[AMD Official Use Only] Ray, That ROCR issue is caused by CGCG(NOT CGPG). With this patch applied, the rocr issue still exists. -- Best Regards Aaron Liu > -----Original Message----- > From: Huang, Ray <Ray.Huang@xxxxxxx> > Sent: Tuesday, November 2, 2021 5:19 PM > To: Liu, Aaron <Aaron.Liu@xxxxxxx> > Cc: amd-gfx@xxxxxxxxxxxxxxxxxxxxx; Deucher, Alexander > <Alexander.Deucher@xxxxxxx> > Subject: Re: [PATCH] drm/amdgpu: update RLC_PG_DELAY_3 Value to 200us > for yellow carp > > On Tue, Nov 02, 2021 at 04:51:18PM +0800, Liu, Aaron wrote: > > For yellow carp, the desired CGPG hysteresis value is 0x4E20. > > > > Signed-off-by: Aaron Liu <aaron.liu@xxxxxxx> > > --- > > drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 5 +---- > > 1 file changed, 1 insertion(+), 4 deletions(-) > > > > diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c > b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c > > index 90a834dc4008..b53b36f5ae92 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c > > +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c > > @@ -8316,11 +8316,8 @@ static void gfx_v10_cntl_power_gating(struct > amdgpu_device *adev, bool enable) > > if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) { > > switch (adev->ip_versions[GC_HWIP][0]) { > > case IP_VERSION(10, 3, 1): > > - data = 0x4E20 & > RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK_Vangogh; > > - WREG32_SOC15(GC, 0, mmRLC_PG_DELAY_3, data); > > - break; > > case IP_VERSION(10, 3, 3): > > - data = 0x1388 & > RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK_Vangogh; > > + data = 0x4E20 & > RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK_Vangogh; > > Acked-by: Huang Rui <ray.huang@xxxxxxx> > > Is this patch able to fix the cgpg issue in ROCr test? > > Thanks, > Ray