Date: Wed, 25 Feb 1998 17:49:00 -0800 (PST)
Message-Id: <199802260149.RAA14184@adit.ap.net>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: ao950 AT FreeNet DOT Carleton DOT CA (Paul Derbyshire), djgpp AT delorie DOT com
From: Nate Eldredge <eldredge AT ap DOT net>
Subject: Re: I'm going slightly mad!!
Precedence: bulk

At 10:42  2/25/1998 GMT, Paul Derbyshire wrote:
>
>Nate Eldredge (eldredge AT ap DOT net) writes:
>> Yes it is.
>> 
>> Weird, irreproducible problems like this are often the sign of some kind of
>> hardware failure. The first thing that comes to mind is bad memory. I'd
>> suggest testing it somehow (maybe someone will follow up and recommend a
>> tester), and replacing it if there is a problem. You might want to back up
>> your data in the meantime.
>
>A tester? He probably already has one, in his BIOS. Nearly all computers
>I've seen (especially with 32 megs of the stuff) do a memory test at
>bootup, of the RAM, CPU caches, and sometimes video memory also.

The BIOS testers are notoriously bad. Usually something simple like "fill
memory with DEADBEEF, then check if it's still there." And memory can fail
in 2^N subtle ways that only show up under certain access patterns. Other
testers than the BIOS are much more thorough. I once had a machine with bad
cache ram, and the BIOS failed to notice it. It showed up only when a disk
cache I was using did a memory test and had it fail.

[snip other possibilities]

IMHO, GCC stack overflow is unlikely because in that case the error should
be reproducible. But changing it is worth a try.

Nate Eldredge
eldredge AT ap DOT net