8.1-RC2 MCE caused by some LAPIC/clock changes?
John Baldwin
jhb at freebsd.org
Wed Jul 21 17:36:25 UTC 2010
On Wednesday, July 21, 2010 5:17:06 am Andriy Gapon wrote:
> on 21/07/2010 11:33 Andriy Gapon said the following:
> > Not sure how to interpret this properly.
> > One possibility is a hardware problem where interrupt message route
between
> > ioapic2 and CPU to which lapic3 belongs is flaky.
>
> Or, I/O path between that CPU and the PCI slot where the device resides. Or
the
> CPU. Or...
> I think that lapic2 and lapic3 reside in a different physical
package/socket,
> given that you have 2x2 CPU/core configuration.
>
> BTW, John, could there be any problem because of this:
> ioapic1: WARNING: intbase 48 != expected base 24
> ioapic2: WARNING: intbase 56 != expected base 55
> ioapic3: WARNING: intbase 24 != expected base 63
You can ignore this, this just means ACPI enumerated the I/O APICs out of
order in the MADT.
--
John Baldwin
More information about the freebsd-stable
mailing list