Re: Illegal instruction errors under XEN3_DOM0 on -CURRENT

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Sun, Apr 19, 2020 at 08:23:30PM -0500, Dustin Marquess wrote:
> I've been playing with a PC Engines apu4d4 box.  Installed -CURRENT
> from the nycdn builds, and everything works great.
> 
> I installed the xenkernel411 and xentools411 packages, copied xen.gz
> to /, and added to boot.cfg:
> 
> menu=Xen:load /netbsd-XEN3_DOM0.gz console=pc;multiboot /xen.gz
> dom0_mem=512M console=com1 com1=115200,8n1 dom0_max_vcpus=1
> 
> The Xen kernel boots fine and then the NetBSD dom0 starts.  However,
> just about everything illegal instruction dumps:
> 
> [   3.5201899] kern.module.path=/stand/amd64-xen/9.99.56/modules
> Mon Apr 20 01:15:00 UTC 2020
> [1]   Illegal instruction     /bin/sh -c "ps -p \$\$ -o ppid="
> ...
> Starting dhcpcd.
> dhcpcd_fork_cb: truncated read 0 (expected 4)
> /etc/rc.d/dhcpcd exited with code 1
> Building databases: dev[1]   Illegal instruction (core dumped) "${@}"
> , dev[1]   Illegal instruction (core dumped) "${@}"
> ...
> Mounting all file systems...
> mount: /var/shm: Illegal instruction
> /etc/rc.d/mountall exited with code 1
> ...
> Starting ntpd.
> [1]   Illegal instruction (core dumped) RC_PID= _rc_pid=
> _rc_original_stdout_fd= _rc_o...
> /etc/rc.d/ntpd exited with code 1
> ...
> Starting sshd.
> [1]   Illegal instruction (core dumped) RC_PID= _rc_pid=
> _rc_original_stdout_fd= _rc_o...
> /etc/rc.d/sshd exited with code 1
> [1]   Illegal instruction (core dumped) awk "/^sendmail[
> \t]/{print\$2}" /etc/mailer.conf
> Apr 20 01:15:15 apu4 init: fatal signal: Illegal instruction
> [  48.1203201] panic: init died (signal 0, exit 4)
> [  48.1203201] cpu0: Begin traceback...
> [  48.1203201] vpanic() at netbsd:vpanic+0x146
> [  48.1203201] snprintf() at netbsd:snprintf
> [  48.1203201] exit1() at netbsd:exit1+0x50b
> [  48.1203201] sys_exit() at netbsd:sys_exit+0x3d
> [  48.1301288] syscall() at netbsd:syscall+0x9c
> [  48.1301288] --- syscall (number 1) ---
> 
> However booting the GENERIC kernel again works just fine.  This is
> from the same build, although I've tried DOM0 kernels from the last
> few days from nycdn.
> 
> Any idea?

It could help to know what intruction is causing it.
If you have core dumps, you should be able to see it with gdb.

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--