diff options
| author | Samuel Pitoiset <[email protected]> | 2017-11-29 14:48:32 +0100 |
|---|---|---|
| committer | Samuel Pitoiset <[email protected]> | 2017-11-30 21:38:30 +0100 |
| commit | ff0f17da1446e7aa965e06c04a6ad5a55d95463d (patch) | |
| tree | 9f55df269d62229335b9a5353e982006b0aae192 /src/amd | |
| parent | f5955c6bf869e323373c59c2687b463b7e16bdf4 (diff) | |
radv: do not allocate CMASK or DCC for small surfaces
The idea is ported from RadeonSI, but using 512x512 instead of
256x256 seems slightly better. This improves dota2 performance
by +2%.
Signed-off-by: Samuel Pitoiset <[email protected]>
Reviewed-by: Dave Airlie <[email protected]>
Tested-by: Dieter Nützel <[email protected]>
Diffstat (limited to 'src/amd')
| -rw-r--r-- | src/amd/vulkan/radv_image.c | 10 |
1 files changed, 10 insertions, 0 deletions
diff --git a/src/amd/vulkan/radv_image.c b/src/amd/vulkan/radv_image.c index b145e81f826..ee03a1dc6a9 100644 --- a/src/amd/vulkan/radv_image.c +++ b/src/amd/vulkan/radv_image.c @@ -806,6 +806,16 @@ radv_image_alloc_htile(struct radv_image *image) static inline bool radv_image_can_enable_dcc_or_cmask(struct radv_image *image) { + if (image->info.samples <= 1 && + image->info.width <= 512 && image->info.height <= 512) { + /* Do not enable CMASK or DCC for small surfaces where the cost + * of the eliminate pass can be higher than the benefit of fast + * clear. RadeonSI does this, but the image threshold is + * different. + */ + return false; + } + return image->usage & VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT && (image->exclusive || image->queue_family_mask == 1); } |