Hello all, happy to be here!
I'm coming from a backend web (technician) formation, but UNIX systems
interest me more. So, I have gotten myself an IBM POWER8 machine to
introduce myself into this new interesting world. I'm not an expert,
but I'm planing to spend some time to giveback to openbsd :)
So, I have an IBM S822L (8247-22L) and I have tried to boot
openbsd/powerpc64 in opal mode and it does not boot. See below for
more information and the possible cause.
system and firmware information:
petitboot : v1.4.2-e1658ec
skiroot : 4.4.70-openpower1
mode : opal
processor : dual 10 core POWER8
ram : 128gb
model : IBM S822L (8247-22L)
mode : OPAL
petitboot menu
------------------------------------------------------------------------
Petitboot (v1.4.2-e1658ec) 8247-22L 213FB3A
───────────────────────────────────────────────────────────────────────
[USB: sda1 / E820-1608]
OpenBSD install
System information
System configuration
System status log
Language
Rescan devices
Retrieve config from URL
*Exit to shell
------------------------------------------------------------------------
when I telnet into the skiroot's shell I see this before the power8 reset's
itself:
skiboot output
------------------------------------------------------------------------
Exiting petitboot. Type 'exit' to return.
You may run 'pb-sos' to gather diagnostic data
/ # uname -a
uname -a
Linux skiroot 4.4.70-openpower1 #1 SMP Wed Jun 14 08:57:46 UTC 2017
ppc64le GNU/Linux
/ #
The system is going down NOW!
Sent SIGTERM to all processes
Sent SIGKILL to all processes
[ 108.051970] kexec_core: Starting new kernel
Connection closed by foreign host.
------------------------------------------------------------------------
After some online searches, I found that this seems to be related to
a kexec-lite regression found by Brandon Bergren from freebsd
(bdra...@freebsd.org).
issue documented on freebsd:
https://www.freebsd.org/platforms/ppc/#issues
> As of 2021-11-30 the firmware of IBM POWER9 and some IBM POWER8
> machines contain a version of kexec-lite with a regression that
> computes FreeBSD kernel entry point erroneously
Upon doing the same test with freebsd, the same behavior happens.
Skiroot outputs the same thing and my power8 reset's itself.
So, it seems as if OpenBSD is facing the same issue.
This seems to be fixed here:
https://github.com/antonblanchard/kexec-lite/pull/9
I have yet to try the patched kexec-lite binary and to update the
firmware on my power8 system if there are any updates available.
I will check this when I will have some more time.
Notice: my power8 system firmware may not be up to date, I will have
to check the logs on the service processor and see when was the last
time that its firmware was updated. If I recall well, it was sometime
in 2021...
Cheers!
