Re: [Fwd: Dual Celery on BP6]

Keith Owens (kaos@ocs.com.au)
Fri, 17 Dec 1999 00:37:53 +1100

On Thu, 16 Dec 1999 18:56:33 +0800,
Jason Jordan <guru@swami.pcguru.com.au> wrote:
>I posted this on alt.comp.periphs.mainboard.abit where someone suggested
>I forward it to this list.
>
>Yesterday, hoping that the problem might be related to 2.2 kernels, I
>compiled 2.3.33 and booted.
>
>Apparently 2.3.33 starts logging APIC errors for the first time:
>
>Dec 15 23:24:10 swami kernel: APIC error interrupt on CPU#0, should
>never happen.
>Dec 15 23:24:10 swami kernel: ... APIC ESR0: 00000000
>Dec 15 23:24:10 swami kernel: ... APIC ESR1: 00000008
>Dec 15 23:24:10 swami kernel: ... bit 3: APIC Receive Accept Error.
>Dec 15 23:24:55 swami kernel: APIC error interrupt on CPU#1, should
>never happen.

You are not alone. Several BP6 users are reporting APIC error
interrupts, some with 0x02, some with 0x08. You see the messages in
2.3.x but not 2.2.x because the debugging messages were only added in
2.3.20. I suspect that if you ported the code for smp_error_interrupt
from 2.3.33 back to 2.2.x then you would see the same messages. Nobody
seems to know if this is a real hardware bug with the BP6 (if it is, it
is a common one) or if there is some error in programming the APIC.

"Maciej W. Rozycki" <macro@ds2.pg.gda.pl> mailed a patch to l-k earlier
today which changes the APIC timing, search recent l-k archives for
get_8254_timer_count. I plan to test that patch tommorow on 2.3.33,
see if it fixes the BP6 problems.

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