View previous topic :: View next topic |
Author |
Message |
xtx Tux's lil' helper
Joined: 27 Dec 2011 Posts: 129
|
Posted: Sun Feb 17, 2013 8:50 pm Post subject: [solved] Segmentation Fault |
|
|
Laptop: Asus UX31e
I had this laptop running perfectly for 8 months but the battery drained and when I booted back up the system was very unstable. I pegged this down to a botched btrfs partition. Decided to just reinstall with ext4. I noticed when reinstalling that many packages were failing with segmentation fault errors. So, for example, when installing kde-meta it would install maybe 5 to 15 packages and then error out. I was fixing this by changing makeopts to -j1 (from -j5) and installing the individual package that was failing, which worked for awhile. qt-webkit eventually failed even with -j1.
Found this stress test
Code: | #!/bin/bash
make dep
while [ "foo" = "foo" ]
do
make clean
make -j2 bzImage
if [ $? -ne 0 ]
then
echo OUCH OUCH OUCH OUCH
exit 1
fi
done |
from http://www.gentoo.org/doc/en/articles/hardware-stability-p1.xml
and ran it. Didn't take long to fail. Exact error is:
Code: | drivers/usb/host/xhci-hub.c:537:5: internal compiler error: Segmentation fault
Please submit a full bug report |
The website says if this fails it is most likely a CPU issue. I just want to get opinions before I send this in to Asus to get repaired... I don't want them to say there's nothing wrong and find out I did something stupid.
Also, how can I prevent this from happening again? Is the stress from gentoo too much for some laptops?
Last edited by xtx on Sat Mar 30, 2013 11:15 pm; edited 1 time in total |
|
Back to top |
|
|
Hu Moderator
Joined: 06 Mar 2007 Posts: 21633
|
Posted: Sun Feb 17, 2013 9:00 pm Post subject: |
|
|
That could be a CPU overheating issue. What is your idle operation temperature? What is the temperature when you observe failures?
Gentoo is not too much for any well built laptop, but some laptops ship with inadequate cooling. Running any CPU intensive workload on such a laptop can cause problems, including Windows games that spin instead of sleeping. |
|
Back to top |
|
|
xtx Tux's lil' helper
Joined: 27 Dec 2011 Posts: 129
|
Posted: Sun Feb 17, 2013 9:03 pm Post subject: |
|
|
Before I tried reinstalling my idle temp would be around 120 or so. When compiling it would show 192.2 and would never report a temp above that, although I have my suspicions. |
|
Back to top |
|
|
Hu Moderator
Joined: 06 Mar 2007 Posts: 21633
|
Posted: Sun Feb 17, 2013 9:58 pm Post subject: |
|
|
I will assume you meant 120F and 192.2F, since I am not aware of any consumer CPU that is designed to survive extended operation in excess of 100C. The latter reading, 192.2F (~89C) is well above where I would initiate an emergency shutdown. I dislike seeing my systems above ~70C. |
|
Back to top |
|
|
xtx Tux's lil' helper
Joined: 27 Dec 2011 Posts: 129
|
Posted: Sun Feb 17, 2013 10:18 pm Post subject: |
|
|
Yes, I did mean Fahrenheit. So I'll assume it's an overheating issue. What should I do about that? Simply initiating a shutdown isn't an option because then I'd never be able to update my system or install new programs... |
|
Back to top |
|
|
Jaglover Watchman
Joined: 29 May 2005 Posts: 8291 Location: Saint Amant, Acadiana
|
|
Back to top |
|
|
xtx Tux's lil' helper
Joined: 27 Dec 2011 Posts: 129
|
Posted: Mon Feb 18, 2013 12:52 am Post subject: |
|
|
It's an ultrabook, encased in aluminum like the mac air. I guess I could try to take it apart to dust...
I'm pretty sure it's run that high since day one when compiling. This is going off the temperature widget in kde btw. |
|
Back to top |
|
|
Hu Moderator
Joined: 06 Mar 2007 Posts: 21633
|
Posted: Mon Feb 18, 2013 1:51 am Post subject: |
|
|
If it has always been this prone to overheating, then the device is simply poorly designed. In my opinion, shipping a machine with components known to be able to easily generate heat well in excess of the ideal cooling capacity is either stupid or negligent. |
|
Back to top |
|
|
DirtyHairy l33t
Joined: 03 Jul 2006 Posts: 608 Location: Würzburg, Deutschland
|
Posted: Mon Feb 18, 2013 11:25 am Post subject: |
|
|
You could try running memtest86 on the machine for a couple of hours. If it fails, you can be pretty sure that you have a hardware flaw. Concerning temperatures: my old Thinkpad T60 (core duo) could well reach about 95 °C during merges, in particular before cleaning the cooling system, and this was still below the emergency trip point (I checked back then) |
|
Back to top |
|
|
Hamlet Apprentice
Joined: 15 Apr 2005 Posts: 151 Location: Turin (Italy)
|
Posted: Tue Feb 19, 2013 7:23 am Post subject: |
|
|
Another paragraph in the "faulty hardware" chapter.
I had a similar problem on my desktop: segmentation faults when compiling (sys-devel/gcc being the favourite crasher), sometimes just during normal usage.
Although I have no final evidence, I strongly suspect it was my motherboard not communicating well with the memory (I tried 3 different memory sets, one of them was even listed among the Blessed... waste of money).
memtest86 presented errors every now and then.
It was definitely an hardware problem. |
|
Back to top |
|
|
xtx Tux's lil' helper
Joined: 27 Dec 2011 Posts: 129
|
Posted: Sat Mar 30, 2013 10:46 pm Post subject: |
|
|
it was faulty hardware, i sent it back to asus and they fixed it for free since the laptop is under 1 year old. although i can't tell you what exactly the problem was... when i got the laptop back the sheet said to check the log.txt file located within C:\some random windows path, despite it not having windows on it.
temps were still definitely reaching 194*F when compiling, so i changed makeopts from -j5 to -j1. now the temp isn't going above 156*F. i'll just have to deal with slow compile times, but better that than risk burning this puppy up after warranty is over. |
|
Back to top |
|
|
|