From: Christian König christian.koenig@amd.com
We shouldn't try to reserve and wait for a BO that isn't bound. Otherwise we can run into a deadlock if we have a fault during binding the BO.
Signed-off-by: Christian König christian.koenig@amd.com --- drivers/gpu/drm/radeon/radeon_mn.c | 3 +++ 1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/radeon/radeon_mn.c b/drivers/gpu/drm/radeon/radeon_mn.c index 535bf40..eef006c 100644 --- a/drivers/gpu/drm/radeon/radeon_mn.c +++ b/drivers/gpu/drm/radeon/radeon_mn.c @@ -142,6 +142,9 @@ static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
list_for_each_entry(bo, &node->bos, mn_list) {
+ if (!bo->tbo.ttm || bo->tbo.ttm->state != tt_bound) + continue; + r = radeon_bo_reserve(bo, true); if (r) { DRM_ERROR("(%ld) failed to reserve user bo\n", r);