fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-07 09:18:04 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

anv: Set recommended values for gfx20 async compute registers in STATE_COMPUTE_MODE

Browse source 
This recommended values should improve the performance of async
compute in gfx20, we may want to tweek this for Linux but at least
this values should give us a better baseline than default values.

Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30796>
This commit is contained in:
José Roberto de Souza 2024-08-22 09:32:04 -07:00 • committed by Marge Bot
parent 2483f8f7cd
commit aa5b2c4165
1 changed files with 8 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
src/intel/vulkan/genX_init_state.c
View file

@ -751,7 +751,14 @@ init_compute_queue_state(struct anv_queue *queue)
}
anv_batch_emit(batch, GENX(STATE_COMPUTE_MODE), cm) {
#if GFX_VER < 20
#if GFX_VER >= 20
cm.AsyncComputeThreadLimit = ACTL_Max8;
cm.ZPassAsyncComputeThreadLimit = ZPACTL_Max60;
cm.ZAsyncThrottlesettings = ZATS_DefertoAsyncComputeThreadLimit;
cm.AsyncComputeThreadLimitMask = 0x7;
cm.ZPassAsyncComputeThreadLimitMask = 0x7;
cm.ZAsyncThrottlesettingsMask = 0x3;
#else
cm.PixelAsyncComputeThreadLimit = PACTL_Max24;
cm.ZPassAsyncComputeThreadLimit = ZPACTL_Max60;
cm.PixelAsyncComputeThreadLimitMask = 0x7;