block: wake up all tasks blocked in get_request()

Once blk_set_queue_dying() is done in blk_cleanup_queue(), we call blk_freeze_queue() and wait for q->q_usage_counter becoming zero. But if there are tasks blocked in get_request(), q->q_usage_counter can never become zero. So we have to wake up all these tasks in blk_set_queue_dying() first.

---
 block/blk-core.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

index 7c54c195e79e..1038706edd87 100644
@@ -637,8 +637,8 @@ void blk_set_queue_dying(struct request_queue *q)
 		spin_lock_irq(q->queue_lock);
 		blk_queue_for_each_rl(rl, q) {
 			if (rl->rq_pool) {
-				wake_up(&rl->wait[BLK_RW_SYNC]);
-				wake_up(&rl->wait[BLK_RW_ASYNC]);
+				wake_up_all(&rl->wait[BLK_RW_SYNC]);
+				wake_up_all(&rl->wait[BLK_RW_ASYNC]);
 			}
 		}
 		spin_unlock_irq(q->queue_lock);

2.9.5

Looks correct to me, since we do exclusive waits on that waitqueue. Applied for this series.

Note that the format here is ("patch title")

Jens Axboe