*BSD News Article 56814


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!spool.mu.edu!howland.reston.ans.net!newsfeed.internetmci.com!news.mathworks.com!fu-berlin.de!zrz.TU-Berlin.DE!isst.fhg.de!berlin.fhg.de!news.fhg.de!blackbush.xlink.net!isar.de!krabat!leo
From: leo@krabat.marco.de (Matthias Pfaller)
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Re: Parity SIMMS really necessary?
Message-ID: <4923@krabat.marco.de>
Date: 6 Dec 95 12:03:10 GMT
References: <49lbnr$4fq@interport.net> <49qabp$efi@zuul.nmti.com> <49spbi$1m8@sixpack.wustl.edu>
Organization: marco GmbH, D-85221 Dachau
Lines: 22
X-Newsreader: TIN [version 1.2 PL2]

Matt Lundberg (ml@sixpack.wustl.edu) wrote:
> As for using non-parity SIMMS, what does that extra bit give you
> anyway?  I agree that ECC is an advantage, but parity will only 
> inform you that you have a memory error, in most cases by locking
> up the machine.  This is no help.

Now for the brain-dead pcs that might be true. On a pc all you get
on a parity error is a NMI. On a real machine you would get an interrupt
*and* you would be able to get the address of the offending memory location.
In alot of cases that's enough to recover:
(1) The address is user process code area             -> page it in again
(2) The address is user process data area (!modified) -> page it in again
(3) The address is user process data area (modified)  -> kill process
(4) The address is in kernel area                     -> panic

Of course you have to mark the offending page if the error is permanent.
Are you still unconvinced that parity is a *good* thing to have?

	Matthias
-- 
leo@dachau.marco.de			in real life: Matthias Pfaller
marco GmbH, 85221 Dachau, Germany	tel: +49 8131 516142