Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 21 Sep 2011 13:02:46 +0300
From:      Andriy Gapon <avg@FreeBSD.org>
To:        Kostik Belousov <kostikbel@gmail.com>
Cc:        Lawrence Stewart <lstewart@FreeBSD.org>, freebsd-x11@FreeBSD.org
Subject:   Re: Deterministic panic 9.0-BETA2 with X + drm/radeon
Message-ID:  <4E79B646.4030904@FreeBSD.org>
In-Reply-To: <20110921095513.GI1511@deviant.kiev.zoral.com.ua>
References:  <4E798E66.1020603@freebsd.org> <4E79AF56.2040703@FreeBSD.org> <20110921095513.GI1511@deviant.kiev.zoral.com.ua>

next in thread | previous in thread | raw e-mail | index | archive | help
on 21/09/2011 12:55 Kostik Belousov said the following:
> On Wed, Sep 21, 2011 at 12:33:10PM +0300, Andriy Gapon wrote:
>> on 21/09/2011 10:12 Lawrence Stewart said the following:
>>> Kernel page fault with the following non-sleepable locks held: 
>>> exclusive sleep mutex drmdev (drmdev) r = 0 (0xfffffe000f831ca0) locked
>>> @ /usr/src/sys/modules/drm/drm/../../../dev/drm/drm_drv.c:791
>> 
>> I think that this is a well-known but not yet fixed issue. I think that
>> it should already be addressed in the general drm part of kib's Intel GEM
>> code. Please see this thread for some earlier history: 
>> http://lists.freebsd.org/pipermail/freebsd-current/2010-November/020914.html
>>
>
>> 
It is, but only for Intel driver. The approach cannot be applied to the
> old dri1 drm drivers, because they need the drm device lock held in the 
> interrupt handler.

Not sure if this is true for radeon though.
I've never had any panics/warnings with the patch and INVARIANTS and WITNESS.


-- 
Andriy Gapon



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4E79B646.4030904>