dma-buf: Discard old fence_excl on retrying get_fences_rcu for realloc

If we have to drop the seqcount & rcu lock to perform a krealloc, we
have to restart the loop. In doing so, be careful not to lose track of
the already acquired exclusive fence.

Fixes: fedf54132d ("dma-buf: Restart reservation_object_get_fences_rcu() after writes")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: stable@vger.kernel.org #v4.10
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190604125323.21396-1-chris@chris-wilson.co.uk
This commit is contained in:
Chris Wilson 2019-06-04 13:53:23 +01:00
parent cbb3207914
commit f5b07b04e5

View File

@ -365,6 +365,10 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj,
GFP_NOWAIT | __GFP_NOWARN);
if (!nshared) {
rcu_read_unlock();
dma_fence_put(fence_excl);
fence_excl = NULL;
nshared = krealloc(shared, sz, GFP_KERNEL);
if (nshared) {
shared = nshared;