Re: 2.6.15-mm3 hangs during boot (raid related?)

[Jens Axboe <axboe@suse.de>]

On Thu, Jan 12 2006, Jurriaan on adsl-gate wrote:
> From: Neil Brown <neilb@suse.de>
> Date: Thu, Jan 12, 2006 at 05:36:42PM +1100
> Sent-To: 
> > On Thursday January 12, thunder7@xs4all.nl wrote:
> > > 
> > > 2.6.15-mm3 hangs during boot for me, after the lines
> > > 
> > > ========
> > > md4: bitmap initialized from disk: read 15/15 pages, set 51 bits, status: 0
> > > created bitmap (224 pages) for device md4
> > > ========
> > > 
> > > ctrl-alt-del to reboot works sometimes (2 out of 3). Below is complete
> > > dmesg (from 2.6.15-mm2, ver_linux output, .config and raid details).
> > 
> > Yep, this is probably a known problem with recent changes to the
> > 'barrier' code.
> > 
> > Try to convince md not to use barrier by changing md_super_write in
> > drivers/md/md.c.  Simply remove
> > 
> > 	if (!test_bit(BarriersNotsupp, &rdev->flags)) {
> > 		struct bio *rbio;
> > 		rw |= (1<<BIO_RW_BARRIER);
> > 		rbio = bio_clone(bio, GFP_NOIO);
> > 		rbio->bi_private = bio;
> > 		rbio->bi_end_io = super_written_barrier;
> > 		submit_bio(rw, rbio);
> > 	} else
> > 
> > leaving the
> > 		submit_bio(rw, rbio);
> > 
> > which comes after it.
> > 
> With this patch, it boots and works just fine.

Please reverse that patch and try this on top of -mm2/3 instead (it's
the patch from Tejun to fix the actual issue, from another similar
thread):

diff --git a/block/elevator.c b/block/elevator.c
index 1b5b5d9..f905e47 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -615,23 +615,23 @@ void elv_completed_request(request_queue
 	 * request is released from the driver, io must be done
 	 */
 	if (blk_account_rq(rq)) {
-		struct request *first_rq = list_entry_rq(q->queue_head.next);
-
 		q->in_flight--;
+		if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn)
+			e->ops->elevator_completed_req_fn(q, rq);
+	}
 
-		/*
-		 * Check if the queue is waiting for fs requests to be
-		 * drained for flush sequence.
-		 */
-		if (q->ordseq && q->in_flight == 0 &&
+	/*
+	 * Check if the queue is waiting for fs requests to be
+	 * drained for flush sequence.
+	 */
+	if (unlikely(q->ordseq)) {
+		struct request *first_rq = list_entry_rq(q->queue_head.next);
+		if (q->in_flight == 0 &&
 		    blk_ordered_cur_seq(q) == QUEUE_ORDSEQ_DRAIN &&
 		    blk_ordered_req_seq(first_rq) > QUEUE_ORDSEQ_DRAIN) {
 			blk_ordered_complete_seq(q, QUEUE_ORDSEQ_DRAIN, 0);
 			q->request_fn(q);
 		}
-
-		if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn)
-			e->ops->elevator_completed_req_fn(q, rq);
 	}
 }


-- 
Jens Axboe