How to get kernel call stack trace in Balena OS following a kernel crash?

Product support balenaOS
1

I am trying to load some out-of-tree kernel modules on balena OS 2.108.26 (recommended for Google Coral) and my application(k3s agent + calico CNI) crashes the host kernel on startup with the nf_conntrack_netlink kernel module enabled. Do you guys know of a way of getting a detailed call trace from the crash that can maybe tell me what kernel call caused this? I’m not sure if anyone has ever attempted starting kdump on balenaOS or anything similar.

dmesg -w log at kernel panic time:

[  256.688073] Unable to handle kernel NULL pointer dereference at virtual address 00000040
[  256.704251] Mem abort info:
[  256.709945]   Exception class = DABT (current EL), IL = 32 bits
[  256.721880]   SET = 0, FnV = 0
[  256.728142]   EA = 0, S1PTW = 0
[  256.734520] Data abort info:
[  256.740377]   ISV = 0, ISS = 0x00000006
[  256.748146]   CM = 0, WnR = 0
[  256.754179] user pgtable: 4k pages, 48-bit VAs, pgd = ffff80001f9f3000
[  256.767329] [0000000000000040] *pgd=000000005f9f8003, *pud=000000005fa76003, *pmd=0000000000000000
[  256.785345] Internal error: Oops: 96000006 [#1] PREEMPT SMP

The issue I am having is described in more detail in my other post

2

Hi,

I can’t help with your issue directly, but I’m curious was the initial install successful? I have the 4GB board and couldn’t get a build after v2.48.0+rev5 to load successfully, even from a fresh install. If this version is a fresh addition, and working, I’ll give it another try.

3

We should look at updating the BSP for that board (and implicitly the kernel version). Could you create an issue about updating the BSP for the Coral board here Issues · balena-os/balena-coral · GitHub? So we can keep track of it
Thank you in advance

4

Hi floion,

That would be much appreciated, any idea on the timeline of such an update, we were hoping to enroll these devices into our edge cloud by end of may ?

I will go ahead and create a request for the BSP and kernel update