drm/amdgpu: fix tear down order in amdgpu_vm_pt_free

When freeing PD/PT with shadows it can happen that the shadow
destruction races with detaching the PD/PT from the VM causing a NULL
pointer dereference in the invalidation code.

Fix this by detaching the the PD/PT from the VM first and then
freeing the shadow instead.

Signed-off-by: Christian König <[email protected]>
Fixes: https://gitlab.freedesktop.org/drm/amd/-/issues/2867
Cc: <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c
index a2287bb252235e8eb9a906a9216bbf568a68fd33..a160265ddc07c141ecde4c6bc973655c1a812563 100644 (file)
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c
@@ -642,13 +642,14 @@ static void amdgpu_vm_pt_free(struct amdgpu_vm_bo_base *entry)
 
        if (!entry->bo)
                return;
+
+       entry->bo->vm_bo = NULL;
        shadow = amdgpu_bo_shadowed(entry->bo);
        if (shadow) {
                ttm_bo_set_bulk_move(&shadow->tbo, NULL);
                amdgpu_bo_unref(&shadow);
        }
        ttm_bo_set_bulk_move(&entry->bo->tbo, NULL);
-       entry->bo->vm_bo = NULL;
 
        spin_lock(&entry->vm->status_lock);
        list_del(&entry->vm_status);