Solved .xsession-errors & computer shutdown

My FreeBSD box keeps shutting down after 20 minutes or less of ontime. Xorg starts just fine but I keep finding .xsession-errors and these weird files that I cant figure out the meaning of like .xsmU97EBA and .prxlzzacC. Are the two issues related? What can I do to get rid of these log files?, and keep my machine running. If the two aren't related then what could be the problem?
 
What .xsession-errors? You gotta give us something.. Is there something relevant in dmesg? Or /var/log/Xorg.0.log? Is it just shutting down, or rebooting (after a kernel panic)?
 
What .xsession-errors? You gotta give us something.. Is there something relevant in dmesg? Or /var/log/Xorg.0.log? Is it just shutting down, or rebooting (after a kernel panic)?
It's just shutting down, I think it might be overheating. No, there's nothing relevant in Xorg.0.log or dmesg. It does this every time I run poudriere testport.
 
There is the possibility of OOM killing your desktop, if you switch to a console (X11 not logged in) and run testport does it still happen. If OOm happens you ought to see logs about which process was killed. If by shutdown you mean entire machine was turned off then could be overheat maybe try enabling powerd and there should be some way to read the temperature sensors (not sure about this, I think there are different kernel modules for Intel/AMD and maybe for different motherboards)
 
It's just shutting down, I think it might be overheating. No, there's nothing relevant in Xorg.0.log or dmesg. It does this every time I run poudriere testport.
Yeah, that sounds like a heat problem. Building ports (and compiling in general) can push a CPU quite hard. Thermal protection will do the rest. Have you cleared out the dust bunnies yet? Checked to make sure your CPU fan(s) are still running? Is it a desktop PC or a laptop? Desktops are easier to service, laptops may need to be taken apart.
 
There is the possibility of OOM killing your desktop, if you switch to a console (X11 not logged in) and run testport does it still happen. If OOm happens you ought to see logs about which process was killed. If by shutdown you mean entire machine was turned off then could be overheat maybe try enabling powerd and there should be some way to read the temperature sensors (not sure about this, I think there are different kernel modules for Intel/AMD and maybe for different motherboards)
What is OOm?

If you mean virtual terminal or command line I will try testport on that... NVM I can't alt+cntl f1 a terminal so I don't know I can't check.
 
OOM = Out Of Memory at this point the kernel will normally abort the process. One of my programs hit this
all the time iuntil I figured out that the way Go allocates/frees memory and FreeBSD's multiple arenas work
I would get an effective malloc of largest buffer * total threads, so that program would use all memory and all swap
kill almost every other process, then crash. Fortunately sshd would survive so I was able to restart things until I figure a fix.
It sounded like something similar was happening on your system.
 
Back
Top