AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

r300: do not limit maximum TEX group for R300/R400

Browse Source 
We do not have the TEX semaphore there anyway so the benefits are not as
high as with R500 and the chances of running out of TEX indirections are
just too high.

This will increase the register pressure in some shaders, but I believe
the gained shaders are worth it and there is also some cycles reduction
in some cases. I'm not sure how to optimize this further without
actually clonning the shader before the pair shceduling and than doing a
trial and error to see if there is some compromise where we can just hit
the indirection limit to not group it too much...

Shader-db RV410:
total instructions in shared programs: 112800 -> 112825 (0.02%)
instructions in affected programs: 5024 -> 5049 (0.50%)
helped: 23
HURT: 19
total temps in shared programs: 18170 -> 18244 (0.41%)
temps in affected programs: 1365 -> 1439 (5.42%)
helped: 39
HURT: 34
total cycles in shared programs: 169535 -> 166806 (-1.61%)
cycles in affected programs: 14229 -> 11500 (-19.18%)
helped: 84
HURT: 4

LOST:   0
GAINED: 8

GAINED: shaders/godot3.4/34-59.shader_test FS
GAINED: shaders/lightsmark/25.shader_test FS
GAINED: shaders/lightsmark/28.shader_test FS
GAINED: shaders/lightsmark/34.shader_test FS
GAINED: shaders/this-war-of-mine/144.shader_test FS
GAINED: shaders/this-war-of-mine/145.shader_test FS
GAINED: shaders/tropics/432.shader_test FS
GAINED: shaders/tropics/462.shader_test FS

Signed-off-by: Pavel Ondračka <pavel.ondracka@gmail.com>
Reviewed-by: Filip Gawin <None>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33275>
This commit is contained in:
Pavel Ondračka
2025-01-28 16:10:17 +01:00
committed by Marge Bot
parent b3bd769957
commit 9835a2df83
2 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/gallium/drivers/r300/ci/r300-rv410-fails.txt
-1
View File
@@ -1032,7 +1032,6 @@ spec@glsl-1.10@execution@fs-sign-times-neg,Fail
spec@glsl-1.10@execution@fs-sign-times-neg-abs,Fail
spec@glsl-1.10@execution@fs-sign-times-sign,Fail
spec@glsl-1.10@execution@glsl-1.10-built-in-matrix-state,Fail
spec@glsl-1.10@execution@glsl-fs-convolution-1,Fail
spec@glsl-1.10@execution@glsl-fs-convolution-2,Fail
spec@glsl-1.10@execution@glsl-fs-if-nested-loop,Fail
spec@glsl-1.10@execution@glsl-vs-if-nested-loop,Fail
src/gallium/drivers/r300/compiler/radeon_pair_schedule.c
+4 -1
View File
@@ -1315,7 +1315,10 @@ rc_pair_schedule(struct radeon_compiler *cc, void *user)
} else {
s.CalcScore = calc_score_r300;
}
s.max_tex_group = debug_get_num_option("RADEON_TEX_GROUP", 8);
/* max_tex_group is mostly R500 optimization, for R300-R400 we want to group as much
* as we can, otherwise we risk running out of TEX indirections.
*/
s.max_tex_group = debug_get_num_option("RADEON_TEX_GROUP", cc->is_r500 ? 8 : UINT_MAX);
/* First go over and count all TEX. */
while (inst != &c->Base.Program.Instructions) {