[PATCH 3.16 316/366] fix __legitimize_mnt()/mntput() race

[Ben Hutchings]

3.16.61-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Al Viro <viro@xxxxxxxxxxxxxxxxxx>

commit 119e1ef80ecfe0d1deb6378d4ab41f5b71519de1 upstream.

__legitimize_mnt() has two problems - one is that in case of success
the check of mount_lock is not ordered wrt preceding increment of
refcount, making it possible to have successful __legitimize_mnt()
on one CPU just before the otherwise final mntpu() on another,
with __legitimize_mnt() not seeing mntput() taking the lock and
mntput() not seeing the increment done by __legitimize_mnt().
Solved by a pair of barriers.

Another is that failure of __legitimize_mnt() on the second
read_seqretry() leaves us with reference that'll need to be
dropped by caller; however, if that races with final mntput()
we can end up with caller dropping rcu_read_lock() and doing
mntput() to release that reference - with the first mntput()
having freed the damn thing just as rcu_read_lock() had been
dropped.  Solution: in "do mntput() yourself" failure case
grab mount_lock, check if MNT_DOOMED has been set by racing
final mntput() that has missed our increment and if it has -
undo the increment and treat that as "failure, caller doesn't
need to drop anything" case.

It's not easy to hit - the final mntput() has to come right
after the first read_seqretry() in __legitimize_mnt() *and*
manage to miss the increment done by __legitimize_mnt() before
the second read_seqretry() in there.  The things that are almost
impossible to hit on bare hardware are not impossible on SMP
KVM, though...

Reported-by: Oleg Nesterov <oleg@xxxxxxxxxx>
Fixes: 48a066e72d97 ("RCU'd vsfmounts")
Signed-off-by: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
[bwh: Backported to 3.16: __legitimize_mnt() has not been split out
 from legitimize_mnt().  Adjust the added return statement and
 comments accordingly.]
Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
---
 fs/namespace.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -592,12 +592,20 @@ bool legitimize_mnt(struct vfsmount *bas
 		return true;
 	mnt = real_mount(bastard);
 	mnt_add_count(mnt, 1);
+	smp_mb();			// see mntput_no_expire()
 	if (likely(!read_seqretry(&mount_lock, seq)))
 		return true;
 	if (bastard->mnt_flags & MNT_SYNC_UMOUNT) {
 		mnt_add_count(mnt, -1);
 		return false;
 	}
+	lock_mount_hash();
+	if (unlikely(bastard->mnt_flags & MNT_DOOMED)) {
+		mnt_add_count(mnt, -1);
+		unlock_mount_hash();
+		return false;
+	}
+	unlock_mount_hash();
 	rcu_read_unlock();
 	mntput(bastard);
 	rcu_read_lock();
@@ -984,6 +992,11 @@ put_again:
 		return;
 	}
 	lock_mount_hash();
+	/*
+	 * make sure that if legitimize_mnt() has not seen us grab
+	 * mount_lock, we'll see their refcount increment here.
+	 */
+	smp_mb();
 	mnt_add_count(mnt, -1);
 	if (mnt_get_count(mnt)) {
 		rcu_read_unlock();