AcoSail/kernel_optimize_test
8
0
Fork 0
forked from luck/tmp_suning_uos_patched
Code Pull Requests Activity

drm/i915: fix scheduling while holding the new active list spinlock

Browse Source 
regression caused by commit 5e118f4139feafe97e913df67b1f7c1e5083e535:
i915_gem_object_move_to_inactive() should be called in task context,
as it calls fput();

Signed-off-by: Shaohua Li<shaohua.li@intel.com>
[anholt: Add more detail to the comment about the lock break that's added]
Signed-off-by: Eric Anholt <eric@anholt.net>
This commit is contained in:
Shaohua Li 2009-04-08 10:58:23 +08:00 committed by Eric Anholt
parent 280b713b5b
commit 68c8434217
1 changed files with 12 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
drivers/gpu/drm/i915/i915_gem.c
View File

@ -1596,8 +1596,19 @@ i915_gem_retire_request(struct drm_device *dev,
if (obj->write_domain != 0)
i915_gem_object_move_to_flushing(obj);
else
else {
/* Take a reference on the object so it won't be
* freed while the spinlock is held. The list
* protection for this spinlock is safe when breaking
* the lock like this since the next thing we do
* is just get the head of the list again.
*/
drm_gem_object_reference(obj);
i915_gem_object_move_to_inactive(obj);
spin_unlock(&dev_priv->mm.active_list_lock);
drm_gem_object_unreference(obj);
spin_lock(&dev_priv->mm.active_list_lock);
}
}
out:
spin_unlock(&dev_priv->mm.active_list_lock);