Subject: Re: Segmentation faults with SMP
To: None <john@johnrshannon.com>
From: Lars Nordlund <lars.nordlund@hem.utfors.se>
List: port-i386
Date: 06/08/2007 23:25:11
On Thu, 19 Apr 2007 05:30:20 -0600
"John R. Shannon" <john@johnrshannon.com> wrote:

> Dual Intel Core 2 (Merom)
> CURRENT (as of this AM)
> dmesg attached
> 
> Segmentation faults, in different processes, occur every few minutes
> when the computer is building software. The same kernel, without options
>          MULTIPROCESSOR, works fine. Also, a 64-bit kernel does not
> display this behavior.
> 
> The core dumps don't reveal anything meaningful (to me):

> Suggestions on how to proceed?

Hello

Have you found a solution to this problem?

I have just managed to solve a mysterious problem I have been suffering
from on my boxed core2duo,

cpu0 at mainbus0 apid 0: (boot processor)
cpu0: Intel(R) Core(TM)2 CPU          6400  @ 2.13GHz, 2133.43 MHz
cpu0: features:
bffbfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR> cpu0:
features: bffbfbff<PGE,MCA,CMOV,PAT,PSE36,CFLUSH,B20,DS,ACPI,MMX> cpu0:
features: bffbfbff<FXSR,SSE,SSE2,SS,HTT,TM,SBF> cpu0: features2:
e3bd<SSE3,MONITOR,DS-CPL,VMX,EST,TM2,xTPR>

on which I have been running NetBSD/amd64 on. I have been experiencing
sudden power-off after days of building software all from the start
when I bought this machine. Especially when using pkg_comp over both
cores. But it was not so bad and did not happen too often.. Recently
this unstable behaviour got worse. Much worse. I could not even run X
without having the machine crash back to console or just hang hard.

I tried the memory modules separately.. Ran the machine headless
without graphics board, and so on. Nothing improved the situation. I
tried a non-SMP kernel and it improved the situation a bit. Perhaps
weak power supply, I thought.

So one day I upgraded my power supply from 350W to 450W by buying a new
case. I did see some improvement. But it did not solve all my problems.
The machine was still unstable during load.

Then I tried re-attaching the CPU heat sink to the motherboard. I did
this 4 or 5 times.. And now the machine is stable as a rock! Those darn
plastic fastenings..!

I have never seen high CPU temperaturess on this computer. I have
checked both envstat and the BIOS. A lost. I suspect that the sensored
area of the chip has had good connection to the heat sink, but not some
other part of the chip?

Well, perhaps both CPU heating problem and weak power supply. I do not
know. I am just happy to be able to use my desktop machine again.. :-)

Good luck with your machine!


Best regards,
	Lars Nordlund