Re: [PATCH v2 1/3] vfs: Check for invalid i_uid in may_follow_link()

[Serge Hallyn]

Quoting Seth Forshee (seth.forshee@xxxxxxxxxxxxx):
> Filesystem uids which don't map into a user namespace may result
> in inode->i_uid being INVALID_UID. A symlink and its parent
> could have different owners in the filesystem can both get
> mapped to INVALID_UID, which may result in following a symlink
> when this would not have otherwise been permitted. Prevent this
> by adding a check that the uid is valid before the comparison.
> 
> Signed-off-by: Seth Forshee <seth.forshee@xxxxxxxxxxxxx>

I'm a bit uncomfortable about this, but I can't put my finger
on why.  Wonder if it could mess up root looking into
a malicious user's task by looking under /proc/self/root.
I suppose not, as this should only be the case (with root in
init_user_ns) for fuse?

Anyway it seems needed for keeping root from falling into a trap.

Acked-by: Serge E. Hallyn <serge.hallyn@xxxxxxxxxx>

> ---
>  fs/namei.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/namei.c b/fs/namei.c
> index a996bb48dfab..193da09e903e 100644
> --- a/fs/namei.c
> +++ b/fs/namei.c
> @@ -741,7 +741,7 @@ static inline int may_follow_link(struct path *link, struct nameidata *nd)
>  		return 0;
>  
>  	/* Allowed if parent directory and link owner match. */
> -	if (uid_eq(parent->i_uid, inode->i_uid))
> +	if (uid_valid(inode->i_uid) && uid_eq(parent->i_uid, inode->i_uid))
>  		return 0;
>  
>  	audit_log_link_denied("follow_link", link);
> -- 
> 1.9.1
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/