VFS: fix a race in lease-breaking during truncate

It is possible that another process could acquire a new file lease right
after break_lease() is called during a truncate, but before lease-granting
is disabled by the subsequent get_write_access().  Merely switching the
order of the break_lease() and get_write_access() calls prevents this race.

Signed-off-by: David M. Richter <richterd@citi.umich.edu>
Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

1 files changed, 9 insertions, 7 deletions

diff --git a/fs/open.c b/fs/open.c
index e27c205364d3..1d9e5e98bf4e 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -256,24 +256,26 @@ static long do_sys_truncate(const char __user * path, loff_t length)
         if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
                 goto dput_and_out;
 
-        /*
-         * Make sure that there are no leases.
-         */
-        error = break_lease(inode, FMODE_WRITE);
+        error = get_write_access(inode);
         if (error)
                 goto dput_and_out;
 
-        error = get_write_access(inode);
+        /*
+         * Make sure that there are no leases.  get_write_access() protects
+         * against the truncate racing with a lease-granting setlease().
+         */
+        error = break_lease(inode, FMODE_WRITE);
         if (error)
-                goto dput_and_out;
+                goto put_write_and_out;
 
         error = locks_verify_truncate(inode, NULL, length);
         if (!error) {
                 DQUOT_INIT(inode);
                 error = do_truncate(nd.dentry, length, 0, NULL);
         }
-        put_write_access(inode);
 
+put_write_and_out:
+        put_write_access(inode);
 dput_and_out:
         path_release(&nd);
 out: