Enable TWL4030_USB which is used at least on Nokia N900/N950/N9 (OMAP3)
and BeagleBoard.
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
[tony@atomide.com: updated comments]
Signed-off-by: Tony Lindgren <tony@atomide.com>
The xhci in Intel Sunrisepoint and Cherryview platforms need a driver
workaround for a Stuck PME that might either block PME events in suspend,
or create spurious PME events preventing runtime suspend.
Workaround is to clear a internal PME flag, BIT(28) in a vendor specific
PMCTRL register at offset 0x80a4, in both suspend resume callbacks
Without this, xhci connected usb devices might never be able to wake up the
system from suspend, or prevent device from going to suspend (xhci d3)
Cc: <stable@vger.kernel.org>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a control transfer has a short data stage, the xHCI controller generates
two transfer events: a COMP_SHORT_TX event that specifies the untransferred
amount, and a COMP_SUCCESS event. But when the data stage is not short, only the
COMP_SUCCESS event occurs. Therefore, xhci-hcd must set urb->actual_length to
urb->transfer_buffer_length while processing the COMP_SUCCESS event, unless
urb->actual_length was set already by a previous COMP_SHORT_TX event.
The driver checks this by seeing whether urb->actual_length == 0, but this alone
is the wrong test, as it is entirely possible for a short transfer to have an
urb->actual_length = 0.
This patch changes the xhci driver to rely on a new td->urb_length_set flag,
which is set to true when a COMP_SHORT_TX event is received and the URB length
updated at that stage.
This fixes a bug which affected the HSO plugin, which relies on URBs with
urb->actual_length == 0 to halt re-submitting the RX URB in the control
endpoint.
Cc: <stable@vger.kernel.org>
Signed-off-by: Aleksander Morgado <aleksander@aleksander.es>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit bca9690b94 ("clk: divider: Make generic for usage elsewhere")
returned only the divider value for read-only dividers instead of the
actual rate.
Fixes: bca9690b94 ("clk: divider: Make generic for usage elsewhere")
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: James Hogan <james.hogan@imgtec.com>
Tested-by: James Hogan <james.hogan@imgtec.com>
Acked-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Michael Turquette <mturquette@linaro.org>
DCAN1 RX and TX lines are internally pulled high according to [1].
While muxing between DCAN mode and SAFE mode we make sure
that the same pull direction is set to minimize opposite
pull contention during the switching window.
[1] in DRA7 data manual, Ball characteristics table 4-2, DSIS colum shows
the state driven to the peripheral input while in the deselcted mode.
DSIS - De-Selected Input State.
Signed-off-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Rev.F onwards ball G19 (dcan1_rx) is used as a GPIO for some other
function so don't include it in DCAN pinctrl node.
Signed-off-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
According to AM437x TRM, Document SPRUHL7B, Revised December 2014,
Section 7.2.1 Pad Control Registers, setting bit 19 of the pad control
registers actually sets the SLEWCTRL value to slow rather than fast as
the current macro indicates. Introduce a new macro, SLEWCTRL_SLOW, that
sets the bit, and modify SLEWCTRL_FAST to 0 but keep it for
completeness.
Current users of the macro (i2c, mdio, and uart) are left unmodified as
SLEWCTRL_FAST was the macro used and actual desired state. Tested on
am437x-gp-evm with no difference in software performance seen.
Signed-off-by: Dave Gerlach <d-gerlach@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
According to AM335x TRM, Document spruh73l, Revised February 2015,
Section 9.2.2 Pad Control Registers, setting bit 6 of the pad control
registers actually sets the SLEWCTRL value to slow rather than fast as
the current macro indicates. Introduce a new macro, SLEWCTRL_SLOW, that
sets the bit, and modify SLEWCTRL_FAST to 0 but keep it for
completeness.
Current users of the macro (i2c and mdio) are left unmodified as
SLEWCTRL_FAST was the macro used and actual desired state. Tested on
am335x-gp-evm with no difference in software performance seen.
Signed-off-by: Dave Gerlach <d-gerlach@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
OMAP4 has a finer counter granularity, which allows for a delay of 1000ms
in the thermal zone polling intervals. OMAP5 has a different counter
mechanism, which allows at maximum a 500ms timer. Adjust the cpu thermal
zone polling interval accordingly.
Without this patch, the polling interval information is simply ignored,
and the following thermal warnings are printed during boot (assuming
thermal is enabled);
[ 1.545343] ti-soc-thermal 4a0021e0.bandgap: Delay 1000 ms is not supported
[ 1.552691] ti-soc-thermal 4a0021e0.bandgap: Delay 1000 ms is not supported
[ 1.560029] ti-soc-thermal 4a0021e0.bandgap: Delay 1000 ms is not supported
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Cc: Tony Lindgren <tony@atomide.com>
Acked-by: Eduardo Valentin <edubezval@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Use external clock for RMII since the internal clock doesn't meet the
jitter requirements.
Signed-off-by: George McCollister <george.mccollister@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Beaglebone Black doesn't have AES and SHAM enabled like the
original Beaglebone White dts. This breaks applications that
leverage the crypto blocks so fix this by enabling these nodes
in the am335x-bone-common.dtsi. With this change, enabling the
nodes in am335x-bone.dts is no longer required so remove them.
Signed-off-by: Matt Porter <mporter@konsulko.com>
Acked-by: Robert Nelson <robertcnelson@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
ehrpwm tbclk is wrongly modelled as deriving from dpll_per_m2_ck.
The TRM says tbclk is derived from SYSCLKOUT. SYSCLKOUT nothing but the
functional clock of pwmss (l4ls_gclk).
Fix this by changing source of ehrpwmx_tbclk to l4ls_gclk.
Fixes: 4da1c67719 ("add tbclk data for ehrpwm")
Signed-off-by: Vignesh R <vigneshr@ti.com>
Acked-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
ehrpwm tbclk is wrongly modelled as deriving from dpll_per_m2_ck.
The TRM says tbclk is derived from SYSCLKOUT. SYSCLKOUT nothing but the
functional clock of pwmss (l4ls_gclk).
Fix this by changing source of ehrpwmx_tbclk to l4ls_gclk.
Fixes: 9e100ebafb: ("Fix ehrpwm tbclk data")
Signed-off-by: Vignesh R <vigneshr@ti.com>
Acked-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Fixes 85dc74e9 (ARM: dts: omap5 clock data)
On OMAP54xx, For DPLL_IVA, the ref clock(CLKINP) is connected to sys_clk1 and
the bypass input(CLKINPULOW) is connected to iva_dpll_hs_clk_div clock.
But the bypass input is not directly routed to bypass clkout instead
both CLKINP and CLKINPULOW are connected to bypass clkout via a mux.
This mux is controlled by the bit - CM_CLKSEL_DPLL_IVA[23]:DPLL_BYP_CLKSEL
and it's POR value is zero which selects the CLKINP as bypass clkout.
which means iva_dpll_hs_clk_div is not the bypass clock for dpll_iva_ck
Fix this by adding another mux clock as parent in bypass mode.
This design is common to most of the PLLs and the rest have only one bypass
clock. Below is a list of the DPLLs that need this fix:
DPLL_IVA,
DPLL_PER,
DPLL_USB and DPLL_CORE
Signed-off-by: Ravikumar Kattekola <rk@ti.com>
Acked-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Fixes: ee6c750761 (ARM: dts: dra7 clock data)
On DRA7x, For DPLL_IVA, the ref clock(CLKINP) is connected to sys_clk1 and
the bypass input(CLKINPULOW) is connected to iva_dpll_hs_clk_div clock.
But the bypass input is not directly routed to bypass clkout instead
both CLKINP and CLKINPULOW are connected to bypass clkout via a mux.
This mux is controlled by the bit - CM_CLKSEL_DPLL_IVA[23]:DPLL_BYP_CLKSEL
and it's POR value is zero which selects the CLKINP as bypass clkout.
which means iva_dpll_hs_clk_div is not the bypass clock for dpll_iva_ck
Fix this by adding another mux clock as parent in bypass mode.
This design is common to most of the PLLs and the rest have only one bypass
clock. Below is a list of the DPLLs that need this fix:
DPLL_IVA, DPLL_DDR,
DPLL_DSP, DPLL_EVE,
DPLL_GMAC, DPLL_PER,
DPLL_USB and DPLL_CORE
Signed-off-by: Ravikumar Kattekola <rk@ti.com>
Acked-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
A few driver specific fixes here, none of them earth shattering in
themselves, that have accumliated since the opening of the merge window.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJU96stAAoJECTWi3JdVIfQLfUH/jBYQqxrPp8qm7Lt9LOuwC4U
SOi0ERIVhLSKZMo7G4wHmbEapx2FBgH+wxqh9kNLaISvOIH4odFS4Fr1XYE70cY9
97M46uzLqdaKCDgvkXdx1NLA0zubEGWLOtdVYlrMAzin0FXNH+KYhxkRQbJ6f0LE
yKmNemnHHtJ4gDl/kMED7bc+mkYSOpwUE/8HLOEA8bcE3luDnj7rvQwOhPIYq0Tz
Ij4GM8HSfWPp/k3YmZVcF8GA6icN3LG5Jg7DOu6S19fMYimTDaHQDCw0AoLfiIOQ
DykyHg7ay/I3B6zUz4JikLIW5KjFxZfeQ818YE0GQ4PgePfb0GSIFt0LnNmq4Oo=
=GIb7
-----END PGP SIGNATURE-----
Merge tag 'asoc-fix-v4.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Fixes for v4.0
A few driver specific fixes here, none of them earth shattering in
themselves, that have accumliated since the opening of the merge window.
Using the pvops kernel a NULL pointer dereference was detected on a
large machine (144 processors) when booting as dom0 in
evtchn_fifo_unmask() during assignment of a pirq.
The event channel in question was the first to need a new entry in
event_array[] in events_fifo.c. Unfortunately xen_irq_info_pirq_setup()
is called with evtchn being 0 for a new pirq and the real event channel
number is assigned to the pirq only during __startup_pirq().
It is mandatory to call xen_evtchn_port_setup() after assigning the
event channel number to the pirq to make sure all memory needed for the
event channel is allocated.
Signed-off-by: Juergen Gross <jgross@suse.com>
Cc: <stable@vger.kernel.org> # 3.14+
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
There is a missing lower bound check on "pitchbend" so it means we can
read up to 6 elements before the start of the opl3_note_table[] array.
Thanks to Clemens Ladisch for his help with this patch.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When CONFIG_DEBUG_SET_MODULE_RONX is enabled, the sizes of
module sections are aligned up so appropriate permissions can
be applied. Adjusting for the symbol table may cause them to
become unaligned. Make sure to re-align the sizes afterward.
Signed-off-by: Laura Abbott <lauraa@codeaurora.org>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
The set_memory_* functions currently only support module
addresses. The addresses are validated using is_module_addr.
That function is special though and relies on internal state
in the module subsystem to work properly. At the time of
module initialization and calling set_memory_*, it's too early
for is_module_addr to work properly so it always returns
false. Rather than be subject to the whims of the module state,
just bounds check against the module virtual address range.
Signed-off-by: Laura Abbott <lauraa@codeaurora.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
The BIOS might reconfigure pins as it needs when S3 is entered. This might
cause drivers using the GPIOs to fail because the state was wrong or
interrupts stopped working.
Fix this by saving and restoring enough pin context over system sleep.
Reported-by: Hans Holmberg <hans.holmberg@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
When DMA descriptor allocation fails we should not try to assign any fields in
the bad descriptor. The patch adds the necessary checks for that.
Fixes: 7063c0d942 (spi/dw_spi: add DMA support)
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org
Instead of handling everything in the driver's first level interrupt
handler, we can take advantage of already existing flow handlers that are
provided by the IRQ core.
This changes the functionality a bit also. Previously the driver looped
over pending interrupts in a single loop, restarting the loop if some
interrupt changed state. This caused problem with Lenovo Thinkpad 10
digitizer that it was not able to deassert the interrupt before the driver
disabled the interrupt for good (looplimit was exhausted).
Rework the interrupt handling logic a bit so that we provide proper mask,
ack and unmask operations in terms of Baytrail GPIO hardware and loop over
pending interrupts only once. If the interrupt remains asserted the first
level handler will be re-triggered automatically.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
If the pin is already configured as GPIO and it has any of the triggering
flags set, we may get spurious interrupts depending on the state of the
pin.
Prevent this by clearing the triggering flags on such pins. However, if the
pin is also configured as "direct IRQ" we leave the flags as is. Otherwise
it will prevent interrupts that are routed directly to IO-APIC.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Zotac ZBOX PI320, a Baytrail based mini-PC, has power button connected to a
GPIO pin and it is exposed to the operating system as Windows 8 button
array. This is implemented in Linux as a driver using gpio_keys.
However, BIOS on this particula machine forgot to mux the pin to be a GPIO
instead of native function, which results following message to be seen on
the console:
byt_gpio INT33FC:02: pin 16 cannot be used as GPIO.
This causes power button to not work as the driver was not able to request
the GPIO it needs.
So instead of completely preventing this we allow turning the pin as GPIO
but issue warning that something might be wrong.
Reported-by: Benjamin Adler <benadler@gmx.net>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
On gcc5 the kernel does not link:
ld: .eh_frame_hdr table[4] FDE at 0000000000000648 overlaps table[5] FDE at 0000000000000670.
Because prior GCC versions always emitted NOPs on ALIGN directives, but
gcc5 started omitting them.
.LSTARTFDEDLSI1 says:
/* HACK: The dwarf2 unwind routines will subtract 1 from the
return address to get an address in the middle of the
presumed call instruction. Since we didn't get here via
a call, we need to include the nop before the real start
to make up for it. */
.long .LSTART_sigreturn-1-. /* PC-relative start address */
But commit 69d0627a7f ("x86 vDSO: reorder vdso32 code") from 2.6.25
replaced .org __kernel_vsyscall+32,0x90 by ALIGN right before
__kernel_sigreturn.
Of course, ALIGN need not generate any NOP in there. Esp. gcc5 collapses
vclock_gettime.o and int80.o together with no generated NOPs as "ALIGN".
So fix this by adding to that point at least a single NOP and make the
function ALIGN possibly with more NOPs then.
Kudos for reporting and diagnosing should go to Richard.
Reported-by: Richard Biener <rguenther@suse.de>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Acked-by: Andy Lutomirski <luto@amacapital.net>
Cc: <stable@vger.kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1425543211-12542-1-git-send-email-jslaby@suse.cz
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In xgene_pcie_map_bus(), we neglected to add in the register offset when
calculating the config space address. This means all config accesses
operated on the first four bytes of config space.
Add the register offset to the config space base address.
Also correct the xgene_pcie_map_bus() prototype to fix a compiler warning.
[bhelgaas: changelog]
Fixes: 350f8be5bb ("PCI: xgene: Convert to use generic config accessors")
Posting: http://lkml.kernel.org/r/1424214840-26498-1-git-send-email-fkan@apm.com
Signed-off-by: Feng Kan <fkan@apm.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Tanmay Inamdar <tinamdar@apm.com>
Acked-by: Rob Herring <robh@kernel.org>
With some mss values, it is possible tcp_xmit_size_goal() puts
one segment more in TSO packet than tcp_tso_autosize().
We send then one TSO packet followed by one single MSS.
It is not a serious bug, but we can do slightly better, especially
for drivers using netif_set_gso_max_size() to lower gso_max_size.
Using same formula avoids these corner cases and makes
tcp_xmit_size_goal() a bit faster.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Fixes: 605ad7f184 ("tcp: refine TSO autosizing")
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the driver is removed (e.g. using unbind through sysfs), the
clocks get disabled twice, once on fec_enet_close and once on
fec_drv_remove. Since the clocks are enabled only once, this leads
to a warning:
WARNING: CPU: 0 PID: 402 at drivers/clk/clk.c:992 clk_core_disable+0x64/0x68()
Remove the call to fec_enet_clk_enable in fec_drv_remove to balance
the clock enable/disable calls again. This has been introduce by
e8fcfcd568 ("net: fec: optimize the clock management to save power").
Signed-off-by: Stefan Agner <stefan@agner.ch>
Acked-by: Fugang Duan <B38611@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The latest spec "I-IPA01-0266-USR Rev 10" limit the MID field length to 12 bit
value. For previous versions it is 16 bit value.
This change will not break the backward compatibility as the latest ID value is
7 and with in the 12 bit value limit.
Signed-off-by: Punnaiah Choudary Kalluri <punnaia@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
eTSEC of-nodes may have children which are not queue-group nodes. For
example new-style fixed-phy declarations. These where incorrectly
assumed to be additional queue-groups.
Change the search to filter out any nodes which are not queue-groups,
or have been disabled.
Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pablo Neira Ayuso says:
====================
Netfilter/IPVS fixes for net
The following patchset contains Netfilter/IPVS fixes for your net tree,
they are:
1) Don't truncate ethernet protocol type to u8 in nft_compat, from
Arturo Borrero.
2) Fix several problems in the addition/deletion of elements in nf_tables.
3) Fix module refcount leak in ip_vs_sync, from Julian Anastasov.
4) Fix a race condition in the abort path in the nf_tables transaction
infrastructure. Basically aborted rules can show up as active rules
until changes are unrolled, oneliner from Patrick McHardy.
5) Check for overflows in the data area of the rule, also from Patrick.
6) Fix off-by-one in the per-rule user data size field. This introduces
a new nft_userdata structure that is placed at the beginning of the
user data area that contains the length to save some bits from the
rule and we only need one bit to indicate its presence, from Patrick.
7) Fix rule replacement error path, the replaced rule is deleted on
error instead of leaving it in place. This has been fixed by relying
on the abort path to undo the incomplete replacement.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
ip_check_defrag() may be used by af_packet to defragment outgoing packets.
skb_network_offset() of af_packet's outgoing packets is not zero.
Signed-off-by: Alexander Drozdov <al.drozdov@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
bcmgenet_set_wol() correctly sets MPD_PW_EN when a password is specified
to match magic packets against, however, when we switch from a
password-matching to a matching without password we would leave this bit
turned on, and GENET would only match magic packets with passwords.
This can be reproduced using the following sequence:
ethtool -s eth0 wol g
ethtool -s eth0 wol s sopass 00:11:22:33:44:55
ethtool -s eth0 wol g
The simple fix is to clear the MPD_PWD_EN bit when WAKE_MAGICSECURE is
not set.
Fixes: c51de7f397 ("net: bcmgenet: add Wake-on-LAN support code")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Improper arithmetics when calculting the address of the extended ref could
lead to an out of bounds memory read and kernel panic.
Signed-off-by: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.cz>
cc: stable@vger.kernel.org # v3.7+
Signed-off-by: Chris Mason <clm@fb.com>
When using the fast file fsync code path we can miss the fact that new
writes happened since the last file fsync and therefore return without
waiting for the IO to finish and write the new extents to the fsync log.
Here's an example scenario where the fsync will miss the fact that new
file data exists that wasn't yet durably persisted:
1. fs_info->last_trans_committed == N - 1 and current transaction is
transaction N (fs_info->generation == N);
2. do a buffered write;
3. fsync our inode, this clears our inode's full sync flag, starts
an ordered extent and waits for it to complete - when it completes
at btrfs_finish_ordered_io(), the inode's last_trans is set to the
value N (via btrfs_update_inode_fallback -> btrfs_update_inode ->
btrfs_set_inode_last_trans);
4. transaction N is committed, so fs_info->last_trans_committed is now
set to the value N and fs_info->generation remains with the value N;
5. do another buffered write, when this happens btrfs_file_write_iter
sets our inode's last_trans to the value N + 1 (that is
fs_info->generation + 1 == N + 1);
6. transaction N + 1 is started and fs_info->generation now has the
value N + 1;
7. transaction N + 1 is committed, so fs_info->last_trans_committed
is set to the value N + 1;
8. fsync our inode - because it doesn't have the full sync flag set,
we only start the ordered extent, we don't wait for it to complete
(only in a later phase) therefore its last_trans field has the
value N + 1 set previously by btrfs_file_write_iter(), and so we
have:
inode->last_trans <= fs_info->last_trans_committed
(N + 1) (N + 1)
Which made us not log the last buffered write and exit the fsync
handler immediately, returning success (0) to user space and resulting
in data loss after a crash.
This can actually be triggered deterministically and the following excerpt
from a testcase I made for xfstests triggers the issue. It moves a dummy
file across directories and then fsyncs the old parent directory - this
is just to trigger a transaction commit, so moving files around isn't
directly related to the issue but it was chosen because running 'sync' for
example does more than just committing the current transaction, as it
flushes/waits for all file data to be persisted. The issue can also happen
at random periods, since the transaction kthread periodicaly commits the
current transaction (about every 30 seconds by default).
The body of the test is:
_scratch_mkfs >> $seqres.full 2>&1
_init_flakey
_mount_flakey
# Create our main test file 'foo', the one we check for data loss.
# By doing an fsync against our file, it makes btrfs clear the 'needs_full_sync'
# bit from its flags (btrfs inode specific flags).
$XFS_IO_PROG -f -c "pwrite -S 0xaa 0 8K" \
-c "fsync" $SCRATCH_MNT/foo | _filter_xfs_io
# Now create one other file and 2 directories. We will move this second file
# from one directory to the other later because it forces btrfs to commit its
# currently open transaction if we fsync the old parent directory. This is
# necessary to trigger the data loss bug that affected btrfs.
mkdir $SCRATCH_MNT/testdir_1
touch $SCRATCH_MNT/testdir_1/bar
mkdir $SCRATCH_MNT/testdir_2
# Make sure everything is durably persisted.
sync
# Write more 8Kb of data to our file.
$XFS_IO_PROG -c "pwrite -S 0xbb 8K 8K" $SCRATCH_MNT/foo | _filter_xfs_io
# Move our 'bar' file into a new directory.
mv $SCRATCH_MNT/testdir_1/bar $SCRATCH_MNT/testdir_2/bar
# Fsync our first directory. Because it had a file moved into some other
# directory, this made btrfs commit the currently open transaction. This is
# a condition necessary to trigger the data loss bug.
$XFS_IO_PROG -c "fsync" $SCRATCH_MNT/testdir_1
# Now fsync our main test file. If the fsync succeeds, we expect the 8Kb of
# data we wrote previously to be persisted and available if a crash happens.
# This did not happen with btrfs, because of the transaction commit that
# happened when we fsynced the parent directory.
$XFS_IO_PROG -c "fsync" $SCRATCH_MNT/foo
# Simulate a crash/power loss.
_load_flakey_table $FLAKEY_DROP_WRITES
_unmount_flakey
_load_flakey_table $FLAKEY_ALLOW_WRITES
_mount_flakey
# Now check that all data we wrote before are available.
echo "File content after log replay:"
od -t x1 $SCRATCH_MNT/foo
status=0
exit
The expected golden output for the test, which is what we get with this
fix applied (or when running against ext3/4 and xfs), is:
wrote 8192/8192 bytes at offset 0
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
wrote 8192/8192 bytes at offset 8192
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
File content after log replay:
0000000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa
*
0020000 bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb
*
0040000
Without this fix applied, the output shows the test file does not have
the second 8Kb extent that we successfully fsynced:
wrote 8192/8192 bytes at offset 0
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
wrote 8192/8192 bytes at offset 8192
XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
File content after log replay:
0000000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa
*
0020000
So fix this by skipping the fsync only if we're doing a full sync and
if the inode's last_trans is <= fs_info->last_trans_committed, or if
the inode is already in the log. Also remove setting the inode's
last_trans in btrfs_file_write_iter since it's useless/unreliable.
Also because btrfs_file_write_iter no longer sets inode->last_trans to
fs_info->generation + 1, don't set last_trans to 0 if we bail out and don't
bail out if last_trans is 0, otherwise something as simple as the following
example wouldn't log the second write on the last fsync:
1. write to file
2. fsync file
3. fsync file
|--> btrfs_inode_in_log() returns true and it set last_trans to 0
4. write to file
|--> btrfs_file_write_iter() no longers sets last_trans, so it
remained with a value of 0
5. fsync
|--> inode->last_trans == 0, so it bails out without logging the
second write
A test case for xfstests will be sent soon.
CC: <stable@vger.kernel.org>
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
With certain restrictions it is possible for a wakeup device to share
an IRQ with an IRQF_NO_SUSPEND user, and the warnings introduced by
commit cab303be91 are spurious. The new
IRQF_COND_SUSPEND flag allows drivers to tell the core when these
restrictions are met, allowing spurious warnings to be silenced.
This patch documents how IRQF_COND_SUSPEND is expected to be used,
updating some of the text now made invalid by its addition.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The IRQ line connected to the DBGU UART is often shared with a timer device
which request the IRQ with IRQF_NO_SUSPEND.
Since the UART driver is correctly disabling IRQs when entering suspend
we can safely request the IRQ with IRQF_COND_SUSPEND so that irq core
will not complain about mixing IRQF_NO_SUSPEND and !IRQF_NO_SUSPEND.
Rework the interrupt handler to wake the system up when an interrupt
happens on the DEBUG_UART while the system is suspended.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Reviewed-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The watchdog interrupt (only used when activating software watchdog)
shouldn't be suspended when entering suspend mode, because it is shared
with a timer device (which request the line with IRQF_NO_SUSPEND) and once
the watchdog "Mode Register" has been written, it cannot be changed (which
means we cannot disable the watchdog interrupt when entering suspend).
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Reviewed-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
* suspend-to-idle:
cpuidle / sleep: Use broadcast timer for states that stop local timer
cpuidle: Clean up fallback handling in cpuidle_idle_call()
cpuidle / sleep: Do sanity checks in cpuidle_enter_freeze() too
idle / sleep: Avoid excessive disabling and enabling interrupts
* acpi-resources:
x86/PCI/ACPI: Relax ACPI resource descriptor checks to work around BIOS bugs
x86/PCI/ACPI: Ignore resources consumed by host bridge itself
PCI: versatile: Update for list_for_each_entry() API change
Commit 3810631332 (PM / sleep: Re-implement suspend-to-idle handling)
overlooked the fact that entering some sufficiently deep idle states
by CPUs may cause their local timers to stop and in those cases it
is necessary to switch over to a broadcast timer prior to entering
the idle state. If the cpuidle driver in use does not provide
the new ->enter_freeze callback for any of the idle states, that
problem affects suspend-to-idle too, but it is not taken into account
after the changes made by commit 3810631332.
Fix that by changing the definition of cpuidle_enter_freeze() and
re-arranging of the code in cpuidle_idle_call(), so the former does
not call cpuidle_enter() any more and the fallback case is handled
by cpuidle_idle_call() directly.
Fixes: 3810631332 (PM / sleep: Re-implement suspend-to-idle handling)
Reported-and-tested-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Commit de7b5b3d79 ("net: eth: xgene: change APM X-Gene SoC platform
ethernet to support ACPI") breaks booting with devicetree with UEFI
firmware. In that case, I get:
Unhandled fault: synchronous external abort (0x96000010) at 0xfffffc0000620010
Internal error: : 96000010 [#1] SMP
Modules linked in: vfat fat xfs libcrc32c ahci_xgene libahci_platform libahci
CPU: 7 PID: 634 Comm: NetworkManager Not tainted 4.0.0-rc1+ #4
Hardware name: AppliedMicro Mustang/Mustang, BIOS 1.1.0-rh-0.14 Mar 1 2015
task: fffffe03d4c7e100 ti: fffffe03d4e24000 task.ti: fffffe03d4e24000
PC is at xgene_enet_rd_mcx_mac.isra.11+0x58/0xd4
LR is at xgene_gmac_tx_enable+0x2c/0x50
pc : [<fffffe000069d6fc>] lr : [<fffffe000069dcc4>] pstate: 80000145
sp : fffffe03d4e27590
x29: fffffe03d4e27590 x28: 0000000000000000
x27: fffffe03d4e277c0 x26: fffffe03da8fda10
x25: fffffe03d4e2760c x24: fffffe03d49e28c0
x23: fffffc0000620004 x22: 0000000000000000
x21: fffffc0000620000 x20: fffffc0000620010
x19: 000000000000000b x18: 000003ffd4a96020
x17: 000003ff7fc1f7a0 x16: fffffe000079b9cc
x15: 0000000000000000 x14: 0000000000000000
x13: 0000000000000000 x12: fffffe03d4e24000
x11: fffffe03d4e27da0 x10: 0000000000000001
x9 : 0000000000000000 x8 : fffffe03d4e27a20
x7 : 0000000000000000 x6 : 00000000ffffffef
x5 : fffffe000105f7d0 x4 : fffffe00007ca8c8
x3 : fffffe03d4e2760c x2 : 0000000000000000
x1 : fffffc0000620000 x0 : 0000000040000000
Process NetworkManager (pid: 634, stack limit = 0xfffffe03d4e24028)
Stack: (0xfffffe03d4e27590 to 0xfffffe03d4e28000)
...
Call trace:
[<fffffe000069d6fc>] xgene_enet_rd_mcx_mac.isra.11+0x58/0xd4
[<fffffe000069dcc0>] xgene_gmac_tx_enable+0x28/0x50
[<fffffe00006a112c>] xgene_enet_open+0x2c/0x130
[<fffffe00007b9254>] __dev_open+0xc8/0x148
[<fffffe00007b956c>] __dev_change_flags+0x90/0x158
[<fffffe00007b9664>] dev_change_flags+0x30/0x70
[<fffffe00007c8ab8>] do_setlink+0x278/0x870
[<fffffe00007c95bc>] rtnl_newlink+0x404/0x6a8
[<fffffe00007c8040>] rtnetlink_rcv_msg+0x98/0x218
[<fffffe00007e78e4>] netlink_rcv_skb+0xe0/0xf8
[<fffffe00007c7f94>] rtnetlink_rcv+0x30/0x44
[<fffffe00007e6f2c>] netlink_unicast+0xfc/0x210
[<fffffe00007e75b8>] netlink_sendmsg+0x498/0x5ac
[<fffffe00007990b8>] do_sock_sendmsg+0xa4/0xcc
[<fffffe000079a958>] ___sys_sendmsg+0x1fc/0x208
[<fffffe000079b984>] __sys_sendmsg+0x4c/0x94
[<fffffe000079b9f8>] SyS_sendmsg+0x2c/0x3c
The problem here is that the enet hw clocks are not getting
initialized because of a test to avoid the initialization if
UEFI is used to boot. This is an incorrect test. When booting
with UEFI and devicetree, the kernel must still initialize
the enet hw clocks. If booting with ACPI, the clock hw is
not exposed to the kernel and it is that case where we want
to avoid initializing clocks.
Signed-off-by: Mark Salter <msalter@redhat.com>
Acked-by: Feng Kan <fkan@apm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
EEH recovery for bnx2x based adapters is not reliable on all Power
systems using the default hot reset, which can result in an
unrecoverable EEH error. Forcing the use of fundamental reset
during EEH recovery fixes this.
Cc: stable<stable@vger.kernel.org>
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
