Created attachment 583600 [details] kernel crash log in /var/log/messages Description of problem: My system runs haldaemon.service daily invoked by systemd. This in turn causes a kernel/udevadm crash that is splashed on the console and logged on messages. I'll attach the crash report. First crash was around 3:30 am on 2012/05/08 after I rebooted with kernel-3.3.4-3.fc16.x86_64 Version-Release number of selected component (if applicable): kernel-3.3.4-3.fc16.x86_64 udev-173-3.fc16.x86_64 hal-0.5.14-6.fc15.x86_64 How reproducible: Happened every night for the last three nights, doesn't bring the system down. Steps to Reproduce: 1. No steps, just running the system Actual results: daily annoying crash report splashed on screen Expected results: no crash Additional info:
Created attachment 583601 [details] kernel crash log in /var/log/messages changing attachment type to text/plain
this isn't the first bug that was hit (Tainted:B indicates a 'bad page' was found earlier). Can you reproduce this without vbox ?
I don't think I ever fired vbox after reboot, only ran the setup for the new kernel, but do have the modules loaded. I'll make sure they are not loaded and reboot sometime today.
Still crashes without vbox. The "Tainted:B" is no longer there, that line containing "Not tainted". I run this kernel on another (very different) machine with no problems, but on this one I do have a Hauppage video capture card. I'll attach the new crash message.
Created attachment 583851 [details] crash log without vbox (not tainted kernel)
RAX: ffefffff81c8bca0 that 'e' should be an 'f'. You had a single bit flip from 1 to 0. This is usually a sign of a hardware problem such as bad memory.
That could be it, though I don't think anything else crashes on my system (other that Adobe's flash, but that's another story). I'll run memtest soon.
memtest did show up a bad memory (but interestingly no crashes with kernel 3.3.5-2.fc16.x86_64). I'll need to find my old 4 GB stick before RMAing the 3x4 now on the system :-(