btrfs: initialize the seq counter in struct btrfs_device
authorSebastian Andrzej Siewior <bigeasy@linutronix.de>
Fri, 15 Jan 2016 13:37:15 +0000 (14:37 +0100)
committerDavid Sterba <dsterba@suse.com>
Fri, 15 Jan 2016 18:28:43 +0000 (19:28 +0100)
I managed to trigger this:
| INFO: trying to register non-static key.
| the code is fine but needs lockdep annotation.
| turning off the locking correctness validator.
| CPU: 1 PID: 781 Comm: systemd-gpt-aut Not tainted 4.4.0-rt2+ #14
| Hardware name: ARM-Versatile Express
| [<80307cec>] (dump_stack)
| [<80070e98>] (__lock_acquire)
| [<8007184c>] (lock_acquire)
| [<80287800>] (btrfs_ioctl)
| [<8012a8d4>] (do_vfs_ioctl)
| [<8012ac14>] (SyS_ioctl)

so I think that btrfs_device_data_ordered_init() is not invoked behind
a macro somewhere.

Fixes: 7cc8e58d53cd ("Btrfs: fix unprotected device's variants on 32bits machine")
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
fs/btrfs/volumes.c

index 574a717fcb6ee347aa3fe895f70acb63294de37a..21fbcedb654258ce3c0185399ff7daa27b325c29 100644 (file)
@@ -232,6 +232,7 @@ static struct btrfs_device *__alloc_device(void)
        spin_lock_init(&dev->reada_lock);
        atomic_set(&dev->reada_in_flight, 0);
        atomic_set(&dev->dev_stats_ccnt, 0);
+       btrfs_device_data_ordered_init(dev);
        INIT_RADIX_TREE(&dev->reada_zones, GFP_NOFS & ~__GFP_DIRECT_RECLAIM);
        INIT_RADIX_TREE(&dev->reada_extents, GFP_NOFS & ~__GFP_DIRECT_RECLAIM);