]> err.no Git - linux-2.6/commitdiff
[PATCH] device-mapper dm-mpath: endio spinlock fix
authorStefan Bader <Stefan.Bader@de.ibm.com>
Tue, 22 Nov 2005 05:32:35 +0000 (21:32 -0800)
committerLinus Torvalds <torvalds@g5.osdl.org>
Tue, 22 Nov 2005 17:14:31 +0000 (09:14 -0800)
do_end_io() can be called without interrupts blocked.

Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
drivers/md/dm-mpath.c

index f9b7b32d5d5c773bea00dfc9e2f198333ea550c1..f72a82fb943489872b3d91abdffe4cb292c92267 100644 (file)
@@ -1000,6 +1000,7 @@ static int do_end_io(struct multipath *m, struct bio *bio,
 {
        struct hw_handler *hwh = &m->hw_handler;
        unsigned err_flags = MP_FAIL_PATH;      /* Default behavior */
+       unsigned long flags;
 
        if (!error)
                return 0;       /* I/O complete */
@@ -1010,17 +1011,17 @@ static int do_end_io(struct multipath *m, struct bio *bio,
        if (error == -EOPNOTSUPP)
                return error;
 
-       spin_lock(&m->lock);
+       spin_lock_irqsave(&m->lock, flags);
        if (!m->nr_valid_paths) {
                if (!m->queue_if_no_path) {
-                       spin_unlock(&m->lock);
+                       spin_unlock_irqrestore(&m->lock, flags);
                        return -EIO;
                } else {
-                       spin_unlock(&m->lock);
+                       spin_unlock_irqrestore(&m->lock, flags);
                        goto requeue;
                }
        }
-       spin_unlock(&m->lock);
+       spin_unlock_irqrestore(&m->lock, flags);
 
        if (hwh->type && hwh->type->error)
                err_flags = hwh->type->error(hwh, bio);
@@ -1040,12 +1041,12 @@ static int do_end_io(struct multipath *m, struct bio *bio,
        dm_bio_restore(&mpio->details, bio);
 
        /* queue for the daemon to resubmit or fail */
-       spin_lock(&m->lock);
+       spin_lock_irqsave(&m->lock, flags);
        bio_list_add(&m->queued_ios, bio);
        m->queue_size++;
        if (!m->queue_io)
                queue_work(kmultipathd, &m->process_queued_ios);
-       spin_unlock(&m->lock);
+       spin_unlock_irqrestore(&m->lock, flags);
 
        return 1;       /* io not complete */
 }