Random programs crashing

Crimsonknave · 2010-03-07 18:51:31

So, short version, computer works fine for 5-30 min or so then programs start to segfault on and off.

Long version:
I've got an up to date arch64 install with raid 5, lvm, nvidia drivers and oss. About a two weeks ago when I couldn't load WoW via wine, it told me various files were corrupted, none of the Blizzard fixing stuff was working and eventually I rebooted to do a memory test which came back clean. I hadn't really fixed anything and magically it worked and after the reboot my raid arrays were resyncing. Since I did a clean shutdown I assumed that somehow they got unsynced and that was part of my issues. And when the issues didn't recur I didn't really think much of it.
A couple of days ago I did a full upgrade and added a new video card, things worked fine for two days. Then when I come back from work I'm sitting at gdm (side note, where the hell did kdm go? I can't seem to find the program or package for it.). I try to log in and things are crashing (can't start up compiz) so I go to do a shutdown and from a virtual terminal shutdown starts and then crashes and spews debug info, but after it killed control to virtual terms, so I do a hard restart and things seem to be fine. I start up WoW later and after a few minutes it crashes like it did the week before. I try and see if I'm getting any errors, don't see any blatant ones and just restart my computer since, hey it worked last time and I didn't want to deal with it then. Well, I can get back into WoW and it works for like 5 minutes then it starts to crash with the same errors, but I can start it up and log back in fine, it just does it after a little while. Then skype crashes while WoW still works, so I restart skype and it won't run complains libxxx-xxxxx-xxx.so isn't there or something like that, the computer keeps locking up and I restart again (can't recall if it was a nice one) and when I'm booting up fsck thinks my main partition (jfs) needs checking which it probably does. How ever it is 4TB and takes forever to check so I go away and when I come back I'm at the log in screen and I can't find any logs about the fsck, so I can only assume it finished alright. Well, this morning I wake up and find that the whole computer is frozen, no signal to my monitors and nothing is giving responses, so I restart and once again things seem to be working I'm digging around seeing if I can find anything that might be meaningful to me; I can't. But I throw up a tail on /var/log/messages.log so I can see if anything wonky seems to be happening. After a bit I see:

Mar  7 13:09:23 shiny kernel: bash[5607]: segfault at 10175fd38 ip 00000000004494a9 sp 00007fff7a292740 error 4 in bash[400000+a6000]
Mar  7 13:09:26 shiny kernel: gnome-terminal[5611]: segfault at 101f64e90 ip 00007f660f81d11b sp 00007fffcfe609c8 error 4 in libX11.so.6.3.0[7f660f7c9000+133000]
Mar  7 13:09:30 shiny kernel: nvidia-settings[5618]: segfault at 7f4aaf5ab500 ip 00007f48b44a7df1 sp 00007fff234410a8 error 6 in libc-2.11.1.so[7f48b4428000+14d000]
Mar  7 13:10:12 shiny kernel: mplayer[5655]: segfault at 1000ca8510 ip 0000000000522c9b sp 00007fff17c0e510 error 4 in mplayer[400000+b87000]
Mar  7 13:10:20 shiny kernel: vlc[5661]: segfault at 400000000 ip 00007fb35020c912 sp 00007fff4b1fbb80 error 4 in ld-2.11.1.so[7fb350207000+1e000]
Mar  7 13:10:25 shiny kernel: nvidia-settings[5667]: segfault at 7f11cf857240 ip 00007f11cf857240 sp 00007fff5702ffd8 error 14
Mar  7 13:10:25 shiny kernel: xmessage[5669]: segfault at 49e9498 ip 00007fd92c03997e sp 00000000009e92b0 error 6 in libXaw7.so.7.0.0[7fd92c011000+64000]

Thing is that after vlc crashed I was able to use mplayer, but the sound seemed to have gone. After that I rebooted and started this post. Any suggestions or ideas on what is going on would be awesome.

Also, what other information would be needed/helpful to figure this out?

JohnVV · 2010-03-07 21:00:25

and added a new video card,

is this a nvidia card ?
and if so are you using the 196 driver?
it is buggy and can kill hardware

Crimsonknave · 2010-03-07 21:11:50

It is an nvidia card: GTS 250
pacman -Qs nvidia gives me nvidia 190.53-1.

I don't think it is just the video card because I've seen stuff like pacman segfault when I'm not running an x server. It seems like a memory or hard drive issue to me, but I'm not sure what else I can try to pinpoint it.

extofme · 2010-03-07 22:17:18

run the system thru some integrity tests, like memtest(memory)/mprime(cpu)/???(harddrive) and friends. i have similar problems (mostly locks tho) when stabilizing overclocks, so maybe some cpu issues. other than that, if every hardware check passes, try eliminating drivers/running minimal.

maybe use a livecd of a different distro to do everything, until you know it's not an Arch package/related.

Last edited by extofme (2010-03-07 23:08:08)

antonikus · 2010-03-07 23:18:17

as above extofme said ^
run memtest from a livecd for atleast half an hour, if anything is red, well... though luck, but then look at FAQ of memtest.

also run some harddrive test for bad sectors

Crimsonknave · 2010-03-08 00:17:11

I ran a memtest two weeks ago for at least 4 hours and there were no errors. Doing a long SMART test on each of my drives, though two of them have stopped reporting % remaining at 10% intervals while the others continued to. Assuming nothing else comes up from the HDD tests I'll let a memtest run over night and see if anything is different.

Arch Linux

#1 2010-03-07 18:51:31

Random programs crashing

#2 2010-03-07 21:00:25

Re: Random programs crashing

#3 2010-03-07 21:11:50

Re: Random programs crashing

#4 2010-03-07 22:17:18

Re: Random programs crashing

#5 2010-03-07 23:18:17

Re: Random programs crashing

#6 2010-03-08 00:17:11

Re: Random programs crashing

Board footer