drm/amdgpu: fix use-after-unlock in eviction fence destroy

The eviction fence destroy path incorrectly calls dma_fence_put() on
evf_mgr->ev_fence after releasing the ev_fence_lock. This introduces a
potential use-after-unlock or race because another thread concurrently
modifies evf_mgr->ev_fence.

Fix this by grabbing a local reference to evf_mgr->ev_fence under the
lock and using that for dma_fence_put() after waiting.

Cc: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Sunil Khatri <sunil.khatri@amd.com>
Signed-off-by: Arvind Yadav <Arvind.Yadav@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Arvind Yadav 2025-05-15 12:44:46 +05:30 committed by Alex Deucher
parent cc473057bb
commit 13d0724f0f

View File

@ -183,7 +183,7 @@ void amdgpu_eviction_fence_destroy(struct amdgpu_eviction_fence_mgr *evf_mgr)
dma_fence_wait(&ev_fence->base, false);
/* Last unref of ev_fence */
dma_fence_put(&evf_mgr->ev_fence->base);
dma_fence_put(&ev_fence->base);
}
int amdgpu_eviction_fence_attach(struct amdgpu_eviction_fence_mgr *evf_mgr,