On Mon, Sep 21, 2026 at 4:11 AM Jesse Zhang <[email protected]> wrote:
>
> Use gfx_supported_reset's PER_PIPE bit to report GFX pipe reset support
> through a shared helper.
>
> Signed-off-by: Jesse Zhang <[email protected]>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 10 ++++++++++
> drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h | 1 +
> 2 files changed, 11 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> index 2da839228911..7c597dab5b2a 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> @@ -2376,6 +2376,16 @@ static void
> amdgpu_gfx_reset_stop_compute_scheds(struct amdgpu_device *adev,
> }
> }
>
> +bool amdgpu_gfx_me_pipe_reset_supported(struct amdgpu_device *adev)
> +{
> + if (!!(adev->gfx.gfx_supported_reset & AMDGPU_RESET_TYPE_PER_PIPE))
> + return true;
> + else
> + dev_warn_once(adev->dev, "Please use the latest MEC version
> to see whether support pipe reset\n");
Should be ME rather than MEC?
Alex
> +
> + return false;
> +}
> +
> /* Caller holds reset_mutex and attempted MES suspend before HQD lookup. */
> void amdgpu_gfx_pipe_reset_prepare(struct amdgpu_device *adev,
> struct amdgpu_ring *guilty_ring,
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
> index 855651f0c930..5939e1fb3b24 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
> @@ -707,6 +707,7 @@ struct amdgpu_gfx_pipe_reset_ctx {
> unsigned long replay_mask;
> };
>
> +bool amdgpu_gfx_me_pipe_reset_supported(struct amdgpu_device *adev);
> void amdgpu_gfx_pipe_reset_prepare(struct amdgpu_device *adev,
> struct amdgpu_ring *guilty_ring,
> u32 me, u32 pipe,
> --
> 2.49.0
>
