From d8396535ce5df234266f0018fb3c338eebb2f61e Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:13 -0700 Subject: [PATCH 01/19] drm/msm: Bring in new register definitions Bring in refreshed register definitions from the rnndb database. Change-Id: Ic0dedbaddd22f6ac0b8cfb6184073968569de22d Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a2xx.xml.h | 19 +- drivers/gpu/drm/msm/adreno/a3xx.xml.h | 42 +- drivers/gpu/drm/msm/adreno/a4xx.xml.h | 114 +- drivers/gpu/drm/msm/adreno/a5xx.xml.h | 3333 +++++++++++++++++ .../gpu/drm/msm/adreno/adreno_common.xml.h | 27 +- drivers/gpu/drm/msm/adreno/adreno_gpu.h | 4 + drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h | 120 +- 7 files changed, 3577 insertions(+), 82 deletions(-) create mode 100644 drivers/gpu/drm/msm/adreno/a5xx.xml.h diff --git a/drivers/gpu/drm/msm/adreno/a2xx.xml.h b/drivers/gpu/drm/msm/adreno/a2xx.xml.h index fee24297fb92..8d16b21ef8a5 100644 --- a/drivers/gpu/drm/msm/adreno/a2xx.xml.h +++ b/drivers/gpu/drm/msm/adreno/a2xx.xml.h @@ -8,16 +8,17 @@ http://github.com/freedreno/envytools/ git clone https://github.com/freedreno/envytools.git The rules-ng-ng source files this header was generated from are: -- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml ( 398 bytes, from 2015-09-24 17:25:31) -- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml ( 1572 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml ( 32901 bytes, from 2015-05-20 20:03:14) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml ( 11518 bytes, from 2016-02-10 21:03:25) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml ( 16166 bytes, from 2016-02-11 21:20:31) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml ( 83967 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml ( 109916 bytes, from 2016-02-20 18:44:48) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/ocmem.xml ( 1773 bytes, from 2015-09-24 17:30:00) +- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) +- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) -Copyright (C) 2013-2015 by the following authors: +Copyright (C) 2013-2016 by the following authors: - Rob Clark (robclark) - Ilia Mirkin (imirkin) diff --git a/drivers/gpu/drm/msm/adreno/a3xx.xml.h b/drivers/gpu/drm/msm/adreno/a3xx.xml.h index 27dabd5e57fb..d521b13bdf9f 100644 --- a/drivers/gpu/drm/msm/adreno/a3xx.xml.h +++ b/drivers/gpu/drm/msm/adreno/a3xx.xml.h @@ -8,14 +8,15 @@ http://github.com/freedreno/envytools/ git clone https://github.com/freedreno/envytools.git The rules-ng-ng source files this header was generated from are: -- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml ( 398 bytes, from 2015-09-24 17:25:31) -- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml ( 1572 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml ( 32901 bytes, from 2015-05-20 20:03:14) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml ( 11518 bytes, from 2016-02-10 21:03:25) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml ( 16166 bytes, from 2016-02-11 21:20:31) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml ( 83967 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml ( 109916 bytes, from 2016-02-20 18:44:48) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/ocmem.xml ( 1773 bytes, from 2015-09-24 17:30:00) +- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) +- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) Copyright (C) 2013-2016 by the following authors: - Rob Clark (robclark) @@ -129,10 +130,14 @@ enum a3xx_tex_fmt { TFMT_Z16_UNORM = 9, TFMT_X8Z24_UNORM = 10, TFMT_Z32_FLOAT = 11, - TFMT_NV12_UV_TILED = 17, - TFMT_NV12_Y_TILED = 19, - TFMT_NV12_UV = 21, - TFMT_NV12_Y = 23, + TFMT_UV_64X32 = 16, + TFMT_VU_64X32 = 17, + TFMT_Y_64X32 = 18, + TFMT_NV12_64X32 = 19, + TFMT_UV_LINEAR = 20, + TFMT_VU_LINEAR = 21, + TFMT_Y_LINEAR = 22, + TFMT_NV12_LINEAR = 23, TFMT_I420_Y = 24, TFMT_I420_U = 26, TFMT_I420_V = 27, @@ -525,14 +530,6 @@ enum a3xx_uche_perfcounter_select { UCHE_UCHEPERF_ACTIVE_CYCLES = 20, }; -enum a3xx_rb_blend_opcode { - BLEND_DST_PLUS_SRC = 0, - BLEND_SRC_MINUS_DST = 1, - BLEND_DST_MINUS_SRC = 2, - BLEND_MIN_DST_SRC = 3, - BLEND_MAX_DST_SRC = 4, -}; - enum a3xx_intp_mode { SMOOTH = 0, FLAT = 1, @@ -1393,13 +1390,14 @@ static inline uint32_t A3XX_RB_COPY_CONTROL_MODE(enum adreno_rb_copy_control_mod { return ((val) << A3XX_RB_COPY_CONTROL_MODE__SHIFT) & A3XX_RB_COPY_CONTROL_MODE__MASK; } +#define A3XX_RB_COPY_CONTROL_MSAA_SRGB_DOWNSAMPLE 0x00000080 #define A3XX_RB_COPY_CONTROL_FASTCLEAR__MASK 0x00000f00 #define A3XX_RB_COPY_CONTROL_FASTCLEAR__SHIFT 8 static inline uint32_t A3XX_RB_COPY_CONTROL_FASTCLEAR(uint32_t val) { return ((val) << A3XX_RB_COPY_CONTROL_FASTCLEAR__SHIFT) & A3XX_RB_COPY_CONTROL_FASTCLEAR__MASK; } -#define A3XX_RB_COPY_CONTROL_UNK12 0x00001000 +#define A3XX_RB_COPY_CONTROL_DEPTH32_RESOLVE 0x00001000 #define A3XX_RB_COPY_CONTROL_GMEM_BASE__MASK 0xffffc000 #define A3XX_RB_COPY_CONTROL_GMEM_BASE__SHIFT 14 static inline uint32_t A3XX_RB_COPY_CONTROL_GMEM_BASE(uint32_t val) @@ -1472,7 +1470,7 @@ static inline uint32_t A3XX_RB_DEPTH_CONTROL_ZFUNC(enum adreno_compare_func val) { return ((val) << A3XX_RB_DEPTH_CONTROL_ZFUNC__SHIFT) & A3XX_RB_DEPTH_CONTROL_ZFUNC__MASK; } -#define A3XX_RB_DEPTH_CONTROL_BF_ENABLE 0x00000080 +#define A3XX_RB_DEPTH_CONTROL_Z_CLAMP_ENABLE 0x00000080 #define A3XX_RB_DEPTH_CONTROL_Z_TEST_ENABLE 0x80000000 #define REG_A3XX_RB_DEPTH_CLEAR 0x00002101 diff --git a/drivers/gpu/drm/msm/adreno/a4xx.xml.h b/drivers/gpu/drm/msm/adreno/a4xx.xml.h index 3220b91f559a..1004d4ecf8fe 100644 --- a/drivers/gpu/drm/msm/adreno/a4xx.xml.h +++ b/drivers/gpu/drm/msm/adreno/a4xx.xml.h @@ -8,14 +8,15 @@ http://github.com/freedreno/envytools/ git clone https://github.com/freedreno/envytools.git The rules-ng-ng source files this header was generated from are: -- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml ( 398 bytes, from 2015-09-24 17:25:31) -- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml ( 1572 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml ( 32901 bytes, from 2015-05-20 20:03:14) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml ( 11518 bytes, from 2016-02-10 21:03:25) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml ( 16166 bytes, from 2016-02-11 21:20:31) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml ( 83967 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml ( 109916 bytes, from 2016-02-20 18:44:48) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/ocmem.xml ( 1773 bytes, from 2015-09-24 17:30:00) +- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) +- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) Copyright (C) 2013-2016 by the following authors: - Rob Clark (robclark) @@ -46,6 +47,9 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. enum a4xx_color_fmt { RB4_A8_UNORM = 1, RB4_R8_UNORM = 2, + RB4_R8_SNORM = 3, + RB4_R8_UINT = 4, + RB4_R8_SINT = 5, RB4_R4G4B4A4_UNORM = 8, RB4_R5G5B5A1_UNORM = 10, RB4_R5G6B5_UNORM = 14, @@ -89,17 +93,10 @@ enum a4xx_color_fmt { enum a4xx_tile_mode { TILE4_LINEAR = 0, + TILE4_2 = 2, TILE4_3 = 3, }; -enum a4xx_rb_blend_opcode { - BLEND_DST_PLUS_SRC = 0, - BLEND_SRC_MINUS_DST = 1, - BLEND_DST_MINUS_SRC = 2, - BLEND_MIN_DST_SRC = 3, - BLEND_MAX_DST_SRC = 4, -}; - enum a4xx_vtx_fmt { VFMT4_32_FLOAT = 1, VFMT4_32_32_FLOAT = 2, @@ -940,6 +937,7 @@ static inline uint32_t A4XX_RB_MODE_CONTROL_HEIGHT(uint32_t val) { return ((val >> 5) << A4XX_RB_MODE_CONTROL_HEIGHT__SHIFT) & A4XX_RB_MODE_CONTROL_HEIGHT__MASK; } +#define A4XX_RB_MODE_CONTROL_ENABLE_GMEM 0x00010000 #define REG_A4XX_RB_RENDER_CONTROL 0x000020a1 #define A4XX_RB_RENDER_CONTROL_BINNING_PASS 0x00000001 @@ -1043,7 +1041,7 @@ static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR(enum adreno_rb_b } #define A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__MASK 0x000000e0 #define A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__SHIFT 5 -static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE(enum a4xx_rb_blend_opcode val) +static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE(enum a3xx_rb_blend_opcode val) { return ((val) << A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__MASK; } @@ -1061,7 +1059,7 @@ static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR(enum adreno_rb } #define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__MASK 0x00e00000 #define A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__SHIFT 21 -static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE(enum a4xx_rb_blend_opcode val) +static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE(enum a3xx_rb_blend_opcode val) { return ((val) << A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__SHIFT) & A4XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__MASK; } @@ -1073,12 +1071,18 @@ static inline uint32_t A4XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR(enum adreno_r } #define REG_A4XX_RB_BLEND_RED 0x000020f0 -#define A4XX_RB_BLEND_RED_UINT__MASK 0x0000ffff +#define A4XX_RB_BLEND_RED_UINT__MASK 0x000000ff #define A4XX_RB_BLEND_RED_UINT__SHIFT 0 static inline uint32_t A4XX_RB_BLEND_RED_UINT(uint32_t val) { return ((val) << A4XX_RB_BLEND_RED_UINT__SHIFT) & A4XX_RB_BLEND_RED_UINT__MASK; } +#define A4XX_RB_BLEND_RED_SINT__MASK 0x0000ff00 +#define A4XX_RB_BLEND_RED_SINT__SHIFT 8 +static inline uint32_t A4XX_RB_BLEND_RED_SINT(uint32_t val) +{ + return ((val) << A4XX_RB_BLEND_RED_SINT__SHIFT) & A4XX_RB_BLEND_RED_SINT__MASK; +} #define A4XX_RB_BLEND_RED_FLOAT__MASK 0xffff0000 #define A4XX_RB_BLEND_RED_FLOAT__SHIFT 16 static inline uint32_t A4XX_RB_BLEND_RED_FLOAT(float val) @@ -1095,12 +1099,18 @@ static inline uint32_t A4XX_RB_BLEND_RED_F32(float val) } #define REG_A4XX_RB_BLEND_GREEN 0x000020f2 -#define A4XX_RB_BLEND_GREEN_UINT__MASK 0x0000ffff +#define A4XX_RB_BLEND_GREEN_UINT__MASK 0x000000ff #define A4XX_RB_BLEND_GREEN_UINT__SHIFT 0 static inline uint32_t A4XX_RB_BLEND_GREEN_UINT(uint32_t val) { return ((val) << A4XX_RB_BLEND_GREEN_UINT__SHIFT) & A4XX_RB_BLEND_GREEN_UINT__MASK; } +#define A4XX_RB_BLEND_GREEN_SINT__MASK 0x0000ff00 +#define A4XX_RB_BLEND_GREEN_SINT__SHIFT 8 +static inline uint32_t A4XX_RB_BLEND_GREEN_SINT(uint32_t val) +{ + return ((val) << A4XX_RB_BLEND_GREEN_SINT__SHIFT) & A4XX_RB_BLEND_GREEN_SINT__MASK; +} #define A4XX_RB_BLEND_GREEN_FLOAT__MASK 0xffff0000 #define A4XX_RB_BLEND_GREEN_FLOAT__SHIFT 16 static inline uint32_t A4XX_RB_BLEND_GREEN_FLOAT(float val) @@ -1117,12 +1127,18 @@ static inline uint32_t A4XX_RB_BLEND_GREEN_F32(float val) } #define REG_A4XX_RB_BLEND_BLUE 0x000020f4 -#define A4XX_RB_BLEND_BLUE_UINT__MASK 0x0000ffff +#define A4XX_RB_BLEND_BLUE_UINT__MASK 0x000000ff #define A4XX_RB_BLEND_BLUE_UINT__SHIFT 0 static inline uint32_t A4XX_RB_BLEND_BLUE_UINT(uint32_t val) { return ((val) << A4XX_RB_BLEND_BLUE_UINT__SHIFT) & A4XX_RB_BLEND_BLUE_UINT__MASK; } +#define A4XX_RB_BLEND_BLUE_SINT__MASK 0x0000ff00 +#define A4XX_RB_BLEND_BLUE_SINT__SHIFT 8 +static inline uint32_t A4XX_RB_BLEND_BLUE_SINT(uint32_t val) +{ + return ((val) << A4XX_RB_BLEND_BLUE_SINT__SHIFT) & A4XX_RB_BLEND_BLUE_SINT__MASK; +} #define A4XX_RB_BLEND_BLUE_FLOAT__MASK 0xffff0000 #define A4XX_RB_BLEND_BLUE_FLOAT__SHIFT 16 static inline uint32_t A4XX_RB_BLEND_BLUE_FLOAT(float val) @@ -1139,12 +1155,18 @@ static inline uint32_t A4XX_RB_BLEND_BLUE_F32(float val) } #define REG_A4XX_RB_BLEND_ALPHA 0x000020f6 -#define A4XX_RB_BLEND_ALPHA_UINT__MASK 0x0000ffff +#define A4XX_RB_BLEND_ALPHA_UINT__MASK 0x000000ff #define A4XX_RB_BLEND_ALPHA_UINT__SHIFT 0 static inline uint32_t A4XX_RB_BLEND_ALPHA_UINT(uint32_t val) { return ((val) << A4XX_RB_BLEND_ALPHA_UINT__SHIFT) & A4XX_RB_BLEND_ALPHA_UINT__MASK; } +#define A4XX_RB_BLEND_ALPHA_SINT__MASK 0x0000ff00 +#define A4XX_RB_BLEND_ALPHA_SINT__SHIFT 8 +static inline uint32_t A4XX_RB_BLEND_ALPHA_SINT(uint32_t val) +{ + return ((val) << A4XX_RB_BLEND_ALPHA_SINT__SHIFT) & A4XX_RB_BLEND_ALPHA_SINT__MASK; +} #define A4XX_RB_BLEND_ALPHA_FLOAT__MASK 0xffff0000 #define A4XX_RB_BLEND_ALPHA_FLOAT__SHIFT 16 static inline uint32_t A4XX_RB_BLEND_ALPHA_FLOAT(float val) @@ -1348,7 +1370,7 @@ static inline uint32_t A4XX_RB_DEPTH_CONTROL_ZFUNC(enum adreno_compare_func val) { return ((val) << A4XX_RB_DEPTH_CONTROL_ZFUNC__SHIFT) & A4XX_RB_DEPTH_CONTROL_ZFUNC__MASK; } -#define A4XX_RB_DEPTH_CONTROL_BF_ENABLE 0x00000080 +#define A4XX_RB_DEPTH_CONTROL_Z_CLAMP_ENABLE 0x00000080 #define A4XX_RB_DEPTH_CONTROL_EARLY_Z_DISABLE 0x00010000 #define A4XX_RB_DEPTH_CONTROL_FORCE_FRAGZ_TO_FS 0x00020000 #define A4XX_RB_DEPTH_CONTROL_Z_TEST_ENABLE 0x80000000 @@ -2177,11 +2199,23 @@ static inline uint32_t REG_A4XX_RBBM_CLOCK_DELAY_RB_MARB_CCU_L1_REG(uint32_t i0) #define REG_A4XX_CP_DRAW_STATE_ADDR 0x00000232 -#define REG_A4XX_CP_PROTECT_REG_0 0x00000240 - static inline uint32_t REG_A4XX_CP_PROTECT(uint32_t i0) { return 0x00000240 + 0x1*i0; } static inline uint32_t REG_A4XX_CP_PROTECT_REG(uint32_t i0) { return 0x00000240 + 0x1*i0; } +#define A4XX_CP_PROTECT_REG_BASE_ADDR__MASK 0x0001ffff +#define A4XX_CP_PROTECT_REG_BASE_ADDR__SHIFT 0 +static inline uint32_t A4XX_CP_PROTECT_REG_BASE_ADDR(uint32_t val) +{ + return ((val) << A4XX_CP_PROTECT_REG_BASE_ADDR__SHIFT) & A4XX_CP_PROTECT_REG_BASE_ADDR__MASK; +} +#define A4XX_CP_PROTECT_REG_MASK_LEN__MASK 0x1f000000 +#define A4XX_CP_PROTECT_REG_MASK_LEN__SHIFT 24 +static inline uint32_t A4XX_CP_PROTECT_REG_MASK_LEN(uint32_t val) +{ + return ((val) << A4XX_CP_PROTECT_REG_MASK_LEN__SHIFT) & A4XX_CP_PROTECT_REG_MASK_LEN__MASK; +} +#define A4XX_CP_PROTECT_REG_TRAP_WRITE 0x20000000 +#define A4XX_CP_PROTECT_REG_TRAP_READ 0x40000000 #define REG_A4XX_CP_PROTECT_CTRL 0x00000250 @@ -2272,7 +2306,7 @@ static inline uint32_t A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT(uint32_t val) { return ((val) << A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT) & A4XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__MASK; } -#define A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__MASK 0x0003fc00 +#define A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__MASK 0x0000fc00 #define A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT 10 static inline uint32_t A4XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT(uint32_t val) { @@ -2420,7 +2454,7 @@ static inline uint32_t A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT(uint32_t val) { return ((val) << A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT) & A4XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__MASK; } -#define A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__MASK 0x0003fc00 +#define A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__MASK 0x0000fc00 #define A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT 10 static inline uint32_t A4XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT(uint32_t val) { @@ -3117,6 +3151,8 @@ static inline uint32_t A4XX_TPL1_TP_TEX_COUNT_GS(uint32_t val) #define REG_A4XX_GRAS_CL_CLIP_CNTL 0x00002000 #define A4XX_GRAS_CL_CLIP_CNTL_CLIP_DISABLE 0x00008000 +#define A4XX_GRAS_CL_CLIP_CNTL_ZNEAR_CLIP_DISABLE 0x00010000 +#define A4XX_GRAS_CL_CLIP_CNTL_ZFAR_CLIP_DISABLE 0x00020000 #define A4XX_GRAS_CL_CLIP_CNTL_ZERO_GB_SCALE_Z 0x00400000 #define REG_A4XX_GRAS_CLEAR_CNTL 0x00002003 @@ -3670,6 +3706,8 @@ static inline uint32_t A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH(uint32_t val) #define REG_A4XX_PC_BINNING_COMMAND 0x00000d00 #define A4XX_PC_BINNING_COMMAND_BINNING_ENABLE 0x00000001 +#define REG_A4XX_PC_TESSFACTOR_ADDR 0x00000d08 + #define REG_A4XX_PC_DRAWCALL_SETUP_OVERRIDE 0x00000d0c #define REG_A4XX_PC_PERFCTR_PC_SEL_0 0x00000d10 @@ -3690,6 +3728,20 @@ static inline uint32_t A4XX_HLSQ_GS_CONTROL_REG_INSTRLENGTH(uint32_t val) #define REG_A4XX_PC_BIN_BASE 0x000021c0 +#define REG_A4XX_PC_VSTREAM_CONTROL 0x000021c2 +#define A4XX_PC_VSTREAM_CONTROL_SIZE__MASK 0x003f0000 +#define A4XX_PC_VSTREAM_CONTROL_SIZE__SHIFT 16 +static inline uint32_t A4XX_PC_VSTREAM_CONTROL_SIZE(uint32_t val) +{ + return ((val) << A4XX_PC_VSTREAM_CONTROL_SIZE__SHIFT) & A4XX_PC_VSTREAM_CONTROL_SIZE__MASK; +} +#define A4XX_PC_VSTREAM_CONTROL_N__MASK 0x07c00000 +#define A4XX_PC_VSTREAM_CONTROL_N__SHIFT 22 +static inline uint32_t A4XX_PC_VSTREAM_CONTROL_N(uint32_t val) +{ + return ((val) << A4XX_PC_VSTREAM_CONTROL_N__SHIFT) & A4XX_PC_VSTREAM_CONTROL_N__MASK; +} + #define REG_A4XX_PC_PRIM_VTX_CNTL 0x000021c4 #define A4XX_PC_PRIM_VTX_CNTL_VAROUT__MASK 0x0000000f #define A4XX_PC_PRIM_VTX_CNTL_VAROUT__SHIFT 0 @@ -3752,12 +3804,8 @@ static inline uint32_t A4XX_PC_HS_PARAM_SPACING(enum a4xx_tess_spacing val) { return ((val) << A4XX_PC_HS_PARAM_SPACING__SHIFT) & A4XX_PC_HS_PARAM_SPACING__MASK; } -#define A4XX_PC_HS_PARAM_PRIMTYPE__MASK 0x01800000 -#define A4XX_PC_HS_PARAM_PRIMTYPE__SHIFT 23 -static inline uint32_t A4XX_PC_HS_PARAM_PRIMTYPE(enum adreno_pa_su_sc_draw val) -{ - return ((val) << A4XX_PC_HS_PARAM_PRIMTYPE__SHIFT) & A4XX_PC_HS_PARAM_PRIMTYPE__MASK; -} +#define A4XX_PC_HS_PARAM_CW 0x00800000 +#define A4XX_PC_HS_PARAM_CONNECTED 0x01000000 #define REG_A4XX_VBIF_VERSION 0x00003000 diff --git a/drivers/gpu/drm/msm/adreno/a5xx.xml.h b/drivers/gpu/drm/msm/adreno/a5xx.xml.h new file mode 100644 index 000000000000..f480c94beb62 --- /dev/null +++ b/drivers/gpu/drm/msm/adreno/a5xx.xml.h @@ -0,0 +1,3333 @@ +#ifndef A5XX_XML +#define A5XX_XML + +/* Autogenerated file, DO NOT EDIT manually! + +This file was generated by the rules-ng-ng headergen tool in this git repository: +http://github.com/freedreno/envytools/ +git clone https://github.com/freedreno/envytools.git + +The rules-ng-ng source files this header was generated from are: +- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) +- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) + +Copyright (C) 2013-2016 by the following authors: +- Rob Clark (robclark) +- Ilia Mirkin (imirkin) + +Permission is hereby granted, free of charge, to any person obtaining +a copy of this software and associated documentation files (the +"Software"), to deal in the Software without restriction, including +without limitation the rights to use, copy, modify, merge, publish, +distribute, sublicense, and/or sell copies of the Software, and to +permit persons to whom the Software is furnished to do so, subject to +the following conditions: + +The above copyright notice and this permission notice (including the +next paragraph) shall be included in all copies or substantial +portions of the Software. + +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 COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS 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. +*/ + + +enum a5xx_color_fmt { + RB5_R8_UNORM = 3, + RB5_R5G5B5A1_UNORM = 10, + RB5_R8G8B8A8_UNORM = 48, + RB5_R8G8B8_UNORM = 49, + RB5_R8G8B8A8_UINT = 51, + RB5_R10G10B10A2_UINT = 58, + RB5_R16G16B16A16_FLOAT = 98, +}; + +enum a5xx_tile_mode { + TILE5_LINEAR = 0, + TILE5_2 = 2, + TILE5_3 = 3, +}; + +enum a5xx_vtx_fmt { + VFMT5_8_UNORM = 3, + VFMT5_8_SNORM = 4, + VFMT5_8_UINT = 5, + VFMT5_8_SINT = 6, + VFMT5_8_8_UNORM = 15, + VFMT5_8_8_SNORM = 16, + VFMT5_8_8_UINT = 17, + VFMT5_8_8_SINT = 18, + VFMT5_16_UNORM = 21, + VFMT5_16_SNORM = 22, + VFMT5_16_FLOAT = 23, + VFMT5_16_UINT = 24, + VFMT5_16_SINT = 25, + VFMT5_8_8_8_UNORM = 33, + VFMT5_8_8_8_SNORM = 34, + VFMT5_8_8_8_UINT = 35, + VFMT5_8_8_8_SINT = 36, + VFMT5_8_8_8_8_UNORM = 48, + VFMT5_8_8_8_8_SNORM = 50, + VFMT5_8_8_8_8_UINT = 51, + VFMT5_8_8_8_8_SINT = 52, + VFMT5_16_16_UNORM = 67, + VFMT5_16_16_SNORM = 68, + VFMT5_16_16_FLOAT = 69, + VFMT5_16_16_UINT = 70, + VFMT5_16_16_SINT = 71, + VFMT5_32_UNORM = 72, + VFMT5_32_SNORM = 73, + VFMT5_32_FLOAT = 74, + VFMT5_32_UINT = 75, + VFMT5_32_SINT = 76, + VFMT5_32_FIXED = 77, + VFMT5_16_16_16_UNORM = 88, + VFMT5_16_16_16_SNORM = 89, + VFMT5_16_16_16_FLOAT = 90, + VFMT5_16_16_16_UINT = 91, + VFMT5_16_16_16_SINT = 92, + VFMT5_16_16_16_16_UNORM = 96, + VFMT5_16_16_16_16_SNORM = 97, + VFMT5_16_16_16_16_FLOAT = 98, + VFMT5_16_16_16_16_UINT = 99, + VFMT5_16_16_16_16_SINT = 100, + VFMT5_32_32_UNORM = 101, + VFMT5_32_32_SNORM = 102, + VFMT5_32_32_FLOAT = 103, + VFMT5_32_32_UINT = 104, + VFMT5_32_32_SINT = 105, + VFMT5_32_32_FIXED = 106, + VFMT5_32_32_32_UNORM = 112, + VFMT5_32_32_32_SNORM = 113, + VFMT5_32_32_32_UINT = 114, + VFMT5_32_32_32_SINT = 115, + VFMT5_32_32_32_FLOAT = 116, + VFMT5_32_32_32_FIXED = 117, + VFMT5_32_32_32_32_UNORM = 128, + VFMT5_32_32_32_32_SNORM = 129, + VFMT5_32_32_32_32_FLOAT = 130, + VFMT5_32_32_32_32_UINT = 131, + VFMT5_32_32_32_32_SINT = 132, + VFMT5_32_32_32_32_FIXED = 133, +}; + +enum a5xx_tex_fmt { + TFMT5_A8_UNORM = 2, + TFMT5_8_UNORM = 3, + TFMT5_4_4_4_4_UNORM = 8, + TFMT5_5_6_5_UNORM = 14, + TFMT5_L8_A8_UNORM = 19, + TFMT5_16_FLOAT = 23, + TFMT5_8_8_8_8_UNORM = 48, + TFMT5_10_10_10_2_UNORM = 54, + TFMT5_16_16_FLOAT = 69, + TFMT5_32_FLOAT = 74, + TFMT5_16_16_16_16_FLOAT = 98, + TFMT5_32_32_FLOAT = 103, + TFMT5_32_32_32_32_FLOAT = 130, + TFMT5_X8Z24_UNORM = 160, +}; + +enum a5xx_tex_fetchsize { + TFETCH5_1_BYTE = 0, + TFETCH5_2_BYTE = 1, + TFETCH5_4_BYTE = 2, + TFETCH5_8_BYTE = 3, + TFETCH5_16_BYTE = 4, +}; + +enum a5xx_depth_format { + DEPTH5_NONE = 0, + DEPTH5_16 = 1, + DEPTH5_24_8 = 2, + DEPTH5_32 = 4, +}; + +enum a5xx_tex_filter { + A5XX_TEX_NEAREST = 0, + A5XX_TEX_LINEAR = 1, + A5XX_TEX_ANISO = 2, +}; + +enum a5xx_tex_clamp { + A5XX_TEX_REPEAT = 0, + A5XX_TEX_CLAMP_TO_EDGE = 1, + A5XX_TEX_MIRROR_REPEAT = 2, + A5XX_TEX_CLAMP_TO_BORDER = 3, + A5XX_TEX_MIRROR_CLAMP = 4, +}; + +enum a5xx_tex_aniso { + A5XX_TEX_ANISO_1 = 0, + A5XX_TEX_ANISO_2 = 1, + A5XX_TEX_ANISO_4 = 2, + A5XX_TEX_ANISO_8 = 3, + A5XX_TEX_ANISO_16 = 4, +}; + +enum a5xx_tex_swiz { + A5XX_TEX_X = 0, + A5XX_TEX_Y = 1, + A5XX_TEX_Z = 2, + A5XX_TEX_W = 3, + A5XX_TEX_ZERO = 4, + A5XX_TEX_ONE = 5, +}; + +enum a5xx_tex_type { + A5XX_TEX_1D = 0, + A5XX_TEX_2D = 1, + A5XX_TEX_CUBE = 2, + A5XX_TEX_3D = 3, +}; + +#define A5XX_INT0_RBBM_GPU_IDLE 0x00000001 +#define A5XX_INT0_RBBM_AHB_ERROR 0x00000002 +#define A5XX_INT0_RBBM_TRANSFER_TIMEOUT 0x00000004 +#define A5XX_INT0_RBBM_ME_MS_TIMEOUT 0x00000008 +#define A5XX_INT0_RBBM_PFP_MS_TIMEOUT 0x00000010 +#define A5XX_INT0_RBBM_ETS_MS_TIMEOUT 0x00000020 +#define A5XX_INT0_RBBM_ATB_ASYNC_OVERFLOW 0x00000040 +#define A5XX_INT0_RBBM_GPC_ERROR 0x00000080 +#define A5XX_INT0_CP_SW 0x00000100 +#define A5XX_INT0_CP_HW_ERROR 0x00000200 +#define A5XX_INT0_CP_CCU_FLUSH_DEPTH_TS 0x00000400 +#define A5XX_INT0_CP_CCU_FLUSH_COLOR_TS 0x00000800 +#define A5XX_INT0_CP_CCU_RESOLVE_TS 0x00001000 +#define A5XX_INT0_CP_IB2 0x00002000 +#define A5XX_INT0_CP_IB1 0x00004000 +#define A5XX_INT0_CP_RB 0x00008000 +#define A5XX_INT0_CP_UNUSED_1 0x00010000 +#define A5XX_INT0_CP_RB_DONE_TS 0x00020000 +#define A5XX_INT0_CP_WT_DONE_TS 0x00040000 +#define A5XX_INT0_UNKNOWN_1 0x00080000 +#define A5XX_INT0_CP_CACHE_FLUSH_TS 0x00100000 +#define A5XX_INT0_UNUSED_2 0x00200000 +#define A5XX_INT0_RBBM_ATB_BUS_OVERFLOW 0x00400000 +#define A5XX_INT0_MISC_HANG_DETECT 0x00800000 +#define A5XX_INT0_UCHE_OOB_ACCESS 0x01000000 +#define A5XX_INT0_UCHE_TRAP_INTR 0x02000000 +#define A5XX_INT0_DEBBUS_INTR_0 0x04000000 +#define A5XX_INT0_DEBBUS_INTR_1 0x08000000 +#define A5XX_INT0_GPMU_VOLTAGE_DROOP 0x10000000 +#define A5XX_INT0_GPMU_FIRMWARE 0x20000000 +#define A5XX_INT0_ISDB_CPU_IRQ 0x40000000 +#define A5XX_INT0_ISDB_UNDER_DEBUG 0x80000000 +#define A5XX_CP_INT_CP_OPCODE_ERROR 0x00000001 +#define A5XX_CP_INT_CP_RESERVED_BIT_ERROR 0x00000002 +#define A5XX_CP_INT_CP_HW_FAULT_ERROR 0x00000004 +#define A5XX_CP_INT_CP_DMA_ERROR 0x00000008 +#define A5XX_CP_INT_CP_REGISTER_PROTECTION_ERROR 0x00000010 +#define A5XX_CP_INT_CP_AHB_ERROR 0x00000020 +#define REG_A5XX_CP_RB_BASE 0x00000800 + +#define REG_A5XX_CP_RB_BASE_HI 0x00000801 + +#define REG_A5XX_CP_RB_CNTL 0x00000802 + +#define REG_A5XX_CP_RB_RPTR_ADDR 0x00000804 + +#define REG_A5XX_CP_RB_RPTR_ADDR_HI 0x00000805 + +#define REG_A5XX_CP_RB_RPTR 0x00000806 + +#define REG_A5XX_CP_RB_WPTR 0x00000807 + +#define REG_A5XX_CP_PFP_STAT_ADDR 0x00000808 + +#define REG_A5XX_CP_PFP_STAT_DATA 0x00000809 + +#define REG_A5XX_CP_DRAW_STATE_ADDR 0x0000080b + +#define REG_A5XX_CP_DRAW_STATE_DATA 0x0000080c + +#define REG_A5XX_CP_CRASH_SCRIPT_BASE_LO 0x00000817 + +#define REG_A5XX_CP_CRASH_SCRIPT_BASE_HI 0x00000818 + +#define REG_A5XX_CP_CRASH_DUMP_CNTL 0x00000819 + +#define REG_A5XX_CP_ME_STAT_ADDR 0x0000081a + +#define REG_A5XX_CP_ROQ_THRESHOLDS_1 0x0000081f + +#define REG_A5XX_CP_ROQ_THRESHOLDS_2 0x00000820 + +#define REG_A5XX_CP_ROQ_DBG_ADDR 0x00000821 + +#define REG_A5XX_CP_ROQ_DBG_DATA 0x00000822 + +#define REG_A5XX_CP_MEQ_DBG_ADDR 0x00000823 + +#define REG_A5XX_CP_MEQ_DBG_DATA 0x00000824 + +#define REG_A5XX_CP_MEQ_THRESHOLDS 0x00000825 + +#define REG_A5XX_CP_MERCIU_SIZE 0x00000826 + +#define REG_A5XX_CP_MERCIU_DBG_ADDR 0x00000827 + +#define REG_A5XX_CP_MERCIU_DBG_DATA_1 0x00000828 + +#define REG_A5XX_CP_MERCIU_DBG_DATA_2 0x00000829 + +#define REG_A5XX_CP_PFP_UCODE_DBG_ADDR 0x0000082a + +#define REG_A5XX_CP_PFP_UCODE_DBG_DATA 0x0000082b + +#define REG_A5XX_CP_ME_UCODE_DBG_ADDR 0x0000082f + +#define REG_A5XX_CP_ME_UCODE_DBG_DATA 0x00000830 + +#define REG_A5XX_CP_CNTL 0x00000831 + +#define REG_A5XX_CP_PFP_ME_CNTL 0x00000832 + +#define REG_A5XX_CP_CHICKEN_DBG 0x00000833 + +#define REG_A5XX_CP_PFP_INSTR_BASE_LO 0x00000835 + +#define REG_A5XX_CP_PFP_INSTR_BASE_HI 0x00000836 + +#define REG_A5XX_CP_ME_INSTR_BASE_LO 0x00000838 + +#define REG_A5XX_CP_ME_INSTR_BASE_HI 0x00000839 + +#define REG_A5XX_CP_CONTEXT_SWITCH_CNTL 0x0000083b + +#define REG_A5XX_CP_CONTEXT_SWITCH_RESTORE_ADDR_LO 0x0000083c + +#define REG_A5XX_CP_CONTEXT_SWITCH_RESTORE_ADDR_HI 0x0000083d + +#define REG_A5XX_CP_CONTEXT_SWITCH_SAVE_ADDR_LO 0x0000083e + +#define REG_A5XX_CP_CONTEXT_SWITCH_SAVE_ADDR_HI 0x0000083f + +#define REG_A5XX_CP_CONTEXT_SWITCH_SMMU_INFO_LO 0x00000840 + +#define REG_A5XX_CP_CONTEXT_SWITCH_SMMU_INFO_HI 0x00000841 + +#define REG_A5XX_CP_ADDR_MODE_CNTL 0x00000860 + +#define REG_A5XX_CP_ME_STAT_DATA 0x00000b14 + +#define REG_A5XX_CP_WFI_PEND_CTR 0x00000b15 + +#define REG_A5XX_CP_INTERRUPT_STATUS 0x00000b18 + +#define REG_A5XX_CP_HW_FAULT 0x00000b1a + +#define REG_A5XX_CP_PROTECT_STATUS 0x00000b1c + +#define REG_A5XX_CP_IB1_BASE 0x00000b1f + +#define REG_A5XX_CP_IB1_BASE_HI 0x00000b20 + +#define REG_A5XX_CP_IB1_BUFSZ 0x00000b21 + +#define REG_A5XX_CP_IB2_BASE 0x00000b22 + +#define REG_A5XX_CP_IB2_BASE_HI 0x00000b23 + +#define REG_A5XX_CP_IB2_BUFSZ 0x00000b24 + +static inline uint32_t REG_A5XX_CP_SCRATCH(uint32_t i0) { return 0x00000b78 + 0x1*i0; } + +static inline uint32_t REG_A5XX_CP_SCRATCH_REG(uint32_t i0) { return 0x00000b78 + 0x1*i0; } + +static inline uint32_t REG_A5XX_CP_PROTECT(uint32_t i0) { return 0x00000880 + 0x1*i0; } + +static inline uint32_t REG_A5XX_CP_PROTECT_REG(uint32_t i0) { return 0x00000880 + 0x1*i0; } +#define A5XX_CP_PROTECT_REG_BASE_ADDR__MASK 0x0001ffff +#define A5XX_CP_PROTECT_REG_BASE_ADDR__SHIFT 0 +static inline uint32_t A5XX_CP_PROTECT_REG_BASE_ADDR(uint32_t val) +{ + return ((val) << A5XX_CP_PROTECT_REG_BASE_ADDR__SHIFT) & A5XX_CP_PROTECT_REG_BASE_ADDR__MASK; +} +#define A5XX_CP_PROTECT_REG_MASK_LEN__MASK 0x1f000000 +#define A5XX_CP_PROTECT_REG_MASK_LEN__SHIFT 24 +static inline uint32_t A5XX_CP_PROTECT_REG_MASK_LEN(uint32_t val) +{ + return ((val) << A5XX_CP_PROTECT_REG_MASK_LEN__SHIFT) & A5XX_CP_PROTECT_REG_MASK_LEN__MASK; +} +#define A5XX_CP_PROTECT_REG_TRAP_WRITE 0x20000000 +#define A5XX_CP_PROTECT_REG_TRAP_READ 0x40000000 + +#define REG_A5XX_CP_PROTECT_CNTL 0x000008a0 + +#define REG_A5XX_CP_AHB_FAULT 0x00000b1b + +#define REG_A5XX_CP_PERFCTR_CP_SEL_0 0x00000bb0 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_1 0x00000bb1 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_2 0x00000bb2 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_3 0x00000bb3 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_4 0x00000bb4 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_5 0x00000bb5 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_6 0x00000bb6 + +#define REG_A5XX_CP_PERFCTR_CP_SEL_7 0x00000bb7 + +#define REG_A5XX_VSC_ADDR_MODE_CNTL 0x00000bc1 + +#define REG_A5XX_CP_POWERCTR_CP_SEL_0 0x00000bba + +#define REG_A5XX_CP_POWERCTR_CP_SEL_1 0x00000bbb + +#define REG_A5XX_CP_POWERCTR_CP_SEL_2 0x00000bbc + +#define REG_A5XX_CP_POWERCTR_CP_SEL_3 0x00000bbd + +#define REG_A5XX_RBBM_CFG_DBGBUS_SEL_A 0x00000004 + +#define REG_A5XX_RBBM_CFG_DBGBUS_SEL_B 0x00000005 + +#define REG_A5XX_RBBM_CFG_DBGBUS_SEL_C 0x00000006 + +#define REG_A5XX_RBBM_CFG_DBGBUS_SEL_D 0x00000007 + +#define REG_A5XX_RBBM_CFG_DBGBUS_CNTLT 0x00000008 + +#define REG_A5XX_RBBM_CFG_DBGBUS_CNTLM 0x00000009 + +#define REG_A5XX_RBBM_CFG_DEBBUS_CTLTM_ENABLE_SHIFT 0x00000018 + +#define REG_A5XX_RBBM_CFG_DBGBUS_OPL 0x0000000a + +#define REG_A5XX_RBBM_CFG_DBGBUS_OPE 0x0000000b + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTL_0 0x0000000c + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTL_1 0x0000000d + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTL_2 0x0000000e + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTL_3 0x0000000f + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKL_0 0x00000010 + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKL_1 0x00000011 + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKL_2 0x00000012 + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKL_3 0x00000013 + +#define REG_A5XX_RBBM_CFG_DBGBUS_BYTEL_0 0x00000014 + +#define REG_A5XX_RBBM_CFG_DBGBUS_BYTEL_1 0x00000015 + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTE_0 0x00000016 + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTE_1 0x00000017 + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTE_2 0x00000018 + +#define REG_A5XX_RBBM_CFG_DBGBUS_IVTE_3 0x00000019 + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKE_0 0x0000001a + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKE_1 0x0000001b + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKE_2 0x0000001c + +#define REG_A5XX_RBBM_CFG_DBGBUS_MASKE_3 0x0000001d + +#define REG_A5XX_RBBM_CFG_DBGBUS_NIBBLEE 0x0000001e + +#define REG_A5XX_RBBM_CFG_DBGBUS_PTRC0 0x0000001f + +#define REG_A5XX_RBBM_CFG_DBGBUS_PTRC1 0x00000020 + +#define REG_A5XX_RBBM_CFG_DBGBUS_LOADREG 0x00000021 + +#define REG_A5XX_RBBM_CFG_DBGBUS_IDX 0x00000022 + +#define REG_A5XX_RBBM_CFG_DBGBUS_CLRC 0x00000023 + +#define REG_A5XX_RBBM_CFG_DBGBUS_LOADIVT 0x00000024 + +#define REG_A5XX_RBBM_INTERFACE_HANG_INT_CNTL 0x0000002f + +#define REG_A5XX_RBBM_INT_CLEAR_CMD 0x00000037 + +#define REG_A5XX_RBBM_INT_0_MASK 0x00000038 +#define A5XX_RBBM_INT_0_MASK_RBBM_GPU_IDLE 0x00000001 +#define A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR 0x00000002 +#define A5XX_RBBM_INT_0_MASK_RBBM_TRANSFER_TIMEOUT 0x00000004 +#define A5XX_RBBM_INT_0_MASK_RBBM_ME_MS_TIMEOUT 0x00000008 +#define A5XX_RBBM_INT_0_MASK_RBBM_PFP_MS_TIMEOUT 0x00000010 +#define A5XX_RBBM_INT_0_MASK_RBBM_ETS_MS_TIMEOUT 0x00000020 +#define A5XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNC_OVERFLOW 0x00000040 +#define A5XX_RBBM_INT_0_MASK_RBBM_GPC_ERROR 0x00000080 +#define A5XX_RBBM_INT_0_MASK_CP_SW 0x00000100 +#define A5XX_RBBM_INT_0_MASK_CP_HW_ERROR 0x00000200 +#define A5XX_RBBM_INT_0_MASK_CP_CCU_FLUSH_DEPTH_TS 0x00000400 +#define A5XX_RBBM_INT_0_MASK_CP_CCU_FLUSH_COLOR_TS 0x00000800 +#define A5XX_RBBM_INT_0_MASK_CP_CCU_RESOLVE_TS 0x00001000 +#define A5XX_RBBM_INT_0_MASK_CP_IB2 0x00002000 +#define A5XX_RBBM_INT_0_MASK_CP_IB1 0x00004000 +#define A5XX_RBBM_INT_0_MASK_CP_RB 0x00008000 +#define A5XX_RBBM_INT_0_MASK_CP_RB_DONE_TS 0x00020000 +#define A5XX_RBBM_INT_0_MASK_CP_WT_DONE_TS 0x00040000 +#define A5XX_RBBM_INT_0_MASK_CP_CACHE_FLUSH_TS 0x00100000 +#define A5XX_RBBM_INT_0_MASK_RBBM_ATB_BUS_OVERFLOW 0x00400000 +#define A5XX_RBBM_INT_0_MASK_MISC_HANG_DETECT 0x00800000 +#define A5XX_RBBM_INT_0_MASK_UCHE_OOB_ACCESS 0x01000000 +#define A5XX_RBBM_INT_0_MASK_UCHE_TRAP_INTR 0x02000000 +#define A5XX_RBBM_INT_0_MASK_DEBBUS_INTR_0 0x04000000 +#define A5XX_RBBM_INT_0_MASK_DEBBUS_INTR_1 0x08000000 +#define A5XX_RBBM_INT_0_MASK_GPMU_VOLTAGE_DROOP 0x10000000 +#define A5XX_RBBM_INT_0_MASK_GPMU_FIRMWARE 0x20000000 +#define A5XX_RBBM_INT_0_MASK_ISDB_CPU_IRQ 0x40000000 +#define A5XX_RBBM_INT_0_MASK_ISDB_UNDER_DEBUG 0x80000000 + +#define REG_A5XX_RBBM_AHB_DBG_CNTL 0x0000003f + +#define REG_A5XX_RBBM_EXT_VBIF_DBG_CNTL 0x00000041 + +#define REG_A5XX_RBBM_SW_RESET_CMD 0x00000043 + +#define REG_A5XX_RBBM_BLOCK_SW_RESET_CMD 0x00000045 + +#define REG_A5XX_RBBM_BLOCK_SW_RESET_CMD2 0x00000046 + +#define REG_A5XX_RBBM_DBG_LO_HI_GPIO 0x00000048 + +#define REG_A5XX_RBBM_EXT_TRACE_BUS_CNTL 0x00000049 + +#define REG_A5XX_RBBM_CLOCK_CNTL_TP0 0x0000004a + +#define REG_A5XX_RBBM_CLOCK_CNTL_TP1 0x0000004b + +#define REG_A5XX_RBBM_CLOCK_CNTL_TP2 0x0000004c + +#define REG_A5XX_RBBM_CLOCK_CNTL_TP3 0x0000004d + +#define REG_A5XX_RBBM_CLOCK_CNTL2_TP0 0x0000004e + +#define REG_A5XX_RBBM_CLOCK_CNTL2_TP1 0x0000004f + +#define REG_A5XX_RBBM_CLOCK_CNTL2_TP2 0x00000050 + +#define REG_A5XX_RBBM_CLOCK_CNTL2_TP3 0x00000051 + +#define REG_A5XX_RBBM_CLOCK_CNTL3_TP0 0x00000052 + +#define REG_A5XX_RBBM_CLOCK_CNTL3_TP1 0x00000053 + +#define REG_A5XX_RBBM_CLOCK_CNTL3_TP2 0x00000054 + +#define REG_A5XX_RBBM_CLOCK_CNTL3_TP3 0x00000055 + +#define REG_A5XX_RBBM_READ_AHB_THROUGH_DBG 0x00000059 + +#define REG_A5XX_RBBM_CLOCK_CNTL_UCHE 0x0000005a + +#define REG_A5XX_RBBM_CLOCK_CNTL2_UCHE 0x0000005b + +#define REG_A5XX_RBBM_CLOCK_CNTL3_UCHE 0x0000005c + +#define REG_A5XX_RBBM_CLOCK_CNTL4_UCHE 0x0000005d + +#define REG_A5XX_RBBM_CLOCK_HYST_UCHE 0x0000005e + +#define REG_A5XX_RBBM_CLOCK_DELAY_UCHE 0x0000005f + +#define REG_A5XX_RBBM_CLOCK_MODE_GPC 0x00000060 + +#define REG_A5XX_RBBM_CLOCK_DELAY_GPC 0x00000061 + +#define REG_A5XX_RBBM_CLOCK_HYST_GPC 0x00000062 + +#define REG_A5XX_RBBM_CLOCK_CNTL_TSE_RAS_RBBM 0x00000063 + +#define REG_A5XX_RBBM_CLOCK_HYST_TSE_RAS_RBBM 0x00000064 + +#define REG_A5XX_RBBM_CLOCK_DELAY_TSE_RAS_RBBM 0x00000065 + +#define REG_A5XX_RBBM_CLOCK_DELAY_HLSQ 0x00000066 + +#define REG_A5XX_RBBM_CLOCK_CNTL 0x00000067 + +#define REG_A5XX_RBBM_CLOCK_CNTL_SP0 0x00000068 + +#define REG_A5XX_RBBM_CLOCK_CNTL_SP1 0x00000069 + +#define REG_A5XX_RBBM_CLOCK_CNTL_SP2 0x0000006a + +#define REG_A5XX_RBBM_CLOCK_CNTL_SP3 0x0000006b + +#define REG_A5XX_RBBM_CLOCK_CNTL2_SP0 0x0000006c + +#define REG_A5XX_RBBM_CLOCK_CNTL2_SP1 0x0000006d + +#define REG_A5XX_RBBM_CLOCK_CNTL2_SP2 0x0000006e + +#define REG_A5XX_RBBM_CLOCK_CNTL2_SP3 0x0000006f + +#define REG_A5XX_RBBM_CLOCK_HYST_SP0 0x00000070 + +#define REG_A5XX_RBBM_CLOCK_HYST_SP1 0x00000071 + +#define REG_A5XX_RBBM_CLOCK_HYST_SP2 0x00000072 + +#define REG_A5XX_RBBM_CLOCK_HYST_SP3 0x00000073 + +#define REG_A5XX_RBBM_CLOCK_DELAY_SP0 0x00000074 + +#define REG_A5XX_RBBM_CLOCK_DELAY_SP1 0x00000075 + +#define REG_A5XX_RBBM_CLOCK_DELAY_SP2 0x00000076 + +#define REG_A5XX_RBBM_CLOCK_DELAY_SP3 0x00000077 + +#define REG_A5XX_RBBM_CLOCK_CNTL_RB0 0x00000078 + +#define REG_A5XX_RBBM_CLOCK_CNTL_RB1 0x00000079 + +#define REG_A5XX_RBBM_CLOCK_CNTL_RB2 0x0000007a + +#define REG_A5XX_RBBM_CLOCK_CNTL_RB3 0x0000007b + +#define REG_A5XX_RBBM_CLOCK_CNTL2_RB0 0x0000007c + +#define REG_A5XX_RBBM_CLOCK_CNTL2_RB1 0x0000007d + +#define REG_A5XX_RBBM_CLOCK_CNTL2_RB2 0x0000007e + +#define REG_A5XX_RBBM_CLOCK_CNTL2_RB3 0x0000007f + +#define REG_A5XX_RBBM_CLOCK_HYST_RAC 0x00000080 + +#define REG_A5XX_RBBM_CLOCK_DELAY_RAC 0x00000081 + +#define REG_A5XX_RBBM_CLOCK_CNTL_CCU0 0x00000082 + +#define REG_A5XX_RBBM_CLOCK_CNTL_CCU1 0x00000083 + +#define REG_A5XX_RBBM_CLOCK_CNTL_CCU2 0x00000084 + +#define REG_A5XX_RBBM_CLOCK_CNTL_CCU3 0x00000085 + +#define REG_A5XX_RBBM_CLOCK_HYST_RB_CCU0 0x00000086 + +#define REG_A5XX_RBBM_CLOCK_HYST_RB_CCU1 0x00000087 + +#define REG_A5XX_RBBM_CLOCK_HYST_RB_CCU2 0x00000088 + +#define REG_A5XX_RBBM_CLOCK_HYST_RB_CCU3 0x00000089 + +#define REG_A5XX_RBBM_CLOCK_CNTL_RAC 0x0000008a + +#define REG_A5XX_RBBM_CLOCK_CNTL2_RAC 0x0000008b + +#define REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_0 0x0000008c + +#define REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_1 0x0000008d + +#define REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_2 0x0000008e + +#define REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_3 0x0000008f + +#define REG_A5XX_RBBM_CLOCK_HYST_VFD 0x00000090 + +#define REG_A5XX_RBBM_CLOCK_MODE_VFD 0x00000091 + +#define REG_A5XX_RBBM_CLOCK_DELAY_VFD 0x00000092 + +#define REG_A5XX_RBBM_AHB_CNTL0 0x00000093 + +#define REG_A5XX_RBBM_AHB_CNTL1 0x00000094 + +#define REG_A5XX_RBBM_AHB_CNTL2 0x00000095 + +#define REG_A5XX_RBBM_AHB_CMD 0x00000096 + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL11 0x0000009c + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL12 0x0000009d + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL13 0x0000009e + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL14 0x0000009f + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL15 0x000000a0 + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL16 0x000000a1 + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL17 0x000000a2 + +#define REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL18 0x000000a3 + +#define REG_A5XX_RBBM_CLOCK_DELAY_TP0 0x000000a4 + +#define REG_A5XX_RBBM_CLOCK_DELAY_TP1 0x000000a5 + +#define REG_A5XX_RBBM_CLOCK_DELAY_TP2 0x000000a6 + +#define REG_A5XX_RBBM_CLOCK_DELAY_TP3 0x000000a7 + +#define REG_A5XX_RBBM_CLOCK_DELAY2_TP0 0x000000a8 + +#define REG_A5XX_RBBM_CLOCK_DELAY2_TP1 0x000000a9 + +#define REG_A5XX_RBBM_CLOCK_DELAY2_TP2 0x000000aa + +#define REG_A5XX_RBBM_CLOCK_DELAY2_TP3 0x000000ab + +#define REG_A5XX_RBBM_CLOCK_DELAY3_TP0 0x000000ac + +#define REG_A5XX_RBBM_CLOCK_DELAY3_TP1 0x000000ad + +#define REG_A5XX_RBBM_CLOCK_DELAY3_TP2 0x000000ae + +#define REG_A5XX_RBBM_CLOCK_DELAY3_TP3 0x000000af + +#define REG_A5XX_RBBM_CLOCK_HYST_TP0 0x000000b0 + +#define REG_A5XX_RBBM_CLOCK_HYST_TP1 0x000000b1 + +#define REG_A5XX_RBBM_CLOCK_HYST_TP2 0x000000b2 + +#define REG_A5XX_RBBM_CLOCK_HYST_TP3 0x000000b3 + +#define REG_A5XX_RBBM_CLOCK_HYST2_TP0 0x000000b4 + +#define REG_A5XX_RBBM_CLOCK_HYST2_TP1 0x000000b5 + +#define REG_A5XX_RBBM_CLOCK_HYST2_TP2 0x000000b6 + +#define REG_A5XX_RBBM_CLOCK_HYST2_TP3 0x000000b7 + +#define REG_A5XX_RBBM_CLOCK_HYST3_TP0 0x000000b8 + +#define REG_A5XX_RBBM_CLOCK_HYST3_TP1 0x000000b9 + +#define REG_A5XX_RBBM_CLOCK_HYST3_TP2 0x000000ba + +#define REG_A5XX_RBBM_CLOCK_HYST3_TP3 0x000000bb + +#define REG_A5XX_RBBM_CLOCK_CNTL_GPMU 0x000000c8 + +#define REG_A5XX_RBBM_CLOCK_DELAY_GPMU 0x000000c9 + +#define REG_A5XX_RBBM_CLOCK_HYST_GPMU 0x000000ca + +#define REG_A5XX_RBBM_PERFCTR_CP_0_LO 0x000003a0 + +#define REG_A5XX_RBBM_PERFCTR_CP_0_HI 0x000003a1 + +#define REG_A5XX_RBBM_PERFCTR_CP_1_LO 0x000003a2 + +#define REG_A5XX_RBBM_PERFCTR_CP_1_HI 0x000003a3 + +#define REG_A5XX_RBBM_PERFCTR_CP_2_LO 0x000003a4 + +#define REG_A5XX_RBBM_PERFCTR_CP_2_HI 0x000003a5 + +#define REG_A5XX_RBBM_PERFCTR_CP_3_LO 0x000003a6 + +#define REG_A5XX_RBBM_PERFCTR_CP_3_HI 0x000003a7 + +#define REG_A5XX_RBBM_PERFCTR_CP_4_LO 0x000003a8 + +#define REG_A5XX_RBBM_PERFCTR_CP_4_HI 0x000003a9 + +#define REG_A5XX_RBBM_PERFCTR_CP_5_LO 0x000003aa + +#define REG_A5XX_RBBM_PERFCTR_CP_5_HI 0x000003ab + +#define REG_A5XX_RBBM_PERFCTR_CP_6_LO 0x000003ac + +#define REG_A5XX_RBBM_PERFCTR_CP_6_HI 0x000003ad + +#define REG_A5XX_RBBM_PERFCTR_CP_7_LO 0x000003ae + +#define REG_A5XX_RBBM_PERFCTR_CP_7_HI 0x000003af + +#define REG_A5XX_RBBM_PERFCTR_RBBM_0_LO 0x000003b0 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_0_HI 0x000003b1 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_1_LO 0x000003b2 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_1_HI 0x000003b3 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_2_LO 0x000003b4 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_2_HI 0x000003b5 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_3_LO 0x000003b6 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_3_HI 0x000003b7 + +#define REG_A5XX_RBBM_PERFCTR_PC_0_LO 0x000003b8 + +#define REG_A5XX_RBBM_PERFCTR_PC_0_HI 0x000003b9 + +#define REG_A5XX_RBBM_PERFCTR_PC_1_LO 0x000003ba + +#define REG_A5XX_RBBM_PERFCTR_PC_1_HI 0x000003bb + +#define REG_A5XX_RBBM_PERFCTR_PC_2_LO 0x000003bc + +#define REG_A5XX_RBBM_PERFCTR_PC_2_HI 0x000003bd + +#define REG_A5XX_RBBM_PERFCTR_PC_3_LO 0x000003be + +#define REG_A5XX_RBBM_PERFCTR_PC_3_HI 0x000003bf + +#define REG_A5XX_RBBM_PERFCTR_PC_4_LO 0x000003c0 + +#define REG_A5XX_RBBM_PERFCTR_PC_4_HI 0x000003c1 + +#define REG_A5XX_RBBM_PERFCTR_PC_5_LO 0x000003c2 + +#define REG_A5XX_RBBM_PERFCTR_PC_5_HI 0x000003c3 + +#define REG_A5XX_RBBM_PERFCTR_PC_6_LO 0x000003c4 + +#define REG_A5XX_RBBM_PERFCTR_PC_6_HI 0x000003c5 + +#define REG_A5XX_RBBM_PERFCTR_PC_7_LO 0x000003c6 + +#define REG_A5XX_RBBM_PERFCTR_PC_7_HI 0x000003c7 + +#define REG_A5XX_RBBM_PERFCTR_VFD_0_LO 0x000003c8 + +#define REG_A5XX_RBBM_PERFCTR_VFD_0_HI 0x000003c9 + +#define REG_A5XX_RBBM_PERFCTR_VFD_1_LO 0x000003ca + +#define REG_A5XX_RBBM_PERFCTR_VFD_1_HI 0x000003cb + +#define REG_A5XX_RBBM_PERFCTR_VFD_2_LO 0x000003cc + +#define REG_A5XX_RBBM_PERFCTR_VFD_2_HI 0x000003cd + +#define REG_A5XX_RBBM_PERFCTR_VFD_3_LO 0x000003ce + +#define REG_A5XX_RBBM_PERFCTR_VFD_3_HI 0x000003cf + +#define REG_A5XX_RBBM_PERFCTR_VFD_4_LO 0x000003d0 + +#define REG_A5XX_RBBM_PERFCTR_VFD_4_HI 0x000003d1 + +#define REG_A5XX_RBBM_PERFCTR_VFD_5_LO 0x000003d2 + +#define REG_A5XX_RBBM_PERFCTR_VFD_5_HI 0x000003d3 + +#define REG_A5XX_RBBM_PERFCTR_VFD_6_LO 0x000003d4 + +#define REG_A5XX_RBBM_PERFCTR_VFD_6_HI 0x000003d5 + +#define REG_A5XX_RBBM_PERFCTR_VFD_7_LO 0x000003d6 + +#define REG_A5XX_RBBM_PERFCTR_VFD_7_HI 0x000003d7 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_0_LO 0x000003d8 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_0_HI 0x000003d9 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_1_LO 0x000003da + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_1_HI 0x000003db + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_2_LO 0x000003dc + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_2_HI 0x000003dd + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_3_LO 0x000003de + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_3_HI 0x000003df + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_4_LO 0x000003e0 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_4_HI 0x000003e1 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_5_LO 0x000003e2 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_5_HI 0x000003e3 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_6_LO 0x000003e4 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_6_HI 0x000003e5 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_7_LO 0x000003e6 + +#define REG_A5XX_RBBM_PERFCTR_HLSQ_7_HI 0x000003e7 + +#define REG_A5XX_RBBM_PERFCTR_VPC_0_LO 0x000003e8 + +#define REG_A5XX_RBBM_PERFCTR_VPC_0_HI 0x000003e9 + +#define REG_A5XX_RBBM_PERFCTR_VPC_1_LO 0x000003ea + +#define REG_A5XX_RBBM_PERFCTR_VPC_1_HI 0x000003eb + +#define REG_A5XX_RBBM_PERFCTR_VPC_2_LO 0x000003ec + +#define REG_A5XX_RBBM_PERFCTR_VPC_2_HI 0x000003ed + +#define REG_A5XX_RBBM_PERFCTR_VPC_3_LO 0x000003ee + +#define REG_A5XX_RBBM_PERFCTR_VPC_3_HI 0x000003ef + +#define REG_A5XX_RBBM_PERFCTR_CCU_0_LO 0x000003f0 + +#define REG_A5XX_RBBM_PERFCTR_CCU_0_HI 0x000003f1 + +#define REG_A5XX_RBBM_PERFCTR_CCU_1_LO 0x000003f2 + +#define REG_A5XX_RBBM_PERFCTR_CCU_1_HI 0x000003f3 + +#define REG_A5XX_RBBM_PERFCTR_CCU_2_LO 0x000003f4 + +#define REG_A5XX_RBBM_PERFCTR_CCU_2_HI 0x000003f5 + +#define REG_A5XX_RBBM_PERFCTR_CCU_3_LO 0x000003f6 + +#define REG_A5XX_RBBM_PERFCTR_CCU_3_HI 0x000003f7 + +#define REG_A5XX_RBBM_PERFCTR_TSE_0_LO 0x000003f8 + +#define REG_A5XX_RBBM_PERFCTR_TSE_0_HI 0x000003f9 + +#define REG_A5XX_RBBM_PERFCTR_TSE_1_LO 0x000003fa + +#define REG_A5XX_RBBM_PERFCTR_TSE_1_HI 0x000003fb + +#define REG_A5XX_RBBM_PERFCTR_TSE_2_LO 0x000003fc + +#define REG_A5XX_RBBM_PERFCTR_TSE_2_HI 0x000003fd + +#define REG_A5XX_RBBM_PERFCTR_TSE_3_LO 0x000003fe + +#define REG_A5XX_RBBM_PERFCTR_TSE_3_HI 0x000003ff + +#define REG_A5XX_RBBM_PERFCTR_RAS_0_LO 0x00000400 + +#define REG_A5XX_RBBM_PERFCTR_RAS_0_HI 0x00000401 + +#define REG_A5XX_RBBM_PERFCTR_RAS_1_LO 0x00000402 + +#define REG_A5XX_RBBM_PERFCTR_RAS_1_HI 0x00000403 + +#define REG_A5XX_RBBM_PERFCTR_RAS_2_LO 0x00000404 + +#define REG_A5XX_RBBM_PERFCTR_RAS_2_HI 0x00000405 + +#define REG_A5XX_RBBM_PERFCTR_RAS_3_LO 0x00000406 + +#define REG_A5XX_RBBM_PERFCTR_RAS_3_HI 0x00000407 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_0_LO 0x00000408 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_0_HI 0x00000409 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_1_LO 0x0000040a + +#define REG_A5XX_RBBM_PERFCTR_UCHE_1_HI 0x0000040b + +#define REG_A5XX_RBBM_PERFCTR_UCHE_2_LO 0x0000040c + +#define REG_A5XX_RBBM_PERFCTR_UCHE_2_HI 0x0000040d + +#define REG_A5XX_RBBM_PERFCTR_UCHE_3_LO 0x0000040e + +#define REG_A5XX_RBBM_PERFCTR_UCHE_3_HI 0x0000040f + +#define REG_A5XX_RBBM_PERFCTR_UCHE_4_LO 0x00000410 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_4_HI 0x00000411 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_5_LO 0x00000412 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_5_HI 0x00000413 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_6_LO 0x00000414 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_6_HI 0x00000415 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_7_LO 0x00000416 + +#define REG_A5XX_RBBM_PERFCTR_UCHE_7_HI 0x00000417 + +#define REG_A5XX_RBBM_PERFCTR_TP_0_LO 0x00000418 + +#define REG_A5XX_RBBM_PERFCTR_TP_0_HI 0x00000419 + +#define REG_A5XX_RBBM_PERFCTR_TP_1_LO 0x0000041a + +#define REG_A5XX_RBBM_PERFCTR_TP_1_HI 0x0000041b + +#define REG_A5XX_RBBM_PERFCTR_TP_2_LO 0x0000041c + +#define REG_A5XX_RBBM_PERFCTR_TP_2_HI 0x0000041d + +#define REG_A5XX_RBBM_PERFCTR_TP_3_LO 0x0000041e + +#define REG_A5XX_RBBM_PERFCTR_TP_3_HI 0x0000041f + +#define REG_A5XX_RBBM_PERFCTR_TP_4_LO 0x00000420 + +#define REG_A5XX_RBBM_PERFCTR_TP_4_HI 0x00000421 + +#define REG_A5XX_RBBM_PERFCTR_TP_5_LO 0x00000422 + +#define REG_A5XX_RBBM_PERFCTR_TP_5_HI 0x00000423 + +#define REG_A5XX_RBBM_PERFCTR_TP_6_LO 0x00000424 + +#define REG_A5XX_RBBM_PERFCTR_TP_6_HI 0x00000425 + +#define REG_A5XX_RBBM_PERFCTR_TP_7_LO 0x00000426 + +#define REG_A5XX_RBBM_PERFCTR_TP_7_HI 0x00000427 + +#define REG_A5XX_RBBM_PERFCTR_SP_0_LO 0x00000428 + +#define REG_A5XX_RBBM_PERFCTR_SP_0_HI 0x00000429 + +#define REG_A5XX_RBBM_PERFCTR_SP_1_LO 0x0000042a + +#define REG_A5XX_RBBM_PERFCTR_SP_1_HI 0x0000042b + +#define REG_A5XX_RBBM_PERFCTR_SP_2_LO 0x0000042c + +#define REG_A5XX_RBBM_PERFCTR_SP_2_HI 0x0000042d + +#define REG_A5XX_RBBM_PERFCTR_SP_3_LO 0x0000042e + +#define REG_A5XX_RBBM_PERFCTR_SP_3_HI 0x0000042f + +#define REG_A5XX_RBBM_PERFCTR_SP_4_LO 0x00000430 + +#define REG_A5XX_RBBM_PERFCTR_SP_4_HI 0x00000431 + +#define REG_A5XX_RBBM_PERFCTR_SP_5_LO 0x00000432 + +#define REG_A5XX_RBBM_PERFCTR_SP_5_HI 0x00000433 + +#define REG_A5XX_RBBM_PERFCTR_SP_6_LO 0x00000434 + +#define REG_A5XX_RBBM_PERFCTR_SP_6_HI 0x00000435 + +#define REG_A5XX_RBBM_PERFCTR_SP_7_LO 0x00000436 + +#define REG_A5XX_RBBM_PERFCTR_SP_7_HI 0x00000437 + +#define REG_A5XX_RBBM_PERFCTR_SP_8_LO 0x00000438 + +#define REG_A5XX_RBBM_PERFCTR_SP_8_HI 0x00000439 + +#define REG_A5XX_RBBM_PERFCTR_SP_9_LO 0x0000043a + +#define REG_A5XX_RBBM_PERFCTR_SP_9_HI 0x0000043b + +#define REG_A5XX_RBBM_PERFCTR_SP_10_LO 0x0000043c + +#define REG_A5XX_RBBM_PERFCTR_SP_10_HI 0x0000043d + +#define REG_A5XX_RBBM_PERFCTR_SP_11_LO 0x0000043e + +#define REG_A5XX_RBBM_PERFCTR_SP_11_HI 0x0000043f + +#define REG_A5XX_RBBM_PERFCTR_RB_0_LO 0x00000440 + +#define REG_A5XX_RBBM_PERFCTR_RB_0_HI 0x00000441 + +#define REG_A5XX_RBBM_PERFCTR_RB_1_LO 0x00000442 + +#define REG_A5XX_RBBM_PERFCTR_RB_1_HI 0x00000443 + +#define REG_A5XX_RBBM_PERFCTR_RB_2_LO 0x00000444 + +#define REG_A5XX_RBBM_PERFCTR_RB_2_HI 0x00000445 + +#define REG_A5XX_RBBM_PERFCTR_RB_3_LO 0x00000446 + +#define REG_A5XX_RBBM_PERFCTR_RB_3_HI 0x00000447 + +#define REG_A5XX_RBBM_PERFCTR_RB_4_LO 0x00000448 + +#define REG_A5XX_RBBM_PERFCTR_RB_4_HI 0x00000449 + +#define REG_A5XX_RBBM_PERFCTR_RB_5_LO 0x0000044a + +#define REG_A5XX_RBBM_PERFCTR_RB_5_HI 0x0000044b + +#define REG_A5XX_RBBM_PERFCTR_RB_6_LO 0x0000044c + +#define REG_A5XX_RBBM_PERFCTR_RB_6_HI 0x0000044d + +#define REG_A5XX_RBBM_PERFCTR_RB_7_LO 0x0000044e + +#define REG_A5XX_RBBM_PERFCTR_RB_7_HI 0x0000044f + +#define REG_A5XX_RBBM_PERFCTR_VSC_0_LO 0x00000450 + +#define REG_A5XX_RBBM_PERFCTR_VSC_0_HI 0x00000451 + +#define REG_A5XX_RBBM_PERFCTR_VSC_1_LO 0x00000452 + +#define REG_A5XX_RBBM_PERFCTR_VSC_1_HI 0x00000453 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_0_LO 0x00000454 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_0_HI 0x00000455 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_1_LO 0x00000456 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_1_HI 0x00000457 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_2_LO 0x00000458 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_2_HI 0x00000459 + +#define REG_A5XX_RBBM_PERFCTR_LRZ_3_LO 0x0000045a + +#define REG_A5XX_RBBM_PERFCTR_LRZ_3_HI 0x0000045b + +#define REG_A5XX_RBBM_PERFCTR_CMP_0_LO 0x0000045c + +#define REG_A5XX_RBBM_PERFCTR_CMP_0_HI 0x0000045d + +#define REG_A5XX_RBBM_PERFCTR_CMP_1_LO 0x0000045e + +#define REG_A5XX_RBBM_PERFCTR_CMP_1_HI 0x0000045f + +#define REG_A5XX_RBBM_PERFCTR_CMP_2_LO 0x00000460 + +#define REG_A5XX_RBBM_PERFCTR_CMP_2_HI 0x00000461 + +#define REG_A5XX_RBBM_PERFCTR_CMP_3_LO 0x00000462 + +#define REG_A5XX_RBBM_PERFCTR_CMP_3_HI 0x00000463 + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_0 0x0000046b + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_1 0x0000046c + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_2 0x0000046d + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_3 0x0000046e + +#define REG_A5XX_RBBM_ALWAYSON_COUNTER_LO 0x000004d2 + +#define REG_A5XX_RBBM_ALWAYSON_COUNTER_HI 0x000004d3 + +#define REG_A5XX_RBBM_STATUS 0x000004f5 +#define A5XX_RBBM_STATUS_GPU_BUSY_IGN_AHB 0x80000000 +#define A5XX_RBBM_STATUS_GPU_BUSY_IGN_AHB_CP 0x40000000 +#define A5XX_RBBM_STATUS_HLSQ_BUSY 0x20000000 +#define A5XX_RBBM_STATUS_VSC_BUSY 0x10000000 +#define A5XX_RBBM_STATUS_TPL1_BUSY 0x08000000 +#define A5XX_RBBM_STATUS_SP_BUSY 0x04000000 +#define A5XX_RBBM_STATUS_UCHE_BUSY 0x02000000 +#define A5XX_RBBM_STATUS_VPC_BUSY 0x01000000 +#define A5XX_RBBM_STATUS_VFDP_BUSY 0x00800000 +#define A5XX_RBBM_STATUS_VFD_BUSY 0x00400000 +#define A5XX_RBBM_STATUS_TESS_BUSY 0x00200000 +#define A5XX_RBBM_STATUS_PC_VSD_BUSY 0x00100000 +#define A5XX_RBBM_STATUS_PC_DCALL_BUSY 0x00080000 +#define A5XX_RBBM_STATUS_GPMU_SLAVE_BUSY 0x00040000 +#define A5XX_RBBM_STATUS_DCOM_BUSY 0x00020000 +#define A5XX_RBBM_STATUS_COM_BUSY 0x00010000 +#define A5XX_RBBM_STATUS_LRZ_BUZY 0x00008000 +#define A5XX_RBBM_STATUS_A2D_DSP_BUSY 0x00004000 +#define A5XX_RBBM_STATUS_CCUFCHE_BUSY 0x00002000 +#define A5XX_RBBM_STATUS_RB_BUSY 0x00001000 +#define A5XX_RBBM_STATUS_RAS_BUSY 0x00000800 +#define A5XX_RBBM_STATUS_TSE_BUSY 0x00000400 +#define A5XX_RBBM_STATUS_VBIF_BUSY 0x00000200 +#define A5XX_RBBM_STATUS_GPU_BUSY_IGN_AHB_HYST 0x00000100 +#define A5XX_RBBM_STATUS_CP_BUSY_IGN_HYST 0x00000080 +#define A5XX_RBBM_STATUS_CP_BUSY 0x00000040 +#define A5XX_RBBM_STATUS_GPMU_MASTER_BUSY 0x00000020 +#define A5XX_RBBM_STATUS_CP_CRASH_BUSY 0x00000010 +#define A5XX_RBBM_STATUS_CP_ETS_BUSY 0x00000008 +#define A5XX_RBBM_STATUS_CP_PFP_BUSY 0x00000004 +#define A5XX_RBBM_STATUS_CP_ME_BUSY 0x00000002 +#define A5XX_RBBM_STATUS_HI_BUSY 0x00000001 + +#define REG_A5XX_RBBM_STATUS3 0x00000530 + +#define REG_A5XX_RBBM_INT_0_STATUS 0x000004e1 + +#define REG_A5XX_RBBM_AHB_ME_SPLIT_STATUS 0x000004f0 + +#define REG_A5XX_RBBM_AHB_PFP_SPLIT_STATUS 0x000004f1 + +#define REG_A5XX_RBBM_AHB_ETS_SPLIT_STATUS 0x000004f3 + +#define REG_A5XX_RBBM_AHB_ERROR_STATUS 0x000004f4 + +#define REG_A5XX_RBBM_PERFCTR_CNTL 0x00000464 + +#define REG_A5XX_RBBM_PERFCTR_LOAD_CMD0 0x00000465 + +#define REG_A5XX_RBBM_PERFCTR_LOAD_CMD1 0x00000466 + +#define REG_A5XX_RBBM_PERFCTR_LOAD_CMD2 0x00000467 + +#define REG_A5XX_RBBM_PERFCTR_LOAD_CMD3 0x00000468 + +#define REG_A5XX_RBBM_PERFCTR_LOAD_VALUE_LO 0x00000469 + +#define REG_A5XX_RBBM_PERFCTR_LOAD_VALUE_HI 0x0000046a + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_0 0x0000046b + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_1 0x0000046c + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_2 0x0000046d + +#define REG_A5XX_RBBM_PERFCTR_RBBM_SEL_3 0x0000046e + +#define REG_A5XX_RBBM_PERFCTR_GPU_BUSY_MASKED 0x0000046f + +#define REG_A5XX_RBBM_CFG_DBGBUS_EVENT_LOGIC 0x00000504 + +#define REG_A5XX_RBBM_CFG_DBGBUS_OVER 0x00000505 + +#define REG_A5XX_RBBM_CFG_DBGBUS_COUNT0 0x00000506 + +#define REG_A5XX_RBBM_CFG_DBGBUS_COUNT1 0x00000507 + +#define REG_A5XX_RBBM_CFG_DBGBUS_COUNT2 0x00000508 + +#define REG_A5XX_RBBM_CFG_DBGBUS_COUNT3 0x00000509 + +#define REG_A5XX_RBBM_CFG_DBGBUS_COUNT4 0x0000050a + +#define REG_A5XX_RBBM_CFG_DBGBUS_COUNT5 0x0000050b + +#define REG_A5XX_RBBM_CFG_DBGBUS_TRACE_ADDR 0x0000050c + +#define REG_A5XX_RBBM_CFG_DBGBUS_TRACE_BUF0 0x0000050d + +#define REG_A5XX_RBBM_CFG_DBGBUS_TRACE_BUF1 0x0000050e + +#define REG_A5XX_RBBM_CFG_DBGBUS_TRACE_BUF2 0x0000050f + +#define REG_A5XX_RBBM_CFG_DBGBUS_TRACE_BUF3 0x00000510 + +#define REG_A5XX_RBBM_CFG_DBGBUS_TRACE_BUF4 0x00000511 + +#define REG_A5XX_RBBM_CFG_DBGBUS_MISR0 0x00000512 + +#define REG_A5XX_RBBM_CFG_DBGBUS_MISR1 0x00000513 + +#define REG_A5XX_RBBM_ISDB_CNT 0x00000533 + +#define REG_A5XX_RBBM_SECVID_TRUST_CONFIG 0x0000f000 + +#define REG_A5XX_RBBM_SECVID_TRUST_CNTL 0x0000f400 + +#define REG_A5XX_RBBM_SECVID_TSB_TRUSTED_BASE_LO 0x0000f800 + +#define REG_A5XX_RBBM_SECVID_TSB_TRUSTED_BASE_HI 0x0000f801 + +#define REG_A5XX_RBBM_SECVID_TSB_TRUSTED_SIZE 0x0000f802 + +#define REG_A5XX_RBBM_SECVID_TSB_CNTL 0x0000f803 + +#define REG_A5XX_RBBM_SECVID_TSB_COMP_STATUS_LO 0x0000f804 + +#define REG_A5XX_RBBM_SECVID_TSB_COMP_STATUS_HI 0x0000f805 + +#define REG_A5XX_RBBM_SECVID_TSB_UCHE_STATUS_LO 0x0000f806 + +#define REG_A5XX_RBBM_SECVID_TSB_UCHE_STATUS_HI 0x0000f807 + +#define REG_A5XX_RBBM_SECVID_TSB_ADDR_MODE_CNTL 0x0000f810 + +#define REG_A5XX_VSC_PIPE_DATA_LENGTH_0 0x00000c00 + +#define REG_A5XX_VSC_PERFCTR_VSC_SEL_0 0x00000c60 + +#define REG_A5XX_VSC_PERFCTR_VSC_SEL_1 0x00000c61 + +#define REG_A5XX_VSC_BIN_SIZE 0x00000cdd +#define A5XX_VSC_BIN_SIZE_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_VSC_BIN_SIZE_X__MASK 0x00007fff +#define A5XX_VSC_BIN_SIZE_X__SHIFT 0 +static inline uint32_t A5XX_VSC_BIN_SIZE_X(uint32_t val) +{ + return ((val) << A5XX_VSC_BIN_SIZE_X__SHIFT) & A5XX_VSC_BIN_SIZE_X__MASK; +} +#define A5XX_VSC_BIN_SIZE_Y__MASK 0x7fff0000 +#define A5XX_VSC_BIN_SIZE_Y__SHIFT 16 +static inline uint32_t A5XX_VSC_BIN_SIZE_Y(uint32_t val) +{ + return ((val) << A5XX_VSC_BIN_SIZE_Y__SHIFT) & A5XX_VSC_BIN_SIZE_Y__MASK; +} + +#define REG_A5XX_GRAS_ADDR_MODE_CNTL 0x00000c81 + +#define REG_A5XX_GRAS_PERFCTR_TSE_SEL_0 0x00000c90 + +#define REG_A5XX_GRAS_PERFCTR_TSE_SEL_1 0x00000c91 + +#define REG_A5XX_GRAS_PERFCTR_TSE_SEL_2 0x00000c92 + +#define REG_A5XX_GRAS_PERFCTR_TSE_SEL_3 0x00000c93 + +#define REG_A5XX_GRAS_PERFCTR_RAS_SEL_0 0x00000c94 + +#define REG_A5XX_GRAS_PERFCTR_RAS_SEL_1 0x00000c95 + +#define REG_A5XX_GRAS_PERFCTR_RAS_SEL_2 0x00000c96 + +#define REG_A5XX_GRAS_PERFCTR_RAS_SEL_3 0x00000c97 + +#define REG_A5XX_GRAS_PERFCTR_LRZ_SEL_0 0x00000c98 + +#define REG_A5XX_GRAS_PERFCTR_LRZ_SEL_1 0x00000c99 + +#define REG_A5XX_GRAS_PERFCTR_LRZ_SEL_2 0x00000c9a + +#define REG_A5XX_GRAS_PERFCTR_LRZ_SEL_3 0x00000c9b + +#define REG_A5XX_RB_DBG_ECO_CNTL 0x00000cc4 + +#define REG_A5XX_RB_ADDR_MODE_CNTL 0x00000cc5 + +#define REG_A5XX_RB_MODE_CNTL 0x00000cc6 + +#define REG_A5XX_RB_CCU_CNTL 0x00000cc7 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_0 0x00000cd0 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_1 0x00000cd1 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_2 0x00000cd2 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_3 0x00000cd3 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_4 0x00000cd4 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_5 0x00000cd5 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_6 0x00000cd6 + +#define REG_A5XX_RB_PERFCTR_RB_SEL_7 0x00000cd7 + +#define REG_A5XX_RB_PERFCTR_CCU_SEL_0 0x00000cd8 + +#define REG_A5XX_RB_PERFCTR_CCU_SEL_1 0x00000cd9 + +#define REG_A5XX_RB_PERFCTR_CCU_SEL_2 0x00000cda + +#define REG_A5XX_RB_PERFCTR_CCU_SEL_3 0x00000cdb + +#define REG_A5XX_RB_POWERCTR_RB_SEL_0 0x00000ce0 + +#define REG_A5XX_RB_POWERCTR_RB_SEL_1 0x00000ce1 + +#define REG_A5XX_RB_POWERCTR_RB_SEL_2 0x00000ce2 + +#define REG_A5XX_RB_POWERCTR_RB_SEL_3 0x00000ce3 + +#define REG_A5XX_RB_POWERCTR_CCU_SEL_0 0x00000ce4 + +#define REG_A5XX_RB_POWERCTR_CCU_SEL_1 0x00000ce5 + +#define REG_A5XX_RB_PERFCTR_CMP_SEL_0 0x00000cec + +#define REG_A5XX_RB_PERFCTR_CMP_SEL_1 0x00000ced + +#define REG_A5XX_RB_PERFCTR_CMP_SEL_2 0x00000cee + +#define REG_A5XX_RB_PERFCTR_CMP_SEL_3 0x00000cef + +#define REG_A5XX_PC_DBG_ECO_CNTL 0x00000d00 +#define A5XX_PC_DBG_ECO_CNTL_TWOPASSUSEWFI 0x00000100 + +#define REG_A5XX_PC_ADDR_MODE_CNTL 0x00000d01 + +#define REG_A5XX_PC_MODE_CNTL 0x00000d02 + +#define REG_A5XX_UNKNOWN_0D08 0x00000d08 + +#define REG_A5XX_UNKNOWN_0D09 0x00000d09 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_0 0x00000d10 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_1 0x00000d11 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_2 0x00000d12 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_3 0x00000d13 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_4 0x00000d14 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_5 0x00000d15 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_6 0x00000d16 + +#define REG_A5XX_PC_PERFCTR_PC_SEL_7 0x00000d17 + +#define REG_A5XX_HLSQ_TIMEOUT_THRESHOLD_0 0x00000e00 + +#define REG_A5XX_HLSQ_TIMEOUT_THRESHOLD_1 0x00000e01 + +#define REG_A5XX_HLSQ_ADDR_MODE_CNTL 0x00000e05 + +#define REG_A5XX_HLSQ_MODE_CNTL 0x00000e06 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_0 0x00000e10 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_1 0x00000e11 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_2 0x00000e12 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_3 0x00000e13 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_4 0x00000e14 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_5 0x00000e15 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_6 0x00000e16 + +#define REG_A5XX_HLSQ_PERFCTR_HLSQ_SEL_7 0x00000e17 + +#define REG_A5XX_HLSQ_SPTP_RDSEL 0x00000f08 + +#define REG_A5XX_HLSQ_DBG_READ_SEL 0x0000bc00 + +#define REG_A5XX_HLSQ_DBG_AHB_READ_APERTURE 0x0000a000 + +#define REG_A5XX_VFD_ADDR_MODE_CNTL 0x00000e41 + +#define REG_A5XX_VFD_MODE_CNTL 0x00000e42 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_0 0x00000e50 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_1 0x00000e51 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_2 0x00000e52 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_3 0x00000e53 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_4 0x00000e54 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_5 0x00000e55 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_6 0x00000e56 + +#define REG_A5XX_VFD_PERFCTR_VFD_SEL_7 0x00000e57 + +#define REG_A5XX_VPC_DBG_ECO_CNTL 0x00000e60 + +#define REG_A5XX_VPC_ADDR_MODE_CNTL 0x00000e61 + +#define REG_A5XX_VPC_MODE_CNTL 0x00000e62 + +#define REG_A5XX_VPC_PERFCTR_VPC_SEL_0 0x00000e64 + +#define REG_A5XX_VPC_PERFCTR_VPC_SEL_1 0x00000e65 + +#define REG_A5XX_VPC_PERFCTR_VPC_SEL_2 0x00000e66 + +#define REG_A5XX_VPC_PERFCTR_VPC_SEL_3 0x00000e67 + +#define REG_A5XX_UCHE_ADDR_MODE_CNTL 0x00000e80 + +#define REG_A5XX_UCHE_SVM_CNTL 0x00000e82 + +#define REG_A5XX_UCHE_WRITE_THRU_BASE_LO 0x00000e87 + +#define REG_A5XX_UCHE_WRITE_THRU_BASE_HI 0x00000e88 + +#define REG_A5XX_UCHE_TRAP_BASE_LO 0x00000e89 + +#define REG_A5XX_UCHE_TRAP_BASE_HI 0x00000e8a + +#define REG_A5XX_UCHE_GMEM_RANGE_MIN_LO 0x00000e8b + +#define REG_A5XX_UCHE_GMEM_RANGE_MIN_HI 0x00000e8c + +#define REG_A5XX_UCHE_GMEM_RANGE_MAX_LO 0x00000e8d + +#define REG_A5XX_UCHE_GMEM_RANGE_MAX_HI 0x00000e8e + +#define REG_A5XX_UCHE_DBG_ECO_CNTL_2 0x00000e8f + +#define REG_A5XX_UCHE_DBG_ECO_CNTL 0x00000e90 + +#define REG_A5XX_UCHE_CACHE_INVALIDATE_MIN_LO 0x00000e91 + +#define REG_A5XX_UCHE_CACHE_INVALIDATE_MIN_HI 0x00000e92 + +#define REG_A5XX_UCHE_CACHE_INVALIDATE_MAX_LO 0x00000e93 + +#define REG_A5XX_UCHE_CACHE_INVALIDATE_MAX_HI 0x00000e94 + +#define REG_A5XX_UCHE_CACHE_INVALIDATE 0x00000e95 + +#define REG_A5XX_UCHE_CACHE_WAYS 0x00000e96 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_0 0x00000ea0 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_1 0x00000ea1 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_2 0x00000ea2 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_3 0x00000ea3 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_4 0x00000ea4 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_5 0x00000ea5 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_6 0x00000ea6 + +#define REG_A5XX_UCHE_PERFCTR_UCHE_SEL_7 0x00000ea7 + +#define REG_A5XX_UCHE_POWERCTR_UCHE_SEL_0 0x00000ea8 + +#define REG_A5XX_UCHE_POWERCTR_UCHE_SEL_1 0x00000ea9 + +#define REG_A5XX_UCHE_POWERCTR_UCHE_SEL_2 0x00000eaa + +#define REG_A5XX_UCHE_POWERCTR_UCHE_SEL_3 0x00000eab + +#define REG_A5XX_UCHE_TRAP_LOG_LO 0x00000eb1 + +#define REG_A5XX_UCHE_TRAP_LOG_HI 0x00000eb2 + +#define REG_A5XX_SP_DBG_ECO_CNTL 0x00000ec0 + +#define REG_A5XX_SP_ADDR_MODE_CNTL 0x00000ec1 + +#define REG_A5XX_SP_MODE_CNTL 0x00000ec2 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_0 0x00000ed0 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_1 0x00000ed1 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_2 0x00000ed2 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_3 0x00000ed3 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_4 0x00000ed4 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_5 0x00000ed5 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_6 0x00000ed6 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_7 0x00000ed7 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_8 0x00000ed8 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_9 0x00000ed9 + +#define REG_A5XX_SP_PERFCTR_SP_SEL_10 0x00000eda + +#define REG_A5XX_SP_PERFCTR_SP_SEL_11 0x00000edb + +#define REG_A5XX_SP_POWERCTR_SP_SEL_0 0x00000edc + +#define REG_A5XX_SP_POWERCTR_SP_SEL_1 0x00000edd + +#define REG_A5XX_SP_POWERCTR_SP_SEL_2 0x00000ede + +#define REG_A5XX_SP_POWERCTR_SP_SEL_3 0x00000edf + +#define REG_A5XX_TPL1_ADDR_MODE_CNTL 0x00000f01 + +#define REG_A5XX_TPL1_MODE_CNTL 0x00000f02 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_0 0x00000f10 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_1 0x00000f11 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_2 0x00000f12 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_3 0x00000f13 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_4 0x00000f14 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_5 0x00000f15 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_6 0x00000f16 + +#define REG_A5XX_TPL1_PERFCTR_TP_SEL_7 0x00000f17 + +#define REG_A5XX_TPL1_POWERCTR_TP_SEL_0 0x00000f18 + +#define REG_A5XX_TPL1_POWERCTR_TP_SEL_1 0x00000f19 + +#define REG_A5XX_TPL1_POWERCTR_TP_SEL_2 0x00000f1a + +#define REG_A5XX_TPL1_POWERCTR_TP_SEL_3 0x00000f1b + +#define REG_A5XX_VBIF_VERSION 0x00003000 + +#define REG_A5XX_VBIF_CLKON 0x00003001 + +#define REG_A5XX_VBIF_ABIT_SORT 0x00003028 + +#define REG_A5XX_VBIF_ABIT_SORT_CONF 0x00003029 + +#define REG_A5XX_VBIF_ROUND_ROBIN_QOS_ARB 0x00003049 + +#define REG_A5XX_VBIF_GATE_OFF_WRREQ_EN 0x0000302a + +#define REG_A5XX_VBIF_IN_RD_LIM_CONF0 0x0000302c + +#define REG_A5XX_VBIF_IN_RD_LIM_CONF1 0x0000302d + +#define REG_A5XX_VBIF_XIN_HALT_CTRL0 0x00003080 + +#define REG_A5XX_VBIF_XIN_HALT_CTRL1 0x00003081 + +#define REG_A5XX_VBIF_TEST_BUS_OUT_CTRL 0x00003084 + +#define REG_A5XX_VBIF_TEST_BUS1_CTRL0 0x00003085 + +#define REG_A5XX_VBIF_TEST_BUS1_CTRL1 0x00003086 + +#define REG_A5XX_VBIF_TEST_BUS2_CTRL0 0x00003087 + +#define REG_A5XX_VBIF_TEST_BUS2_CTRL1 0x00003088 + +#define REG_A5XX_VBIF_TEST_BUS_OUT 0x0000308c + +#define REG_A5XX_VBIF_PERF_CNT_SEL0 0x000030d0 + +#define REG_A5XX_VBIF_PERF_CNT_SEL1 0x000030d1 + +#define REG_A5XX_VBIF_PERF_CNT_SEL2 0x000030d2 + +#define REG_A5XX_VBIF_PERF_CNT_SEL3 0x000030d3 + +#define REG_A5XX_VBIF_PERF_CNT_LOW0 0x000030d8 + +#define REG_A5XX_VBIF_PERF_CNT_LOW1 0x000030d9 + +#define REG_A5XX_VBIF_PERF_CNT_LOW2 0x000030da + +#define REG_A5XX_VBIF_PERF_CNT_LOW3 0x000030db + +#define REG_A5XX_VBIF_PERF_CNT_HIGH0 0x000030e0 + +#define REG_A5XX_VBIF_PERF_CNT_HIGH1 0x000030e1 + +#define REG_A5XX_VBIF_PERF_CNT_HIGH2 0x000030e2 + +#define REG_A5XX_VBIF_PERF_CNT_HIGH3 0x000030e3 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_EN0 0x00003100 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_EN1 0x00003101 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_EN2 0x00003102 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_LOW0 0x00003110 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_LOW1 0x00003111 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_LOW2 0x00003112 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_HIGH0 0x00003118 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_HIGH1 0x00003119 + +#define REG_A5XX_VBIF_PERF_PWR_CNT_HIGH2 0x0000311a + +#define REG_A5XX_GPMU_INST_RAM_BASE 0x00008800 + +#define REG_A5XX_GPMU_DATA_RAM_BASE 0x00009800 + +#define REG_A5XX_GPMU_SP_POWER_CNTL 0x0000a881 + +#define REG_A5XX_GPMU_RBCCU_CLOCK_CNTL 0x0000a886 + +#define REG_A5XX_GPMU_RBCCU_POWER_CNTL 0x0000a887 + +#define REG_A5XX_GPMU_SP_PWR_CLK_STATUS 0x0000a88b +#define A5XX_GPMU_SP_PWR_CLK_STATUS_PWR_ON 0x00100000 + +#define REG_A5XX_GPMU_RBCCU_PWR_CLK_STATUS 0x0000a88d +#define A5XX_GPMU_RBCCU_PWR_CLK_STATUS_PWR_ON 0x00100000 + +#define REG_A5XX_GPMU_PWR_COL_STAGGER_DELAY 0x0000a891 + +#define REG_A5XX_GPMU_PWR_COL_INTER_FRAME_CTRL 0x0000a892 + +#define REG_A5XX_GPMU_PWR_COL_INTER_FRAME_HYST 0x0000a893 + +#define REG_A5XX_GPMU_PWR_COL_BINNING_CTRL 0x0000a894 + +#define REG_A5XX_GPMU_CLOCK_THROTTLE_CTRL 0x0000a8a3 + +#define REG_A5XX_GPMU_WFI_CONFIG 0x0000a8c1 + +#define REG_A5XX_GPMU_RBBM_INTR_INFO 0x0000a8d6 + +#define REG_A5XX_GPMU_CM3_SYSRESET 0x0000a8d8 + +#define REG_A5XX_GPMU_GENERAL_0 0x0000a8e0 + +#define REG_A5XX_GPMU_GENERAL_1 0x0000a8e1 + +#define REG_A5XX_SP_POWER_COUNTER_0_LO 0x0000a840 + +#define REG_A5XX_SP_POWER_COUNTER_0_HI 0x0000a841 + +#define REG_A5XX_SP_POWER_COUNTER_1_LO 0x0000a842 + +#define REG_A5XX_SP_POWER_COUNTER_1_HI 0x0000a843 + +#define REG_A5XX_SP_POWER_COUNTER_2_LO 0x0000a844 + +#define REG_A5XX_SP_POWER_COUNTER_2_HI 0x0000a845 + +#define REG_A5XX_SP_POWER_COUNTER_3_LO 0x0000a846 + +#define REG_A5XX_SP_POWER_COUNTER_3_HI 0x0000a847 + +#define REG_A5XX_TP_POWER_COUNTER_0_LO 0x0000a848 + +#define REG_A5XX_TP_POWER_COUNTER_0_HI 0x0000a849 + +#define REG_A5XX_TP_POWER_COUNTER_1_LO 0x0000a84a + +#define REG_A5XX_TP_POWER_COUNTER_1_HI 0x0000a84b + +#define REG_A5XX_TP_POWER_COUNTER_2_LO 0x0000a84c + +#define REG_A5XX_TP_POWER_COUNTER_2_HI 0x0000a84d + +#define REG_A5XX_TP_POWER_COUNTER_3_LO 0x0000a84e + +#define REG_A5XX_TP_POWER_COUNTER_3_HI 0x0000a84f + +#define REG_A5XX_RB_POWER_COUNTER_0_LO 0x0000a850 + +#define REG_A5XX_RB_POWER_COUNTER_0_HI 0x0000a851 + +#define REG_A5XX_RB_POWER_COUNTER_1_LO 0x0000a852 + +#define REG_A5XX_RB_POWER_COUNTER_1_HI 0x0000a853 + +#define REG_A5XX_RB_POWER_COUNTER_2_LO 0x0000a854 + +#define REG_A5XX_RB_POWER_COUNTER_2_HI 0x0000a855 + +#define REG_A5XX_RB_POWER_COUNTER_3_LO 0x0000a856 + +#define REG_A5XX_RB_POWER_COUNTER_3_HI 0x0000a857 + +#define REG_A5XX_CCU_POWER_COUNTER_0_LO 0x0000a858 + +#define REG_A5XX_CCU_POWER_COUNTER_0_HI 0x0000a859 + +#define REG_A5XX_CCU_POWER_COUNTER_1_LO 0x0000a85a + +#define REG_A5XX_CCU_POWER_COUNTER_1_HI 0x0000a85b + +#define REG_A5XX_UCHE_POWER_COUNTER_0_LO 0x0000a85c + +#define REG_A5XX_UCHE_POWER_COUNTER_0_HI 0x0000a85d + +#define REG_A5XX_UCHE_POWER_COUNTER_1_LO 0x0000a85e + +#define REG_A5XX_UCHE_POWER_COUNTER_1_HI 0x0000a85f + +#define REG_A5XX_UCHE_POWER_COUNTER_2_LO 0x0000a860 + +#define REG_A5XX_UCHE_POWER_COUNTER_2_HI 0x0000a861 + +#define REG_A5XX_UCHE_POWER_COUNTER_3_LO 0x0000a862 + +#define REG_A5XX_UCHE_POWER_COUNTER_3_HI 0x0000a863 + +#define REG_A5XX_CP_POWER_COUNTER_0_LO 0x0000a864 + +#define REG_A5XX_CP_POWER_COUNTER_0_HI 0x0000a865 + +#define REG_A5XX_CP_POWER_COUNTER_1_LO 0x0000a866 + +#define REG_A5XX_CP_POWER_COUNTER_1_HI 0x0000a867 + +#define REG_A5XX_CP_POWER_COUNTER_2_LO 0x0000a868 + +#define REG_A5XX_CP_POWER_COUNTER_2_HI 0x0000a869 + +#define REG_A5XX_CP_POWER_COUNTER_3_LO 0x0000a86a + +#define REG_A5XX_CP_POWER_COUNTER_3_HI 0x0000a86b + +#define REG_A5XX_GPMU_POWER_COUNTER_0_LO 0x0000a86c + +#define REG_A5XX_GPMU_POWER_COUNTER_0_HI 0x0000a86d + +#define REG_A5XX_GPMU_POWER_COUNTER_1_LO 0x0000a86e + +#define REG_A5XX_GPMU_POWER_COUNTER_1_HI 0x0000a86f + +#define REG_A5XX_GPMU_POWER_COUNTER_2_LO 0x0000a870 + +#define REG_A5XX_GPMU_POWER_COUNTER_2_HI 0x0000a871 + +#define REG_A5XX_GPMU_POWER_COUNTER_3_LO 0x0000a872 + +#define REG_A5XX_GPMU_POWER_COUNTER_3_HI 0x0000a873 + +#define REG_A5XX_GPMU_POWER_COUNTER_4_LO 0x0000a874 + +#define REG_A5XX_GPMU_POWER_COUNTER_4_HI 0x0000a875 + +#define REG_A5XX_GPMU_POWER_COUNTER_5_LO 0x0000a876 + +#define REG_A5XX_GPMU_POWER_COUNTER_5_HI 0x0000a877 + +#define REG_A5XX_GPMU_POWER_COUNTER_ENABLE 0x0000a878 + +#define REG_A5XX_GPMU_ALWAYS_ON_COUNTER_LO 0x0000a879 + +#define REG_A5XX_GPMU_ALWAYS_ON_COUNTER_HI 0x0000a87a + +#define REG_A5XX_GPMU_ALWAYS_ON_COUNTER_RESET 0x0000a87b + +#define REG_A5XX_GPMU_POWER_COUNTER_SELECT_0 0x0000a87c + +#define REG_A5XX_GPMU_POWER_COUNTER_SELECT_1 0x0000a87d + +#define REG_A5XX_GPMU_CLOCK_THROTTLE_CTRL 0x0000a8a3 + +#define REG_A5XX_GPMU_THROTTLE_UNMASK_FORCE_CTRL 0x0000a8a8 + +#define REG_A5XX_GPMU_TEMP_SENSOR_ID 0x0000ac00 + +#define REG_A5XX_GPMU_TEMP_SENSOR_CONFIG 0x0000ac01 + +#define REG_A5XX_GPMU_TEMP_VAL 0x0000ac02 + +#define REG_A5XX_GPMU_DELTA_TEMP_THRESHOLD 0x0000ac03 + +#define REG_A5XX_GPMU_TEMP_THRESHOLD_INTR_STATUS 0x0000ac05 + +#define REG_A5XX_GPMU_TEMP_THRESHOLD_INTR_EN_MASK 0x0000ac06 + +#define REG_A5XX_GPMU_LEAKAGE_TEMP_COEFF_0_1 0x0000ac40 + +#define REG_A5XX_GPMU_LEAKAGE_TEMP_COEFF_2_3 0x0000ac41 + +#define REG_A5XX_GPMU_LEAKAGE_VTG_COEFF_0_1 0x0000ac42 + +#define REG_A5XX_GPMU_LEAKAGE_VTG_COEFF_2_3 0x0000ac43 + +#define REG_A5XX_GPMU_BASE_LEAKAGE 0x0000ac46 + +#define REG_A5XX_GPMU_GPMU_VOLTAGE 0x0000ac60 + +#define REG_A5XX_GPMU_GPMU_VOLTAGE_INTR_STATUS 0x0000ac61 + +#define REG_A5XX_GPMU_GPMU_VOLTAGE_INTR_EN_MASK 0x0000ac62 + +#define REG_A5XX_GPMU_GPMU_PWR_THRESHOLD 0x0000ac80 + +#define REG_A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL 0x0000acc4 + +#define REG_A5XX_GPMU_GPMU_LLM_GLM_SLEEP_STATUS 0x0000acc5 + +#define REG_A5XX_GDPM_CONFIG1 0x0000b80c + +#define REG_A5XX_GDPM_CONFIG2 0x0000b80d + +#define REG_A5XX_GDPM_INT_EN 0x0000b80f + +#define REG_A5XX_GDPM_INT_MASK 0x0000b811 + +#define REG_A5XX_GPMU_BEC_ENABLE 0x0000b9a0 + +#define REG_A5XX_GPU_CS_SENSOR_GENERAL_STATUS 0x0000c41a + +#define REG_A5XX_GPU_CS_AMP_CALIBRATION_STATUS1_0 0x0000c41d + +#define REG_A5XX_GPU_CS_AMP_CALIBRATION_STATUS1_2 0x0000c41f + +#define REG_A5XX_GPU_CS_AMP_CALIBRATION_STATUS1_4 0x0000c421 + +#define REG_A5XX_GPU_CS_ENABLE_REG 0x0000c520 + +#define REG_A5XX_GPU_CS_AMP_CALIBRATION_CONTROL1 0x0000c557 + +#define REG_A5XX_GRAS_CL_CNTL 0x0000e000 + +#define REG_A5XX_UNKNOWN_E001 0x0000e001 + +#define REG_A5XX_UNKNOWN_E004 0x0000e004 + +#define REG_A5XX_GRAS_CLEAR_CNTL 0x0000e005 +#define A5XX_GRAS_CLEAR_CNTL_NOT_FASTCLEAR 0x00000001 + +#define REG_A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ 0x0000e006 +#define A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_HORZ__MASK 0x000003ff +#define A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_HORZ__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_HORZ(uint32_t val) +{ + return ((val) << A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_HORZ__SHIFT) & A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_HORZ__MASK; +} +#define A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_VERT__MASK 0x000ffc00 +#define A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_VERT__SHIFT 10 +static inline uint32_t A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_VERT(uint32_t val) +{ + return ((val) << A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_VERT__SHIFT) & A5XX_GRAS_CL_GUARDBAND_CLIP_ADJ_VERT__MASK; +} + +#define REG_A5XX_GRAS_CL_VPORT_XOFFSET_0 0x0000e010 +#define A5XX_GRAS_CL_VPORT_XOFFSET_0__MASK 0xffffffff +#define A5XX_GRAS_CL_VPORT_XOFFSET_0__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_VPORT_XOFFSET_0(float val) +{ + return ((fui(val)) << A5XX_GRAS_CL_VPORT_XOFFSET_0__SHIFT) & A5XX_GRAS_CL_VPORT_XOFFSET_0__MASK; +} + +#define REG_A5XX_GRAS_CL_VPORT_XSCALE_0 0x0000e011 +#define A5XX_GRAS_CL_VPORT_XSCALE_0__MASK 0xffffffff +#define A5XX_GRAS_CL_VPORT_XSCALE_0__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_VPORT_XSCALE_0(float val) +{ + return ((fui(val)) << A5XX_GRAS_CL_VPORT_XSCALE_0__SHIFT) & A5XX_GRAS_CL_VPORT_XSCALE_0__MASK; +} + +#define REG_A5XX_GRAS_CL_VPORT_YOFFSET_0 0x0000e012 +#define A5XX_GRAS_CL_VPORT_YOFFSET_0__MASK 0xffffffff +#define A5XX_GRAS_CL_VPORT_YOFFSET_0__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_VPORT_YOFFSET_0(float val) +{ + return ((fui(val)) << A5XX_GRAS_CL_VPORT_YOFFSET_0__SHIFT) & A5XX_GRAS_CL_VPORT_YOFFSET_0__MASK; +} + +#define REG_A5XX_GRAS_CL_VPORT_YSCALE_0 0x0000e013 +#define A5XX_GRAS_CL_VPORT_YSCALE_0__MASK 0xffffffff +#define A5XX_GRAS_CL_VPORT_YSCALE_0__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_VPORT_YSCALE_0(float val) +{ + return ((fui(val)) << A5XX_GRAS_CL_VPORT_YSCALE_0__SHIFT) & A5XX_GRAS_CL_VPORT_YSCALE_0__MASK; +} + +#define REG_A5XX_GRAS_CL_VPORT_ZOFFSET_0 0x0000e014 +#define A5XX_GRAS_CL_VPORT_ZOFFSET_0__MASK 0xffffffff +#define A5XX_GRAS_CL_VPORT_ZOFFSET_0__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_VPORT_ZOFFSET_0(float val) +{ + return ((fui(val)) << A5XX_GRAS_CL_VPORT_ZOFFSET_0__SHIFT) & A5XX_GRAS_CL_VPORT_ZOFFSET_0__MASK; +} + +#define REG_A5XX_GRAS_CL_VPORT_ZSCALE_0 0x0000e015 +#define A5XX_GRAS_CL_VPORT_ZSCALE_0__MASK 0xffffffff +#define A5XX_GRAS_CL_VPORT_ZSCALE_0__SHIFT 0 +static inline uint32_t A5XX_GRAS_CL_VPORT_ZSCALE_0(float val) +{ + return ((fui(val)) << A5XX_GRAS_CL_VPORT_ZSCALE_0__SHIFT) & A5XX_GRAS_CL_VPORT_ZSCALE_0__MASK; +} + +#define REG_A5XX_GRAS_SU_CNTL 0x0000e090 +#define A5XX_GRAS_SU_CNTL_FRONT_CW 0x00000004 +#define A5XX_GRAS_SU_CNTL_POLY_OFFSET 0x00000800 + +#define REG_A5XX_GRAS_SU_POINT_MINMAX 0x0000e091 +#define A5XX_GRAS_SU_POINT_MINMAX_MIN__MASK 0x0000ffff +#define A5XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT 0 +static inline uint32_t A5XX_GRAS_SU_POINT_MINMAX_MIN(float val) +{ + return ((((uint32_t)(val * 16.0))) << A5XX_GRAS_SU_POINT_MINMAX_MIN__SHIFT) & A5XX_GRAS_SU_POINT_MINMAX_MIN__MASK; +} +#define A5XX_GRAS_SU_POINT_MINMAX_MAX__MASK 0xffff0000 +#define A5XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT 16 +static inline uint32_t A5XX_GRAS_SU_POINT_MINMAX_MAX(float val) +{ + return ((((uint32_t)(val * 16.0))) << A5XX_GRAS_SU_POINT_MINMAX_MAX__SHIFT) & A5XX_GRAS_SU_POINT_MINMAX_MAX__MASK; +} + +#define REG_A5XX_GRAS_SU_POINT_SIZE 0x0000e092 +#define A5XX_GRAS_SU_POINT_SIZE__MASK 0xffffffff +#define A5XX_GRAS_SU_POINT_SIZE__SHIFT 0 +static inline uint32_t A5XX_GRAS_SU_POINT_SIZE(float val) +{ + return ((((int32_t)(val * 16.0))) << A5XX_GRAS_SU_POINT_SIZE__SHIFT) & A5XX_GRAS_SU_POINT_SIZE__MASK; +} + +#define REG_A5XX_UNKNOWN_E093 0x0000e093 + +#define REG_A5XX_GRAS_SU_DEPTH_PLANE_CNTL 0x0000e094 +#define A5XX_GRAS_SU_DEPTH_PLANE_CNTL_ALPHA_TEST_ENABLE 0x00000001 + +#define REG_A5XX_GRAS_SU_POLY_OFFSET_SCALE 0x0000e095 +#define A5XX_GRAS_SU_POLY_OFFSET_SCALE__MASK 0xffffffff +#define A5XX_GRAS_SU_POLY_OFFSET_SCALE__SHIFT 0 +static inline uint32_t A5XX_GRAS_SU_POLY_OFFSET_SCALE(float val) +{ + return ((fui(val)) << A5XX_GRAS_SU_POLY_OFFSET_SCALE__SHIFT) & A5XX_GRAS_SU_POLY_OFFSET_SCALE__MASK; +} + +#define REG_A5XX_GRAS_SU_POLY_OFFSET_OFFSET 0x0000e096 +#define A5XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK 0xffffffff +#define A5XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT 0 +static inline uint32_t A5XX_GRAS_SU_POLY_OFFSET_OFFSET(float val) +{ + return ((fui(val)) << A5XX_GRAS_SU_POLY_OFFSET_OFFSET__SHIFT) & A5XX_GRAS_SU_POLY_OFFSET_OFFSET__MASK; +} + +#define REG_A5XX_GRAS_SU_POLY_OFFSET_OFFSET_CLAMP 0x0000e097 +#define A5XX_GRAS_SU_POLY_OFFSET_OFFSET_CLAMP__MASK 0xffffffff +#define A5XX_GRAS_SU_POLY_OFFSET_OFFSET_CLAMP__SHIFT 0 +static inline uint32_t A5XX_GRAS_SU_POLY_OFFSET_OFFSET_CLAMP(float val) +{ + return ((fui(val)) << A5XX_GRAS_SU_POLY_OFFSET_OFFSET_CLAMP__SHIFT) & A5XX_GRAS_SU_POLY_OFFSET_OFFSET_CLAMP__MASK; +} + +#define REG_A5XX_GRAS_SU_DEPTH_BUFFER_INFO 0x0000e098 +#define A5XX_GRAS_SU_DEPTH_BUFFER_INFO_DEPTH_FORMAT__MASK 0x00000007 +#define A5XX_GRAS_SU_DEPTH_BUFFER_INFO_DEPTH_FORMAT__SHIFT 0 +static inline uint32_t A5XX_GRAS_SU_DEPTH_BUFFER_INFO_DEPTH_FORMAT(enum a5xx_depth_format val) +{ + return ((val) << A5XX_GRAS_SU_DEPTH_BUFFER_INFO_DEPTH_FORMAT__SHIFT) & A5XX_GRAS_SU_DEPTH_BUFFER_INFO_DEPTH_FORMAT__MASK; +} + +#define REG_A5XX_GRAS_SU_CONSERVATIVE_RAS_CNTL 0x0000e099 + +#define REG_A5XX_GRAS_SC_CNTL 0x0000e0a0 + +#define REG_A5XX_GRAS_SC_BIN_CNTL 0x0000e0a1 + +#define REG_A5XX_GRAS_SC_RAS_MSAA_CNTL 0x0000e0a2 +#define A5XX_GRAS_SC_RAS_MSAA_CNTL_SAMPLES__MASK 0x00000003 +#define A5XX_GRAS_SC_RAS_MSAA_CNTL_SAMPLES__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_RAS_MSAA_CNTL_SAMPLES(enum a3xx_msaa_samples val) +{ + return ((val) << A5XX_GRAS_SC_RAS_MSAA_CNTL_SAMPLES__SHIFT) & A5XX_GRAS_SC_RAS_MSAA_CNTL_SAMPLES__MASK; +} + +#define REG_A5XX_GRAS_SC_DEST_MSAA_CNTL 0x0000e0a3 +#define A5XX_GRAS_SC_DEST_MSAA_CNTL_SAMPLES__MASK 0x00000003 +#define A5XX_GRAS_SC_DEST_MSAA_CNTL_SAMPLES__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_DEST_MSAA_CNTL_SAMPLES(enum a3xx_msaa_samples val) +{ + return ((val) << A5XX_GRAS_SC_DEST_MSAA_CNTL_SAMPLES__SHIFT) & A5XX_GRAS_SC_DEST_MSAA_CNTL_SAMPLES__MASK; +} +#define A5XX_GRAS_SC_DEST_MSAA_CNTL_MSAA_DISABLE 0x00000004 + +#define REG_A5XX_GRAS_SC_SCREEN_SCISSOR_CNTL 0x0000e0a4 + +#define REG_A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0 0x0000e0aa +#define A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_X__MASK 0x00007fff +#define A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_X__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_X(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_X__SHIFT) & A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_X__MASK; +} +#define A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_Y__MASK 0x7fff0000 +#define A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_Y__SHIFT 16 +static inline uint32_t A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_Y(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_Y__SHIFT) & A5XX_GRAS_SC_SCREEN_SCISSOR_TL_0_Y__MASK; +} + +#define REG_A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0 0x0000e0ab +#define A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_X__MASK 0x00007fff +#define A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_X__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_X(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_X__SHIFT) & A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_X__MASK; +} +#define A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_Y__MASK 0x7fff0000 +#define A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_Y__SHIFT 16 +static inline uint32_t A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_Y(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_Y__SHIFT) & A5XX_GRAS_SC_SCREEN_SCISSOR_BR_0_Y__MASK; +} + +#define REG_A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0 0x0000e0ca +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_X__MASK 0x00007fff +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_X__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_X(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_X__SHIFT) & A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_X__MASK; +} +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_Y__MASK 0x7fff0000 +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_Y__SHIFT 16 +static inline uint32_t A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_Y(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_Y__SHIFT) & A5XX_GRAS_SC_VIEWPORT_SCISSOR_TL_0_Y__MASK; +} + +#define REG_A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0 0x0000e0cb +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_X__MASK 0x00007fff +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_X__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_X(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_X__SHIFT) & A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_X__MASK; +} +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_Y__MASK 0x7fff0000 +#define A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_Y__SHIFT 16 +static inline uint32_t A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_Y(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_Y__SHIFT) & A5XX_GRAS_SC_VIEWPORT_SCISSOR_BR_0_Y__MASK; +} + +#define REG_A5XX_GRAS_SC_WINDOW_SCISSOR_TL 0x0000e0ea +#define A5XX_GRAS_SC_WINDOW_SCISSOR_TL_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_GRAS_SC_WINDOW_SCISSOR_TL_X__MASK 0x00007fff +#define A5XX_GRAS_SC_WINDOW_SCISSOR_TL_X__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_WINDOW_SCISSOR_TL_X(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_WINDOW_SCISSOR_TL_X__SHIFT) & A5XX_GRAS_SC_WINDOW_SCISSOR_TL_X__MASK; +} +#define A5XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__MASK 0x7fff0000 +#define A5XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__SHIFT 16 +static inline uint32_t A5XX_GRAS_SC_WINDOW_SCISSOR_TL_Y(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__SHIFT) & A5XX_GRAS_SC_WINDOW_SCISSOR_TL_Y__MASK; +} + +#define REG_A5XX_GRAS_SC_WINDOW_SCISSOR_BR 0x0000e0eb +#define A5XX_GRAS_SC_WINDOW_SCISSOR_BR_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_GRAS_SC_WINDOW_SCISSOR_BR_X__MASK 0x00007fff +#define A5XX_GRAS_SC_WINDOW_SCISSOR_BR_X__SHIFT 0 +static inline uint32_t A5XX_GRAS_SC_WINDOW_SCISSOR_BR_X(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_WINDOW_SCISSOR_BR_X__SHIFT) & A5XX_GRAS_SC_WINDOW_SCISSOR_BR_X__MASK; +} +#define A5XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__MASK 0x7fff0000 +#define A5XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__SHIFT 16 +static inline uint32_t A5XX_GRAS_SC_WINDOW_SCISSOR_BR_Y(uint32_t val) +{ + return ((val) << A5XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__SHIFT) & A5XX_GRAS_SC_WINDOW_SCISSOR_BR_Y__MASK; +} + +#define REG_A5XX_GRAS_LRZ_CNTL 0x0000e100 + +#define REG_A5XX_GRAS_LRZ_BUFFER_BASE_LO 0x0000e101 + +#define REG_A5XX_GRAS_LRZ_BUFFER_BASE_HI 0x0000e102 + +#define REG_A5XX_GRAS_LRZ_BUFFER_PITCH 0x0000e103 + +#define REG_A5XX_GRAS_LRZ_FAST_CLEAR_BUFFER_BASE_LO 0x0000e104 + +#define REG_A5XX_GRAS_LRZ_FAST_CLEAR_BUFFER_BASE_HI 0x0000e105 + +#define REG_A5XX_RB_CNTL 0x0000e140 +#define A5XX_RB_CNTL_WIDTH__MASK 0x000000ff +#define A5XX_RB_CNTL_WIDTH__SHIFT 0 +static inline uint32_t A5XX_RB_CNTL_WIDTH(uint32_t val) +{ + return ((val >> 5) << A5XX_RB_CNTL_WIDTH__SHIFT) & A5XX_RB_CNTL_WIDTH__MASK; +} +#define A5XX_RB_CNTL_HEIGHT__MASK 0x0001fe00 +#define A5XX_RB_CNTL_HEIGHT__SHIFT 9 +static inline uint32_t A5XX_RB_CNTL_HEIGHT(uint32_t val) +{ + return ((val >> 5) << A5XX_RB_CNTL_HEIGHT__SHIFT) & A5XX_RB_CNTL_HEIGHT__MASK; +} +#define A5XX_RB_CNTL_BYPASS 0x00020000 + +#define REG_A5XX_RB_RENDER_CNTL 0x0000e141 +#define A5XX_RB_RENDER_CNTL_ENABLED_MRTS__MASK 0x00ff0000 +#define A5XX_RB_RENDER_CNTL_ENABLED_MRTS__SHIFT 16 +static inline uint32_t A5XX_RB_RENDER_CNTL_ENABLED_MRTS(uint32_t val) +{ + return ((val) << A5XX_RB_RENDER_CNTL_ENABLED_MRTS__SHIFT) & A5XX_RB_RENDER_CNTL_ENABLED_MRTS__MASK; +} +#define A5XX_RB_RENDER_CNTL_ENABLED_MRTS2__MASK 0xff000000 +#define A5XX_RB_RENDER_CNTL_ENABLED_MRTS2__SHIFT 24 +static inline uint32_t A5XX_RB_RENDER_CNTL_ENABLED_MRTS2(uint32_t val) +{ + return ((val) << A5XX_RB_RENDER_CNTL_ENABLED_MRTS2__SHIFT) & A5XX_RB_RENDER_CNTL_ENABLED_MRTS2__MASK; +} + +#define REG_A5XX_RB_RAS_MSAA_CNTL 0x0000e142 +#define A5XX_RB_RAS_MSAA_CNTL_SAMPLES__MASK 0x00000003 +#define A5XX_RB_RAS_MSAA_CNTL_SAMPLES__SHIFT 0 +static inline uint32_t A5XX_RB_RAS_MSAA_CNTL_SAMPLES(enum a3xx_msaa_samples val) +{ + return ((val) << A5XX_RB_RAS_MSAA_CNTL_SAMPLES__SHIFT) & A5XX_RB_RAS_MSAA_CNTL_SAMPLES__MASK; +} + +#define REG_A5XX_RB_DEST_MSAA_CNTL 0x0000e143 +#define A5XX_RB_DEST_MSAA_CNTL_SAMPLES__MASK 0x00000003 +#define A5XX_RB_DEST_MSAA_CNTL_SAMPLES__SHIFT 0 +static inline uint32_t A5XX_RB_DEST_MSAA_CNTL_SAMPLES(enum a3xx_msaa_samples val) +{ + return ((val) << A5XX_RB_DEST_MSAA_CNTL_SAMPLES__SHIFT) & A5XX_RB_DEST_MSAA_CNTL_SAMPLES__MASK; +} +#define A5XX_RB_DEST_MSAA_CNTL_MSAA_DISABLE 0x00000004 + +#define REG_A5XX_RB_RENDER_CONTROL0 0x0000e144 +#define A5XX_RB_RENDER_CONTROL0_VARYING 0x00000001 +#define A5XX_RB_RENDER_CONTROL0_XCOORD 0x00000040 +#define A5XX_RB_RENDER_CONTROL0_YCOORD 0x00000080 +#define A5XX_RB_RENDER_CONTROL0_ZCOORD 0x00000100 +#define A5XX_RB_RENDER_CONTROL0_WCOORD 0x00000200 + +#define REG_A5XX_RB_RENDER_CONTROL1 0x0000e145 +#define A5XX_RB_RENDER_CONTROL1_FACENESS 0x00000002 + +#define REG_A5XX_RB_FS_OUTPUT_CNTL 0x0000e146 +#define A5XX_RB_FS_OUTPUT_CNTL_MRT__MASK 0x0000000f +#define A5XX_RB_FS_OUTPUT_CNTL_MRT__SHIFT 0 +static inline uint32_t A5XX_RB_FS_OUTPUT_CNTL_MRT(uint32_t val) +{ + return ((val) << A5XX_RB_FS_OUTPUT_CNTL_MRT__SHIFT) & A5XX_RB_FS_OUTPUT_CNTL_MRT__MASK; +} +#define A5XX_RB_FS_OUTPUT_CNTL_FRAG_WRITES_Z 0x00000020 + +static inline uint32_t REG_A5XX_RB_MRT(uint32_t i0) { return 0x0000e150 + 0x7*i0; } + +static inline uint32_t REG_A5XX_RB_MRT_CONTROL(uint32_t i0) { return 0x0000e150 + 0x7*i0; } +#define A5XX_RB_MRT_CONTROL_BLEND 0x00000001 +#define A5XX_RB_MRT_CONTROL_BLEND2 0x00000002 +#define A5XX_RB_MRT_CONTROL_COMPONENT_ENABLE__MASK 0x00000780 +#define A5XX_RB_MRT_CONTROL_COMPONENT_ENABLE__SHIFT 7 +static inline uint32_t A5XX_RB_MRT_CONTROL_COMPONENT_ENABLE(uint32_t val) +{ + return ((val) << A5XX_RB_MRT_CONTROL_COMPONENT_ENABLE__SHIFT) & A5XX_RB_MRT_CONTROL_COMPONENT_ENABLE__MASK; +} + +static inline uint32_t REG_A5XX_RB_MRT_BLEND_CONTROL(uint32_t i0) { return 0x0000e151 + 0x7*i0; } +#define A5XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__MASK 0x0000001f +#define A5XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__SHIFT 0 +static inline uint32_t A5XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR(enum adreno_rb_blend_factor val) +{ + return ((val) << A5XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__SHIFT) & A5XX_RB_MRT_BLEND_CONTROL_RGB_SRC_FACTOR__MASK; +} +#define A5XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__MASK 0x000000e0 +#define A5XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__SHIFT 5 +static inline uint32_t A5XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE(enum a3xx_rb_blend_opcode val) +{ + return ((val) << A5XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__SHIFT) & A5XX_RB_MRT_BLEND_CONTROL_RGB_BLEND_OPCODE__MASK; +} +#define A5XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__MASK 0x00001f00 +#define A5XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__SHIFT 8 +static inline uint32_t A5XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR(enum adreno_rb_blend_factor val) +{ + return ((val) << A5XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__SHIFT) & A5XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR__MASK; +} +#define A5XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__MASK 0x001f0000 +#define A5XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__SHIFT 16 +static inline uint32_t A5XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR(enum adreno_rb_blend_factor val) +{ + return ((val) << A5XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__SHIFT) & A5XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR__MASK; +} +#define A5XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__MASK 0x00e00000 +#define A5XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__SHIFT 21 +static inline uint32_t A5XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE(enum a3xx_rb_blend_opcode val) +{ + return ((val) << A5XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__SHIFT) & A5XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE__MASK; +} +#define A5XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__MASK 0x1f000000 +#define A5XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__SHIFT 24 +static inline uint32_t A5XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR(enum adreno_rb_blend_factor val) +{ + return ((val) << A5XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__SHIFT) & A5XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR__MASK; +} + +static inline uint32_t REG_A5XX_RB_MRT_BUF_INFO(uint32_t i0) { return 0x0000e152 + 0x7*i0; } +#define A5XX_RB_MRT_BUF_INFO_COLOR_FORMAT__MASK 0x0000007f +#define A5XX_RB_MRT_BUF_INFO_COLOR_FORMAT__SHIFT 0 +static inline uint32_t A5XX_RB_MRT_BUF_INFO_COLOR_FORMAT(enum a5xx_color_fmt val) +{ + return ((val) << A5XX_RB_MRT_BUF_INFO_COLOR_FORMAT__SHIFT) & A5XX_RB_MRT_BUF_INFO_COLOR_FORMAT__MASK; +} +#define A5XX_RB_MRT_BUF_INFO_COLOR_TILE_MODE__MASK 0x00000300 +#define A5XX_RB_MRT_BUF_INFO_COLOR_TILE_MODE__SHIFT 8 +static inline uint32_t A5XX_RB_MRT_BUF_INFO_COLOR_TILE_MODE(enum a5xx_tile_mode val) +{ + return ((val) << A5XX_RB_MRT_BUF_INFO_COLOR_TILE_MODE__SHIFT) & A5XX_RB_MRT_BUF_INFO_COLOR_TILE_MODE__MASK; +} +#define A5XX_RB_MRT_BUF_INFO_COLOR_SRGB 0x00008000 + +static inline uint32_t REG_A5XX_RB_MRT_PITCH(uint32_t i0) { return 0x0000e153 + 0x7*i0; } +#define A5XX_RB_MRT_PITCH_COLOR_BUF_PITCH__MASK 0x0007ffff +#define A5XX_RB_MRT_PITCH_COLOR_BUF_PITCH__SHIFT 0 +static inline uint32_t A5XX_RB_MRT_PITCH_COLOR_BUF_PITCH(uint32_t val) +{ + return ((val >> 4) << A5XX_RB_MRT_PITCH_COLOR_BUF_PITCH__SHIFT) & A5XX_RB_MRT_PITCH_COLOR_BUF_PITCH__MASK; +} + +static inline uint32_t REG_A5XX_RB_MRT_ARRAY_PITCH(uint32_t i0) { return 0x0000e154 + 0x7*i0; } +#define A5XX_RB_MRT_ARRAY_PITCH_COLOR_BUF_SIZE__MASK 0x01ffffff +#define A5XX_RB_MRT_ARRAY_PITCH_COLOR_BUF_SIZE__SHIFT 0 +static inline uint32_t A5XX_RB_MRT_ARRAY_PITCH_COLOR_BUF_SIZE(uint32_t val) +{ + return ((val >> 6) << A5XX_RB_MRT_ARRAY_PITCH_COLOR_BUF_SIZE__SHIFT) & A5XX_RB_MRT_ARRAY_PITCH_COLOR_BUF_SIZE__MASK; +} + +static inline uint32_t REG_A5XX_RB_MRT_BASE_LO(uint32_t i0) { return 0x0000e155 + 0x7*i0; } + +static inline uint32_t REG_A5XX_RB_MRT_BASE_HI(uint32_t i0) { return 0x0000e156 + 0x7*i0; } + +#define REG_A5XX_RB_BLEND_RED 0x0000e1a0 +#define A5XX_RB_BLEND_RED_UINT__MASK 0x000000ff +#define A5XX_RB_BLEND_RED_UINT__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_RED_UINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_RED_UINT__SHIFT) & A5XX_RB_BLEND_RED_UINT__MASK; +} +#define A5XX_RB_BLEND_RED_SINT__MASK 0x0000ff00 +#define A5XX_RB_BLEND_RED_SINT__SHIFT 8 +static inline uint32_t A5XX_RB_BLEND_RED_SINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_RED_SINT__SHIFT) & A5XX_RB_BLEND_RED_SINT__MASK; +} +#define A5XX_RB_BLEND_RED_FLOAT__MASK 0xffff0000 +#define A5XX_RB_BLEND_RED_FLOAT__SHIFT 16 +static inline uint32_t A5XX_RB_BLEND_RED_FLOAT(float val) +{ + return ((util_float_to_half(val)) << A5XX_RB_BLEND_RED_FLOAT__SHIFT) & A5XX_RB_BLEND_RED_FLOAT__MASK; +} + +#define REG_A5XX_RB_BLEND_RED_F32 0x0000e1a1 +#define A5XX_RB_BLEND_RED_F32__MASK 0xffffffff +#define A5XX_RB_BLEND_RED_F32__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_RED_F32(float val) +{ + return ((fui(val)) << A5XX_RB_BLEND_RED_F32__SHIFT) & A5XX_RB_BLEND_RED_F32__MASK; +} + +#define REG_A5XX_RB_BLEND_GREEN 0x0000e1a2 +#define A5XX_RB_BLEND_GREEN_UINT__MASK 0x000000ff +#define A5XX_RB_BLEND_GREEN_UINT__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_GREEN_UINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_GREEN_UINT__SHIFT) & A5XX_RB_BLEND_GREEN_UINT__MASK; +} +#define A5XX_RB_BLEND_GREEN_SINT__MASK 0x0000ff00 +#define A5XX_RB_BLEND_GREEN_SINT__SHIFT 8 +static inline uint32_t A5XX_RB_BLEND_GREEN_SINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_GREEN_SINT__SHIFT) & A5XX_RB_BLEND_GREEN_SINT__MASK; +} +#define A5XX_RB_BLEND_GREEN_FLOAT__MASK 0xffff0000 +#define A5XX_RB_BLEND_GREEN_FLOAT__SHIFT 16 +static inline uint32_t A5XX_RB_BLEND_GREEN_FLOAT(float val) +{ + return ((util_float_to_half(val)) << A5XX_RB_BLEND_GREEN_FLOAT__SHIFT) & A5XX_RB_BLEND_GREEN_FLOAT__MASK; +} + +#define REG_A5XX_RB_BLEND_GREEN_F32 0x0000e1a3 +#define A5XX_RB_BLEND_GREEN_F32__MASK 0xffffffff +#define A5XX_RB_BLEND_GREEN_F32__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_GREEN_F32(float val) +{ + return ((fui(val)) << A5XX_RB_BLEND_GREEN_F32__SHIFT) & A5XX_RB_BLEND_GREEN_F32__MASK; +} + +#define REG_A5XX_RB_BLEND_BLUE 0x0000e1a4 +#define A5XX_RB_BLEND_BLUE_UINT__MASK 0x000000ff +#define A5XX_RB_BLEND_BLUE_UINT__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_BLUE_UINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_BLUE_UINT__SHIFT) & A5XX_RB_BLEND_BLUE_UINT__MASK; +} +#define A5XX_RB_BLEND_BLUE_SINT__MASK 0x0000ff00 +#define A5XX_RB_BLEND_BLUE_SINT__SHIFT 8 +static inline uint32_t A5XX_RB_BLEND_BLUE_SINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_BLUE_SINT__SHIFT) & A5XX_RB_BLEND_BLUE_SINT__MASK; +} +#define A5XX_RB_BLEND_BLUE_FLOAT__MASK 0xffff0000 +#define A5XX_RB_BLEND_BLUE_FLOAT__SHIFT 16 +static inline uint32_t A5XX_RB_BLEND_BLUE_FLOAT(float val) +{ + return ((util_float_to_half(val)) << A5XX_RB_BLEND_BLUE_FLOAT__SHIFT) & A5XX_RB_BLEND_BLUE_FLOAT__MASK; +} + +#define REG_A5XX_RB_BLEND_BLUE_F32 0x0000e1a5 +#define A5XX_RB_BLEND_BLUE_F32__MASK 0xffffffff +#define A5XX_RB_BLEND_BLUE_F32__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_BLUE_F32(float val) +{ + return ((fui(val)) << A5XX_RB_BLEND_BLUE_F32__SHIFT) & A5XX_RB_BLEND_BLUE_F32__MASK; +} + +#define REG_A5XX_RB_BLEND_ALPHA 0x0000e1a6 +#define A5XX_RB_BLEND_ALPHA_UINT__MASK 0x000000ff +#define A5XX_RB_BLEND_ALPHA_UINT__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_ALPHA_UINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_ALPHA_UINT__SHIFT) & A5XX_RB_BLEND_ALPHA_UINT__MASK; +} +#define A5XX_RB_BLEND_ALPHA_SINT__MASK 0x0000ff00 +#define A5XX_RB_BLEND_ALPHA_SINT__SHIFT 8 +static inline uint32_t A5XX_RB_BLEND_ALPHA_SINT(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_ALPHA_SINT__SHIFT) & A5XX_RB_BLEND_ALPHA_SINT__MASK; +} +#define A5XX_RB_BLEND_ALPHA_FLOAT__MASK 0xffff0000 +#define A5XX_RB_BLEND_ALPHA_FLOAT__SHIFT 16 +static inline uint32_t A5XX_RB_BLEND_ALPHA_FLOAT(float val) +{ + return ((util_float_to_half(val)) << A5XX_RB_BLEND_ALPHA_FLOAT__SHIFT) & A5XX_RB_BLEND_ALPHA_FLOAT__MASK; +} + +#define REG_A5XX_RB_BLEND_ALPHA_F32 0x0000e1a7 +#define A5XX_RB_BLEND_ALPHA_F32__MASK 0xffffffff +#define A5XX_RB_BLEND_ALPHA_F32__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_ALPHA_F32(float val) +{ + return ((fui(val)) << A5XX_RB_BLEND_ALPHA_F32__SHIFT) & A5XX_RB_BLEND_ALPHA_F32__MASK; +} + +#define REG_A5XX_RB_ALPHA_CONTROL 0x0000e1a8 +#define A5XX_RB_ALPHA_CONTROL_ALPHA_REF__MASK 0x000000ff +#define A5XX_RB_ALPHA_CONTROL_ALPHA_REF__SHIFT 0 +static inline uint32_t A5XX_RB_ALPHA_CONTROL_ALPHA_REF(uint32_t val) +{ + return ((val) << A5XX_RB_ALPHA_CONTROL_ALPHA_REF__SHIFT) & A5XX_RB_ALPHA_CONTROL_ALPHA_REF__MASK; +} +#define A5XX_RB_ALPHA_CONTROL_ALPHA_TEST 0x00000100 +#define A5XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__MASK 0x00000e00 +#define A5XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__SHIFT 9 +static inline uint32_t A5XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC(enum adreno_compare_func val) +{ + return ((val) << A5XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__SHIFT) & A5XX_RB_ALPHA_CONTROL_ALPHA_TEST_FUNC__MASK; +} + +#define REG_A5XX_RB_BLEND_CNTL 0x0000e1a9 +#define A5XX_RB_BLEND_CNTL_ENABLE_BLEND__MASK 0x000000ff +#define A5XX_RB_BLEND_CNTL_ENABLE_BLEND__SHIFT 0 +static inline uint32_t A5XX_RB_BLEND_CNTL_ENABLE_BLEND(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_CNTL_ENABLE_BLEND__SHIFT) & A5XX_RB_BLEND_CNTL_ENABLE_BLEND__MASK; +} +#define A5XX_RB_BLEND_CNTL_INDEPENDENT_BLEND 0x00000100 +#define A5XX_RB_BLEND_CNTL_SAMPLE_MASK__MASK 0xffff0000 +#define A5XX_RB_BLEND_CNTL_SAMPLE_MASK__SHIFT 16 +static inline uint32_t A5XX_RB_BLEND_CNTL_SAMPLE_MASK(uint32_t val) +{ + return ((val) << A5XX_RB_BLEND_CNTL_SAMPLE_MASK__SHIFT) & A5XX_RB_BLEND_CNTL_SAMPLE_MASK__MASK; +} + +#define REG_A5XX_RB_DEPTH_PLANE_CNTL 0x0000e1b0 +#define A5XX_RB_DEPTH_PLANE_CNTL_FRAG_WRITES_Z 0x00000001 + +#define REG_A5XX_RB_DEPTH_CNTL 0x0000e1b1 +#define A5XX_RB_DEPTH_CNTL_Z_ENABLE 0x00000001 +#define A5XX_RB_DEPTH_CNTL_Z_WRITE_ENABLE 0x00000002 +#define A5XX_RB_DEPTH_CNTL_ZFUNC__MASK 0x0000001c +#define A5XX_RB_DEPTH_CNTL_ZFUNC__SHIFT 2 +static inline uint32_t A5XX_RB_DEPTH_CNTL_ZFUNC(enum adreno_compare_func val) +{ + return ((val) << A5XX_RB_DEPTH_CNTL_ZFUNC__SHIFT) & A5XX_RB_DEPTH_CNTL_ZFUNC__MASK; +} +#define A5XX_RB_DEPTH_CNTL_Z_TEST_ENABLE 0x00000040 + +#define REG_A5XX_RB_DEPTH_BUFFER_INFO 0x0000e1b2 +#define A5XX_RB_DEPTH_BUFFER_INFO_DEPTH_FORMAT__MASK 0x00000007 +#define A5XX_RB_DEPTH_BUFFER_INFO_DEPTH_FORMAT__SHIFT 0 +static inline uint32_t A5XX_RB_DEPTH_BUFFER_INFO_DEPTH_FORMAT(enum a5xx_depth_format val) +{ + return ((val) << A5XX_RB_DEPTH_BUFFER_INFO_DEPTH_FORMAT__SHIFT) & A5XX_RB_DEPTH_BUFFER_INFO_DEPTH_FORMAT__MASK; +} + +#define REG_A5XX_RB_DEPTH_BUFFER_BASE_LO 0x0000e1b3 + +#define REG_A5XX_RB_DEPTH_BUFFER_BASE_HI 0x0000e1b4 + +#define REG_A5XX_RB_DEPTH_BUFFER_PITCH 0x0000e1b5 +#define A5XX_RB_DEPTH_BUFFER_PITCH__MASK 0xffffffff +#define A5XX_RB_DEPTH_BUFFER_PITCH__SHIFT 0 +static inline uint32_t A5XX_RB_DEPTH_BUFFER_PITCH(uint32_t val) +{ + return ((val >> 5) << A5XX_RB_DEPTH_BUFFER_PITCH__SHIFT) & A5XX_RB_DEPTH_BUFFER_PITCH__MASK; +} + +#define REG_A5XX_RB_DEPTH_BUFFER_ARRAY_PITCH 0x0000e1b6 +#define A5XX_RB_DEPTH_BUFFER_ARRAY_PITCH__MASK 0xffffffff +#define A5XX_RB_DEPTH_BUFFER_ARRAY_PITCH__SHIFT 0 +static inline uint32_t A5XX_RB_DEPTH_BUFFER_ARRAY_PITCH(uint32_t val) +{ + return ((val >> 5) << A5XX_RB_DEPTH_BUFFER_ARRAY_PITCH__SHIFT) & A5XX_RB_DEPTH_BUFFER_ARRAY_PITCH__MASK; +} + +#define REG_A5XX_RB_STENCIL_CONTROL 0x0000e1c0 +#define A5XX_RB_STENCIL_CONTROL_STENCIL_ENABLE 0x00000001 +#define A5XX_RB_STENCIL_CONTROL_STENCIL_ENABLE_BF 0x00000002 +#define A5XX_RB_STENCIL_CONTROL_STENCIL_READ 0x00000004 +#define A5XX_RB_STENCIL_CONTROL_FUNC__MASK 0x00000700 +#define A5XX_RB_STENCIL_CONTROL_FUNC__SHIFT 8 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_FUNC(enum adreno_compare_func val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_FUNC__SHIFT) & A5XX_RB_STENCIL_CONTROL_FUNC__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_FAIL__MASK 0x00003800 +#define A5XX_RB_STENCIL_CONTROL_FAIL__SHIFT 11 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_FAIL(enum adreno_stencil_op val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_FAIL__SHIFT) & A5XX_RB_STENCIL_CONTROL_FAIL__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_ZPASS__MASK 0x0001c000 +#define A5XX_RB_STENCIL_CONTROL_ZPASS__SHIFT 14 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_ZPASS(enum adreno_stencil_op val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_ZPASS__SHIFT) & A5XX_RB_STENCIL_CONTROL_ZPASS__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_ZFAIL__MASK 0x000e0000 +#define A5XX_RB_STENCIL_CONTROL_ZFAIL__SHIFT 17 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_ZFAIL(enum adreno_stencil_op val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_ZFAIL__SHIFT) & A5XX_RB_STENCIL_CONTROL_ZFAIL__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_FUNC_BF__MASK 0x00700000 +#define A5XX_RB_STENCIL_CONTROL_FUNC_BF__SHIFT 20 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_FUNC_BF(enum adreno_compare_func val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_FUNC_BF__SHIFT) & A5XX_RB_STENCIL_CONTROL_FUNC_BF__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_FAIL_BF__MASK 0x03800000 +#define A5XX_RB_STENCIL_CONTROL_FAIL_BF__SHIFT 23 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_FAIL_BF(enum adreno_stencil_op val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_FAIL_BF__SHIFT) & A5XX_RB_STENCIL_CONTROL_FAIL_BF__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_ZPASS_BF__MASK 0x1c000000 +#define A5XX_RB_STENCIL_CONTROL_ZPASS_BF__SHIFT 26 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_ZPASS_BF(enum adreno_stencil_op val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_ZPASS_BF__SHIFT) & A5XX_RB_STENCIL_CONTROL_ZPASS_BF__MASK; +} +#define A5XX_RB_STENCIL_CONTROL_ZFAIL_BF__MASK 0xe0000000 +#define A5XX_RB_STENCIL_CONTROL_ZFAIL_BF__SHIFT 29 +static inline uint32_t A5XX_RB_STENCIL_CONTROL_ZFAIL_BF(enum adreno_stencil_op val) +{ + return ((val) << A5XX_RB_STENCIL_CONTROL_ZFAIL_BF__SHIFT) & A5XX_RB_STENCIL_CONTROL_ZFAIL_BF__MASK; +} + +#define REG_A5XX_RB_STENCIL_INFO 0x0000e1c2 +#define A5XX_RB_STENCIL_INFO_SEPARATE_STENCIL 0x00000001 +#define A5XX_RB_STENCIL_INFO_STENCIL_BASE__MASK 0xfffff000 +#define A5XX_RB_STENCIL_INFO_STENCIL_BASE__SHIFT 12 +static inline uint32_t A5XX_RB_STENCIL_INFO_STENCIL_BASE(uint32_t val) +{ + return ((val >> 12) << A5XX_RB_STENCIL_INFO_STENCIL_BASE__SHIFT) & A5XX_RB_STENCIL_INFO_STENCIL_BASE__MASK; +} + +#define REG_A5XX_UNKNOWN_E1C3 0x0000e1c3 + +#define REG_A5XX_RB_STENCILREFMASK 0x0000e1c6 +#define A5XX_RB_STENCILREFMASK_STENCILREF__MASK 0x000000ff +#define A5XX_RB_STENCILREFMASK_STENCILREF__SHIFT 0 +static inline uint32_t A5XX_RB_STENCILREFMASK_STENCILREF(uint32_t val) +{ + return ((val) << A5XX_RB_STENCILREFMASK_STENCILREF__SHIFT) & A5XX_RB_STENCILREFMASK_STENCILREF__MASK; +} +#define A5XX_RB_STENCILREFMASK_STENCILMASK__MASK 0x0000ff00 +#define A5XX_RB_STENCILREFMASK_STENCILMASK__SHIFT 8 +static inline uint32_t A5XX_RB_STENCILREFMASK_STENCILMASK(uint32_t val) +{ + return ((val) << A5XX_RB_STENCILREFMASK_STENCILMASK__SHIFT) & A5XX_RB_STENCILREFMASK_STENCILMASK__MASK; +} +#define A5XX_RB_STENCILREFMASK_STENCILWRITEMASK__MASK 0x00ff0000 +#define A5XX_RB_STENCILREFMASK_STENCILWRITEMASK__SHIFT 16 +static inline uint32_t A5XX_RB_STENCILREFMASK_STENCILWRITEMASK(uint32_t val) +{ + return ((val) << A5XX_RB_STENCILREFMASK_STENCILWRITEMASK__SHIFT) & A5XX_RB_STENCILREFMASK_STENCILWRITEMASK__MASK; +} + +#define REG_A5XX_RB_WINDOW_OFFSET 0x0000e1d0 +#define A5XX_RB_WINDOW_OFFSET_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_RB_WINDOW_OFFSET_X__MASK 0x00007fff +#define A5XX_RB_WINDOW_OFFSET_X__SHIFT 0 +static inline uint32_t A5XX_RB_WINDOW_OFFSET_X(uint32_t val) +{ + return ((val) << A5XX_RB_WINDOW_OFFSET_X__SHIFT) & A5XX_RB_WINDOW_OFFSET_X__MASK; +} +#define A5XX_RB_WINDOW_OFFSET_Y__MASK 0x7fff0000 +#define A5XX_RB_WINDOW_OFFSET_Y__SHIFT 16 +static inline uint32_t A5XX_RB_WINDOW_OFFSET_Y(uint32_t val) +{ + return ((val) << A5XX_RB_WINDOW_OFFSET_Y__SHIFT) & A5XX_RB_WINDOW_OFFSET_Y__MASK; +} + +#define REG_A5XX_RB_RESOLVE_CNTL_1 0x0000e211 +#define A5XX_RB_RESOLVE_CNTL_1_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_RB_RESOLVE_CNTL_1_X__MASK 0x00007fff +#define A5XX_RB_RESOLVE_CNTL_1_X__SHIFT 0 +static inline uint32_t A5XX_RB_RESOLVE_CNTL_1_X(uint32_t val) +{ + return ((val) << A5XX_RB_RESOLVE_CNTL_1_X__SHIFT) & A5XX_RB_RESOLVE_CNTL_1_X__MASK; +} +#define A5XX_RB_RESOLVE_CNTL_1_Y__MASK 0x7fff0000 +#define A5XX_RB_RESOLVE_CNTL_1_Y__SHIFT 16 +static inline uint32_t A5XX_RB_RESOLVE_CNTL_1_Y(uint32_t val) +{ + return ((val) << A5XX_RB_RESOLVE_CNTL_1_Y__SHIFT) & A5XX_RB_RESOLVE_CNTL_1_Y__MASK; +} + +#define REG_A5XX_RB_RESOLVE_CNTL_2 0x0000e212 +#define A5XX_RB_RESOLVE_CNTL_2_WINDOW_OFFSET_DISABLE 0x80000000 +#define A5XX_RB_RESOLVE_CNTL_2_X__MASK 0x00007fff +#define A5XX_RB_RESOLVE_CNTL_2_X__SHIFT 0 +static inline uint32_t A5XX_RB_RESOLVE_CNTL_2_X(uint32_t val) +{ + return ((val) << A5XX_RB_RESOLVE_CNTL_2_X__SHIFT) & A5XX_RB_RESOLVE_CNTL_2_X__MASK; +} +#define A5XX_RB_RESOLVE_CNTL_2_Y__MASK 0x7fff0000 +#define A5XX_RB_RESOLVE_CNTL_2_Y__SHIFT 16 +static inline uint32_t A5XX_RB_RESOLVE_CNTL_2_Y(uint32_t val) +{ + return ((val) << A5XX_RB_RESOLVE_CNTL_2_Y__SHIFT) & A5XX_RB_RESOLVE_CNTL_2_Y__MASK; +} + +#define REG_A5XX_RB_DEPTH_FLAG_BUFFER_BASE_LO 0x0000e240 + +#define REG_A5XX_RB_DEPTH_FLAG_BUFFER_BASE_HI 0x0000e241 + +#define REG_A5XX_RB_DEPTH_FLAG_BUFFER_PITCH 0x0000e242 + +#define REG_A5XX_VPC_CNTL_0 0x0000e280 +#define A5XX_VPC_CNTL_0_STRIDE_IN_VPC__MASK 0x0000007f +#define A5XX_VPC_CNTL_0_STRIDE_IN_VPC__SHIFT 0 +static inline uint32_t A5XX_VPC_CNTL_0_STRIDE_IN_VPC(uint32_t val) +{ + return ((val) << A5XX_VPC_CNTL_0_STRIDE_IN_VPC__SHIFT) & A5XX_VPC_CNTL_0_STRIDE_IN_VPC__MASK; +} + +static inline uint32_t REG_A5XX_VPC_VARYING_INTERP(uint32_t i0) { return 0x0000e282 + 0x1*i0; } + +static inline uint32_t REG_A5XX_VPC_VARYING_INTERP_MODE(uint32_t i0) { return 0x0000e282 + 0x1*i0; } + +static inline uint32_t REG_A5XX_VPC_VARYING_PS_REPL(uint32_t i0) { return 0x0000e28a + 0x1*i0; } + +static inline uint32_t REG_A5XX_VPC_VARYING_PS_REPL_MODE(uint32_t i0) { return 0x0000e28a + 0x1*i0; } + +static inline uint32_t REG_A5XX_VPC_VAR(uint32_t i0) { return 0x0000e294 + 0x1*i0; } + +static inline uint32_t REG_A5XX_VPC_VAR_DISABLE(uint32_t i0) { return 0x0000e294 + 0x1*i0; } + +#define REG_A5XX_VPC_GS_SIV_CNTL 0x0000e298 + +#define REG_A5XX_VPC_PACK 0x0000e29d +#define A5XX_VPC_PACK_NUMNONPOSVAR__MASK 0x000000ff +#define A5XX_VPC_PACK_NUMNONPOSVAR__SHIFT 0 +static inline uint32_t A5XX_VPC_PACK_NUMNONPOSVAR(uint32_t val) +{ + return ((val) << A5XX_VPC_PACK_NUMNONPOSVAR__SHIFT) & A5XX_VPC_PACK_NUMNONPOSVAR__MASK; +} + +#define REG_A5XX_VPC_FS_PRIMITIVEID_CNTL 0x0000e2a0 + +#define REG_A5XX_VPC_SO_OVERRIDE 0x0000e2a2 + +#define REG_A5XX_VPC_SO_BUFFER_BASE_LO_0 0x0000e2a7 + +#define REG_A5XX_VPC_SO_BUFFER_BASE_HI_0 0x0000e2a8 + +#define REG_A5XX_VPC_SO_BUFFER_SIZE_0 0x0000e2a9 + +#define REG_A5XX_VPC_SO_FLUSH_BASE_LO_0 0x0000e2ac + +#define REG_A5XX_VPC_SO_FLUSH_BASE_HI_0 0x0000e2ad + +#define REG_A5XX_PC_PRIMITIVE_CNTL 0x0000e384 +#define A5XX_PC_PRIMITIVE_CNTL_STRIDE_IN_VPC__MASK 0x0000007f +#define A5XX_PC_PRIMITIVE_CNTL_STRIDE_IN_VPC__SHIFT 0 +static inline uint32_t A5XX_PC_PRIMITIVE_CNTL_STRIDE_IN_VPC(uint32_t val) +{ + return ((val) << A5XX_PC_PRIMITIVE_CNTL_STRIDE_IN_VPC__SHIFT) & A5XX_PC_PRIMITIVE_CNTL_STRIDE_IN_VPC__MASK; +} + +#define REG_A5XX_PC_RASTER_CNTL 0x0000e388 + +#define REG_A5XX_PC_RESTART_INDEX 0x0000e38c + +#define REG_A5XX_PC_GS_PARAM 0x0000e38e + +#define REG_A5XX_PC_HS_PARAM 0x0000e38f + +#define REG_A5XX_PC_POWER_CNTL 0x0000e3b0 + +#define REG_A5XX_VFD_CONTROL_0 0x0000e400 +#define A5XX_VFD_CONTROL_0_STRMDECINSTRCNT__MASK 0x0000003f +#define A5XX_VFD_CONTROL_0_STRMDECINSTRCNT__SHIFT 0 +static inline uint32_t A5XX_VFD_CONTROL_0_STRMDECINSTRCNT(uint32_t val) +{ + return ((val) << A5XX_VFD_CONTROL_0_STRMDECINSTRCNT__SHIFT) & A5XX_VFD_CONTROL_0_STRMDECINSTRCNT__MASK; +} + +#define REG_A5XX_VFD_CONTROL_1 0x0000e401 +#define A5XX_VFD_CONTROL_1_REGID4INST__MASK 0x0000ff00 +#define A5XX_VFD_CONTROL_1_REGID4INST__SHIFT 8 +static inline uint32_t A5XX_VFD_CONTROL_1_REGID4INST(uint32_t val) +{ + return ((val) << A5XX_VFD_CONTROL_1_REGID4INST__SHIFT) & A5XX_VFD_CONTROL_1_REGID4INST__MASK; +} +#define A5XX_VFD_CONTROL_1_REGID4VTX__MASK 0x00ff0000 +#define A5XX_VFD_CONTROL_1_REGID4VTX__SHIFT 16 +static inline uint32_t A5XX_VFD_CONTROL_1_REGID4VTX(uint32_t val) +{ + return ((val) << A5XX_VFD_CONTROL_1_REGID4VTX__SHIFT) & A5XX_VFD_CONTROL_1_REGID4VTX__MASK; +} + +#define REG_A5XX_VFD_CONTROL_2 0x0000e402 + +#define REG_A5XX_VFD_CONTROL_3 0x0000e403 + +#define REG_A5XX_VFD_CONTROL_4 0x0000e404 + +#define REG_A5XX_VFD_CONTROL_5 0x0000e405 + +#define REG_A5XX_VFD_INDEX_OFFSET 0x0000e408 + +#define REG_A5XX_VFD_INSTANCE_START_OFFSET 0x0000e409 + +static inline uint32_t REG_A5XX_VFD_FETCH(uint32_t i0) { return 0x0000e40a + 0x4*i0; } + +static inline uint32_t REG_A5XX_VFD_FETCH_BASE_LO(uint32_t i0) { return 0x0000e40a + 0x4*i0; } + +static inline uint32_t REG_A5XX_VFD_FETCH_BASE_HI(uint32_t i0) { return 0x0000e40b + 0x4*i0; } + +static inline uint32_t REG_A5XX_VFD_FETCH_SIZE(uint32_t i0) { return 0x0000e40c + 0x4*i0; } + +static inline uint32_t REG_A5XX_VFD_FETCH_STRIDE(uint32_t i0) { return 0x0000e40d + 0x4*i0; } + +static inline uint32_t REG_A5XX_VFD_DECODE(uint32_t i0) { return 0x0000e48a + 0x2*i0; } + +static inline uint32_t REG_A5XX_VFD_DECODE_INSTR(uint32_t i0) { return 0x0000e48a + 0x2*i0; } +#define A5XX_VFD_DECODE_INSTR_IDX__MASK 0x0000001f +#define A5XX_VFD_DECODE_INSTR_IDX__SHIFT 0 +static inline uint32_t A5XX_VFD_DECODE_INSTR_IDX(uint32_t val) +{ + return ((val) << A5XX_VFD_DECODE_INSTR_IDX__SHIFT) & A5XX_VFD_DECODE_INSTR_IDX__MASK; +} +#define A5XX_VFD_DECODE_INSTR_FORMAT__MASK 0x3ff00000 +#define A5XX_VFD_DECODE_INSTR_FORMAT__SHIFT 20 +static inline uint32_t A5XX_VFD_DECODE_INSTR_FORMAT(enum a5xx_vtx_fmt val) +{ + return ((val) << A5XX_VFD_DECODE_INSTR_FORMAT__SHIFT) & A5XX_VFD_DECODE_INSTR_FORMAT__MASK; +} + +static inline uint32_t REG_A5XX_VFD_DECODE_STEP_RATE(uint32_t i0) { return 0x0000e48b + 0x2*i0; } + +static inline uint32_t REG_A5XX_VFD_DEST_CNTL(uint32_t i0) { return 0x0000e4ca + 0x1*i0; } + +static inline uint32_t REG_A5XX_VFD_DEST_CNTL_INSTR(uint32_t i0) { return 0x0000e4ca + 0x1*i0; } +#define A5XX_VFD_DEST_CNTL_INSTR_WRITEMASK__MASK 0x0000000f +#define A5XX_VFD_DEST_CNTL_INSTR_WRITEMASK__SHIFT 0 +static inline uint32_t A5XX_VFD_DEST_CNTL_INSTR_WRITEMASK(uint32_t val) +{ + return ((val) << A5XX_VFD_DEST_CNTL_INSTR_WRITEMASK__SHIFT) & A5XX_VFD_DEST_CNTL_INSTR_WRITEMASK__MASK; +} +#define A5XX_VFD_DEST_CNTL_INSTR_REGID__MASK 0x00000ff0 +#define A5XX_VFD_DEST_CNTL_INSTR_REGID__SHIFT 4 +static inline uint32_t A5XX_VFD_DEST_CNTL_INSTR_REGID(uint32_t val) +{ + return ((val) << A5XX_VFD_DEST_CNTL_INSTR_REGID__SHIFT) & A5XX_VFD_DEST_CNTL_INSTR_REGID__MASK; +} + +#define REG_A5XX_VFD_POWER_CNTL 0x0000e4f0 + +#define REG_A5XX_SP_SP_CNTL 0x0000e580 + +#define REG_A5XX_SP_VS_CONTROL_REG 0x0000e584 +#define A5XX_SP_VS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_SP_VS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_SP_VS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_SP_VS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_VS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_SP_VS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_SP_VS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_SP_VS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_SP_VS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_VS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_SP_VS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_SP_FS_CONTROL_REG 0x0000e585 +#define A5XX_SP_FS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_SP_FS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_SP_FS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_SP_FS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_FS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_SP_FS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_SP_FS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_SP_FS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_SP_FS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_FS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_SP_FS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_SP_HS_CONTROL_REG 0x0000e586 +#define A5XX_SP_HS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_SP_HS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_SP_HS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_SP_HS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_HS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_SP_HS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_SP_HS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_SP_HS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_SP_HS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_HS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_SP_HS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_SP_DS_CONTROL_REG 0x0000e587 +#define A5XX_SP_DS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_SP_DS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_SP_DS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_SP_DS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_DS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_SP_DS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_SP_DS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_SP_DS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_SP_DS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_DS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_SP_DS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_SP_GS_CONTROL_REG 0x0000e588 +#define A5XX_SP_GS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_SP_GS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_SP_GS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_SP_GS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_GS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_SP_GS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_SP_GS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_SP_GS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_SP_GS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_SP_GS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_SP_GS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_SP_CS_CONFIG 0x0000e589 + +#define REG_A5XX_SP_VS_CONFIG_MAX_CONST 0x0000e58a + +#define REG_A5XX_SP_FS_CONFIG_MAX_CONST 0x0000e58b + +#define REG_A5XX_SP_VS_CTRL_REG0 0x0000e590 +#define A5XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__MASK 0x000003f0 +#define A5XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT 4 +static inline uint32_t A5XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT(uint32_t val) +{ + return ((val) << A5XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT) & A5XX_SP_VS_CTRL_REG0_HALFREGFOOTPRINT__MASK; +} +#define A5XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__MASK 0x0000fc00 +#define A5XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT 10 +static inline uint32_t A5XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT(uint32_t val) +{ + return ((val) << A5XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT) & A5XX_SP_VS_CTRL_REG0_FULLREGFOOTPRINT__MASK; +} +#define A5XX_SP_VS_CTRL_REG0_VARYING 0x00010000 +#define A5XX_SP_VS_CTRL_REG0_PIXLODENABLE 0x00100000 + +static inline uint32_t REG_A5XX_SP_VS_OUT(uint32_t i0) { return 0x0000e593 + 0x1*i0; } + +static inline uint32_t REG_A5XX_SP_VS_OUT_REG(uint32_t i0) { return 0x0000e593 + 0x1*i0; } +#define A5XX_SP_VS_OUT_REG_A_REGID__MASK 0x000000ff +#define A5XX_SP_VS_OUT_REG_A_REGID__SHIFT 0 +static inline uint32_t A5XX_SP_VS_OUT_REG_A_REGID(uint32_t val) +{ + return ((val) << A5XX_SP_VS_OUT_REG_A_REGID__SHIFT) & A5XX_SP_VS_OUT_REG_A_REGID__MASK; +} +#define A5XX_SP_VS_OUT_REG_A_COMPMASK__MASK 0x00000f00 +#define A5XX_SP_VS_OUT_REG_A_COMPMASK__SHIFT 8 +static inline uint32_t A5XX_SP_VS_OUT_REG_A_COMPMASK(uint32_t val) +{ + return ((val) << A5XX_SP_VS_OUT_REG_A_COMPMASK__SHIFT) & A5XX_SP_VS_OUT_REG_A_COMPMASK__MASK; +} +#define A5XX_SP_VS_OUT_REG_B_REGID__MASK 0x00ff0000 +#define A5XX_SP_VS_OUT_REG_B_REGID__SHIFT 16 +static inline uint32_t A5XX_SP_VS_OUT_REG_B_REGID(uint32_t val) +{ + return ((val) << A5XX_SP_VS_OUT_REG_B_REGID__SHIFT) & A5XX_SP_VS_OUT_REG_B_REGID__MASK; +} +#define A5XX_SP_VS_OUT_REG_B_COMPMASK__MASK 0x0f000000 +#define A5XX_SP_VS_OUT_REG_B_COMPMASK__SHIFT 24 +static inline uint32_t A5XX_SP_VS_OUT_REG_B_COMPMASK(uint32_t val) +{ + return ((val) << A5XX_SP_VS_OUT_REG_B_COMPMASK__SHIFT) & A5XX_SP_VS_OUT_REG_B_COMPMASK__MASK; +} + +static inline uint32_t REG_A5XX_SP_VS_VPC_DST(uint32_t i0) { return 0x0000e5a3 + 0x1*i0; } + +static inline uint32_t REG_A5XX_SP_VS_VPC_DST_REG(uint32_t i0) { return 0x0000e5a3 + 0x1*i0; } +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC0__MASK 0x000000ff +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC0__SHIFT 0 +static inline uint32_t A5XX_SP_VS_VPC_DST_REG_OUTLOC0(uint32_t val) +{ + return ((val) << A5XX_SP_VS_VPC_DST_REG_OUTLOC0__SHIFT) & A5XX_SP_VS_VPC_DST_REG_OUTLOC0__MASK; +} +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC1__MASK 0x0000ff00 +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC1__SHIFT 8 +static inline uint32_t A5XX_SP_VS_VPC_DST_REG_OUTLOC1(uint32_t val) +{ + return ((val) << A5XX_SP_VS_VPC_DST_REG_OUTLOC1__SHIFT) & A5XX_SP_VS_VPC_DST_REG_OUTLOC1__MASK; +} +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC2__MASK 0x00ff0000 +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC2__SHIFT 16 +static inline uint32_t A5XX_SP_VS_VPC_DST_REG_OUTLOC2(uint32_t val) +{ + return ((val) << A5XX_SP_VS_VPC_DST_REG_OUTLOC2__SHIFT) & A5XX_SP_VS_VPC_DST_REG_OUTLOC2__MASK; +} +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC3__MASK 0xff000000 +#define A5XX_SP_VS_VPC_DST_REG_OUTLOC3__SHIFT 24 +static inline uint32_t A5XX_SP_VS_VPC_DST_REG_OUTLOC3(uint32_t val) +{ + return ((val) << A5XX_SP_VS_VPC_DST_REG_OUTLOC3__SHIFT) & A5XX_SP_VS_VPC_DST_REG_OUTLOC3__MASK; +} + +#define REG_A5XX_SP_VS_OBJ_START_LO 0x0000e5ac + +#define REG_A5XX_SP_VS_OBJ_START_HI 0x0000e5ad + +#define REG_A5XX_SP_FS_CTRL_REG0 0x0000e5c0 +#define A5XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__MASK 0x000003f0 +#define A5XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT 4 +static inline uint32_t A5XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT(uint32_t val) +{ + return ((val) << A5XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__SHIFT) & A5XX_SP_FS_CTRL_REG0_HALFREGFOOTPRINT__MASK; +} +#define A5XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__MASK 0x0000fc00 +#define A5XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT 10 +static inline uint32_t A5XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT(uint32_t val) +{ + return ((val) << A5XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__SHIFT) & A5XX_SP_FS_CTRL_REG0_FULLREGFOOTPRINT__MASK; +} +#define A5XX_SP_FS_CTRL_REG0_VARYING 0x00010000 +#define A5XX_SP_FS_CTRL_REG0_PIXLODENABLE 0x00100000 + +#define REG_A5XX_SP_FS_OBJ_START_LO 0x0000e5c3 + +#define REG_A5XX_SP_FS_OBJ_START_HI 0x0000e5c4 + +#define REG_A5XX_SP_BLEND_CNTL 0x0000e5c9 + +#define REG_A5XX_SP_FS_OUTPUT_CNTL 0x0000e5ca +#define A5XX_SP_FS_OUTPUT_CNTL_MRT__MASK 0x0000000f +#define A5XX_SP_FS_OUTPUT_CNTL_MRT__SHIFT 0 +static inline uint32_t A5XX_SP_FS_OUTPUT_CNTL_MRT(uint32_t val) +{ + return ((val) << A5XX_SP_FS_OUTPUT_CNTL_MRT__SHIFT) & A5XX_SP_FS_OUTPUT_CNTL_MRT__MASK; +} +#define A5XX_SP_FS_OUTPUT_CNTL_DEPTH_REGID__MASK 0x00001fe0 +#define A5XX_SP_FS_OUTPUT_CNTL_DEPTH_REGID__SHIFT 5 +static inline uint32_t A5XX_SP_FS_OUTPUT_CNTL_DEPTH_REGID(uint32_t val) +{ + return ((val) << A5XX_SP_FS_OUTPUT_CNTL_DEPTH_REGID__SHIFT) & A5XX_SP_FS_OUTPUT_CNTL_DEPTH_REGID__MASK; +} +#define A5XX_SP_FS_OUTPUT_CNTL_SAMPLEMASK_REGID__MASK 0x001fe000 +#define A5XX_SP_FS_OUTPUT_CNTL_SAMPLEMASK_REGID__SHIFT 13 +static inline uint32_t A5XX_SP_FS_OUTPUT_CNTL_SAMPLEMASK_REGID(uint32_t val) +{ + return ((val) << A5XX_SP_FS_OUTPUT_CNTL_SAMPLEMASK_REGID__SHIFT) & A5XX_SP_FS_OUTPUT_CNTL_SAMPLEMASK_REGID__MASK; +} + +static inline uint32_t REG_A5XX_SP_FS_OUTPUT(uint32_t i0) { return 0x0000e5cb + 0x1*i0; } + +static inline uint32_t REG_A5XX_SP_FS_OUTPUT_REG(uint32_t i0) { return 0x0000e5cb + 0x1*i0; } +#define A5XX_SP_FS_OUTPUT_REG_REGID__MASK 0x000000ff +#define A5XX_SP_FS_OUTPUT_REG_REGID__SHIFT 0 +static inline uint32_t A5XX_SP_FS_OUTPUT_REG_REGID(uint32_t val) +{ + return ((val) << A5XX_SP_FS_OUTPUT_REG_REGID__SHIFT) & A5XX_SP_FS_OUTPUT_REG_REGID__MASK; +} +#define A5XX_SP_FS_OUTPUT_REG_HALF_PRECISION 0x00000100 + +static inline uint32_t REG_A5XX_SP_FS_MRT(uint32_t i0) { return 0x0000e5d3 + 0x1*i0; } + +static inline uint32_t REG_A5XX_SP_FS_MRT_REG(uint32_t i0) { return 0x0000e5d3 + 0x1*i0; } +#define A5XX_SP_FS_MRT_REG_COLOR_FORMAT__MASK 0x0000007f +#define A5XX_SP_FS_MRT_REG_COLOR_FORMAT__SHIFT 0 +static inline uint32_t A5XX_SP_FS_MRT_REG_COLOR_FORMAT(enum a5xx_color_fmt val) +{ + return ((val) << A5XX_SP_FS_MRT_REG_COLOR_FORMAT__SHIFT) & A5XX_SP_FS_MRT_REG_COLOR_FORMAT__MASK; +} + +#define REG_A5XX_SP_CS_CNTL_0 0x0000e5f0 + +#define REG_A5XX_TPL1_TP_RAS_MSAA_CNTL 0x0000e704 + +#define REG_A5XX_TPL1_TP_DEST_MSAA_CNTL 0x0000e705 + +#define REG_A5XX_TPL1_VS_TEX_SAMP_LO 0x0000e722 + +#define REG_A5XX_TPL1_VS_TEX_SAMP_HI 0x0000e723 + +#define REG_A5XX_TPL1_VS_TEX_CONST_LO 0x0000e72a + +#define REG_A5XX_TPL1_VS_TEX_CONST_HI 0x0000e72b + +#define REG_A5XX_TPL1_FS_TEX_CONST_LO 0x0000e75a + +#define REG_A5XX_TPL1_FS_TEX_CONST_HI 0x0000e75b + +#define REG_A5XX_TPL1_FS_TEX_SAMP_LO 0x0000e75e + +#define REG_A5XX_TPL1_FS_TEX_SAMP_HI 0x0000e75f + +#define REG_A5XX_TPL1_TP_FS_ROTATION_CNTL 0x0000e764 + +#define REG_A5XX_HLSQ_CONTROL_0_REG 0x0000e784 + +#define REG_A5XX_HLSQ_CONTROL_1_REG 0x0000e785 + +#define REG_A5XX_HLSQ_CONTROL_2_REG 0x0000e786 +#define A5XX_HLSQ_CONTROL_2_REG_FACEREGID__MASK 0x000000ff +#define A5XX_HLSQ_CONTROL_2_REG_FACEREGID__SHIFT 0 +static inline uint32_t A5XX_HLSQ_CONTROL_2_REG_FACEREGID(uint32_t val) +{ + return ((val) << A5XX_HLSQ_CONTROL_2_REG_FACEREGID__SHIFT) & A5XX_HLSQ_CONTROL_2_REG_FACEREGID__MASK; +} + +#define REG_A5XX_HLSQ_CONTROL_3_REG 0x0000e787 +#define A5XX_HLSQ_CONTROL_3_REG_REGID__MASK 0x000000ff +#define A5XX_HLSQ_CONTROL_3_REG_REGID__SHIFT 0 +static inline uint32_t A5XX_HLSQ_CONTROL_3_REG_REGID(uint32_t val) +{ + return ((val) << A5XX_HLSQ_CONTROL_3_REG_REGID__SHIFT) & A5XX_HLSQ_CONTROL_3_REG_REGID__MASK; +} + +#define REG_A5XX_HLSQ_CONTROL_4_REG 0x0000e788 +#define A5XX_HLSQ_CONTROL_4_REG_XYCOORDREGID__MASK 0x00ff0000 +#define A5XX_HLSQ_CONTROL_4_REG_XYCOORDREGID__SHIFT 16 +static inline uint32_t A5XX_HLSQ_CONTROL_4_REG_XYCOORDREGID(uint32_t val) +{ + return ((val) << A5XX_HLSQ_CONTROL_4_REG_XYCOORDREGID__SHIFT) & A5XX_HLSQ_CONTROL_4_REG_XYCOORDREGID__MASK; +} +#define A5XX_HLSQ_CONTROL_4_REG_ZWCOORDREGID__MASK 0xff000000 +#define A5XX_HLSQ_CONTROL_4_REG_ZWCOORDREGID__SHIFT 24 +static inline uint32_t A5XX_HLSQ_CONTROL_4_REG_ZWCOORDREGID(uint32_t val) +{ + return ((val) << A5XX_HLSQ_CONTROL_4_REG_ZWCOORDREGID__SHIFT) & A5XX_HLSQ_CONTROL_4_REG_ZWCOORDREGID__MASK; +} + +#define REG_A5XX_HLSQ_UPDATE_CNTL 0x0000e78a + +#define REG_A5XX_HLSQ_VS_CONTROL_REG 0x0000e78b +#define A5XX_HLSQ_VS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_HLSQ_VS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_HLSQ_VS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_HLSQ_FS_CONTROL_REG 0x0000e78c +#define A5XX_HLSQ_FS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_HLSQ_FS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_HLSQ_FS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_HLSQ_HS_CONTROL_REG 0x0000e78d +#define A5XX_HLSQ_HS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_HLSQ_HS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_HLSQ_HS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_HLSQ_DS_CONTROL_REG 0x0000e78e +#define A5XX_HLSQ_DS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_HLSQ_DS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_HLSQ_DS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_HLSQ_GS_CONTROL_REG 0x0000e78f +#define A5XX_HLSQ_GS_CONTROL_REG_ENABLED 0x00000001 +#define A5XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__MASK 0x000000fe +#define A5XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT 1 +static inline uint32_t A5XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__SHIFT) & A5XX_HLSQ_GS_CONTROL_REG_CONSTOBJECTOFFSET__MASK; +} +#define A5XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__MASK 0x00007f00 +#define A5XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__SHIFT 8 +static inline uint32_t A5XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET(uint32_t val) +{ + return ((val) << A5XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__SHIFT) & A5XX_HLSQ_GS_CONTROL_REG_SHADEROBJOFFSET__MASK; +} + +#define REG_A5XX_HLSQ_CS_CONFIG 0x0000e790 + +#define REG_A5XX_HLSQ_VS_CNTL 0x0000e791 + +#define REG_A5XX_HLSQ_FS_CNTL 0x0000e792 + +#define REG_A5XX_HLSQ_CS_CNTL 0x0000e796 + +#define REG_A5XX_HLSQ_CS_KERNEL_GROUP_X 0x0000e7b9 + +#define REG_A5XX_HLSQ_CS_KERNEL_GROUP_Y 0x0000e7ba + +#define REG_A5XX_HLSQ_CS_KERNEL_GROUP_Z 0x0000e7bb + +#define REG_A5XX_HLSQ_CS_NDRANGE_0 0x0000e7b0 + +#define REG_A5XX_HLSQ_CS_NDRANGE_1 0x0000e7b1 + +#define REG_A5XX_HLSQ_CS_NDRANGE_2 0x0000e7b2 + +#define REG_A5XX_HLSQ_CS_NDRANGE_3 0x0000e7b3 + +#define REG_A5XX_HLSQ_CS_NDRANGE_4 0x0000e7b4 + +#define REG_A5XX_HLSQ_CS_NDRANGE_5 0x0000e7b5 + +#define REG_A5XX_HLSQ_CS_NDRANGE_6 0x0000e7b6 + +#define REG_A5XX_HLSQ_CS_CNTL_0 0x0000e7b7 + +#define REG_A5XX_HLSQ_CS_CNTL_1 0x0000e7b8 + +#define REG_A5XX_HLSQ_VS_CONSTLEN 0x0000e7c3 + +#define REG_A5XX_HLSQ_VS_INSTRLEN 0x0000e7c4 + +#define REG_A5XX_HLSQ_FS_CONSTLEN 0x0000e7d7 + +#define REG_A5XX_HLSQ_FS_INSTRLEN 0x0000e7d8 + +#define REG_A5XX_HLSQ_CONTEXT_SWITCH_CS_SW_3 0x0000e7dc + +#define REG_A5XX_HLSQ_CONTEXT_SWITCH_CS_SW_4 0x0000e7dd + +#define REG_A5XX_TEX_SAMP_0 0x00000000 +#define A5XX_TEX_SAMP_0_MIPFILTER_LINEAR_NEAR 0x00000001 +#define A5XX_TEX_SAMP_0_XY_MAG__MASK 0x00000006 +#define A5XX_TEX_SAMP_0_XY_MAG__SHIFT 1 +static inline uint32_t A5XX_TEX_SAMP_0_XY_MAG(enum a5xx_tex_filter val) +{ + return ((val) << A5XX_TEX_SAMP_0_XY_MAG__SHIFT) & A5XX_TEX_SAMP_0_XY_MAG__MASK; +} +#define A5XX_TEX_SAMP_0_XY_MIN__MASK 0x00000018 +#define A5XX_TEX_SAMP_0_XY_MIN__SHIFT 3 +static inline uint32_t A5XX_TEX_SAMP_0_XY_MIN(enum a5xx_tex_filter val) +{ + return ((val) << A5XX_TEX_SAMP_0_XY_MIN__SHIFT) & A5XX_TEX_SAMP_0_XY_MIN__MASK; +} +#define A5XX_TEX_SAMP_0_WRAP_S__MASK 0x000000e0 +#define A5XX_TEX_SAMP_0_WRAP_S__SHIFT 5 +static inline uint32_t A5XX_TEX_SAMP_0_WRAP_S(enum a5xx_tex_clamp val) +{ + return ((val) << A5XX_TEX_SAMP_0_WRAP_S__SHIFT) & A5XX_TEX_SAMP_0_WRAP_S__MASK; +} +#define A5XX_TEX_SAMP_0_WRAP_T__MASK 0x00000700 +#define A5XX_TEX_SAMP_0_WRAP_T__SHIFT 8 +static inline uint32_t A5XX_TEX_SAMP_0_WRAP_T(enum a5xx_tex_clamp val) +{ + return ((val) << A5XX_TEX_SAMP_0_WRAP_T__SHIFT) & A5XX_TEX_SAMP_0_WRAP_T__MASK; +} +#define A5XX_TEX_SAMP_0_WRAP_R__MASK 0x00003800 +#define A5XX_TEX_SAMP_0_WRAP_R__SHIFT 11 +static inline uint32_t A5XX_TEX_SAMP_0_WRAP_R(enum a5xx_tex_clamp val) +{ + return ((val) << A5XX_TEX_SAMP_0_WRAP_R__SHIFT) & A5XX_TEX_SAMP_0_WRAP_R__MASK; +} +#define A5XX_TEX_SAMP_0_ANISO__MASK 0x0001c000 +#define A5XX_TEX_SAMP_0_ANISO__SHIFT 14 +static inline uint32_t A5XX_TEX_SAMP_0_ANISO(enum a5xx_tex_aniso val) +{ + return ((val) << A5XX_TEX_SAMP_0_ANISO__SHIFT) & A5XX_TEX_SAMP_0_ANISO__MASK; +} +#define A5XX_TEX_SAMP_0_LOD_BIAS__MASK 0xfff80000 +#define A5XX_TEX_SAMP_0_LOD_BIAS__SHIFT 19 +static inline uint32_t A5XX_TEX_SAMP_0_LOD_BIAS(float val) +{ + return ((((int32_t)(val * 256.0))) << A5XX_TEX_SAMP_0_LOD_BIAS__SHIFT) & A5XX_TEX_SAMP_0_LOD_BIAS__MASK; +} + +#define REG_A5XX_TEX_SAMP_1 0x00000001 +#define A5XX_TEX_SAMP_1_COMPARE_FUNC__MASK 0x0000000e +#define A5XX_TEX_SAMP_1_COMPARE_FUNC__SHIFT 1 +static inline uint32_t A5XX_TEX_SAMP_1_COMPARE_FUNC(enum adreno_compare_func val) +{ + return ((val) << A5XX_TEX_SAMP_1_COMPARE_FUNC__SHIFT) & A5XX_TEX_SAMP_1_COMPARE_FUNC__MASK; +} +#define A5XX_TEX_SAMP_1_CUBEMAPSEAMLESSFILTOFF 0x00000010 +#define A5XX_TEX_SAMP_1_UNNORM_COORDS 0x00000020 +#define A5XX_TEX_SAMP_1_MIPFILTER_LINEAR_FAR 0x00000040 +#define A5XX_TEX_SAMP_1_MAX_LOD__MASK 0x000fff00 +#define A5XX_TEX_SAMP_1_MAX_LOD__SHIFT 8 +static inline uint32_t A5XX_TEX_SAMP_1_MAX_LOD(float val) +{ + return ((((uint32_t)(val * 256.0))) << A5XX_TEX_SAMP_1_MAX_LOD__SHIFT) & A5XX_TEX_SAMP_1_MAX_LOD__MASK; +} +#define A5XX_TEX_SAMP_1_MIN_LOD__MASK 0xfff00000 +#define A5XX_TEX_SAMP_1_MIN_LOD__SHIFT 20 +static inline uint32_t A5XX_TEX_SAMP_1_MIN_LOD(float val) +{ + return ((((uint32_t)(val * 256.0))) << A5XX_TEX_SAMP_1_MIN_LOD__SHIFT) & A5XX_TEX_SAMP_1_MIN_LOD__MASK; +} + +#define REG_A5XX_TEX_SAMP_2 0x00000002 + +#define REG_A5XX_TEX_SAMP_3 0x00000003 + +#define REG_A5XX_TEX_CONST_0 0x00000000 +#define A5XX_TEX_CONST_0_TILED 0x00000001 +#define A5XX_TEX_CONST_0_SRGB 0x00000004 +#define A5XX_TEX_CONST_0_SWIZ_X__MASK 0x00000070 +#define A5XX_TEX_CONST_0_SWIZ_X__SHIFT 4 +static inline uint32_t A5XX_TEX_CONST_0_SWIZ_X(enum a5xx_tex_swiz val) +{ + return ((val) << A5XX_TEX_CONST_0_SWIZ_X__SHIFT) & A5XX_TEX_CONST_0_SWIZ_X__MASK; +} +#define A5XX_TEX_CONST_0_SWIZ_Y__MASK 0x00000380 +#define A5XX_TEX_CONST_0_SWIZ_Y__SHIFT 7 +static inline uint32_t A5XX_TEX_CONST_0_SWIZ_Y(enum a5xx_tex_swiz val) +{ + return ((val) << A5XX_TEX_CONST_0_SWIZ_Y__SHIFT) & A5XX_TEX_CONST_0_SWIZ_Y__MASK; +} +#define A5XX_TEX_CONST_0_SWIZ_Z__MASK 0x00001c00 +#define A5XX_TEX_CONST_0_SWIZ_Z__SHIFT 10 +static inline uint32_t A5XX_TEX_CONST_0_SWIZ_Z(enum a5xx_tex_swiz val) +{ + return ((val) << A5XX_TEX_CONST_0_SWIZ_Z__SHIFT) & A5XX_TEX_CONST_0_SWIZ_Z__MASK; +} +#define A5XX_TEX_CONST_0_SWIZ_W__MASK 0x0000e000 +#define A5XX_TEX_CONST_0_SWIZ_W__SHIFT 13 +static inline uint32_t A5XX_TEX_CONST_0_SWIZ_W(enum a5xx_tex_swiz val) +{ + return ((val) << A5XX_TEX_CONST_0_SWIZ_W__SHIFT) & A5XX_TEX_CONST_0_SWIZ_W__MASK; +} +#define A5XX_TEX_CONST_0_FMT__MASK 0x3fc00000 +#define A5XX_TEX_CONST_0_FMT__SHIFT 22 +static inline uint32_t A5XX_TEX_CONST_0_FMT(enum a5xx_tex_fmt val) +{ + return ((val) << A5XX_TEX_CONST_0_FMT__SHIFT) & A5XX_TEX_CONST_0_FMT__MASK; +} + +#define REG_A5XX_TEX_CONST_1 0x00000001 +#define A5XX_TEX_CONST_1_WIDTH__MASK 0x00007fff +#define A5XX_TEX_CONST_1_WIDTH__SHIFT 0 +static inline uint32_t A5XX_TEX_CONST_1_WIDTH(uint32_t val) +{ + return ((val) << A5XX_TEX_CONST_1_WIDTH__SHIFT) & A5XX_TEX_CONST_1_WIDTH__MASK; +} +#define A5XX_TEX_CONST_1_HEIGHT__MASK 0x3fff8000 +#define A5XX_TEX_CONST_1_HEIGHT__SHIFT 15 +static inline uint32_t A5XX_TEX_CONST_1_HEIGHT(uint32_t val) +{ + return ((val) << A5XX_TEX_CONST_1_HEIGHT__SHIFT) & A5XX_TEX_CONST_1_HEIGHT__MASK; +} + +#define REG_A5XX_TEX_CONST_2 0x00000002 +#define A5XX_TEX_CONST_2_FETCHSIZE__MASK 0x0000000f +#define A5XX_TEX_CONST_2_FETCHSIZE__SHIFT 0 +static inline uint32_t A5XX_TEX_CONST_2_FETCHSIZE(enum a5xx_tex_fetchsize val) +{ + return ((val) << A5XX_TEX_CONST_2_FETCHSIZE__SHIFT) & A5XX_TEX_CONST_2_FETCHSIZE__MASK; +} +#define A5XX_TEX_CONST_2_PITCH__MASK 0x1fffff00 +#define A5XX_TEX_CONST_2_PITCH__SHIFT 8 +static inline uint32_t A5XX_TEX_CONST_2_PITCH(uint32_t val) +{ + return ((val) << A5XX_TEX_CONST_2_PITCH__SHIFT) & A5XX_TEX_CONST_2_PITCH__MASK; +} +#define A5XX_TEX_CONST_2_TYPE__MASK 0x60000000 +#define A5XX_TEX_CONST_2_TYPE__SHIFT 29 +static inline uint32_t A5XX_TEX_CONST_2_TYPE(enum a5xx_tex_type val) +{ + return ((val) << A5XX_TEX_CONST_2_TYPE__SHIFT) & A5XX_TEX_CONST_2_TYPE__MASK; +} + +#define REG_A5XX_TEX_CONST_3 0x00000003 +#define A5XX_TEX_CONST_3_LAYERSZ__MASK 0x00003fff +#define A5XX_TEX_CONST_3_LAYERSZ__SHIFT 0 +static inline uint32_t A5XX_TEX_CONST_3_LAYERSZ(uint32_t val) +{ + return ((val >> 12) << A5XX_TEX_CONST_3_LAYERSZ__SHIFT) & A5XX_TEX_CONST_3_LAYERSZ__MASK; +} +#define A5XX_TEX_CONST_3_LAYERSZ2__MASK 0xff800000 +#define A5XX_TEX_CONST_3_LAYERSZ2__SHIFT 23 +static inline uint32_t A5XX_TEX_CONST_3_LAYERSZ2(uint32_t val) +{ + return ((val >> 12) << A5XX_TEX_CONST_3_LAYERSZ2__SHIFT) & A5XX_TEX_CONST_3_LAYERSZ2__MASK; +} + +#define REG_A5XX_TEX_CONST_4 0x00000004 +#define A5XX_TEX_CONST_4_BASE__MASK 0xffffffe0 +#define A5XX_TEX_CONST_4_BASE__SHIFT 5 +static inline uint32_t A5XX_TEX_CONST_4_BASE(uint32_t val) +{ + return ((val >> 5) << A5XX_TEX_CONST_4_BASE__SHIFT) & A5XX_TEX_CONST_4_BASE__MASK; +} + +#define REG_A5XX_TEX_CONST_5 0x00000005 +#define A5XX_TEX_CONST_5_DEPTH__MASK 0x3ffe0000 +#define A5XX_TEX_CONST_5_DEPTH__SHIFT 17 +static inline uint32_t A5XX_TEX_CONST_5_DEPTH(uint32_t val) +{ + return ((val) << A5XX_TEX_CONST_5_DEPTH__SHIFT) & A5XX_TEX_CONST_5_DEPTH__MASK; +} + +#define REG_A5XX_TEX_CONST_6 0x00000006 + +#define REG_A5XX_TEX_CONST_7 0x00000007 + +#define REG_A5XX_TEX_CONST_8 0x00000008 + +#define REG_A5XX_TEX_CONST_9 0x00000009 + +#define REG_A5XX_TEX_CONST_10 0x0000000a + +#define REG_A5XX_TEX_CONST_11 0x0000000b + + +#endif /* A5XX_XML */ diff --git a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h index e81481d1b7df..1cf84479e447 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h +++ b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h @@ -8,14 +8,15 @@ http://github.com/freedreno/envytools/ git clone https://github.com/freedreno/envytools.git The rules-ng-ng source files this header was generated from are: -- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml ( 398 bytes, from 2015-09-24 17:25:31) -- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml ( 1572 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml ( 32901 bytes, from 2015-05-20 20:03:14) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml ( 11518 bytes, from 2016-02-10 21:03:25) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml ( 16166 bytes, from 2016-02-11 21:20:31) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml ( 83967 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml ( 109916 bytes, from 2016-02-20 18:44:48) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/ocmem.xml ( 1773 bytes, from 2015-09-24 17:30:00) +- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) +- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) Copyright (C) 2013-2016 by the following authors: - Rob Clark (robclark) @@ -127,11 +128,13 @@ enum a3xx_rop_code { ROP_COPY_INVERTED = 3, ROP_AND_REVERSE = 4, ROP_INVERT = 5, + ROP_XOR = 6, ROP_NAND = 7, ROP_AND = 8, ROP_EQUIV = 9, ROP_NOOP = 10, ROP_OR_INVERTED = 11, + ROP_COPY = 12, ROP_OR_REVERSE = 13, ROP_OR = 14, ROP_SET = 15, @@ -172,6 +175,14 @@ enum a3xx_color_swap { XYZW = 3, }; +enum a3xx_rb_blend_opcode { + BLEND_DST_PLUS_SRC = 0, + BLEND_SRC_MINUS_DST = 1, + BLEND_DST_MINUS_SRC = 2, + BLEND_MIN_DST_SRC = 3, + BLEND_MAX_DST_SRC = 4, +}; + #define REG_AXXX_CP_RB_BASE 0x000001c0 #define REG_AXXX_CP_RB_CNTL 0x000001c1 diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index aea39d357dc0..e0647121a41f 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -24,6 +24,10 @@ #include "msm_gpu.h" +/* arrg, somehow fb.h is getting pulled in: */ +#undef ROP_COPY +#undef ROP_XOR + #include "adreno_common.xml.h" #include "adreno_pm4.xml.h" diff --git a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h index d7477ff867c9..9911a181f9c2 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h +++ b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h @@ -8,14 +8,15 @@ http://github.com/freedreno/envytools/ git clone https://github.com/freedreno/envytools.git The rules-ng-ng source files this header was generated from are: -- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml ( 398 bytes, from 2015-09-24 17:25:31) -- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml ( 1572 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml ( 32901 bytes, from 2015-05-20 20:03:14) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml ( 11518 bytes, from 2016-02-10 21:03:25) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml ( 16166 bytes, from 2016-02-11 21:20:31) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml ( 83967 bytes, from 2016-02-10 17:07:21) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml ( 109916 bytes, from 2016-02-20 18:44:48) -- /home/robclark/src/freedreno/envytools/rnndb/adreno/ocmem.xml ( 1773 bytes, from 2015-09-24 17:30:00) +- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) +- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) Copyright (C) 2013-2016 by the following authors: - Rob Clark (robclark) @@ -58,6 +59,7 @@ enum vgt_event_type { RST_PIX_CNT = 13, RST_VTX_CNT = 14, TILE_FLUSH = 15, + STAT_EVENT = 16, CACHE_FLUSH_AND_INV_TS_EVENT = 20, ZPASS_DONE = 21, CACHE_FLUSH_AND_INV_EVENT = 22, @@ -82,7 +84,6 @@ enum pc_di_primtype { DI_PT_LINESTRIP_ADJ = 11, DI_PT_TRI_ADJ = 12, DI_PT_TRISTRIP_ADJ = 13, - DI_PT_PATCHES = 34, }; enum pc_di_src_sel { @@ -110,11 +111,15 @@ enum adreno_pm4_packet_type { CP_TYPE1_PKT = 0x40000000, CP_TYPE2_PKT = 0x80000000, CP_TYPE3_PKT = 0xc0000000, + CP_TYPE4_PKT = 0x40000000, + CP_TYPE7_PKT = 0x70000000, }; enum adreno_pm4_type3_packets { CP_ME_INIT = 72, CP_NOP = 16, + CP_PREEMPT_ENABLE = 28, + CP_PREEMPT_TOKEN = 30, CP_INDIRECT_BUFFER = 63, CP_INDIRECT_BUFFER_PFD = 55, CP_WAIT_FOR_IDLE = 38, @@ -163,6 +168,7 @@ enum adreno_pm4_type3_packets { CP_TEST_TWO_MEMS = 113, CP_REG_WR_NO_CTXT = 120, CP_RECORD_PFP_TIMESTAMP = 17, + CP_SET_SECURE_MODE = 102, CP_WAIT_FOR_ME = 19, CP_SET_DRAW_STATE = 67, CP_DRAW_INDX_OFFSET = 56, @@ -178,6 +184,21 @@ enum adreno_pm4_type3_packets { CP_WAIT_MEM_WRITES = 18, CP_COND_REG_EXEC = 71, CP_MEM_TO_REG = 66, + CP_EXEC_CS = 51, + CP_PERFCOUNTER_ACTION = 80, + CP_SMMU_TABLE_UPDATE = 83, + CP_CONTEXT_REG_BUNCH = 92, + CP_YIELD_ENABLE = 28, + CP_SKIP_IB2_ENABLE_GLOBAL = 29, + CP_SKIP_IB2_ENABLE_LOCAL = 35, + CP_SET_SUBDRAW_SIZE = 53, + CP_SET_VISIBILITY_OVERRIDE = 100, + CP_PREEMPT_ENABLE_GLOBAL = 105, + CP_PREEMPT_ENABLE_LOCAL = 106, + CP_CONTEXT_SWITCH_YIELD = 107, + CP_SET_RENDER_MODE = 108, + CP_COMPUTE_CHECKPOINT = 110, + CP_MEM_TO_MEM = 115, IN_IB_PREFETCH_END = 23, IN_SUBBLK_PREFETCH = 31, IN_INSTR_PREFETCH = 32, @@ -196,6 +217,7 @@ enum adreno_state_block { SB_VERT_SHADER = 4, SB_GEOM_SHADER = 5, SB_FRAG_SHADER = 6, + SB_COMPUTE_SHADER = 7, }; enum adreno_state_type { @@ -389,7 +411,12 @@ static inline uint32_t CP_DRAW_INDX_OFFSET_0_SOURCE_SELECT(enum pc_di_src_sel va { return ((val) << CP_DRAW_INDX_OFFSET_0_SOURCE_SELECT__SHIFT) & CP_DRAW_INDX_OFFSET_0_SOURCE_SELECT__MASK; } -#define CP_DRAW_INDX_OFFSET_0_TESSELLATE 0x00000100 +#define CP_DRAW_INDX_OFFSET_0_VIS_CULL__MASK 0x00000300 +#define CP_DRAW_INDX_OFFSET_0_VIS_CULL__SHIFT 8 +static inline uint32_t CP_DRAW_INDX_OFFSET_0_VIS_CULL(enum pc_di_vis_cull_mode val) +{ + return ((val) << CP_DRAW_INDX_OFFSET_0_VIS_CULL__SHIFT) & CP_DRAW_INDX_OFFSET_0_VIS_CULL__MASK; +} #define CP_DRAW_INDX_OFFSET_0_INDEX_SIZE__MASK 0x00000c00 #define CP_DRAW_INDX_OFFSET_0_INDEX_SIZE__SHIFT 10 static inline uint32_t CP_DRAW_INDX_OFFSET_0_INDEX_SIZE(enum a4xx_index_size val) @@ -533,5 +560,78 @@ static inline uint32_t CP_REG_TO_MEM_1_DEST(uint32_t val) return ((val) << CP_REG_TO_MEM_1_DEST__SHIFT) & CP_REG_TO_MEM_1_DEST__MASK; } +#define REG_CP_DISPATCH_COMPUTE_0 0x00000000 + +#define REG_CP_DISPATCH_COMPUTE_1 0x00000001 +#define CP_DISPATCH_COMPUTE_1_X__MASK 0xffffffff +#define CP_DISPATCH_COMPUTE_1_X__SHIFT 0 +static inline uint32_t CP_DISPATCH_COMPUTE_1_X(uint32_t val) +{ + return ((val) << CP_DISPATCH_COMPUTE_1_X__SHIFT) & CP_DISPATCH_COMPUTE_1_X__MASK; +} + +#define REG_CP_DISPATCH_COMPUTE_2 0x00000002 +#define CP_DISPATCH_COMPUTE_2_Y__MASK 0xffffffff +#define CP_DISPATCH_COMPUTE_2_Y__SHIFT 0 +static inline uint32_t CP_DISPATCH_COMPUTE_2_Y(uint32_t val) +{ + return ((val) << CP_DISPATCH_COMPUTE_2_Y__SHIFT) & CP_DISPATCH_COMPUTE_2_Y__MASK; +} + +#define REG_CP_DISPATCH_COMPUTE_3 0x00000003 +#define CP_DISPATCH_COMPUTE_3_Z__MASK 0xffffffff +#define CP_DISPATCH_COMPUTE_3_Z__SHIFT 0 +static inline uint32_t CP_DISPATCH_COMPUTE_3_Z(uint32_t val) +{ + return ((val) << CP_DISPATCH_COMPUTE_3_Z__SHIFT) & CP_DISPATCH_COMPUTE_3_Z__MASK; +} + +#define REG_CP_SET_RENDER_MODE_0 0x00000000 + +#define REG_CP_SET_RENDER_MODE_1 0x00000001 +#define CP_SET_RENDER_MODE_1_ADDR_0_LO__MASK 0xffffffff +#define CP_SET_RENDER_MODE_1_ADDR_0_LO__SHIFT 0 +static inline uint32_t CP_SET_RENDER_MODE_1_ADDR_0_LO(uint32_t val) +{ + return ((val) << CP_SET_RENDER_MODE_1_ADDR_0_LO__SHIFT) & CP_SET_RENDER_MODE_1_ADDR_0_LO__MASK; +} + +#define REG_CP_SET_RENDER_MODE_2 0x00000002 +#define CP_SET_RENDER_MODE_2_ADDR_0_HI__MASK 0xffffffff +#define CP_SET_RENDER_MODE_2_ADDR_0_HI__SHIFT 0 +static inline uint32_t CP_SET_RENDER_MODE_2_ADDR_0_HI(uint32_t val) +{ + return ((val) << CP_SET_RENDER_MODE_2_ADDR_0_HI__SHIFT) & CP_SET_RENDER_MODE_2_ADDR_0_HI__MASK; +} + +#define REG_CP_SET_RENDER_MODE_3 0x00000003 +#define CP_SET_RENDER_MODE_3_GMEM_ENABLE 0x00000010 + +#define REG_CP_SET_RENDER_MODE_4 0x00000004 + +#define REG_CP_SET_RENDER_MODE_5 0x00000005 +#define CP_SET_RENDER_MODE_5_ADDR_1_LEN__MASK 0xffffffff +#define CP_SET_RENDER_MODE_5_ADDR_1_LEN__SHIFT 0 +static inline uint32_t CP_SET_RENDER_MODE_5_ADDR_1_LEN(uint32_t val) +{ + return ((val) << CP_SET_RENDER_MODE_5_ADDR_1_LEN__SHIFT) & CP_SET_RENDER_MODE_5_ADDR_1_LEN__MASK; +} + +#define REG_CP_SET_RENDER_MODE_6 0x00000006 +#define CP_SET_RENDER_MODE_6_ADDR_1_LO__MASK 0xffffffff +#define CP_SET_RENDER_MODE_6_ADDR_1_LO__SHIFT 0 +static inline uint32_t CP_SET_RENDER_MODE_6_ADDR_1_LO(uint32_t val) +{ + return ((val) << CP_SET_RENDER_MODE_6_ADDR_1_LO__SHIFT) & CP_SET_RENDER_MODE_6_ADDR_1_LO__MASK; +} + +#define REG_CP_SET_RENDER_MODE_7 0x00000007 +#define CP_SET_RENDER_MODE_7_ADDR_1_HI__MASK 0xffffffff +#define CP_SET_RENDER_MODE_7_ADDR_1_HI__SHIFT 0 +static inline uint32_t CP_SET_RENDER_MODE_7_ADDR_1_HI(uint32_t val) +{ + return ((val) << CP_SET_RENDER_MODE_7_ADDR_1_HI__SHIFT) & CP_SET_RENDER_MODE_7_ADDR_1_HI__MASK; +} + #endif /* ADRENO_PM4_XML */ From 97fc325a319f25b4e34435aff6262c987fd79ba7 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:13 -0700 Subject: [PATCH 02/19] drm/msm: gpu: Add OUT_TYPE4 and OUT_TYPE7 Add helper functions for TYPE4 and TYPE7 ME opcodes that replace TYPE0 and TYPE3 starting on the A5XX targets. Change-Id: Ic0dedbad114e28bdbcba55a788c6307b48e14675 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/adreno_gpu.h | 32 ++++++++++++++++++++++++- 1 file changed, 31 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index e0647121a41f..3915c737cdd6 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -2,7 +2,7 @@ * Copyright (C) 2013 Red Hat * Author: Rob Clark * - * Copyright (c) 2014 The Linux Foundation. All rights reserved. + * Copyright (c) 2014,2016 The Linux Foundation. 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 version 2 as published by @@ -282,6 +282,36 @@ OUT_PKT3(struct msm_ringbuffer *ring, uint8_t opcode, uint16_t cnt) OUT_RING(ring, CP_TYPE3_PKT | ((cnt-1) << 16) | ((opcode & 0xFF) << 8)); } +static inline u32 PM4_PARITY(u32 val) +{ + return (0x9669 >> (0xF & (val ^ + (val >> 4) ^ (val >> 8) ^ (val >> 12) ^ + (val >> 16) ^ ((val) >> 20) ^ (val >> 24) ^ + (val >> 28)))) & 1; +} + +/* Maximum number of values that can be executed for one opcode */ +#define TYPE4_MAX_PAYLOAD 127 + +#define PKT4(_reg, _cnt) \ + (CP_TYPE4_PKT | ((_cnt) << 0) | (PM4_PARITY((_cnt)) << 7) | \ + (((_reg) & 0x3FFFF) << 8) | (PM4_PARITY((_reg)) << 27)) + +static inline void +OUT_PKT4(struct msm_ringbuffer *ring, uint16_t regindx, uint16_t cnt) +{ + adreno_wait_ring(ring->gpu, cnt + 1); + OUT_RING(ring, PKT4(regindx, cnt)); +} + +static inline void +OUT_PKT7(struct msm_ringbuffer *ring, uint8_t opcode, uint16_t cnt) +{ + adreno_wait_ring(ring->gpu, cnt + 1); + OUT_RING(ring, CP_TYPE7_PKT | (cnt << 0) | (PM4_PARITY(cnt) << 15) | + ((opcode & 0x7F) << 16) | (PM4_PARITY(opcode) << 23)); +} + /* * adreno_checkreg_off() - Checks the validity of a register enum * @gpu: Pointer to struct adreno_gpu From 6740cb63dd5e53a89477026c95a883bbcdbb9ada Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:14 -0700 Subject: [PATCH 03/19] drm/msm: gpu: Cut down the list of "generic" registers to the ones we use There are very few register accesses in the common code. Cut down the list of common registers to just those that are used. This saves const space and saves us the effort of maintaining registers for A3XX and A4XX that don't exist or are unused. Change-Id: Ic0dedbadb4dccbba284e9badf2f52f3a72594581 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a3xx_gpu.c | 82 +------------------------ drivers/gpu/drm/msm/adreno/a4xx_gpu.c | 78 +---------------------- drivers/gpu/drm/msm/adreno/adreno_gpu.h | 76 +++++------------------ 3 files changed, 20 insertions(+), 216 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c index f3a959eedbe0..5a061ad6225a 100644 --- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c @@ -422,91 +422,13 @@ static void a3xx_dump(struct msm_gpu *gpu) } /* Register offset defines for A3XX */ static const unsigned int a3xx_register_offsets[REG_ADRENO_REGISTER_MAX] = { - REG_ADRENO_DEFINE(REG_ADRENO_CP_DEBUG, REG_AXXX_CP_DEBUG), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_WADDR, REG_AXXX_CP_ME_RAM_WADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_DATA, REG_AXXX_CP_ME_RAM_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_DATA, - REG_A3XX_CP_PFP_UCODE_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_ADDR, - REG_A3XX_CP_PFP_UCODE_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_WFI_PEND_CTR, REG_A3XX_CP_WFI_PEND_CTR), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE, REG_AXXX_CP_RB_BASE), + REG_ADRENO_SKIP(REG_ADRENO_CP_RB_BASE_HI), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR_ADDR, REG_AXXX_CP_RB_RPTR_ADDR), + REG_ADRENO_SKIP(REG_ADRENO_CP_RB_RPTR_ADDR_HI), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR, REG_AXXX_CP_RB_RPTR), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_WPTR, REG_AXXX_CP_RB_WPTR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_CTRL, REG_A3XX_CP_PROTECT_CTRL), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_CNTL, REG_AXXX_CP_ME_CNTL), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_CNTL, REG_AXXX_CP_RB_CNTL), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BASE, REG_AXXX_CP_IB1_BASE), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BUFSZ, REG_AXXX_CP_IB1_BUFSZ), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BASE, REG_AXXX_CP_IB2_BASE), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BUFSZ, REG_AXXX_CP_IB2_BUFSZ), - REG_ADRENO_DEFINE(REG_ADRENO_CP_TIMESTAMP, REG_AXXX_CP_SCRATCH_REG0), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_RADDR, REG_AXXX_CP_ME_RAM_RADDR), - REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_ADDR, REG_AXXX_SCRATCH_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_UMSK, REG_AXXX_SCRATCH_UMSK), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_ADDR, REG_A3XX_CP_ROQ_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_DATA, REG_A3XX_CP_ROQ_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_ADDR, REG_A3XX_CP_MERCIU_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA, REG_A3XX_CP_MERCIU_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA2, REG_A3XX_CP_MERCIU_DATA2), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_ADDR, REG_A3XX_CP_MEQ_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_DATA, REG_A3XX_CP_MEQ_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_HW_FAULT, REG_A3XX_CP_HW_FAULT), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_STATUS, - REG_A3XX_CP_PROTECT_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_STATUS, REG_A3XX_RBBM_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_CTL, - REG_A3XX_RBBM_PERFCTR_CTL), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD0, - REG_A3XX_RBBM_PERFCTR_LOAD_CMD0), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD1, - REG_A3XX_RBBM_PERFCTR_LOAD_CMD1), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_PWR_1_LO, - REG_A3XX_RBBM_PERFCTR_PWR_1_LO), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_MASK, REG_A3XX_RBBM_INT_0_MASK), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_STATUS, - REG_A3XX_RBBM_INT_0_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_ERROR_STATUS, - REG_A3XX_RBBM_AHB_ERROR_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_CMD, REG_A3XX_RBBM_AHB_CMD), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_CLEAR_CMD, - REG_A3XX_RBBM_INT_CLEAR_CMD), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_CLOCK_CTL, REG_A3XX_RBBM_CLOCK_CTL), - REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_SEL, - REG_A3XX_VPC_VPC_DEBUG_RAM_SEL), - REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_READ, - REG_A3XX_VPC_VPC_DEBUG_RAM_READ), - REG_ADRENO_DEFINE(REG_ADRENO_VSC_SIZE_ADDRESS, - REG_A3XX_VSC_SIZE_ADDRESS), - REG_ADRENO_DEFINE(REG_ADRENO_VFD_CONTROL_0, REG_A3XX_VFD_CONTROL_0), - REG_ADRENO_DEFINE(REG_ADRENO_VFD_INDEX_MAX, REG_A3XX_VFD_INDEX_MAX), - REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_PVT_MEM_ADDR_REG, - REG_A3XX_SP_VS_PVT_MEM_ADDR_REG), - REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_PVT_MEM_ADDR_REG, - REG_A3XX_SP_FS_PVT_MEM_ADDR_REG), - REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_OBJ_START_REG, - REG_A3XX_SP_VS_OBJ_START_REG), - REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_OBJ_START_REG, - REG_A3XX_SP_FS_OBJ_START_REG), - REG_ADRENO_DEFINE(REG_ADRENO_PA_SC_AA_CONFIG, REG_A3XX_PA_SC_AA_CONFIG), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PM_OVERRIDE2, - REG_A3XX_RBBM_PM_OVERRIDE2), - REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_REG2, REG_AXXX_CP_SCRATCH_REG2), - REG_ADRENO_DEFINE(REG_ADRENO_SQ_GPR_MANAGEMENT, - REG_A3XX_SQ_GPR_MANAGEMENT), - REG_ADRENO_DEFINE(REG_ADRENO_SQ_INST_STORE_MANAGMENT, - REG_A3XX_SQ_INST_STORE_MANAGMENT), - REG_ADRENO_DEFINE(REG_ADRENO_TP0_CHICKEN, REG_A3XX_TP0_CHICKEN), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_RBBM_CTL, REG_A3XX_RBBM_RBBM_CTL), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_SW_RESET_CMD, - REG_A3XX_RBBM_SW_RESET_CMD), - REG_ADRENO_DEFINE(REG_ADRENO_UCHE_INVALIDATE0, - REG_A3XX_UCHE_CACHE_INVALIDATE0_REG), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_LO, - REG_A3XX_RBBM_PERFCTR_LOAD_VALUE_LO), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_HI, - REG_A3XX_RBBM_PERFCTR_LOAD_VALUE_HI), }; static const struct adreno_gpu_funcs funcs = { diff --git a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c index f430b291916d..dde9dc1333e1 100644 --- a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c @@ -455,87 +455,13 @@ static void a4xx_show(struct msm_gpu *gpu, struct seq_file *m) /* Register offset defines for A4XX, in order of enum adreno_regs */ static const unsigned int a4xx_register_offsets[REG_ADRENO_REGISTER_MAX] = { - REG_ADRENO_DEFINE(REG_ADRENO_CP_DEBUG, REG_A4XX_CP_DEBUG), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_WADDR, REG_A4XX_CP_ME_RAM_WADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_DATA, REG_A4XX_CP_ME_RAM_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_DATA, - REG_A4XX_CP_PFP_UCODE_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PFP_UCODE_ADDR, - REG_A4XX_CP_PFP_UCODE_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_WFI_PEND_CTR, REG_A4XX_CP_WFI_PEND_CTR), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE, REG_A4XX_CP_RB_BASE), + REG_ADRENO_SKIP(REG_ADRENO_CP_RB_BASE_HI), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR_ADDR, REG_A4XX_CP_RB_RPTR_ADDR), + REG_ADRENO_SKIP(REG_ADRENO_CP_RB_RPTR_ADDR_HI), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR, REG_A4XX_CP_RB_RPTR), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_WPTR, REG_A4XX_CP_RB_WPTR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_CTRL, REG_A4XX_CP_PROTECT_CTRL), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_CNTL, REG_A4XX_CP_ME_CNTL), REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_CNTL, REG_A4XX_CP_RB_CNTL), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BASE, REG_A4XX_CP_IB1_BASE), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB1_BUFSZ, REG_A4XX_CP_IB1_BUFSZ), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BASE, REG_A4XX_CP_IB2_BASE), - REG_ADRENO_DEFINE(REG_ADRENO_CP_IB2_BUFSZ, REG_A4XX_CP_IB2_BUFSZ), - REG_ADRENO_DEFINE(REG_ADRENO_CP_TIMESTAMP, REG_AXXX_CP_SCRATCH_REG0), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ME_RAM_RADDR, REG_A4XX_CP_ME_RAM_RADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_ADDR, REG_A4XX_CP_ROQ_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_ROQ_DATA, REG_A4XX_CP_ROQ_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_ADDR, REG_A4XX_CP_MERCIU_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA, REG_A4XX_CP_MERCIU_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MERCIU_DATA2, REG_A4XX_CP_MERCIU_DATA2), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_ADDR, REG_A4XX_CP_MEQ_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_CP_MEQ_DATA, REG_A4XX_CP_MEQ_DATA), - REG_ADRENO_DEFINE(REG_ADRENO_CP_HW_FAULT, REG_A4XX_CP_HW_FAULT), - REG_ADRENO_DEFINE(REG_ADRENO_CP_PROTECT_STATUS, - REG_A4XX_CP_PROTECT_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_ADDR, REG_A4XX_CP_SCRATCH_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_SCRATCH_UMSK, REG_A4XX_CP_SCRATCH_UMASK), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_STATUS, REG_A4XX_RBBM_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_CTL, - REG_A4XX_RBBM_PERFCTR_CTL), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD0, - REG_A4XX_RBBM_PERFCTR_LOAD_CMD0), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD1, - REG_A4XX_RBBM_PERFCTR_LOAD_CMD1), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_CMD2, - REG_A4XX_RBBM_PERFCTR_LOAD_CMD2), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_PWR_1_LO, - REG_A4XX_RBBM_PERFCTR_PWR_1_LO), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_MASK, REG_A4XX_RBBM_INT_0_MASK), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_0_STATUS, - REG_A4XX_RBBM_INT_0_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_ERROR_STATUS, - REG_A4XX_RBBM_AHB_ERROR_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_CMD, REG_A4XX_RBBM_AHB_CMD), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_CLOCK_CTL, REG_A4XX_RBBM_CLOCK_CTL), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_ME_SPLIT_STATUS, - REG_A4XX_RBBM_AHB_ME_SPLIT_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_AHB_PFP_SPLIT_STATUS, - REG_A4XX_RBBM_AHB_PFP_SPLIT_STATUS), - REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_SEL, - REG_A4XX_VPC_DEBUG_RAM_SEL), - REG_ADRENO_DEFINE(REG_ADRENO_VPC_DEBUG_RAM_READ, - REG_A4XX_VPC_DEBUG_RAM_READ), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_INT_CLEAR_CMD, - REG_A4XX_RBBM_INT_CLEAR_CMD), - REG_ADRENO_DEFINE(REG_ADRENO_VSC_SIZE_ADDRESS, - REG_A4XX_VSC_SIZE_ADDRESS), - REG_ADRENO_DEFINE(REG_ADRENO_VFD_CONTROL_0, REG_A4XX_VFD_CONTROL_0), - REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_PVT_MEM_ADDR_REG, - REG_A4XX_SP_VS_PVT_MEM_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_PVT_MEM_ADDR_REG, - REG_A4XX_SP_FS_PVT_MEM_ADDR), - REG_ADRENO_DEFINE(REG_ADRENO_SP_VS_OBJ_START_REG, - REG_A4XX_SP_VS_OBJ_START), - REG_ADRENO_DEFINE(REG_ADRENO_SP_FS_OBJ_START_REG, - REG_A4XX_SP_FS_OBJ_START), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_RBBM_CTL, REG_A4XX_RBBM_RBBM_CTL), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_SW_RESET_CMD, - REG_A4XX_RBBM_SW_RESET_CMD), - REG_ADRENO_DEFINE(REG_ADRENO_UCHE_INVALIDATE0, - REG_A4XX_UCHE_INVALIDATE0), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_LO, - REG_A4XX_RBBM_PERFCTR_LOAD_VALUE_LO), - REG_ADRENO_DEFINE(REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_HI, - REG_A4XX_RBBM_PERFCTR_LOAD_VALUE_HI), }; static void a4xx_dump(struct msm_gpu *gpu) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index 3915c737cdd6..67ceeb00f98e 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -32,6 +32,9 @@ #include "adreno_pm4.xml.h" #define REG_ADRENO_DEFINE(_offset, _reg) [_offset] = (_reg) + 1 +#define REG_SKIP ~0 +#define REG_ADRENO_SKIP(_offset) [_offset] = REG_SKIP + /** * adreno_regs: List of registers that are used in across all * 3D devices. Each device type has different offset value for the same @@ -39,70 +42,13 @@ * and are indexed by the enumeration values defined in this enum */ enum adreno_regs { - REG_ADRENO_CP_DEBUG, - REG_ADRENO_CP_ME_RAM_WADDR, - REG_ADRENO_CP_ME_RAM_DATA, - REG_ADRENO_CP_PFP_UCODE_DATA, - REG_ADRENO_CP_PFP_UCODE_ADDR, - REG_ADRENO_CP_WFI_PEND_CTR, REG_ADRENO_CP_RB_BASE, + REG_ADRENO_CP_RB_BASE_HI, REG_ADRENO_CP_RB_RPTR_ADDR, + REG_ADRENO_CP_RB_RPTR_ADDR_HI, REG_ADRENO_CP_RB_RPTR, REG_ADRENO_CP_RB_WPTR, - REG_ADRENO_CP_PROTECT_CTRL, - REG_ADRENO_CP_ME_CNTL, REG_ADRENO_CP_RB_CNTL, - REG_ADRENO_CP_IB1_BASE, - REG_ADRENO_CP_IB1_BUFSZ, - REG_ADRENO_CP_IB2_BASE, - REG_ADRENO_CP_IB2_BUFSZ, - REG_ADRENO_CP_TIMESTAMP, - REG_ADRENO_CP_ME_RAM_RADDR, - REG_ADRENO_CP_ROQ_ADDR, - REG_ADRENO_CP_ROQ_DATA, - REG_ADRENO_CP_MERCIU_ADDR, - REG_ADRENO_CP_MERCIU_DATA, - REG_ADRENO_CP_MERCIU_DATA2, - REG_ADRENO_CP_MEQ_ADDR, - REG_ADRENO_CP_MEQ_DATA, - REG_ADRENO_CP_HW_FAULT, - REG_ADRENO_CP_PROTECT_STATUS, - REG_ADRENO_SCRATCH_ADDR, - REG_ADRENO_SCRATCH_UMSK, - REG_ADRENO_SCRATCH_REG2, - REG_ADRENO_RBBM_STATUS, - REG_ADRENO_RBBM_PERFCTR_CTL, - REG_ADRENO_RBBM_PERFCTR_LOAD_CMD0, - REG_ADRENO_RBBM_PERFCTR_LOAD_CMD1, - REG_ADRENO_RBBM_PERFCTR_LOAD_CMD2, - REG_ADRENO_RBBM_PERFCTR_PWR_1_LO, - REG_ADRENO_RBBM_INT_0_MASK, - REG_ADRENO_RBBM_INT_0_STATUS, - REG_ADRENO_RBBM_AHB_ERROR_STATUS, - REG_ADRENO_RBBM_PM_OVERRIDE2, - REG_ADRENO_RBBM_AHB_CMD, - REG_ADRENO_RBBM_INT_CLEAR_CMD, - REG_ADRENO_RBBM_SW_RESET_CMD, - REG_ADRENO_RBBM_CLOCK_CTL, - REG_ADRENO_RBBM_AHB_ME_SPLIT_STATUS, - REG_ADRENO_RBBM_AHB_PFP_SPLIT_STATUS, - REG_ADRENO_VPC_DEBUG_RAM_SEL, - REG_ADRENO_VPC_DEBUG_RAM_READ, - REG_ADRENO_VSC_SIZE_ADDRESS, - REG_ADRENO_VFD_CONTROL_0, - REG_ADRENO_VFD_INDEX_MAX, - REG_ADRENO_SP_VS_PVT_MEM_ADDR_REG, - REG_ADRENO_SP_FS_PVT_MEM_ADDR_REG, - REG_ADRENO_SP_VS_OBJ_START_REG, - REG_ADRENO_SP_FS_OBJ_START_REG, - REG_ADRENO_PA_SC_AA_CONFIG, - REG_ADRENO_SQ_GPR_MANAGEMENT, - REG_ADRENO_SQ_INST_STORE_MANAGMENT, - REG_ADRENO_TP0_CHICKEN, - REG_ADRENO_RBBM_RBBM_CTL, - REG_ADRENO_UCHE_INVALIDATE0, - REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_LO, - REG_ADRENO_RBBM_PERFCTR_LOAD_VALUE_HI, REG_ADRENO_REGISTER_MAX, }; @@ -313,7 +259,7 @@ OUT_PKT7(struct msm_ringbuffer *ring, uint8_t opcode, uint16_t cnt) } /* - * adreno_checkreg_off() - Checks the validity of a register enum + * adreno_reg_check() - Checks the validity of a register enum * @gpu: Pointer to struct adreno_gpu * @offset_name: The register enum that is checked */ @@ -324,6 +270,16 @@ static inline bool adreno_reg_check(struct adreno_gpu *gpu, !gpu->reg_offsets[offset_name]) { BUG(); } + + /* + * REG_SKIP is a special value that tell us that the register in + * question isn't implemented on target but don't trigger a BUG(). This + * is used to cleanly implement adreno_gpu_write64() and + * adreno_gpu_read64() in a generic fashion + */ + if (gpu->reg_offsets[offset_name] == REG_SKIP) + return false; + return true; } From 6c523382a937fb185f30b4d1a7ef63ecb79d3059 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:15 -0700 Subject: [PATCH 04/19] drm/msm: gpu Add new gpu register read/write functions Add some new functions to manipulate GPU registers. gpu_read64 and gpu_write64 can read/write a 64 bit value to two 32 bit registers. For 4XX and older these are normally perfcounter registers, but future targets will use 64 bit addressing so there will be many more spots where a 64 bit read and write are needed. gpu_rmw() does a read/modify/write on a 32 bit register given a mask and bits to OR in. Change-Id: Ic0dedbadb83d3ac46f7e463c9c901d4f94a7bb58 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a4xx_gpu.c | 12 ++------- drivers/gpu/drm/msm/msm_gpu.h | 39 +++++++++++++++++++++++++++ 2 files changed, 41 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c index dde9dc1333e1..47c9b22b0801 100644 --- a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c @@ -508,16 +508,8 @@ static int a4xx_pm_suspend(struct msm_gpu *gpu) { static int a4xx_get_timestamp(struct msm_gpu *gpu, uint64_t *value) { - uint32_t hi, lo, tmp; - - tmp = gpu_read(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_HI); - do { - hi = tmp; - lo = gpu_read(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_LO); - tmp = gpu_read(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_HI); - } while (tmp != hi); - - *value = (((uint64_t)hi) << 32) | lo; + *value = gpu_read64(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_LO, + REG_A4XX_RBBM_PERFCTR_CP_0_HI); return 0; } diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index 309422b73a49..c4aff2e0829b 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -151,6 +151,45 @@ static inline u32 gpu_read(struct msm_gpu *gpu, u32 reg) return msm_readl(gpu->mmio + (reg << 2)); } +static inline void gpu_rmw(struct msm_gpu *gpu, u32 reg, u32 mask, u32 or) +{ + uint32_t val = gpu_read(gpu, reg); + + val &= ~mask; + gpu_write(gpu, reg, val | or); +} + +static inline u64 gpu_read64(struct msm_gpu *gpu, u32 lo, u32 hi) +{ + u64 val; + + /* + * Why not a readq here? Two reasons: 1) many of the LO registers are + * not quad word aligned and 2) the GPU hardware designers have a bit + * of a history of putting registers where they fit, especially in + * spins. The longer a GPU family goes the higher the chance that + * we'll get burned. We could do a series of validity checks if we + * wanted to, but really is a readq() that much better? Nah. + */ + + /* + * For some lo/hi registers (like perfcounters), the hi value is latched + * when the lo is read, so make sure to read the lo first to trigger + * that + */ + val = (u64) msm_readl(gpu->mmio + (lo << 2)); + val |= ((u64) msm_readl(gpu->mmio + (hi << 2)) << 32); + + return val; +} + +static inline void gpu_write64(struct msm_gpu *gpu, u32 lo, u32 hi, u64 val) +{ + /* Why not a writeq here? Read the screed above */ + msm_writel(lower_32_bits(val), gpu->mmio + (lo << 2)); + msm_writel(upper_32_bits(val), gpu->mmio + (hi << 2)); +} + int msm_gpu_pm_suspend(struct msm_gpu *gpu); int msm_gpu_pm_resume(struct msm_gpu *gpu); From 49892bb8a55b2ce10f62d6303cf2ddf7e5675fe6 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:15 -0700 Subject: [PATCH 05/19] drm/msm: Disable interrupts during init Disable the interrupt during the init sequence to avoid having interrupts fired for errors and other things that we are not ready to handle while initializing. Change-Id: Ic0dedbad972f25586e792478f9c96c4af7c31d17 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/adreno_device.c | 4 ++++ drivers/gpu/drm/msm/adreno/adreno_gpu.c | 3 +++ 2 files changed, 7 insertions(+) diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c index 5127b75dbf40..3321d33ae6a9 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_device.c +++ b/drivers/gpu/drm/msm/adreno/adreno_device.c @@ -148,12 +148,16 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev) mutex_lock(&dev->struct_mutex); gpu->funcs->pm_resume(gpu); mutex_unlock(&dev->struct_mutex); + + disable_irq(gpu->irq); + ret = gpu->funcs->hw_init(gpu); if (ret) { dev_err(dev->dev, "gpu hw init failed: %d\n", ret); gpu->funcs->destroy(gpu); gpu = NULL; } else { + enable_irq(gpu->irq); /* give inactive pm a chance to kick in: */ msm_gpu_retire(gpu); } diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 4b518bfdeee4..478a41181083 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -126,11 +126,14 @@ void adreno_recover(struct msm_gpu *gpu) adreno_gpu->memptrs->wptr = 0; gpu->funcs->pm_resume(gpu); + + disable_irq(gpu->irq); ret = gpu->funcs->hw_init(gpu); if (ret) { dev_err(dev->dev, "gpu hw init failed: %d\n", ret); /* hmm, oh well? */ } + enable_irq(gpu->irq); } int adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, From 3aea14ccac22a874ae7666cec1b11a9ab4a82f02 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:16 -0700 Subject: [PATCH 06/19] drm/msm: Ask PM_QOS to set the interrupt affnity PM_QOS will put the interrupt where it wants to and we're okay with that. Change-Id: Ic0dedbad5294d51a55125a0021f7dcc3b185de02 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/msm_gpu.c | 8 ++++++++ drivers/gpu/drm/msm/msm_gpu.h | 3 +++ 2 files changed, 11 insertions(+) diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 7b8f2d02b56d..1dc82145d3d0 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -681,6 +681,14 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev, goto fail; } +#ifdef CONFIG_SMP + gpu->pm_qos_req_dma.type = PM_QOS_REQ_AFFINE_IRQ; + gpu->pm_qos_req_dma.irq = gpu->irq; +#endif + + pm_qos_add_request(&gpu->pm_qos_req_dma, PM_QOS_CPU_DMA_LATENCY, + PM_QOS_DEFAULT_VALUE); + bs_init(gpu); return 0; diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index c4aff2e0829b..401f74075485 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -19,6 +19,7 @@ #define __MSM_GPU_H__ #include +#include #include #include "msm_drv.h" @@ -103,6 +104,8 @@ struct msm_gpu { struct clk *ebi1_clk, *grp_clks[6]; uint32_t fast_rate, slow_rate, bus_freq; + struct pm_qos_request pm_qos_req_dma; + #ifdef DOWNSTREAM_CONFIG_MSM_BUS_SCALING struct msm_bus_scale_pdata *bus_scale_table; uint32_t bsc; From 1314edd5c5a4c42e5159285de31814ec6ff8beba Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:16 -0700 Subject: [PATCH 07/19] drm/msm: get an iova from the address space instead of an id In the future we won't have a fixed set of addresses spaces. Instead of going through the effort of assigning a ID for each address space just use the address space itself as a token for getting / putting an iova. This forces a few changes in the gem object however: instead of using a simple index into a list of domains, we need to maintain a list of them. Luckily the list will be pretty small; even with dynamic address spaces we wouldn't ever see more than two or three. Change-Id: Ic0dedbad4495f02a21135217f3605b93f8b8dfea Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 6 +- drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c | 7 +- drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c | 11 +- drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h | 2 - drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c | 12 +- drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c | 4 +- drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c | 7 - drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h | 1 - drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c | 12 +- drivers/gpu/drm/msm/msm_drv.c | 14 -- drivers/gpu/drm/msm/msm_drv.h | 24 +-- drivers/gpu/drm/msm/msm_fb.c | 15 +- drivers/gpu/drm/msm/msm_gem.c | 138 +++++++++++++----- drivers/gpu/drm/msm/msm_gem.h | 4 +- drivers/gpu/drm/msm/msm_gem_submit.c | 4 +- drivers/gpu/drm/msm/msm_gpu.c | 8 +- drivers/gpu/drm/msm/msm_gpu.h | 1 - drivers/gpu/drm/msm/sde/sde_connector.c | 34 ++--- drivers/gpu/drm/msm/sde/sde_connector.h | 8 +- drivers/gpu/drm/msm/sde/sde_encoder_phys.h | 4 +- drivers/gpu/drm/msm/sde/sde_encoder_phys_wb.c | 29 ++-- drivers/gpu/drm/msm/sde/sde_formats.c | 20 +-- drivers/gpu/drm/msm/sde/sde_formats.h | 11 +- drivers/gpu/drm/msm/sde/sde_kms.c | 21 +-- drivers/gpu/drm/msm/sde/sde_kms.h | 1 - drivers/gpu/drm/msm/sde/sde_plane.c | 12 +- 26 files changed, 220 insertions(+), 190 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 478a41181083..392622953df9 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -64,7 +64,7 @@ int adreno_hw_init(struct msm_gpu *gpu) DBG("%s", gpu->name); - ret = msm_gem_get_iova(gpu->rb->bo, gpu->id, &gpu->rb_iova); + ret = msm_gem_get_iova(gpu->rb->bo, gpu->aspace, &gpu->rb_iova); if (ret) { gpu->rb_iova = 0; dev_err(gpu->dev->dev, "could not map ringbuffer: %d\n", ret); @@ -417,7 +417,7 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, return -ENOMEM; } - ret = msm_gem_get_iova(adreno_gpu->memptrs_bo, gpu->id, + ret = msm_gem_get_iova(adreno_gpu->memptrs_bo, gpu->aspace, &adreno_gpu->memptrs_iova); if (ret) { dev_err(drm->dev, "could not map memptrs: %d\n", ret); @@ -433,7 +433,7 @@ void adreno_gpu_cleanup(struct adreno_gpu *gpu) if (gpu->memptrs_bo) { if (gpu->memptrs_iova) - msm_gem_put_iova(gpu->memptrs_bo, gpu->base.id); + msm_gem_put_iova(gpu->memptrs_bo, gpu->base.aspace); drm_gem_object_unreference_unlocked(gpu->memptrs_bo); } release_firmware(gpu->pm4); diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c index 6ac9aa165768..dff43329c020 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c @@ -133,7 +133,7 @@ static void unref_cursor_worker(struct drm_flip_work *work, void *val) container_of(work, struct mdp4_crtc, unref_cursor_work); struct mdp4_kms *mdp4_kms = get_kms(&mdp4_crtc->base); - msm_gem_put_iova(val, mdp4_kms->id); + msm_gem_put_iova(val, mdp4_kms->aspace); drm_gem_object_unreference_unlocked(val); } @@ -392,7 +392,8 @@ static void update_cursor(struct drm_crtc *crtc) if (next_bo) { /* take a obj ref + iova ref when we start scanning out: */ drm_gem_object_reference(next_bo); - msm_gem_get_iova_locked(next_bo, mdp4_kms->id, &iova); + msm_gem_get_iova_locked(next_bo, mdp4_kms->aspace, + &iova); /* enable cursor: */ mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_SIZE(dma), @@ -449,7 +450,7 @@ static int mdp4_crtc_cursor_set(struct drm_crtc *crtc, } if (cursor_bo) { - ret = msm_gem_get_iova(cursor_bo, mdp4_kms->id, &iova); + ret = msm_gem_get_iova(cursor_bo, mdp4_kms->aspace, &iova); if (ret) goto fail; } else { diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c index 6d8e174236ea..da9d58eccc3a 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c @@ -197,7 +197,7 @@ static void mdp4_destroy(struct msm_kms *kms) struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms)); if (mdp4_kms->blank_cursor_iova) - msm_gem_put_iova(mdp4_kms->blank_cursor_bo, mdp4_kms->id); + msm_gem_put_iova(mdp4_kms->blank_cursor_bo, mdp4_kms->aspace); if (mdp4_kms->blank_cursor_bo) drm_gem_object_unreference_unlocked(mdp4_kms->blank_cursor_bo); @@ -541,13 +541,6 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev) aspace = NULL; } - mdp4_kms->id = msm_register_address_space(dev, aspace); - if (mdp4_kms->id < 0) { - ret = mdp4_kms->id; - dev_err(dev->dev, "failed to register mdp4 iommu: %d\n", ret); - goto fail; - } - ret = modeset_init(mdp4_kms); if (ret) { dev_err(dev->dev, "modeset_init failed: %d\n", ret); @@ -564,7 +557,7 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev) goto fail; } - ret = msm_gem_get_iova(mdp4_kms->blank_cursor_bo, mdp4_kms->id, + ret = msm_gem_get_iova(mdp4_kms->blank_cursor_bo, mdp4_kms->aspace, &mdp4_kms->blank_cursor_iova); if (ret) { dev_err(dev->dev, "could not pin blank-cursor bo: %d\n", ret); diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h index 923a028db452..aa060e71ec91 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h @@ -33,8 +33,6 @@ struct mdp4_kms { int rev; /* mapper-id used to request GEM buffer mapped for scanout: */ - int id; - void __iomem *mmio; struct regulator *dsi_pll_vdda; diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c index 30d57e74c42f..bc1ece2a5b7e 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c @@ -109,7 +109,7 @@ static int mdp4_plane_prepare_fb(struct drm_plane *plane, return 0; DBG("%s: prepare: FB[%u]", mdp4_plane->name, fb->base.id); - return msm_framebuffer_prepare(fb, mdp4_kms->id); + return msm_framebuffer_prepare(fb, mdp4_kms->aspace); } static void mdp4_plane_cleanup_fb(struct drm_plane *plane, @@ -123,7 +123,7 @@ static void mdp4_plane_cleanup_fb(struct drm_plane *plane, return; DBG("%s: cleanup: FB[%u]", mdp4_plane->name, fb->base.id); - msm_framebuffer_cleanup(fb, mdp4_kms->id); + msm_framebuffer_cleanup(fb, mdp4_kms->aspace); } @@ -172,13 +172,13 @@ static void mdp4_plane_set_scanout(struct drm_plane *plane, MDP4_PIPE_SRC_STRIDE_B_P3(fb->pitches[3])); mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRCP0_BASE(pipe), - msm_framebuffer_iova(fb, mdp4_kms->id, 0)); + msm_framebuffer_iova(fb, mdp4_kms->aspace, 0)); mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRCP1_BASE(pipe), - msm_framebuffer_iova(fb, mdp4_kms->id, 1)); + msm_framebuffer_iova(fb, mdp4_kms->aspace, 1)); mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRCP2_BASE(pipe), - msm_framebuffer_iova(fb, mdp4_kms->id, 2)); + msm_framebuffer_iova(fb, mdp4_kms->aspace, 2)); mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRCP3_BASE(pipe), - msm_framebuffer_iova(fb, mdp4_kms->id, 3)); + msm_framebuffer_iova(fb, mdp4_kms->aspace, 3)); plane->fb = fb; } diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c index 7f9f4ac88029..56c1f6c4ab96 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c @@ -171,7 +171,7 @@ static void unref_cursor_worker(struct drm_flip_work *work, void *val) container_of(work, struct mdp5_crtc, unref_cursor_work); struct mdp5_kms *mdp5_kms = get_kms(&mdp5_crtc->base); - msm_gem_put_iova(val, mdp5_kms->id); + msm_gem_put_iova(val, mdp5_kms->aspace); drm_gem_object_unreference_unlocked(val); } @@ -536,7 +536,7 @@ static int mdp5_crtc_cursor_set(struct drm_crtc *crtc, if (!cursor_bo) return -ENOENT; - ret = msm_gem_get_iova(cursor_bo, mdp5_kms->id, &cursor_addr); + ret = msm_gem_get_iova(cursor_bo, mdp5_kms->aspace, &cursor_addr); if (ret) return -EINVAL; diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c index cc5a73505537..ede681e12a68 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c @@ -627,13 +627,6 @@ struct msm_kms *mdp5_kms_init(struct drm_device *dev) aspace = NULL; } - mdp5_kms->id = msm_register_address_space(dev, aspace); - if (mdp5_kms->id < 0) { - ret = mdp5_kms->id; - dev_err(dev->dev, "failed to register mdp5 iommu: %d\n", ret); - goto fail; - } - ret = modeset_init(mdp5_kms); if (ret) { dev_err(dev->dev, "modeset_init failed: %d\n", ret); diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h index 6cb43ad0c1d7..c1aa86d21416 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h @@ -36,7 +36,6 @@ struct mdp5_kms { /* mapper-id used to request GEM buffer mapped for scanout: */ - int id; struct msm_gem_address_space *aspace; struct mdp5_smp *smp; diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c index 81cd49045ffc..873ab11d34d2 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c @@ -260,7 +260,7 @@ static int mdp5_plane_prepare_fb(struct drm_plane *plane, return 0; DBG("%s: prepare: FB[%u]", mdp5_plane->name, fb->base.id); - return msm_framebuffer_prepare(fb, mdp5_kms->id); + return msm_framebuffer_prepare(fb, mdp5_kms->aspace); } static void mdp5_plane_cleanup_fb(struct drm_plane *plane, @@ -274,7 +274,7 @@ static void mdp5_plane_cleanup_fb(struct drm_plane *plane, return; DBG("%s: cleanup: FB[%u]", mdp5_plane->name, fb->base.id); - msm_framebuffer_cleanup(fb, mdp5_kms->id); + msm_framebuffer_cleanup(fb, mdp5_kms->aspace); } static int mdp5_plane_atomic_check(struct drm_plane *plane, @@ -400,13 +400,13 @@ static void set_scanout_locked(struct drm_plane *plane, MDP5_PIPE_SRC_STRIDE_B_P3(fb->pitches[3])); mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC0_ADDR(pipe), - msm_framebuffer_iova(fb, mdp5_kms->id, 0)); + msm_framebuffer_iova(fb, mdp5_kms->aspace, 0)); mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC1_ADDR(pipe), - msm_framebuffer_iova(fb, mdp5_kms->id, 1)); + msm_framebuffer_iova(fb, mdp5_kms->aspace, 1)); mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC2_ADDR(pipe), - msm_framebuffer_iova(fb, mdp5_kms->id, 2)); + msm_framebuffer_iova(fb, mdp5_kms->aspace, 2)); mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC3_ADDR(pipe), - msm_framebuffer_iova(fb, mdp5_kms->id, 3)); + msm_framebuffer_iova(fb, mdp5_kms->aspace, 3)); plane->fb = fb; } diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index 4b67edf5a77b..6d3cf1accdb9 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -38,20 +38,6 @@ static const struct drm_mode_config_funcs mode_config_funcs = { .atomic_commit = msm_atomic_commit, }; -int msm_register_address_space(struct drm_device *dev, - struct msm_gem_address_space *aspace) -{ - struct msm_drm_private *priv = dev->dev_private; - int idx = priv->num_aspaces++; - - if (WARN_ON(idx >= ARRAY_SIZE(priv->aspace))) - return -EINVAL; - - priv->aspace[idx] = aspace; - - return idx; -} - #ifdef CONFIG_DRM_MSM_REGISTER_LOGGING static bool reglog = false; MODULE_PARM_DESC(reglog, "Enable register read/write logging"); diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h index 4ed5953d07f9..f59bd5fd4d37 100644 --- a/drivers/gpu/drm/msm/msm_drv.h +++ b/drivers/gpu/drm/msm/msm_drv.h @@ -374,8 +374,6 @@ int msm_queue_fence_cb(struct drm_device *dev, struct msm_fence_cb *cb, uint32_t fence); void msm_update_fence(struct drm_device *dev, uint32_t fence); -int msm_register_address_space(struct drm_device *dev, - struct msm_gem_address_space *aspace); void msm_gem_unmap_vma(struct msm_gem_address_space *aspace, struct msm_gem_vma *vma, struct sg_table *sgt, void *priv); @@ -402,13 +400,16 @@ int msm_gem_mmap_obj(struct drm_gem_object *obj, int msm_gem_mmap(struct file *filp, struct vm_area_struct *vma); int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf); uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj); -int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id, - uint32_t *iova); -int msm_gem_get_iova(struct drm_gem_object *obj, int id, uint32_t *iova); -uint32_t msm_gem_iova(struct drm_gem_object *obj, int id); +int msm_gem_get_iova_locked(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace, uint32_t *iova); +int msm_gem_get_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace, uint32_t *iova); +uint32_t msm_gem_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace); struct page **msm_gem_get_pages(struct drm_gem_object *obj); void msm_gem_put_pages(struct drm_gem_object *obj); -void msm_gem_put_iova(struct drm_gem_object *obj, int id); +void msm_gem_put_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace); int msm_gem_dumb_create(struct drm_file *file, struct drm_device *dev, struct drm_mode_create_dumb *args); int msm_gem_dumb_map_offset(struct drm_file *file, struct drm_device *dev, @@ -439,9 +440,12 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev, struct drm_gem_object *msm_gem_import(struct drm_device *dev, uint32_t size, struct sg_table *sgt); -int msm_framebuffer_prepare(struct drm_framebuffer *fb, int id); -void msm_framebuffer_cleanup(struct drm_framebuffer *fb, int id); -uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, int id, int plane); +int msm_framebuffer_prepare(struct drm_framebuffer *fb, + struct msm_gem_address_space *aspace); +void msm_framebuffer_cleanup(struct drm_framebuffer *fb, + struct msm_gem_address_space *aspace); +uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, + struct msm_gem_address_space *aspace, int plane); struct drm_gem_object *msm_framebuffer_bo(struct drm_framebuffer *fb, int plane); const struct msm_format *msm_framebuffer_format(struct drm_framebuffer *fb); struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev, diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c index dca4de382581..28395ea13235 100644 --- a/drivers/gpu/drm/msm/msm_fb.c +++ b/drivers/gpu/drm/msm/msm_fb.c @@ -92,14 +92,15 @@ void msm_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m) * should be fine, since only the scanout (mdpN) side of things needs * this, the gpu doesn't care about fb's. */ -int msm_framebuffer_prepare(struct drm_framebuffer *fb, int id) +int msm_framebuffer_prepare(struct drm_framebuffer *fb, + struct msm_gem_address_space *aspace) { struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb); int ret, i, n = drm_format_num_planes(fb->pixel_format); uint32_t iova; for (i = 0; i < n; i++) { - ret = msm_gem_get_iova(msm_fb->planes[i], id, &iova); + ret = msm_gem_get_iova(msm_fb->planes[i], aspace, &iova); DBG("FB[%u]: iova[%d]: %08x (%d)", fb->base.id, i, iova, ret); if (ret) return ret; @@ -108,21 +109,23 @@ int msm_framebuffer_prepare(struct drm_framebuffer *fb, int id) return 0; } -void msm_framebuffer_cleanup(struct drm_framebuffer *fb, int id) +void msm_framebuffer_cleanup(struct drm_framebuffer *fb, + struct msm_gem_address_space *aspace) { struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb); int i, n = drm_format_num_planes(fb->pixel_format); for (i = 0; i < n; i++) - msm_gem_put_iova(msm_fb->planes[i], id); + msm_gem_put_iova(msm_fb->planes[i], aspace); } -uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, int id, int plane) +uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, + struct msm_gem_address_space *aspace, int plane) { struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb); if (!msm_fb->planes[plane]) return 0; - return msm_gem_iova(msm_fb->planes[plane], id) + fb->offsets[plane]; + return msm_gem_iova(msm_fb->planes[plane], aspace) + fb->offsets[plane]; } struct drm_gem_object *msm_framebuffer_bo(struct drm_framebuffer *fb, int plane) diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index 5aa08cf4d6d8..ca24cc433afc 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -273,22 +273,63 @@ uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj) return offset; } +static void obj_remove_domain(struct msm_gem_vma *domain) +{ + if (domain) { + list_del(&domain->list); + kfree(domain); + } +} + static void put_iova(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct msm_drm_private *priv = obj->dev->dev_private; struct msm_gem_object *msm_obj = to_msm_bo(obj); - int id; + struct msm_gem_vma *domain, *tmp; WARN_ON(!mutex_is_locked(&dev->struct_mutex)); - for (id = 0; id < ARRAY_SIZE(msm_obj->domain); id++) { - msm_gem_unmap_vma(priv->aspace[id], &msm_obj->domain[id], - msm_obj->sgt, get_dmabuf_ptr(obj)); + list_for_each_entry_safe(domain, tmp, &msm_obj->domains, list) { + if (iommu_present(&platform_bus_type)) { + msm_gem_unmap_vma(domain->aspace, domain, + msm_obj->sgt, get_dmabuf_ptr(obj)); + } + + obj_remove_domain(domain); } } +static struct msm_gem_vma *obj_add_domain(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace) +{ + struct msm_gem_object *msm_obj = to_msm_bo(obj); + struct msm_gem_vma *domain = kzalloc(sizeof(*domain), GFP_KERNEL); + + if (!domain) + return ERR_PTR(-ENOMEM); + + domain->aspace = aspace; + + list_add_tail(&domain->list, &msm_obj->domains); + + return domain; +} + +static struct msm_gem_vma *obj_get_domain(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace) +{ + struct msm_gem_object *msm_obj = to_msm_bo(obj); + struct msm_gem_vma *domain; + + list_for_each_entry(domain, &msm_obj->domains, list) { + if (domain->aspace == aspace) + return domain; + } + + return NULL; +} + /* should be called under struct_mutex.. although it can be called * from atomic context without struct_mutex to acquire an extra * iova ref if you know one is already held. @@ -296,49 +337,64 @@ put_iova(struct drm_gem_object *obj) * That means when I do eventually need to add support for unpinning * the refcnt counter needs to be atomic_t. */ -int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id, - uint32_t *iova) +int msm_gem_get_iova_locked(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace, uint32_t *iova) { struct msm_gem_object *msm_obj = to_msm_bo(obj); + struct page **pages; + struct msm_gem_vma *domain; int ret = 0; - if (!msm_obj->domain[id].iova) { - struct msm_drm_private *priv = obj->dev->dev_private; - struct page **pages = get_pages(obj); + if (!iommu_present(&platform_bus_type)) { + pages = get_pages(obj); if (IS_ERR(pages)) return PTR_ERR(pages); - if (iommu_present(&platform_bus_type)) { - ret = msm_gem_map_vma(priv->aspace[id], - &msm_obj->domain[id], msm_obj->sgt, - get_dmabuf_ptr(obj)); - } else - msm_obj->domain[id].iova = physaddr(obj); + *iova = physaddr(obj); + return 0; + } + + domain = obj_get_domain(obj, aspace); + + if (!domain) { + domain = obj_add_domain(obj, aspace); + if (IS_ERR(domain)) + return PTR_ERR(domain); + + pages = get_pages(obj); + if (IS_ERR(pages)) { + obj_remove_domain(domain); + return PTR_ERR(pages); + } + + ret = msm_gem_map_vma(aspace, domain, msm_obj->sgt, + get_dmabuf_ptr(obj)); } if (!ret) - *iova = msm_obj->domain[id].iova; + *iova = domain->iova; + else + obj_remove_domain(domain); return ret; } /* get iova, taking a reference. Should have a matching put */ -int msm_gem_get_iova(struct drm_gem_object *obj, int id, uint32_t *iova) +int msm_gem_get_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace, uint32_t *iova) { - struct msm_gem_object *msm_obj = to_msm_bo(obj); + struct msm_gem_vma *domain; int ret; - /* this is safe right now because we don't unmap until the - * bo is deleted: - */ - if (msm_obj->domain[id].iova) { - *iova = msm_obj->domain[id].iova; + domain = obj_get_domain(obj, aspace); + if (domain) { + *iova = domain->iova; return 0; } mutex_lock(&obj->dev->struct_mutex); - ret = msm_gem_get_iova_locked(obj, id, iova); + ret = msm_gem_get_iova_locked(obj, aspace, iova); mutex_unlock(&obj->dev->struct_mutex); return ret; } @@ -346,14 +402,18 @@ int msm_gem_get_iova(struct drm_gem_object *obj, int id, uint32_t *iova) /* get iova without taking a reference, used in places where you have * already done a 'msm_gem_get_iova()'. */ -uint32_t msm_gem_iova(struct drm_gem_object *obj, int id) +uint32_t msm_gem_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace) { - struct msm_gem_object *msm_obj = to_msm_bo(obj); - WARN_ON(!msm_obj->domain[id].iova); - return msm_obj->domain[id].iova; + struct msm_gem_vma *domain = obj_get_domain(obj, aspace); + + WARN_ON(!domain); + + return domain ? domain->iova : 0; } -void msm_gem_put_iova(struct drm_gem_object *obj, int id) +void msm_gem_put_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace) { // XXX TODO .. // NOTE: probably don't need a _locked() version.. we wouldn't @@ -487,9 +547,8 @@ void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m) { struct drm_device *dev = obj->dev; struct msm_gem_object *msm_obj = to_msm_bo(obj); - struct msm_drm_private *priv = obj->dev->dev_private; + struct msm_gem_vma *domain; uint64_t off = drm_vma_node_start(&obj->vma_node); - int id; WARN_ON(!mutex_is_locked(&dev->struct_mutex)); seq_printf(m, "%08x: %c(r=%u,w=%u) %2d (%2d) %08llx %p\t", @@ -498,8 +557,9 @@ void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m) obj->name, obj->refcount.refcount.counter, off, msm_obj->vaddr); - for (id = 0; id < priv->num_aspaces; id++) - seq_printf(m, " %08llx", msm_obj->domain[id].iova); + /* FIXME: we need to print the address space here too */ + list_for_each_entry(domain, &msm_obj->domains, list) + seq_printf(m, " %08llx", domain->iova); seq_puts(m, "\n"); } @@ -618,8 +678,12 @@ static int msm_gem_new_impl(struct drm_device *dev, if (!msm_obj) return -ENOMEM; - if (use_vram) - msm_obj->vram_node = &msm_obj->domain[0].node; + if (use_vram) { + struct msm_gem_vma *domain = obj_add_domain(&msm_obj->base, 0); + + if (!IS_ERR(domain)) + msm_obj->vram_node = &domain->node; + } msm_obj->flags = flags; @@ -627,6 +691,8 @@ static int msm_gem_new_impl(struct drm_device *dev, reservation_object_init(msm_obj->resv); INIT_LIST_HEAD(&msm_obj->submit_entry); + INIT_LIST_HEAD(&msm_obj->domains); + list_add_tail(&msm_obj->mm_list, &priv->inactive_list); *obj = &msm_obj->base; diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h index 86c297c6de32..937e80c12554 100644 --- a/drivers/gpu/drm/msm/msm_gem.h +++ b/drivers/gpu/drm/msm/msm_gem.h @@ -43,7 +43,9 @@ struct msm_gem_address_space { struct msm_gem_vma { /* Node used by the GPU address space, but not the SDE address space */ struct drm_mm_node node; + struct msm_gem_address_space *aspace; uint64_t iova; + struct list_head list; }; struct msm_gem_object { @@ -74,7 +76,7 @@ struct msm_gem_object { struct sg_table *sgt; void *vaddr; - struct msm_gem_vma domain[NUM_DOMAINS]; + struct list_head domains; /* normally (resv == &_resv) except for imported bo's */ struct reservation_object *resv; diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c index 1847f83b1e33..f277f075d704 100644 --- a/drivers/gpu/drm/msm/msm_gem_submit.c +++ b/drivers/gpu/drm/msm/msm_gem_submit.c @@ -141,7 +141,7 @@ static void submit_unlock_unpin_bo(struct msm_gem_submit *submit, int i) struct msm_gem_object *msm_obj = submit->bos[i].obj; if (submit->bos[i].flags & BO_PINNED) - msm_gem_put_iova(&msm_obj->base, submit->gpu->id); + msm_gem_put_iova(&msm_obj->base, submit->gpu->aspace); if (submit->bos[i].flags & BO_LOCKED) ww_mutex_unlock(&msm_obj->resv->lock); @@ -180,7 +180,7 @@ retry: /* if locking succeeded, pin bo: */ ret = msm_gem_get_iova_locked(&msm_obj->base, - submit->gpu->id, &iova); + submit->gpu->aspace, &iova); /* this would break the logic in the fail path.. there is no * reason for this to happen, but just to be on the safe side diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 1dc82145d3d0..9881461d7d24 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -474,7 +474,7 @@ static void retire_worker(struct work_struct *work) (obj->write_fence <= fence)) { /* move to inactive: */ msm_gem_move_to_inactive(&obj->base); - msm_gem_put_iova(&obj->base, gpu->id); + msm_gem_put_iova(&obj->base, gpu->aspace); drm_gem_object_unreference(&obj->base); } else { break; @@ -533,7 +533,7 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, /* ring takes a reference to the bo and iova: */ drm_gem_object_reference(&msm_obj->base); msm_gem_get_iova_locked(&msm_obj->base, - submit->gpu->id, &iova); + submit->gpu->aspace, &iova); } if (submit->bos[i].flags & MSM_SUBMIT_BO_READ) @@ -667,8 +667,6 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev, } else { dev_info(drm->dev, "%s: no IOMMU, fallback to VRAM carveout!\n", name); } - gpu->id = msm_register_address_space(drm, gpu->aspace); - /* Create ringbuffer: */ mutex_lock(&drm->struct_mutex); @@ -707,7 +705,7 @@ void msm_gpu_cleanup(struct msm_gpu *gpu) if (gpu->rb) { if (gpu->rb_iova) - msm_gem_put_iova(gpu->rb->bo, gpu->id); + msm_gem_put_iova(gpu->rb->bo, gpu->aspace); msm_ringbuffer_destroy(gpu->rb); } } diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index 401f74075485..206814f33ca9 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -97,7 +97,6 @@ struct msm_gpu { int irq; struct msm_gem_address_space *aspace; - int id; /* Power Control: */ struct regulator *gpu_reg, *gpu_cx; diff --git a/drivers/gpu/drm/msm/sde/sde_connector.c b/drivers/gpu/drm/msm/sde/sde_connector.c index ac9997c238cd..6e27ac09c7a4 100644 --- a/drivers/gpu/drm/msm/sde/sde_connector.c +++ b/drivers/gpu/drm/msm/sde/sde_connector.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2016, The Linux Foundation. All rights reserved. +/* Copyright (c) 2016-2017, The Linux Foundation. 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 version 2 and @@ -88,8 +88,7 @@ static void _sde_connector_destroy_fb(struct sde_connector *c_conn, return; } - msm_framebuffer_cleanup(c_state->out_fb, - c_state->mmu_id); + msm_framebuffer_cleanup(c_state->out_fb, c_state->aspace); drm_framebuffer_unreference(c_state->out_fb); c_state->out_fb = NULL; @@ -193,7 +192,7 @@ sde_connector_atomic_duplicate_state(struct drm_connector *connector) if (c_state->out_fb) { drm_framebuffer_reference(c_state->out_fb); rc = msm_framebuffer_prepare(c_state->out_fb, - c_state->mmu_id); + c_state->aspace); if (rc) SDE_ERROR("failed to prepare fb, %d\n", rc); } @@ -241,14 +240,14 @@ static int sde_connector_atomic_set_property(struct drm_connector *connector, rc = -EFAULT; } else { if (c_state->out_fb->flags & DRM_MODE_FB_SECURE) - c_state->mmu_id = - c_conn->mmu_id[SDE_IOMMU_DOMAIN_SECURE]; + c_state->aspace = + c_conn->aspace[SDE_IOMMU_DOMAIN_SECURE]; else - c_state->mmu_id = - c_conn->mmu_id[SDE_IOMMU_DOMAIN_UNSECURE]; + c_state->aspace = + c_conn->aspace[SDE_IOMMU_DOMAIN_UNSECURE]; rc = msm_framebuffer_prepare(c_state->out_fb, - c_state->mmu_id); + c_state->aspace); if (rc) SDE_ERROR("prep fb failed, %d\n", rc); } @@ -492,18 +491,17 @@ struct drm_connector *sde_connector_init(struct drm_device *dev, c_conn->panel = panel; c_conn->display = display; - /* cache mmu_id's for later */ sde_kms = to_sde_kms(priv->kms); if (sde_kms->vbif[VBIF_NRT]) { - c_conn->mmu_id[SDE_IOMMU_DOMAIN_UNSECURE] = - sde_kms->mmu_id[MSM_SMMU_DOMAIN_NRT_UNSECURE]; - c_conn->mmu_id[SDE_IOMMU_DOMAIN_SECURE] = - sde_kms->mmu_id[MSM_SMMU_DOMAIN_NRT_SECURE]; + c_conn->aspace[SDE_IOMMU_DOMAIN_UNSECURE] = + sde_kms->aspace[MSM_SMMU_DOMAIN_NRT_UNSECURE]; + c_conn->aspace[SDE_IOMMU_DOMAIN_SECURE] = + sde_kms->aspace[MSM_SMMU_DOMAIN_NRT_SECURE]; } else { - c_conn->mmu_id[SDE_IOMMU_DOMAIN_UNSECURE] = - sde_kms->mmu_id[MSM_SMMU_DOMAIN_UNSECURE]; - c_conn->mmu_id[SDE_IOMMU_DOMAIN_SECURE] = - sde_kms->mmu_id[MSM_SMMU_DOMAIN_SECURE]; + c_conn->aspace[SDE_IOMMU_DOMAIN_UNSECURE] = + sde_kms->aspace[MSM_SMMU_DOMAIN_UNSECURE]; + c_conn->aspace[SDE_IOMMU_DOMAIN_SECURE] = + sde_kms->aspace[MSM_SMMU_DOMAIN_SECURE]; } if (ops) diff --git a/drivers/gpu/drm/msm/sde/sde_connector.h b/drivers/gpu/drm/msm/sde/sde_connector.h index 9580282291db..7d86528c81bc 100644 --- a/drivers/gpu/drm/msm/sde/sde_connector.h +++ b/drivers/gpu/drm/msm/sde/sde_connector.h @@ -1,4 +1,4 @@ -/* Copyright (c) 2016, The Linux Foundation. All rights reserved. +/* Copyright (c) 2016-2017, The Linux Foundation. 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 version 2 and @@ -140,7 +140,7 @@ struct sde_connector { struct drm_panel *panel; void *display; - int mmu_id[SDE_IOMMU_DOMAIN_MAX]; + struct msm_gem_address_space *aspace[SDE_IOMMU_DOMAIN_MAX]; char name[SDE_CONNECTOR_NAME_SIZE]; @@ -195,13 +195,13 @@ struct sde_connector { * struct sde_connector_state - private connector status structure * @base: Base drm connector structure * @out_fb: Pointer to output frame buffer, if applicable - * @mmu_id: MMU ID for accessing frame buffer objects, if applicable + * @aspace: Address space for accessing frame buffer objects, if applicable * @property_values: Local cache of current connector property values */ struct sde_connector_state { struct drm_connector_state base; struct drm_framebuffer *out_fb; - int mmu_id; + struct msm_gem_address_space *aspace; uint64_t property_values[CONNECTOR_PROP_COUNT]; }; diff --git a/drivers/gpu/drm/msm/sde/sde_encoder_phys.h b/drivers/gpu/drm/msm/sde/sde_encoder_phys.h index ed4b7be34281..2205dd98a927 100644 --- a/drivers/gpu/drm/msm/sde/sde_encoder_phys.h +++ b/drivers/gpu/drm/msm/sde/sde_encoder_phys.h @@ -264,7 +264,7 @@ struct sde_encoder_phys_cmd { * @wb_fmt: Writeback pixel format * @frame_count: Counter of completed writeback operations * @kickoff_count: Counter of issued writeback operations - * @mmu_id: mmu identifier for non-secure/secure domain + * @aspace: address space identifier for non-secure/secure domain * @wb_dev: Pointer to writeback device * @start_time: Start time of writeback latest request * @end_time: End time of writeback latest request @@ -285,7 +285,7 @@ struct sde_encoder_phys_wb { const struct sde_format *wb_fmt; u32 frame_count; u32 kickoff_count; - int mmu_id[SDE_IOMMU_DOMAIN_MAX]; + struct msm_gem_address_space *aspace[SDE_IOMMU_DOMAIN_MAX]; struct sde_wb_device *wb_dev; ktime_t start_time; ktime_t end_time; diff --git a/drivers/gpu/drm/msm/sde/sde_encoder_phys_wb.c b/drivers/gpu/drm/msm/sde/sde_encoder_phys_wb.c index 9943e3906df0..9368c4974126 100644 --- a/drivers/gpu/drm/msm/sde/sde_encoder_phys_wb.c +++ b/drivers/gpu/drm/msm/sde/sde_encoder_phys_wb.c @@ -180,7 +180,8 @@ static void sde_encoder_phys_wb_setup_fb(struct sde_encoder_phys *phys_enc, struct sde_hw_wb *hw_wb; struct sde_hw_wb_cfg *wb_cfg; const struct msm_format *format; - int ret, mmu_id; + int ret; + struct msm_gem_address_space *aspace; if (!phys_enc) { SDE_ERROR("invalid encoder\n"); @@ -193,9 +194,9 @@ static void sde_encoder_phys_wb_setup_fb(struct sde_encoder_phys *phys_enc, wb_cfg->intf_mode = phys_enc->intf_mode; wb_cfg->is_secure = (fb->flags & DRM_MODE_FB_SECURE) ? true : false; - mmu_id = (wb_cfg->is_secure) ? - wb_enc->mmu_id[SDE_IOMMU_DOMAIN_SECURE] : - wb_enc->mmu_id[SDE_IOMMU_DOMAIN_UNSECURE]; + aspace = (wb_cfg->is_secure) ? + wb_enc->aspace[SDE_IOMMU_DOMAIN_SECURE] : + wb_enc->aspace[SDE_IOMMU_DOMAIN_UNSECURE]; SDE_DEBUG("[fb_secure:%d]\n", wb_cfg->is_secure); @@ -217,7 +218,7 @@ static void sde_encoder_phys_wb_setup_fb(struct sde_encoder_phys *phys_enc, wb_cfg->roi = *wb_roi; if (hw_wb->caps->features & BIT(SDE_WB_XY_ROI_OFFSET)) { - ret = sde_format_populate_layout(mmu_id, fb, &wb_cfg->dest); + ret = sde_format_populate_layout(aspace, fb, &wb_cfg->dest); if (ret) { SDE_DEBUG("failed to populate layout %d\n", ret); return; @@ -226,7 +227,7 @@ static void sde_encoder_phys_wb_setup_fb(struct sde_encoder_phys *phys_enc, wb_cfg->dest.height = fb->height; wb_cfg->dest.num_planes = wb_cfg->dest.format->num_planes; } else { - ret = sde_format_populate_layout_with_roi(mmu_id, fb, wb_roi, + ret = sde_format_populate_layout_with_roi(aspace, fb, wb_roi, &wb_cfg->dest); if (ret) { /* this error should be detected during atomic_check */ @@ -1017,15 +1018,15 @@ struct sde_encoder_phys *sde_encoder_phys_wb_init( phys_enc = &wb_enc->base; if (p->sde_kms->vbif[VBIF_NRT]) { - wb_enc->mmu_id[SDE_IOMMU_DOMAIN_UNSECURE] = - p->sde_kms->mmu_id[MSM_SMMU_DOMAIN_NRT_UNSECURE]; - wb_enc->mmu_id[SDE_IOMMU_DOMAIN_SECURE] = - p->sde_kms->mmu_id[MSM_SMMU_DOMAIN_NRT_SECURE]; + wb_enc->aspace[SDE_IOMMU_DOMAIN_UNSECURE] = + p->sde_kms->aspace[MSM_SMMU_DOMAIN_NRT_UNSECURE]; + wb_enc->aspace[SDE_IOMMU_DOMAIN_SECURE] = + p->sde_kms->aspace[MSM_SMMU_DOMAIN_NRT_SECURE]; } else { - wb_enc->mmu_id[SDE_IOMMU_DOMAIN_UNSECURE] = - p->sde_kms->mmu_id[MSM_SMMU_DOMAIN_UNSECURE]; - wb_enc->mmu_id[SDE_IOMMU_DOMAIN_SECURE] = - p->sde_kms->mmu_id[MSM_SMMU_DOMAIN_SECURE]; + wb_enc->aspace[SDE_IOMMU_DOMAIN_UNSECURE] = + p->sde_kms->aspace[MSM_SMMU_DOMAIN_UNSECURE]; + wb_enc->aspace[SDE_IOMMU_DOMAIN_SECURE] = + p->sde_kms->aspace[MSM_SMMU_DOMAIN_SECURE]; } hw_mdp = sde_rm_get_mdp(&p->sde_kms->rm); diff --git a/drivers/gpu/drm/msm/sde/sde_formats.c b/drivers/gpu/drm/msm/sde/sde_formats.c index 41180f5dec12..42bbbdcab2c9 100644 --- a/drivers/gpu/drm/msm/sde/sde_formats.c +++ b/drivers/gpu/drm/msm/sde/sde_formats.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2015-2016, The Linux Foundation. All rights reserved. +/* Copyright (c) 2015-2017, The Linux Foundation. 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 version 2 and @@ -630,7 +630,7 @@ static int _sde_format_get_plane_sizes( } static int _sde_format_populate_addrs_ubwc( - int mmu_id, + struct msm_gem_address_space *aspace, struct drm_framebuffer *fb, struct sde_hw_fmt_layout *layout) { @@ -641,7 +641,7 @@ static int _sde_format_populate_addrs_ubwc( return -EINVAL; } - base_addr = msm_framebuffer_iova(fb, mmu_id, 0); + base_addr = msm_framebuffer_iova(fb, aspace, 0); if (!base_addr) { DRM_ERROR("failed to retrieve base addr\n"); return -EFAULT; @@ -711,7 +711,7 @@ static int _sde_format_populate_addrs_ubwc( } static int _sde_format_populate_addrs_linear( - int mmu_id, + struct msm_gem_address_space *aspace, struct drm_framebuffer *fb, struct sde_hw_fmt_layout *layout) { @@ -728,7 +728,7 @@ static int _sde_format_populate_addrs_linear( /* Populate addresses for simple formats here */ for (i = 0; i < layout->num_planes; ++i) { - layout->plane_addr[i] = msm_framebuffer_iova(fb, mmu_id, i); + layout->plane_addr[i] = msm_framebuffer_iova(fb, aspace, i); if (!layout->plane_addr[i]) { DRM_ERROR("failed to retrieve base addr\n"); return -EFAULT; @@ -739,7 +739,7 @@ static int _sde_format_populate_addrs_linear( } int sde_format_populate_layout( - int mmu_id, + struct msm_gem_address_space *aspace, struct drm_framebuffer *fb, struct sde_hw_fmt_layout *layout) { @@ -770,9 +770,9 @@ int sde_format_populate_layout( /* Populate the addresses given the fb */ if (SDE_FORMAT_IS_UBWC(layout->format)) - ret = _sde_format_populate_addrs_ubwc(mmu_id, fb, layout); + ret = _sde_format_populate_addrs_ubwc(aspace, fb, layout); else - ret = _sde_format_populate_addrs_linear(mmu_id, fb, layout); + ret = _sde_format_populate_addrs_linear(aspace, fb, layout); /* check if anything changed */ if (!ret && !memcmp(plane_addr, layout->plane_addr, sizeof(plane_addr))) @@ -814,14 +814,14 @@ static void _sde_format_calc_offset_linear(struct sde_hw_fmt_layout *source, } int sde_format_populate_layout_with_roi( - int mmu_id, + struct msm_gem_address_space *aspace, struct drm_framebuffer *fb, struct sde_rect *roi, struct sde_hw_fmt_layout *layout) { int ret; - ret = sde_format_populate_layout(mmu_id, fb, layout); + ret = sde_format_populate_layout(aspace, fb, layout); if (ret || !roi) return ret; diff --git a/drivers/gpu/drm/msm/sde/sde_formats.h b/drivers/gpu/drm/msm/sde/sde_formats.h index 5dcdfbb653ed..0de081d619b7 100644 --- a/drivers/gpu/drm/msm/sde/sde_formats.h +++ b/drivers/gpu/drm/msm/sde/sde_formats.h @@ -1,4 +1,4 @@ -/* Copyright (c) 2015-2016, The Linux Foundation. All rights reserved. +/* Copyright (c) 2015-2017, The Linux Foundation. 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 version 2 and @@ -14,6 +14,7 @@ #define _SDE_FORMATS_H #include +#include "msm_gem.h" #include "sde_hw_mdss.h" /** @@ -76,7 +77,7 @@ int sde_format_check_modified_format( /** * sde_format_populate_layout - populate the given format layout based on * mmu, fb, and format found in the fb - * @mmu_id: mmu id handle + * @aspace: address space pointer * @fb: framebuffer pointer * @fmtl: format layout structure to populate * @@ -84,14 +85,14 @@ int sde_format_check_modified_format( * are the same as before or 0 if new addresses were populated */ int sde_format_populate_layout( - int mmu_id, + struct msm_gem_address_space *aspace, struct drm_framebuffer *fb, struct sde_hw_fmt_layout *fmtl); /** * sde_format_populate_layout_with_roi - populate the given format layout * based on mmu, fb, roi, and format found in the fb - * @mmu_id: mmu id handle + * @aspace: mmu id handle * @fb: framebuffer pointer * @roi: region of interest (optional) * @fmtl: format layout structure to populate @@ -99,7 +100,7 @@ int sde_format_populate_layout( * Return: error code on failure, 0 on success */ int sde_format_populate_layout_with_roi( - int mmu_id, + struct msm_gem_address_space *aspace, struct drm_framebuffer *fb, struct sde_rect *roi, struct sde_hw_fmt_layout *fmtl); diff --git a/drivers/gpu/drm/msm/sde/sde_kms.c b/drivers/gpu/drm/msm/sde/sde_kms.c index e40e33f11fd9..e906b66dcd95 100644 --- a/drivers/gpu/drm/msm/sde/sde_kms.c +++ b/drivers/gpu/drm/msm/sde/sde_kms.c @@ -940,17 +940,17 @@ static int _sde_kms_mmu_destroy(struct sde_kms *sde_kms) struct msm_mmu *mmu; int i; - for (i = ARRAY_SIZE(sde_kms->mmu_id) - 1; i >= 0; i--) { - mmu = sde_kms->aspace[i]->mmu; - - if (!mmu) + for (i = ARRAY_SIZE(sde_kms->aspace) - 1; i >= 0; i--) { + if (!sde_kms->aspace[i]) continue; + mmu = sde_kms->aspace[i]->mmu; + mmu->funcs->detach(mmu, (const char **)iommu_ports, ARRAY_SIZE(iommu_ports)); msm_gem_address_space_destroy(sde_kms->aspace[i]); - sde_kms->mmu_id[i] = 0; + sde_kms->aspace[i] = NULL; } return 0; @@ -991,17 +991,6 @@ static int _sde_kms_mmu_init(struct sde_kms *sde_kms) goto fail; } - sde_kms->mmu_id[i] = msm_register_address_space(sde_kms->dev, - aspace); - if (sde_kms->mmu_id[i] < 0) { - ret = sde_kms->mmu_id[i]; - SDE_ERROR("failed to register sde iommu %d: %d\n", - i, ret); - mmu->funcs->detach(mmu, (const char **)iommu_ports, - ARRAY_SIZE(iommu_ports)); - msm_gem_address_space_destroy(aspace); - goto fail; - } } return 0; diff --git a/drivers/gpu/drm/msm/sde/sde_kms.h b/drivers/gpu/drm/msm/sde/sde_kms.h index 8663c632699b..cd3722fa7df8 100644 --- a/drivers/gpu/drm/msm/sde/sde_kms.h +++ b/drivers/gpu/drm/msm/sde/sde_kms.h @@ -123,7 +123,6 @@ struct sde_kms { struct sde_mdss_cfg *catalog; struct msm_gem_address_space *aspace[MSM_SMMU_DOMAIN_MAX]; - int mmu_id[MSM_SMMU_DOMAIN_MAX]; struct sde_power_client *core_client; /* directory entry for debugfs */ diff --git a/drivers/gpu/drm/msm/sde/sde_plane.c b/drivers/gpu/drm/msm/sde/sde_plane.c index b3de45302dea..114acfd7a173 100644 --- a/drivers/gpu/drm/msm/sde/sde_plane.c +++ b/drivers/gpu/drm/msm/sde/sde_plane.c @@ -86,7 +86,7 @@ enum sde_plane_qos { struct sde_plane { struct drm_plane base; - int mmu_id; + struct msm_gem_address_space *aspace; struct mutex lock; @@ -580,7 +580,7 @@ static inline void _sde_plane_set_scanout(struct drm_plane *plane, return; } - ret = sde_format_populate_layout(psde->mmu_id, fb, &pipe_cfg->layout); + ret = sde_format_populate_layout(psde->aspace, fb, &pipe_cfg->layout); if (ret == -EAGAIN) SDE_DEBUG_PLANE(psde, "not updating same src addrs\n"); else if (ret) @@ -1285,7 +1285,7 @@ static int sde_plane_prepare_fb(struct drm_plane *plane, return 0; SDE_DEBUG_PLANE(psde, "FB[%u]\n", fb->base.id); - return msm_framebuffer_prepare(fb, psde->mmu_id); + return msm_framebuffer_prepare(fb, psde->aspace); } static void sde_plane_cleanup_fb(struct drm_plane *plane, @@ -1294,11 +1294,11 @@ static void sde_plane_cleanup_fb(struct drm_plane *plane, struct drm_framebuffer *fb = old_state ? old_state->fb : NULL; struct sde_plane *psde = plane ? to_sde_plane(plane) : NULL; - if (!fb) + if (!fb || !psde) return; SDE_DEBUG_PLANE(psde, "FB[%u]\n", fb->base.id); - msm_framebuffer_cleanup(fb, psde->mmu_id); + msm_framebuffer_cleanup(fb, psde->aspace); } static void _sde_plane_atomic_check_mode_changed(struct sde_plane *psde, @@ -2384,7 +2384,7 @@ struct drm_plane *sde_plane_init(struct drm_device *dev, /* cache local stuff for later */ plane = &psde->base; psde->pipe = pipe; - psde->mmu_id = kms->mmu_id[MSM_SMMU_DOMAIN_UNSECURE]; + psde->aspace = kms->aspace[MSM_SMMU_DOMAIN_UNSECURE]; /* initialize underlying h/w driver */ psde->pipe_hw = sde_hw_sspp_init(pipe, kms->mmio, kms->catalog); From 758a96f0b825a25cabb3bd074864c9d850722f3a Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:17 -0700 Subject: [PATCH 08/19] drm/msm: Use 64 bit containers for iovas Newer Adreno GPUs are able to support 64 bit virtual addressing. To prepare for this brave new world switch all IOVA related variables and members to a uint64_t container. This is harmless for the display and older targets that do not have 64 bit addressing because a 32 bit address is just a 64 bit address with lots of zeros. To avoid ambiguity and compiler oddness make sure to use lower_32_bits() and upper_32_bits() everywhere the IOVA is used to ensure that you get what you expect. Change-Id: Ic0dedbad23322fae32509c1f4d75d9b4e2863081 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 4 ++-- drivers/gpu/drm/msm/adreno/adreno_gpu.h | 2 +- drivers/gpu/drm/msm/dsi-staging/dsi_ctrl.c | 4 ++-- drivers/gpu/drm/msm/dsi/dsi.h | 6 +++--- drivers/gpu/drm/msm/dsi/dsi_host.c | 11 ++++++----- drivers/gpu/drm/msm/dsi/dsi_manager.c | 4 ++-- drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c | 12 +++++++----- drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h | 2 +- drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c | 3 ++- drivers/gpu/drm/msm/msm_drv.h | 6 +++--- drivers/gpu/drm/msm/msm_fb.c | 13 ++++++++++--- drivers/gpu/drm/msm/msm_fbdev.c | 7 ++++--- drivers/gpu/drm/msm/msm_gem.c | 8 ++++---- drivers/gpu/drm/msm/msm_gem.h | 4 ++-- drivers/gpu/drm/msm/msm_gem_submit.c | 9 +++++---- drivers/gpu/drm/msm/msm_gpu.c | 2 +- drivers/gpu/drm/msm/msm_gpu.h | 2 +- drivers/gpu/drm/msm/msm_iommu.c | 14 +++++++------- drivers/gpu/drm/msm/msm_mmu.h | 4 ++-- drivers/gpu/drm/msm/msm_rd.c | 6 +++--- drivers/gpu/drm/msm/msm_smmu.c | 12 ++++++------ 21 files changed, 74 insertions(+), 61 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 392622953df9..f6871b54c24a 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -2,7 +2,7 @@ * Copyright (C) 2013 Red Hat * Author: Rob Clark * - * Copyright (c) 2014 The Linux Foundation. All rights reserved. + * Copyright (c) 2014,2016 The Linux Foundation. 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 version 2 as published by @@ -156,7 +156,7 @@ int adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, case MSM_SUBMIT_CMD_BUF: OUT_PKT3(ring, adreno_is_a430(adreno_gpu) ? CP_INDIRECT_BUFFER_PFE : CP_INDIRECT_BUFFER_PFD, 2); - OUT_RING(ring, submit->cmd[i].iova); + OUT_RING(ring, lower_32_bits(submit->cmd[i].iova)); OUT_RING(ring, submit->cmd[i].size); ibs++; break; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index 67ceeb00f98e..7d3553ac5224 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -103,7 +103,7 @@ struct adreno_gpu { // different for z180.. struct adreno_rbmemptrs *memptrs; struct drm_gem_object *memptrs_bo; - uint32_t memptrs_iova; + uint64_t memptrs_iova; /* * Register offsets are different between some GPUs. diff --git a/drivers/gpu/drm/msm/dsi-staging/dsi_ctrl.c b/drivers/gpu/drm/msm/dsi-staging/dsi_ctrl.c index b8520aadbc0c..fb9617ce572a 100644 --- a/drivers/gpu/drm/msm/dsi-staging/dsi_ctrl.c +++ b/drivers/gpu/drm/msm/dsi-staging/dsi_ctrl.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2016, The Linux Foundation. All rights reserved. + * Copyright (c) 2016-2017, The Linux Foundation. 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 version 2 and @@ -1186,7 +1186,7 @@ static int dsi_ctrl_buffer_deinit(struct dsi_ctrl *dsi_ctrl) int dsi_ctrl_buffer_init(struct dsi_ctrl *dsi_ctrl) { int rc = 0; - u32 iova = 0; + u64 iova = 0; dsi_ctrl->tx_cmd_buf = msm_gem_new(dsi_ctrl->drm_dev, SZ_4K, diff --git a/drivers/gpu/drm/msm/dsi/dsi.h b/drivers/gpu/drm/msm/dsi/dsi.h index 5f5a3732cdf6..4cb4764e7492 100644 --- a/drivers/gpu/drm/msm/dsi/dsi.h +++ b/drivers/gpu/drm/msm/dsi/dsi.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2015, The Linux Foundation. All rights reserved. + * Copyright (c) 2015,2017, The Linux Foundation. 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 version 2 and @@ -89,7 +89,7 @@ int msm_dsi_manager_phy_enable(int id, u32 *clk_pre, u32 *clk_post); void msm_dsi_manager_phy_disable(int id); int msm_dsi_manager_cmd_xfer(int id, const struct mipi_dsi_msg *msg); -bool msm_dsi_manager_cmd_xfer_trigger(int id, u32 iova, u32 len); +bool msm_dsi_manager_cmd_xfer_trigger(int id, u64 iova, u32 len); int msm_dsi_manager_register(struct msm_dsi *msm_dsi); void msm_dsi_manager_unregister(struct msm_dsi *msm_dsi); @@ -143,7 +143,7 @@ int msm_dsi_host_cmd_tx(struct mipi_dsi_host *host, int msm_dsi_host_cmd_rx(struct mipi_dsi_host *host, const struct mipi_dsi_msg *msg); void msm_dsi_host_cmd_xfer_commit(struct mipi_dsi_host *host, - u32 iova, u32 len); + u64 iova, u32 len); int msm_dsi_host_enable(struct mipi_dsi_host *host); int msm_dsi_host_disable(struct mipi_dsi_host *host); int msm_dsi_host_power_on(struct mipi_dsi_host *host); diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c index 4c49868efcda..4580a6e3c877 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_host.c +++ b/drivers/gpu/drm/msm/dsi/dsi_host.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2015, The Linux Foundation. All rights reserved. + * Copyright (c) 2015,2017 The Linux Foundation. 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 version 2 and @@ -836,7 +836,7 @@ static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size) { struct drm_device *dev = msm_host->dev; int ret; - u32 iova; + u64 iova; mutex_lock(&dev->struct_mutex); msm_host->tx_gem_obj = msm_gem_new(dev, size, MSM_BO_UNCACHED); @@ -974,7 +974,7 @@ static int dsi_long_read_resp(u8 *buf, const struct mipi_dsi_msg *msg) static int dsi_cmd_dma_tx(struct msm_dsi_host *msm_host, int len) { int ret; - u32 iova; + uint64_t iova; bool triggered; ret = msm_gem_get_iova(msm_host->tx_gem_obj, 0, &iova); @@ -1750,11 +1750,12 @@ int msm_dsi_host_cmd_rx(struct mipi_dsi_host *host, return ret; } -void msm_dsi_host_cmd_xfer_commit(struct mipi_dsi_host *host, u32 iova, u32 len) +void msm_dsi_host_cmd_xfer_commit(struct mipi_dsi_host *host, u64 iova, u32 len) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); - dsi_write(msm_host, REG_DSI_DMA_BASE, iova); + /* FIXME: Verify that the iova < 32 bits? */ + dsi_write(msm_host, REG_DSI_DMA_BASE, lower_32_bits(iova)); dsi_write(msm_host, REG_DSI_DMA_LEN, len); dsi_write(msm_host, REG_DSI_TRIG_DMA, 1); diff --git a/drivers/gpu/drm/msm/dsi/dsi_manager.c b/drivers/gpu/drm/msm/dsi/dsi_manager.c index 0455ff75074a..2091b748abbb 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_manager.c +++ b/drivers/gpu/drm/msm/dsi/dsi_manager.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2015, The Linux Foundation. All rights reserved. + * Copyright (c) 2015,2017 The Linux Foundation. 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 version 2 and @@ -774,7 +774,7 @@ restore_host0: return ret; } -bool msm_dsi_manager_cmd_xfer_trigger(int id, u32 iova, u32 len) +bool msm_dsi_manager_cmd_xfer_trigger(int id, u64 iova, u32 len) { struct msm_dsi *msm_dsi = dsi_mgr_get_dsi(id); struct msm_dsi *msm_dsi0 = dsi_mgr_get_dsi(DSI_0); diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c index dff43329c020..caad2be87ae4 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c @@ -40,7 +40,7 @@ struct mdp4_crtc { uint32_t x, y; /* next cursor to scan-out: */ - uint32_t next_iova; + uint64_t next_iova; struct drm_gem_object *next_bo; /* current cursor being scanned out: */ @@ -387,7 +387,7 @@ static void update_cursor(struct drm_crtc *crtc) if (mdp4_crtc->cursor.stale) { struct drm_gem_object *next_bo = mdp4_crtc->cursor.next_bo; struct drm_gem_object *prev_bo = mdp4_crtc->cursor.scanout_bo; - uint32_t iova = mdp4_crtc->cursor.next_iova; + uint64_t iova = mdp4_crtc->cursor.next_iova; if (next_bo) { /* take a obj ref + iova ref when we start scanning out: */ @@ -399,14 +399,16 @@ static void update_cursor(struct drm_crtc *crtc) mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_SIZE(dma), MDP4_DMA_CURSOR_SIZE_WIDTH(mdp4_crtc->cursor.width) | MDP4_DMA_CURSOR_SIZE_HEIGHT(mdp4_crtc->cursor.height)); - mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BASE(dma), iova); + /* FIXME: Make sure iova < 32 bits */ + mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BASE(dma), + lower_32_bits(iova)); mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BLEND_CONFIG(dma), MDP4_DMA_CURSOR_BLEND_CONFIG_FORMAT(CURSOR_ARGB) | MDP4_DMA_CURSOR_BLEND_CONFIG_CURSOR_EN); } else { /* disable cursor: */ mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BASE(dma), - mdp4_kms->blank_cursor_iova); + lower_32_bits(mdp4_kms->blank_cursor_iova)); } /* and drop the iova ref + obj rev when done scanning out: */ @@ -433,7 +435,7 @@ static int mdp4_crtc_cursor_set(struct drm_crtc *crtc, struct drm_device *dev = crtc->dev; struct drm_gem_object *cursor_bo, *old_bo; unsigned long flags; - uint32_t iova; + uint64_t iova; int ret; if ((width > CURSOR_WIDTH) || (height > CURSOR_HEIGHT)) { diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h index aa060e71ec91..5cf03e58a4f4 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h @@ -49,7 +49,7 @@ struct mdp4_kms { /* empty/blank cursor bo to use when cursor is "disabled" */ struct drm_gem_object *blank_cursor_bo; - uint32_t blank_cursor_iova; + uint64_t blank_cursor_iova; }; #define to_mdp4_kms(x) container_of(x, struct mdp4_kms, base) diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c index 56c1f6c4ab96..422a9a78f3b2 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c @@ -509,7 +509,8 @@ static int mdp5_crtc_cursor_set(struct drm_crtc *crtc, struct drm_device *dev = crtc->dev; struct mdp5_kms *mdp5_kms = get_kms(crtc); struct drm_gem_object *cursor_bo, *old_bo = NULL; - uint32_t blendcfg, cursor_addr, stride; + uint32_t blendcfg, stride; + uint64_t cursor_addr; int ret, bpp, lm; unsigned int depth; enum mdp5_cursor_alpha cur_alpha = CURSOR_ALPHA_PER_PIXEL; diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h index f59bd5fd4d37..406b40f0f8de 100644 --- a/drivers/gpu/drm/msm/msm_drv.h +++ b/drivers/gpu/drm/msm/msm_drv.h @@ -401,10 +401,10 @@ int msm_gem_mmap(struct file *filp, struct vm_area_struct *vma); int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf); uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj); int msm_gem_get_iova_locked(struct drm_gem_object *obj, - struct msm_gem_address_space *aspace, uint32_t *iova); + struct msm_gem_address_space *aspace, uint64_t *iova); int msm_gem_get_iova(struct drm_gem_object *obj, - struct msm_gem_address_space *aspace, uint32_t *iova); -uint32_t msm_gem_iova(struct drm_gem_object *obj, + struct msm_gem_address_space *aspace, uint64_t *iova); +uint64_t msm_gem_iova(struct drm_gem_object *obj, struct msm_gem_address_space *aspace); struct page **msm_gem_get_pages(struct drm_gem_object *obj); void msm_gem_put_pages(struct drm_gem_object *obj); diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c index 28395ea13235..a3f0392c2f88 100644 --- a/drivers/gpu/drm/msm/msm_fb.c +++ b/drivers/gpu/drm/msm/msm_fb.c @@ -97,11 +97,11 @@ int msm_framebuffer_prepare(struct drm_framebuffer *fb, { struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb); int ret, i, n = drm_format_num_planes(fb->pixel_format); - uint32_t iova; + uint64_t iova; for (i = 0; i < n; i++) { ret = msm_gem_get_iova(msm_fb->planes[i], aspace, &iova); - DBG("FB[%u]: iova[%d]: %08x (%d)", fb->base.id, i, iova, ret); + DBG("FB[%u]: iova[%d]: %08llx (%d)", fb->base.id, i, iova, ret); if (ret) return ret; } @@ -119,13 +119,20 @@ void msm_framebuffer_cleanup(struct drm_framebuffer *fb, msm_gem_put_iova(msm_fb->planes[i], aspace); } +/* FIXME: Leave this as a uint32_t and just return the lower 32 bits? */ uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb, struct msm_gem_address_space *aspace, int plane) { struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb); + uint64_t iova; + if (!msm_fb->planes[plane]) return 0; - return msm_gem_iova(msm_fb->planes[plane], aspace) + fb->offsets[plane]; + + iova = msm_gem_iova(msm_fb->planes[plane], aspace) + fb->offsets[plane]; + + /* FIXME: Make sure it is < 32 bits */ + return lower_32_bits(iova); } struct drm_gem_object *msm_framebuffer_bo(struct drm_framebuffer *fb, int plane) diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c index 3f6ec077b51d..cf127250c0d0 100644 --- a/drivers/gpu/drm/msm/msm_fbdev.c +++ b/drivers/gpu/drm/msm/msm_fbdev.c @@ -85,7 +85,7 @@ static int msm_fbdev_create(struct drm_fb_helper *helper, struct drm_framebuffer *fb = NULL; struct fb_info *fbi = NULL; struct drm_mode_fb_cmd2 mode_cmd = {0}; - uint32_t paddr; + uint64_t paddr; int ret, size; DBG("create fbdev: %dx%d@%d (%dx%d)", sizes->surface_width, @@ -160,11 +160,12 @@ static int msm_fbdev_create(struct drm_fb_helper *helper, drm_fb_helper_fill_fix(fbi, fb->pitches[0], fb->depth); drm_fb_helper_fill_var(fbi, helper, sizes->fb_width, sizes->fb_height); - dev->mode_config.fb_base = paddr; + /* FIXME: Verify paddr < 32 bits? */ + dev->mode_config.fb_base = lower_32_bits(paddr); fbi->screen_base = msm_gem_vaddr_locked(fbdev->bo); fbi->screen_size = fbdev->bo->size; - fbi->fix.smem_start = paddr; + fbi->fix.smem_start = lower_32_bits(paddr); fbi->fix.smem_len = fbdev->bo->size; DBG("par=%p, %dx%d", fbi->par, fbi->var.xres, fbi->var.yres); diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index ca24cc433afc..f42c7bff18fb 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -338,7 +338,7 @@ static struct msm_gem_vma *obj_get_domain(struct drm_gem_object *obj, * the refcnt counter needs to be atomic_t. */ int msm_gem_get_iova_locked(struct drm_gem_object *obj, - struct msm_gem_address_space *aspace, uint32_t *iova) + struct msm_gem_address_space *aspace, uint64_t *iova) { struct msm_gem_object *msm_obj = to_msm_bo(obj); struct page **pages; @@ -351,7 +351,7 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, if (IS_ERR(pages)) return PTR_ERR(pages); - *iova = physaddr(obj); + *iova = (uint64_t) physaddr(obj); return 0; } @@ -382,7 +382,7 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, /* get iova, taking a reference. Should have a matching put */ int msm_gem_get_iova(struct drm_gem_object *obj, - struct msm_gem_address_space *aspace, uint32_t *iova) + struct msm_gem_address_space *aspace, uint64_t *iova) { struct msm_gem_vma *domain; int ret; @@ -402,7 +402,7 @@ int msm_gem_get_iova(struct drm_gem_object *obj, /* get iova without taking a reference, used in places where you have * already done a 'msm_gem_get_iova()'. */ -uint32_t msm_gem_iova(struct drm_gem_object *obj, +uint64_t msm_gem_iova(struct drm_gem_object *obj, struct msm_gem_address_space *aspace) { struct msm_gem_vma *domain = obj_get_domain(obj, aspace); diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h index 937e80c12554..d9c3977434d9 100644 --- a/drivers/gpu/drm/msm/msm_gem.h +++ b/drivers/gpu/drm/msm/msm_gem.h @@ -127,13 +127,13 @@ struct msm_gem_submit { struct { uint32_t type; uint32_t size; /* in dwords */ - uint32_t iova; + uint64_t iova; uint32_t idx; /* cmdstream buffer idx in bos[] */ } cmd[MAX_CMDS]; struct { uint32_t flags; struct msm_gem_object *obj; - uint32_t iova; + uint64_t iova; } bos[0]; }; diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c index f277f075d704..4fdf3a793814 100644 --- a/drivers/gpu/drm/msm/msm_gem_submit.c +++ b/drivers/gpu/drm/msm/msm_gem_submit.c @@ -162,7 +162,7 @@ retry: for (i = 0; i < submit->nr_bos; i++) { struct msm_gem_object *msm_obj = submit->bos[i].obj; - uint32_t iova; + uint64_t iova; if (slow_locked == i) slow_locked = -1; @@ -229,7 +229,7 @@ fail: } static int submit_bo(struct msm_gem_submit *submit, uint32_t idx, - struct msm_gem_object **obj, uint32_t *iova, bool *valid) + struct msm_gem_object **obj, uint64_t *iova, bool *valid) { if (idx >= submit->nr_bos) { DRM_ERROR("invalid buffer index: %u (out of %u)\n", @@ -275,7 +275,8 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob struct drm_msm_gem_submit_reloc submit_reloc; void __user *userptr = to_user_ptr(relocs + (i * sizeof(submit_reloc))); - uint32_t iova, off; + uint64_t iova; + uint32_t off; bool valid; ret = copy_from_user(&submit_reloc, userptr, sizeof(submit_reloc)); @@ -376,7 +377,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data, void __user *userptr = to_user_ptr(args->cmds + (i * sizeof(submit_cmd))); struct msm_gem_object *msm_obj; - uint32_t iova; + uint64_t iova; ret = copy_from_user(&submit_cmd, userptr, sizeof(submit_cmd)); if (ret) { diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 9881461d7d24..19c08ac07632 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -528,7 +528,7 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, WARN_ON(is_active(msm_obj) && (msm_obj->gpu != gpu)); if (!is_active(msm_obj)) { - uint32_t iova; + uint64_t iova; /* ring takes a reference to the bo and iova: */ drm_gem_object_reference(&msm_obj->base); diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index 206814f33ca9..5e6d137140d0 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -79,7 +79,7 @@ struct msm_gpu { uint32_t num_perfcntrs; struct msm_ringbuffer *rb; - uint32_t rb_iova; + uint64_t rb_iova; /* list of GEM active objects: */ struct list_head active_list; diff --git a/drivers/gpu/drm/msm/msm_iommu.c b/drivers/gpu/drm/msm/msm_iommu.c index 7e64a6255c11..253b85b917c3 100644 --- a/drivers/gpu/drm/msm/msm_iommu.c +++ b/drivers/gpu/drm/msm/msm_iommu.c @@ -27,7 +27,7 @@ struct msm_iommu { static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev, unsigned long iova, int flags, void *arg) { - pr_warn_ratelimited("*** fault: iova=%08lx, flags=%d\n", iova, flags); + pr_warn_ratelimited("*** fault: iova=%16llX, flags=%d\n", (u64) iova, flags); return 0; } @@ -43,13 +43,13 @@ static void msm_iommu_detach(struct msm_mmu *mmu, const char **names, int cnt) iommu_detach_device(iommu->domain, mmu->dev); } -static int msm_iommu_map(struct msm_mmu *mmu, uint32_t iova, +static int msm_iommu_map(struct msm_mmu *mmu, uint64_t iova, struct sg_table *sgt, int prot) { struct msm_iommu *iommu = to_msm_iommu(mmu); struct iommu_domain *domain = iommu->domain; struct scatterlist *sg; - unsigned int da = iova; + uint64_t da = iova; unsigned int i, j; int ret; @@ -60,7 +60,7 @@ static int msm_iommu_map(struct msm_mmu *mmu, uint32_t iova, phys_addr_t pa = sg_phys(sg) - sg->offset; size_t bytes = sg->length + sg->offset; - VERB("map[%d]: %08x %pa(%zx)", i, iova, &pa, bytes); + VERB("map[%d]: %016llx %pa(%zx)", i, iova, &pa, bytes); ret = iommu_map(domain, da, pa, bytes, prot); if (ret) @@ -82,13 +82,13 @@ fail: return ret; } -static int msm_iommu_unmap(struct msm_mmu *mmu, uint32_t iova, +static int msm_iommu_unmap(struct msm_mmu *mmu, uint64_t iova, struct sg_table *sgt) { struct msm_iommu *iommu = to_msm_iommu(mmu); struct iommu_domain *domain = iommu->domain; struct scatterlist *sg; - unsigned int da = iova; + uint64_t da = iova; int i; for_each_sg(sgt->sgl, sg, sgt->nents, i) { @@ -99,7 +99,7 @@ static int msm_iommu_unmap(struct msm_mmu *mmu, uint32_t iova, if (unmapped < bytes) return unmapped; - VERB("unmap[%d]: %08x(%zx)", i, iova, bytes); + VERB("unmap[%d]: %016llx(%zx)", i, iova, bytes); BUG_ON(!PAGE_ALIGNED(bytes)); diff --git a/drivers/gpu/drm/msm/msm_mmu.h b/drivers/gpu/drm/msm/msm_mmu.h index 3d750eca5cae..8464b9e04b63 100644 --- a/drivers/gpu/drm/msm/msm_mmu.h +++ b/drivers/gpu/drm/msm/msm_mmu.h @@ -33,9 +33,9 @@ enum msm_mmu_domain_type { struct msm_mmu_funcs { int (*attach)(struct msm_mmu *mmu, const char **names, int cnt); void (*detach)(struct msm_mmu *mmu, const char **names, int cnt); - int (*map)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt, + int (*map)(struct msm_mmu *mmu, uint64_t iova, struct sg_table *sgt, int prot); - int (*unmap)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt); + int (*unmap)(struct msm_mmu *mmu, uint64_t iova, struct sg_table *sgt); int (*map_sg)(struct msm_mmu *mmu, struct sg_table *sgt, enum dma_data_direction dir); void (*unmap_sg)(struct msm_mmu *mmu, struct sg_table *sgt, diff --git a/drivers/gpu/drm/msm/msm_rd.c b/drivers/gpu/drm/msm/msm_rd.c index 9a78c48817c6..6dbb516cd4dc 100644 --- a/drivers/gpu/drm/msm/msm_rd.c +++ b/drivers/gpu/drm/msm/msm_rd.c @@ -307,7 +307,7 @@ void msm_rd_dump_submit(struct msm_gem_submit *submit) for (i = 0; i < submit->nr_cmds; i++) { uint32_t idx = submit->cmd[i].idx; - uint32_t iova = submit->cmd[i].iova; + uint64_t iova = submit->cmd[i].iova; uint32_t szd = submit->cmd[i].size; /* in dwords */ struct msm_gem_object *obj = submit->bos[idx].obj; const char *buf = msm_gem_vaddr_locked(&obj->base); @@ -315,7 +315,7 @@ void msm_rd_dump_submit(struct msm_gem_submit *submit) buf += iova - submit->bos[idx].iova; rd_write_section(rd, RD_GPUADDR, - (uint32_t[2]){ iova, szd * 4 }, 8); + (uint64_t[2]) { iova, szd * 4 }, 16); rd_write_section(rd, RD_BUFFER_CONTENTS, buf, szd * 4); @@ -329,7 +329,7 @@ void msm_rd_dump_submit(struct msm_gem_submit *submit) case MSM_SUBMIT_CMD_CTX_RESTORE_BUF: case MSM_SUBMIT_CMD_BUF: rd_write_section(rd, RD_CMDSTREAM_ADDR, - (uint32_t[2]){ iova, szd }, 8); + (uint64_t[2]) { iova, szd }, 16); break; } } diff --git a/drivers/gpu/drm/msm/msm_smmu.c b/drivers/gpu/drm/msm/msm_smmu.c index e3db6690e38a..500e8a5e6247 100644 --- a/drivers/gpu/drm/msm/msm_smmu.c +++ b/drivers/gpu/drm/msm/msm_smmu.c @@ -104,14 +104,14 @@ static void msm_smmu_detach(struct msm_mmu *mmu, const char **names, int cnt) dev_dbg(client->dev, "iommu domain detached\n"); } -static int msm_smmu_map(struct msm_mmu *mmu, uint32_t iova, +static int msm_smmu_map(struct msm_mmu *mmu, uint64_t iova, struct sg_table *sgt, int prot) { struct msm_smmu *smmu = to_msm_smmu(mmu); struct msm_smmu_client *client = msm_smmu_to_client(smmu); struct iommu_domain *domain; struct scatterlist *sg; - unsigned int da = iova; + uint64_t da = iova; unsigned int i, j; int ret; @@ -126,7 +126,7 @@ static int msm_smmu_map(struct msm_mmu *mmu, uint32_t iova, u32 pa = sg_phys(sg) - sg->offset; size_t bytes = sg->length + sg->offset; - VERB("map[%d]: %08x %08x(%zx)", i, iova, pa, bytes); + VERB("map[%d]: %16llx %08x(%zx)", i, iova, pa, bytes); ret = iommu_map(domain, da, pa, bytes, prot); if (ret) @@ -172,14 +172,14 @@ static void msm_smmu_unmap_sg(struct msm_mmu *mmu, struct sg_table *sgt, dma_unmap_sg(client->dev, sgt->sgl, sgt->nents, dir); } -static int msm_smmu_unmap(struct msm_mmu *mmu, uint32_t iova, +static int msm_smmu_unmap(struct msm_mmu *mmu, uint64_t iova, struct sg_table *sgt) { struct msm_smmu *smmu = to_msm_smmu(mmu); struct msm_smmu_client *client = msm_smmu_to_client(smmu); struct iommu_domain *domain; struct scatterlist *sg; - unsigned int da = iova; + uint64_t da = iova; int i; if (!client) @@ -197,7 +197,7 @@ static int msm_smmu_unmap(struct msm_mmu *mmu, uint32_t iova, if (unmapped < bytes) return unmapped; - VERB("unmap[%d]: %08x(%zx)", i, iova, bytes); + VERB("unmap[%d]: %16llx(%zx)", i, iova, bytes); WARN_ON(!PAGE_ALIGNED(bytes)); From 4629fd01af62bc9a57b5536ff19b397e1111ac65 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:17 -0700 Subject: [PATCH 09/19] drm/msm: Search for the SMMU device in the node sub-devices The upstream DT model presumes that the iommu phandle is in the main GPU device node. The downstream model defines sub-nodes for the IOMMU devices. In order to get the right device to attach the IOMMU domain to we need to do a bit of searching. First, check in the top level node. If that fails, then start walking the port list passed in the attached function - hopefully one of those will have the right device - if found, use that to attach the domain and breathe normally. Change-Id: Ic0dedbad4a1852d2ff204031bd360e0b7960805a Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 3 +- drivers/gpu/drm/msm/msm_iommu.c | 38 ++++++++++++++++++++++++- 2 files changed, 38 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index f6871b54c24a..7afec61cf387 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -344,8 +344,7 @@ void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords) } static const char *iommu_ports[] = { - "gfx3d_user", "gfx3d_priv", - "gfx3d1_user", "gfx3d1_priv", + "gfx3d_user", }; int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, diff --git a/drivers/gpu/drm/msm/msm_iommu.c b/drivers/gpu/drm/msm/msm_iommu.c index 253b85b917c3..b4a5015ff922 100644 --- a/drivers/gpu/drm/msm/msm_iommu.c +++ b/drivers/gpu/drm/msm/msm_iommu.c @@ -15,6 +15,7 @@ * this program. If not, see . */ +#include #include "msm_drv.h" #include "msm_mmu.h" @@ -34,7 +35,42 @@ static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev, static int msm_iommu_attach(struct msm_mmu *mmu, const char **names, int cnt) { struct msm_iommu *iommu = to_msm_iommu(mmu); - return iommu_attach_device(iommu->domain, mmu->dev); + int i; + + /* See if there is a iommus member in the current device. If not, look + * for the names and see if there is one in there. + */ + + if (of_find_property(mmu->dev->of_node, "iommus", NULL)) + return iommu_attach_device(iommu->domain, mmu->dev); + + /* Look through the list of names for a target */ + for (i = 0; i < cnt; i++) { + struct device_node *node = + of_find_node_by_name(mmu->dev->of_node, names[i]); + + if (!node) + continue; + + if (of_find_property(node, "iommus", NULL)) { + struct platform_device *pdev; + + /* Get the platform device for the node */ + of_platform_populate(node->parent, NULL, NULL, + mmu->dev); + + pdev = of_find_device_by_node(node); + + if (!pdev) + continue; + + mmu->dev = &pdev->dev; + return iommu_attach_device(iommu->domain, mmu->dev); + } + } + + dev_err(mmu->dev, "Couldn't find a IOMMU device\n"); + return -ENODEV; } static void msm_iommu_detach(struct msm_mmu *mmu, const char **names, int cnt) From ef41564a3e4af0c3c101fae76f77b3ea65aca5be Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:18 -0700 Subject: [PATCH 10/19] drm/msm: Make sure that MSM_SUBMIT_BO_FLAGS are set For every submission buffer object one of MSM_SUBMIT_BO_WRITE and MSM_SUBMIT_BO_READ must be set (and nothing else). If we allowed zero then the buffer object would never get queued to be unreferenced. Change-Id: Ic0dedbad8a4543428372fe3fef9503cedd1eef6e Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/msm_gem_submit.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c index 4fdf3a793814..0aea6377b449 100644 --- a/drivers/gpu/drm/msm/msm_gem_submit.c +++ b/drivers/gpu/drm/msm/msm_gem_submit.c @@ -90,7 +90,8 @@ static int submit_lookup_objects(struct msm_gem_submit *submit, pagefault_disable(); } - if (submit_bo.flags & ~MSM_SUBMIT_BO_FLAGS) { + if ((submit_bo.flags & ~MSM_SUBMIT_BO_FLAGS) || + !(submit_bo.flags & MSM_SUBMIT_BO_FLAGS)) { DRM_ERROR("invalid flags: %x\n", submit_bo.flags); ret = -EINVAL; goto out_unlock; From 7d080736a4b4601a16a2d81a4537d0202fc05157 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:18 -0700 Subject: [PATCH 11/19] drm/msm: Make sure that WPTR stays in bounds Currently the value written to CP_RB_WPTR is calculated on the fly as (rb->next - rb->start). But as the code is designed rb->next is wrapped before writing the commands so if a series of commands happened to fit perfectly in the ringbuffer, rb->next would end up being equal to rb->size / 4 and thus result in an out of bounds address to CP_RB_WPTR. The easiest way to fix this is to mask WPTR when writing it to the hardware; it makes the hardware happy and the rest of the ringbuffer math appears to work and there isn't any point in messing with things. Change-Id: Ic0dedbade543a2253b88acf46a8f8e062ae7faf9 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 7afec61cf387..46c2fc96a4cb 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -222,7 +222,14 @@ int adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, void adreno_flush(struct msm_gpu *gpu) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); - uint32_t wptr = get_wptr(gpu->rb); + uint32_t wptr; + + /* + * Mask the wptr value that we calculate to fit in the HW range. This is + * to account for the possibility that the last command fit exactly into + * the ringbuffer and rb->next hasn't wrapped to zero yet + */ + wptr = get_wptr(gpu->rb) & ((gpu->rb->size / 4) - 1); /* ensure writes to ringbuffer have hit system memory: */ mb(); From 88de027b795e9dfe640b896295801cf8c094c3bd Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:19 -0700 Subject: [PATCH 12/19] drm/msm: Make sure the GPU is valid before we use it Verify that the GPU is valid before calling any ioctl functions that might need the priv->gpu member. Standardize the return code (-ENXIO). Change-Id: Ic0dedbad015a531b792e0b94cb61e0570b691de8 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/msm_drv.c | 2 +- drivers/gpu/drm/msm/msm_gem_submit.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index 6d3cf1accdb9..e223f376c908 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -942,7 +942,7 @@ int msm_wait_fence(struct drm_device *dev, uint32_t fence, int ret; if (!priv->gpu) - return 0; + return -ENXIO; if (fence > priv->gpu->submitted_fence) { DRM_ERROR("waiting on invalid fence: %u (of %u)\n", diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c index 0aea6377b449..f7b5e30b41eb 100644 --- a/drivers/gpu/drm/msm/msm_gem_submit.c +++ b/drivers/gpu/drm/msm/msm_gem_submit.c @@ -353,6 +353,8 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data, return -EINVAL; gpu = priv->gpu; + if (!gpu) + return -ENXIO; if (args->nr_cmds > MAX_CMDS) return -EINVAL; From 045b6f0aabaa281c7fb1b529355d310561b0d942 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:19 -0700 Subject: [PATCH 13/19] drm/msm: gpu: Add A5XX target support Add support for the A5XX family of Adreno GPUs. Change-Id: Ic0dedbad665ef9d0f2cdb32a33eef9fe3fd7aa5c Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/Makefile | 3 +- drivers/gpu/drm/msm/adreno/a5xx_gpu.c | 861 +++++++++++++++++++++ drivers/gpu/drm/msm/adreno/a5xx_gpu.h | 37 + drivers/gpu/drm/msm/adreno/adreno_device.c | 25 +- drivers/gpu/drm/msm/adreno/adreno_gpu.c | 16 +- drivers/gpu/drm/msm/adreno/adreno_gpu.h | 47 ++ drivers/gpu/drm/msm/msm_gpu.c | 4 +- drivers/gpu/drm/msm/msm_gpu.h | 2 +- 8 files changed, 981 insertions(+), 14 deletions(-) create mode 100644 drivers/gpu/drm/msm/adreno/a5xx_gpu.c create mode 100644 drivers/gpu/drm/msm/adreno/a5xx_gpu.h diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile index 442c5d9711a9..5c587d291198 100644 --- a/drivers/gpu/drm/msm/Makefile +++ b/drivers/gpu/drm/msm/Makefile @@ -52,7 +52,8 @@ ifneq ($(CONFIG_QCOM_KGSL),y) msm_drm-y += adreno/adreno_device.o \ adreno/adreno_gpu.o \ adreno/a3xx_gpu.o \ - adreno/a4xx_gpu.o + adreno/a4xx_gpu.o \ + adreno/a5xx_gpu.o endif msm_drm-$(CONFIG_DRM_MSM_MDP4) += mdp/mdp4/mdp4_crtc.o \ diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c new file mode 100644 index 000000000000..9c2a818620b6 --- /dev/null +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -0,0 +1,861 @@ +/* Copyright (c) 2016-2017 The Linux Foundation. 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 version 2 and + * only 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. + * + */ + +#include "msm_gem.h" +#include "a5xx_gpu.h" + +extern bool hang_debug; +static void a5xx_dump(struct msm_gpu *gpu); + +static int a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, + struct msm_file_private *ctx) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct msm_drm_private *priv = gpu->dev->dev_private; + struct msm_ringbuffer *ring = gpu->rb; + unsigned int i, ibs = 0; + + for (i = 0; i < submit->nr_cmds; i++) { + switch (submit->cmd[i].type) { + case MSM_SUBMIT_CMD_IB_TARGET_BUF: + break; + case MSM_SUBMIT_CMD_CTX_RESTORE_BUF: + if (priv->lastctx == ctx) + break; + case MSM_SUBMIT_CMD_BUF: + OUT_PKT7(ring, CP_INDIRECT_BUFFER_PFE, 3); + OUT_RING(ring, lower_32_bits(submit->cmd[i].iova)); + OUT_RING(ring, upper_32_bits(submit->cmd[i].iova)); + OUT_RING(ring, submit->cmd[i].size); + ibs++; + break; + } + } + + OUT_PKT4(ring, REG_A5XX_CP_SCRATCH_REG(2), 1); + OUT_RING(ring, submit->fence); + + OUT_PKT7(ring, CP_EVENT_WRITE, 4); + OUT_RING(ring, CACHE_FLUSH_TS | (1 << 31)); + OUT_RING(ring, lower_32_bits(rbmemptr(adreno_gpu, fence))); + OUT_RING(ring, upper_32_bits(rbmemptr(adreno_gpu, fence))); + OUT_RING(ring, submit->fence); + + gpu->funcs->flush(gpu); + + return 0; +} + +struct a5xx_hwcg { + u32 offset; + u32 value; +}; + +static const struct a5xx_hwcg a530_hwcg[] = { + {REG_A5XX_RBBM_CLOCK_CNTL_SP0, 0x02222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_SP1, 0x02222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_SP2, 0x02222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_SP3, 0x02222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_SP0, 0x02222220}, + {REG_A5XX_RBBM_CLOCK_CNTL2_SP1, 0x02222220}, + {REG_A5XX_RBBM_CLOCK_CNTL2_SP2, 0x02222220}, + {REG_A5XX_RBBM_CLOCK_CNTL2_SP3, 0x02222220}, + {REG_A5XX_RBBM_CLOCK_HYST_SP0, 0x0000F3CF}, + {REG_A5XX_RBBM_CLOCK_HYST_SP1, 0x0000F3CF}, + {REG_A5XX_RBBM_CLOCK_HYST_SP2, 0x0000F3CF}, + {REG_A5XX_RBBM_CLOCK_HYST_SP3, 0x0000F3CF}, + {REG_A5XX_RBBM_CLOCK_DELAY_SP0, 0x00000080}, + {REG_A5XX_RBBM_CLOCK_DELAY_SP1, 0x00000080}, + {REG_A5XX_RBBM_CLOCK_DELAY_SP2, 0x00000080}, + {REG_A5XX_RBBM_CLOCK_DELAY_SP3, 0x00000080}, + {REG_A5XX_RBBM_CLOCK_CNTL_TP0, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_TP1, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_TP2, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_TP3, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_TP0, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_TP1, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_TP2, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_TP3, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL3_TP0, 0x00002222}, + {REG_A5XX_RBBM_CLOCK_CNTL3_TP1, 0x00002222}, + {REG_A5XX_RBBM_CLOCK_CNTL3_TP2, 0x00002222}, + {REG_A5XX_RBBM_CLOCK_CNTL3_TP3, 0x00002222}, + {REG_A5XX_RBBM_CLOCK_HYST_TP0, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST_TP1, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST_TP2, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST_TP3, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST2_TP0, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST2_TP1, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST2_TP2, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST2_TP3, 0x77777777}, + {REG_A5XX_RBBM_CLOCK_HYST3_TP0, 0x00007777}, + {REG_A5XX_RBBM_CLOCK_HYST3_TP1, 0x00007777}, + {REG_A5XX_RBBM_CLOCK_HYST3_TP2, 0x00007777}, + {REG_A5XX_RBBM_CLOCK_HYST3_TP3, 0x00007777}, + {REG_A5XX_RBBM_CLOCK_DELAY_TP0, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY_TP1, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY_TP2, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY_TP3, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY2_TP0, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY2_TP1, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY2_TP2, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY2_TP3, 0x11111111}, + {REG_A5XX_RBBM_CLOCK_DELAY3_TP0, 0x00001111}, + {REG_A5XX_RBBM_CLOCK_DELAY3_TP1, 0x00001111}, + {REG_A5XX_RBBM_CLOCK_DELAY3_TP2, 0x00001111}, + {REG_A5XX_RBBM_CLOCK_DELAY3_TP3, 0x00001111}, + {REG_A5XX_RBBM_CLOCK_CNTL_UCHE, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_UCHE, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL3_UCHE, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL4_UCHE, 0x00222222}, + {REG_A5XX_RBBM_CLOCK_HYST_UCHE, 0x00444444}, + {REG_A5XX_RBBM_CLOCK_DELAY_UCHE, 0x00000002}, + {REG_A5XX_RBBM_CLOCK_CNTL_RB0, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_RB1, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_RB2, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_RB3, 0x22222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_RB0, 0x00222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_RB1, 0x00222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_RB2, 0x00222222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_RB3, 0x00222222}, + {REG_A5XX_RBBM_CLOCK_CNTL_CCU0, 0x00022220}, + {REG_A5XX_RBBM_CLOCK_CNTL_CCU1, 0x00022220}, + {REG_A5XX_RBBM_CLOCK_CNTL_CCU2, 0x00022220}, + {REG_A5XX_RBBM_CLOCK_CNTL_CCU3, 0x00022220}, + {REG_A5XX_RBBM_CLOCK_CNTL_RAC, 0x05522222}, + {REG_A5XX_RBBM_CLOCK_CNTL2_RAC, 0x00505555}, + {REG_A5XX_RBBM_CLOCK_HYST_RB_CCU0, 0x04040404}, + {REG_A5XX_RBBM_CLOCK_HYST_RB_CCU1, 0x04040404}, + {REG_A5XX_RBBM_CLOCK_HYST_RB_CCU2, 0x04040404}, + {REG_A5XX_RBBM_CLOCK_HYST_RB_CCU3, 0x04040404}, + {REG_A5XX_RBBM_CLOCK_HYST_RAC, 0x07444044}, + {REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_0, 0x00000002}, + {REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_1, 0x00000002}, + {REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_2, 0x00000002}, + {REG_A5XX_RBBM_CLOCK_DELAY_RB_CCU_L1_3, 0x00000002}, + {REG_A5XX_RBBM_CLOCK_DELAY_RAC, 0x00010011}, + {REG_A5XX_RBBM_CLOCK_CNTL_TSE_RAS_RBBM, 0x04222222}, + {REG_A5XX_RBBM_CLOCK_MODE_GPC, 0x02222222}, + {REG_A5XX_RBBM_CLOCK_MODE_VFD, 0x00002222}, + {REG_A5XX_RBBM_CLOCK_HYST_TSE_RAS_RBBM, 0x00000000}, + {REG_A5XX_RBBM_CLOCK_HYST_GPC, 0x04104004}, + {REG_A5XX_RBBM_CLOCK_HYST_VFD, 0x00000000}, + {REG_A5XX_RBBM_CLOCK_DELAY_HLSQ, 0x00000000}, + {REG_A5XX_RBBM_CLOCK_DELAY_TSE_RAS_RBBM, 0x00004000}, + {REG_A5XX_RBBM_CLOCK_DELAY_GPC, 0x00000200}, + {REG_A5XX_RBBM_CLOCK_DELAY_VFD, 0x00002222} +}; + +static const struct { + int (*test)(struct adreno_gpu *gpu); + const struct a5xx_hwcg *regs; + unsigned int count; +} a5xx_hwcg_regs[] = { + { adreno_is_a530, a530_hwcg, ARRAY_SIZE(a530_hwcg), }, +}; + +static void _a5xx_enable_hwcg(struct msm_gpu *gpu, + const struct a5xx_hwcg *regs, unsigned int count) +{ + unsigned int i; + + for (i = 0; i < count; i++) + gpu_write(gpu, regs[i].offset, regs[i].value); + + gpu_write(gpu, REG_A5XX_RBBM_CLOCK_CNTL, 0xAAA8AA00); + gpu_write(gpu, REG_A5XX_RBBM_ISDB_CNT, 0x182); +} + +static void a5xx_enable_hwcg(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(a5xx_hwcg_regs); i++) { + if (a5xx_hwcg_regs[i].test(adreno_gpu)) { + _a5xx_enable_hwcg(gpu, a5xx_hwcg_regs[i].regs, + a5xx_hwcg_regs[i].count); + return; + } + } +} + +static int a5xx_me_init(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct msm_ringbuffer *ring = gpu->rb; + + OUT_PKT7(ring, CP_ME_INIT, 8); + + OUT_RING(ring, 0x0000002F); + + /* Enable multiple hardware contexts */ + OUT_RING(ring, 0x00000003); + + /* Enable error detection */ + OUT_RING(ring, 0x20000000); + + /* Don't enable header dump */ + OUT_RING(ring, 0x00000000); + OUT_RING(ring, 0x00000000); + + /* Specify workarounds for various microcode issues */ + if (adreno_is_a530(adreno_gpu)) { + /* Workaround for token end syncs + * Force a WFI after every direct-render 3D mode draw and every + * 2D mode 3 draw + */ + OUT_RING(ring, 0x0000000B); + } else { + /* No workarounds enabled */ + OUT_RING(ring, 0x00000000); + } + + OUT_RING(ring, 0x00000000); + OUT_RING(ring, 0x00000000); + + gpu->funcs->flush(gpu); + + return gpu->funcs->idle(gpu) ? 0 : -EINVAL; +} + +static struct drm_gem_object *a5xx_ucode_load_bo(struct msm_gpu *gpu, + const struct firmware *fw, u64 *iova) +{ + struct drm_device *drm = gpu->dev; + struct drm_gem_object *bo; + void *ptr; + + mutex_lock(&drm->struct_mutex); + bo = msm_gem_new(drm, fw->size - 4, MSM_BO_UNCACHED); + mutex_unlock(&drm->struct_mutex); + + if (IS_ERR(bo)) + return bo; + + ptr = msm_gem_vaddr(bo); + if (!ptr) { + drm_gem_object_unreference_unlocked(bo); + return ERR_PTR(-ENOMEM); + } + + if (iova) { + int ret = msm_gem_get_iova(bo, gpu->aspace, iova); + + if (ret) { + drm_gem_object_unreference_unlocked(bo); + return ERR_PTR(ret); + } + } + + memcpy(ptr, &fw->data[4], fw->size - 4); + return bo; +} + +static int a5xx_ucode_init(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + int ret; + + if (!a5xx_gpu->pm4_bo) { + a5xx_gpu->pm4_bo = a5xx_ucode_load_bo(gpu, adreno_gpu->pm4, + &a5xx_gpu->pm4_iova); + + if (IS_ERR(a5xx_gpu->pm4_bo)) { + ret = PTR_ERR(a5xx_gpu->pm4_bo); + a5xx_gpu->pm4_bo = NULL; + dev_err(gpu->dev->dev, "could not allocate PM4: %d\n", + ret); + return ret; + } + } + + if (!a5xx_gpu->pfp_bo) { + a5xx_gpu->pfp_bo = a5xx_ucode_load_bo(gpu, adreno_gpu->pfp, + &a5xx_gpu->pfp_iova); + + if (IS_ERR(a5xx_gpu->pfp_bo)) { + ret = PTR_ERR(a5xx_gpu->pfp_bo); + a5xx_gpu->pfp_bo = NULL; + dev_err(gpu->dev->dev, "could not allocate PFP: %d\n", + ret); + return ret; + } + } + + gpu_write64(gpu, REG_A5XX_CP_ME_INSTR_BASE_LO, + REG_A5XX_CP_ME_INSTR_BASE_HI, a5xx_gpu->pm4_iova); + + gpu_write64(gpu, REG_A5XX_CP_PFP_INSTR_BASE_LO, + REG_A5XX_CP_PFP_INSTR_BASE_HI, a5xx_gpu->pfp_iova); + + return 0; +} + +#define A5XX_INT_MASK (A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR | \ + A5XX_RBBM_INT_0_MASK_RBBM_TRANSFER_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_ME_MS_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_PFP_MS_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_ETS_MS_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNC_OVERFLOW | \ + A5XX_RBBM_INT_0_MASK_CP_HW_ERROR | \ + A5XX_RBBM_INT_0_MASK_MISC_HANG_DETECT | \ + A5XX_RBBM_INT_0_MASK_CP_CACHE_FLUSH_TS | \ + A5XX_RBBM_INT_0_MASK_UCHE_OOB_ACCESS | \ + A5XX_RBBM_INT_0_MASK_GPMU_VOLTAGE_DROOP) + +static int a5xx_hw_init(struct msm_gpu *gpu) +{ + struct msm_drm_private *priv = gpu->dev->dev_private; + struct platform_device *pdev = priv->gpu_pdev; + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + int ret, bit = 0; + + pm_qos_update_request(&gpu->pm_qos_req_dma, 101); + + gpu_write(gpu, REG_A5XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x00000003); + + /* Make all blocks contribute to the GPU BUSY perf counter */ + gpu_write(gpu, REG_A5XX_RBBM_PERFCTR_GPU_BUSY_MASKED, 0xFFFFFFFF); + + /* Enable RBBM error reporting bits */ + gpu_write(gpu, REG_A5XX_RBBM_AHB_CNTL0, 0x00000001); + + if (adreno_gpu->quirks & ADRENO_QUIRK_FAULT_DETECT_MASK) { + /* + * Mask out the activity signals from RB1-3 to avoid false + * positives + */ + + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL11, + 0xF0000000); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL12, + 0xFFFFFFFF); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL13, + 0xFFFFFFFF); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL14, + 0xFFFFFFFF); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL15, + 0xFFFFFFFF); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL16, + 0xFFFFFFFF); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL17, + 0xFFFFFFFF); + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_MASK_CNTL18, + 0xFFFFFFFF); + } + + /* Enable fault detection */ + gpu_write(gpu, REG_A5XX_RBBM_INTERFACE_HANG_INT_CNTL, + (1 << 30) | 0xFFFF); + + /* Turn on performance counters */ + gpu_write(gpu, REG_A5XX_RBBM_PERFCTR_CNTL, 0x01); + + /* Increase VFD cache access so LRZ and other data gets evicted less */ + gpu_write(gpu, REG_A5XX_UCHE_CACHE_WAYS, 0x02); + + /* Disable L2 bypass in the UCHE */ + gpu_write(gpu, REG_A5XX_UCHE_TRAP_BASE_LO, 0xFFFF0000); + gpu_write(gpu, REG_A5XX_UCHE_TRAP_BASE_HI, 0x0001FFFF); + gpu_write(gpu, REG_A5XX_UCHE_WRITE_THRU_BASE_LO, 0xFFFF0000); + gpu_write(gpu, REG_A5XX_UCHE_WRITE_THRU_BASE_HI, 0x0001FFFF); + + /* Set the GMEM VA range [0x100000:0x100000 + gpu->gmem - 1] */ + gpu_write64(gpu, REG_A5XX_UCHE_GMEM_RANGE_MIN_LO, + REG_A5XX_UCHE_GMEM_RANGE_MIN_LO, 0x00100000); + + gpu_write64(gpu, REG_A5XX_UCHE_GMEM_RANGE_MAX_LO, + REG_A5XX_UCHE_GMEM_RANGE_MAX_HI, + 0x00100000 + adreno_gpu->gmem - 1); + + gpu_write(gpu, REG_A5XX_CP_MEQ_THRESHOLDS, 0x40); + gpu_write(gpu, REG_A5XX_CP_MERCIU_SIZE, 0x40); + gpu_write(gpu, REG_A5XX_CP_ROQ_THRESHOLDS_2, 0x80000060); + gpu_write(gpu, REG_A5XX_CP_ROQ_THRESHOLDS_1, 0x40201B16); + + gpu_write(gpu, REG_A5XX_PC_DBG_ECO_CNTL, (0x400 << 11 | 0x300 << 22)); + + if (adreno_gpu->quirks & ADRENO_QUIRK_TWO_PASS_USE_WFI) + gpu_rmw(gpu, REG_A5XX_PC_DBG_ECO_CNTL, 0, (1 << 8)); + + gpu_write(gpu, REG_A5XX_PC_DBG_ECO_CNTL, 0xc0200100); + + /* Enable USE_RETENTION_FLOPS */ + gpu_write(gpu, REG_A5XX_CP_CHICKEN_DBG, 0x02000000); + + /* Enable ME/PFP split notification */ + gpu_write(gpu, REG_A5XX_RBBM_AHB_CNTL1, 0xA6FFFFFF); + + /* Enable HWCG */ + a5xx_enable_hwcg(gpu); + + gpu_write(gpu, REG_A5XX_RBBM_AHB_CNTL2, 0x0000003F); + + /* Set the highest bank bit if specified in the device tree */ + if (!of_property_read_u32(pdev->dev.of_node, "qcom,highest-bank-bit", + &bit)) { + if (bit >= 13 && bit <= 16) { + bit -= 13; + + gpu_write(gpu, REG_A5XX_TPL1_MODE_CNTL, bit << 7); + gpu_write(gpu, REG_A5XX_RB_MODE_CNTL, bit << 1); + } + } + + /* Protect registers from the CP */ + gpu_write(gpu, REG_A5XX_CP_PROTECT_CNTL, 0x00000007); + + /* RBBM */ + gpu_write(gpu, REG_A5XX_CP_PROTECT(0), ADRENO_PROTECT_RW(0x04, 4)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(1), ADRENO_PROTECT_RW(0x08, 8)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(2), ADRENO_PROTECT_RW(0x10, 16)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(3), ADRENO_PROTECT_RW(0x20, 32)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(4), ADRENO_PROTECT_RW(0x40, 64)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(5), ADRENO_PROTECT_RW(0x80, 64)); + + /* Content protect */ + gpu_write(gpu, REG_A5XX_CP_PROTECT(6), + ADRENO_PROTECT_RW(REG_A5XX_RBBM_SECVID_TSB_TRUSTED_BASE_LO, + 16)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(7), + ADRENO_PROTECT_RW(REG_A5XX_RBBM_SECVID_TRUST_CNTL, 2)); + + /* CP */ + gpu_write(gpu, REG_A5XX_CP_PROTECT(8), ADRENO_PROTECT_RW(0x800, 64)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(9), ADRENO_PROTECT_RW(0x840, 8)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(10), ADRENO_PROTECT_RW(0x880, 32)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(11), ADRENO_PROTECT_RW(0xAA0, 1)); + + /* RB */ + gpu_write(gpu, REG_A5XX_CP_PROTECT(12), ADRENO_PROTECT_RW(0xCC0, 1)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(13), ADRENO_PROTECT_RW(0xCF0, 2)); + + /* VPC */ + gpu_write(gpu, REG_A5XX_CP_PROTECT(14), ADRENO_PROTECT_RW(0xE68, 8)); + gpu_write(gpu, REG_A5XX_CP_PROTECT(15), ADRENO_PROTECT_RW(0xE70, 4)); + + /* UCHE */ + gpu_write(gpu, REG_A5XX_CP_PROTECT(16), ADRENO_PROTECT_RW(0xE80, 16)); + + if (adreno_is_a530(adreno_gpu)) + gpu_write(gpu, REG_A5XX_CP_PROTECT(17), + ADRENO_PROTECT_RW(0x10000, 0x8000)); + + gpu_write(gpu, REG_A5XX_RBBM_SECVID_TSB_CNTL, 0); + /* + * Disable the trusted memory range - we don't actually supported secure + * memory rendering at this point in time and we don't want to block off + * part of the virtual memory space. + */ + gpu_write64(gpu, REG_A5XX_RBBM_SECVID_TSB_TRUSTED_BASE_LO, + REG_A5XX_RBBM_SECVID_TSB_TRUSTED_BASE_HI, 0x00000000); + gpu_write(gpu, REG_A5XX_RBBM_SECVID_TSB_TRUSTED_SIZE, 0x00000000); + + ret = adreno_hw_init(gpu); + if (ret) + return ret; + + ret = a5xx_ucode_init(gpu); + if (ret) + return ret; + + /* Disable the interrupts through the initial bringup stage */ + gpu_write(gpu, REG_A5XX_RBBM_INT_0_MASK, A5XX_INT_MASK); + + /* Clear ME_HALT to start the micro engine */ + gpu_write(gpu, REG_A5XX_CP_PFP_ME_CNTL, 0); + ret = a5xx_me_init(gpu); + if (ret) + return ret; + + /* + * Send a pipeline event stat to get misbehaving counters to start + * ticking correctly + */ + if (adreno_is_a530(adreno_gpu)) { + OUT_PKT7(gpu->rb, CP_EVENT_WRITE, 1); + OUT_RING(gpu->rb, 0x0F); + + gpu->funcs->flush(gpu); + if (!gpu->funcs->idle(gpu)) + return -EINVAL; + } + + /* Put the GPU into insecure mode */ + gpu_write(gpu, REG_A5XX_RBBM_SECVID_TRUST_CNTL, 0x0); + + pm_qos_update_request(&gpu->pm_qos_req_dma, 501); + + return 0; +} + +static void a5xx_recover(struct msm_gpu *gpu) +{ + adreno_dump_info(gpu); + + if (hang_debug) + a5xx_dump(gpu); + + /* Reset the GPU so it can work again */ + gpu_write(gpu, REG_A5XX_RBBM_SW_RESET_CMD, 1); + gpu_read(gpu, REG_A5XX_RBBM_SW_RESET_CMD); + gpu_write(gpu, REG_A5XX_RBBM_SW_RESET_CMD, 0); + + adreno_recover(gpu); +} + +static void a5xx_destroy(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + + DBG("%s", gpu->name); + + if (a5xx_gpu->pm4_bo) { + if (a5xx_gpu->pm4_iova) + msm_gem_put_iova(a5xx_gpu->pm4_bo, gpu->aspace); + drm_gem_object_unreference_unlocked(a5xx_gpu->pm4_bo); + } + + if (a5xx_gpu->pfp_bo) { + if (a5xx_gpu->pfp_iova) + msm_gem_put_iova(a5xx_gpu->pfp_bo, gpu->aspace); + drm_gem_object_unreference_unlocked(a5xx_gpu->pfp_bo); + } + + adreno_gpu_cleanup(adreno_gpu); + kfree(a5xx_gpu); +} + +static inline bool _a5xx_check_idle(struct msm_gpu *gpu) +{ + if (gpu_read(gpu, REG_A5XX_RBBM_STATUS) & ~A5XX_RBBM_STATUS_HI_BUSY) + return false; + + /* + * Nearly every abnormality ends up pausing the GPU and triggering a + * fault so we can safely just watch for this one interrupt to fire + */ + return !(gpu_read(gpu, REG_A5XX_RBBM_INT_0_STATUS) & + A5XX_RBBM_INT_0_MASK_MISC_HANG_DETECT); +} + +static bool a5xx_idle(struct msm_gpu *gpu) +{ + /* wait for CP to drain ringbuffer: */ + if (!adreno_idle(gpu)) + return false; + + if (spin_until(_a5xx_check_idle(gpu))) { + DRM_ERROR("%s: timeout waiting for GPU to idle: status %8.8X irq %8.8X\n", + gpu->name, + gpu_read(gpu, REG_A5XX_RBBM_STATUS), + gpu_read(gpu, REG_A5XX_RBBM_INT_0_STATUS)); + + return false; + } + + return true; +} + +static void a5xx_cp_err_irq(struct msm_gpu *gpu) +{ + u32 status = gpu_read(gpu, REG_A5XX_CP_INTERRUPT_STATUS); + + if (status & A5XX_CP_INT_CP_OPCODE_ERROR) { + u32 val; + + gpu_write(gpu, REG_A5XX_CP_PFP_STAT_ADDR, 0); + + /* + * REG_A5XX_CP_PFP_STAT_DATA is indexed, and we want index 1 so + * read it twice + */ + + gpu_read(gpu, REG_A5XX_CP_PFP_STAT_DATA); + val = gpu_read(gpu, REG_A5XX_CP_PFP_STAT_DATA); + + dev_err_ratelimited(gpu->dev->dev, "CP | opcode error | possible opcode=0x%8.8X\n", + val); + } + + if (status & A5XX_CP_INT_CP_HW_FAULT_ERROR) + dev_err_ratelimited(gpu->dev->dev, "CP | HW fault | status=0x%8.8X\n", + gpu_read(gpu, REG_A5XX_CP_HW_FAULT)); + + if (status & A5XX_CP_INT_CP_DMA_ERROR) + dev_err_ratelimited(gpu->dev->dev, "CP | DMA error\n"); + + if (status & A5XX_CP_INT_CP_REGISTER_PROTECTION_ERROR) { + u32 val = gpu_read(gpu, REG_A5XX_CP_PROTECT_STATUS); + + dev_err_ratelimited(gpu->dev->dev, + "CP | protected mode error | %s | addr=0x%8.8X | status=0x%8.8X\n", + val & (1 << 24) ? "WRITE" : "READ", + (val & 0xFFFFF) >> 2, val); + } + + if (status & A5XX_CP_INT_CP_AHB_ERROR) { + u32 status = gpu_read(gpu, REG_A5XX_CP_AHB_FAULT); + const char *access[16] = { "reserved", "reserved", + "timestamp lo", "timestamp hi", "pfp read", "pfp write", + "", "", "me read", "me write", "", "", "crashdump read", + "crashdump write" }; + + dev_err_ratelimited(gpu->dev->dev, + "CP | AHB error | addr=%X access=%s error=%d | status=0x%8.8X\n", + status & 0xFFFFF, access[(status >> 24) & 0xF], + (status & (1 << 31)), status); + } +} + +static void a5xx_rbbm_err_irq(struct msm_gpu *gpu, u32 status) +{ + if (status & A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR) { + u32 val = gpu_read(gpu, REG_A5XX_RBBM_AHB_ERROR_STATUS); + + dev_err_ratelimited(gpu->dev->dev, + "RBBM | AHB bus error | %s | addr=0x%X | ports=0x%X:0x%X\n", + val & (1 << 28) ? "WRITE" : "READ", + (val & 0xFFFFF) >> 2, (val >> 20) & 0x3, + (val >> 24) & 0xF); + + /* Clear the error */ + gpu_write(gpu, REG_A5XX_RBBM_AHB_CMD, (1 << 4)); + + /* Clear the interrupt */ + gpu_write(gpu, REG_A5XX_RBBM_INT_CLEAR_CMD, + A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR); + } + + if (status & A5XX_RBBM_INT_0_MASK_RBBM_TRANSFER_TIMEOUT) + dev_err_ratelimited(gpu->dev->dev, "RBBM | AHB transfer timeout\n"); + + if (status & A5XX_RBBM_INT_0_MASK_RBBM_ME_MS_TIMEOUT) + dev_err_ratelimited(gpu->dev->dev, "RBBM | ME master split | status=0x%X\n", + gpu_read(gpu, REG_A5XX_RBBM_AHB_ME_SPLIT_STATUS)); + + if (status & A5XX_RBBM_INT_0_MASK_RBBM_PFP_MS_TIMEOUT) + dev_err_ratelimited(gpu->dev->dev, "RBBM | PFP master split | status=0x%X\n", + gpu_read(gpu, REG_A5XX_RBBM_AHB_PFP_SPLIT_STATUS)); + + if (status & A5XX_RBBM_INT_0_MASK_RBBM_ETS_MS_TIMEOUT) + dev_err_ratelimited(gpu->dev->dev, "RBBM | ETS master split | status=0x%X\n", + gpu_read(gpu, REG_A5XX_RBBM_AHB_ETS_SPLIT_STATUS)); + + if (status & A5XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNC_OVERFLOW) + dev_err_ratelimited(gpu->dev->dev, "RBBM | ATB ASYNC overflow\n"); + + if (status & A5XX_RBBM_INT_0_MASK_RBBM_ATB_BUS_OVERFLOW) + dev_err_ratelimited(gpu->dev->dev, "RBBM | ATB bus overflow\n"); +} + +static void a5xx_uche_err_irq(struct msm_gpu *gpu) +{ + uint64_t addr = (uint64_t) gpu_read(gpu, REG_A5XX_UCHE_TRAP_LOG_HI); + + addr |= gpu_read(gpu, REG_A5XX_UCHE_TRAP_LOG_LO); + + dev_err_ratelimited(gpu->dev->dev, "UCHE | Out of bounds access | addr=0x%llX\n", + addr); +} + +static void a5xx_gpmu_err_irq(struct msm_gpu *gpu) +{ + dev_err_ratelimited(gpu->dev->dev, "GPMU | voltage droop\n"); +} + +static void a5xx_fault_detect_irq(struct msm_gpu *gpu) +{ + struct drm_device *dev = gpu->dev; + struct msm_drm_private *priv = dev->dev_private; + + queue_work(priv->wq, &gpu->recover_work); +} + +#define RBBM_ERROR_MASK \ + (A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR | \ + A5XX_RBBM_INT_0_MASK_RBBM_TRANSFER_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_ME_MS_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_PFP_MS_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_ETS_MS_TIMEOUT | \ + A5XX_RBBM_INT_0_MASK_RBBM_ATB_ASYNC_OVERFLOW) + +static irqreturn_t a5xx_irq(struct msm_gpu *gpu) +{ + u32 status = gpu_read(gpu, REG_A5XX_RBBM_INT_0_STATUS); + + /* + * Clear all the interrupts except for RBBM_AHB_ERROR + * which needs to be cleared after the error condition + * is cleared otherwise it will storm + */ + gpu_write(gpu, REG_A5XX_RBBM_INT_CLEAR_CMD, + status & ~A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR); + + if (status & RBBM_ERROR_MASK) + a5xx_rbbm_err_irq(gpu, status); + + if (status & A5XX_RBBM_INT_0_MASK_CP_HW_ERROR) + a5xx_cp_err_irq(gpu); + + if (status & A5XX_RBBM_INT_0_MASK_MISC_HANG_DETECT) + a5xx_fault_detect_irq(gpu); + + if (status & A5XX_RBBM_INT_0_MASK_UCHE_OOB_ACCESS) + a5xx_uche_err_irq(gpu); + + if (status & A5XX_RBBM_INT_0_MASK_GPMU_VOLTAGE_DROOP) + a5xx_gpmu_err_irq(gpu); + + if (status & A5XX_RBBM_INT_0_MASK_CP_CACHE_FLUSH_TS) + msm_gpu_retire(gpu); + + return IRQ_HANDLED; +} + +static const u32 a5xx_register_offsets[REG_ADRENO_REGISTER_MAX] = { + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE, REG_A5XX_CP_RB_BASE), + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE_HI, REG_A5XX_CP_RB_BASE_HI), + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR_ADDR, REG_A5XX_CP_RB_RPTR_ADDR), + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR_ADDR_HI, + REG_A5XX_CP_RB_RPTR_ADDR_HI), + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_RPTR, REG_A5XX_CP_RB_RPTR), + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_WPTR, REG_A5XX_CP_RB_WPTR), + REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_CNTL, REG_A5XX_CP_RB_CNTL), +}; + +static const u32 a5xx_registers[] = { + 0x0000, 0x0002, 0x0004, 0x0020, 0x0022, 0x0026, 0x0029, 0x002B, + 0x002E, 0x0035, 0x0038, 0x0042, 0x0044, 0x0044, 0x0047, 0x0095, + 0x0097, 0x00BB, 0x03A0, 0x0464, 0x0469, 0x046F, 0x04D2, 0x04D3, + 0x04E0, 0x0533, 0x0540, 0x0555, 0xF400, 0xF400, 0xF800, 0xF807, + 0x0800, 0x081A, 0x081F, 0x0841, 0x0860, 0x0860, 0x0880, 0x08A0, + 0x0B00, 0x0B12, 0x0B15, 0x0B28, 0x0B78, 0x0B7F, 0x0BB0, 0x0BBD, + 0x0BC0, 0x0BC6, 0x0BD0, 0x0C53, 0x0C60, 0x0C61, 0x0C80, 0x0C82, + 0x0C84, 0x0C85, 0x0C90, 0x0C98, 0x0CA0, 0x0CA0, 0x0CB0, 0x0CB2, + 0x2180, 0x2185, 0x2580, 0x2585, 0x0CC1, 0x0CC1, 0x0CC4, 0x0CC7, + 0x0CCC, 0x0CCC, 0x0CD0, 0x0CD8, 0x0CE0, 0x0CE5, 0x0CE8, 0x0CE8, + 0x0CEC, 0x0CF1, 0x0CFB, 0x0D0E, 0x2100, 0x211E, 0x2140, 0x2145, + 0x2500, 0x251E, 0x2540, 0x2545, 0x0D10, 0x0D17, 0x0D20, 0x0D23, + 0x0D30, 0x0D30, 0x20C0, 0x20C0, 0x24C0, 0x24C0, 0x0E40, 0x0E43, + 0x0E4A, 0x0E4A, 0x0E50, 0x0E57, 0x0E60, 0x0E7C, 0x0E80, 0x0E8E, + 0x0E90, 0x0E96, 0x0EA0, 0x0EA8, 0x0EB0, 0x0EB2, 0xE140, 0xE147, + 0xE150, 0xE187, 0xE1A0, 0xE1A9, 0xE1B0, 0xE1B6, 0xE1C0, 0xE1C7, + 0xE1D0, 0xE1D1, 0xE200, 0xE201, 0xE210, 0xE21C, 0xE240, 0xE268, + 0xE000, 0xE006, 0xE010, 0xE09A, 0xE0A0, 0xE0A4, 0xE0AA, 0xE0EB, + 0xE100, 0xE105, 0xE380, 0xE38F, 0xE3B0, 0xE3B0, 0xE400, 0xE405, + 0xE408, 0xE4E9, 0xE4F0, 0xE4F0, 0xE280, 0xE280, 0xE282, 0xE2A3, + 0xE2A5, 0xE2C2, 0xE940, 0xE947, 0xE950, 0xE987, 0xE9A0, 0xE9A9, + 0xE9B0, 0xE9B6, 0xE9C0, 0xE9C7, 0xE9D0, 0xE9D1, 0xEA00, 0xEA01, + 0xEA10, 0xEA1C, 0xEA40, 0xEA68, 0xE800, 0xE806, 0xE810, 0xE89A, + 0xE8A0, 0xE8A4, 0xE8AA, 0xE8EB, 0xE900, 0xE905, 0xEB80, 0xEB8F, + 0xEBB0, 0xEBB0, 0xEC00, 0xEC05, 0xEC08, 0xECE9, 0xECF0, 0xECF0, + 0xEA80, 0xEA80, 0xEA82, 0xEAA3, 0xEAA5, 0xEAC2, 0xA800, 0xA8FF, + 0xAC60, 0xAC60, 0xB000, 0xB97F, 0xB9A0, 0xB9BF, + ~0 +}; + +static void a5xx_dump(struct msm_gpu *gpu) +{ + dev_info(gpu->dev->dev, "status: %08x\n", + gpu_read(gpu, REG_A5XX_RBBM_STATUS)); + adreno_dump(gpu); +} + +static int a5xx_pm_resume(struct msm_gpu *gpu) +{ + return msm_gpu_pm_resume(gpu); +} + +static int a5xx_pm_suspend(struct msm_gpu *gpu) +{ + return msm_gpu_pm_suspend(gpu); +} + +static int a5xx_get_timestamp(struct msm_gpu *gpu, uint64_t *value) +{ + *value = gpu_read64(gpu, REG_A5XX_RBBM_PERFCTR_CP_0_LO, + REG_A5XX_RBBM_PERFCTR_CP_0_HI); + + return 0; +} + +#ifdef CONFIG_DEBUG_FS +static void a5xx_show(struct msm_gpu *gpu, struct seq_file *m) +{ + gpu->funcs->pm_resume(gpu); + + seq_printf(m, "status: %08x\n", + gpu_read(gpu, REG_A5XX_RBBM_STATUS)); + gpu->funcs->pm_suspend(gpu); + + adreno_show(gpu, m); +} +#endif + +static const struct adreno_gpu_funcs funcs = { + .base = { + .get_param = adreno_get_param, + .hw_init = a5xx_hw_init, + .pm_suspend = a5xx_pm_suspend, + .pm_resume = a5xx_pm_resume, + .recover = a5xx_recover, + .last_fence = adreno_last_fence, + .submit = a5xx_submit, + .flush = adreno_flush, + .idle = a5xx_idle, + .irq = a5xx_irq, + .destroy = a5xx_destroy, +#ifdef CONFIG_DEBUG_FS + .show = a5xx_show, +#endif + }, + .get_timestamp = a5xx_get_timestamp, +}; + +struct msm_gpu *a5xx_gpu_init(struct drm_device *dev) +{ + struct msm_drm_private *priv = dev->dev_private; + struct platform_device *pdev = priv->gpu_pdev; + struct a5xx_gpu *a5xx_gpu = NULL; + struct adreno_gpu *adreno_gpu; + struct msm_gpu *gpu; + int ret; + + if (!pdev) { + dev_err(dev->dev, "No A5XX device is defined\n"); + return ERR_PTR(-ENXIO); + } + + a5xx_gpu = kzalloc(sizeof(*a5xx_gpu), GFP_KERNEL); + if (!a5xx_gpu) + return ERR_PTR(-ENOMEM); + + adreno_gpu = &a5xx_gpu->base; + gpu = &adreno_gpu->base; + + a5xx_gpu->pdev = pdev; + adreno_gpu->registers = a5xx_registers; + adreno_gpu->reg_offsets = a5xx_register_offsets; + + ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs); + if (ret) { + a5xx_destroy(&(a5xx_gpu->base.base)); + return ERR_PTR(ret); + } + + return gpu; +} diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h new file mode 100644 index 000000000000..39a07f400b35 --- /dev/null +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h @@ -0,0 +1,37 @@ +/* Copyright (c) 2016 The Linux Foundation. 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 version 2 and + * only 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. + * + */ +#ifndef __A5XX_GPU_H__ +#define __A5XX_GPU_H__ + +#include "adreno_gpu.h" + +/* Bringing over the hack from the previous targets */ +#undef ROP_COPY +#undef ROP_XOR + +#include "a5xx.xml.h" + +struct a5xx_gpu { + struct adreno_gpu base; + struct platform_device *pdev; + + struct drm_gem_object *pm4_bo; + uint64_t pm4_iova; + + struct drm_gem_object *pfp_bo; + uint64_t pfp_iova; +}; + +#define to_a5xx_gpu(x) container_of(x, struct a5xx_gpu, base) + +#endif /* __A5XX_GPU_H__ */ diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c index 3321d33ae6a9..7749b60a876c 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_device.c +++ b/drivers/gpu/drm/msm/adreno/adreno_device.c @@ -27,6 +27,7 @@ module_param_named(hang_debug, hang_debug, bool, 0600); struct msm_gpu *a3xx_gpu_init(struct drm_device *dev); struct msm_gpu *a4xx_gpu_init(struct drm_device *dev); +struct msm_gpu *a5xx_gpu_init(struct drm_device *dev); static const struct adreno_info gpulist[] = { { @@ -77,6 +78,14 @@ static const struct adreno_info gpulist[] = { .pfpfw = "a420_pfp.fw", .gmem = (SZ_1M + SZ_512K), .init = a4xx_gpu_init, + }, { + .rev = ADRENO_REV(5, 3, 0, ANY_ID), + .revn = 530, + .name = "A530", + .pm4fw = "a530_pm4.fw", + .pfpfw = "a530_pfp.fw", + .gmem = SZ_1M, + .init = a5xx_gpu_init, }, }; @@ -86,6 +95,8 @@ MODULE_FIRMWARE("a330_pm4.fw"); MODULE_FIRMWARE("a330_pfp.fw"); MODULE_FIRMWARE("a420_pm4.fw"); MODULE_FIRMWARE("a420_pfp.fw"); +MODULE_FIRMWARE("a530_fm4.fw"); +MODULE_FIRMWARE("a530_pfp.fw"); static inline bool _rev_match(uint8_t entry, uint8_t id) { @@ -173,12 +184,20 @@ static void set_gpu_pdev(struct drm_device *dev, priv->gpu_pdev = pdev; } +static const struct { + const char *str; + uint32_t flag; +} quirks[] = { + { "qcom,gpu-quirk-two-pass-use-wfi", ADRENO_QUIRK_TWO_PASS_USE_WFI }, + { "qcom,gpu-quirk-fault-detect-mask", ADRENO_QUIRK_FAULT_DETECT_MASK }, +}; + static int adreno_bind(struct device *dev, struct device *master, void *data) { static struct adreno_platform_config config = {}; struct device_node *child, *node = dev->of_node; u32 val; - int ret; + int ret, i; ret = of_property_read_u32(node, "qcom,chipid", &val); if (ret) { @@ -212,6 +231,10 @@ static int adreno_bind(struct device *dev, struct device *master, void *data) return -ENXIO; } + for (i = 0; i < ARRAY_SIZE(quirks); i++) + if (of_property_read_bool(node, quirks[i].str)) + config.quirks |= quirks[i].flag; + dev->platform_data = &config; set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev)); return 0; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 46c2fc96a4cb..49f01c2e8a8d 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -22,7 +22,7 @@ #include "msm_mmu.h" #define RB_SIZE SZ_32K -#define RB_BLKSIZE 16 +#define RB_BLKSIZE 32 int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value) { @@ -54,9 +54,6 @@ int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value) } } -#define rbmemptr(adreno_gpu, member) \ - ((adreno_gpu)->memptrs_iova + offsetof(struct adreno_rbmemptrs, member)) - int adreno_hw_init(struct msm_gpu *gpu) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); @@ -78,12 +75,12 @@ int adreno_hw_init(struct msm_gpu *gpu) AXXX_CP_RB_CNTL_BLKSZ(ilog2(RB_BLKSIZE / 8)) | (adreno_is_a430(adreno_gpu) ? AXXX_CP_RB_CNTL_NO_UPDATE : 0)); - /* Setup ringbuffer address: */ - adreno_gpu_write(adreno_gpu, REG_ADRENO_CP_RB_BASE, gpu->rb_iova); + /* Setup ringbuffer address */ + adreno_gpu_write64(adreno_gpu, REG_ADRENO_CP_RB_BASE, + REG_ADRENO_CP_RB_BASE_HI, gpu->rb_iova); - if (!adreno_is_a430(adreno_gpu)) - adreno_gpu_write(adreno_gpu, REG_ADRENO_CP_RB_RPTR_ADDR, - rbmemptr(adreno_gpu, rptr)); + adreno_gpu_write64(adreno_gpu, REG_ADRENO_CP_RB_RPTR_ADDR, + REG_ADRENO_CP_RB_RPTR_ADDR_HI, rbmemptr(adreno_gpu, rptr)); return 0; } @@ -367,6 +364,7 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, adreno_gpu->gmem = adreno_gpu->info->gmem; adreno_gpu->revn = adreno_gpu->info->revn; adreno_gpu->rev = config->rev; + adreno_gpu->quirks = config->quirks; gpu->fast_rate = config->fast_rate; gpu->slow_rate = config->slow_rate; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index 7d3553ac5224..190e864a2cf9 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -52,6 +52,11 @@ enum adreno_regs { REG_ADRENO_REGISTER_MAX, }; +enum adreno_quirks { + ADRENO_QUIRK_TWO_PASS_USE_WFI = 1, + ADRENO_QUIRK_FAULT_DETECT_MASK = 2, +}; + struct adreno_rev { uint8_t core; uint8_t major; @@ -78,6 +83,9 @@ struct adreno_info { const struct adreno_info *adreno_info(struct adreno_rev rev); +#define rbmemptr(adreno_gpu, member) \ + ((adreno_gpu)->memptrs_iova + offsetof(struct adreno_rbmemptrs, member)) + struct adreno_rbmemptrs { volatile uint32_t rptr; volatile uint32_t wptr; @@ -111,6 +119,8 @@ struct adreno_gpu { * code (a3xx_gpu.c) and stored in this common location. */ const unsigned int *reg_offsets; + + uint32_t quirks; }; #define to_adreno_gpu(x) container_of(x, struct adreno_gpu, base) @@ -121,6 +131,7 @@ struct adreno_platform_config { #ifdef DOWNSTREAM_CONFIG_MSM_BUS_SCALING struct msm_bus_scale_pdata *bus_scale_table; #endif + uint32_t quirks; }; #define ADRENO_IDLE_TIMEOUT msecs_to_jiffies(1000) @@ -184,6 +195,11 @@ static inline int adreno_is_a430(struct adreno_gpu *gpu) return gpu->revn == 430; } +static inline int adreno_is_a530(struct adreno_gpu *gpu) +{ + return gpu->revn == 530; +} + int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value); int adreno_hw_init(struct msm_gpu *gpu); uint32_t adreno_last_fence(struct msm_gpu *gpu); @@ -301,4 +317,35 @@ static inline void adreno_gpu_write(struct adreno_gpu *gpu, gpu_write(&gpu->base, reg - 1, data); } +static inline void adreno_gpu_write64(struct adreno_gpu *gpu, + enum adreno_regs lo, enum adreno_regs hi, u64 data) +{ + adreno_gpu_write(gpu, lo, lower_32_bits(data)); + adreno_gpu_write(gpu, hi, upper_32_bits(data)); +} + +/* + * Given a register and a count, return a value to program into + * REG_CP_PROTECT_REG(n) - this will block both reads and writes for _len + * registers starting at _reg. + * + * The register base needs to be a multiple of the length. If it is not, the + * hardware will quietly mask off the bits for you and shift the size. For + * example, if you intend the protection to start at 0x07 for a length of 4 + * (0x07-0x0A) the hardware will actually protect (0x04-0x07) which might + * expose registers you intended to protect! + */ +#define ADRENO_PROTECT_RW(_reg, _len) \ + ((1 << 30) | (1 << 29) | \ + ((ilog2((_len)) & 0x1F) << 24) | (((_reg) << 2) & 0xFFFFF)) + +/* + * Same as above, but allow reads over the range. For areas of mixed use (such + * as performance counters) this allows us to protect a much larger range with a + * single register + */ +#define ADRENO_PROTECT_RDONLY(_reg, _len) \ + ((1 << 29) \ + ((ilog2((_len)) & 0x1F) << 24) | (((_reg) << 2) & 0xFFFFF)) + #endif /* __ADRENO_GPU_H__ */ diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 19c08ac07632..30d2fc91c84d 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -562,8 +562,8 @@ static irqreturn_t irq_handler(int irq, void *data) } static const char *clk_names[] = { - "src_clk", "core_clk", "iface_clk", "mem_clk", "mem_iface_clk", - "alt_mem_iface_clk", + "src_clk", "core_clk", "iface_clk", "rbbmtimer_clk", + "mem_clk", "mem_iface_clk", "alt_mem_iface_clk", }; int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev, diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index 5e6d137140d0..a45faf2d06e5 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -100,7 +100,7 @@ struct msm_gpu { /* Power Control: */ struct regulator *gpu_reg, *gpu_cx; - struct clk *ebi1_clk, *grp_clks[6]; + struct clk *ebi1_clk, *grp_clks[7]; uint32_t fast_rate, slow_rate, bus_freq; struct pm_qos_request pm_qos_req_dma; From 0a7274232bb86ef80e4402e5eaffd59edbcda85d Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:20 -0700 Subject: [PATCH 14/19] drm/msm: gpu: Add support for the GPMU Most 5XX targets have GPMU (Graphics Power Management Unit) that handles a lot of the heavy lifting for power management including thermal and limits management and dynamic power collapse. While the GPMU itself is optional, it is usually nessesary to hit aggressive power targets. If the GPMU is to be used a filename and minimum version are defined in the device tree. The GPMU firmware needs to be loaded into the GPMU at init time via a shared hardware block of registers. Using the GPU to write the microcode is more efficient than using the CPU so at first load create an indirect buffer that can be executed during subsequent initalization sequences. After loading the GPMU gets initalized through a shared register interface and then we mostly get out of its way and let it do its thing. Change-Id: Ic0dedbad8d899177919b71500f2e944b187e87c0 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/Makefile | 3 +- drivers/gpu/drm/msm/adreno/a5xx_gpu.c | 123 ++++++- drivers/gpu/drm/msm/adreno/a5xx_gpu.h | 23 ++ drivers/gpu/drm/msm/adreno/a5xx_power.c | 426 ++++++++++++++++++++++++ drivers/gpu/drm/msm/adreno/adreno_gpu.h | 3 + 5 files changed, 576 insertions(+), 2 deletions(-) create mode 100644 drivers/gpu/drm/msm/adreno/a5xx_power.c diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile index 5c587d291198..9f035b10875b 100644 --- a/drivers/gpu/drm/msm/Makefile +++ b/drivers/gpu/drm/msm/Makefile @@ -53,7 +53,8 @@ msm_drm-y += adreno/adreno_device.o \ adreno/adreno_gpu.o \ adreno/a3xx_gpu.o \ adreno/a4xx_gpu.o \ - adreno/a5xx_gpu.o + adreno/a5xx_gpu.o \ + adreno/a5xx_power.o endif msm_drm-$(CONFIG_DRM_MSM_MDP4) += mdp/mdp4/mdp4_crtc.o \ diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c index 9c2a818620b6..fd92312529a6 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -463,6 +463,9 @@ static int a5xx_hw_init(struct msm_gpu *gpu) REG_A5XX_RBBM_SECVID_TSB_TRUSTED_BASE_HI, 0x00000000); gpu_write(gpu, REG_A5XX_RBBM_SECVID_TSB_TRUSTED_SIZE, 0x00000000); + /* Load the GPMU firmware before starting the HW init */ + a5xx_gpmu_ucode_init(gpu); + ret = adreno_hw_init(gpu); if (ret) return ret; @@ -480,6 +483,10 @@ static int a5xx_hw_init(struct msm_gpu *gpu) if (ret) return ret; + ret = a5xx_power_init(gpu); + if (ret) + return ret; + /* * Send a pipeline event stat to get misbehaving counters to start * ticking correctly @@ -535,6 +542,12 @@ static void a5xx_destroy(struct msm_gpu *gpu) drm_gem_object_unreference_unlocked(a5xx_gpu->pfp_bo); } + if (a5xx_gpu->gpmu_bo) { + if (a5xx_gpu->gpmu_iova) + msm_gem_put_iova(a5xx_gpu->gpmu_bo, gpu->aspace); + drm_gem_object_unreference_unlocked(a5xx_gpu->gpmu_bo); + } + adreno_gpu_cleanup(adreno_gpu); kfree(a5xx_gpu); } @@ -777,11 +790,55 @@ static void a5xx_dump(struct msm_gpu *gpu) static int a5xx_pm_resume(struct msm_gpu *gpu) { - return msm_gpu_pm_resume(gpu); + int ret; + + /* Turn on the core power */ + ret = msm_gpu_pm_resume(gpu); + if (ret) + return ret; + + /* Turn the RBCCU domain first to limit the chances of voltage droop */ + gpu_write(gpu, REG_A5XX_GPMU_RBCCU_POWER_CNTL, 0x778000); + + /* Wait 3 usecs before polling */ + udelay(3); + + ret = spin_usecs(gpu, 20, REG_A5XX_GPMU_RBCCU_PWR_CLK_STATUS, + (1 << 20), (1 << 20)); + if (ret) { + DRM_ERROR("%s: timeout waiting for RBCCU GDSC enable: %X\n", + gpu->name, + gpu_read(gpu, REG_A5XX_GPMU_RBCCU_PWR_CLK_STATUS)); + return ret; + } + + /* Turn on the SP domain */ + gpu_write(gpu, REG_A5XX_GPMU_SP_POWER_CNTL, 0x778000); + ret = spin_usecs(gpu, 20, REG_A5XX_GPMU_SP_PWR_CLK_STATUS, + (1 << 20), (1 << 20)); + if (ret) + DRM_ERROR("%s: timeout waiting for SP GDSC enable\n", + gpu->name); + + return ret; } static int a5xx_pm_suspend(struct msm_gpu *gpu) { + /* Clear the VBIF pipe before shutting down */ + + gpu_write(gpu, REG_A5XX_VBIF_XIN_HALT_CTRL0, 0xF); + spin_until((gpu_read(gpu, REG_A5XX_VBIF_XIN_HALT_CTRL1) & 0xF) == 0xF); + + gpu_write(gpu, REG_A5XX_VBIF_XIN_HALT_CTRL0, 0); + + /* + * Reset the VBIF before power collapse to avoid issue with FIFO + * entries + */ + gpu_write(gpu, REG_A5XX_RBBM_BLOCK_SW_RESET_CMD, 0x003C0000); + gpu_write(gpu, REG_A5XX_RBBM_BLOCK_SW_RESET_CMD, 0x00000000); + return msm_gpu_pm_suspend(gpu); } @@ -826,6 +883,65 @@ static const struct adreno_gpu_funcs funcs = { .get_timestamp = a5xx_get_timestamp, }; +/* Read the limits management leakage from the efuses */ +static void a530_efuse_leakage(struct platform_device *pdev, + struct adreno_gpu *adreno_gpu, void *base, + size_t size) +{ + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + unsigned int row0, row2; + unsigned int leakage_pwr_on, coeff; + + if (size < 0x148) + return; + + /* Leakage */ + row0 = readl_relaxed(base + 0x134); + row2 = readl_relaxed(base + 0x144); + + /* Read barrier to get the previous two reads */ + rmb(); + + /* Get the leakage coefficient from device tree */ + if (of_property_read_u32(pdev->dev.of_node, + "qcom,base-leakage-coefficent", &coeff)) + return; + + leakage_pwr_on = ((row2 >> 2) & 0xFF) * (1 << (row0 >> 1) & 0x03); + a5xx_gpu->lm_leakage = (leakage_pwr_on << 16) | + ((leakage_pwr_on * coeff) / 100); +} + +/* Read target specific configuration from the efuses */ +static void a5xx_efuses_read(struct platform_device *pdev, + struct adreno_gpu *adreno_gpu) +{ + struct adreno_platform_config *config = pdev->dev.platform_data; + const struct adreno_info *info = adreno_info(config->rev); + struct resource *res; + void *base; + + /* + * The adreno_gpu->revn mechanism isn't set up yet so we need to check + * it directly here + */ + if (info->revn != 530) + return; + + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, + "qfprom_memory"); + if (!res) + return; + + base = ioremap(res->start, resource_size(res)); + if (!base) + return; + + a530_efuse_leakage(pdev, adreno_gpu, base, resource_size(res)); + + iounmap(base); +} + struct msm_gpu *a5xx_gpu_init(struct drm_device *dev) { struct msm_drm_private *priv = dev->dev_private; @@ -851,6 +967,11 @@ struct msm_gpu *a5xx_gpu_init(struct drm_device *dev) adreno_gpu->registers = a5xx_registers; adreno_gpu->reg_offsets = a5xx_register_offsets; + a5xx_gpu->lm_leakage = 0x4E001A; + + /* Check the efuses for some configuration */ + a5xx_efuses_read(pdev, adreno_gpu); + ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs); if (ret) { a5xx_destroy(&(a5xx_gpu->base.base)); diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h index 39a07f400b35..1590f845d554 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h @@ -30,8 +30,31 @@ struct a5xx_gpu { struct drm_gem_object *pfp_bo; uint64_t pfp_iova; + + struct drm_gem_object *gpmu_bo; + uint64_t gpmu_iova; + uint32_t gpmu_dwords; + + uint32_t lm_leakage; }; #define to_a5xx_gpu(x) container_of(x, struct a5xx_gpu, base) +int a5xx_power_init(struct msm_gpu *gpu); +void a5xx_gpmu_ucode_init(struct msm_gpu *gpu); + +static inline int spin_usecs(struct msm_gpu *gpu, uint32_t usecs, + uint32_t reg, uint32_t mask, uint32_t value) +{ + while (usecs--) { + udelay(1); + if ((gpu_read(gpu, reg) & mask) == value) + return 0; + cpu_relax(); + } + + return -ETIMEDOUT; +} + + #endif /* __A5XX_GPU_H__ */ diff --git a/drivers/gpu/drm/msm/adreno/a5xx_power.c b/drivers/gpu/drm/msm/adreno/a5xx_power.c new file mode 100644 index 000000000000..6bc1d5cbb4fb --- /dev/null +++ b/drivers/gpu/drm/msm/adreno/a5xx_power.c @@ -0,0 +1,426 @@ +/* Copyright (c) 2016 The Linux Foundation. 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 version 2 and + * only 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. + * + */ + +#include +#include "a5xx_gpu.h" + +/* + * The GPMU data block is a block of shared registers that can be used to + * communicate back and forth. These "registers" are by convention with the GPMU + * firwmare and not bound to any specific hardware design + */ + +#define AGC_INIT_BASE REG_A5XX_GPMU_DATA_RAM_BASE +#define AGC_INIT_MSG_MAGIC (AGC_INIT_BASE + 5) +#define AGC_MSG_BASE (AGC_INIT_BASE + 7) + +#define AGC_MSG_STATE (AGC_MSG_BASE + 0) +#define AGC_MSG_COMMAND (AGC_MSG_BASE + 1) +#define AGC_MSG_PAYLOAD_SIZE (AGC_MSG_BASE + 3) +#define AGC_MSG_PAYLOAD(_o) ((AGC_MSG_BASE + 5) + (_o)) + +#define AGC_POWER_CONFIG_PRODUCTION_ID 1 +#define AGC_INIT_MSG_VALUE 0xBABEFACE + +static struct { + uint32_t reg; + uint32_t value; +} a5xx_sequence_regs[] = { + { 0xB9A1, 0x00010303 }, + { 0xB9A2, 0x13000000 }, + { 0xB9A3, 0x00460020 }, + { 0xB9A4, 0x10000000 }, + { 0xB9A5, 0x040A1707 }, + { 0xB9A6, 0x00010000 }, + { 0xB9A7, 0x0E000904 }, + { 0xB9A8, 0x10000000 }, + { 0xB9A9, 0x01165000 }, + { 0xB9AA, 0x000E0002 }, + { 0xB9AB, 0x03884141 }, + { 0xB9AC, 0x10000840 }, + { 0xB9AD, 0x572A5000 }, + { 0xB9AE, 0x00000003 }, + { 0xB9AF, 0x00000000 }, + { 0xB9B0, 0x10000000 }, + { 0xB828, 0x6C204010 }, + { 0xB829, 0x6C204011 }, + { 0xB82A, 0x6C204012 }, + { 0xB82B, 0x6C204013 }, + { 0xB82C, 0x6C204014 }, + { 0xB90F, 0x00000004 }, + { 0xB910, 0x00000002 }, + { 0xB911, 0x00000002 }, + { 0xB912, 0x00000002 }, + { 0xB913, 0x00000002 }, + { 0xB92F, 0x00000004 }, + { 0xB930, 0x00000005 }, + { 0xB931, 0x00000005 }, + { 0xB932, 0x00000005 }, + { 0xB933, 0x00000005 }, + { 0xB96F, 0x00000001 }, + { 0xB970, 0x00000003 }, + { 0xB94F, 0x00000004 }, + { 0xB950, 0x0000000B }, + { 0xB951, 0x0000000B }, + { 0xB952, 0x0000000B }, + { 0xB953, 0x0000000B }, + { 0xB907, 0x00000019 }, + { 0xB927, 0x00000019 }, + { 0xB947, 0x00000019 }, + { 0xB967, 0x00000019 }, + { 0xB987, 0x00000019 }, + { 0xB906, 0x00220001 }, + { 0xB926, 0x00220001 }, + { 0xB946, 0x00220001 }, + { 0xB966, 0x00220001 }, + { 0xB986, 0x00300000 }, + { 0xAC40, 0x0340FF41 }, + { 0xAC41, 0x03BEFED0 }, + { 0xAC42, 0x00331FED }, + { 0xAC43, 0x021FFDD3 }, + { 0xAC44, 0x5555AAAA }, + { 0xAC45, 0x5555AAAA }, + { 0xB9BA, 0x00000008 }, +}; + +/* + * Get the actual voltage value for the operating point at the specified + * frequency + */ +static inline uint32_t _get_mvolts(struct msm_gpu *gpu, uint32_t freq) +{ + struct drm_device *dev = gpu->dev; + struct msm_drm_private *priv = dev->dev_private; + struct platform_device *pdev = priv->gpu_pdev; + struct dev_pm_opp *opp; + + opp = dev_pm_opp_find_freq_exact(&pdev->dev, freq, true); + + return (!IS_ERR(opp)) ? dev_pm_opp_get_voltage(opp) / 1000 : 0; +} + +/* Setup thermal limit management */ +static void a5xx_lm_setup(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + uint32_t tsens = 0; + uint32_t lm_limit = 6000; + uint32_t max_power = 0; + unsigned int i; + + /* Write the block of sequence registers */ + for (i = 0; i < ARRAY_SIZE(a5xx_sequence_regs); i++) + gpu_write(gpu, a5xx_sequence_regs[i].reg, + a5xx_sequence_regs[i].value); + + of_property_read_u32(GPU_OF_NODE(gpu), "qcom,gpmu-tsens", &tsens); + + gpu_write(gpu, REG_A5XX_GPMU_TEMP_SENSOR_ID, tsens); + gpu_write(gpu, REG_A5XX_GPMU_DELTA_TEMP_THRESHOLD, 0x01); + gpu_write(gpu, REG_A5XX_GPMU_TEMP_SENSOR_CONFIG, 0x01); + + /* Until we get clock scaling 0 is always the active power level */ + gpu_write(gpu, REG_A5XX_GPMU_GPMU_VOLTAGE, 0x80000000 | 0); + + gpu_write(gpu, REG_A5XX_GPMU_BASE_LEAKAGE, a5xx_gpu->lm_leakage); + + of_property_read_u32(GPU_OF_NODE(gpu), "qcom,lm-limit", &lm_limit); + + gpu_write(gpu, REG_A5XX_GPMU_GPMU_PWR_THRESHOLD, 0x80000000 | lm_limit); + + gpu_write(gpu, REG_A5XX_GPMU_BEC_ENABLE, 0x10001FFF); + gpu_write(gpu, REG_A5XX_GDPM_CONFIG1, 0x00201FF1); + + /* Write the voltage table */ + + /* Get the max-power from the device tree */ + of_property_read_u32(GPU_OF_NODE(gpu), "qcom,lm-max-power", &max_power); + + gpu_write(gpu, REG_A5XX_GPMU_BEC_ENABLE, 0x10001FFF); + gpu_write(gpu, REG_A5XX_GDPM_CONFIG1, 0x201FF1); + + gpu_write(gpu, AGC_MSG_STATE, 1); + gpu_write(gpu, AGC_MSG_COMMAND, AGC_POWER_CONFIG_PRODUCTION_ID); + + /* + * For now just write the one voltage level - we will do more when we + * can do scaling + */ + gpu_write(gpu, AGC_MSG_PAYLOAD(0), max_power); + gpu_write(gpu, AGC_MSG_PAYLOAD(1), 1); + + gpu_write(gpu, AGC_MSG_PAYLOAD(2), _get_mvolts(gpu, gpu->fast_rate)); + gpu_write(gpu, AGC_MSG_PAYLOAD(3), gpu->fast_rate / 1000000); + + gpu_write(gpu, AGC_MSG_PAYLOAD_SIZE, 4 * sizeof(uint32_t)); + gpu_write(gpu, AGC_INIT_MSG_MAGIC, AGC_INIT_MSG_VALUE); +} + +/* Enable SP/TP cpower collapse */ +static void a5xx_pc_init(struct msm_gpu *gpu) +{ + gpu_write(gpu, REG_A5XX_GPMU_PWR_COL_INTER_FRAME_CTRL, 0x7F); + gpu_write(gpu, REG_A5XX_GPMU_PWR_COL_BINNING_CTRL, 0); + gpu_write(gpu, REG_A5XX_GPMU_PWR_COL_INTER_FRAME_HYST, 0xA0080); + gpu_write(gpu, REG_A5XX_GPMU_PWR_COL_STAGGER_DELAY, 0x600040); +} + +/* Enable the GPMU microcontroller */ +static int a5xx_gpmu_init(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + struct msm_ringbuffer *ring = gpu->rb; + + if (!a5xx_gpu->gpmu_dwords) + return 0; + + /* Turn off protected mode for this operation */ + OUT_PKT7(ring, CP_SET_PROTECTED_MODE, 1); + OUT_RING(ring, 0); + + /* Kick off the IB to load the GPMU microcode */ + OUT_PKT7(ring, CP_INDIRECT_BUFFER_PFE, 3); + OUT_RING(ring, lower_32_bits(a5xx_gpu->gpmu_iova)); + OUT_RING(ring, upper_32_bits(a5xx_gpu->gpmu_iova)); + OUT_RING(ring, a5xx_gpu->gpmu_dwords); + + /* Turn back on protected mode */ + OUT_PKT7(ring, CP_SET_PROTECTED_MODE, 1); + OUT_RING(ring, 1); + + gpu->funcs->flush(gpu); + + if (!gpu->funcs->idle(gpu)) { + DRM_ERROR("%s: Unable to load GPMU firmware. GPMU will not be active\n", + gpu->name); + return -EINVAL; + } + + gpu_write(gpu, REG_A5XX_GPMU_WFI_CONFIG, 0x4014); + + /* Kick off the GPMU */ + gpu_write(gpu, REG_A5XX_GPMU_CM3_SYSRESET, 0x0); + + /* + * Wait for the GPMU to respond. It isn't fatal if it doesn't, we just + * won't have advanced power collapse. + */ + if (spin_usecs(gpu, 25, REG_A5XX_GPMU_GENERAL_0, 0xFFFFFFFF, + 0xBABEFACE)) + DRM_ERROR("%s: GPMU firmware initialization timed out\n", + gpu->name); + + return 0; +} + +/* Enable limits management */ +static void a5xx_lm_enable(struct msm_gpu *gpu) +{ + gpu_write(gpu, REG_A5XX_GDPM_INT_MASK, 0x0); + gpu_write(gpu, REG_A5XX_GDPM_INT_EN, 0x0A); + gpu_write(gpu, REG_A5XX_GPMU_GPMU_VOLTAGE_INTR_EN_MASK, 0x01); + gpu_write(gpu, REG_A5XX_GPMU_TEMP_THRESHOLD_INTR_EN_MASK, 0x50000); + gpu_write(gpu, REG_A5XX_GPMU_THROTTLE_UNMASK_FORCE_CTRL, 0x30000); + + gpu_write(gpu, REG_A5XX_GPMU_CLOCK_THROTTLE_CTRL, 0x011); +} + +int a5xx_power_init(struct msm_gpu *gpu) +{ + int ret; + + /* Set up the limits management */ + a5xx_lm_setup(gpu); + + /* Set up SP/TP power collpase */ + a5xx_pc_init(gpu); + + /* Start the GPMU */ + ret = a5xx_gpmu_init(gpu); + if (ret) + return ret; + + /* Start the limits management */ + a5xx_lm_enable(gpu); + + return 0; +} + +static int _read_header(unsigned int *data, uint32_t fwsize, + unsigned int *major, unsigned int *minor) +{ + uint32_t size; + unsigned int i; + + /* First dword of the header is the header size */ + if (fwsize < 4) + return -EINVAL; + + size = data[0]; + + /* Make sure the header isn't too big and is a multiple of two */ + if ((size % 2) || (size > 10) || size > (fwsize >> 2)) + return -EINVAL; + + /* Read the values in pairs */ + for (i = 1; i < size; i += 2) { + switch (data[i]) { + case 1: + *major = data[i + 1]; + break; + case 2: + *minor = data[i + 1]; + break; + default: + /* Invalid values are non fatal */ + break; + } + } + + return 0; +} + +/* + * Make sure cur_major and cur_minor are greater than or equal to the minimum + * allowable major/minor + */ +static inline bool _check_gpmu_version(uint32_t cur_major, uint32_t cur_minor, + uint32_t min_major, uint32_t min_minor) +{ + return ((cur_major > min_major) || + ((cur_major == min_major) && (cur_minor >= min_minor))); +} + +void a5xx_gpmu_ucode_init(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + struct drm_device *drm = gpu->dev; + const char *name; + const struct firmware *fw; + uint32_t version[2] = { 0, 0 }; + uint32_t dwords = 0, offset = 0; + uint32_t major = 0, minor = 0, bosize; + unsigned int *data, *ptr, *cmds; + unsigned int cmds_size; + + if (a5xx_gpu->gpmu_bo) + return; + + /* + * Read the firmware name from the device tree - if it doesn't exist + * then don't initialize the GPMU for this target + */ + if (of_property_read_string(GPU_OF_NODE(gpu), "qcom,gpmu-firmware", + &name)) + return; + + /* + * The version isn't mandatory, but if it exists, we need to enforce + * that the version of the GPMU firmware matches or is newer than the + * value + */ + of_property_read_u32_array(GPU_OF_NODE(gpu), "qcom,gpmu-version", + version, 2); + + /* Get the firmware */ + if (request_firmware(&fw, name, drm->dev)) { + DRM_ERROR("%s: Could not get GPMU firmware. GPMU will not be active\n", + gpu->name); + return; + } + + data = (unsigned int *) fw->data; + + /* + * The first dword is the size of the remaining data in dwords. Use it + * as a checksum of sorts and make sure it matches the actual size of + * the firmware that we read + */ + + if (fw->size < 8 || (data[0] < 2) || (data[0] >= (fw->size >> 2))) + goto out; + + /* The second dword is an ID - look for 2 (GPMU_FIRMWARE_ID) */ + if (data[1] != 2) + goto out; + + /* Read the header and get the major/minor of the read firmware */ + if (_read_header(&data[2], fw->size - 8, &major, &minor)) + goto out; + + if (!_check_gpmu_version(major, minor, version[0], version[1])) { + DRM_ERROR("%s: Loaded GPMU version %d.%d is too old\n", + gpu->name, major, minor); + goto out; + } + + cmds = data + data[2] + 3; + cmds_size = data[0] - data[2] - 2; + + /* + * A single type4 opcode can only have so many values attached so + * add enough opcodes to load the all the commands + */ + bosize = (cmds_size + (cmds_size / TYPE4_MAX_PAYLOAD) + 1) << 2; + + mutex_lock(&drm->struct_mutex); + a5xx_gpu->gpmu_bo = msm_gem_new(drm, bosize, MSM_BO_UNCACHED); + mutex_unlock(&drm->struct_mutex); + + if (IS_ERR(a5xx_gpu->gpmu_bo)) + goto err; + + if (msm_gem_get_iova(a5xx_gpu->gpmu_bo, gpu->aspace, + &a5xx_gpu->gpmu_iova)) + goto err; + + ptr = msm_gem_vaddr(a5xx_gpu->gpmu_bo); + if (!ptr) + goto err; + + while (cmds_size > 0) { + int i; + uint32_t _size = cmds_size > TYPE4_MAX_PAYLOAD ? + TYPE4_MAX_PAYLOAD : cmds_size; + + ptr[dwords++] = PKT4(REG_A5XX_GPMU_INST_RAM_BASE + offset, + _size); + + for (i = 0; i < _size; i++) + ptr[dwords++] = *cmds++; + + offset += _size; + cmds_size -= _size; + } + + a5xx_gpu->gpmu_dwords = dwords; + + goto out; + +err: + if (a5xx_gpu->gpmu_iova) + msm_gem_put_iova(a5xx_gpu->gpmu_bo, gpu->aspace); + if (a5xx_gpu->gpmu_bo) + drm_gem_object_unreference_unlocked(a5xx_gpu->gpmu_bo); + + a5xx_gpu->gpmu_bo = NULL; + a5xx_gpu->gpmu_iova = 0; + a5xx_gpu->gpmu_dwords = 0; + +out: + /* No need to keep that firmware laying around anymore */ + release_firmware(fw); +} diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index 190e864a2cf9..a3073ef38795 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -148,6 +148,9 @@ struct adreno_platform_config { __ret; \ }) +#define GPU_OF_NODE(_g) \ + (((struct msm_drm_private *) \ + ((_g)->dev->dev_private))->gpu_pdev->dev.of_node) static inline bool adreno_is_a3xx(struct adreno_gpu *gpu) { From ef3370e0ca6fa8d4becc7e61c13cc004cd1c6fc8 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:20 -0700 Subject: [PATCH 15/19] drm/msm: gpu: Use the zap shader on 5XX if we can The A5XX GPU powers on in "secure" mode. In secure mode the GPU can only render to buffers that are marked as secure and inaccessible to the kernel and user through a series of hardware protections. In practice secure mode is used to draw things like a UI on a secure video frame. In order to switch out of secure mode the GPU executes a special shader that clears out the GMEM and other sensitve registers and then writes a register. Because the kernel can't be trusted the shader binary is signed and verified and programmed by the trustzone using the PIL loader to upload the binary and access to the special register is blocked with hardware protection. So in summary, to do secure mode correctly you need 1) a friendly trustzone, 2) PIL loader support and 3) a verified zap shader. For targets without secure support there is an out: if the trustzone doesn't support secure then there are no hardware protections and we can freely write the SECVID_TRUST register from the CPU. Unfortunately we don't have any good way to figure out at runtime if the trustzone is secure or not so we use a cheat. If there is a zap shader specified in the device tree for the target, we assume that the trustzone is secure. No zap shader definition means that the target is not secure. The downside of course is that if you guess wrong you trigger a XPU violation which usually ends up in a system crash but thats a problem that shows up immediately. Change-Id: Ic0dedbad8738ad1fac8a7bb8d76e1988aa49f2c8 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a5xx_gpu.c | 90 ++++++++++++++++++++++++++- drivers/gpu/drm/msm/adreno/a5xx_gpu.h | 6 ++ 2 files changed, 94 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c index fd92312529a6..056b6c89d877 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -303,6 +303,67 @@ static int a5xx_ucode_init(struct msm_gpu *gpu) return 0; } +#ifdef CONFIG_MSM_SUBSYSTEM_RESTART + +#include +#include + +static void a5xx_zap_shader_init(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + const char *name; + void *ptr; + + /* If no zap shader was defined, we'll assume that none is needed */ + if (of_property_read_string(GPU_OF_NODE(gpu), "qcom,zap-shader", &name)) + return; + + /* + * If the zap shader has already been loaded then just ask the SCM to + * re-initialize the registers (not needed if CPZ retention is a thing) + */ + if (test_bit(A5XX_ZAP_SHADER_LOADED, &a5xx_gpu->flags)) { + int ret; + struct scm_desc desc = { 0 }; + + if (of_property_read_bool(GPU_OF_NODE(gpu), + "qcom,cpz-retention")) + return; + + desc.args[0] = 0; + desc.args[1] = 13; + desc.arginfo = SCM_ARGS(2); + + ret = scm_call2(SCM_SIP_FNID(SCM_SVC_BOOT, 0x0A), &desc); + if (ret) + DRM_ERROR( + "%s: zap-shader resume failed with error %d\n", + gpu->name, ret); + + return; + } + + ptr = subsystem_get(name); + + if (IS_ERR_OR_NULL(ptr)) { + DRM_ERROR("%s: Unable to load the zap shader: %ld\n", gpu->name, + IS_ERR(ptr) ? PTR_ERR(ptr) : -ENODEV); + } else { + set_bit(A5XX_ZAP_SHADER_LOADED, &a5xx_gpu->flags); + } +} +#else +static void a5xx_zap_shader_init(struct msm_gpu *gpu) +{ + if (of_find_property(GPU_OF_NODE(gpu), "qcom,zap-shader", NULL)) + return; + + DRM_INFO_ONCE("%s: Zap shader is defined but loader isn't available\n", + gpu->name); +} +#endif + #define A5XX_INT_MASK (A5XX_RBBM_INT_0_MASK_RBBM_AHB_ERROR | \ A5XX_RBBM_INT_0_MASK_RBBM_TRANSFER_TIMEOUT | \ A5XX_RBBM_INT_0_MASK_RBBM_ME_MS_TIMEOUT | \ @@ -320,6 +381,7 @@ static int a5xx_hw_init(struct msm_gpu *gpu) struct msm_drm_private *priv = gpu->dev->dev_private; struct platform_device *pdev = priv->gpu_pdev; struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); int ret, bit = 0; pm_qos_update_request(&gpu->pm_qos_req_dma, 101); @@ -414,6 +476,9 @@ static int a5xx_hw_init(struct msm_gpu *gpu) } } + /* Try to load and initialize the zap shader if applicable */ + a5xx_zap_shader_init(gpu); + /* Protect registers from the CP */ gpu_write(gpu, REG_A5XX_CP_PROTECT_CNTL, 0x00000007); @@ -500,8 +565,29 @@ static int a5xx_hw_init(struct msm_gpu *gpu) return -EINVAL; } - /* Put the GPU into insecure mode */ - gpu_write(gpu, REG_A5XX_RBBM_SECVID_TRUST_CNTL, 0x0); + /* + * If a zap shader was specified in the device tree, assume that we are + * on a secure device that blocks access to the RBBM_SECVID registers + * so we need to use the CP to switch out of secure mode. If a zap + * shader was NOT specified then we assume we are on an unlocked device. + * If we guessed wrong then the access to the register will probably + * cause a XPU violation. + */ + if (test_bit(A5XX_ZAP_SHADER_LOADED, &a5xx_gpu->flags)) { + struct msm_ringbuffer *ring = gpu->rb; + + OUT_PKT7(ring, CP_SET_SECURE_MODE, 1); + OUT_RING(ring, 0x00000000); + + gpu->funcs->flush(gpu); + if (!gpu->funcs->idle(gpu)) + return -EINVAL; + } else { + /* Print a warning so if we die, we know why */ + dev_warn_once(gpu->dev->dev, + "Zap shader not enabled - using SECVID_TRUST_CNTL instead\n"); + gpu_write(gpu, REG_A5XX_RBBM_SECVID_TRUST_CNTL, 0x0); + } pm_qos_update_request(&gpu->pm_qos_req_dma, 501); diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h index 1590f845d554..e82f54063877 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.h +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.h @@ -21,7 +21,13 @@ #include "a5xx.xml.h" +enum { + A5XX_ZAP_SHADER_LOADED = 1, +}; + struct a5xx_gpu { + unsigned long flags; + struct adreno_gpu base; struct platform_device *pdev; From 482c4d814b1bcab47aa89c5d49b6cad4c59ff423 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:21 -0700 Subject: [PATCH 16/19] drm/msm: gpu: Move DT probing and configuration to adreno_gpu In the current code the probed configuration is pretty light and it can be easily probed at boot time and stored in a config struct. As more features come online the configuration becomes tougher and more platform specific. It makes sense to do the configuration in real time so move the majority of the config over to adreno_gpu and probe at load. Change-Id: Ic0dedbade871aae2d39e6b036a5e73e4347fc7ca Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a5xx_gpu.c | 21 ++++ drivers/gpu/drm/msm/adreno/a5xx_power.c | 5 +- drivers/gpu/drm/msm/adreno/adreno_device.c | 48 ++------ drivers/gpu/drm/msm/adreno/adreno_gpu.c | 127 +++++++++++++++++++-- drivers/gpu/drm/msm/adreno/adreno_gpu.h | 6 +- drivers/gpu/drm/msm/msm_gpu.c | 42 ++++--- drivers/gpu/drm/msm/msm_gpu.h | 6 +- 7 files changed, 175 insertions(+), 80 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c index 056b6c89d877..a66aa1065e51 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -998,6 +998,26 @@ static void a530_efuse_leakage(struct platform_device *pdev, ((leakage_pwr_on * coeff) / 100); } +/* Read the speed bin from the efuses */ +static void a530_efuse_bin(struct platform_device *pdev, + struct adreno_gpu *adreno_gpu, void *base, + size_t size) +{ + uint32_t speed_bin[3]; + uint32_t val; + + if (of_property_read_u32_array(pdev->dev.of_node, + "qcom,gpu-speed-bin", speed_bin, 3)) + return; + + if (size < speed_bin[0] + 4) + return; + + val = readl_relaxed(base + speed_bin[0]); + + adreno_gpu->speed_bin = (val & speed_bin[1]) >> speed_bin[2]; +} + /* Read target specific configuration from the efuses */ static void a5xx_efuses_read(struct platform_device *pdev, struct adreno_gpu *adreno_gpu) @@ -1023,6 +1043,7 @@ static void a5xx_efuses_read(struct platform_device *pdev, if (!base) return; + a530_efuse_bin(pdev, adreno_gpu, base, resource_size(res)); a530_efuse_leakage(pdev, adreno_gpu, base, resource_size(res)); iounmap(base); diff --git a/drivers/gpu/drm/msm/adreno/a5xx_power.c b/drivers/gpu/drm/msm/adreno/a5xx_power.c index 6bc1d5cbb4fb..38260241a672 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_power.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_power.c @@ -114,6 +114,7 @@ static void a5xx_lm_setup(struct msm_gpu *gpu) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); + uint32_t rate = gpu->gpufreq[gpu->active_level]; uint32_t tsens = 0; uint32_t lm_limit = 6000; uint32_t max_power = 0; @@ -160,8 +161,8 @@ static void a5xx_lm_setup(struct msm_gpu *gpu) gpu_write(gpu, AGC_MSG_PAYLOAD(0), max_power); gpu_write(gpu, AGC_MSG_PAYLOAD(1), 1); - gpu_write(gpu, AGC_MSG_PAYLOAD(2), _get_mvolts(gpu, gpu->fast_rate)); - gpu_write(gpu, AGC_MSG_PAYLOAD(3), gpu->fast_rate / 1000000); + gpu_write(gpu, AGC_MSG_PAYLOAD(2), _get_mvolts(gpu, rate)); + gpu_write(gpu, AGC_MSG_PAYLOAD(3), rate / 1000000); gpu_write(gpu, AGC_MSG_PAYLOAD_SIZE, 4 * sizeof(uint32_t)); gpu_write(gpu, AGC_INIT_MSG_MAGIC, AGC_INIT_MSG_VALUE); diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c index 7749b60a876c..5ce5f5c7d0b2 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_device.c +++ b/drivers/gpu/drm/msm/adreno/adreno_device.c @@ -184,22 +184,19 @@ static void set_gpu_pdev(struct drm_device *dev, priv->gpu_pdev = pdev; } -static const struct { - const char *str; - uint32_t flag; -} quirks[] = { - { "qcom,gpu-quirk-two-pass-use-wfi", ADRENO_QUIRK_TWO_PASS_USE_WFI }, - { "qcom,gpu-quirk-fault-detect-mask", ADRENO_QUIRK_FAULT_DETECT_MASK }, -}; - static int adreno_bind(struct device *dev, struct device *master, void *data) { static struct adreno_platform_config config = {}; - struct device_node *child, *node = dev->of_node; - u32 val; - int ret, i; + uint32_t val = 0; + int ret; - ret = of_property_read_u32(node, "qcom,chipid", &val); + /* + * Read the chip ID from the device tree at bind time - we use this + * information to load the correct functions. All the rest of the + * (extensive) device tree probing should happen in the GPU specific + * code + */ + ret = of_property_read_u32(dev->of_node, "qcom,chipid", &val); if (ret) { dev_err(dev, "could not find chipid: %d\n", ret); return ret; @@ -208,33 +205,6 @@ static int adreno_bind(struct device *dev, struct device *master, void *data) config.rev = ADRENO_REV((val >> 24) & 0xff, (val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff); - /* find clock rates: */ - config.fast_rate = 0; - config.slow_rate = ~0; - for_each_child_of_node(node, child) { - if (of_device_is_compatible(child, "qcom,gpu-pwrlevels")) { - struct device_node *pwrlvl; - for_each_child_of_node(child, pwrlvl) { - ret = of_property_read_u32(pwrlvl, "qcom,gpu-freq", &val); - if (ret) { - dev_err(dev, "could not find gpu-freq: %d\n", ret); - return ret; - } - config.fast_rate = max(config.fast_rate, val); - config.slow_rate = min(config.slow_rate, val); - } - } - } - - if (!config.fast_rate) { - dev_err(dev, "could not find clk rates\n"); - return -ENXIO; - } - - for (i = 0; i < ARRAY_SIZE(quirks); i++) - if (of_property_read_bool(node, quirks[i].str)) - config.quirks |= quirks[i].flag; - dev->platform_data = &config; set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev)); return 0; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 49f01c2e8a8d..e67c63c9a3ac 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -42,7 +42,7 @@ int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value) (adreno_gpu->rev.core << 24); return 0; case MSM_PARAM_MAX_FREQ: - *value = adreno_gpu->base.fast_rate; + *value = gpu->gpufreq[gpu->active_level]; return 0; case MSM_PARAM_TIMESTAMP: if (adreno_gpu->funcs->get_timestamp) @@ -351,6 +351,119 @@ static const char *iommu_ports[] = { "gfx3d_user", }; +/* Read the set of powerlevels */ +static int _adreno_get_pwrlevels(struct msm_gpu *gpu, struct device_node *node) +{ + struct device_node *child; + + gpu->active_level = 1; + + /* The device tree will tell us the best clock to initialize with */ + of_property_read_u32(node, "qcom,initial-pwrlevel", &gpu->active_level); + + if (gpu->active_level >= ARRAY_SIZE(gpu->gpufreq)) + gpu->active_level = 1; + + for_each_child_of_node(node, child) { + unsigned int index; + + if (of_property_read_u32(child, "reg", &index)) + return -EINVAL; + + if (index >= ARRAY_SIZE(gpu->gpufreq)) + continue; + + gpu->nr_pwrlevels = max(gpu->nr_pwrlevels, index + 1); + + of_property_read_u32(child, "qcom,gpu-freq", + &gpu->gpufreq[index]); + of_property_read_u32(child, "qcom,bus-freq", + &gpu->busfreq[index]); + } + + DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u", + gpu->gpufreq[gpu->active_level], + gpu->gpufreq[gpu->nr_pwrlevels - 1], + gpu->busfreq[gpu->active_level]); + + return 0; +} + +/* + * Escape valve for targets that don't define the binning nodes. Get the + * first powerlevel node and parse it + */ +static int adreno_get_legacy_pwrlevels(struct msm_gpu *gpu, + struct device_node *parent) +{ + struct device_node *child; + + child = of_find_node_by_name(parent, "qcom,gpu-pwrlevels"); + if (child) + return _adreno_get_pwrlevels(gpu, child); + + dev_err(gpu->dev->dev, "Unable to parse any powerlevels\n"); + return -EINVAL; +} + +/* Get the powerlevels for the target */ +static int adreno_get_pwrlevels(struct msm_gpu *gpu, struct device_node *parent) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct device_node *node, *child; + + /* See if the target has defined a number of power bins */ + node = of_find_node_by_name(parent, "qcom,gpu-pwrlevel-bins"); + if (!node) { + /* If not look for the qcom,gpu-pwrlevels node */ + return adreno_get_legacy_pwrlevels(gpu, parent); + } + + for_each_child_of_node(node, child) { + unsigned int bin; + + if (of_property_read_u32(child, "qcom,speed-bin", &bin)) + continue; + + /* + * If the bin matches the bin specified by the fuses, then we + * have a winner - parse it + */ + if (adreno_gpu->speed_bin == bin) + return _adreno_get_pwrlevels(gpu, child); + } + + return -ENODEV; +} + +static const struct { + const char *str; + uint32_t flag; +} quirks[] = { + { "qcom,gpu-quirk-two-pass-use-wfi", ADRENO_QUIRK_TWO_PASS_USE_WFI }, + { "qcom,gpu-quirk-fault-detect-mask", ADRENO_QUIRK_FAULT_DETECT_MASK }, +}; + +/* Parse the statistics from the device tree */ +static int adreno_of_parse(struct platform_device *pdev, struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + struct device_node *node = pdev->dev.of_node; + int i, ret; + + /* Probe the powerlevels */ + ret = adreno_get_pwrlevels(gpu, node); + if (ret) + return ret; + + /* Check to see if any quirks were specified in the device tree */ + for (i = 0; i < ARRAY_SIZE(quirks); i++) + if (of_property_read_bool(node, quirks[i].str)) + adreno_gpu->quirks |= quirks[i].flag; + + return 0; +} + int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, struct adreno_gpu *adreno_gpu, const struct adreno_gpu_funcs *funcs) { @@ -364,17 +477,9 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, adreno_gpu->gmem = adreno_gpu->info->gmem; adreno_gpu->revn = adreno_gpu->info->revn; adreno_gpu->rev = config->rev; - adreno_gpu->quirks = config->quirks; - gpu->fast_rate = config->fast_rate; - gpu->slow_rate = config->slow_rate; - gpu->bus_freq = config->bus_freq; -#ifdef DOWNSTREAM_CONFIG_MSM_BUS_SCALING - gpu->bus_scale_table = config->bus_scale_table; -#endif - - DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u", - gpu->fast_rate, gpu->slow_rate, gpu->bus_freq); + /* Get the rest of the target configuration from the device tree */ + adreno_of_parse(pdev, gpu); ret = msm_gpu_init(drm, pdev, &adreno_gpu->base, &funcs->base, adreno_gpu->info->name, "kgsl_3d0_reg_memory", "kgsl_3d0_irq", diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index a3073ef38795..b9e1ea094c71 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -121,17 +121,13 @@ struct adreno_gpu { const unsigned int *reg_offsets; uint32_t quirks; + uint32_t speed_bin; }; #define to_adreno_gpu(x) container_of(x, struct adreno_gpu, base) /* platform config data (ie. from DT, or pdata) */ struct adreno_platform_config { struct adreno_rev rev; - uint32_t fast_rate, slow_rate, bus_freq; -#ifdef DOWNSTREAM_CONFIG_MSM_BUS_SCALING - struct msm_bus_scale_pdata *bus_scale_table; -#endif - uint32_t quirks; }; #define ADRENO_IDLE_TIMEOUT msecs_to_jiffies(1000) diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 30d2fc91c84d..08ecc089611f 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -90,19 +90,18 @@ static int disable_pwrrail(struct msm_gpu *gpu) static int enable_clk(struct msm_gpu *gpu) { - struct clk *rate_clk = NULL; + uint32_t rate = gpu->gpufreq[gpu->active_level]; int i; - /* NOTE: kgsl_pwrctrl_clk() ignores grp_clks[0].. */ - for (i = ARRAY_SIZE(gpu->grp_clks) - 1; i > 0; i--) { - if (gpu->grp_clks[i]) { - clk_prepare(gpu->grp_clks[i]); - rate_clk = gpu->grp_clks[i]; - } - } + clk_set_rate(gpu->grp_clks[0], rate); - if (rate_clk && gpu->fast_rate) - clk_set_rate(rate_clk, gpu->fast_rate); + if (gpu->grp_clks[3]) + clk_set_rate(gpu->grp_clks[3], 19200000); + + /* NOTE: kgsl_pwrctrl_clk() ignores grp_clks[0].. */ + for (i = ARRAY_SIZE(gpu->grp_clks) - 1; i > 0; i--) + if (gpu->grp_clks[i]) + clk_prepare(gpu->grp_clks[i]); for (i = ARRAY_SIZE(gpu->grp_clks) - 1; i > 0; i--) if (gpu->grp_clks[i]) @@ -113,24 +112,19 @@ static int enable_clk(struct msm_gpu *gpu) static int disable_clk(struct msm_gpu *gpu) { - struct clk *rate_clk = NULL; + uint32_t rate = gpu->gpufreq[gpu->nr_pwrlevels - 1]; int i; /* NOTE: kgsl_pwrctrl_clk() ignores grp_clks[0].. */ - for (i = ARRAY_SIZE(gpu->grp_clks) - 1; i > 0; i--) { - if (gpu->grp_clks[i]) { + for (i = ARRAY_SIZE(gpu->grp_clks) - 1; i > 0; i--) + if (gpu->grp_clks[i]) clk_disable(gpu->grp_clks[i]); - rate_clk = gpu->grp_clks[i]; - } - } - - if (rate_clk && gpu->slow_rate) - clk_set_rate(rate_clk, gpu->slow_rate); for (i = ARRAY_SIZE(gpu->grp_clks) - 1; i > 0; i--) if (gpu->grp_clks[i]) clk_unprepare(gpu->grp_clks[i]); + clk_set_rate(gpu->grp_clks[0], rate); return 0; } @@ -138,8 +132,9 @@ static int enable_axi(struct msm_gpu *gpu) { if (gpu->ebi1_clk) clk_prepare_enable(gpu->ebi1_clk); - if (gpu->bus_freq) - bs_set(gpu, gpu->bus_freq); + + if (gpu->busfreq[gpu->active_level]) + bs_set(gpu, gpu->busfreq[gpu->active_level]); return 0; } @@ -147,7 +142,8 @@ static int disable_axi(struct msm_gpu *gpu) { if (gpu->ebi1_clk) clk_disable_unprepare(gpu->ebi1_clk); - if (gpu->bus_freq) + + if (gpu->busfreq[gpu->active_level]) bs_set(gpu, 0); return 0; } @@ -627,6 +623,8 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev, gpu->grp_clks[i] = NULL; } + gpu->grp_clks[0] = gpu->grp_clks[1]; + gpu->ebi1_clk = devm_clk_get(&pdev->dev, "bus_clk"); DBG("ebi1_clk: %p", gpu->ebi1_clk); if (IS_ERR(gpu->ebi1_clk)) diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index a45faf2d06e5..c6f1d3bd36e9 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -101,7 +101,11 @@ struct msm_gpu { /* Power Control: */ struct regulator *gpu_reg, *gpu_cx; struct clk *ebi1_clk, *grp_clks[7]; - uint32_t fast_rate, slow_rate, bus_freq; + + uint32_t gpufreq[10]; + uint32_t busfreq[10]; + uint32_t nr_pwrlevels; + uint32_t active_level; struct pm_qos_request pm_qos_req_dma; From 6645ddebe87e9a02dd119c66ba8f5a2a975d6af2 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:21 -0700 Subject: [PATCH 17/19] drm/msm: gpu: Bring in new registers for A540 Bring in a new revision of the rnndb register to fill out some A540 registers. Change-Id: Ic0dedbad0daf9ad2301479354a83c05bdc301163 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a5xx.xml.h | 35 ++++++++++++++++++++------- 1 file changed, 26 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a5xx.xml.h b/drivers/gpu/drm/msm/adreno/a5xx.xml.h index f480c94beb62..bfee2fd83462 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx.xml.h +++ b/drivers/gpu/drm/msm/adreno/a5xx.xml.h @@ -8,15 +8,15 @@ http://github.com/freedreno/envytools/ git clone https://github.com/freedreno/envytools.git The rules-ng-ng source files this header was generated from are: -- ./adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) -- ./freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) -- ./adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) -- ./adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) -- ./adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) -- ./adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) -- ./adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) -- ./adreno/a5xx.xml ( 81207 bytes, from 2016-10-26 19:36:59) -- ./adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno.xml ( 431 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//freedreno_copyright.xml ( 1572 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno/a2xx.xml ( 32901 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno/adreno_common.xml ( 12025 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno/adreno_pm4.xml ( 19684 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno/a3xx.xml ( 83840 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno/a4xx.xml ( 110708 bytes, from 2016-10-24 21:12:27) +- /local3/projects/drm/envytools/rnndb//adreno/a5xx.xml ( 81546 bytes, from 2016-10-31 16:38:41) +- /local3/projects/drm/envytools/rnndb//adreno/ocmem.xml ( 1773 bytes, from 2016-10-24 21:12:27) Copyright (C) 2013-2016 by the following authors: - Rob Clark (robclark) @@ -1821,6 +1821,14 @@ static inline uint32_t A5XX_VSC_BIN_SIZE_Y(uint32_t val) #define REG_A5XX_GPMU_TEMP_SENSOR_ID 0x0000ac00 #define REG_A5XX_GPMU_TEMP_SENSOR_CONFIG 0x0000ac01 +#define A5XX_GPMU_TEMP_SENSOR_CONFIG_ISENSE_STATUS__MASK 0x0000000f +#define A5XX_GPMU_TEMP_SENSOR_CONFIG_ISENSE_STATUS__SHIFT 0 +static inline uint32_t A5XX_GPMU_TEMP_SENSOR_CONFIG_ISENSE_STATUS(uint32_t val) +{ + return ((val) << A5XX_GPMU_TEMP_SENSOR_CONFIG_ISENSE_STATUS__SHIFT) & A5XX_GPMU_TEMP_SENSOR_CONFIG_ISENSE_STATUS__MASK; +} +#define A5XX_GPMU_TEMP_SENSOR_CONFIG_BCL_ENABLED 0x00000002 +#define A5XX_GPMU_TEMP_SENSOR_CONFIG_LLM_ENABLED 0x00000200 #define REG_A5XX_GPMU_TEMP_VAL 0x0000ac02 @@ -1849,8 +1857,17 @@ static inline uint32_t A5XX_VSC_BIN_SIZE_Y(uint32_t val) #define REG_A5XX_GPMU_GPMU_PWR_THRESHOLD 0x0000ac80 #define REG_A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL 0x0000acc4 +#define A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL_IDLE_FULL_LM 0x00000001 +#define A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL_STATE_OF_CHILD__MASK 0x00000030 +#define A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL_STATE_OF_CHILD__SHIFT 4 +static inline uint32_t A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL_STATE_OF_CHILD(uint32_t val) +{ + return ((val) << A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL_STATE_OF_CHILD__SHIFT) & A5XX_GPMU_GPMU_LLM_GLM_SLEEP_CTRL_STATE_OF_CHILD__MASK; +} #define REG_A5XX_GPMU_GPMU_LLM_GLM_SLEEP_STATUS 0x0000acc5 +#define A5XX_GPMU_GPMU_LLM_GLM_SLEEP_STATUS_IDLE_FULL_ACK 0x00000001 +#define A5XX_GPMU_GPMU_LLM_GLM_SLEEP_STATUS_WAKEUP_ACK 0x00000002 #define REG_A5XX_GDPM_CONFIG1 0x0000b80c From d82284934ca3c606a3aa2e498f412066ab723891 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:22 -0700 Subject: [PATCH 18/19] drm/msm: adreno: Add A540 support Add support for the A540 GPU which is a revision of the A530 target with somewhat more complicated power management support. Change-Id: Ic0dedbadeca628809b6de7c1843f665ecdc427fe Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/adreno/a5xx_gpu.c | 58 +++++------ drivers/gpu/drm/msm/adreno/a5xx_power.c | 106 ++++++++++++++++++--- drivers/gpu/drm/msm/adreno/adreno_device.c | 8 ++ drivers/gpu/drm/msm/adreno/adreno_gpu.h | 5 + 4 files changed, 132 insertions(+), 45 deletions(-) diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c index a66aa1065e51..8bc3c7bee3fb 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -56,12 +56,10 @@ static int a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, return 0; } -struct a5xx_hwcg { +static const struct { u32 offset; u32 value; -}; - -static const struct a5xx_hwcg a530_hwcg[] = { +} a5xx_hwcg[] = { {REG_A5XX_RBBM_CLOCK_CNTL_SP0, 0x02222222}, {REG_A5XX_RBBM_CLOCK_CNTL_SP1, 0x02222222}, {REG_A5XX_RBBM_CLOCK_CNTL_SP2, 0x02222222}, @@ -156,38 +154,22 @@ static const struct a5xx_hwcg a530_hwcg[] = { {REG_A5XX_RBBM_CLOCK_DELAY_VFD, 0x00002222} }; -static const struct { - int (*test)(struct adreno_gpu *gpu); - const struct a5xx_hwcg *regs; - unsigned int count; -} a5xx_hwcg_regs[] = { - { adreno_is_a530, a530_hwcg, ARRAY_SIZE(a530_hwcg), }, -}; - -static void _a5xx_enable_hwcg(struct msm_gpu *gpu, - const struct a5xx_hwcg *regs, unsigned int count) -{ - unsigned int i; - - for (i = 0; i < count; i++) - gpu_write(gpu, regs[i].offset, regs[i].value); - - gpu_write(gpu, REG_A5XX_RBBM_CLOCK_CNTL, 0xAAA8AA00); - gpu_write(gpu, REG_A5XX_RBBM_ISDB_CNT, 0x182); -} - static void a5xx_enable_hwcg(struct msm_gpu *gpu) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); unsigned int i; - for (i = 0; i < ARRAY_SIZE(a5xx_hwcg_regs); i++) { - if (a5xx_hwcg_regs[i].test(adreno_gpu)) { - _a5xx_enable_hwcg(gpu, a5xx_hwcg_regs[i].regs, - a5xx_hwcg_regs[i].count); - return; - } + for (i = 0; i < ARRAY_SIZE(a5xx_hwcg); i++) + gpu_write(gpu, a5xx_hwcg[i].offset, a5xx_hwcg[i].value); + + /* There are a few additional registers just for A540 */ + if (adreno_is_a540(adreno_gpu)) { + gpu_write(gpu, REG_A5XX_RBBM_CLOCK_DELAY_GPMU, 0x770); + gpu_write(gpu, REG_A5XX_RBBM_CLOCK_HYST_GPMU, 0x004); } + + gpu_write(gpu, REG_A5XX_RBBM_CLOCK_CNTL, 0xAAA8AA00); + gpu_write(gpu, REG_A5XX_RBBM_ISDB_CNT, 0x182); } static int a5xx_me_init(struct msm_gpu *gpu) @@ -387,6 +369,8 @@ static int a5xx_hw_init(struct msm_gpu *gpu) pm_qos_update_request(&gpu->pm_qos_req_dma, 101); gpu_write(gpu, REG_A5XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x00000003); + if (adreno_is_a540(adreno_gpu)) + gpu_write(gpu, REG_A5XX_VBIF_GATE_OFF_WRREQ_EN, 0x00000009); /* Make all blocks contribute to the GPU BUSY perf counter */ gpu_write(gpu, REG_A5XX_RBBM_PERFCTR_GPU_BUSY_MASKED, 0xFFFFFFFF); @@ -473,6 +457,10 @@ static int a5xx_hw_init(struct msm_gpu *gpu) gpu_write(gpu, REG_A5XX_TPL1_MODE_CNTL, bit << 7); gpu_write(gpu, REG_A5XX_RB_MODE_CNTL, bit << 1); + + if (adreno_is_a540(adreno_gpu)) + gpu_write(gpu, REG_A5XX_UCHE_DBG_ECO_CNTL_2, + bit); } } @@ -911,6 +899,8 @@ static int a5xx_pm_resume(struct msm_gpu *gpu) static int a5xx_pm_suspend(struct msm_gpu *gpu) { + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + /* Clear the VBIF pipe before shutting down */ gpu_write(gpu, REG_A5XX_VBIF_XIN_HALT_CTRL0, 0xF); @@ -922,8 +912,12 @@ static int a5xx_pm_suspend(struct msm_gpu *gpu) * Reset the VBIF before power collapse to avoid issue with FIFO * entries */ - gpu_write(gpu, REG_A5XX_RBBM_BLOCK_SW_RESET_CMD, 0x003C0000); - gpu_write(gpu, REG_A5XX_RBBM_BLOCK_SW_RESET_CMD, 0x00000000); + + if (adreno_is_a530(adreno_gpu)) { + /* These only need to be done for A530 */ + gpu_write(gpu, REG_A5XX_RBBM_BLOCK_SW_RESET_CMD, 0x003C0000); + gpu_write(gpu, REG_A5XX_RBBM_BLOCK_SW_RESET_CMD, 0x00000000); + } return msm_gpu_pm_suspend(gpu); } diff --git a/drivers/gpu/drm/msm/adreno/a5xx_power.c b/drivers/gpu/drm/msm/adreno/a5xx_power.c index 38260241a672..18bca141b425 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_power.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_power.c @@ -32,6 +32,18 @@ #define AGC_POWER_CONFIG_PRODUCTION_ID 1 #define AGC_INIT_MSG_VALUE 0xBABEFACE +/* AGC_LM_CONFIG (A540+) */ +#define AGC_LM_CONFIG (136/4) +#define AGC_LM_CONFIG_GPU_VERSION_SHIFT 17 +#define AGC_LM_CONFIG_ENABLE_GPMU_ADAPTIVE 1 +#define AGC_LM_CONFIG_THROTTLE_DISABLE (2 << 8) +#define AGC_LM_CONFIG_ISENSE_ENABLE (1 << 4) +#define AGC_LM_CONFIG_ENABLE_ERROR (3 << 4) +#define AGC_LM_CONFIG_LLM_ENABLED (1 << 16) +#define AGC_LM_CONFIG_BCL_DISABLED (1 << 24) + +#define AGC_LEVEL_CONFIG (140/4) + static struct { uint32_t reg; uint32_t value; @@ -109,14 +121,56 @@ static inline uint32_t _get_mvolts(struct msm_gpu *gpu, uint32_t freq) return (!IS_ERR(opp)) ? dev_pm_opp_get_voltage(opp) / 1000 : 0; } -/* Setup thermal limit management */ -static void a5xx_lm_setup(struct msm_gpu *gpu) +#define PAYLOAD_SIZE(_size) ((_size) * sizeof(u32)) +#define LM_DCVS_LIMIT 1 +#define LEVEL_CONFIG ~(0x303) + +/* Setup thermal limit management for A540 */ +static void a540_lm_setup(struct msm_gpu *gpu) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + u32 max_power = 0; + u32 rate = gpu->gpufreq[gpu->active_level]; + u32 config; + + /* The battery current limiter isn't enabled for A540 */ + config = AGC_LM_CONFIG_BCL_DISABLED; + config |= adreno_gpu->rev.patchid << AGC_LM_CONFIG_GPU_VERSION_SHIFT; + + /* For now disable GPMU side throttling */ + config |= AGC_LM_CONFIG_THROTTLE_DISABLE; + + /* Get the max-power from the device tree */ + of_property_read_u32(GPU_OF_NODE(gpu), "qcom,lm-max-power", &max_power); + + gpu_write(gpu, AGC_MSG_STATE, 0x80000001); + gpu_write(gpu, AGC_MSG_COMMAND, AGC_POWER_CONFIG_PRODUCTION_ID); + + /* + * For now just write the one voltage level - we will do more when we + * can do scaling + */ + gpu_write(gpu, AGC_MSG_PAYLOAD(0), max_power); + gpu_write(gpu, AGC_MSG_PAYLOAD(1), 1); + + gpu_write(gpu, AGC_MSG_PAYLOAD(2), _get_mvolts(gpu, rate)); + gpu_write(gpu, AGC_MSG_PAYLOAD(3), rate / 1000000); + + gpu_write(gpu, AGC_MSG_PAYLOAD(AGC_LM_CONFIG), config); + gpu_write(gpu, AGC_MSG_PAYLOAD(AGC_LEVEL_CONFIG), LEVEL_CONFIG); + gpu_write(gpu, AGC_MSG_PAYLOAD_SIZE, + PAYLOAD_SIZE(AGC_LEVEL_CONFIG + 1)); + + gpu_write(gpu, AGC_INIT_MSG_MAGIC, AGC_INIT_MSG_VALUE); +} + +/* Setup thermal limit management for A530 */ +static void a530_lm_setup(struct msm_gpu *gpu) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu); uint32_t rate = gpu->gpufreq[gpu->active_level]; uint32_t tsens = 0; - uint32_t lm_limit = 6000; uint32_t max_power = 0; unsigned int i; @@ -131,15 +185,8 @@ static void a5xx_lm_setup(struct msm_gpu *gpu) gpu_write(gpu, REG_A5XX_GPMU_DELTA_TEMP_THRESHOLD, 0x01); gpu_write(gpu, REG_A5XX_GPMU_TEMP_SENSOR_CONFIG, 0x01); - /* Until we get clock scaling 0 is always the active power level */ - gpu_write(gpu, REG_A5XX_GPMU_GPMU_VOLTAGE, 0x80000000 | 0); - gpu_write(gpu, REG_A5XX_GPMU_BASE_LEAKAGE, a5xx_gpu->lm_leakage); - of_property_read_u32(GPU_OF_NODE(gpu), "qcom,lm-limit", &lm_limit); - - gpu_write(gpu, REG_A5XX_GPMU_GPMU_PWR_THRESHOLD, 0x80000000 | lm_limit); - gpu_write(gpu, REG_A5XX_GPMU_BEC_ENABLE, 0x10001FFF); gpu_write(gpu, REG_A5XX_GDPM_CONFIG1, 0x00201FF1); @@ -209,7 +256,9 @@ static int a5xx_gpmu_init(struct msm_gpu *gpu) return -EINVAL; } - gpu_write(gpu, REG_A5XX_GPMU_WFI_CONFIG, 0x4014); + /* Clock gating setup for A530 targets */ + if (adreno_is_a530(adreno_gpu)) + gpu_write(gpu, REG_A5XX_GPMU_WFI_CONFIG, 0x4014); /* Kick off the GPMU */ gpu_write(gpu, REG_A5XX_GPMU_CM3_SYSRESET, 0x0); @@ -223,12 +272,29 @@ static int a5xx_gpmu_init(struct msm_gpu *gpu) DRM_ERROR("%s: GPMU firmware initialization timed out\n", gpu->name); + if (!adreno_is_a530(adreno_gpu)) { + u32 val = gpu_read(gpu, REG_A5XX_GPMU_GENERAL_1); + + if (val) { + DRM_ERROR("%s: GPMU firmare initialization failed: %d\n", + gpu->name, val); + return -EIO; + } + } + + /* FIXME: Clear GPMU interrupts? */ return 0; } /* Enable limits management */ static void a5xx_lm_enable(struct msm_gpu *gpu) { + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + + /* This init sequence only applies to A530 */ + if (!adreno_is_a530(adreno_gpu)) + return; + gpu_write(gpu, REG_A5XX_GDPM_INT_MASK, 0x0); gpu_write(gpu, REG_A5XX_GDPM_INT_EN, 0x0A); gpu_write(gpu, REG_A5XX_GPMU_GPMU_VOLTAGE_INTR_EN_MASK, 0x01); @@ -240,10 +306,24 @@ static void a5xx_lm_enable(struct msm_gpu *gpu) int a5xx_power_init(struct msm_gpu *gpu) { + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); int ret; + u32 lm_limit = 6000; - /* Set up the limits management */ - a5xx_lm_setup(gpu); + /* + * Set up the limit management + * first, do some generic setup: + */ + gpu_write(gpu, REG_A5XX_GPMU_GPMU_VOLTAGE, 0x80000000 | 0); + + of_property_read_u32(GPU_OF_NODE(gpu), "qcom,lm-limit", &lm_limit); + gpu_write(gpu, REG_A5XX_GPMU_GPMU_PWR_THRESHOLD, 0x80000000 | lm_limit); + + /* Now do the target specific setup */ + if (adreno_is_a530(adreno_gpu)) + a530_lm_setup(gpu); + else + a540_lm_setup(gpu); /* Set up SP/TP power collpase */ a5xx_pc_init(gpu); diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c index 5ce5f5c7d0b2..a498a60cd52d 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_device.c +++ b/drivers/gpu/drm/msm/adreno/adreno_device.c @@ -86,6 +86,14 @@ static const struct adreno_info gpulist[] = { .pfpfw = "a530_pfp.fw", .gmem = SZ_1M, .init = a5xx_gpu_init, + }, { + .rev = ADRENO_REV(5, 4, 0, ANY_ID), + .revn = 540, + .name = "A540", + .pm4fw = "a530_pm4.fw", + .pfpfw = "a530_pfp.fw", + .gmem = SZ_1M, + .init = a5xx_gpu_init, }, }; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index b9e1ea094c71..9bcf4bb705bd 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -199,6 +199,11 @@ static inline int adreno_is_a530(struct adreno_gpu *gpu) return gpu->revn == 530; } +static inline int adreno_is_a540(struct adreno_gpu *gpu) +{ + return gpu->revn == 540; +} + int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value); int adreno_hw_init(struct msm_gpu *gpu); uint32_t adreno_last_fence(struct msm_gpu *gpu); From a679678669843294c1381913bdbf5620c1229ba2 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 13 Feb 2017 10:14:22 -0700 Subject: [PATCH 19/19] drm/msm: Add hint to DRM_IOCTL_MSM_GEM_INFO to return an object IOVA Modify the 'pad' member of struct drm_msm_gem_info to 'hint'. If the user sets 'hint' to non-zero it means that they want a IOVA for the GEM object instead of a mmap() offset. Return the iova in the 'offset' member. Change-Id: Ic0dedbad543df80fdc4b74cd1cd924e9b8534b44 Signed-off-by: Jordan Crouse --- drivers/gpu/drm/msm/msm_drv.c | 23 +++++++++++++++++++++-- include/uapi/drm/msm_drm.h | 8 ++++++-- 2 files changed, 27 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index e223f376c908..9163b90981a2 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -1124,6 +1124,17 @@ static int msm_ioctl_gem_cpu_fini(struct drm_device *dev, void *data, return ret; } +static int msm_ioctl_gem_info_iova(struct drm_device *dev, + struct drm_gem_object *obj, uint64_t *iova) +{ + struct msm_drm_private *priv = dev->dev_private; + + if (!priv->gpu) + return -EINVAL; + + return msm_gem_get_iova(obj, priv->gpu->aspace, iova); +} + static int msm_ioctl_gem_info(struct drm_device *dev, void *data, struct drm_file *file) { @@ -1131,14 +1142,22 @@ static int msm_ioctl_gem_info(struct drm_device *dev, void *data, struct drm_gem_object *obj; int ret = 0; - if (args->pad) + if (args->flags & ~MSM_INFO_FLAGS) return -EINVAL; obj = drm_gem_object_lookup(dev, file, args->handle); if (!obj) return -ENOENT; - args->offset = msm_gem_mmap_offset(obj); + if (args->flags & MSM_INFO_IOVA) { + uint64_t iova; + + ret = msm_ioctl_gem_info_iova(dev, obj, &iova); + if (!ret) + args->offset = iova; + } else { + args->offset = msm_gem_mmap_offset(obj); + } drm_gem_object_unreference_unlocked(obj); diff --git a/include/uapi/drm/msm_drm.h b/include/uapi/drm/msm_drm.h index 4fe580f0680d..587d35ce1638 100644 --- a/include/uapi/drm/msm_drm.h +++ b/include/uapi/drm/msm_drm.h @@ -85,10 +85,14 @@ struct drm_msm_gem_new { __u32 handle; /* out */ }; +#define MSM_INFO_IOVA 0x01 + +#define MSM_INFO_FLAGS (MSM_INFO_IOVA) + struct drm_msm_gem_info { __u32 handle; /* in */ - __u32 pad; - __u64 offset; /* out, offset to pass to mmap() */ + __u32 flags; /* in - combination of MSM_INFO_* flags */ + __u64 offset; /* out, mmap() offset or iova */ }; #define MSM_PREP_READ 0x01