Odd memtest86 behaviour

[Lennart Sorensen]

On Sat, Aug 30, 2008 at 03:43:10PM -0400, Alex Beamish wrote:
> I've been running memtest86 on a hand-me-down PC and getting strange results.
> 
> Originally, it had 512M of RAM and it would fail memtest86 after 13
> seconds. I bought an additional 1G today, installed it (1G first slot,
> 512M second slot), ran memtest86 and it failed after about 43 seconds.
> 
> Thinking I'd isolated the problem (or symptom) to the 512M RAM, I took
> that out and ran memtest86 again -- and it failed after 32 seconds.
> (I've defined failing as having the wall time clock stop, and the
> system refuse to respond to the ESC key to reboot.)
> 
> The system is a fairly new white box, P4-2GHz.
> 
> I'm now trying to decide if the processor (or something else) is
> marginal, or if I'm worrying about nothing -- it had an old Windows XP
> installation that I'm planning on overwriting, possibly with
> openSolaris, and I'd like to be sure that the hardware is OK before I
> start storing data on this box (I'm looking at making it a file
> server/data warehouse).

Well it could be cache failure on the CPU.  After all memtest can't tell
the difference between ram and cache, it can only tell something changes
where it shouldn't have.

So could be bad cpu, bad motherboard (chipset, traces, etc), bad power
supply, etc.

I didn't think there was any such thing as a new P4 these days.  P4s
have been obsolete for years.

-- 
Len Sorensen
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists