From owner-freebsd-hardware Tue Dec 28 20:36:11 1999 Delivered-To: freebsd-hardware@freebsd.org Received: from panzer.kdm.org (panzer.kdm.org [216.160.178.169]) by hub.freebsd.org (Postfix) with ESMTP id CF4951564E; Tue, 28 Dec 1999 20:36:07 -0800 (PST) (envelope-from ken@panzer.kdm.org) Received: (from ken@localhost) by panzer.kdm.org (8.9.3/8.9.1) id VAA94167; Tue, 28 Dec 1999 21:36:05 -0700 (MST) (envelope-from ken) Date: Tue, 28 Dec 1999 21:36:05 -0700 From: "Kenneth D. Merry" To: Mike Smith Cc: Lance Costanzo , freebsd-hardware@FreeBSD.ORG Subject: Re: ECC RAM useless with FreeBSD? Message-ID: <19991228213605.A94126@panzer.kdm.org> References: <3.0.32.19991228154805.006cc640@costanzo.net> <199912290118.RAA04995@mass.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0i In-Reply-To: <199912290118.RAA04995@mass.cdrom.com>; from msmith@FreeBSD.ORG on Tue, Dec 28, 1999 at 05:18:05PM -0800 Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, Dec 28, 1999 at 17:18:05 -0800, Mike Smith wrote: > > I guess the question is- > > Is not being able to log memory errors a problem with the PC hardware > > being inadequate, or feature missing from the operating system? > > I'm fairly sure that most of the chipsets on the market at the moment > that support ECC also support notifying the OS. We don't have any > support for any of these (but would happily integrate it if someone were > to come up with such a beast). I think Oliver's problem was that he was having memory trouble, but wasn't being notified about it. That is slightly different than notifying the OS about bad bits in memory that have been corrected with ECC. In general, with parity errors, I've seen NMI's with the "ram parity error" message, which triggers a panic. You would think that if the system sees a parity error that it can't correct with ECC (2-bit), it would generate the same sort of NMI that it generates for a standard parity error. FWIW, I generally run with parity detection turned on, but not ECC, since I've heard (i.e. I haven't looked in any Intel datasheets to verify this) that there may be a performance penalty for running with ECC turned on. You could probably verify the performance penalty by doing a dd test for memory bandwidth with ECC enabled and simple parity checking enabled. (e.g. "dd if=/dev/zero of=/dev/null bs=1m count=1024") Ken -- Kenneth Merry ken@kdm.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message