OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

radv: Slightly refactor the determination of max_ps_params.

Browse Source 
It now uses has_dedicated_vram and gfx_level to detect GFX10.3+
discrete GPUs, which should also include GFX11 now.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27751>
This commit is contained in:
Timur Kristóf
2024-02-22 16:05:13 +01:00
committed by Marge Bot
parent dbfb96f08f
commit 4464e6baff
1 changed files with 3 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
src/amd/vulkan/radv_shader.c
View File

@@ -740,14 +740,10 @@ radv_consider_culling(const struct radv_physical_device *pdev, struct nir_shader
/* Shader based culling efficiency can depend on PS throughput.
* Estimate an upper limit for PS input param count based on GPU info.
*/
unsigned max_ps_params;
unsigned max_render_backends = pdev->info.max_render_backends;
unsigned max_se = pdev->info.max_se;
unsigned max_ps_params = 4;
if (max_render_backends / max_se == 4)
max_ps_params = 6; /* Navi21 and other GFX10.3 dGPUs. */
else
max_ps_params = 4; /* Navi 1x. */
if (pdev->info.gfx_level >= GFX10_3 && pdev->info.has_dedicated_vram)
max_ps_params = 6; /* GFX10.3 and newer discrete GPUs. */
/* TODO: consider other heuristics here, such as PS execution time */
if (util_bitcount64(ps_inputs_read & ~VARYING_BIT_POS) > max_ps_params)