FS#55816 - Kernel panic: BUG: unable to handle kernel paging request / acpi_safe_halt
Attached to Project:
Arch Linux
Opened by Jakub Okoński (farnoy) - Saturday, 30 September 2017, 16:27 GMT
Last edited by Jan de Groot (JGC) - Friday, 06 October 2017, 11:10 GMT
Opened by Jakub Okoński (farnoy) - Saturday, 30 September 2017, 16:27 GMT
Last edited by Jan de Groot (JGC) - Friday, 06 October 2017, 11:10 GMT
|
Details
Description:
My system freezes, sometimes after a couple of hours, sometimes after 20+ hrs of uptime. Graphics output is frozen, I can no longer ssh into my machine etc. I was able to get an output from the kernel by using netconsole and sending it to another computer. Additional info: * package version(s) core/linux 4.13-3-1 * config and/or log files etc. I attached the dmesg output that was sent through netconsole, it shows the kernel panic details I am using an AMD Ryzen Threadripper 1950X on a X399 Taichi motherboard, I have heard of such problems before and already tried disabling a couple of options in BIOS: "Global C-state Control", SMT, OpCache. For what it's worth, I also get blue screens on Windows 10 with messages like DRIVER_IRQL_LESS_THAN_OR_EQUAL_TO or page fault in non paged area. Steps to reproduce: I don't have any reliable way of reproducing it, it just happens after some time. 
kernel-log
(4.5 KiB)
|
This task depends upon
I had systems in the past that would run memtest86 for hours without issues, but as soon as I started compiling it would crash: the chipset couldn't handle 3 memory sticks without becoming unstable.
Same with another system, memtest86 passes fine, but as soon as video drivers were upgraded, the system would freeze after rendering 2 frames in 3D. Replacing memory solved that issue.
Anyways, the trace you attached, is it always the same type of trace, or is it random? You could try to install linux-lts to see if an older kernel is also affected.