Re: Memory Trauam

Richard B. Johnson (
Tue, 19 Nov 1996 20:58:55 -0500 (EST)

On Tue, 19 Nov 1996, Greg Alexander wrote:

> You write something out, you read it back in, it's still in the
> cache, so you're not actually ever reading from the SIMM again. There is
> an instruction to clear the cache, but I don't know if it would be
> implemented on all platforms because I really don't see the point in
> having such an instruction. :)

Well sometimes. Many caches are write-through, which solves a lot of
problems. Some Cyrix chips can have write-behind enabled, but due to other
hardware "bugs" (reserved word, normally used for software), many such
systems fail when running multitasking operating systems.

One thing is certainly true. Power-on-self-test and check-memory routines
are just for "show". You could lean a lot more by seeing if the system
boots and can then do something useful for several days.

I designed the operating system for a CAT-Scanner. The "Program managers"
insisted that, before anything useful is accomplished, a complete suite
of test routines be run. I have never, ever, seen these routines report
any errors. However, I have seen many systems fail because of errors that
the test routines were designed to find!

Dick Johnson
Richard B. Johnson
Project Engineer
Analogic Corporation
Voice : (508) 977-3000 ext. 3754
Fax : (508) 532-6097
Modem : (508) 977-6870
Ftp :
Email :,
Penguin : Linux version 2.1.11 on an i586 machine.
Warning : It's hard to remain at the trailing edge of technology.