On Fri, 19 Oct 2018, Tracy Bales wrote:
> I have an AMD FX-8300 CPU.
There's been exactly one dmesg submitted for that CPU, but it was from
back in 2016 with OpenBSD 6.0, so it doesn't really tell us much about why
your machine is unhappy now.
Is this a new box for you, or has it successfully run _any_ previous
versions of OpenBSD?
> I can boot and finish the OpenBSD 6.4 AMD64 bsd.rd installation dated
> October 11, 2018 with no problems. However, on reboot the machine with the
> bsd.mp image gets stuck in a reboot loop as soon as the first line of the
> bsd image is displayed. I have the same result with bsd.sp as well.
So: bsd.rd is fine, but bsd.mp and bsd.sp both fail early. By "first
line" do you mean the "OpenBSD 6.4" line, or the number+number+number+...
line before that?
> On the same machine I can boot and finish the OpenBSD 6.4 i386
> installation dated October 11,2018 with no problems. However, on reboot
> the machine gives me the following error:
>
> kernel: protection fault trap, code=0
> Stopped at _kernel_lock+0x45: pushfl
> ddb{0}> trace
> _kernel_lock(d0adf750,2af) at _kernel_lock+0x45
> proc_trampoline_mp(fe7000,fdb000,0,d020053a,0) at proc_trampoline_mp+0x7d
> ddb{0}>
>
> Any suggestions would be appreciated! Thanks...
Running sendbug as root on that (i386) install to capture the dmesg,
pcidump(8), and acpidump(8) output, and submitting that would help. In
addition, when an MP install panics, you should follow the "Note for SMP
systems" section at https://www.openbsd.org/ddb.html and capture the
backtraces from _all_ the CPUs.
Philip Guenther
