Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 93803e01 authored by Jens Axboe's avatar Jens Axboe
Browse files

cfq-iosched: fix race in cfq_set_request() 



We need to hold the queue lock over the reference increment,
it's not atomic anymore.

Signed-off-by: default avatarJens Axboe <jaxboe@fusionio.com>
parent da527770

block/cfq-iosched.c

+1 −2
Original line number Original line Diff line number Diff line
@@ -3675,12 +3675,11 @@ new_queue: 




	cfqq->allocated[rw]++;

	cfqq->allocated[rw]++;





	spin_unlock_irqrestore(q->queue_lock, flags);



	cfqq->ref++;

	cfqq->ref++;

	rq->elevator_private[0] = cic;

	rq->elevator_private[0] = cic;

	rq->elevator_private[1] = cfqq;

	rq->elevator_private[1] = cfqq;

	rq->elevator_private[2] = cfq_ref_get_cfqg(cfqq->cfqg);

	rq->elevator_private[2] = cfq_ref_get_cfqg(cfqq->cfqg);

	spin_unlock_irqrestore(q->queue_lock, flags);

	return 0;

	return 0;





queue_fail:

queue_fail: