APIC error on CPU1: 00(02) && APIC error on CPU0: 00(02)

APIC error on CPU1: 00(02)
APIC error on CPU0: 00(02)

I've spent some time trying to research these error messages, but I'm not exactly sure what I'm looking at. I've read that these may indicate a faulty or soon to go faulty processor. However, when I do get these errors, it always involves both processors. What are the odds both would be dying at the same time?

Please CC, thanks!
Steve

Is this a Dual AMD Platform? If yes, are this real MPs or just modded XPs? I've had this kind of error for years on a dual AMD with modded XPs on a Tyan 2460 - until now, no processor broke down. The APIC is very picky on this board, judging different posts from the internet.

Jan

They are genuine AMD Athlon 1900 MPs (never over-clocked) on a Giga-byte 7DPXDW-P. The only time these errors seem to occur is when the system is under heavy disk activity. Could there be a bug with the APIC timer? This motherboard has two ide-controllers on board. I've noticed in the log:

Enabling APIC mode: Flat. Using 1 I/O APICs

Is this correct, or should it be 2 I/O APICs?

I'm really fishing for answers. I'd like to know if there really is something wrong with my system (although it still seems rock stable) or is this possible a kernel bug?

I'm currently using 2.6.7-rc3 compiled with gcc 3.4.0.

Thanks,
Steve

-----Original Message-----
From: Jan Dittmer [mailto:[email protected]]
Sent: Thursday, June 10, 2004 1:10 PM
To: Steve Lee
Cc: [email protected]
Subject: Re: APIC error on CPU1: 00(02) && APIC error on CPU0: 00(02)

Is this a Dual AMD Platform? If yes, are this real MPs or just modded XPs? I've had this kind of error for years on a dual AMD with modded XPs

on a Tyan 2460 - until now, no processor broke down. The APIC is very picky on this board, judging different posts from the internet.

Jan

02 == receive checksum error on the local APIC bus. This tells us that you have an MP board with P6 or AMD processors, and that the system has hardware problems. Likely problems: bad board design (BP6 anyone?), inadequate cooling, inadequate power supply, overclocked processors, or using processors not validated for MP.