xfs
[Top] [All Lists]

[PATCH] xfs: Correctly lock inode when removing suid and security marks

To: xfs@xxxxxxxxxxx
Subject: [PATCH] xfs: Correctly lock inode when removing suid and security marks
From: Jan Kara <jack@xxxxxxx>
Date: Tue, 2 Dec 2014 16:01:29 +0100
Cc: Jan Kara <jack@xxxxxxx>
Delivered-to: xfs@xxxxxxxxxxx
Currently XFS calls file_remove_suid() without holding i_mutex. This is
wrong because that function can end up messing with file permissions and
security xattrs for which we need i_mutex held.

Fix the problem by grabbing iolock exclusively when we will need to
change anything in permissions / xattrs.

Signed-off-by: Jan Kara <jack@xxxxxxx>
---
 fs/xfs/xfs_file.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index eb596b419942..ad6636ac4943 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -521,6 +521,18 @@ restart:
        if (error)
                return error;
 
+       /* For changing security info in file_remove_suid() we need i_mutex */
+       if (!IS_NOSEC(inode) && *iolock == XFS_IOLOCK_SHARED) {
+               struct dentry *dentry = file->f_path.dentry;
+
+               if (should_remove_suid(dentry) ||
+                   security_inode_need_killpriv(dentry)) {
+                       xfs_rw_iunlock(ip, *iolock);
+                       *iolock = XFS_IOLOCK_EXCL;
+                       xfs_rw_ilock(ip, *iolock);
+                       goto restart;
+               }
+       }
        /*
         * If the offset is beyond the size of the file, we need to zero any
         * blocks that fall between the existing EOF and the start of this
-- 
1.8.1.4

<Prev in Thread] Current Thread [Next in Thread>