diff options
| author | YuBiao Wang <YuBiao.Wang@amd.com> | 2023-03-16 06:30:32 +0300 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2023-03-23 16:31:30 +0300 |
| commit | 033c56474acf567a450f8bafca50e0b610f2b716 (patch) | |
| tree | 066f7492c49b71aaebf13c654f99942be0182382 /drivers/gpu | |
| parent | 4eb0b49a0ad3e004a6a65b84efe37bc7e66d560f (diff) | |
| download | linux-033c56474acf567a450f8bafca50e0b610f2b716.tar.xz | |
drm/amdgpu: Force signal hw_fences that are embedded in non-sched jobs
[Why]
For engines not supporting soft reset, i.e. VCN, there will be a failed
ib test before mode 1 reset during asic reset. The fences in this case
are never signaled and next time when we try to free the sa_bo, kernel
will hang.
[How]
During pre_asic_reset, driver will clear job fences and afterwards the
fences' refcount will be reduced to 1. For drm_sched_jobs it will be
released in job_free_cb, and for non-sched jobs like ib_test, it's meant
to be released in sa_bo_free but only when the fences are signaled. So
we have to force signal the non_sched bad job's fence during
pre_asic_reset or the clear is not complete.
Signed-off-by: YuBiao Wang <YuBiao.Wang@amd.com>
Acked-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Diffstat (limited to 'drivers/gpu')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 9 |
1 files changed, 9 insertions, 0 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index faff4a3f96e6..f52d0ba91a77 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -678,6 +678,15 @@ void amdgpu_fence_driver_clear_job_fences(struct amdgpu_ring *ring) ptr = &ring->fence_drv.fences[i]; old = rcu_dereference_protected(*ptr, 1); if (old && old->ops == &amdgpu_job_fence_ops) { + struct amdgpu_job *job; + + /* For non-scheduler bad job, i.e. failed ib test, we need to signal + * it right here or we won't be able to track them in fence_drv + * and they will remain unsignaled during sa_bo free. + */ + job = container_of(old, struct amdgpu_job, hw_fence); + if (!job->base.s_fence && !dma_fence_is_signaled(old)) + dma_fence_signal(old); RCU_INIT_POINTER(*ptr, NULL); dma_fence_put(old); } |