Container segfaults ending up in host's kernel log?

LXD Support
1

How normal is that container segfault messages end up in host’s dmesg?

alex@alexsv:~$ lxc launch ubuntu:22.04 segfaulter
alex@alexsv:~$ lxc exec segfaulter -- bash -i
root@segfaulter# apt -y install gcc
...
root@segfaulter:~# gcc -o segfaulter -x c -
void main() {
    int *pointer;
    *pointer = 1;
}^D
root@segfaulter:~# ./segfaulter 
Segmentation fault (core dumped)
^D
alex@alexsv:~$ sudo dmesg | tail -2
[2264660.993706] segfaulter[197529]: segfault at 560ebc28d040 ip 0000560ebc28d135 sp 00007ffe5e5915d0 error 7 in segfaulter[560ebc28d000+1000]
[2264660.993722] Code: ff c6 05 fd 2e 00 00 01 5d c3 0f 1f 00 c3 0f 1f 80 00 00 00 00 f3 0f 1e fa e9 77 ff ff ff f3 0f 1e fa 55 48 89 e5 48 8b 45 f8 <c7> 00 01 00 00 00 90 5d c3 00 00 f3 0f 1e fa 48 83 ec 08 48 83 c4
2

I would expect that this is normal inside a container since it is using the host’s kernel. The same would be true if you were running a docker container or similar.

3

Yes this is because printk in the kernel isn’t namespace aware yet.