summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMiklos Szeredi <miklos@szeredi.hu>2012-09-17 22:23:30 +0200
committerLinus Torvalds <torvalds@linux-foundation.org>2012-09-29 17:41:40 -0700
commit8110e16d42d587997bcaee0c864179e6d93603fe (patch)
tree210b83419cc6464b1a1faff58c47d8316f6e5939
parent6a3e3dbee62a8fcf13c349b141b3cedf35a65c41 (diff)
vfs: dcache: fix deadlock in tree traversal
IBM reported a deadlock in select_parent(). This was found to be caused by taking rename_lock when already locked when restarting the tree traversal. There are two cases when the traversal needs to be restarted: 1) concurrent d_move(); this can only happen when not already locked, since taking rename_lock protects against concurrent d_move(). 2) racing with final d_put() on child just at the moment of ascending to parent; rename_lock doesn't protect against this rare race, so it can happen when already locked. Because of case 2, we need to be able to handle restarting the traversal when rename_lock is already held. This patch fixes all three callers of try_to_ascend(). IBM reported that the deadlock is gone with this patch. [ I rewrote the patch to be smaller and just do the "goto again" if the lock was already held, but credit goes to Miklos for the real work. - Linus ] Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Cc: Al Viro <viro@ZenIV.linux.org.uk> Cc: stable@vger.kernel.org Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-rw-r--r--fs/dcache.c6
1 files changed, 6 insertions, 0 deletions
diff --git a/fs/dcache.c b/fs/dcache.c
index 0364af2311f..693f95bf1ca 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1134,6 +1134,8 @@ positive:
return 1;
rename_retry:
+ if (locked)
+ goto again;
locked = 1;
write_seqlock(&rename_lock);
goto again;
@@ -1236,6 +1238,8 @@ out:
rename_retry:
if (found)
return found;
+ if (locked)
+ goto again;
locked = 1;
write_seqlock(&rename_lock);
goto again;
@@ -3035,6 +3039,8 @@ resume:
return;
rename_retry:
+ if (locked)
+ goto again;
locked = 1;
write_seqlock(&rename_lock);
goto again;