On Tuesday, July 7, 2026 3:44:28 PM Central European Summer Time Alex Deucher 
wrote:
> On Tue, Jul 7, 2026 at 4:06 AM Timur Kristóf <[email protected]> 
wrote:
> > On 2026. június 26., péntek 22:40:58 közép-európai nyári idő Alex Deucher
> > 
> > wrote:
> > > This is needed to allocate occlusion counters across
> > > both gfx pipes.
> > > 
> > > Fixes: b7a1a0ef12b8 ("drm/amd/amdgpu: add pipe1 hardware support")
> > > Signed-off-by: Alex Deucher <[email protected]>
> > > ---
> > > 
> > >  drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 3 +++
> > >  1 file changed, 3 insertions(+)
> > > 
> > > diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
> > > b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index
> > > d72ecf5dab09e..6ff7a8a700939
> > > 100644
> > > --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
> > > +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
> > > @@ -5352,6 +5352,9 @@ static void gfx_v10_0_constants_init(struct
> > > amdgpu_device *adev) gfx_v10_0_get_tcc_info(adev);
> > > 
> > >       adev->gfx.config.pa_sc_tile_steering_override =
> > >       
> > >               gfx_v10_0_init_pa_sc_tile_steering_override(adev);
> > > 
> > > +     /* program DB_RING_CONTROL for multiple GFX pipes */
> > > +     WREG32_FIELD15(GC, 0, DB_RING_CONTROL, COUNTER_CONTROL,
> > > +                    (adev->gfx.me.num_pipe_per_me > 1) ? 0 : 1);
> > 
> > Hi Alex,
> > 
> > Why do you set this to zero when the number of pipes is more than 1?
> > Wouldn't it need to be the other way around and set to the number of pipes
> > (or number of rings)?
> 
> The hardware default is 1.  The other settings are as follows for this
> field: 0 - split occlusion counters between gfx pipes
> 1 - all occlusion counters to pipe 0
> 2 - all occlusion counters to pipe 1
> 

I see, thanks!

Could you please mention that in a comment or maybe add an enum for the 
possible values of that field?

With that, the series is:
Reviewed-by: Timur Kristóf <[email protected]>


> > 
> > >       /* XXX SH_MEM regs */
> > >       /* where to put LDS, scratch, GPUVM in FSA64 space */




Reply via email to