From mboxrd@z Thu Jan  1 00:00:00 1970
From: Marcin Slusarz <marcin.slusarz@gmail.com>
Subject: Re: drm/nouveau: crash regression in 3.5
Date: Thu, 26 Jul 2012 17:47:09 +0200
Message-ID: <20120726154709.GA3220@joi.lan>
References: <500D916A.60703@odi.ch>
 <20120724170002.GA3129@joi.lan>
 <500ED9EC.5040309@odi.ch>
 <20120724205746.GA8707@joi.lan>
 <500FB279.1020904@odi.ch>
 <20120725184205.GA3119@joi.lan>
 <50113E76.9090706@odi.ch>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: QUOTED-PRINTABLE
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <50113E76.9090706@odi.ch>
Sender: linux-kernel-owner@vger.kernel.org
To: Ortwin =?utf-8?B?R2zDvGNr?= <odi@odi.ch>
Cc: linux-kernel@vger.kernel.org, bskeggs@redhat.com, dri-devel@lists.freedesktop.org, airlied@redhat.com
List-Id: dri-devel@lists.freedesktop.org

On Thu, Jul 26, 2012 at 02:56:22PM +0200, Ortwin Gl=C3=BCck wrote:
> On 25.07.2012 20:42, Marcin Slusarz wrote:
> > Good, below patch should fix this panic.
> >
> > Note that you can hit an oops in drm_handle_vblank because patch fr=
om
> > http://lists.freedesktop.org/archives/dri-devel/2012-May/023498.htm=
l
> > has not been applied (yet?).
>=20
> After applying your patch, it still crashes, although with a slightly=
=20
> different stack trace. I then also applied the second patch, but that=
=20
> doesn't make any difference. New log attached.
>=20
> Looks like interrupt occurs before nouveau_software_context_new() is=20
> called? Shouldn't the initialization be done from=20
> nouveau_irq_preinstall() so it is available when the irq occurs? Agai=
n,=20
> I am not an expert here. Just guessing...

I didn't look at it yet, but can you catch panic with maxcpus=3D1?

Marcin