[PATCH 3.8 097/124] drm/mgag200: do not attempt to acquire a reservation while in an interrupt handler

[Kamal Mostafa]

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

------------------

From: Maarten Lankhorst <m.b.lankhorst at gmail.com>

commit a06b9a74c73750835b8fd69fe0d0bd7877da111b upstream.

Mutexes should not be acquired in interrupt context. While the trylock
fastpath is arguably safe on all implementations, the slowpath
unlock path definitely isn't.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst at canonical.com>
Signed-off-by: Dave Airlie <airlied at redhat.com>
Signed-off-by: Kamal Mostafa <kamal at canonical.com>
---
 drivers/gpu/drm/mgag200/mgag200_fb.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/mgag200/mgag200_fb.c b/drivers/gpu/drm/mgag200/mgag200_fb.c
index 41eefc4..0cbd607 100644
--- a/drivers/gpu/drm/mgag200/mgag200_fb.c
+++ b/drivers/gpu/drm/mgag200/mgag200_fb.c
@@ -26,7 +26,7 @@ static void mga_dirty_update(struct mga_fbdev *mfbdev,
 	struct mgag200_bo *bo;
 	int src_offset, dst_offset;
 	int bpp = (mfbdev->mfb.base.bits_per_pixel + 7)/8;
-	int ret;
+	int ret = -EBUSY;
 	bool unmap = false;
 	bool store_for_later = false;
 	int x2, y2;
@@ -40,7 +40,8 @@ static void mga_dirty_update(struct mga_fbdev *mfbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	ret = mgag200_bo_reserve(bo, true);
+	if (!in_interrupt())
+		ret = mgag200_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
 			return;
-- 
1.8.3.2