You are not logged in.
Pages: 1
Hi, I'm experiencing kernel panics randomly while kernel updates (like when mkinitcpio gets updated or nvidia driver)
it's third time it's happening and i have no idea why it's happening.
Last edited by HJavid010 (2024-03-04 21:06:52)
Offline
An illustration of the actual panic might be useful, is it
a) actually a panic or does the system just freeze somehow?
b) when the systemd daemon reloads?
c) only w/ the testing repo?
Offline
It's a kernel panic, system completely stops and caps lock flashes.(even no loges are not complete because system doesn't sync files to filesystem after panic and logs are 1 or 2 minutes before panic occurs.
Kernel panic have happends 5 times till know
3 times while updating kernel(i thought the problem was the update)
But know it happens right after the login!
First time i logged in and connected to bluetooth then suddenly a kernel panic occurred.
And now again right after login. Before the kde splash screen goes away.
I disabled kde-unstable and core-testing and downgraded before last kernel panic.
Offline
I have three guesses
First is a problem with bluetooth headphone i'm connecting becuase i'm pretty sure it was connected when kernel panics occured(but it has a very low probability)
Second is a hardware issuse and now i'm running memtest86+
Third is a problem with graphics driver(probably nvidia)
Offline
Last resort: https://wiki.archlinux.org/title/Kdump
How long is this been going on?
There seems to be a severe memory management bug in recent kernels, see eg. https://bbs.archlinux.org/viewtopic.php?id=293400 and try to add "numa=off" to the kernel parameters.
nb. that meaningful memtest86+ checks are measured in days, at least run 16h or until something shows up.
Offline
Thanks
i don't think it's a hardware problem.
first occurrence was 3 days ago.
Offline
Last resort: https://wiki.archlinux.org/title/Kdump
How long is this been going on?
There seems to be a severe memory management bug in recent kernels, see eg. https://bbs.archlinux.org/viewtopic.php?id=293400 and try to add "numa=off" to the kernel parameters.nb. that meaningful memtest86+ checks are measured in days, at least run 16h or until something shows up.
Does this problem affect lts kernels? because the last one happend with lts kernel
Offline
Try the parameter.
Edit: potentially, the commit message cites a 6.6.0 oops, but it seems to have flared up only very recently
https://github.com/torvalds/linux/commi … bdabc4f974
Last edited by seth (2024-03-05 19:11:22)
Offline
Pages: 1