lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

drm/ttm: Improved fencing of buffer object lists

Browse Source 
Drastically reduce the number of spin lock / unlock operations by performing
unreserving and fencing under global locks.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jerome Glisse <j.glisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This commit is contained in:
Thomas Hellstrom
2010-11-17 12:28:30 +00:00
committed by Dave Airlie
parent 702adba224
commit 95762c2b34
4 changed files with 45 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
drivers/gpu/drm/ttm/ttm_bo.c
View File

@@ -299,14 +299,19 @@ int ttm_bo_reserve(struct ttm_buffer_object *bo,
return ret;
}
void ttm_bo_unreserve_locked(struct ttm_buffer_object *bo)
{
ttm_bo_add_to_lru(bo);
atomic_set(&bo->reserved, 0);
wake_up_all(&bo->event_queue);
}
void ttm_bo_unreserve(struct ttm_buffer_object *bo)
{
struct ttm_bo_global *glob = bo->glob;
spin_lock(&glob->lru_lock);
ttm_bo_add_to_lru(bo);
atomic_set(&bo->reserved, 0);
wake_up_all(&bo->event_queue);
ttm_bo_unreserve_locked(bo);
spin_unlock(&glob->lru_lock);
}
EXPORT_SYMBOL(ttm_bo_unreserve);