RE: [PATCH 6/8] x86, topology.c: Enable CPU0 online/offline

[Yu, Fenghua]

> >> And you forgot poweroff and reboot, which have similar dependencies
> on
> >> some machines. That whole low level ACPI stuff is sensitive.
> >
> > I tested poweroff, shutdown, and reboot with various reboot_type on a
> few different platforms. I haven't seen poweroff and reboot issues
> after CPU0 is offline.
> >
> > Do you have specific platforms that I can test poweroff and reboot
> dependency on CPU0?
> 
> > Or we just assume there are some platforms out there that depend on
> CPU0 for poweroff/reboot?
> 
> A classic Linux/ACPI failure happened when HT first shipped.
> Some platforms stopped powering off or rebooting 50% of the time.
> 
> It turned out that SMM on those machines assumed
> it would be triggered from CPU0 and not CPU1.
> 
> the original code should have worked, of course,
> and most of the time it did -- but some systems broke.
> The same will probably happen here.

Thanks for your info on poweroff/reboot and CPU0.

I wrote a statement in the cpu-hotplug.txt doc in newer patch set:

"It's said poweroff/reboot may depend on BSP on some machines although I haven't
seen any poweroff/reboot failure so far after BSP is offline on a few tested
machines."
"Please let me know if you know or see any other dependencies of BSP."

If we know a specific model of machine can't poweroff/reboot after CPU0 is offline, I can do something like put that machine in a black list in the poweroff/reboot path to let user online CPU0 before poweroff/reboot.

Hopefully this is a temp OK solution for the patch set to get in kernel and let people report any poweroff/reboot or any other dependencies on CPU0.

Thanks.

-Fenghua
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/