[PATCH] drm/amdgpu: fix mqd corruption for gfx12

Alex Deucher alexdeucher at gmail.com
Tue May 7 16:57:46 UTC 2024


On Tue, May 7, 2024 at 12:12 PM Min, Frank <Frank.Min at amd.com> wrote:
>
> [AMD Official Use Only - General]
>
> From: Frank Min <Frank.Min at amd.com>
>
> 1. restore mqd from backup while resuming
>
> 2. use copy_toio and copy_fromio while mqd in vram
>
> Signed-off-by: Frank Min <Frank.Min at amd.com>

Reviewed-by: Alex Deucher <alexander.deucher at amd.com>

> ---
>  drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c | 21 ++++++++-------------
>  1 file changed, 8 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
> index 0bb119be2284..974219386e47 100644
> --- a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
> @@ -2720,17 +2720,15 @@ static int gfx_v12_0_gfx_init_queue(struct amdgpu_ring *ring)
>                 soc24_grbm_select(adev, 0, 0, 0, 0);
>                 mutex_unlock(&adev->srbm_mutex);
>                 if (adev->gfx.me.mqd_backup[mqd_idx])
> -                       memcpy(adev->gfx.me.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
> -       } else if (amdgpu_in_reset(adev)) {
> -               /* reset mqd with the backup copy */
> +                       memcpy_fromio(adev->gfx.me.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
> +       } else {
> +               /* restore mqd with the backup copy */
>                 if (adev->gfx.me.mqd_backup[mqd_idx])
> -                       memcpy(mqd, adev->gfx.me.mqd_backup[mqd_idx], sizeof(*mqd));
> +                       memcpy_toio(mqd, adev->gfx.me.mqd_backup[mqd_idx], sizeof(*mqd));
>                 /* reset the ring */
>                 ring->wptr = 0;
>                 *ring->wptr_cpu_addr = 0;
>                 amdgpu_ring_clear_ring(ring);
> -       } else {
> -               amdgpu_ring_clear_ring(ring);
>         }
>
>         return 0;
> @@ -3091,18 +3089,15 @@ static int gfx_v12_0_kcq_init_queue(struct amdgpu_ring *ring)
>                 mutex_unlock(&adev->srbm_mutex);
>
>                 if (adev->gfx.mec.mqd_backup[mqd_idx])
> -                       memcpy(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
> -       } else if (amdgpu_in_reset(adev)) { /* for GPU_RESET case */
> -               /* reset MQD to a clean status */
> +                       memcpy_fromio(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(*mqd));
> +       } else {
> +               /* restore MQD to a clean status */
>                 if (adev->gfx.mec.mqd_backup[mqd_idx])
> -                       memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(*mqd));
> -
> +                       memcpy_toio(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(*mqd));
>                 /* reset ring buffer */
>                 ring->wptr = 0;
>                 atomic64_set((atomic64_t *)ring->wptr_cpu_addr, 0);
>                 amdgpu_ring_clear_ring(ring);
> -       } else {
> -               amdgpu_ring_clear_ring(ring);
>         }
>
>         return 0;
> --
> 2.34.1
>


More information about the amd-gfx mailing list