Celeron N2940 Segmentation faults & application crashes

hrobeers · 2015-07-29 08:15:32

Hi all,

I bought a new low budget laptop (https://www.bto.eu/bto-v-book-15cl07.html).
I'm very happy with it, not considering the crashes that occur from time to time.

Crashes happen frequently when compiling source code with gcc.
Most of the time these are SEGMENTATION FAULTs, and rarely a BUS ERROR appears.

Also Firefox and Chromium crash regularly. And sometimes the whole system hangs (kernel panic) and a hard reboot is needed.

I was thinking that it is a bad processor causing these issues.
But I'm not sure, since the last linux-firmware update seems to have increased the stability (but crashes still happen).

I've been using the 4.1.2-2 kernel, but the issues also happen using the lts kernel.

Can someone help me troubleshoot these issues?

Thanks for any clues!
Hans

hrobeers · 2015-07-29 09:48:52

I'm thinking that increasing the CPU voltage can increase the system stability?

Does anyone know how i can increase the CPU voltage with a Celeron N2940?

I never did that before and can't find these settings in the bios or using cpupower.

mauritiusdadd · 2015-07-29 09:58:41

Did you install the micocode updates? And did they apply correctly? https://wiki.archlinux.org/index.php/Microcode

hrobeers · 2015-07-29 10:38:13

No, I didn't do that.

Is it in the beginners guide? I must have missed that part.

Thanks, I'll try that later today. I have a Memtest running at the moment.

Edit: Ok it is in the guide, but since this is not my first installation, I read over it.

Last edited by hrobeers (2015-07-29 10:39:50)

mauritiusdadd · 2015-07-29 10:40:28

hrobeers wrote:

Is it in the beginners guide? I must have missed that part.

https://wiki.archlinux.org/index.php/Be … bootloader

hrobeers · 2015-07-29 11:10:31

ucode installed.

Unfortunately it looks like there are no updates for my hardware yet:

$ dmesg | grep microcode
[    1.431173] microcode: CPU0 sig=0x30678, pf=0x8, revision=0x809
[    1.431189] microcode: CPU1 sig=0x30678, pf=0x8, revision=0x809
[    1.431200] microcode: CPU2 sig=0x30678, pf=0x8, revision=0x809
[    1.431216] microcode: CPU3 sig=0x30678, pf=0x8, revision=0x809
[    1.431337] microcode: Microcode Update Driver: v2.00 <tigran@aivazian.fsnet.co.uk>, Peter Oruba

See: https://wiki.archlinux.org/index.php/Mi … ed_on_boot

I'll report on stability later, no crashes yet, but still early.

hrobeers · 2015-07-29 11:25:02

Unfortunately, this didn't fix the issues.

I still get a SEGMENTATION FAULT when comiling source code using gcc.

Any ideas for further troubleshooting? Do I need to wait for ucode updates?

mauritiusdadd · 2015-07-29 12:39:37

What are you trying to compile and what version of gcc are you using?
Can you post the actual compilation output?
Did the memtest give any result?
Is there anything interesting in the journal?

The SIG_BUS error indicates that probably your CPU can't handle a misaligned pointer (see this other thread https://bbs.archlinux.org/viewtopic.php … 2#p1536852).

hrobeers · 2015-07-29 14:37:26

These segmentation faults are not really application specific, I'm just using GCC since it reproduces the crashes consistently (high CPU load for long time).

I'm compiling a personal project: https://github.com/hrobeers/finFoil
And it gives errors on random moments in the compilation process like below:

/home/hans/development/ffserver/src/hrlib/version.cpp: In constructor ‘hrlib::Version::Version(int, int, int, int, QString, hrlib::ReleaseType::e)’:
/home/hans/development/ffserver/src/hrlib/version.cpp:61:1: internal compiler error: Segmentation fault
 }
 ^
Please submit a full bug report,
with preprocessed source if appropriate.
See <https://bugs.archlinux.org/> for instructions.
../src/hrlib/CMakeFiles/hrlib.dir/build.make:123: recipe for target '../src/hrlib/CMakeFiles/hrlib.dir/version.cpp.o' failed
make[2]: *** [../src/hrlib/CMakeFiles/hrlib.dir/version.cpp.o] Error 1
CMakeFiles/Makefile2:185: recipe for target '../src/hrlib/CMakeFiles/hrlib.dir/all' failed
make[1]: *** [../src/hrlib/CMakeFiles/hrlib.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....

I'm using GCC 5.2.0 now (but this happend with older versions too)

memtest did not give any errors

The bus error only happened once, so it's not as consistent as the segmentation fault.

How do I read the journal?
Edit: I assume you mean this: https://wiki.archlinux.org/index.php/Systemd#Journal ? Any clues what to look for?

Many thanks for your help!

Last edited by hrobeers (2015-07-29 14:41:35)

mauritiusdadd · 2015-07-30 06:11:52

hrobeers wrote:

I assume you mean this: https://wiki.archlinux.org/index.php/Systemd#Journal ? Any clues what to look for?

Yes, I mean that. Is there any error that can be related to your problem? Can you post the full journal for the current boot after a crash has happened?

I don't think that overvolting the CPU helps in this case and messing up with voltages (especially when overvolting) can be dangerous and unfortunately I don't know a method to do it if the bios does not allow it. I'm sorry but I have no clue on what causes you problem.

--edit: typo

Last edited by mauritiusdadd (2015-07-30 08:54:20)

Arch Linux

#1 2015-07-29 08:15:32

Celeron N2940 Segmentation faults & application crashes

#2 2015-07-29 09:48:52

Re: Celeron N2940 Segmentation faults & application crashes

#3 2015-07-29 09:58:41

Re: Celeron N2940 Segmentation faults & application crashes

#4 2015-07-29 10:38:13

Re: Celeron N2940 Segmentation faults & application crashes

#5 2015-07-29 10:40:28

Re: Celeron N2940 Segmentation faults & application crashes

#6 2015-07-29 11:10:31

Re: Celeron N2940 Segmentation faults & application crashes

#7 2015-07-29 11:25:02

Re: Celeron N2940 Segmentation faults & application crashes

#8 2015-07-29 12:39:37

Re: Celeron N2940 Segmentation faults & application crashes

#9 2015-07-29 14:37:26

Re: Celeron N2940 Segmentation faults & application crashes

#10 2015-07-30 06:11:52

Re: Celeron N2940 Segmentation faults & application crashes

Board footer