Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 07 Apr 2000 11:17:54 -0700
From:      Parag Patel <parag@cgt.com>
To:        wc.bulte@chello.nl
Cc:        "Koster, K.J." <K.J.Koster@kpn.com>, "'Bob.Gorichanaz@midata.com'" <Bob.Gorichanaz@midata.com>, "'FreeBSD Hackers mailing list'" <freebsd-hackers@FreeBSD.ORG>
Subject:   Re: bad memory patch? 
Message-ID:  <76320.955131474@pinhead.parag.codegen.com>
In-Reply-To: Message from Wilko Bulte <wkb@chello.nl>  of "Fri, 07 Apr 2000 15:36:46 %2B0200." <20000407153646.A7558@yedi.wbnet> 

next in thread | previous in thread | raw e-mail | index | archive | help

Not necessarily a bad idea, but certainly incomplete.  Being able to
detect and run with bad RAM is only about halfway there.  The other half
is being able to remove and replace the bad RAM without taking the
machine down.  Like some of the old mainframes (ie TOPS-10, I think).

This not only needs quite a bit more hardware support, but the software
also has to be able to shuffle everything else out of the bad
DIMM/SIMM/thunk into good memory before it's pulled and replaced.

If you have to take the machine down anyway to replace the bad RAM once
it's detected, you may as well just use a good ECC design instead.


	-- Parag Patel


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?76320.955131474>