netdev - Re: [RFC] bonding: fix workqueue re-arming races

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:	Wed, 1 Sep 2010 15:30:56 +0200
From:	Jiri Bohac <jbohac@...e.cz>
To:	Jarek Poplawski <jarkao2@...il.com>
Cc:	Jay Vosburgh <fubar@...ibm.com>, Jiri Bohac <jbohac@...e.cz>,
	bonding-devel@...ts.sourceforge.net, markine@...gle.com,
	chavey@...gle.com, netdev@...r.kernel.org
Subject: Re: [RFC] bonding: fix workqueue re-arming races

On Wed, Sep 01, 2010 at 12:23:56PM +0000, Jarek Poplawski wrote:
> On 2010-08-31 22:54, Jay Vosburgh wrote:
> > 	What prevents this from deadlocking such that cpu A is in
> > bond_close, holding RTNL and in cancel_delayed_work_sync, while cpu B is
> > in the above function, trying to acquire RTNL?
> 
> I guess this one isn't cancelled in bond_close, so it should be safe.

Nah, Jay was correct. Although this work item is not explicitly
cancelled with cancel_delayed_work_sync(), it is on the same
workqueue as work items that are being cancelled with
cancel_delayed_work_sync(), so this can still cause a deadlock.
Fixed in the new version of the patch by putting these on a
separate workqueue.

> > 	Also, assuming for the moment that the above isn't a problem,
> > curr_active_slave may be NULL if the last slave is removed between the
> > time bond_alb_promisc_disable is scheduled and when it runs.  I'm not
> > sure that the alb_bond_info can be guaranteed to be valid, either, if
> > the mode changed.
> 
> Probably some or all of these work functions should test for closing
> eg. with netif_running() or some other flag/variable under rtnl_lock().
> 
> ....
> >> static void bond_work_cancel_all(struct bonding *bond)
> >> {
> >> -	write_lock_bh(&bond->lock);
> >> -	bond->kill_timers = 1;
> >> -	write_unlock_bh(&bond->lock);
> >> -
> >> 	if (bond->params.miimon && delayed_work_pending(&bond->mii_work))
> 
> I'd suggest to skip these delayed_work_pending() tests.

I agree, these should not be needed anymore, since bond_close()
now makes sure the re-arming work is cancelled. I'll update the
patch if Jay thinks it's otherwise good.

-- 
Jiri Bohac <jbohac@...e.cz>
SUSE Labs, SUSE CZ

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html