linux-kernel - Re: [PATCH 04/11] vfs: check unlinked ancestors before mount

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 5 Sep 2013 17:52:51 +0200
From:	Miklos Szeredi <miklos@...redi.hu>
To:	Al Viro <viro@...iv.linux.org.uk>
Cc:	Linux-Fsdevel <linux-fsdevel@...r.kernel.org>,
	Kernel Mailing List <linux-kernel@...r.kernel.org>,
	"mszeredi@...e.cz" <mszeredi@...e.cz>,
	David Howells <dhowells@...hat.com>,
	Steven Whitehouse <swhiteho@...hat.com>,
	Trond Myklebust <Trond.Myklebust@...app.com>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: [PATCH 04/11] vfs: check unlinked ancestors before mount

On Thu, Sep 5, 2013 at 4:56 PM, Al Viro <viro@...iv.linux.org.uk> wrote:
> I'd probably just do this, and to hell with helper functions...
>
> int d_set_mounted(struct dentry *dentry)
> {
>         struct dentry *p;
>         int ret = 0;
>         write_seqlock(&rename_lock);
>         for (p = dentry; !IS_ROOT(p); p = p->d_parent) {
>                 /* Need exclusion wrt. check_submounts_and_drop() */
>                 spin_lock(&p->d_lock);
>                 if (unlikely(d_unhashed(p))) {
>                         spin_unlock(&p->d_lock);
>                         ret = -ENOENT;
>                         goto out;
>                 }
>                 spin_unlock(&p->d_lock);
>         }
>         spin_lock(&dentry->d_lock);
>         dentry->d_flags |= DCACHE_MOUNTED;
>         spin_unlock(&dentry->d_lock);
> out:
>         write_sequnlock(&rename_lock);
>         return ret;
> }

One issue with that: the dentry should be checked and marked within
the same d_locked region.   Because e.g. d_invalidate() relies solely
on d_lock for non-dir mounts and d_mountpoint() checking, no
rename_lock protection there.

Thanks,
Miklos
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/