The patch below does not apply to the .38-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

>From 03885ac3c79ab7f3f4a8e502486be2ea6c85c3e3 Mon Sep 17 00:00:00 2001
From: Josh Hunt <[email protected]>
Date: Thu, 24 Feb 2011 11:48:22 +0100
Subject: [PATCH] ext2: Fix link count corruption under heavy link+rename load

vfs_rename_other() does not lock renamed inode with i_mutex. Thus changing
i_nlink in a non-atomic manner (which happens in ext2_rename()) can corrupt
it as reported and analyzed by Josh.

In fact, there is no good reason to mess with i_nlink of the moved file.
We did it presumably to simulate linking into the new directory and unlinking
from an old one. But the practical effect of this is disputable because fsck
can possibly treat file as being properly linked into both directories without
writing any error which is confusing. So we just stop increment-decrement
games with i_nlink which also fixes the corruption.

CC: [email protected]
CC: Al Viro <[email protected]>
Signed-off-by: Josh Hunt <[email protected]>
Signed-off-by: Jan Kara <[email protected]>

diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c
index 2e1d834..adb9185 100644
--- a/fs/ext2/namei.c
+++ b/fs/ext2/namei.c
@@ -344,7 +344,6 @@ static int ext2_rename (struct inode * old_dir, struct 
dentry * old_dentry,
                new_de = ext2_find_entry (new_dir, &new_dentry->d_name, 
&new_page);
                if (!new_de)
                        goto out_dir;
-               inode_inc_link_count(old_inode);
                ext2_set_link(new_dir, new_de, new_page, old_inode, 1);
                new_inode->i_ctime = CURRENT_TIME_SEC;
                if (dir_de)
@@ -356,12 +355,9 @@ static int ext2_rename (struct inode * old_dir, struct 
dentry * old_dentry,
                        if (new_dir->i_nlink >= EXT2_LINK_MAX)
                                goto out_dir;
                }
-               inode_inc_link_count(old_inode);
                err = ext2_add_link(new_dentry, old_inode);
-               if (err) {
-                       inode_dec_link_count(old_inode);
+               if (err)
                        goto out_dir;
-               }
                if (dir_de)
                        inode_inc_link_count(new_dir);
        }
@@ -369,12 +365,11 @@ static int ext2_rename (struct inode * old_dir, struct 
dentry * old_dentry,
        /*
         * Like most other Unix systems, set the ctime for inodes on a
         * rename.
-        * inode_dec_link_count() will mark the inode dirty.
         */
        old_inode->i_ctime = CURRENT_TIME_SEC;
+       mark_inode_dirty(old_inode);
 
        ext2_delete_entry (old_de, old_page);
-       inode_dec_link_count(old_inode);
 
        if (dir_de) {
                if (old_dir != new_dir)

_______________________________________________
stable mailing list
[email protected]
http://linux.kernel.org/mailman/listinfo/stable

Reply via email to