3.16.7-ckt26 -stable review patch.  If anyone has any objections, please let me know.

---8<------------------------------------------------------------

From: Luis Henriques <redacted>

This reverts commit 73187980dfefe5198aadcfdf0a377e461eed2bfa, which was
commit f6ff4f67cdf8455d0a4226eeeaf5af17c37d05eb upstream.

This patch was triggering a Oops in stable kernel 3.10.99.  Christian
agrees that the patch is correct but "assumes that radeon_fence_unref()
can safely take NULL as the fence which is not the case for older
kernels."

Reported-by: Erik Andersen <redacted>
Acked-by: Christian König <christian.koenig@amd.com>
Cc: Nicolai Hähnle <redacted>
Signed-off-by: Luis Henriques <redacted>
---
 drivers/gpu/drm/radeon/radeon_sa.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
index 15fd57296081..adcf3e2f07da 100644
--- a/drivers/gpu/drm/radeon/radeon_sa.c
+++ b/drivers/gpu/drm/radeon/radeon_sa.c

@@ -349,13 +349,8 @@ int radeon_sa_bo_new(struct radeon_device *rdev,
 			/* see if we can skip over some allocations */
 		} while (radeon_sa_bo_next_hole(sa_manager, fences, tries));
 
-		for (i = 0; i < RADEON_NUM_RINGS; ++i)
-			radeon_fence_ref(fences[i]);
-
 		spin_unlock(&sa_manager->wq.lock);
 		r = radeon_fence_wait_any(rdev, fences, false);
-		for (i = 0; i < RADEON_NUM_RINGS; ++i)
-			radeon_fence_unref(&fences[i]);
 		spin_lock(&sa_manager->wq.lock);
 		/* if we have nothing to wait for block */
 		if (r == -ENOENT) {