From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Thu, 11 Jul 2013 21:11:32 +0000 (-0700)
Subject: Fix zfsctl_expire_snapshot() deadlock
X-Git-Tag: zfs-0.6.2~37
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=76351672c222f28ea1b681097a9eff58a6791555;p=zfs

Fix zfsctl_expire_snapshot() deadlock

It is possible for an automounted snapshot which is expiring to
deadlock with a manual unmount of the snapshot.  This can occur
because taskq_cancel_id() will block if the task is currently
executing until it completes.  But it will never complete because
zfsctl_unmount_snapshot() is holding the zsb->z_ctldir_lock which
zfsctl_expire_snapshot() must acquire.

---------------------- z_unmount/0:2153 ---------------------
  mutex_lock                <blocking on zsb->z_ctldir_lock>
  zfsctl_unmount_snapshot
  zfsctl_expire_snapshot
  taskq_thread

------------------------- zfs:10690 -------------------------
  taskq_wait_id             <waiting for z_unmount to exit>
  taskq_cancel_id
  __zfsctl_unmount_snapshot
  zfsctl_unmount_snapshot   <takes zsb->z_ctldir_lock>
  zfs_unmount_snap
  zfs_ioc_destroy_snaps_nvl
  zfsdev_ioctl
  do_vfs_ioctl

We resolve the deadlock by dropping the zsb->z_ctldir_lock before
calling __zfsctl_unmount_snapshot().  The lock is only there to
prevent concurrent modification to the zsb->z_ctldir_snaps AVL
tree.  Moreover, we're careful to remove the zfs_snapentry_t from
the AVL tree before dropping the lock which ensures no other tasks
can find it.  On failure it's added back to the tree.

Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chris Dunlap <cdunlap@llnl.gov>
Closes #1527
---

diff --git a/module/zfs/zfs_ctldir.c b/module/zfs/zfs_ctldir.c
index 4fa530b6c..168f853f1 100644
--- a/module/zfs/zfs_ctldir.c
+++ b/module/zfs/zfs_ctldir.c
@@ -732,7 +732,11 @@ zfsctl_unmount_snapshot(zfs_sb_t *zsb, char *name, int flags)
 	sep = avl_find(&zsb->z_ctldir_snaps, &search, NULL);
 	if (sep) {
 		avl_remove(&zsb->z_ctldir_snaps, sep);
+		mutex_exit(&zsb->z_ctldir_lock);
+
 		error = __zfsctl_unmount_snapshot(sep, flags);
+
+		mutex_enter(&zsb->z_ctldir_lock);
 		if (error == EBUSY)
 			avl_add(&zsb->z_ctldir_snaps, sep);
 		else
@@ -767,7 +771,11 @@ zfsctl_unmount_snapshots(zfs_sb_t *zsb, int flags, int *count)
 	while (sep != NULL) {
 		next = AVL_NEXT(&zsb->z_ctldir_snaps, sep);
 		avl_remove(&zsb->z_ctldir_snaps, sep);
+		mutex_exit(&zsb->z_ctldir_lock);
+
 		error = __zfsctl_unmount_snapshot(sep, flags);
+
+		mutex_enter(&zsb->z_ctldir_lock);
 		if (error == EBUSY) {
 			avl_add(&zsb->z_ctldir_snaps, sep);
 			(*count)++;