[PATCH 2/2] drm/amdgpu/gfx9: put queue resets behind a debug option
Felix Kuehling
felix.kuehling at amd.com
Wed Aug 21 21:59:48 UTC 2024
On 2024-08-20 16:25, Alex Deucher wrote:
> Pending extended validation.
>
> Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c | 4 ++++
> drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 4 ++++
> drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c | 6 ++++++
> 3 files changed, 14 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
> index c63528a4e8941..1254a43ec96b6 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
> @@ -1151,6 +1151,10 @@ uint64_t kgd_gfx_v9_hqd_get_pq_addr(struct amdgpu_device *adev,
> uint32_t low, high;
> uint64_t queue_addr = 0;
>
> + if (!adev->debug_exp_resets &&
> + !adev->gfx.num_gfx_rings)
> + return 0;
> +
Did you put this in the HW-specific code path intentionally? If you want
this check to apply to all ASICs, you should put it into
detect_queue_hang in kfd_device_queue_manager.c. But maybe the extended
validation is HW-specific.
Either way, the patch is
Acked-by: Felix Kuehling <felix.kuehling at amd.com>
> kgd_gfx_v9_acquire_queue(adev, pipe_id, queue_id, inst);
> amdgpu_gfx_rlc_enter_safe_mode(adev, inst);
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
> index 21089aadbb7b4..8cf5d7925b51c 100644
> --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
> @@ -7233,6 +7233,10 @@ static int gfx_v9_0_reset_kcq(struct amdgpu_ring *ring,
> unsigned long flags;
> int i, r;
>
> + if (!adev->debug_exp_resets &&
> + !adev->gfx.num_gfx_rings)
> + return -EINVAL;
> +
> if (amdgpu_sriov_vf(adev))
> return -EINVAL;
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
> index 2067f26d3a9d8..f8649546b9c4c 100644
> --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
> +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
> @@ -3052,6 +3052,9 @@ static void gfx_v9_4_3_ring_soft_recovery(struct amdgpu_ring *ring,
> struct amdgpu_device *adev = ring->adev;
> uint32_t value = 0;
>
> + if (!adev->debug_exp_resets)
> + return;
> +
> value = REG_SET_FIELD(value, SQ_CMD, CMD, 0x03);
> value = REG_SET_FIELD(value, SQ_CMD, MODE, 0x01);
> value = REG_SET_FIELD(value, SQ_CMD, CHECK_VMID, 1);
> @@ -3475,6 +3478,9 @@ static int gfx_v9_4_3_reset_kcq(struct amdgpu_ring *ring,
> unsigned long flags;
> int r, i;
>
> + if (!adev->debug_exp_resets)
> + return -EINVAL;
> +
> if (amdgpu_sriov_vf(adev))
> return -EINVAL;
>
More information about the amd-gfx
mailing list