netdev - Re: [PATCH net] packet: call fanout_release, while UNREGISTERING a netdev

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:   Mon, 30 Jan 2017 09:26:29 -0800
From:   Eric Dumazet <eric.dumazet@...il.com>
To:     David Miller <davem@...emloft.net>
Cc:     anoob.soman@...rix.com, netdev@...r.kernel.org
Subject: Re: [PATCH net] packet: call fanout_release, while UNREGISTERING a
 netdev

On Thu, 2016-10-06 at 20:50 -0400, David Miller wrote:
> From: Anoob Soman <anoob.soman@...rix.com>
> Date: Wed, 5 Oct 2016 15:12:54 +0100
> 
> > If a socket has FANOUT sockopt set, a new proto_hook is registered
> > as part of fanout_add(). When processing a NETDEV_UNREGISTER event in
> > af_packet, __fanout_unlink is called for all sockets, but prot_hook which was
> > registered as part of fanout_add is not removed. Call fanout_release, on a
> > NETDEV_UNREGISTER, which removes prot_hook and removes fanout from the
> > fanout_list.
> > 
> > This fixes BUG_ON(!list_empty(&dev->ptype_specific)) in netdev_run_todo()
> > 
> > Signed-off-by: Anoob Soman <anoob.soman@...rix.com>
> 
> Applied and queued up for -stable, thanks.

This commit (6664498280cf "packet: call fanout_release, while
UNREGISTERING a netdev")
looks buggy :

We end up calling fanout_release() while holding a spinlock
( spin_lock(&po->bind_lock); )

But fanout_release() grabs a mutex ( mutex_lock(&fanout_mutex) ), and
this is absolutely not valid while holding a spinlock.

Anoob, can you cook a fix, I guess you have a way to reproduce the thing
that wanted a kernel patch ?

(Please build your test kernel with CONFIG_LOCKDEP=y)

Thanks.