Erez Zadok <ezk@fsl.cs.sunysb.edu> writes:
> Niklos, I forgot to mention that I had to apply a small fix to a VFS
> bug in fs/namei.c:1362, where a mix of symlinks and renames, using
> racer, triggers an BUG_ON at the VFS layer (very reproducible).
> Without this fix, racer oopses in the VFS well before it gets to
> trigger overlayfs bugs.  It's a pity 2.6.39 was released with this
> very reproducible *VFS* level bug (doesn't anyone run racer+fsx+ltp
> before releasing a new kernel?)
>
> Anyway, here's the small patch here.  Clearly viro/hch need to review
> this "fix" b/c I'm not sure it's really the right one

Patch looks good to me.

Acked-by: Miklos Szeredi <mszeredi@suse.cz>

If you've got a patch fixing a serious bug, you'd better send it to
Linus and Andrew (as well as the VFS maintainers), where it'll get more
prompt evaluation.

You can still do that, also adding "CC: stable@kernel.org" to the SOB
block in the patch so that the fix quickly makes it into the 2.6.39
stable series.

Thanks,
Miklos


>
> VFS: move BUG_ON test for symlink nd->depth after current->link_count test
>
> This solves a bug in nested_symlink (which was rewritten from
> do_follow_link), and follows the order of depth tests that existed before.
> The bug triggers a BUG_ON in fs/namei.c:1346, when running racer with
> symlink and rename ops.
>
> Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu>
> diff --git a/fs/namei.c b/fs/namei.c
> index 017c3fa..7a93387 100644
> --- a/fs/namei.c
> +++ b/fs/namei.c
> @@ -1343,12 +1343,12 @@ static inline int nested_symlink(struct path *path, struct nameidata *nd)
>  {
>  	int res;
> 
> -	BUG_ON(nd->depth >= MAX_NESTED_LINKS);
>  	if (unlikely(current->link_count >= MAX_NESTED_LINKS)) {
>  		path_put_conditional(path, nd);
>  		path_put(&nd->path);
>  		return -ELOOP;
>  	}
> +	BUG_ON(nd->depth >= MAX_NESTED_LINKS);
> 
>  	nd->depth++;
>  	current->link_count++;