fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2025-12-22 22:10:10 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

freedreno: Specify GMEM tile alignment per GPU

Browse source 
They differ presumably based on # of CCU/SP and DDR bus topology.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21573>
This commit is contained in:
Rob Clark 2023-01-22 10:57:30 -08:00 • committed by Marge Bot
parent ad4271439c
commit 60bc7c0e22
2 changed files with 23 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

38
src/freedreno/common/freedreno_devices.py
View file

@ -108,19 +108,17 @@ class A6xxGPUInfo(GPUInfo):
into distinct sub-generations. The template parameter avoids into distinct sub-generations. The template parameter avoids
duplication of parameters that are unique to the sub-generation. duplication of parameters that are unique to the sub-generation.
""" """
def __init__(self, template, num_sp_cores, num_ccu, magic_regs): def __init__(self, template, num_ccu, tile_align_w, tile_align_h, magic_regs):
super().__init__(gmem_align_w = 16, gmem_align_h = 4, super().__init__(gmem_align_w = 16, gmem_align_h = 4,
tile_align_w = 32, tile_align_h = 32, tile_align_w = tile_align_w,
tile_align_h = tile_align_h,
tile_max_w = 1024, # max_bitfield_val(5, 0, 5) tile_max_w = 1024, # max_bitfield_val(5, 0, 5)
tile_max_h = max_bitfield_val(14, 8, 4), tile_max_h = max_bitfield_val(14, 8, 4),
num_vsc_pipes = 32) num_vsc_pipes = 32)
assert(num_sp_cores == num_ccu)
self.num_sp_cores = num_sp_cores # The # of SP cores seems to always match # of CCU
self.num_sp_cores = num_ccu
# 96 tile alignment seems correlated to 3 CCU self.num_ccu = num_ccu
if num_ccu == 3:
self.tile_align_w = 96
self.a6xx = Struct() self.a6xx = Struct()
self.a6xx.magic = Struct() self.a6xx.magic = Struct()
@ -273,8 +271,9 @@ add_gpus([
GPUId(619), GPUId(619),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen1, a6xx_gen1,
num_sp_cores = 1,
num_ccu = 1, num_ccu = 1,
tile_align_w = 32,
tile_align_h = 16,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 0, PC_POWER_CNTL = 0,
TPL1_DBG_ECO_CNTL = 0x00108000, TPL1_DBG_ECO_CNTL = 0x00108000,
@ -296,8 +295,9 @@ add_gpus([
GPUId(620), GPUId(620),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen1, a6xx_gen1,
num_sp_cores = 1,
num_ccu = 1, num_ccu = 1,
tile_align_w = 32,
tile_align_h = 16,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 0, PC_POWER_CNTL = 0,
TPL1_DBG_ECO_CNTL = 0x01008000, TPL1_DBG_ECO_CNTL = 0x01008000,
@ -319,8 +319,9 @@ add_gpus([
GPUId(630), GPUId(630),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen1, a6xx_gen1,
num_sp_cores = 2,
num_ccu = 2, num_ccu = 2,
tile_align_w = 32,
tile_align_h = 16,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 1, PC_POWER_CNTL = 1,
TPL1_DBG_ECO_CNTL = 0x00108000, TPL1_DBG_ECO_CNTL = 0x00108000,
@ -342,8 +343,9 @@ add_gpus([
GPUId(640), GPUId(640),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen2, a6xx_gen2,
num_sp_cores = 2,
num_ccu = 2, num_ccu = 2,
tile_align_w = 32,
tile_align_h = 16,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 1, PC_POWER_CNTL = 1,
TPL1_DBG_ECO_CNTL = 0x00008000, TPL1_DBG_ECO_CNTL = 0x00008000,
@ -365,8 +367,9 @@ add_gpus([
GPUId(680), GPUId(680),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen2, a6xx_gen2,
num_sp_cores = 4,
num_ccu = 4, num_ccu = 4,
tile_align_w = 64,
tile_align_h = 32,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 3, PC_POWER_CNTL = 3,
TPL1_DBG_ECO_CNTL = 0x00108000, TPL1_DBG_ECO_CNTL = 0x00108000,
@ -388,8 +391,9 @@ add_gpus([
GPUId(650), GPUId(650),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen3, a6xx_gen3,
num_sp_cores = 3,
num_ccu = 3, num_ccu = 3,
tile_align_w = 96,
tile_align_h = 48,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 2, PC_POWER_CNTL = 2,
# this seems to be a chicken bit that fixes cubic filtering: # this seems to be a chicken bit that fixes cubic filtering:
@ -416,8 +420,9 @@ add_gpus([
GPUId(chip_id=0xffff06030500, name="Adreno 7c+ Gen 3"), GPUId(chip_id=0xffff06030500, name="Adreno 7c+ Gen 3"),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen4, a6xx_gen4,
num_sp_cores = 2,
num_ccu = 2, num_ccu = 2,
tile_align_w = 32,
tile_align_h = 16,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 1, PC_POWER_CNTL = 1,
TPL1_DBG_ECO_CNTL = 0x05008000, TPL1_DBG_ECO_CNTL = 0x05008000,
@ -439,8 +444,9 @@ add_gpus([
GPUId(660), GPUId(660),
], A6xxGPUInfo( ], A6xxGPUInfo(
a6xx_gen4, a6xx_gen4,
num_sp_cores = 3,
num_ccu = 3, num_ccu = 3,
tile_align_w = 96,
tile_align_h = 16,
magic_regs = dict( magic_regs = dict(
PC_POWER_CNTL = 2, PC_POWER_CNTL = 2,
TPL1_DBG_ECO_CNTL = 0x05008000, TPL1_DBG_ECO_CNTL = 0x05008000,

2
src/gallium/drivers/freedreno/freedreno_gmem.c
View file

@ -549,7 +549,7 @@ gmem_key_init(struct fd_batch *batch, bool assume_zs, bool no_scis_opt)
*/ */
key->gmem_page_align = 8; key->gmem_page_align = 8;
} else if (is_a6xx(screen)) { } else if (is_a6xx(screen)) {
key->gmem_page_align = (screen->info->tile_align_w == 96) ? 3 : 1; key->gmem_page_align = screen->info->num_ccu;
} else { } else {
// TODO re-check this across gens.. maybe it should only // TODO re-check this across gens.. maybe it should only
// be a single page in some cases: // be a single page in some cases: