Subject: Re: Crashes in 1.3.3 indicate faulty hardware?
To: None <port-i386@netbsd.org>
From: Bernd Sieker <bsieker@techfak.uni-bielefeld.de>
List: port-i386
Date: 02/22/1999 17:08:51
On 22.02.99, 06:50:28, Michael Graff wrote:
> Bernd Sieker <bsieker@techfak.uni-bielefeld.de> writes:
> 
> > Hopefully the problem will go away.
> 
> Don't count on it.  I've found that this is almost always a faulty bit
> of hardware somewhere.  For instance, even though there are more SIMM
> slots on my motherboard, putting more than 128M in two of my boxes
> will cause this to happen within a day or so of boot.

You left out the most important bit of my original message where it
said that I would get a new main board.

A lot of the hardware that is faulty or just flaky (which amounts to
faulty if it's a critical piece, even it only fails once in two weeks)
sits on the mainboard itself. Even mainboards with the same chip set
can be very different.

> 
> Many motherboards only cache up to the first N megabytes of RAM.  The
> ones in question (my 128M-limited ones) cache the first 64M, I
> believe.

Uncached main memory will not cause panics, only reduce system
performance. My current mainboard has 512K of L2-Cache and is able to
cache 128MB of main memory, the Tyan will have 1M of L2-cache and able
to cache 256MB, which will be about enough for me for at least one
year. After which I will consider a new system (K7 maybe) anyway.

But now I'm really curious about K6-III performance with 256k of
core-speed on-chip L2-cache plus 1M 100MHz-mainboard-L3-Cache.

> 
> --Michael

-- 
Bernd Sieker

Unix, the solution to the W2k problem.