Subject: Re: amap corrupt panic?
To: None <david@fundy.ca>
From: Robert Alexander Baxter <rbaxte03@harris.com>
List: netbsd-help
Date: 06/09/1999 08:09:26
David Maxwell <david@fundy.ca> wrote...

>
>As a further datapoint, the machine has paniced twice more,
>with different panics, uvm_fault, and pmap-something.
>
>cc1 also gets a lot of SEGVs. I'm going to try the hd in
>another machine to see if the problem moves or stays.
>

I had problems like that on a Pentium computer into which I
inserted a cache module.  No task which was more than basic
would finish without some fault, panic, etc.  For example,
installing, or compiling, as you've listed.

In my case, the cache module (not the main system RAM) was
at fault.  Sadly, it somehow damaged the on-board cache, and
now I must disable the entire L2 cache in BIOS SETUP for this
computer to work.

I made sure it was the right COAST revision, and everything,
but it sure caused trouble!

Anyway, those errors you've gotten really look like memory-
related errors.  I think there's even a paragraph somewhere
in the gcc documentation which talks about how lots of
segmentation faults are likely caused by your memory, not a
bug in gcc.

-Alex (:-)
 rbaxte03@harris.com