From mboxrd@z Thu Jan  1 00:00:00 1970
From: Eric Dumazet <eric.dumazet@gmail.com>
Subject: Re: [PATCH] pkt_sched: Fix lockdep warning on est_tree_lock in
 gen_estimator
Date: Fri, 03 Sep 2010 07:26:45 +0200
Message-ID: <1283491605.3699.1378.camel@edumazet-laptop>
References: <20100902201257.GA3089@del.dom.local>
	 <20100902.132248.229742285.davem@davemloft.net>
	 <1283459521.3699.193.camel@edumazet-laptop>
	 <20100902205535.GB3089@del.dom.local>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: David Miller <davem@davemloft.net>, netdev@vger.kernel.org
To: Jarek Poplawski <jarkao2@gmail.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-ww0-f44.google.com ([74.125.82.44]:40901 "EHLO
	mail-ww0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1750852Ab0ICF0v (ORCPT
	<rfc822;netdev@vger.kernel.org>); Fri, 3 Sep 2010 01:26:51 -0400
Received: by wwj40 with SMTP id 40so1981457wwj.1
        for <netdev@vger.kernel.org>; Thu, 02 Sep 2010 22:26:50 -0700 (PDT)
In-Reply-To: <20100902205535.GB3089@del.dom.local>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

Le jeudi 02 septembre 2010 =C3=A0 22:55 +0200, Jarek Poplawski a =C3=A9=
crit :
> On Thu, Sep 02, 2010 at 10:32:01PM +0200, Eric Dumazet wrote:
> > Le jeudi 02 septembre 2010 ?? 13:22 -0700, David Miller a =C3=A9cri=
t :
> > > From: Jarek Poplawski <jarkao2@gmail.com>
> > > Date: Thu, 2 Sep 2010 22:12:58 +0200
> > >=20
> > > > This patch fixes a lockdep warning:
> > > >=20
> > > > [  516.287584] =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> > > > [  516.288386] [ INFO: possible irq lock inversion dependency d=
etected ]
> > > > [  516.288386] 2.6.35b #7
> > > > [  516.288386] ------------------------------------------------=
---------
> > > > [  516.288386] swapper/0 just changed the state of lock:
> > > > [  516.288386]  (&qdisc_tx_lock){+.-...}, at: [<c12eacda>] est_=
timer+0x62/0x1b4
> > > > [  516.288386] but this lock took another, SOFTIRQ-unsafe lock =
in the past:
> > > > [  516.288386]  (est_tree_lock){+.+...}
> > > > [  516.288386]=20
> > > > [  516.288386] and interrupts could create inverse lock orderin=
g between them.
> > > > ...
> > > >=20
> > > > So, est_tree_lock needs BH protection because it's taken by
> > > > qdisc_tx_lock, which is used both in BH and process contexts.
> > > > (Full warning with this patch at netdev, 02 Sep 2010.)
> > > >=20
> > > > Fixes commit: ae638c47dc040b8def16d05dc6acdd527628f231
> > > >=20
> > > > Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
> > >=20
> > > Applied, thanks Jarek.
> > >=20
> > > Eric, please make an effort to run with lockdep enabled when maki=
ng
> > > locking changes, just for fun :-)
> >=20
> > Hmm... I still trying to understand the problem.
> >=20
> > I dont understand Jarek analysis yet.
> >=20
>=20
> cpu1:					cpu2:
> (process)				(process)
> qdisc_tx_lock				est_tree_lock
> (waiting for est_tree_lock)		(bh)
> 					(waiting for qdisc_tx_lock)
>=20

Cool. Reading commit ae638c47dc040b, I thought only RTNL was possibly
hold in these paths.

Thanks Jarek !