APIC error on CPU0

APIC error on CPU0

[Kristian]

Hello.

Since I converted one box to Debian 3.0 I'm getting strange errors related to APM.
When I push the power button (or do "apm -s") the box suspends normally. But it comes back almost immediately with this error:

APIC error on CPU0: 00(40)

I grepped the kernel source and found that "40" means reserved APIC error. What does that mean exactly ? BTW: I enabled "local APIC on uniprocessors" but this behaviour also appears without enabling that option.

Strangely this error won't appear with Redhat 7.2 and earlier versions. (I've never had any problems.)

$ dmesg:

Local APIC disabled by BIOS -- reenabling.
Found and enabled local APIC!
Kernel command line: auto BOOT_IMAGE=linux ro root=305 console=ttyS0,57600n8,vt1
02 console=tty0
Initializing CPU#0
Detected 349.187 MHz processor.
Console: colour VGA+ 80x25
Calibrating delay loop... 696.32 BogoMIPS
Memory: 61808k/65536k available (1354k kernel code, 3340k reserved, 480k data, 2
32k init, 0k highmem)
[...]

$ lspci:

00:00.0 Host bridge: Intel Corp. 440BX/ZX - 82443BX/ZX Host bridge (rev 02)
00:01.0 PCI bridge: Intel Corp. 440BX/ZX - 82443BX/ZX AGP bridge (rev 02)
00:0e.0 Ethernet controller: 3Com Corporation 3c905B 100BaseTX [Cyclone]
00:0f.0 Token ring network controller: IBM 16/4 Token ring UTP/STP controller (rev 05)
00:14.0 ISA bridge: Intel Corp. 82371AB PIIX4 ISA (rev 02)
00:14.1 IDE interface: Intel Corp. 82371AB PIIX4 IDE (rev 01)
00:14.2 USB Controller: Intel Corp. 82371AB PIIX4 USB (rev 01)
00:14.3 Bridge: Intel Corp. 82371AB PIIX4 ACPI (rev 02)
01:00.0 VGA compatible controller: Matrox Graphics, Inc. MGA G200 AGP (rev 01)

BTW: I already tried the specified list for apmd but it seems to be down since years..

*Kristian

  :... [snd.science] ...:
 ::
 :: http://www.korseby.net
 :: http://gsmp.sf.net
  :.........................:: ~/$ kristian@korseby.net :

Re: APIC error on CPU0

[Zwane Mwaikambo]

On Mon, 11 Feb 2002, Kristian wrote:

> Hello.
> 
> Since I converted one box to Debian 3.0 I'm getting strange errors related to APM.
> When I push the power button (or do "apm -s") the box suspends normally. But it comes back almost immediately with this error:
> 
> APIC error on CPU0: 00(40)
> 
> I grepped the kernel source and found that "40" means reserved APIC error. What does that mean exactly ? BTW: I enabled "local APIC on uniprocessors" but this behaviour also appears without enabling that option.
> 
> Strangely this error won't appear with Redhat 7.2 and earlier versions. (I've never had any problems.)

Mikael Pettersson submitted a patch which fixed that for me, its 
definately in 2.4 mainline. Which kernel is in Debian 3.0?

Regards,
	Zwane Mwaikambo

Re: APIC error on CPU0

[Kristian]

Zwane Mwaikambo <zwane@linux.realnet.co.sz> wrote:
>
> Mikael Pettersson submitted a patch which fixed that for me, its 
> definately in 2.4 mainline. Which kernel is in Debian 3.0?
> 

apm is disabled per default on Debian's default kernel config. I compiled 2.4.17 myself which is running now.

*Kristian

  :... [snd.science] ...:
 ::
 :: http://www.korseby.net
 :: http://gsmp.sf.net
  :.........................:: ~/$ kristian@korseby.net :

Re: APIC error on CPU0

[Kristian]

Zwane Mwaikambo <zwane@linux.realnet.co.sz> wrote:
> 
> Mikael Pettersson submitted a patch which fixed that for me, its 
> definately in 2.4 mainline. Which kernel is in Debian 3.0?

Hm... That's strange. When I disable serial console on ttyS0 (in inittab) linux will suspend correctly. Although this APIC error on CPU0: 00(40) appears though. Also my one and only fan in my power supply spins down completely. So the box will get quite hot after some minutes in standby mode... Maybe a hardware problem...

Thanks anyway. ;-)

*Kristian

  :... [snd.science] ...:
 ::
 :: http://www.korseby.net
 :: http://gsmp.sf.net
  :.........................:: ~/$ kristian@korseby.net :

APIC error on CPU0

[Michael Buesch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

What does this kernel message mean?

Apr  5 23:16:20 lfs kernel: APIC error on CPU0: 60(60)
Apr  5 23:16:31 lfs kernel: APIC error on CPU0: 60(60)

kernel is 2.6.5 with reiser4 patch.
I can post .config if needed, but I don't want to abuse
the list, if nobody is interrested in my configuration.

Thanks.

- -- 
Regards Michael Buesch  [ http://www.tuxsoft.de.vu ]

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAcdFWFGK1OIvVOP4RAg9JAJkBImV4TpaExUhg/AfATSzkENKxLwCdHU5a
fY7cKIaNDVYbRasKIgQODX8=
=mLkL
-----END PGP SIGNATURE-----

Re: APIC error on CPU0

[Steve Youngs]

[-- Attachment #1: Type: text/plain, Size: 598 bytes --]

* Michael Buesch <mbuesch@freenet.de> writes:

  > What does this kernel message mean?

  > Apr  5 23:16:20 lfs kernel: APIC error on CPU0: 60(60)
  > Apr  5 23:16:31 lfs kernel: APIC error on CPU0: 60(60)

My logs are filling up with these as well, although I'm getting
40(40), and the very occasional 00(40).

I'm wondering what it means as well.

-- 
|---<Steve Youngs>---------------<GnuPG KeyID: A94B3003>---|
|              Ashes to ashes, dust to dust.               |
|      The proof of the pudding, is under the crust.       |
|------------------------------<sryoungs@bigpond.net.au>---|

[-- Attachment #2: Type: application/pgp-signature, Size: 256 bytes --]

Re: APIC error on CPU0

[Mikael Pettersson]

On Mon, 5 Apr 2004 23:35:59 +0200, Michael Buesch wrote:
>What does this kernel message mean?
>
>Apr  5 23:16:20 lfs kernel: APIC error on CPU0: 60(60)
>Apr  5 23:16:31 lfs kernel: APIC error on CPU0: 60(60)
>
>kernel is 2.6.5 with reiser4 patch.

Send Illegal Vector and Receive Illegal Vector at the same time.
This is more interesting than the usual 40(40) errors
(just Receive Illegal Vector), since 60 implies that the
CPU itself is the source of the bogus vector, whereas 40
usually implies a crap mainboard.

My guess is that either your hardware (whatever it is,
you didn't leave any clues) has problems with a noisy
APIC bus, broken chipset, or something like that, or
you enabled ACPI and the ACPI tables are crap.

In any event, Linux is probably not the source of the
problem. You may need to run without I/O APIC ("noapic"
kernel param), no ACPI-based PCI routing ("pci=noapci"),
or completely without ACPI ("acpi=off").

Re: APIC error on CPU0

[Michael Buesch]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 06 April 2004 02:57, you wrote:
> On Mon, 5 Apr 2004 23:35:59 +0200, Michael Buesch wrote:
> >What does this kernel message mean?
> >
> >Apr  5 23:16:20 lfs kernel: APIC error on CPU0: 60(60)
> >Apr  5 23:16:31 lfs kernel: APIC error on CPU0: 60(60)
> >
> >kernel is 2.6.5 with reiser4 patch.
> 
> Send Illegal Vector and Receive Illegal Vector at the same time.
> This is more interesting than the usual 40(40) errors
> (just Receive Illegal Vector), since 60 implies that the
> CPU itself is the source of the bogus vector, whereas 40
> usually implies a crap mainboard.
> 
> My guess is that either your hardware (whatever it is,
> you didn't leave any clues) has problems with a noisy
> APIC bus, broken chipset, or something like that, or
> you enabled ACPI and the ACPI tables are crap.
> 
> In any event, Linux is probably not the source of the
> problem. You may need to run without I/O APIC ("noapic"
> kernel param), no ACPI-based PCI routing ("pci=noapci"),
> or completely without ACPI ("acpi=off").

Ok, thanks for this long explanation.
I think it's possible, that the reason is bad hardware.

- -- 
Regards Michael Buesch  [ http://www.tuxsoft.de.vu ]

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAcsdTFGK1OIvVOP4RAji5AJ9G4ms5U3ktWNDV0zih30XnPAMwxQCgqkxI
5bDE3Tki3a+BSSyWS1kw9+Q=
=Oqt5
-----END PGP SIGNATURE-----