nfs: remove unnecessary check from nfs_rename()
authorMiklos Szeredi <mszeredi@suse.cz>
Thu, 3 Dec 2009 20:58:56 +0000 (15:58 -0500)
committerTrond Myklebust <Trond.Myklebust@netapp.com>
Thu, 3 Dec 2009 20:58:56 +0000 (15:58 -0500)
VFS already checks if both source and target are directories.

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
fs/nfs/dir.c

index 7cb2985..b5fae19 100644 (file)
@@ -1601,13 +1601,8 @@ static int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
         * silly-rename. If the silly-rename succeeds, the
         * copied dentry is hashed and becomes the new target.
         */
-       if (!new_inode)
-               goto go_ahead;
-       if (S_ISDIR(new_inode->i_mode)) {
-               error = -EISDIR;
-               if (!S_ISDIR(old_inode->i_mode))
-                       goto out;
-       } else if (atomic_read(&new_dentry->d_count) > 2) {
+       if (new_inode && !S_ISDIR(new_inode->i_mode) &&
+           atomic_read(&new_dentry->d_count) > 2) {
                int err;
                /* copy the target dentry's name */
                dentry = d_alloc(new_dentry->d_parent,
@@ -1627,7 +1622,6 @@ static int nfs_rename(struct inode *old_dir, struct dentry *old_dentry,
                        goto out;
        }
 
-go_ahead:
        /*
         * ... prune child dentries and writebacks if needed.
         */