Getting kernel crash dump

What’s the easiest way to get kernel crash dumps on EOS?

kdump on archwiki assumes the system is using mkinitcpio and initramfs/vmlinuz in /boot. Same with simple-kdump. kdumpts assumes the system uses grub.

I’m trying to debug whats causing a fatal kernel panic when resuming from suspend, interrupt error. This started after some update. My hunch is either GPU drivers or kernel, but this is just guessing. The problem is a bit sporadic, with uptime seemingly increasing chance of it appearing. Thus I’d rather look at the memory dump directly / bisect kernel rather than guess where the problem is. dmesg/journalctl do not show anything related to this.

I’m more used to F around memory dumps on Gentoo using classical /boot etc, never had to run these on EOS :sweat_smile:

System is EOS base install with systemd-boot. Dracut running on default settings. Beyond that, the system is quite heavily customized. HW tested OK. I just need that kernel dump.

Kernel configs and crashkernel= required are set.

What is the hardware? You could post the url.

inxi -Faz | eos-sendlog

https://0x0.st/8Ebk.txt

Booted on default kernel as a test. Normally I run bore scheduler instead of eevdf with a custom kernel. Also I’m aware theres newer BIOS available.

Solved this by converting the system to mkinitcpio and grub, then using kdumpst.

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.