Browse Source

xen-blkfront: fix a deadlock while handling discard response

When we get -EOPNOTSUPP response for a discard request, we will clear
the discard flag on the request queue so we won't attempt to send discard
requests to backend again, and this should be protected under rq->queue_lock.
However, when we setup the request queue, we pass blkif_io_lock to
blk_init_queue so rq->queue_lock is blkif_io_lock indeed, and this lock
is already taken when we are in blkif_interrpt, so remove the
spin_lock/spin_unlock when we clear the discard flag or we will end up
with deadlock here

Signed-off-by: Li Dongyang <lidongyang@novell.com>
[v1: Updated description a bit and removed comment from source]
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Li Dongyang 14 years ago
parent
commit
69ef68cef9
1 changed files with 0 additions and 2 deletions
  1. 0 2
      drivers/block/xen-blkfront.c

+ 0 - 2
drivers/block/xen-blkfront.c

@@ -747,9 +747,7 @@ static irqreturn_t blkif_interrupt(int irq, void *dev_id)
 					   info->gd->disk_name);
 				error = -EOPNOTSUPP;
 				info->feature_discard = 0;
-				spin_lock(rq->queue_lock);
 				queue_flag_clear(QUEUE_FLAG_DISCARD, rq);
-				spin_unlock(rq->queue_lock);
 			}
 			__blk_end_request_all(req, error);
 			break;