block: silently error unsupported empty barriers too
Mark McLoughlin [Sat, 24 Oct 2009 12:14:31 +0000 (14:14 +0200)]
With 2.6.32-rc5 in a KVM guest using dm and virtio_blk, we see the
following errors:

  end_request: I/O error, dev vda, sector 0
  end_request: I/O error, dev vda, sector 0

The errors go away if dm stops submitting empty barriers, by reverting:

  commit 52b1fd5a27c625c78373e024bf570af3c9d44a79
  Author: Mikulas Patocka <mpatocka@redhat.com>
    dm: send empty barriers to targets in dm_flush

We should silently error all barriers, even empty barriers, on devices
like virtio_blk which don't support them.

See also:

  https://bugzilla.redhat.com/514901

Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Acked-by: Alasdair G Kergon <agk@redhat.com>
Acked-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Neil Brown <neilb@suse.de>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

block/blk-core.c

index ac0fa10..71da511 100644 (file)
@@ -1161,7 +1161,7 @@ static int __make_request(struct request_queue *q, struct bio *bio)
        const unsigned int ff = bio->bi_rw & REQ_FAILFAST_MASK;
        int rw_flags;
 
-       if (bio_rw_flagged(bio, BIO_RW_BARRIER) && bio_has_data(bio) &&
+       if (bio_rw_flagged(bio, BIO_RW_BARRIER) &&
            (q->next_ordered == QUEUE_ORDERED_NONE)) {
                bio_endio(bio, -EOPNOTSUPP);
                return 0;