- issue warning when bad inode found via ext3_lookup

I have a hand-crafted bad filesystem image which has a corrupted directory entry:

[root@inode ~]# ls mnt/Picture
LINKS_20 OBEN_20 VORNE_20 VORN_LINKS_20 VORN_RECHTS_20 [root@inode ~]# ls mnt/Picture/VORN_LINKS_20
ls: cannot access mnt/Picture/VORN_LINKS_20: No such file or directory [root@inode ~]# stat mnt/Picture/VORN_LINKS_20
stat: cannot stat `mnt/Picture/VORN_LINKS_20': No such file or directory [root@inode ~]# ls -l mnt/Picture
ls: cannot access mnt/Picture/VORN_LINKS_20: No such file or directory total 8
drwxr-xr-x 2 root root 1024 2007-09-04 13:36 LINKS_20 drwxr-xr-x 2 root root 1024 2007-09-04 13:36 OBEN_20 drwxr-xr-x 2 root root 1024 2007-09-04 13:36 VORNE_20
d????????? ? ? ? ? ? VORN_LINKS_20
drwxr-xr-x 2 root root 1024 2007-09-04 13:36 VORN_RECHTS_20

e2fsck also knows it's corrupted:
Pass 2: Checking directory structure
Entry 'VORN_LINKS_20' in /Picture (2049) has deleted/unused inode 13. Clear? no

Entry 'VORN_LINKS_20' in /Picture (2049) has an incorrect filetype (was 2, should be 1). Fix? no

Pass 3: Checking directory connectivity
Unconnected directory inode 2053 (/Picture/???)

BUT there are no kernel messages anywhere.

I think the below makes sense; I don't think there are any instances where an inode found via a filename passed to ext3_lookup should be returning a bad inode without warning; NFS may pass in a stale filehandle which also makes a bad inode, but that shouldn't go through lookup...

Comments? (if it looks good I'll resubmit for ext*_lookup)

---

Index: linux-2.6.24-rc3/fs/ext3/namei.c
===================================================================
@@ -1049,6 +1049,10 @@ static struct dentry *ext3_lookup(struct
 			return ERR_PTR(-EACCES);
 
 		if (is_bad_inode(inode)) {
+			ext3_warning(inode->i_sb, __FUNCTION__,
+				   "bad inode %lu for file %s in dir #%lu",
+				   inode->i_ino, dentry->d_name.name,
+				   dir->i_ino);
 			iput(inode);
 			return ERR_PTR(-ENOENT);
 		}