reassign 651532 src:linux-2.6 3.1.1-1
quit

Hi,

Camaleón wrote:

> As soon as the firmware package is removed (which means no more 3D
> accelerated graphics capabilities) the system runs normally and no
> kernel oops presents.
>
> root@linux:~# lspci | grep VGA
> 01:00.0 VGA compatible controller: ATI Technologies Inc M56P [Radeon Mobility X1600]
[...]
> radeon 0000:01:00.0: GPU lockup CP stall for more than 10040msec
> ------------[ cut here ]------------
> WARNING: at /build/buildd-linux-2.6_3.1.1-1-i386-6vi5gY/linux-2.6-3.1.1/debian/build/source_i386_none/drivers/gpu/drm/radeon/radeon_fence.c:267 radeon_fence_wait+0x22e/0x298 [radeon]()
> Hardware name: iMac5,1
> GPU lockup (waiting for 0x00005C23 last fence id 0x00005C20)

Thanks for reporting it. Can you reproduce this without the virtualbox
and wl drivers?

There have been some changes in this area recently upstream (e.g.,
v3.2-rc1~135^2~5 and v3.2-rc1~11^2~10). Please test with v3.2-rc4
or later from experimental, and if it still exhibits the problem,
send a summary to https://bugs.freedesktop.org/, product DRI,
component DRM/Radeon and let us know the bug number so we can track it.

In your summary, be sure to include steps to reproduce the trouble,
any visible symptoms, /var/log/Xorg.0.log, full "dmesg" output from
booting and reproducing the bug, and a link to
http://bugs.debian.org/651532 in case readers want the full story.

Hope that helps,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111210004817.GA3062@elie.hsd1.il.comcast.net">ht tp://lists.debian.org/20111210004817.GA3062@elie.hsd1.il.comcast.net

Camaleón wrote:

> The user has provided the following information:
>
> - The crash is still reproducible with Debian's stock kernel 3.1 with no
> VirtualBox and wl modules loaded.
>
> - The crash is also happening with Debian's experimental kernel
> (3.2-rc4). I'm attaching the full syslog and a snapshot for the crash
> (what follows is a small snippet for the kernel crash):
>
> Dec 11 21:00:55 linux kernel: [ 39.546036] upowerd: Corrupted page table at address b77df424
[...]

Hmm, this time there are no messages about the Radeon CP. Does
uninstalling the radeon firmware (or disabling the radeon driver some
other way) still avoid trouble?



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111212174940.GA31793@elie.hsd1.il.comcast.net">h ttp://lists.debian.org/20111212174940.GA31793@elie.hsd1.il.comcast.net

El 2011-12-12 a las 11:49 -0600, Jonathan Nieder escribió:

> Camaleón wrote:
>
> > The user has provided the following information:
> >
> > - The crash is still reproducible with Debian's stock kernel 3.1 with no
> > VirtualBox and wl modules loaded.
> >
> > - The crash is also happening with Debian's experimental kernel
> > (3.2-rc4). I'm attaching the full syslog and a snapshot for the crash
> > (what follows is a small snippet for the kernel crash):
> >
> > Dec 11 21:00:55 linux kernel: [ 39.546036] upowerd: Corrupted page table at address b77df424
> [...]
>
> Hmm, this time there are no messages about the Radeon CP. Does
> uninstalling the radeon firmware (or disabling the radeon driver some
> other way) still avoid trouble?

The user tried the latest kernel (3.2-rc4) with the firmware blob
package installed to enable 3D acceleration. Do you want he tries again
this kernel with no firmware package at all?

--
Camaleón



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111212183624.GA8202@stt008.linux.site">http://lists.debian.org/20111212183624.GA8202@stt008.linux.site

Camaleón wrote:

> The user tried the latest kernel (3.2-rc4) with the firmware blob
> package installed to enable 3D acceleration. Do you want he tries again
> this kernel with no firmware package at all?

Yes, that would help to figure out whether these new crashes are due
to the radeon driver as before or some independent bug.

Even better would be to try booting without loading the radeon module,
by putting

blacklist radeon

in a file /etc/modprobe.d/oa-blacklist-radeon.conf and booting in
"recovery mode" so X does not get started.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111212184543.GI31793@elie.hsd1.il.comcast.net">h ttp://lists.debian.org/20111212184543.GI31793@elie.hsd1.il.comcast.net

El 2011-12-12 a las 12:45 -0600, Jonathan Nieder escribió:

> Camaleón wrote:
>
> > The user tried the latest kernel (3.2-rc4) with the firmware blob
> > package installed to enable 3D acceleration. Do you want he tries again
> > this kernel with no firmware package at all?
>
> Yes, that would help to figure out whether these new crashes are due
> to the radeon driver as before or some independent bug.
>
> Even better would be to try booting without loading the radeon module,
> by putting
>
> blacklist radeon
>
> in a file /etc/modprobe.d/oa-blacklist-radeon.conf and booting in
> "recovery mode" so X does not get started.

Would appending "nomodeset" to the kernel's GRUB line (→ disabling KMS)
do the same job or you need the module is expressly blacklisted?

--
Camaleón



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111212185603.GA8297@stt008.linux.site">http://lists.debian.org/20111212185603.GA8297@stt008.linux.site

Camaleón wrote:
> El 2011-12-12 a las 12:45 -0600, Jonathan Nieder escribió:

>> Even better would be to try booting without loading the radeon module,
>> by putting
>>
>> blacklist radeon
>>
>> in a file /etc/modprobe.d/oa-blacklist-radeon.conf and booting in
>> "recovery mode" so X does not get started.
>
> Would appending "nomodeset" to the kernel's GRUB line (→ disabling KMS)
> do the same job or you need the module is expressly blacklisted?

I haven't had good luck with "nomodeset" generally. But really, if you
can do anything (like removing the firmware package) to the radeon
module to make the problem go away, that will be enough to put the
ball in their court, and I will be happy.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111212190725.GK31793@elie.hsd1.il.comcast.net">h ttp://lists.debian.org/20111212190725.GK31793@elie.hsd1.il.comcast.net

El 2011-12-12 a las 13:07 -0600, Jonathan Nieder escribió:

> I haven't had good luck with "nomodeset" generally. But really, if you
> can do anything (like removing the firmware package) to the radeon
> module to make the problem go away, that will be enough to put the
> ball in their court, and I will be happy.

The user reports that kernel 3.2-rc4 works fine when package "firmware-
linux-nonfree" is not installed.

--
Camaleón



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111213125208.GA9068@stt008.linux.site">http://lists.debian.org/20111213125208.GA9068@stt008.linux.site

Camaleón wrote:

> The user reports that kernel 3.2-rc4 works fine when package "firmware-
> linux-nonfree" is not installed.

Excellent. The next step is indeed to report this upstream.

Be sure to mention:

- steps to reproduce the problem
- symptoms and how they differ from the expected result (that should
be simple enough in this case )
- which other versions you tried and the result for each
- a link to http://bugs.debian.org/651532 for the full story
- any other hints or weird observations

Attach:

- full "dmesg" output and Xorg.0.log from booting and reproducing the
problem with 3.2-rc4, with the firmware available
- full "dmesg" output and Xorg.0.log from booting 3.2-rc4 without the
firmware available, for contrast

And please let us know the bug number so we can track it.

Thanks very much,
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111213195213.GA2011@elie.hsd1.il.comcast.net">ht tp://lists.debian.org/20111213195213.GA2011@elie.hsd1.il.comcast.net

El 2011-12-13 a las 13:52 -0600, Jonathan Nieder escribió:

> Excellent. The next step is indeed to report this upstream.

(...)

> And please let us know the bug number so we can track it.

Done:

https://bugs.freedesktop.org/show_bug.cgi?id=43835

--
Camaleón



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111214170954.GA8692@stt008.linux.site">http://lists.debian.org/20111214170954.GA8692@stt008.linux.site