[Intel-gfx] [PATCH 11/37] drm/i915: Only update LRU position of fence on success.

[Chris Wilson]

If we fail to set the domain when preparing to mmap a fenced object,
don't bother to more it to the front of the LRU fence list.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem.c |   15 +++++++--------
 1 files changed, 7 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index b93f7e1..892be92 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1046,20 +1046,19 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	if (read_domains & I915_GEM_DOMAIN_GTT) {
 		ret = i915_gem_object_set_to_gtt_domain(obj, write_domain != 0);
 
-		/* Update the LRU on the fence for the CPU access that's
-		 * about to occur.
-		 */
-		if (obj_priv->fence_reg != I915_FENCE_REG_NONE) {
-			list_move_tail(&obj_priv->fence_list,
-				       &dev_priv->mm.fence_list);
-		}
-
 		/* Silently promote "you're not bound, there was nothing to do"
 		 * to success, since the client was just asking us to
 		 * make sure everything was done.
 		 */
 		if (ret == -EINVAL)
 			ret = 0;
+
+		/* Update the LRU on the fence for the CPU access that's
+		 * about to occur.
+		 */
+		if (ret == 0 && obj_priv->fence_reg != I915_FENCE_REG_NONE)
+			list_move_tail(&obj_priv->fence_list,
+				       &dev_priv->mm.fence_list);
 	} else {
 		ret = i915_gem_object_set_to_cpu_domain(obj, write_domain != 0);
 	}
-- 
1.7.0