2007-11-13 10:05:48

by Adrian Bunk

[permalink] [raw]
Subject: [RFC: 2.6.16 patch] return ENOENT from ext3_link when racing with unlink

It seems the commit below (I've removed the ext4 part that doesn't
apply) should also be included in the 2.6.16 branch, or do I miss
anything?

TIA
Adrian


commit 2988a7740dc0dd9a0cb56576e8fe1d777dff0db3
Author: Eric Sandeen <[email protected]>
Date: Sat Feb 10 01:46:16 2007 -0800

[PATCH] return ENOENT from ext3_link when racing with unlink

Return -ENOENT from ext[34]_link if we've raced with unlink and i_nlink is
0. Doing otherwise has the potential to corrupt the orphan inode list,
because we'd wind up with an inode with a non-zero link count on the list,
and it will never get properly cleaned up & removed from the orphan list
before it is freed.

[[email protected]: build fix]
Signed-off-by: Eric Sandeen <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c
index 4df39c4..57402b5 100644
--- a/fs/ext3/namei.c
+++ b/fs/ext3/namei.c
@@ -2191,6 +2191,12 @@ static int ext3_link (struct dentry * old_dentry,

if (inode->i_nlink >= EXT3_LINK_MAX)
return -EMLINK;
+ /*
+ * Return -ENOENT if we've raced with unlink and i_nlink is 0. Doing
+ * otherwise has the potential to corrupt the orphan inode list.
+ */
+ if (inode->i_nlink == 0)
+ return -ENOENT;

retry:
handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) +