdep snd_pcm shpchp pcmcia_core k8temp snd_timer atl1c snd psmouse hwmon i2c_piix4 i2c_algo_bit soundcore evdev i2c_core ehci_hcd sg serio_raw snd_page_alloc loop btrfs Pid: 1008, comm: modprobe Not tainted 3.3.0-rc1 #21 LENOVO 20046 /AMD CRB RIP: 0010:[] [] memcpy+0xb/0x120 RSP: 0018:ffff8800aa72db00 EFLAGS: 00010246 RAX: ffff8800a4150000 RBX: 0000000000001000 RCX: 0000000000000087 RDX: 0000000000000000 RSI: ffff8800a4244000 RDI: ffff8800a4150bc8 RBP: ffff8800aa72db78 R08: 0000000000000010 R09: ffffffff8174bbec R10: ffffffff812ee010 R11: 0000000000000001 R12: 0000000000001000 R13: 0000000000010000 R14: ffff8800a4140000 R15: ffff8800aaba1800 FS: 00007ff9a3bd4720(0000) GS:ffff8800afa00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: ffff8800a4244000 CR3: 00000000a9c18000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process modprobe (pid: 1008, threadinfo ffff8800aa72c000, task ffff8800aa0e4000) Stack: ffffffffa04e7c7b 0000000000000001 0000000000010000 ffff8800aa72db28 ffffffff00000001 0000000000001000 ffffffff8113cbef 0000000000000020 ffff8800a4243420 ffff880000000002 ffff8800aa72db08 ffff8800a9d42000 Call Trace: [] ? radeon_atrm_get_bios_chunk+0x8b/0xd0 [radeon] [] ? kmalloc_order_trace+0x3f/0xb0 [] radeon_get_bios+0x68/0x2f0 [radeon] [] rv770_init+0x40/0x280 [radeon] [] radeon_device_init+0x560/0x600 [radeon] [] radeon_driver_load_kms+0xaf/0x170 [radeon] [] drm_get_pci_dev+0x18e/0x2c0 [drm] [] radeon_pci_probe+0xad/0xb5 [radeon] [] local_pci_probe+0x5f/0xd0 [] pci_device_probe+0x88/0xb0 [] ? driver_sysfs_add+0x7a/0xb0 [] really_probe+0x68/0x180 [] driver_probe_device+0x45/0x70 [] __driver_attach+0xa3/0xb0 [] ? driver_probe_device+0x70/0x70 [] bus_for_each_dev+0x5e/0x90 [] driver_attach+0x1e/0x20 [] bus_add_driver+0xc8/0x280 [] driver_register+0x76/0x140 [] __pci_register_driver+0x66/0xe0 [] drm_pci_init+0x111/0x120 [drm] [] ? vga_switcheroo_register_handler+0x3a/0x60 [] ? 0xffffffffa0228fff [] radeon_init+0xec/0xee [radeon] [] do_one_initcall+0x42/0x180 [] sys_init_module+0x92/0x1e0 [] system_call_fastpath+0x16/0x1b Code: 58 2a 43 50 88 43 4e 48 83 c4 08 5b c9 c3 66 90 e8 cb fd ff ff eb e6 90 90 90 90 90 90 90 90 90 48 89 f8 89 d1 c1 e9 03 83 e2 07 48 a5 89 d1 f3 a4 c3 20 48 83 ea 20 4c 8b 06 4c 8b 4e 08 4c RIP [] memcpy+0xb/0x120 RSP CR2: ffff8800a4244000 ---[ end trace fcffa1599cf56382 ]--- Call to acpi_evaluate_object() not always returns 4096 bytes chunks, on my system it can return 2048 bytes chunk, so pass the length of retrieved chunk to memcpy(), not the length of the recieving buffer. Signed-off-by: Igor Murzov Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon_atpx_handler.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c index 9d95792bea3e..c666a5b5e505 100644 --- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c +++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c @@ -58,7 +58,7 @@ static int radeon_atrm_call(acpi_handle atrm_handle, uint8_t *bios, } obj = (union acpi_object *)buffer.pointer; - memcpy(bios+offset, obj->buffer.pointer, len); + memcpy(bios+offset, obj->buffer.pointer, obj->buffer.length); kfree(buffer.pointer); return len; } From 211fa4fc4e13492151e698d92b0dff56b29928ec Mon Sep 17 00:00:00 2001 From: Igor Murzov Date: Sun, 22 Jan 2012 18:47:28 +0400 Subject: [PATCH 773/820] drm/radeon: finish getting bios earlier Return a number of bytes read in radeon_atrm_get_bios_chunk() and properly check this value in radeon_atrm_get_bios(). If radeon_atrm_get_bios_chunk() read less bytes then were requested, it means that it finished reading bios data. Prior to this patch, condition in radeon_atrm_get_bios() was always equivalent to "if (ATRM_BIOS_PAGE <= 0)", so it was always false, thus radeon_atrm_get_bios() was trying to read past the bios data wasting boot time. On my lenovo ideapad u455 laptop this patch drops bios reading time from ~5.5s to ~1.5s. Signed-off-by: Igor Murzov Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon_atpx_handler.c | 2 +- drivers/gpu/drm/radeon/radeon_bios.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c index c666a5b5e505..13ac63ba6075 100644 --- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c +++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c @@ -60,7 +60,7 @@ static int radeon_atrm_call(acpi_handle atrm_handle, uint8_t *bios, obj = (union acpi_object *)buffer.pointer; memcpy(bios+offset, obj->buffer.pointer, obj->buffer.length); kfree(buffer.pointer); - return len; + return obj->buffer.length; } bool radeon_atrm_supported(struct pci_dev *pdev) diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c index 229a20f10e2b..501f4881e5aa 100644 --- a/drivers/gpu/drm/radeon/radeon_bios.c +++ b/drivers/gpu/drm/radeon/radeon_bios.c @@ -120,7 +120,7 @@ static bool radeon_atrm_get_bios(struct radeon_device *rdev) ret = radeon_atrm_get_bios_chunk(rdev->bios, (i * ATRM_BIOS_PAGE), ATRM_BIOS_PAGE); - if (ret <= 0) + if (ret < ATRM_BIOS_PAGE) break; } From 3fa47d9efa6a0f5123e26e2c3ad54e3e1a1d108d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 20 Jan 2012 14:56:39 -0500 Subject: [PATCH 774/820] drm/radeon/kms: move disp eng pll setup to init path We really only need to set it up once on init or resume rather than on every mode set. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/atombios_crtc.c | 56 ++++++++++++------------- drivers/gpu/drm/radeon/radeon_device.c | 6 ++- drivers/gpu/drm/radeon/radeon_display.c | 6 ++- drivers/gpu/drm/radeon/radeon_mode.h | 1 + 4 files changed, 36 insertions(+), 33 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index 0fda830ef806..807b89b4933f 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -355,15 +355,12 @@ static void atombios_crtc_set_timing(struct drm_crtc *crtc, atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); } -static void atombios_disable_ss(struct drm_crtc *crtc) +static void atombios_disable_ss(struct radeon_device *rdev, int pll_id) { - struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); - struct drm_device *dev = crtc->dev; - struct radeon_device *rdev = dev->dev_private; u32 ss_cntl; if (ASIC_IS_DCE4(rdev)) { - switch (radeon_crtc->pll_id) { + switch (pll_id) { case ATOM_PPLL1: ss_cntl = RREG32(EVERGREEN_P1PLL_SS_CNTL); ss_cntl &= ~EVERGREEN_PxPLL_SS_EN; @@ -379,7 +376,7 @@ static void atombios_disable_ss(struct drm_crtc *crtc) return; } } else if (ASIC_IS_AVIVO(rdev)) { - switch (radeon_crtc->pll_id) { + switch (pll_id) { case ATOM_PPLL1: ss_cntl = RREG32(AVIVO_P1PLL_INT_SS_CNTL); ss_cntl &= ~1; @@ -406,13 +403,11 @@ union atom_enable_ss { ENABLE_SPREAD_SPECTRUM_ON_PPLL_V3 v3; }; -static void atombios_crtc_program_ss(struct drm_crtc *crtc, +static void atombios_crtc_program_ss(struct radeon_device *rdev, int enable, int pll_id, struct radeon_atom_ss *ss) { - struct drm_device *dev = crtc->dev; - struct radeon_device *rdev = dev->dev_private; int index = GetIndexIntoMasterTable(COMMAND, EnableSpreadSpectrumOnPPLL); union atom_enable_ss args; @@ -479,7 +474,7 @@ static void atombios_crtc_program_ss(struct drm_crtc *crtc, } else if (ASIC_IS_AVIVO(rdev)) { if ((enable == ATOM_DISABLE) || (ss->percentage == 0) || (ss->type & ATOM_EXTERNAL_SS_MASK)) { - atombios_disable_ss(crtc); + atombios_disable_ss(rdev, pll_id); return; } args.lvds_ss_2.usSpreadSpectrumPercentage = cpu_to_le16(ss->percentage); @@ -491,7 +486,7 @@ static void atombios_crtc_program_ss(struct drm_crtc *crtc, } else { if ((enable == ATOM_DISABLE) || (ss->percentage == 0) || (ss->type & ATOM_EXTERNAL_SS_MASK)) { - atombios_disable_ss(crtc); + atombios_disable_ss(rdev, pll_id); return; } args.lvds_ss.usSpreadSpectrumPercentage = cpu_to_le16(ss->percentage); @@ -702,11 +697,9 @@ union set_pixel_clock { /* on DCE5, make sure the voltage is high enough to support the * required disp clk. */ -static void atombios_crtc_set_dcpll(struct drm_crtc *crtc, +static void atombios_crtc_set_dcpll(struct radeon_device *rdev, u32 dispclk) { - struct drm_device *dev = crtc->dev; - struct radeon_device *rdev = dev->dev_private; u8 frev, crev; int index; union set_pixel_clock args; @@ -996,7 +989,7 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode radeon_compute_pll_legacy(pll, adjusted_clock, &pll_clock, &fb_div, &frac_fb_div, &ref_div, &post_div); - atombios_crtc_program_ss(crtc, ATOM_DISABLE, radeon_crtc->pll_id, &ss); + atombios_crtc_program_ss(rdev, ATOM_DISABLE, radeon_crtc->pll_id, &ss); atombios_crtc_program_pll(crtc, radeon_crtc->crtc_id, radeon_crtc->pll_id, encoder_mode, radeon_encoder->encoder_id, mode->clock, @@ -1019,7 +1012,7 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode ss.step = step_size; } - atombios_crtc_program_ss(crtc, ATOM_ENABLE, radeon_crtc->pll_id, &ss); + atombios_crtc_program_ss(rdev, ATOM_ENABLE, radeon_crtc->pll_id, &ss); } } @@ -1494,6 +1487,24 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc) } +void radeon_atom_dcpll_init(struct radeon_device *rdev) +{ + /* always set DCPLL */ + if (ASIC_IS_DCE4(rdev)) { + struct radeon_atom_ss ss; + bool ss_enabled = radeon_atombios_get_asic_ss_info(rdev, &ss, + ASIC_INTERNAL_SS_ON_DCPLL, + rdev->clock.default_dispclk); + if (ss_enabled) + atombios_crtc_program_ss(rdev, ATOM_DISABLE, ATOM_DCPLL, &ss); + /* XXX: DCE5, make sure voltage, dispclk is high enough */ + atombios_crtc_set_dcpll(rdev, rdev->clock.default_dispclk); + if (ss_enabled) + atombios_crtc_program_ss(rdev, ATOM_ENABLE, ATOM_DCPLL, &ss); + } + +} + int atombios_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode, @@ -1515,19 +1526,6 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc, } } - /* always set DCPLL */ - if (ASIC_IS_DCE4(rdev)) { - struct radeon_atom_ss ss; - bool ss_enabled = radeon_atombios_get_asic_ss_info(rdev, &ss, - ASIC_INTERNAL_SS_ON_DCPLL, - rdev->clock.default_dispclk); - if (ss_enabled) - atombios_crtc_program_ss(crtc, ATOM_DISABLE, ATOM_DCPLL, &ss); - /* XXX: DCE5, make sure voltage, dispclk is high enough */ - atombios_crtc_set_dcpll(crtc, rdev->clock.default_dispclk); - if (ss_enabled) - atombios_crtc_program_ss(crtc, ATOM_ENABLE, ATOM_DCPLL, &ss); - } atombios_crtc_set_pll(crtc, adjusted_mode); if (ASIC_IS_DCE4(rdev)) diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 0afb13bd8dca..a811bc64ad5c 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -959,9 +959,11 @@ int radeon_resume_kms(struct drm_device *dev) radeon_fbdev_set_suspend(rdev, 0); console_unlock(); - /* init dig PHYs */ - if (rdev->is_atom_bios) + /* init dig PHYs, disp eng pll */ + if (rdev->is_atom_bios) { radeon_atom_encoder_init(rdev); + radeon_atom_dcpll_init(rdev); + } /* reset hpd state */ radeon_hpd_init(rdev); /* blat the mode back in */ diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index d3ffc18774a6..8c49fef1ce78 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1305,9 +1305,11 @@ int radeon_modeset_init(struct radeon_device *rdev) return ret; } - /* init dig PHYs */ - if (rdev->is_atom_bios) + /* init dig PHYs, disp eng pll */ + if (rdev->is_atom_bios) { radeon_atom_encoder_init(rdev); + radeon_atom_dcpll_init(rdev); + } /* initialize hpd */ radeon_hpd_init(rdev); diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index 08ff857c8fd6..8cb19f38f8da 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -484,6 +484,7 @@ extern u8 radeon_dp_getsinktype(struct radeon_connector *radeon_connector); extern bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector); extern void atombios_dig_encoder_setup(struct drm_encoder *encoder, int action, int panel_mode); extern void radeon_atom_encoder_init(struct radeon_device *rdev); +extern void radeon_atom_dcpll_init(struct radeon_device *rdev); extern void atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t lane_num, uint8_t lane_set); From 386d4d751e8e0b4b693bb724f09aae064ee5297d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 20 Jan 2012 15:01:29 -0500 Subject: [PATCH 775/820] drm/radeon/kms: move panel mode setup into encoder mode set Needs to happen earlier in the mode set. Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/atombios_dp.c | 22 ++++++++++------------ drivers/gpu/drm/radeon/atombios_encoders.c | 11 +++++++++++ drivers/gpu/drm/radeon/radeon_mode.h | 3 +++ 3 files changed, 24 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index 6fb335a4fdda..a71557ce01dc 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -549,8 +549,8 @@ bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector) return false; } -static void radeon_dp_set_panel_mode(struct drm_encoder *encoder, - struct drm_connector *connector) +int radeon_dp_get_panel_mode(struct drm_encoder *encoder, + struct drm_connector *connector) { struct drm_device *dev = encoder->dev; struct radeon_device *rdev = dev->dev_private; @@ -558,7 +558,7 @@ static void radeon_dp_set_panel_mode(struct drm_encoder *encoder, int panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE; if (!ASIC_IS_DCE4(rdev)) - return; + return panel_mode; if (radeon_connector_encoder_get_dp_bridge_encoder_id(connector) == ENCODER_OBJECT_ID_NUTMEG) @@ -572,14 +572,7 @@ static void radeon_dp_set_panel_mode(struct drm_encoder *encoder, panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE; } - atombios_dig_encoder_setup(encoder, - ATOM_ENCODER_CMD_SETUP_PANEL_MODE, - panel_mode); - - if ((connector->connector_type == DRM_MODE_CONNECTOR_eDP) && - (panel_mode == DP_PANEL_MODE_INTERNAL_DP2_MODE)) { - radeon_write_dpcd_reg(radeon_connector, DP_EDP_CONFIGURATION_SET, 1); - } + return panel_mode; } void radeon_dp_set_link_config(struct drm_connector *connector, @@ -717,6 +710,8 @@ static void radeon_dp_set_tp(struct radeon_dp_link_train_info *dp_info, int tp) static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info) { + struct radeon_encoder *radeon_encoder = to_radeon_encoder(dp_info->encoder); + struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; u8 tmp; /* power up the sink */ @@ -732,7 +727,10 @@ static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info) radeon_write_dpcd_reg(dp_info->radeon_connector, DP_DOWNSPREAD_CTRL, 0); - radeon_dp_set_panel_mode(dp_info->encoder, dp_info->connector); + if ((dp_info->connector->connector_type == DRM_MODE_CONNECTOR_eDP) && + (dig->panel_mode == DP_PANEL_MODE_INTERNAL_DP2_MODE)) { + radeon_write_dpcd_reg(dp_info->radeon_connector, DP_EDP_CONFIGURATION_SET, 1); + } /* set the lane count on the sink */ tmp = dp_info->dp_lane_count; diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index f1f06ca9f1f5..a3a9166f1ecb 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -1811,10 +1811,21 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder, case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: if (ASIC_IS_DCE4(rdev)) { + struct drm_connector *connector = radeon_get_connector_for_encoder(encoder); + struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; + + if (!connector) + dig->panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE; + else + dig->panel_mode = radeon_dp_get_panel_mode(encoder, connector); + /* disable the transmitter */ atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0); /* setup and enable the encoder */ atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_SETUP, 0); + atombios_dig_encoder_setup(encoder, + ATOM_ENCODER_CMD_SETUP_PANEL_MODE, + dig->panel_mode); /* enable the transmitter */ atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0); diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index 8cb19f38f8da..d34dcb6ac384 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -362,6 +362,7 @@ struct radeon_encoder_atom_dig { struct backlight_device *bl_dev; int dpms_mode; uint8_t backlight_level; + int panel_mode; }; struct radeon_encoder_atom_dac { @@ -482,6 +483,8 @@ extern void radeon_dp_link_train(struct drm_encoder *encoder, extern bool radeon_dp_needs_link_train(struct radeon_connector *radeon_connector); extern u8 radeon_dp_getsinktype(struct radeon_connector *radeon_connector); extern bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector); +extern int radeon_dp_get_panel_mode(struct drm_encoder *encoder, + struct drm_connector *connector); extern void atombios_dig_encoder_setup(struct drm_encoder *encoder, int action, int panel_mode); extern void radeon_atom_encoder_init(struct radeon_device *rdev); extern void radeon_atom_dcpll_init(struct radeon_device *rdev); From 3a47824d85eeca122895646f027dc63480994199 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 20 Jan 2012 15:01:30 -0500 Subject: [PATCH 776/820] drm/radeon/kms: rework modeset sequence for DCE41 and DCE5 dig transmitter control table only has ENABLE/DISABLE actions on DCE4.1/DCE5. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=44955 Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/atombios_encoders.c | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index a3a9166f1ecb..79a4880e8800 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -1341,7 +1341,8 @@ radeon_atom_encoder_dpms_dig(struct drm_encoder *encoder, int mode) switch (mode) { case DRM_MODE_DPMS_ON: /* some early dce3.2 boards have a bug in their transmitter control table */ - if ((rdev->family == CHIP_RV710) || (rdev->family == CHIP_RV730)) + if ((rdev->family == CHIP_RV710) || (rdev->family == CHIP_RV730) || + ASIC_IS_DCE41(rdev) || ASIC_IS_DCE5(rdev)) atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0); else atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE_OUTPUT, 0, 0); @@ -1351,8 +1352,6 @@ radeon_atom_encoder_dpms_dig(struct drm_encoder *encoder, int mode) ATOM_TRANSMITTER_ACTION_POWER_ON); radeon_dig_connector->edp_on = true; } - if (ASIC_IS_DCE4(rdev)) - atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_OFF, 0); radeon_dp_link_train(encoder, connector); if (ASIC_IS_DCE4(rdev)) atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_ON, 0); @@ -1363,7 +1362,10 @@ radeon_atom_encoder_dpms_dig(struct drm_encoder *encoder, int mode) case DRM_MODE_DPMS_STANDBY: case DRM_MODE_DPMS_SUSPEND: case DRM_MODE_DPMS_OFF: - atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE_OUTPUT, 0, 0); + if (ASIC_IS_DCE41(rdev) || ASIC_IS_DCE5(rdev)) + atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0); + else + atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE_OUTPUT, 0, 0); if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(encoder)) && connector) { if (ASIC_IS_DCE4(rdev)) atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_OFF, 0); @@ -1810,7 +1812,7 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder, case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: - if (ASIC_IS_DCE4(rdev)) { + if (ASIC_IS_DCE41(rdev) || ASIC_IS_DCE5(rdev)) { struct drm_connector *connector = radeon_get_connector_for_encoder(encoder); struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; @@ -1819,13 +1821,16 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder, else dig->panel_mode = radeon_dp_get_panel_mode(encoder, connector); - /* disable the transmitter */ - atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0); /* setup and enable the encoder */ atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_SETUP, 0); atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_SETUP_PANEL_MODE, dig->panel_mode); + } else if (ASIC_IS_DCE4(rdev)) { + /* disable the transmitter */ + atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0); + /* setup and enable the encoder */ + atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_SETUP, 0); /* enable the transmitter */ atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0); From 27d9cc8428367e0fec2fc0fc6385e9241b079c3a Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 20 Jan 2012 15:03:29 -0500 Subject: [PATCH 777/820] drm/radeon/kms: use drm_detect_hdmi_monitor for picking encoder mode We were previously just checking for audio. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/atombios_encoders.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index 79a4880e8800..f2f14a20a493 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -432,7 +432,7 @@ atombios_get_encoder_mode(struct drm_encoder *encoder) switch (connector->connector_type) { case DRM_MODE_CONNECTOR_DVII: case DRM_MODE_CONNECTOR_HDMIB: /* HDMI-B is basically DL-DVI; analog works fine */ - if (drm_detect_monitor_audio(radeon_connector->edid) && + if (drm_detect_hdmi_monitor(radeon_connector->edid) && radeon_audio) return ATOM_ENCODER_MODE_HDMI; else if (radeon_connector->use_digital) @@ -443,7 +443,7 @@ atombios_get_encoder_mode(struct drm_encoder *encoder) case DRM_MODE_CONNECTOR_DVID: case DRM_MODE_CONNECTOR_HDMIA: default: - if (drm_detect_monitor_audio(radeon_connector->edid) && + if (drm_detect_hdmi_monitor(radeon_connector->edid) && radeon_audio) return ATOM_ENCODER_MODE_HDMI; else @@ -457,7 +457,7 @@ atombios_get_encoder_mode(struct drm_encoder *encoder) if ((dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT) || (dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_eDP)) return ATOM_ENCODER_MODE_DP; - else if (drm_detect_monitor_audio(radeon_connector->edid) && + else if (drm_detect_hdmi_monitor(radeon_connector->edid) && radeon_audio) return ATOM_ENCODER_MODE_HDMI; else From 9aa59993e226af94088adaee993eb8cfd33ae295 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 20 Jan 2012 15:03:30 -0500 Subject: [PATCH 778/820] drm/radeon/kms: refine TMDS dual link checks HDMI 1.3 defines single link clocks up to 340 Mhz. Refine the current dual link checks to only enable dual link for DVI > 165 Mhz or HDMI > 340 Mhz if the hw supports HDMI 1.3 (DCE3+). Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=44755 Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/atombios_crtc.c | 4 +- drivers/gpu/drm/radeon/atombios_encoders.c | 57 ++++++---------- drivers/gpu/drm/radeon/radeon_encoders.c | 77 ++++++++++++++++++++++ drivers/gpu/drm/radeon/radeon_mode.h | 4 ++ 4 files changed, 105 insertions(+), 37 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index 807b89b4933f..891935271d34 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -518,6 +518,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, int encoder_mode = 0; u32 dp_clock = mode->clock; int bpc = 8; + bool is_duallink = false; /* reset the pll flags */ pll->flags = 0; @@ -552,6 +553,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, if (connector && connector->display_info.bpc) bpc = connector->display_info.bpc; encoder_mode = atombios_get_encoder_mode(encoder); + is_duallink = radeon_dig_monitor_is_duallink(encoder, mode->clock); if ((radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT | ATOM_DEVICE_DFP_SUPPORT)) || (radeon_encoder_get_dp_bridge_encoder_id(encoder) != ENCODER_OBJECT_ID_NONE)) { if (connector) { @@ -647,7 +649,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, if (dig->coherent_mode) args.v3.sInput.ucDispPllConfig |= DISPPLL_CONFIG_COHERENT_MODE; - if (mode->clock > 165000) + if (is_duallink) args.v3.sInput.ucDispPllConfig |= DISPPLL_CONFIG_DUAL_LINK; } diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index f2f14a20a493..b88c4608731b 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -57,22 +57,6 @@ static inline bool radeon_encoder_is_digital(struct drm_encoder *encoder) } } -static struct drm_connector * -radeon_get_connector_for_encoder_init(struct drm_encoder *encoder) -{ - struct drm_device *dev = encoder->dev; - struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); - struct drm_connector *connector; - struct radeon_connector *radeon_connector; - - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - radeon_connector = to_radeon_connector(connector); - if (radeon_encoder->devices & radeon_connector->devices) - return connector; - } - return NULL; -} - static bool radeon_atom_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) @@ -253,7 +237,7 @@ atombios_dvo_setup(struct drm_encoder *encoder, int action) /* R4xx, R5xx */ args.ext_tmds.sXTmdsEncoder.ucEnable = action; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.ext_tmds.sXTmdsEncoder.ucMisc |= PANEL_ENCODER_MISC_DUAL; args.ext_tmds.sXTmdsEncoder.ucMisc |= ATOM_PANEL_MISC_888RGB; @@ -265,7 +249,7 @@ atombios_dvo_setup(struct drm_encoder *encoder, int action) /* DFP1, CRT1, TV1 depending on the type of port */ args.dvo.sDVOEncoder.ucDeviceType = ATOM_DEVICE_DFP1_INDEX; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.dvo.sDVOEncoder.usDevAttr.sDigAttrib.ucAttribute |= PANEL_ENCODER_MISC_DUAL; break; case 3: @@ -349,7 +333,7 @@ atombios_digital_setup(struct drm_encoder *encoder, int action) } else { if (dig->linkb) args.v1.ucMisc |= PANEL_ENCODER_MISC_TMDS_LINKB; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v1.ucMisc |= PANEL_ENCODER_MISC_DUAL; /*if (pScrn->rgbBits == 8) */ args.v1.ucMisc |= ATOM_PANEL_MISC_888RGB; @@ -388,7 +372,7 @@ atombios_digital_setup(struct drm_encoder *encoder, int action) } else { if (dig->linkb) args.v2.ucMisc |= PANEL_ENCODER_MISC_TMDS_LINKB; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v2.ucMisc |= PANEL_ENCODER_MISC_DUAL; } break; @@ -587,7 +571,7 @@ atombios_dig_encoder_setup(struct drm_encoder *encoder, int action, int panel_mo if (ENCODER_MODE_IS_DP(args.v1.ucEncoderMode)) args.v1.ucLaneNum = dp_lane_count; - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v1.ucLaneNum = 8; else args.v1.ucLaneNum = 4; @@ -622,7 +606,7 @@ atombios_dig_encoder_setup(struct drm_encoder *encoder, int action, int panel_mo if (ENCODER_MODE_IS_DP(args.v1.ucEncoderMode)) args.v3.ucLaneNum = dp_lane_count; - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v3.ucLaneNum = 8; else args.v3.ucLaneNum = 4; @@ -662,7 +646,7 @@ atombios_dig_encoder_setup(struct drm_encoder *encoder, int action, int panel_mo if (ENCODER_MODE_IS_DP(args.v1.ucEncoderMode)) args.v4.ucLaneNum = dp_lane_count; - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v4.ucLaneNum = 8; else args.v4.ucLaneNum = 4; @@ -806,7 +790,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if (is_dp) args.v1.usPixelClock = cpu_to_le16(dp_clock / 10); - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v1.usPixelClock = cpu_to_le16((radeon_encoder->pixel_clock / 2) / 10); else args.v1.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10); @@ -821,7 +805,8 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if ((rdev->flags & RADEON_IS_IGP) && (radeon_encoder->encoder_id == ENCODER_OBJECT_ID_INTERNAL_UNIPHY)) { - if (is_dp || (radeon_encoder->pixel_clock <= 165000)) { + if (is_dp || + !radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) { if (igp_lane_info & 0x1) args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_0_3; else if (igp_lane_info & 0x2) @@ -848,7 +833,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t else if (radeon_encoder->devices & (ATOM_DEVICE_DFP_SUPPORT)) { if (dig->coherent_mode) args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_COHERENT; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_8LANE_LINK; } break; @@ -863,7 +848,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if (is_dp) args.v2.usPixelClock = cpu_to_le16(dp_clock / 10); - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v2.usPixelClock = cpu_to_le16((radeon_encoder->pixel_clock / 2) / 10); else args.v2.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10); @@ -891,7 +876,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t } else if (radeon_encoder->devices & (ATOM_DEVICE_DFP_SUPPORT)) { if (dig->coherent_mode) args.v2.acConfig.fCoherentMode = 1; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v2.acConfig.fDualLinkConnector = 1; } break; @@ -906,7 +891,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if (is_dp) args.v3.usPixelClock = cpu_to_le16(dp_clock / 10); - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v3.usPixelClock = cpu_to_le16((radeon_encoder->pixel_clock / 2) / 10); else args.v3.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10); @@ -914,7 +899,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if (is_dp) args.v3.ucLaneNum = dp_lane_count; - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v3.ucLaneNum = 8; else args.v3.ucLaneNum = 4; @@ -951,7 +936,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t else if (radeon_encoder->devices & (ATOM_DEVICE_DFP_SUPPORT)) { if (dig->coherent_mode) args.v3.acConfig.fCoherentMode = 1; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v3.acConfig.fDualLinkConnector = 1; } break; @@ -966,7 +951,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if (is_dp) args.v4.usPixelClock = cpu_to_le16(dp_clock / 10); - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v4.usPixelClock = cpu_to_le16((radeon_encoder->pixel_clock / 2) / 10); else args.v4.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10); @@ -974,7 +959,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t if (is_dp) args.v4.ucLaneNum = dp_lane_count; - else if (radeon_encoder->pixel_clock > 165000) + else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v4.ucLaneNum = 8; else args.v4.ucLaneNum = 4; @@ -1014,7 +999,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t else if (radeon_encoder->devices & (ATOM_DEVICE_DFP_SUPPORT)) { if (dig->coherent_mode) args.v4.acConfig.fCoherentMode = 1; - if (radeon_encoder->pixel_clock > 165000) + if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v4.acConfig.fDualLinkConnector = 1; } break; @@ -1137,7 +1122,7 @@ atombios_external_encoder_setup(struct drm_encoder *encoder, if (dp_clock == 270000) args.v1.sDigEncoder.ucConfig |= ATOM_ENCODER_CONFIG_DPLINKRATE_2_70GHZ; args.v1.sDigEncoder.ucLaneNum = dp_lane_count; - } else if (radeon_encoder->pixel_clock > 165000) + } else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v1.sDigEncoder.ucLaneNum = 8; else args.v1.sDigEncoder.ucLaneNum = 4; @@ -1156,7 +1141,7 @@ atombios_external_encoder_setup(struct drm_encoder *encoder, else if (dp_clock == 540000) args.v3.sExtEncoder.ucConfig |= EXTERNAL_ENCODER_CONFIG_V3_DPLINKRATE_5_40GHZ; args.v3.sExtEncoder.ucLaneNum = dp_lane_count; - } else if (radeon_encoder->pixel_clock > 165000) + } else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock)) args.v3.sExtEncoder.ucLaneNum = 8; else args.v3.sExtEncoder.ucLaneNum = 4; diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 4b27efa4405b..9419c51bcf50 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -202,6 +202,22 @@ radeon_get_connector_for_encoder(struct drm_encoder *encoder) return NULL; } +struct drm_connector * +radeon_get_connector_for_encoder_init(struct drm_encoder *encoder) +{ + struct drm_device *dev = encoder->dev; + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + struct drm_connector *connector; + struct radeon_connector *radeon_connector; + + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + radeon_connector = to_radeon_connector(connector); + if (radeon_encoder->devices & radeon_connector->devices) + return connector; + } + return NULL; +} + struct drm_encoder *radeon_get_external_encoder(struct drm_encoder *encoder) { struct drm_device *dev = encoder->dev; @@ -288,3 +304,64 @@ void radeon_panel_mode_fixup(struct drm_encoder *encoder, } +bool radeon_dig_monitor_is_duallink(struct drm_encoder *encoder, + u32 pixel_clock) +{ + struct drm_device *dev = encoder->dev; + struct radeon_device *rdev = dev->dev_private; + struct drm_connector *connector; + struct radeon_connector *radeon_connector; + struct radeon_connector_atom_dig *dig_connector; + + connector = radeon_get_connector_for_encoder(encoder); + /* if we don't have an active device yet, just use one of + * the connectors tied to the encoder. + */ + if (!connector) + connector = radeon_get_connector_for_encoder_init(encoder); + radeon_connector = to_radeon_connector(connector); + + switch (connector->connector_type) { + case DRM_MODE_CONNECTOR_DVII: + case DRM_MODE_CONNECTOR_HDMIB: + if (radeon_connector->use_digital) { + /* HDMI 1.3 supports up to 340 Mhz over single link */ + if (ASIC_IS_DCE3(rdev) && drm_detect_hdmi_monitor(radeon_connector->edid)) { + if (pixel_clock > 340000) + return true; + else + return false; + } else { + if (pixel_clock > 165000) + return true; + else + return false; + } + } else + return false; + case DRM_MODE_CONNECTOR_DVID: + case DRM_MODE_CONNECTOR_HDMIA: + case DRM_MODE_CONNECTOR_DisplayPort: + dig_connector = radeon_connector->con_priv; + if ((dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT) || + (dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_eDP)) + return false; + else { + /* HDMI 1.3 supports up to 340 Mhz over single link */ + if (ASIC_IS_DCE3(rdev) && drm_detect_hdmi_monitor(radeon_connector->edid)) { + if (pixel_clock > 340000) + return true; + else + return false; + } else { + if (pixel_clock > 165000) + return true; + else + return false; + } + } + default: + return false; + } +} + diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index d34dcb6ac384..4330e3253573 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -467,6 +467,10 @@ radeon_atombios_get_tv_info(struct radeon_device *rdev); extern struct drm_connector * radeon_get_connector_for_encoder(struct drm_encoder *encoder); +extern struct drm_connector * +radeon_get_connector_for_encoder_init(struct drm_encoder *encoder); +extern bool radeon_dig_monitor_is_duallink(struct drm_encoder *encoder, + u32 pixel_clock); extern u16 radeon_encoder_get_dp_bridge_encoder_id(struct drm_encoder *encoder); extern u16 radeon_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *connector); From 15b63d35261ba3351d07e7937252f18bb6cbf814 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Tue, 24 Jan 2012 16:57:42 +0000 Subject: [PATCH 779/820] gma500: Fix shmem mapping GMA500 did it the old way and it's been on the TODO list to fix. Current kernels now blow up if we use the old way so we'd better do the work ! Signed-off-by: Alan Cox Signed-off-by: Dave Airlie --- drivers/gpu/drm/gma500/gtt.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/gma500/gtt.c b/drivers/gpu/drm/gma500/gtt.c index e770bd190a5c..5d5330f667f1 100644 --- a/drivers/gpu/drm/gma500/gtt.c +++ b/drivers/gpu/drm/gma500/gtt.c @@ -20,6 +20,7 @@ */ #include +#include #include "psb_drv.h" @@ -203,9 +204,7 @@ static int psb_gtt_attach_pages(struct gtt_range *gt) gt->npage = pages; for (i = 0; i < pages; i++) { - /* FIXME: needs updating as per mail from Hugh Dickins */ - p = read_cache_page_gfp(mapping, i, - __GFP_COLD | GFP_KERNEL); + p = shmem_read_mapping_page(mapping, i); if (IS_ERR(p)) goto err; gt->pages[i] = p; From e912b6d27cea198980132f012d14f22247e19ad6 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Tue, 24 Jan 2012 16:57:42 +0000 Subject: [PATCH 780/820] gma500: Fix shmem mapping GMA500 did it the old way and it's been on the TODO list to fix. Current kernels now blow up if we use the old way so we'd better do the work! Signed-off-by: Alan Cox Signed-off-by: Linus Torvalds --- drivers/gpu/drm/gma500/gtt.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/gma500/gtt.c b/drivers/gpu/drm/gma500/gtt.c index e770bd190a5c..5d5330f667f1 100644 --- a/drivers/gpu/drm/gma500/gtt.c +++ b/drivers/gpu/drm/gma500/gtt.c @@ -20,6 +20,7 @@ */ #include +#include #include "psb_drv.h" @@ -203,9 +204,7 @@ static int psb_gtt_attach_pages(struct gtt_range *gt) gt->npage = pages; for (i = 0; i < pages; i++) { - /* FIXME: needs updating as per mail from Hugh Dickins */ - p = read_cache_page_gfp(mapping, i, - __GFP_COLD | GFP_KERNEL); + p = shmem_read_mapping_page(mapping, i); if (IS_ERR(p)) goto err; gt->pages[i] = p; From 4a7cbb56fdbd92a47f57ca8b25bf5db35f0d6518 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 24 Jan 2012 11:17:26 +0000 Subject: [PATCH 781/820] regulator: Fix documentation for of_node parameter of regulator_register() Commit 5bc75a886353 ("kernel-doc: fix new warning in regulator core") added documentation for of_node to address a warning but the documentation didn't explain what the parameter is for so would be likely to be unhelpful for users. Clarify that. Signed-off-by: Mark Brown Signed-off-by: Linus Torvalds --- drivers/regulator/core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index b1fa25d6583a..e9a83f84adaf 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -2731,7 +2731,8 @@ static void rdev_init_debugfs(struct regulator_dev *rdev) * @dev: struct device for the regulator * @init_data: platform provided init data, passed through by driver * @driver_data: private regulator data - * @of_node: target open firmware device structure (may be NULL) + * @of_node: OpenFirmware node to parse for device tree bindings (may be + * NULL). * * Called by regulator drivers to register a regulator. * Returns 0 on success. From 90b9a5454fd2e626aa1614fe9ece6b63a0dc37af Mon Sep 17 00:00:00 2001 From: Alessandro Rubini Date: Mon, 23 Jan 2012 23:26:48 +0000 Subject: [PATCH 782/820] stmmac: fix phy naming inconsistency After commit "db8857b stmmac: use an unique MDIO bus name" my device stopped being probed because two different names were being used in different places. This fixes the inconsistency. Signed-off-by: Alessandro Rubini Acked-by: Giancarlo Asnaghi Cc: Giuseppe Cavallaro Cc: Florian Fainelli Acked-by: Florian Fainelli Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c index da4a1042523a..73195329aa46 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c @@ -154,7 +154,7 @@ int stmmac_mdio_register(struct net_device *ndev) else irqlist = priv->mii_irq; - new_bus->name = "STMMAC MII Bus"; + new_bus->name = "stmmac"; new_bus->read = &stmmac_mdio_read; new_bus->write = &stmmac_mdio_write; new_bus->reset = &stmmac_mdio_reset; From 56ac11cf2f21366ad48b356f7a0d1af8cff3588e Mon Sep 17 00:00:00 2001 From: Radu Iliescu Date: Thu, 19 Jan 2012 03:57:57 +0000 Subject: [PATCH 783/820] llc: Fix race condition in llc_ui_recvmsg There is a race on sk_receive_queue between llc_ui_recvmsg and sock_queue_rcv_skb. Our current solution is to protect skb_eat in llc_ui_recvmsg with the queue spinlock. Signed-off-by: Radu Iliescu Signed-off-by: David S. Miller --- net/llc/af_llc.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c index a18e6c3d36e3..b9bef2c75026 100644 --- a/net/llc/af_llc.c +++ b/net/llc/af_llc.c @@ -713,6 +713,7 @@ static int llc_ui_recvmsg(struct kiocb *iocb, struct socket *sock, struct sk_buff *skb = NULL; struct sock *sk = sock->sk; struct llc_sock *llc = llc_sk(sk); + unsigned long cpu_flags; size_t copied = 0; u32 peek_seq = 0; u32 *seq; @@ -838,7 +839,9 @@ static int llc_ui_recvmsg(struct kiocb *iocb, struct socket *sock, goto copy_uaddr; if (!(flags & MSG_PEEK)) { + spin_lock_irqsave(&sk->sk_receive_queue.lock, cpu_flags); sk_eat_skb(sk, skb, 0); + spin_unlock_irqrestore(&sk->sk_receive_queue.lock, cpu_flags); *seq = 0; } @@ -859,7 +862,9 @@ copy_uaddr: llc_cmsg_rcv(msg, skb); if (!(flags & MSG_PEEK)) { + spin_lock_irqsave(&sk->sk_receive_queue.lock, cpu_flags); sk_eat_skb(sk, skb, 0); + spin_unlock_irqrestore(&sk->sk_receive_queue.lock, cpu_flags); *seq = 0; } From 5437f4b2576f1763a27bc4c0e7f7c280220ba1aa Mon Sep 17 00:00:00 2001 From: Alessandro Rubini Date: Mon, 23 Jan 2012 23:08:56 +0000 Subject: [PATCH 784/820] stmmac: added PCI identifiers STM has a device ID within its own VENDOR space, and it is being used in the STA2X11 I/O Hub. Signed-off-by: Alessandro Rubini Acked-by: Giancarlo Asnaghi Acked-by: Giuseppe Cavallaro Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c index 54a819a36487..c796de9eed72 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c @@ -170,9 +170,9 @@ static int stmmac_pci_resume(struct pci_dev *pdev) #define STMMAC_DEVICE_ID 0x1108 static DEFINE_PCI_DEVICE_TABLE(stmmac_id_table) = { - { - PCI_DEVICE(STMMAC_VENDOR_ID, STMMAC_DEVICE_ID)}, { - } + {PCI_DEVICE(STMMAC_VENDOR_ID, STMMAC_DEVICE_ID)}, + {PCI_DEVICE(PCI_VENDOR_ID_STMICRO, PCI_DEVICE_ID_STMICRO_MAC)}, + {} }; MODULE_DEVICE_TABLE(pci, stmmac_id_table); From 2bbba277a554431a426e81f37d5c50ab6216c07d Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Tue, 24 Jan 2012 10:41:40 +0000 Subject: [PATCH 785/820] drivers/net: dsa/mv88e6xxx.c files need linux/module.h An implicit instance of module.h leaked back into existence and was masking the fact that these drivers weren't calling out the include for itself. Fix the drivers before we remove the implicit include path via net/netprio_cgroup.h file. Signed-off-by: Paul Gortmaker Signed-off-by: David S. Miller --- drivers/net/dsa/mv88e6060.c | 1 + drivers/net/dsa/mv88e6123_61_65.c | 1 + drivers/net/dsa/mv88e6131.c | 1 + drivers/net/dsa/mv88e6xxx.c | 1 + 4 files changed, 4 insertions(+) diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index 7fc4e81d4d43..325391d19bad 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -9,6 +9,7 @@ */ #include +#include #include #include #include diff --git a/drivers/net/dsa/mv88e6123_61_65.c b/drivers/net/dsa/mv88e6123_61_65.c index 6f23c9521f0e..c17c75b9f531 100644 --- a/drivers/net/dsa/mv88e6123_61_65.c +++ b/drivers/net/dsa/mv88e6123_61_65.c @@ -9,6 +9,7 @@ */ #include +#include #include #include #include diff --git a/drivers/net/dsa/mv88e6131.c b/drivers/net/dsa/mv88e6131.c index e0eb68243834..55888b06d8b4 100644 --- a/drivers/net/dsa/mv88e6131.c +++ b/drivers/net/dsa/mv88e6131.c @@ -9,6 +9,7 @@ */ #include +#include #include #include #include diff --git a/drivers/net/dsa/mv88e6xxx.c b/drivers/net/dsa/mv88e6xxx.c index 5467c040824a..a2c62c2f30ee 100644 --- a/drivers/net/dsa/mv88e6xxx.c +++ b/drivers/net/dsa/mv88e6xxx.c @@ -9,6 +9,7 @@ */ #include +#include #include #include #include From c11bf1c8baff170fa478adc04964da519d160e62 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Tue, 24 Jan 2012 10:21:28 +0000 Subject: [PATCH 786/820] net/hyperv: fix possible memory leak in do_set_multicast() do_set_multicast() may not free the memory malloc in netvsc_set_multicast_list(). Signed-off-by: Wei Yongjun Signed-off-by: Haiyang Zhang Signed-off-by: K. Y. Srinivasan Signed-off-by: David S. Miller --- drivers/net/hyperv/netvsc_drv.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c index 462d05f05e84..1a1ca6cfc74a 100644 --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@ -68,11 +68,11 @@ static void do_set_multicast(struct work_struct *w) nvdev = hv_get_drvdata(ndevctx->device_ctx); if (nvdev == NULL) - return; + goto out; rdev = nvdev->extension; if (rdev == NULL) - return; + goto out; if (net->flags & IFF_PROMISC) rndis_filter_set_packet_filter(rdev, @@ -83,6 +83,7 @@ static void do_set_multicast(struct work_struct *w) NDIS_PACKET_TYPE_ALL_MULTICAST | NDIS_PACKET_TYPE_DIRECTED); +out: kfree(w); } From b82b9183d4f18f9b8c4bb31f223eb6c79b734eb0 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Tue, 24 Jan 2012 05:16:00 +0000 Subject: [PATCH 787/820] team: send only changed options/ports via netlink This patch changes event message behaviour to send only updated records instead of whole list. This fixes bug on which userspace receives non-actual data in case multiple events occur in row. Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller --- drivers/net/team/team.c | 136 ++++++++++++++++++++++++++-------------- include/linux/if_team.h | 10 +++ 2 files changed, 100 insertions(+), 46 deletions(-) diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c index ed2a862b835d..6b678f38e5ce 100644 --- a/drivers/net/team/team.c +++ b/drivers/net/team/team.c @@ -92,9 +92,9 @@ struct team_option *__team_find_option(struct team *team, const char *opt_name) return NULL; } -int team_options_register(struct team *team, - const struct team_option *option, - size_t option_count) +int __team_options_register(struct team *team, + const struct team_option *option, + size_t option_count) { int i; struct team_option **dst_opts; @@ -116,8 +116,11 @@ int team_options_register(struct team *team, } } - for (i = 0; i < option_count; i++) + for (i = 0; i < option_count; i++) { + dst_opts[i]->changed = true; + dst_opts[i]->removed = false; list_add_tail(&dst_opts[i]->list, &team->option_list); + } kfree(dst_opts); return 0; @@ -130,10 +133,22 @@ rollback: return err; } -EXPORT_SYMBOL(team_options_register); +static void __team_options_mark_removed(struct team *team, + const struct team_option *option, + size_t option_count) +{ + int i; -static void __team_options_change_check(struct team *team, - struct team_option *changed_option); + for (i = 0; i < option_count; i++, option++) { + struct team_option *del_opt; + + del_opt = __team_find_option(team, option->name); + if (del_opt) { + del_opt->changed = true; + del_opt->removed = true; + } + } +} static void __team_options_unregister(struct team *team, const struct team_option *option, @@ -152,12 +167,29 @@ static void __team_options_unregister(struct team *team, } } +static void __team_options_change_check(struct team *team); + +int team_options_register(struct team *team, + const struct team_option *option, + size_t option_count) +{ + int err; + + err = __team_options_register(team, option, option_count); + if (err) + return err; + __team_options_change_check(team); + return 0; +} +EXPORT_SYMBOL(team_options_register); + void team_options_unregister(struct team *team, const struct team_option *option, size_t option_count) { + __team_options_mark_removed(team, option, option_count); + __team_options_change_check(team); __team_options_unregister(team, option, option_count); - __team_options_change_check(team, NULL); } EXPORT_SYMBOL(team_options_unregister); @@ -176,7 +208,8 @@ static int team_option_set(struct team *team, struct team_option *option, if (err) return err; - __team_options_change_check(team, option); + option->changed = true; + __team_options_change_check(team); return err; } @@ -653,6 +686,7 @@ static int team_port_del(struct team *team, struct net_device *port_dev) return -ENOENT; } + port->removed = true; __team_port_change_check(port, false); team_port_list_del_port(team, port); team_adjust_ops(team); @@ -1200,10 +1234,9 @@ err_fill: return err; } -static int team_nl_fill_options_get_changed(struct sk_buff *skb, - u32 pid, u32 seq, int flags, - struct team *team, - struct team_option *changed_option) +static int team_nl_fill_options_get(struct sk_buff *skb, + u32 pid, u32 seq, int flags, + struct team *team, bool fillall) { struct nlattr *option_list; void *hdr; @@ -1223,12 +1256,19 @@ static int team_nl_fill_options_get_changed(struct sk_buff *skb, struct nlattr *option_item; long arg; + /* Include only changed options if fill all mode is not on */ + if (!fillall && !option->changed) + continue; option_item = nla_nest_start(skb, TEAM_ATTR_ITEM_OPTION); if (!option_item) goto nla_put_failure; NLA_PUT_STRING(skb, TEAM_ATTR_OPTION_NAME, option->name); - if (option == changed_option) + if (option->changed) { NLA_PUT_FLAG(skb, TEAM_ATTR_OPTION_CHANGED); + option->changed = false; + } + if (option->removed) + NLA_PUT_FLAG(skb, TEAM_ATTR_OPTION_REMOVED); switch (option->type) { case TEAM_OPTION_TYPE_U32: NLA_PUT_U8(skb, TEAM_ATTR_OPTION_TYPE, NLA_U32); @@ -1255,13 +1295,13 @@ nla_put_failure: return -EMSGSIZE; } -static int team_nl_fill_options_get(struct sk_buff *skb, - struct genl_info *info, int flags, - struct team *team) +static int team_nl_fill_options_get_all(struct sk_buff *skb, + struct genl_info *info, int flags, + struct team *team) { - return team_nl_fill_options_get_changed(skb, info->snd_pid, - info->snd_seq, NLM_F_ACK, - team, NULL); + return team_nl_fill_options_get(skb, info->snd_pid, + info->snd_seq, NLM_F_ACK, + team, true); } static int team_nl_cmd_options_get(struct sk_buff *skb, struct genl_info *info) @@ -1273,7 +1313,7 @@ static int team_nl_cmd_options_get(struct sk_buff *skb, struct genl_info *info) if (!team) return -EINVAL; - err = team_nl_send_generic(info, team, team_nl_fill_options_get); + err = team_nl_send_generic(info, team, team_nl_fill_options_get_all); team_nl_team_put(team); @@ -1365,10 +1405,10 @@ team_put: return err; } -static int team_nl_fill_port_list_get_changed(struct sk_buff *skb, - u32 pid, u32 seq, int flags, - struct team *team, - struct team_port *changed_port) +static int team_nl_fill_port_list_get(struct sk_buff *skb, + u32 pid, u32 seq, int flags, + struct team *team, + bool fillall) { struct nlattr *port_list; void *hdr; @@ -1387,12 +1427,19 @@ static int team_nl_fill_port_list_get_changed(struct sk_buff *skb, list_for_each_entry(port, &team->port_list, list) { struct nlattr *port_item; + /* Include only changed ports if fill all mode is not on */ + if (!fillall && !port->changed) + continue; port_item = nla_nest_start(skb, TEAM_ATTR_ITEM_PORT); if (!port_item) goto nla_put_failure; NLA_PUT_U32(skb, TEAM_ATTR_PORT_IFINDEX, port->dev->ifindex); - if (port == changed_port) + if (port->changed) { NLA_PUT_FLAG(skb, TEAM_ATTR_PORT_CHANGED); + port->changed = false; + } + if (port->removed) + NLA_PUT_FLAG(skb, TEAM_ATTR_PORT_REMOVED); if (port->linkup) NLA_PUT_FLAG(skb, TEAM_ATTR_PORT_LINKUP); NLA_PUT_U32(skb, TEAM_ATTR_PORT_SPEED, port->speed); @@ -1408,13 +1455,13 @@ nla_put_failure: return -EMSGSIZE; } -static int team_nl_fill_port_list_get(struct sk_buff *skb, - struct genl_info *info, int flags, - struct team *team) +static int team_nl_fill_port_list_get_all(struct sk_buff *skb, + struct genl_info *info, int flags, + struct team *team) { - return team_nl_fill_port_list_get_changed(skb, info->snd_pid, - info->snd_seq, NLM_F_ACK, - team, NULL); + return team_nl_fill_port_list_get(skb, info->snd_pid, + info->snd_seq, NLM_F_ACK, + team, true); } static int team_nl_cmd_port_list_get(struct sk_buff *skb, @@ -1427,7 +1474,7 @@ static int team_nl_cmd_port_list_get(struct sk_buff *skb, if (!team) return -EINVAL; - err = team_nl_send_generic(info, team, team_nl_fill_port_list_get); + err = team_nl_send_generic(info, team, team_nl_fill_port_list_get_all); team_nl_team_put(team); @@ -1464,8 +1511,7 @@ static struct genl_multicast_group team_change_event_mcgrp = { .name = TEAM_GENL_CHANGE_EVENT_MC_GRP_NAME, }; -static int team_nl_send_event_options_get(struct team *team, - struct team_option *changed_option) +static int team_nl_send_event_options_get(struct team *team) { struct sk_buff *skb; int err; @@ -1475,8 +1521,7 @@ static int team_nl_send_event_options_get(struct team *team, if (!skb) return -ENOMEM; - err = team_nl_fill_options_get_changed(skb, 0, 0, 0, team, - changed_option); + err = team_nl_fill_options_get(skb, 0, 0, 0, team, false); if (err < 0) goto err_fill; @@ -1489,18 +1534,17 @@ err_fill: return err; } -static int team_nl_send_event_port_list_get(struct team_port *port) +static int team_nl_send_event_port_list_get(struct team *team) { struct sk_buff *skb; int err; - struct net *net = dev_net(port->team->dev); + struct net *net = dev_net(team->dev); skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL); if (!skb) return -ENOMEM; - err = team_nl_fill_port_list_get_changed(skb, 0, 0, 0, - port->team, port); + err = team_nl_fill_port_list_get(skb, 0, 0, 0, team, false); if (err < 0) goto err_fill; @@ -1544,12 +1588,11 @@ static void team_nl_fini(void) * Change checkers ******************/ -static void __team_options_change_check(struct team *team, - struct team_option *changed_option) +static void __team_options_change_check(struct team *team) { int err; - err = team_nl_send_event_options_get(team, changed_option); + err = team_nl_send_event_options_get(team); if (err) netdev_warn(team->dev, "Failed to send options change via netlink\n"); } @@ -1559,9 +1602,10 @@ static void __team_port_change_check(struct team_port *port, bool linkup) { int err; - if (port->linkup == linkup) + if (!port->removed && port->linkup == linkup) return; + port->changed = true; port->linkup = linkup; if (linkup) { struct ethtool_cmd ecmd; @@ -1577,7 +1621,7 @@ static void __team_port_change_check(struct team_port *port, bool linkup) port->duplex = 0; send_event: - err = team_nl_send_event_port_list_get(port); + err = team_nl_send_event_port_list_get(port->team); if (err) netdev_warn(port->team->dev, "Failed to send port change of device %s via netlink\n", port->dev->name); diff --git a/include/linux/if_team.h b/include/linux/if_team.h index 828181fbad5d..58404b0c5010 100644 --- a/include/linux/if_team.h +++ b/include/linux/if_team.h @@ -46,6 +46,10 @@ struct team_port { u32 speed; u8 duplex; + /* Custom gennetlink interface related flags */ + bool changed; + bool removed; + struct rcu_head rcu; }; @@ -72,6 +76,10 @@ struct team_option { enum team_option_type type; int (*getter)(struct team *team, void *arg); int (*setter)(struct team *team, void *arg); + + /* Custom gennetlink interface related flags */ + bool changed; + bool removed; }; struct team_mode { @@ -207,6 +215,7 @@ enum { TEAM_ATTR_OPTION_CHANGED, /* flag */ TEAM_ATTR_OPTION_TYPE, /* u8 */ TEAM_ATTR_OPTION_DATA, /* dynamic */ + TEAM_ATTR_OPTION_REMOVED, /* flag */ __TEAM_ATTR_OPTION_MAX, TEAM_ATTR_OPTION_MAX = __TEAM_ATTR_OPTION_MAX - 1, @@ -227,6 +236,7 @@ enum { TEAM_ATTR_PORT_LINKUP, /* flag */ TEAM_ATTR_PORT_SPEED, /* u32 */ TEAM_ATTR_PORT_DUPLEX, /* u8 */ + TEAM_ATTR_PORT_REMOVED, /* flag */ __TEAM_ATTR_PORT_MAX, TEAM_ATTR_PORT_MAX = __TEAM_ATTR_PORT_MAX - 1, From c452ed70771cea3af73d21a5914989137fbd28b8 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Tue, 24 Jan 2012 16:03:33 -0500 Subject: [PATCH 788/820] net: flow_dissector.c missing include linux/export.h The file net/core/flow_dissector.c seems to be missing including linux/export.h. Signed-off-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller --- net/core/flow_dissector.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c index 0985b9b14b80..a225089df5b6 100644 --- a/net/core/flow_dissector.c +++ b/net/core/flow_dissector.c @@ -1,4 +1,5 @@ #include +#include #include #include #include From 36a1211970193ce215de50ed1e4e1272bc814df1 Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Tue, 24 Jan 2012 11:33:19 +0000 Subject: [PATCH 789/820] netprio_cgroup.h: dont include module.h from other includes A considerable effort was invested in wiping out module.h from being present in all the other standard includes. This one leaked back in, but once again isn't strictly necessary, so remove it. Signed-off-by: Paul Gortmaker Signed-off-by: David S. Miller --- include/net/netprio_cgroup.h | 1 - 1 file changed, 1 deletion(-) diff --git a/include/net/netprio_cgroup.h b/include/net/netprio_cgroup.h index e503b87c4c1b..7b2d43139c8e 100644 --- a/include/net/netprio_cgroup.h +++ b/include/net/netprio_cgroup.h @@ -13,7 +13,6 @@ #ifndef _NETPRIO_CGROUP_H #define _NETPRIO_CGROUP_H -#include #include #include #include From efc3dbc37412c027e363736b4f4c74ee5e8ecffc Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 24 Jan 2012 17:03:44 -0500 Subject: [PATCH 790/820] rds: Make rds_sock_lock BH rather than IRQ safe. rds_sock_info() triggers locking warnings because we try to perform a local_bh_enable() (via sock_i_ino()) while hardware interrupts are disabled (via taking rds_sock_lock). There is no reason for rds_sock_lock to be a hardware IRQ disabling lock, none of these access paths run in hardware interrupt context. Therefore making it a BH disabling lock is safe and sufficient to fix this bug. Reported-by: Kumar Sanghvi Reported-by: Josh Boyer Signed-off-by: David S. Miller --- net/rds/af_rds.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-) diff --git a/net/rds/af_rds.c b/net/rds/af_rds.c index bb6ad81b671d..424ff622ab5f 100644 --- a/net/rds/af_rds.c +++ b/net/rds/af_rds.c @@ -68,7 +68,6 @@ static int rds_release(struct socket *sock) { struct sock *sk = sock->sk; struct rds_sock *rs; - unsigned long flags; if (!sk) goto out; @@ -94,10 +93,10 @@ static int rds_release(struct socket *sock) rds_rdma_drop_keys(rs); rds_notify_queue_get(rs, NULL); - spin_lock_irqsave(&rds_sock_lock, flags); + spin_lock_bh(&rds_sock_lock); list_del_init(&rs->rs_item); rds_sock_count--; - spin_unlock_irqrestore(&rds_sock_lock, flags); + spin_unlock_bh(&rds_sock_lock); rds_trans_put(rs->rs_transport); @@ -409,7 +408,6 @@ static const struct proto_ops rds_proto_ops = { static int __rds_create(struct socket *sock, struct sock *sk, int protocol) { - unsigned long flags; struct rds_sock *rs; sock_init_data(sock, sk); @@ -426,10 +424,10 @@ static int __rds_create(struct socket *sock, struct sock *sk, int protocol) spin_lock_init(&rs->rs_rdma_lock); rs->rs_rdma_keys = RB_ROOT; - spin_lock_irqsave(&rds_sock_lock, flags); + spin_lock_bh(&rds_sock_lock); list_add_tail(&rs->rs_item, &rds_sock_list); rds_sock_count++; - spin_unlock_irqrestore(&rds_sock_lock, flags); + spin_unlock_bh(&rds_sock_lock); return 0; } @@ -471,12 +469,11 @@ static void rds_sock_inc_info(struct socket *sock, unsigned int len, { struct rds_sock *rs; struct rds_incoming *inc; - unsigned long flags; unsigned int total = 0; len /= sizeof(struct rds_info_message); - spin_lock_irqsave(&rds_sock_lock, flags); + spin_lock_bh(&rds_sock_lock); list_for_each_entry(rs, &rds_sock_list, rs_item) { read_lock(&rs->rs_recv_lock); @@ -492,7 +489,7 @@ static void rds_sock_inc_info(struct socket *sock, unsigned int len, read_unlock(&rs->rs_recv_lock); } - spin_unlock_irqrestore(&rds_sock_lock, flags); + spin_unlock_bh(&rds_sock_lock); lens->nr = total; lens->each = sizeof(struct rds_info_message); @@ -504,11 +501,10 @@ static void rds_sock_info(struct socket *sock, unsigned int len, { struct rds_info_socket sinfo; struct rds_sock *rs; - unsigned long flags; len /= sizeof(struct rds_info_socket); - spin_lock_irqsave(&rds_sock_lock, flags); + spin_lock_bh(&rds_sock_lock); if (len < rds_sock_count) goto out; @@ -529,7 +525,7 @@ out: lens->nr = rds_sock_count; lens->each = sizeof(struct rds_info_socket); - spin_unlock_irqrestore(&rds_sock_lock, flags); + spin_unlock_bh(&rds_sock_lock); } static void rds_exit(void) From f7aa554510f8ff4b1411c31332713ccd8c244ace Mon Sep 17 00:00:00 2001 From: Deepthi Dharwar Date: Thu, 12 Jan 2012 03:05:55 +0000 Subject: [PATCH 791/820] powerpc/cpuidle: Make it a bool, not a tristate As pointed out, asm/system.h has empty inline implementations for update_smt_snooze_delay and pseries_notify_cpuidle_add_cpu, which are used when CONFIG_PSERIES_IDLE is undefined. Since those two functions are used in core power architecture functions (store_smt_snooze_delay at kernel/sysfs.c and smp_xics_setup_cpu at platforms/pseries/smp.c), Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/pseries/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/pseries/Kconfig b/arch/powerpc/platforms/pseries/Kconfig index ae7b6d41fed3..31f22c1f657d 100644 --- a/arch/powerpc/platforms/pseries/Kconfig +++ b/arch/powerpc/platforms/pseries/Kconfig @@ -122,7 +122,7 @@ config DTL Say N if you are unsure. config PSERIES_IDLE - tristate "Cpuidle driver for pSeries platforms" + bool "Cpuidle driver for pSeries platforms" depends on CPU_IDLE depends on PPC_PSERIES default y From 897e01a08c08d86bc76bebb0ca14588b406500e5 Mon Sep 17 00:00:00 2001 From: Christian Kujau Date: Tue, 17 Jan 2012 19:13:05 +0000 Subject: [PATCH 792/820] powerpc/crash: Fix build error without SMP I could not find cpus_in_crash anywhere in the sourcetree, except for arch/powerpc/kernel/crash.c. Moving the definition into the CONFIG_SMP fixes it. Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/crash.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/crash.c b/arch/powerpc/kernel/crash.c index 28be3452e67a..abef75176c07 100644 --- a/arch/powerpc/kernel/crash.c +++ b/arch/powerpc/kernel/crash.c @@ -46,7 +46,6 @@ /* This keeps a track of which one is the crashing cpu. */ int crashing_cpu = -1; -static atomic_t cpus_in_crash; static int time_to_dump; #define CRASH_HANDLER_MAX 3 @@ -66,6 +65,7 @@ static int handle_fault(struct pt_regs *regs) #ifdef CONFIG_SMP +static atomic_t cpus_in_crash; void crash_ipi_callback(struct pt_regs *regs) { static cpumask_t cpus_state_saved = CPU_MASK_NONE; From f7ea82beb24f65e2058fa4029edefa0949e872b1 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 25 Jan 2012 13:31:56 +1100 Subject: [PATCH 793/820] powerpc/powernv: Fix PCI resource handling Recent changes to the handling of PCI resources for host bridges are breaking the PowerNV code for assigning resources on IODA. The root of the problem is that the pci_bus attached to a host bridge no longer has its "legacy" resource pointers populated but only uses the newer list instead. Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/powernv/pci-ioda.c | 43 ++++++++++++++--------- 1 file changed, 27 insertions(+), 16 deletions(-) diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c b/arch/powerpc/platforms/powernv/pci-ioda.c index f31162cfdaa9..5e155dfc4320 100644 --- a/arch/powerpc/platforms/powernv/pci-ioda.c +++ b/arch/powerpc/platforms/powernv/pci-ioda.c @@ -204,11 +204,10 @@ static void __devinit pnv_ioda_offset_bus(struct pci_bus *bus, pr_devel(" -> OBR %s [%x] +%016llx\n", bus->self ? pci_name(bus->self) : "root", flags, offset); - for (i = 0; i < 2; i++) { - r = bus->resource[i]; + pci_bus_for_each_resource(bus, r, i) { if (r && (r->flags & flags)) { - bus->resource[i]->start += offset; - bus->resource[i]->end += offset; + r->start += offset; + r->end += offset; } } list_for_each_entry(dev, &bus->devices, bus_list) @@ -288,12 +287,17 @@ static void __devinit pnv_ioda_calc_bus(struct pci_bus *bus, unsigned int flags, * assignment algorithm is going to be uber-trivial for now, we * can try to be smarter later at filling out holes. */ - start = bus->self ? 0 : bus->resource[bres]->start; - - /* Don't hand out IO 0 */ - if ((flags & IORESOURCE_IO) && !bus->self) - start += 0x1000; - + if (bus->self) { + /* No offset for downstream bridges */ + start = 0; + } else { + /* Offset from the root */ + if (flags & IORESOURCE_IO) + /* Don't hand out IO 0 */ + start = hose->io_resource.start + 0x1000; + else + start = hose->mem_resources[0].start; + } while(!list_empty(&head)) { w = list_first_entry(&head, struct resource_wrap, link); list_del(&w->link); @@ -321,13 +325,20 @@ static void __devinit pnv_ioda_calc_bus(struct pci_bus *bus, unsigned int flags, empty: /* Only setup P2P's, not the PHB itself */ if (bus->self) { - WARN_ON(bus->resource[bres] == NULL); - bus->resource[bres]->start = 0; - bus->resource[bres]->flags = (*size) ? flags : 0; - bus->resource[bres]->end = (*size) ? (*size - 1) : 0; + struct resource *res = bus->resource[bres]; - /* Clear prefetch bus resources for now */ - bus->resource[2]->flags = 0; + if (WARN_ON(res == NULL)) + return; + + /* + * FIXME: We should probably export and call + * pci_bridge_check_ranges() to properly re-initialize + * the PCI portion of the flags here, and to detect + * what the bridge actually supports. + */ + res->start = 0; + res->flags = (*size) ? flags : 0; + res->end = (*size) ? (*size - 1) : 0; } pr_devel("<- CBR %s [%x] *size=%016llx *align=%016llx\n", From 3493c85366ba09c9d0972c919e7123367a39982a Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 25 Jan 2012 13:33:22 +1100 Subject: [PATCH 794/820] powerpc: Fix build on some non-freescale platforms Commit 9deaa53ac7fa373623123aa4f18828dd62292b1a broke build on platforms that use legacy_serial.c without also having CONFIG_SERIAL_8250_FSL enabled due to an unconditional code to a routine in that module. Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/kernel/legacy_serial.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/powerpc/kernel/legacy_serial.c b/arch/powerpc/kernel/legacy_serial.c index 3fea3689527e..bedd12e1cfbc 100644 --- a/arch/powerpc/kernel/legacy_serial.c +++ b/arch/powerpc/kernel/legacy_serial.c @@ -442,8 +442,10 @@ static void __init fixup_port_irq(int index, port->irq = virq; +#ifdef CONFIG_SERIAL_8250_FSL if (of_device_is_compatible(np, "fsl,ns16550")) port->handle_irq = fsl8250_handle_irq; +#endif } static void __init fixup_port_pio(int index, From 598781d71119827b454fd75d46f84755bca6f0c6 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Tue, 24 Jan 2012 18:54:21 +0100 Subject: [PATCH 795/820] drm: Fix authentication kernel crash If the master tries to authenticate a client using drm_authmagic and that client has already closed its drm file descriptor, either wilfully or because it was terminated, the call to drm_authmagic will dereference a stale pointer into kmalloc'ed memory and corrupt it. Typically this results in a hard system hang. This patch fixes that problem by removing any authentication tokens (struct drm_magic_entry) open for a file descriptor when that file descriptor is closed. Signed-off-by: Thomas Hellstrom Reviewed-by: Daniel Vetter Cc: stable@vger.kernel.org Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_auth.c | 6 +++++- drivers/gpu/drm/drm_fops.c | 5 +++++ include/drm/drmP.h | 1 + 3 files changed, 11 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_auth.c b/drivers/gpu/drm/drm_auth.c index 3f46772f0cb2..ba23790450e9 100644 --- a/drivers/gpu/drm/drm_auth.c +++ b/drivers/gpu/drm/drm_auth.c @@ -101,7 +101,7 @@ static int drm_add_magic(struct drm_master *master, struct drm_file *priv, * Searches and unlinks the entry in drm_device::magiclist with the magic * number hash key, while holding the drm_device::struct_mutex lock. */ -static int drm_remove_magic(struct drm_master *master, drm_magic_t magic) +int drm_remove_magic(struct drm_master *master, drm_magic_t magic) { struct drm_magic_entry *pt; struct drm_hash_item *hash; @@ -136,6 +136,8 @@ static int drm_remove_magic(struct drm_master *master, drm_magic_t magic) * If there is a magic number in drm_file::magic then use it, otherwise * searches an unique non-zero magic number and add it associating it with \p * file_priv. + * This ioctl needs protection by the drm_global_mutex, which protects + * struct drm_file::magic and struct drm_magic_entry::priv. */ int drm_getmagic(struct drm_device *dev, void *data, struct drm_file *file_priv) { @@ -173,6 +175,8 @@ int drm_getmagic(struct drm_device *dev, void *data, struct drm_file *file_priv) * \return zero if authentication successed, or a negative number otherwise. * * Checks if \p file_priv is associated with the magic number passed in \arg. + * This ioctl needs protection by the drm_global_mutex, which protects + * struct drm_file::magic and struct drm_magic_entry::priv. */ int drm_authmagic(struct drm_device *dev, void *data, struct drm_file *file_priv) diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c index c00cf154cc0b..6263b0147598 100644 --- a/drivers/gpu/drm/drm_fops.c +++ b/drivers/gpu/drm/drm_fops.c @@ -487,6 +487,11 @@ int drm_release(struct inode *inode, struct file *filp) (long)old_encode_dev(file_priv->minor->device), dev->open_count); + /* Release any auth tokens that might point to this file_priv, + (do that under the drm_global_mutex) */ + if (file_priv->magic) + (void) drm_remove_magic(file_priv->master, file_priv->magic); + /* if the master has gone away we can't do anything with the lock */ if (file_priv->minor->master) drm_master_release(dev, filp); diff --git a/include/drm/drmP.h b/include/drm/drmP.h index 76caa67c22e2..92f0981b5fb8 100644 --- a/include/drm/drmP.h +++ b/include/drm/drmP.h @@ -1328,6 +1328,7 @@ extern int drm_getmagic(struct drm_device *dev, void *data, struct drm_file *file_priv); extern int drm_authmagic(struct drm_device *dev, void *data, struct drm_file *file_priv); +extern int drm_remove_magic(struct drm_master *master, drm_magic_t magic); /* Cache management (drm_cache.c) */ void drm_clflush_pages(struct page *pages[], unsigned long num_pages); From d54fbd49efe5c75bc7cf963bf065aef3fd22417a Mon Sep 17 00:00:00 2001 From: Jerome Glisse Date: Tue, 24 Jan 2012 12:08:52 -0500 Subject: [PATCH 796/820] drm/radeon: silence out possible lock dependency warning Silence out the lock dependency warning by moving bo allocation out of ib mutex protected section. Might lead to useless temporary allocation but it's not harmful as such things only happen at initialization. Signed-off-by: Jerome Glisse Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon_ring.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index e8bc70933d1b..1cb4b941be47 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -204,22 +204,25 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib) int radeon_ib_pool_init(struct radeon_device *rdev) { + struct radeon_sa_manager tmp; int i, r; + r = radeon_sa_bo_manager_init(rdev, &tmp, + RADEON_IB_POOL_SIZE*64*1024, + RADEON_GEM_DOMAIN_GTT); + if (r) { + return r; + } + mutex_lock(&rdev->ib_pool.mutex); if (rdev->ib_pool.ready) { mutex_unlock(&rdev->ib_pool.mutex); + radeon_sa_bo_manager_fini(rdev, &tmp); return 0; } - r = radeon_sa_bo_manager_init(rdev, &rdev->ib_pool.sa_manager, - RADEON_IB_POOL_SIZE*64*1024, - RADEON_GEM_DOMAIN_GTT); - if (r) { - mutex_unlock(&rdev->ib_pool.mutex); - return r; - } - + rdev->ib_pool.sa_manager = tmp; + INIT_LIST_HEAD(&rdev->ib_pool.sa_manager.sa_bo); for (i = 0; i < RADEON_IB_POOL_SIZE; i++) { rdev->ib_pool.ibs[i].fence = NULL; rdev->ib_pool.ibs[i].idx = i; From 9fc04b503df9a34ec1a691225445c5b7dfd022e7 Mon Sep 17 00:00:00 2001 From: Jerome Glisse Date: Mon, 23 Jan 2012 11:52:15 -0500 Subject: [PATCH 797/820] drm/radeon: avoid deadlock if GPU lockup is detected in ib_pool_get If GPU lockup is detected in ib_pool get we are holding the ib_pool mutex that will be needed by the GPU reset code. As ib_pool code is safe to be reentrant from GPU reset code we should not block if we are trying to get the ib pool lock on the behalf of the same userspace caller, thus use the radeon_mutex_lock helper. Signed-off-by: Jerome Glisse Signed-off-by: Dave Airlie --- drivers/gpu/drm/radeon/radeon.h | 84 +++++++++++++------------- drivers/gpu/drm/radeon/radeon_device.c | 2 +- drivers/gpu/drm/radeon/radeon_ring.c | 22 +++---- 3 files changed, 54 insertions(+), 54 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 73e05cb85eca..1668ec1ee770 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -156,6 +156,47 @@ static inline int radeon_atrm_get_bios_chunk(uint8_t *bios, int offset, int len) bool radeon_get_bios(struct radeon_device *rdev); +/* + * Mutex which allows recursive locking from the same process. + */ +struct radeon_mutex { + struct mutex mutex; + struct task_struct *owner; + int level; +}; + +static inline void radeon_mutex_init(struct radeon_mutex *mutex) +{ + mutex_init(&mutex->mutex); + mutex->owner = NULL; + mutex->level = 0; +} + +static inline void radeon_mutex_lock(struct radeon_mutex *mutex) +{ + if (mutex_trylock(&mutex->mutex)) { + /* The mutex was unlocked before, so it's ours now */ + mutex->owner = current; + } else if (mutex->owner != current) { + /* Another process locked the mutex, take it */ + mutex_lock(&mutex->mutex); + mutex->owner = current; + } + /* Otherwise the mutex was already locked by this process */ + + mutex->level++; +} + +static inline void radeon_mutex_unlock(struct radeon_mutex *mutex) +{ + if (--mutex->level > 0) + return; + + mutex->owner = NULL; + mutex_unlock(&mutex->mutex); +} + + /* * Dummy page */ @@ -598,7 +639,7 @@ struct radeon_ib { * mutex protects scheduled_ibs, ready, alloc_bm */ struct radeon_ib_pool { - struct mutex mutex; + struct radeon_mutex mutex; struct radeon_sa_manager sa_manager; struct radeon_ib ibs[RADEON_IB_POOL_SIZE]; bool ready; @@ -1354,47 +1395,6 @@ struct r600_vram_scratch { }; -/* - * Mutex which allows recursive locking from the same process. - */ -struct radeon_mutex { - struct mutex mutex; - struct task_struct *owner; - int level; -}; - -static inline void radeon_mutex_init(struct radeon_mutex *mutex) -{ - mutex_init(&mutex->mutex); - mutex->owner = NULL; - mutex->level = 0; -} - -static inline void radeon_mutex_lock(struct radeon_mutex *mutex) -{ - if (mutex_trylock(&mutex->mutex)) { - /* The mutex was unlocked before, so it's ours now */ - mutex->owner = current; - } else if (mutex->owner != current) { - /* Another process locked the mutex, take it */ - mutex_lock(&mutex->mutex); - mutex->owner = current; - } - /* Otherwise the mutex was already locked by this process */ - - mutex->level++; -} - -static inline void radeon_mutex_unlock(struct radeon_mutex *mutex) -{ - if (--mutex->level > 0) - return; - - mutex->owner = NULL; - mutex_unlock(&mutex->mutex); -} - - /* * Core structure, functions and helpers. */ diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index a811bc64ad5c..cec51a5b69dd 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -720,7 +720,7 @@ int radeon_device_init(struct radeon_device *rdev, /* mutex initialization are all done here so we * can recall function without having locking issues */ radeon_mutex_init(&rdev->cs_mutex); - mutex_init(&rdev->ib_pool.mutex); + radeon_mutex_init(&rdev->ib_pool.mutex); for (i = 0; i < RADEON_NUM_RINGS; ++i) mutex_init(&rdev->ring[i].mutex); mutex_init(&rdev->dc_hw_i2c_mutex); diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index 1cb4b941be47..30a4c5014c8b 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -109,12 +109,12 @@ int radeon_ib_get(struct radeon_device *rdev, int ring, return r; } - mutex_lock(&rdev->ib_pool.mutex); + radeon_mutex_lock(&rdev->ib_pool.mutex); idx = rdev->ib_pool.head_id; retry: if (cretry > 5) { dev_err(rdev->dev, "failed to get an ib after 5 retry\n"); - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); radeon_fence_unref(&fence); return -ENOMEM; } @@ -139,7 +139,7 @@ retry: */ rdev->ib_pool.head_id = (1 + idx); rdev->ib_pool.head_id &= (RADEON_IB_POOL_SIZE - 1); - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); return 0; } } @@ -158,7 +158,7 @@ retry: } idx = (idx + 1) & (RADEON_IB_POOL_SIZE - 1); } - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); radeon_fence_unref(&fence); return r; } @@ -171,12 +171,12 @@ void radeon_ib_free(struct radeon_device *rdev, struct radeon_ib **ib) if (tmp == NULL) { return; } - mutex_lock(&rdev->ib_pool.mutex); + radeon_mutex_lock(&rdev->ib_pool.mutex); if (tmp->fence && !tmp->fence->emitted) { radeon_sa_bo_free(rdev, &tmp->sa_bo); radeon_fence_unref(&tmp->fence); } - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); } int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib) @@ -214,9 +214,9 @@ int radeon_ib_pool_init(struct radeon_device *rdev) return r; } - mutex_lock(&rdev->ib_pool.mutex); + radeon_mutex_lock(&rdev->ib_pool.mutex); if (rdev->ib_pool.ready) { - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); radeon_sa_bo_manager_fini(rdev, &tmp); return 0; } @@ -239,7 +239,7 @@ int radeon_ib_pool_init(struct radeon_device *rdev) if (radeon_debugfs_ring_init(rdev)) { DRM_ERROR("Failed to register debugfs file for rings !\n"); } - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); return 0; } @@ -247,7 +247,7 @@ void radeon_ib_pool_fini(struct radeon_device *rdev) { unsigned i; - mutex_lock(&rdev->ib_pool.mutex); + radeon_mutex_lock(&rdev->ib_pool.mutex); if (rdev->ib_pool.ready) { for (i = 0; i < RADEON_IB_POOL_SIZE; i++) { radeon_sa_bo_free(rdev, &rdev->ib_pool.ibs[i].sa_bo); @@ -256,7 +256,7 @@ void radeon_ib_pool_fini(struct radeon_device *rdev) radeon_sa_bo_manager_fini(rdev, &rdev->ib_pool.sa_manager); rdev->ib_pool.ready = false; } - mutex_unlock(&rdev->ib_pool.mutex); + radeon_mutex_unlock(&rdev->ib_pool.mutex); } int radeon_ib_pool_start(struct radeon_device *rdev) From 9b025eb3a89e041bab6698e3858706be2385d692 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 11 Jan 2012 18:52:10 +0000 Subject: [PATCH 798/820] xfs: Fix missing xfs_iunlock() on error recovery path in xfs_readlink() Commit b52a360b forgot to call xfs_iunlock() when it detected corrupted symplink and bailed out. Fix it by jumping to 'out' instead of doing return. CC: stable@kernel.org CC: Carlos Maiolino Signed-off-by: Jan Kara Reviewed-by: Alex Elder Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_vnodeops.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 0cf52da9d246..ebdb88840a47 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -131,7 +131,8 @@ xfs_readlink( __func__, (unsigned long long) ip->i_ino, (long long) pathlen); ASSERT(0); - return XFS_ERROR(EFSCORRUPTED); + error = XFS_ERROR(EFSCORRUPTED); + goto out; } From 9f1feed2e16652a6e599ed4a73b4c501bb3d4568 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Wed, 25 Jan 2012 15:34:22 +1000 Subject: [PATCH 799/820] drm/ttm: fix two regressions since move_notify changes Both changes in dc97b3409a790d2a21aac6e5cdb99558b5944119 cause serious regressions in the nouveau driver. move_notify() was originally able to presume that bo->mem is the old node, and new_mem is the new node. The above commit moves the call to move_notify() to after move() has been done, which means that now, sometimes, new_mem isn't the new node at all, bo->mem is, and new_mem points at a stale, possibly-just-been-killed-by-move node. This is clearly not a good situation. This patch reverts this change, and replaces it with a cleanup in the move() failure path instead. The second issue is that the call to move_notify() from cleanup_memtype_use() causes the TTM ghost objects to get passed into the driver. This is clearly bad as the driver knows nothing about these "fake" TTM BOs, and ends up accessing uninitialised memory. I worked around this in nouveau's move_notify() hook by ensuring the BO destructor was nouveau's. I don't particularly like this solution, and would rather TTM never pass the driver these objects. However, I don't clearly understand the reason why we're calling move_notify() here anyway and am happy to work around the problem in nouveau instead of breaking the behaviour expected by other drivers. Signed-off-by: Ben Skeggs Reviewed-by: Thomas Hellstrom Cc: Jerome Glisse Signed-off-by: Dave Airlie --- drivers/gpu/drm/nouveau/nouveau_bo.c | 4 ++++ drivers/gpu/drm/ttm/ttm_bo.c | 17 +++++++++++++---- 2 files changed, 17 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 724b41a2b9e9..ec54364ac828 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -812,6 +812,10 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem) struct nouveau_bo *nvbo = nouveau_bo(bo); struct nouveau_vma *vma; + /* ttm can now (stupidly) pass the driver bos it didn't create... */ + if (bo->destroy != nouveau_bo_del_ttm) + return; + list_for_each_entry(vma, &nvbo->vma_list, head) { if (new_mem && new_mem->mem_type == TTM_PL_VRAM) { nouveau_vm_map(vma, new_mem->mm_node); diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 2f0eab66ece6..7c3a57de8187 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -404,6 +404,9 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo, } } + if (bdev->driver->move_notify) + bdev->driver->move_notify(bo, mem); + if (!(old_man->flags & TTM_MEMTYPE_FLAG_FIXED) && !(new_man->flags & TTM_MEMTYPE_FLAG_FIXED)) ret = ttm_bo_move_ttm(bo, evict, no_wait_reserve, no_wait_gpu, mem); @@ -413,11 +416,17 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo, else ret = ttm_bo_move_memcpy(bo, evict, no_wait_reserve, no_wait_gpu, mem); - if (ret) - goto out_err; + if (ret) { + if (bdev->driver->move_notify) { + struct ttm_mem_reg tmp_mem = *mem; + *mem = bo->mem; + bo->mem = tmp_mem; + bdev->driver->move_notify(bo, mem); + bo->mem = *mem; + } - if (bdev->driver->move_notify) - bdev->driver->move_notify(bo, mem); + goto out_err; + } moved: if (bo->evicted) { From 140941987f8f85ee7c7d4592ecb6667b32f3485d Mon Sep 17 00:00:00 2001 From: Dustin Kirkland Date: Wed, 7 Dec 2011 08:56:49 -0600 Subject: [PATCH 800/820] MAINTAINERS: Update eCryptfs maintainer address Update my email address in MAINTAINERS. Signed-off-by: Dustin Kirkland Signed-off-by: Tyler Hicks --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 89b70df91f4f..b63c181b1ed6 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2391,7 +2391,7 @@ F: net/bridge/netfilter/ebt*.c ECRYPT FILE SYSTEM M: Tyler Hicks -M: Dustin Kirkland +M: Dustin Kirkland L: ecryptfs@vger.kernel.org W: https://launchpad.net/ecryptfs S: Supported From 30373dc0c87ffef68d5628e77d56ffb1fa22e1ee Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Thu, 12 Jan 2012 16:31:55 +0100 Subject: [PATCH 801/820] ecryptfs: Improve metadata read failure logging Print inode on metadata read failure. The only real way of dealing with metadata read failures is to delete the underlying file system file. Having the inode allows one to 'find . -inum INODE`. [tyhicks@canonical.com: Removed some minor not-for-stable parts] Signed-off-by: Tim Gardner Reviewed-by: Kees Cook Cc: stable@vger.kernel.org Signed-off-by: Tyler Hicks --- fs/ecryptfs/crypto.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 2a834255c75d..2bf52033538b 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1620,7 +1620,8 @@ int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry) rc = ecryptfs_read_xattr_region(page_virt, ecryptfs_inode); if (rc) { printk(KERN_DEBUG "Valid eCryptfs headers not found in " - "file header region or xattr region\n"); + "file header region or xattr region, inode %lu\n", + ecryptfs_inode->i_ino); rc = -EINVAL; goto out; } @@ -1629,7 +1630,8 @@ int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry) ECRYPTFS_DONT_VALIDATE_HEADER_SIZE); if (rc) { printk(KERN_DEBUG "Valid eCryptfs headers not found in " - "file xattr region either\n"); + "file xattr region either, inode %lu\n", + ecryptfs_inode->i_ino); rc = -EINVAL; } if (crypt_stat->mount_crypt_stat->flags @@ -1640,7 +1642,8 @@ int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry) "crypto metadata only in the extended attribute " "region, but eCryptfs was mounted without " "xattr support enabled. eCryptfs will not treat " - "this like an encrypted file.\n"); + "this like an encrypted file, inode %lu\n", + ecryptfs_inode->i_ino); rc = -EINVAL; } } From bb4503615d95d6826b7907986ad574e3157877e8 Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Thu, 12 Jan 2012 16:31:55 +0100 Subject: [PATCH 802/820] ecryptfs: Remove unnecessary variable initialization Removes unneeded variable initialization in ecryptfs_read_metadata(). Also adds a small comment to help explain metadata reading logic. [tyhicks@canonical.com: Pulled out of for-stable patch and wrote commit msg] Signed-off-by: Tim Gardner Signed-off-by: Tyler Hicks --- fs/ecryptfs/crypto.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 2bf52033538b..ff981503b3e3 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1590,8 +1590,8 @@ int ecryptfs_read_and_validate_xattr_region(struct dentry *dentry, */ int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry) { - int rc = 0; - char *page_virt = NULL; + int rc; + char *page_virt; struct inode *ecryptfs_inode = ecryptfs_dentry->d_inode; struct ecryptfs_crypt_stat *crypt_stat = &ecryptfs_inode_to_private(ecryptfs_inode)->crypt_stat; @@ -1616,6 +1616,7 @@ int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry) ecryptfs_dentry, ECRYPTFS_VALIDATE_HEADER_SIZE); if (rc) { + /* metadata is not in the file header, so try xattrs */ memset(page_virt, 0, PAGE_CACHE_SIZE); rc = ecryptfs_read_xattr_region(page_virt, ecryptfs_inode); if (rc) { From db10e556518eb9d21ee92ff944530d84349684f4 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 12 Jan 2012 11:30:44 +0100 Subject: [PATCH 803/820] eCryptfs: Sanitize write counts of /dev/ecryptfs A malicious count value specified when writing to /dev/ecryptfs may result in a a very large kernel memory allocation. This patch peeks at the specified packet payload size, adds that to the size of the packet headers and compares the result with the write count value. The resulting maximum memory allocation size is approximately 532 bytes. Signed-off-by: Tyler Hicks Reported-by: Sasha Levin Cc: --- fs/ecryptfs/miscdev.c | 56 +++++++++++++++++++++++++++++-------------- 1 file changed, 38 insertions(+), 18 deletions(-) diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index 940a82e63dc3..0dc5a3d554a4 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -409,11 +409,47 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, ssize_t sz = 0; char *data; uid_t euid = current_euid(); + unsigned char packet_size_peek[3]; int rc; - if (count == 0) + if (count == 0) { goto out; + } else if (count == (1 + 4)) { + /* Likely a harmless MSG_HELO or MSG_QUIT - no packet length */ + goto memdup; + } else if (count < (1 + 4 + 1) + || count > (1 + 4 + 2 + sizeof(struct ecryptfs_message) + 4 + + ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES)) { + printk(KERN_WARNING "%s: Acceptable packet size range is " + "[%d-%lu], but amount of data written is [%zu].", + __func__, (1 + 4 + 1), + (1 + 4 + 2 + sizeof(struct ecryptfs_message) + 4 + + ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES), count); + return -EINVAL; + } + if (copy_from_user(packet_size_peek, (buf + 1 + 4), + sizeof(packet_size_peek))) { + printk(KERN_WARNING "%s: Error while inspecting packet size\n", + __func__); + return -EFAULT; + } + + rc = ecryptfs_parse_packet_length(packet_size_peek, &packet_size, + &packet_size_length); + if (rc) { + printk(KERN_WARNING "%s: Error parsing packet length; " + "rc = [%d]\n", __func__, rc); + return rc; + } + + if ((1 + 4 + packet_size_length + packet_size) != count) { + printk(KERN_WARNING "%s: Invalid packet size [%zu]\n", __func__, + packet_size); + return -EINVAL; + } + +memdup: data = memdup_user(buf, count); if (IS_ERR(data)) { printk(KERN_ERR "%s: memdup_user returned error [%ld]\n", @@ -435,23 +471,7 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, } memcpy(&counter_nbo, &data[i], 4); seq = be32_to_cpu(counter_nbo); - i += 4; - rc = ecryptfs_parse_packet_length(&data[i], &packet_size, - &packet_size_length); - if (rc) { - printk(KERN_WARNING "%s: Error parsing packet length; " - "rc = [%d]\n", __func__, rc); - goto out_free; - } - i += packet_size_length; - if ((1 + 4 + packet_size_length + packet_size) != count) { - printk(KERN_WARNING "%s: (1 + packet_size_length([%zd])" - " + packet_size([%zd]))([%zd]) != " - "count([%zd]). Invalid packet format.\n", - __func__, packet_size_length, packet_size, - (1 + packet_size_length + packet_size), count); - goto out_free; - } + i += 4 + packet_size_length; rc = ecryptfs_miscdev_response(&data[i], packet_size, euid, current_user_ns(), task_pid(current), seq); From 7f133504249afa48618becac546ce3c35c9f0185 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Sat, 14 Jan 2012 15:51:37 +0100 Subject: [PATCH 804/820] eCryptfs: Report errors in writes to /dev/ecryptfs Errors in writes to /dev/ecryptfs were being incorrectly reported by returning 0 or the value of the original write count. This patch clears up the return code assignment in error paths. Signed-off-by: Tyler Hicks --- fs/ecryptfs/miscdev.c | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-) diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index 0dc5a3d554a4..1145c58103e2 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -406,14 +406,13 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, __be32 counter_nbo; u32 seq; size_t packet_size, packet_size_length, i; - ssize_t sz = 0; char *data; uid_t euid = current_euid(); unsigned char packet_size_peek[3]; - int rc; + ssize_t rc; if (count == 0) { - goto out; + return 0; } else if (count == (1 + 4)) { /* Likely a harmless MSG_HELO or MSG_QUIT - no packet length */ goto memdup; @@ -439,7 +438,7 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, &packet_size_length); if (rc) { printk(KERN_WARNING "%s: Error parsing packet length; " - "rc = [%d]\n", __func__, rc); + "rc = [%zd]\n", __func__, rc); return rc; } @@ -454,9 +453,8 @@ memdup: if (IS_ERR(data)) { printk(KERN_ERR "%s: memdup_user returned error [%ld]\n", __func__, PTR_ERR(data)); - goto out; + return PTR_ERR(data); } - sz = count; i = 0; switch (data[i++]) { case ECRYPTFS_MSG_RESPONSE: @@ -467,6 +465,7 @@ memdup: __func__, (1 + 4 + 1 + sizeof(struct ecryptfs_message)), count); + rc = -EINVAL; goto out_free; } memcpy(&counter_nbo, &data[i], 4); @@ -475,10 +474,12 @@ memdup: rc = ecryptfs_miscdev_response(&data[i], packet_size, euid, current_user_ns(), task_pid(current), seq); - if (rc) + if (rc) { printk(KERN_WARNING "%s: Failed to deliver miscdev " - "response to requesting operation; rc = [%d]\n", + "response to requesting operation; rc = [%zd]\n", __func__, rc); + goto out_free; + } break; case ECRYPTFS_MSG_HELO: case ECRYPTFS_MSG_QUIT: @@ -487,12 +488,13 @@ memdup: ecryptfs_printk(KERN_WARNING, "Dropping miscdev " "message of unrecognized type [%d]\n", data[0]); - break; + rc = -EINVAL; + goto out_free; } + rc = count; out_free: kfree(data); -out: - return sz; + return rc; } From 48399c0b0e6172888a2e2e36df1595ab1e049ba8 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Sat, 14 Jan 2012 16:46:46 +0100 Subject: [PATCH 805/820] eCryptfs: Replace miscdev read/write magic numbers ecryptfs_miscdev_read() and ecryptfs_miscdev_write() contained many magic numbers for specifying packet header field sizes and offsets. This patch defines those values and replaces the magic values. Signed-off-by: Tyler Hicks --- fs/ecryptfs/ecryptfs_kernel.h | 5 ++ fs/ecryptfs/keystore.c | 5 +- fs/ecryptfs/miscdev.c | 86 +++++++++++++++++++---------------- 3 files changed, 55 insertions(+), 41 deletions(-) diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index a9f29b12fbf2..a2362df58ae8 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -151,6 +151,11 @@ ecryptfs_get_key_payload_data(struct key *key) * dentry name */ #define ECRYPTFS_TAG_73_PACKET_TYPE 0x49 /* FEK-encrypted filename as * metadata */ +#define ECRYPTFS_MIN_PKT_LEN_SIZE 1 /* Min size to specify packet length */ +#define ECRYPTFS_MAX_PKT_LEN_SIZE 2 /* Pass at least this many bytes to + * ecryptfs_parse_packet_length() and + * ecryptfs_write_packet_length() + */ /* Constraint: ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES >= * ECRYPTFS_MAX_IV_BYTES */ #define ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES 16 diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index ac1ad48c2376..8e3b943e330f 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -109,7 +109,7 @@ int ecryptfs_parse_packet_length(unsigned char *data, size_t *size, (*size) += ((unsigned char)(data[1]) + 192); (*length_size) = 2; } else if (data[0] == 255) { - /* Five-byte length; we're not supposed to see this */ + /* If support is added, adjust ECRYPTFS_MAX_PKT_LEN_SIZE */ ecryptfs_printk(KERN_ERR, "Five-byte packet length not " "supported\n"); rc = -EINVAL; @@ -126,7 +126,7 @@ out: /** * ecryptfs_write_packet_length * @dest: The byte array target into which to write the length. Must - * have at least 5 bytes allocated. + * have at least ECRYPTFS_MAX_PKT_LEN_SIZE bytes allocated. * @size: The length to write. * @packet_size_length: The number of bytes used to encode the packet * length is written to this address. @@ -146,6 +146,7 @@ int ecryptfs_write_packet_length(char *dest, size_t size, dest[1] = ((size - 192) % 256); (*packet_size_length) = 2; } else { + /* If support is added, adjust ECRYPTFS_MAX_PKT_LEN_SIZE */ rc = -EINVAL; ecryptfs_printk(KERN_WARNING, "Unsupported packet size: [%zd]\n", size); diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index 1145c58103e2..349209dc6a91 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -218,6 +218,29 @@ out_unlock: return rc; } +/* + * miscdevfs packet format: + * Octet 0: Type + * Octets 1-4: network byte order msg_ctx->counter + * Octets 5-N0: Size of struct ecryptfs_message to follow + * Octets N0-N1: struct ecryptfs_message (including data) + * + * Octets 5-N1 not written if the packet type does not include a message + */ +#define PKT_TYPE_SIZE 1 +#define PKT_CTR_SIZE 4 +#define MIN_NON_MSG_PKT_SIZE (PKT_TYPE_SIZE + PKT_CTR_SIZE) +#define MIN_MSG_PKT_SIZE (PKT_TYPE_SIZE + PKT_CTR_SIZE \ + + ECRYPTFS_MIN_PKT_LEN_SIZE) +/* 4 + ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES comes from tag 65 packet format */ +#define MAX_MSG_PKT_SIZE (PKT_TYPE_SIZE + PKT_CTR_SIZE \ + + ECRYPTFS_MAX_PKT_LEN_SIZE \ + + sizeof(struct ecryptfs_message) \ + + 4 + ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES) +#define PKT_TYPE_OFFSET 0 +#define PKT_CTR_OFFSET PKT_TYPE_SIZE +#define PKT_LEN_OFFSET (PKT_TYPE_SIZE + PKT_CTR_SIZE) + /** * ecryptfs_miscdev_read - format and send message from queue * @file: fs/ecryptfs/euid miscdevfs handle (ignored) @@ -237,7 +260,7 @@ ecryptfs_miscdev_read(struct file *file, char __user *buf, size_t count, struct ecryptfs_daemon *daemon; struct ecryptfs_msg_ctx *msg_ctx; size_t packet_length_size; - char packet_length[3]; + char packet_length[ECRYPTFS_MAX_PKT_LEN_SIZE]; size_t i; size_t total_length; uid_t euid = current_euid(); @@ -305,15 +328,8 @@ check_list: packet_length_size = 0; msg_ctx->msg_size = 0; } - /* miscdevfs packet format: - * Octet 0: Type - * Octets 1-4: network byte order msg_ctx->counter - * Octets 5-N0: Size of struct ecryptfs_message to follow - * Octets N0-N1: struct ecryptfs_message (including data) - * - * Octets 5-N1 not written if the packet type does not - * include a message */ - total_length = (1 + 4 + packet_length_size + msg_ctx->msg_size); + total_length = (PKT_TYPE_SIZE + PKT_CTR_SIZE + packet_length_size + + msg_ctx->msg_size); if (count < total_length) { rc = 0; printk(KERN_WARNING "%s: Only given user buffer of " @@ -324,9 +340,10 @@ check_list: rc = -EFAULT; if (put_user(msg_ctx->type, buf)) goto out_unlock_msg_ctx; - if (put_user(cpu_to_be32(msg_ctx->counter), (__be32 __user *)(buf + 1))) + if (put_user(cpu_to_be32(msg_ctx->counter), + (__be32 __user *)(&buf[PKT_CTR_OFFSET]))) goto out_unlock_msg_ctx; - i = 5; + i = PKT_TYPE_SIZE + PKT_CTR_SIZE; if (msg_ctx->msg) { if (copy_to_user(&buf[i], packet_length, packet_length_size)) goto out_unlock_msg_ctx; @@ -391,12 +408,6 @@ out: * @count: Amount of data in @buf * @ppos: Pointer to offset in file (ignored) * - * miscdevfs packet format: - * Octet 0: Type - * Octets 1-4: network byte order msg_ctx->counter (0's for non-response) - * Octets 5-N0: Size of struct ecryptfs_message to follow - * Octets N0-N1: struct ecryptfs_message (including data) - * * Returns the number of bytes read from @buf */ static ssize_t @@ -405,29 +416,25 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, { __be32 counter_nbo; u32 seq; - size_t packet_size, packet_size_length, i; + size_t packet_size, packet_size_length; char *data; uid_t euid = current_euid(); - unsigned char packet_size_peek[3]; + unsigned char packet_size_peek[ECRYPTFS_MAX_PKT_LEN_SIZE]; ssize_t rc; if (count == 0) { return 0; - } else if (count == (1 + 4)) { + } else if (count == MIN_NON_MSG_PKT_SIZE) { /* Likely a harmless MSG_HELO or MSG_QUIT - no packet length */ goto memdup; - } else if (count < (1 + 4 + 1) - || count > (1 + 4 + 2 + sizeof(struct ecryptfs_message) + 4 - + ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES)) { + } else if (count < MIN_MSG_PKT_SIZE || count > MAX_MSG_PKT_SIZE) { printk(KERN_WARNING "%s: Acceptable packet size range is " "[%d-%lu], but amount of data written is [%zu].", - __func__, (1 + 4 + 1), - (1 + 4 + 2 + sizeof(struct ecryptfs_message) + 4 - + ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES), count); + __func__, MIN_MSG_PKT_SIZE, MAX_MSG_PKT_SIZE, count); return -EINVAL; } - if (copy_from_user(packet_size_peek, (buf + 1 + 4), + if (copy_from_user(packet_size_peek, &buf[PKT_LEN_OFFSET], sizeof(packet_size_peek))) { printk(KERN_WARNING "%s: Error while inspecting packet size\n", __func__); @@ -442,7 +449,8 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, return rc; } - if ((1 + 4 + packet_size_length + packet_size) != count) { + if ((PKT_TYPE_SIZE + PKT_CTR_SIZE + packet_size_length + packet_size) + != count) { printk(KERN_WARNING "%s: Invalid packet size [%zu]\n", __func__, packet_size); return -EINVAL; @@ -455,25 +463,25 @@ memdup: __func__, PTR_ERR(data)); return PTR_ERR(data); } - i = 0; - switch (data[i++]) { + switch (data[PKT_TYPE_OFFSET]) { case ECRYPTFS_MSG_RESPONSE: - if (count < (1 + 4 + 1 + sizeof(struct ecryptfs_message))) { + if (count < (MIN_MSG_PKT_SIZE + + sizeof(struct ecryptfs_message))) { printk(KERN_WARNING "%s: Minimum acceptable packet " "size is [%zd], but amount of data written is " "only [%zd]. Discarding response packet.\n", __func__, - (1 + 4 + 1 + sizeof(struct ecryptfs_message)), - count); + (MIN_MSG_PKT_SIZE + + sizeof(struct ecryptfs_message)), count); rc = -EINVAL; goto out_free; } - memcpy(&counter_nbo, &data[i], 4); + memcpy(&counter_nbo, &data[PKT_CTR_OFFSET], PKT_CTR_SIZE); seq = be32_to_cpu(counter_nbo); - i += 4 + packet_size_length; - rc = ecryptfs_miscdev_response(&data[i], packet_size, - euid, current_user_ns(), - task_pid(current), seq); + rc = ecryptfs_miscdev_response( + &data[PKT_LEN_OFFSET + packet_size_length], + packet_size, euid, current_user_ns(), + task_pid(current), seq); if (rc) { printk(KERN_WARNING "%s: Failed to deliver miscdev " "response to requesting operation; rc = [%zd]\n", From 684a3ff7e69acc7c678d1a1394fe9e757993fd34 Mon Sep 17 00:00:00 2001 From: Li Wang Date: Thu, 19 Jan 2012 09:44:36 +0800 Subject: [PATCH 806/820] eCryptfs: Infinite loop due to overflow in ecryptfs_write() ecryptfs_write() can enter an infinite loop when truncating a file to a size larger than 4G. This only happens on architectures where size_t is represented by 32 bits. This was caused by a size_t overflow due to it incorrectly being used to store the result of a calculation which uses potentially large values of type loff_t. [tyhicks@canonical.com: rewrite subject and commit message] Signed-off-by: Li Wang Signed-off-by: Yunchuan Wen Reviewed-by: Cong Wang Cc: Signed-off-by: Tyler Hicks --- fs/ecryptfs/read_write.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/ecryptfs/read_write.c b/fs/ecryptfs/read_write.c index 3745f7c2b9c2..ec3d9368dc5b 100644 --- a/fs/ecryptfs/read_write.c +++ b/fs/ecryptfs/read_write.c @@ -130,13 +130,13 @@ int ecryptfs_write(struct inode *ecryptfs_inode, char *data, loff_t offset, pgoff_t ecryptfs_page_idx = (pos >> PAGE_CACHE_SHIFT); size_t start_offset_in_page = (pos & ~PAGE_CACHE_MASK); size_t num_bytes = (PAGE_CACHE_SIZE - start_offset_in_page); - size_t total_remaining_bytes = ((offset + size) - pos); + loff_t total_remaining_bytes = ((offset + size) - pos); if (num_bytes > total_remaining_bytes) num_bytes = total_remaining_bytes; if (pos < offset) { /* remaining zeros to write, up to destination offset */ - size_t total_remaining_zeros = (offset - pos); + loff_t total_remaining_zeros = (offset - pos); if (num_bytes > total_remaining_zeros) num_bytes = total_remaining_zeros; From 5e6f0d769017cc49207ef56996e42363ec26c1f0 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Wed, 18 Jan 2012 18:30:04 -0600 Subject: [PATCH 807/820] eCryptfs: Make truncate path killable ecryptfs_write() handles the truncation of eCryptfs inodes. It grabs a page, zeroes out the appropriate portions, and then encrypts the page before writing it to the lower filesystem. It was unkillable and due to the lack of sparse file support could result in tying up a large portion of system resources, while encrypting pages of zeros, with no way for the truncate operation to be stopped from userspace. This patch adds the ability for ecryptfs_write() to detect a pending fatal signal and return as gracefully as possible. The intent is to leave the lower file in a useable state, while still allowing a user to break out of the encryption loop. If a pending fatal signal is detected, the eCryptfs inode size is updated to reflect the modified inode size and then -EINTR is returned. Signed-off-by: Tyler Hicks Cc: --- fs/ecryptfs/read_write.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/fs/ecryptfs/read_write.c b/fs/ecryptfs/read_write.c index ec3d9368dc5b..608c1c3fde1b 100644 --- a/fs/ecryptfs/read_write.c +++ b/fs/ecryptfs/read_write.c @@ -132,6 +132,11 @@ int ecryptfs_write(struct inode *ecryptfs_inode, char *data, loff_t offset, size_t num_bytes = (PAGE_CACHE_SIZE - start_offset_in_page); loff_t total_remaining_bytes = ((offset + size) - pos); + if (fatal_signal_pending(current)) { + rc = -EINTR; + break; + } + if (num_bytes > total_remaining_bytes) num_bytes = total_remaining_bytes; if (pos < offset) { @@ -193,15 +198,19 @@ int ecryptfs_write(struct inode *ecryptfs_inode, char *data, loff_t offset, } pos += num_bytes; } - if ((offset + size) > ecryptfs_file_size) { - i_size_write(ecryptfs_inode, (offset + size)); + if (pos > ecryptfs_file_size) { + i_size_write(ecryptfs_inode, pos); if (crypt_stat->flags & ECRYPTFS_ENCRYPTED) { - rc = ecryptfs_write_inode_size_to_metadata( + int rc2; + + rc2 = ecryptfs_write_inode_size_to_metadata( ecryptfs_inode); - if (rc) { + if (rc2) { printk(KERN_ERR "Problem with " "ecryptfs_write_inode_size_to_metadata; " - "rc = [%d]\n", rc); + "rc = [%d]\n", rc2); + if (!rc) + rc = rc2; goto out; } } From a261a03904849c3df50bd0300efb7fb3f865137d Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 19 Jan 2012 20:33:44 -0600 Subject: [PATCH 808/820] eCryptfs: Check inode changes in setattr Most filesystems call inode_change_ok() very early in ->setattr(), but eCryptfs didn't call it at all. It allowed the lower filesystem to make the call in its ->setattr() function. Then, eCryptfs would copy the appropriate inode attributes from the lower inode to the eCryptfs inode. This patch changes that and actually calls inode_change_ok() on the eCryptfs inode, fairly early in ecryptfs_setattr(). Ideally, the call would happen earlier in ecryptfs_setattr(), but there are some possible inode initialization steps that must happen first. Since the call was already being made on the lower inode, the change in functionality should be minimal, except for the case of a file extending truncate call. In that case, inode_newsize_ok() was never being called on the eCryptfs inode. Rather than inode_newsize_ok() catching maximum file size errors early on, eCryptfs would encrypt zeroed pages and write them to the lower filesystem until the lower filesystem's write path caught the error in generic_write_checks(). This patch introduces a new function, called ecryptfs_inode_newsize_ok(), which checks if the new lower file size is within the appropriate limits when the truncate operation will be growing the lower file. In summary this change prevents eCryptfs truncate operations (and the resulting page encryptions), which would exceed the lower filesystem limits or FSIZE rlimits, from ever starting. Signed-off-by: Tyler Hicks Reviewed-by: Li Wang Cc: --- fs/ecryptfs/inode.c | 48 +++++++++++++++++++++++++++++++++------------ 1 file changed, 36 insertions(+), 12 deletions(-) diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 19a8ca4ab1dd..19892d7d2ed1 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -822,18 +822,6 @@ static int truncate_upper(struct dentry *dentry, struct iattr *ia, size_t num_zeros = (PAGE_CACHE_SIZE - (ia->ia_size & ~PAGE_CACHE_MASK)); - - /* - * XXX(truncate) this should really happen at the begginning - * of ->setattr. But the code is too messy to that as part - * of a larger patch. ecryptfs is also totally missing out - * on the inode_change_ok check at the beginning of - * ->setattr while would include this. - */ - rc = inode_newsize_ok(inode, ia->ia_size); - if (rc) - goto out; - if (!(crypt_stat->flags & ECRYPTFS_ENCRYPTED)) { truncate_setsize(inode, ia->ia_size); lower_ia->ia_size = ia->ia_size; @@ -883,6 +871,28 @@ out: return rc; } +static int ecryptfs_inode_newsize_ok(struct inode *inode, loff_t offset) +{ + struct ecryptfs_crypt_stat *crypt_stat; + loff_t lower_oldsize, lower_newsize; + + crypt_stat = &ecryptfs_inode_to_private(inode)->crypt_stat; + lower_oldsize = upper_size_to_lower_size(crypt_stat, + i_size_read(inode)); + lower_newsize = upper_size_to_lower_size(crypt_stat, offset); + if (lower_newsize > lower_oldsize) { + /* + * The eCryptfs inode and the new *lower* size are mixed here + * because we may not have the lower i_mutex held and/or it may + * not be appropriate to call inode_newsize_ok() with inodes + * from other filesystems. + */ + return inode_newsize_ok(inode, lower_newsize); + } + + return 0; +} + /** * ecryptfs_truncate * @dentry: The ecryptfs layer dentry @@ -899,6 +909,10 @@ int ecryptfs_truncate(struct dentry *dentry, loff_t new_length) struct iattr lower_ia = { .ia_valid = 0 }; int rc; + rc = ecryptfs_inode_newsize_ok(dentry->d_inode, new_length); + if (rc) + return rc; + rc = truncate_upper(dentry, &ia, &lower_ia); if (!rc && lower_ia.ia_valid & ATTR_SIZE) { struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry); @@ -978,6 +992,16 @@ static int ecryptfs_setattr(struct dentry *dentry, struct iattr *ia) } } mutex_unlock(&crypt_stat->cs_mutex); + + rc = inode_change_ok(inode, ia); + if (rc) + goto out; + if (ia->ia_valid & ATTR_SIZE) { + rc = ecryptfs_inode_newsize_ok(inode, ia->ia_size); + if (rc) + goto out; + } + if (S_ISREG(inode->i_mode)) { rc = filemap_write_and_wait(inode->i_mapping); if (rc) From f2cb933501ebc066bf3c4b1836fd8428f8fe9863 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Wed, 18 Jan 2012 15:09:43 -0600 Subject: [PATCH 809/820] eCryptfs: Remove unused ecryptfs_read() ecryptfs_read() has been ifdef'ed out for years now and it was apparently unused before then. It is time to get rid of it for good. Signed-off-by: Tyler Hicks --- fs/ecryptfs/read_write.c | 73 ---------------------------------------- 1 file changed, 73 deletions(-) diff --git a/fs/ecryptfs/read_write.c b/fs/ecryptfs/read_write.c index 608c1c3fde1b..5c0106f75775 100644 --- a/fs/ecryptfs/read_write.c +++ b/fs/ecryptfs/read_write.c @@ -282,76 +282,3 @@ int ecryptfs_read_lower_page_segment(struct page *page_for_ecryptfs, flush_dcache_page(page_for_ecryptfs); return rc; } - -#if 0 -/** - * ecryptfs_read - * @data: The virtual address into which to write the data read (and - * possibly decrypted) from the lower file - * @offset: The offset in the decrypted view of the file from which to - * read into @data - * @size: The number of bytes to read into @data - * @ecryptfs_file: The eCryptfs file from which to read - * - * Read an arbitrary amount of data from an arbitrary location in the - * eCryptfs page cache. This is done on an extent-by-extent basis; - * individual extents are decrypted and read from the lower page - * cache (via VFS reads). This function takes care of all the - * address translation to locations in the lower filesystem. - * - * Returns zero on success; non-zero otherwise - */ -int ecryptfs_read(char *data, loff_t offset, size_t size, - struct file *ecryptfs_file) -{ - struct inode *ecryptfs_inode = ecryptfs_file->f_dentry->d_inode; - struct page *ecryptfs_page; - char *ecryptfs_page_virt; - loff_t ecryptfs_file_size = i_size_read(ecryptfs_inode); - loff_t data_offset = 0; - loff_t pos; - int rc = 0; - - if ((offset + size) > ecryptfs_file_size) { - rc = -EINVAL; - printk(KERN_ERR "%s: Attempt to read data past the end of the " - "file; offset = [%lld]; size = [%td]; " - "ecryptfs_file_size = [%lld]\n", - __func__, offset, size, ecryptfs_file_size); - goto out; - } - pos = offset; - while (pos < (offset + size)) { - pgoff_t ecryptfs_page_idx = (pos >> PAGE_CACHE_SHIFT); - size_t start_offset_in_page = (pos & ~PAGE_CACHE_MASK); - size_t num_bytes = (PAGE_CACHE_SIZE - start_offset_in_page); - size_t total_remaining_bytes = ((offset + size) - pos); - - if (num_bytes > total_remaining_bytes) - num_bytes = total_remaining_bytes; - ecryptfs_page = ecryptfs_get_locked_page(ecryptfs_inode, - ecryptfs_page_idx); - if (IS_ERR(ecryptfs_page)) { - rc = PTR_ERR(ecryptfs_page); - printk(KERN_ERR "%s: Error getting page at " - "index [%ld] from eCryptfs inode " - "mapping; rc = [%d]\n", __func__, - ecryptfs_page_idx, rc); - goto out; - } - ecryptfs_page_virt = kmap_atomic(ecryptfs_page, KM_USER0); - memcpy((data + data_offset), - ((char *)ecryptfs_page_virt + start_offset_in_page), - num_bytes); - kunmap_atomic(ecryptfs_page_virt, KM_USER0); - flush_dcache_page(ecryptfs_page); - SetPageUptodate(ecryptfs_page); - unlock_page(ecryptfs_page); - page_cache_release(ecryptfs_page); - pos += num_bytes; - data_offset += num_bytes; - } -out: - return rc; -} -#endif /* 0 */ From 58ded24f0fcb85bddb665baba75892f6ad0f4b8a Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Tue, 24 Jan 2012 10:02:22 -0600 Subject: [PATCH 810/820] eCryptfs: Fix oops when printing debug info in extent crypto functions If pages passed to the eCryptfs extent-based crypto functions are not mapped and the module parameter ecryptfs_verbosity=1 was specified at loading time, a NULL pointer dereference will occur. Note that this wouldn't happen on a production system, as you wouldn't pass ecryptfs_verbosity=1 on a production system. It leaks private information to the system logs and is for debugging only. The debugging info printed in these messages is no longer very useful and rather than doing a kmap() in these debugging paths, it will be better to simply remove the debugging paths completely. https://launchpad.net/bugs/913651 Signed-off-by: Tyler Hicks Reported-by: Daniel DeFreez Cc: --- fs/ecryptfs/crypto.c | 40 ---------------------------------------- 1 file changed, 40 deletions(-) diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index ff981503b3e3..63ab24510649 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -417,17 +417,6 @@ static int ecryptfs_encrypt_extent(struct page *enc_extent_page, (unsigned long long)(extent_base + extent_offset), rc); goto out; } - if (unlikely(ecryptfs_verbosity > 0)) { - ecryptfs_printk(KERN_DEBUG, "Encrypting extent " - "with iv:\n"); - ecryptfs_dump_hex(extent_iv, crypt_stat->iv_bytes); - ecryptfs_printk(KERN_DEBUG, "First 8 bytes before " - "encryption:\n"); - ecryptfs_dump_hex((char *) - (page_address(page) - + (extent_offset * crypt_stat->extent_size)), - 8); - } rc = ecryptfs_encrypt_page_offset(crypt_stat, enc_extent_page, 0, page, (extent_offset * crypt_stat->extent_size), @@ -440,14 +429,6 @@ static int ecryptfs_encrypt_extent(struct page *enc_extent_page, goto out; } rc = 0; - if (unlikely(ecryptfs_verbosity > 0)) { - ecryptfs_printk(KERN_DEBUG, "Encrypt extent [0x%.16llx]; " - "rc = [%d]\n", - (unsigned long long)(extent_base + extent_offset), rc); - ecryptfs_printk(KERN_DEBUG, "First 8 bytes after " - "encryption:\n"); - ecryptfs_dump_hex((char *)(page_address(enc_extent_page)), 8); - } out: return rc; } @@ -543,17 +524,6 @@ static int ecryptfs_decrypt_extent(struct page *page, (unsigned long long)(extent_base + extent_offset), rc); goto out; } - if (unlikely(ecryptfs_verbosity > 0)) { - ecryptfs_printk(KERN_DEBUG, "Decrypting extent " - "with iv:\n"); - ecryptfs_dump_hex(extent_iv, crypt_stat->iv_bytes); - ecryptfs_printk(KERN_DEBUG, "First 8 bytes before " - "decryption:\n"); - ecryptfs_dump_hex((char *) - (page_address(enc_extent_page) - + (extent_offset * crypt_stat->extent_size)), - 8); - } rc = ecryptfs_decrypt_page_offset(crypt_stat, page, (extent_offset * crypt_stat->extent_size), @@ -567,16 +537,6 @@ static int ecryptfs_decrypt_extent(struct page *page, goto out; } rc = 0; - if (unlikely(ecryptfs_verbosity > 0)) { - ecryptfs_printk(KERN_DEBUG, "Decrypt extent [0x%.16llx]; " - "rc = [%d]\n", - (unsigned long long)(extent_base + extent_offset), rc); - ecryptfs_printk(KERN_DEBUG, "First 8 bytes after " - "decryption:\n"); - ecryptfs_dump_hex((char *)(page_address(page) - + (extent_offset - * crypt_stat->extent_size)), 8); - } out: return rc; } From 1589cb1a94c381579a0235ca708d9e2dca6d3a39 Mon Sep 17 00:00:00 2001 From: Li Wang Date: Wed, 25 Jan 2012 15:40:31 +0800 Subject: [PATCH 811/820] eCryptfs: move misleading function comments The data encryption was moved from ecryptfs_write_end into ecryptfs_writepage, this patch moves the corresponding function comments to be consistent with the modification. Signed-off-by: Li Wang Signed-off-by: Linus Torvalds --- fs/ecryptfs/mmap.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c index 6a44148c5fb9..10ec695ccd68 100644 --- a/fs/ecryptfs/mmap.c +++ b/fs/ecryptfs/mmap.c @@ -57,6 +57,10 @@ struct page *ecryptfs_get_locked_page(struct inode *inode, loff_t index) * @page: Page that is locked before this call is made * * Returns zero on success; non-zero otherwise + * + * This is where we encrypt the data and pass the encrypted data to + * the lower filesystem. In OpenPGP-compatible mode, we operate on + * entire underlying packets. */ static int ecryptfs_writepage(struct page *page, struct writeback_control *wbc) { @@ -481,10 +485,6 @@ int ecryptfs_write_inode_size_to_metadata(struct inode *ecryptfs_inode) * @copied: The amount of data copied * @page: The eCryptfs page * @fsdata: The fsdata (unused) - * - * This is where we encrypt the data and pass the encrypted data to - * the lower filesystem. In OpenPGP-compatible mode, we operate on - * entire underlying packets. */ static int ecryptfs_write_end(struct file *file, struct address_space *mapping, From 5a51467b146ab7948d2f6812892eac120a30529c Mon Sep 17 00:00:00 2001 From: Russ Anderson Date: Wed, 18 Jan 2012 20:07:54 -0600 Subject: [PATCH 812/820] x86/uv: Fix uv_gpa_to_soc_phys_ram() shift uv_gpa_to_soc_phys_ram() was inadvertently ignoring the shift values. This fix takes the shift into account. Signed-off-by: Russ Anderson Cc: Link: http://lkml.kernel.org/r/20120119020753.GA7228@sgi.com Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uv/uv_hub.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h index 54a13aaebc40..21f7385badb8 100644 --- a/arch/x86/include/asm/uv/uv_hub.h +++ b/arch/x86/include/asm/uv/uv_hub.h @@ -318,13 +318,13 @@ uv_gpa_in_mmr_space(unsigned long gpa) /* UV global physical address --> socket phys RAM */ static inline unsigned long uv_gpa_to_soc_phys_ram(unsigned long gpa) { - unsigned long paddr = gpa & uv_hub_info->gpa_mask; + unsigned long paddr; unsigned long remap_base = uv_hub_info->lowmem_remap_base; unsigned long remap_top = uv_hub_info->lowmem_remap_top; gpa = ((gpa << uv_hub_info->m_shift) >> uv_hub_info->m_shift) | ((gpa >> uv_hub_info->n_lshift) << uv_hub_info->m_val); - gpa = gpa & uv_hub_info->gpa_mask; + paddr = gpa & uv_hub_info->gpa_mask; if (paddr >= remap_base && paddr < remap_base + remap_top) paddr -= remap_base; return paddr; From d2ebc71d472020bc30e29afe8c4d2a85a5b41f56 Mon Sep 17 00:00:00 2001 From: Cliff Wickman Date: Wed, 18 Jan 2012 09:40:47 -0600 Subject: [PATCH 813/820] x86/uv: Fix uninitialized spinlocks Initialize two spinlocks in tlb_uv.c and also properly define/initialize the uv_irq_lock. The lack of explicit initialization seems to be functionally harmless, but it is diagnosed when these are turned on: CONFIG_DEBUG_SPINLOCK=y CONFIG_DEBUG_MUTEXES=y CONFIG_DEBUG_LOCK_ALLOC=y CONFIG_LOCKDEP=y Signed-off-by: Cliff Wickman Cc: Cc: Dimitri Sivanich Link: http://lkml.kernel.org/r/E1RnXd1-0003wU-PM@eag09.americas.sgi.com [ Added the uv_irq_lock initialization fix by Dimitri Sivanich ] Signed-off-by: Ingo Molnar --- arch/x86/platform/uv/tlb_uv.c | 2 ++ arch/x86/platform/uv/uv_irq.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/x86/platform/uv/tlb_uv.c b/arch/x86/platform/uv/tlb_uv.c index 9be4cff00a2d..3ae0e61abd23 100644 --- a/arch/x86/platform/uv/tlb_uv.c +++ b/arch/x86/platform/uv/tlb_uv.c @@ -1851,6 +1851,8 @@ static void __init init_per_cpu_tunables(void) bcp->cong_reps = congested_reps; bcp->cong_period = congested_period; bcp->clocks_per_100_usec = usec_2_cycles(100); + spin_lock_init(&bcp->queue_lock); + spin_lock_init(&bcp->uvhub_lock); } } diff --git a/arch/x86/platform/uv/uv_irq.c b/arch/x86/platform/uv/uv_irq.c index 374a05d8ad22..f25c2765a5c9 100644 --- a/arch/x86/platform/uv/uv_irq.c +++ b/arch/x86/platform/uv/uv_irq.c @@ -25,7 +25,7 @@ struct uv_irq_2_mmr_pnode{ int irq; }; -static spinlock_t uv_irq_lock; +static DEFINE_SPINLOCK(uv_irq_lock); static struct rb_root uv_irq_root; static int uv_set_irq_affinity(struct irq_data *, const struct cpumask *, bool); From 3fe54564a61f72982032423d24041dca30617ca2 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Wed, 25 Jan 2012 14:35:49 +0800 Subject: [PATCH 814/820] x86/numachip: Drop unnecessary conflict with EDAC EDAC detection no longer crashes multi-node systems, so don't conflict on it with NumaChip. Signed-off-by: Daniel J Blueman Cc: Steffen Persvold Link: http://lkml.kernel.org/r/1327473349-28395-1-git-send-email-daniel@numascale-asia.com Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 864cc6e6ac8e..5bed94e189fa 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -360,7 +360,6 @@ config X86_NUMACHIP depends on NUMA depends on SMP depends on X86_X2APIC - depends on !EDAC_AMD64 ---help--- Adds support for Numascale NumaChip large-SMP systems. Needed to enable more than ~168 cores. From 5067cf53cac9b36d42ebb3a45bb12259d0bc1e68 Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Mon, 23 Jan 2012 23:34:59 +0100 Subject: [PATCH 815/820] x86/boot-image: Don't leak phdrs in arch/x86/boot/compressed/misc.c::Parse_elf() We allocate memory with malloc(), but neglect to free it before the variable 'phdrs' goes out of scope --> leak. Signed-off-by: Jesper Juhl Link: http://lkml.kernel.org/r/alpine.LNX.2.00.1201232332590.8772@swampdragon.chaosbits.net [ Mostly harmless. ] Signed-off-by: Ingo Molnar --- arch/x86/boot/compressed/misc.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index 3a19d04cebeb..7116dcba0c9e 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -321,6 +321,8 @@ static void parse_elf(void *output) default: /* Ignore other PT_* */ break; } } + + free(phdrs); } asmlinkage void decompress_kernel(void *rmode, memptr heap, From 652847aa449cfe364d40018849223f57f31a38e2 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Fri, 20 Jan 2012 17:38:23 +0100 Subject: [PATCH 816/820] x86/amd: Add missing feature flag for fam15h models 10h-1fh processors That is the last one missing for those CPUs. Others were recently added with commits fb215366b3c7320ac25dca766a0152df16534932 (KVM: expose latest Intel cpu new features (BMI1/BMI2/FMA/AVX2) to guest) and commit 969df4b82904a30fef19a67398a0c854d223ea67 (x86: Report cpb and eff_freq_ro flags correctly) Signed-off-by: Andreas Herrmann Link: http://lkml.kernel.org/r/20120120163823.GC24508@alberich.amd.com Signed-off-by: Ingo Molnar --- arch/x86/include/asm/cpufeature.h | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index 17c5d4bdee5e..8d67d428b0f9 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -159,6 +159,7 @@ #define X86_FEATURE_WDT (6*32+13) /* Watchdog timer */ #define X86_FEATURE_LWP (6*32+15) /* Light Weight Profiling */ #define X86_FEATURE_FMA4 (6*32+16) /* 4 operands MAC instructions */ +#define X86_FEATURE_TCE (6*32+17) /* translation cache extension */ #define X86_FEATURE_NODEID_MSR (6*32+19) /* NodeId MSR */ #define X86_FEATURE_TBM (6*32+21) /* trailing bit manipulations */ #define X86_FEATURE_TOPOEXT (6*32+22) /* topology extensions CPUID leafs */ From 5b68edc91cdc972c46f76f85eded7ffddc3ff5c2 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Fri, 20 Jan 2012 17:44:12 +0100 Subject: [PATCH 817/820] x86/microcode_amd: Add support for CPU family specific container files We've decided to provide CPU family specific container files (starting with CPU family 15h). E.g. for family 15h we have to load microcode_amd_fam15h.bin instead of microcode_amd.bin Rationale is that starting with family 15h patch size is larger than 2KB which was hard coded as maximum patch size in various microcode loaders (not just Linux). Container files which include patches larger than 2KB cause different kinds of trouble with such old patch loaders. Thus we have to ensure that the default container file provides only patches with size less than 2KB. Signed-off-by: Andreas Herrmann Cc: Borislav Petkov Cc: Link: http://lkml.kernel.org/r/20120120164412.GD24508@alberich.amd.com [ documented the naming convention and tidied the code a bit. ] Signed-off-by: Ingo Molnar --- arch/x86/kernel/microcode_amd.c | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-) diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c index fe86493f3ed1..ac0417be9131 100644 --- a/arch/x86/kernel/microcode_amd.c +++ b/arch/x86/kernel/microcode_amd.c @@ -311,13 +311,33 @@ out: return state; } +/* + * AMD microcode firmware naming convention, up to family 15h they are in + * the legacy file: + * + * amd-ucode/microcode_amd.bin + * + * This legacy file is always smaller than 2K in size. + * + * Starting at family 15h they are in family specific firmware files: + * + * amd-ucode/microcode_amd_fam15h.bin + * amd-ucode/microcode_amd_fam16h.bin + * ... + * + * These might be larger than 2K. + */ static enum ucode_state request_microcode_amd(int cpu, struct device *device) { - const char *fw_name = "amd-ucode/microcode_amd.bin"; + char fw_name[36] = "amd-ucode/microcode_amd.bin"; const struct firmware *fw; enum ucode_state ret = UCODE_NFOUND; + struct cpuinfo_x86 *c = &cpu_data(cpu); - if (request_firmware(&fw, fw_name, device)) { + if (c->x86 >= 0x15) + snprintf(fw_name, sizeof(fw_name), "amd-ucode/microcode_amd_fam%.2xh.bin", c->x86); + + if (request_firmware(&fw, (const char *)fw_name, device)) { pr_err("failed to load file %s\n", fw_name); goto out; } From a1ad803322a904a250fa901020b4a4dfaf51a829 Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Wed, 25 Jan 2012 22:07:05 +0100 Subject: [PATCH 818/820] sh: se7724: fix compile breakage Fix compilation breakage arch/sh/boards/mach-se/7724/setup.c:182: error: 'V4L2_PIX_FMT_RGB565' undeclared here (not in a function) make[3]: *** [arch/sh/boards/mach-se/7724/setup.o] Error 1 caused by commit "fbdev: sh_mobile_lcdc: Support FOURCC-based format API" Also add other missing headers, even if compilation currently succeeds because of their indirect inclusion via other headers. Signed-off-by: Guennadi Liakhovetski Acked-by: Laurent Pinchart Signed-off-by: Paul Mundt --- arch/sh/boards/mach-ap325rxa/setup.c | 1 + arch/sh/boards/mach-ecovec24/setup.c | 2 ++ arch/sh/boards/mach-kfr2r09/setup.c | 1 + arch/sh/boards/mach-migor/setup.c | 2 ++ arch/sh/boards/mach-se/7724/setup.c | 1 + 5 files changed, 7 insertions(+) diff --git a/arch/sh/boards/mach-ap325rxa/setup.c b/arch/sh/boards/mach-ap325rxa/setup.c index 6418e95c2b6b..ebd0f818a25f 100644 --- a/arch/sh/boards/mach-ap325rxa/setup.c +++ b/arch/sh/boards/mach-ap325rxa/setup.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include diff --git a/arch/sh/boards/mach-ecovec24/setup.c b/arch/sh/boards/mach-ecovec24/setup.c index 033ef2ba621f..cde7c0085ced 100644 --- a/arch/sh/boards/mach-ecovec24/setup.c +++ b/arch/sh/boards/mach-ecovec24/setup.c @@ -29,9 +29,11 @@ #include #include #include +#include #include