Re: kernel 3.11.6 general protection fault

From: Borislav Petkov <bp@alien8.de>
To: "MPhil. Emanoil Kotsev" <Emanoil.Kotsev@fincom.at>
Cc: intel-gfx@lists.freedesktop.org, linux-kernel@vger.kernel.org
Subject: Re: kernel 3.11.6 general protection fault
Date: Sun, 17 Nov 2013 16:06:07 +0100	[thread overview]
Message-ID: <20131117150607.GF27323@pd.tnic> (raw)
In-Reply-To: <201311171545.35217.Emanoil.Kotsev@fincom.at>

On Sun, Nov 17, 2013 at 03:45:34PM +0100, MPhil. Emanoil Kotsev wrote:
> this is also true - which makes me sad as the notebook was working
> thgreat in e past 7y

Hmm, maybe it is heading slowly for the eternal hunting fields... :-)

> > What kind of upgrade exactly did you do to a laptop?
> 
> I was using debian squeeze with trinity desktop (KDE 3.5.10) and upgraded to 
> debian wheeze with TDE (3.5.13)

Oh ok, so I thought you were talking about a hw upgrade, like adding
more RAM, hew hdd, etc.

Ok, can you try this: boot without X and try overloading the machine on
the console, i.e. do

while true; do make clean && make -j64; done

or similar in your kernel repository. Does it trigger then?

Although I can't imagine how a software upgrade would cause the
overheating... :-\.

> > Can you revert the upgrade and see whether it still happens?
> This would be hard - no impossible as I have a backup but it will be
> time consuming

You could try booting a distro from a livecd and see any change there...

> $ sensors
> acpitz-virtual-0
> Adapter: Virtual device
> temp1:        +47.5°C  (crit = +126.0°C)

That's some ACPI timezone thing. So what happens if you do

$ watch -n 1 sensors

and you incur the load? Do you hit the critical temperature?

> grep . -EriIn /sys/devices/system/cpu/cpu0/cpufreq
> /sys/devices/system/cpu/cpu0/cpufreq/bios_limit:1:2000000
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor:1:ondemand
> /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_transition_latency:1:10000
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_frequencies:1:2000000 1667000 1333000 1000000
> /sys/devices/system/cpu/cpu0/cpufreq/freqdomain_cpus:1:0 1
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver:1:acpi-cpufreq
> /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_cur_freq:1:1000000
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_governors:1:ondemand powersave performance conservative userspace
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq:1:1000000
> /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_max_freq:1:2000000
> /sys/devices/system/cpu/cpu0/cpufreq/cpuinfo_min_freq:1:1000000
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq:1:2000000
> /sys/devices/system/cpu/cpu0/cpufreq/affected_cpus:1:0
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq:1:1000000
> /sys/devices/system/cpu/cpu0/cpufreq/related_cpus:1:0
> /sys/devices/system/cpu/cpu0/cpufreq/scaling_setspeed:1:<unsupported>

Yeah, I don't see anything wrong with that output.

> I could try this. I guess this assumes I have to have another machine
> running in paralell, but this can be arranged with a little effort

Yep.

> Thanks for the hints. As I never had to do with overheating or
> similar issues, your help is very precious to me. Unfortunately we
> have a little child on board and time is limitted :) to a couple of
> hours daily, where I can work at home which means even less time for
> debugging. But I never give up. I just want to be sure that it is not
> a hardware issue

No worries, take care of the child first - the laptop and everyone else
can wait :-)

-- 
Regards/Gruss,
    Boris.

Sent from a fat crate under my desk. Formatting is fine.
--
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx