From: Marek Olšák
This can result in 2x increase in performance on non-harvested Kaveris.
v2: don't do it on radeon
Tested-by: Michel Dänzer
---
src/amd/common/ac_gpu_info.c | 15 +--
1 file changed, 13 insertions(+), 2 deletions(-)
diff --git a/src/amd/common/ac_gpu_info.c b/src
On 2018-05-02 09:11 PM, Marek Olšák wrote:
> If you are shader-bound, you won't see any difference. In order to become
> RB-bound, you need more work in RBs than shaders. Blending or MSAA might
> help add that.
Indeed,
vblank_mode=0 glxgears -samples 8
shows a significant difference in performa
The changed field in PA_SC_RASTER_CONFIG is pretty clear here:
0 - RB0 renders all pixels (default for single rb per pkr configs)
1 - RB0 renders rb_tile_id==1, RB1 renders rb_tile_id==0
2 - RB0 renders rb_tile_id==0, RB1 renders rb_tile_id==1
3 - RB1 renders all pixels
So you see that values 0 a
If you are shader-bound, you won't see any difference. In order to become
RB-bound, you need more work in RBs than shaders. Blending or MSAA might
help add that.
Marek
On Wed, May 2, 2018 at 5:31 AM, Michel Dänzer wrote:
> On 2018-05-02 06:00 AM, Marek Olšák wrote:
> > From: Marek Olšák
> >
>
On 2018-05-02 06:00 AM, Marek Olšák wrote:
> From: Marek Olšák
>
> This can result in 2x increase in performance on non-harvested Kaveris.
FWIW, no difference with glxgears on my (non-harvested AFAIK) Kaveris,
neither with amdgpu nor radeon. Should I try something else?
Tested-by: Michel Dänzer
From: Marek Olšák
This can result in 2x increase in performance on non-harvested Kaveris.
---
src/amd/common/ac_gpu_info.c | 8 ++--
src/gallium/winsys/radeon/drm/radeon_drm_winsys.c | 4
2 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/src/amd/common
