From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Cooke Subject: Re: 2.4.26rc1 / HPT 374 / RAID = data read corruption with disks onprimary channels. Date: Wed, 31 Mar 2004 16:01:12 +0100 Sender: linux-ide-owner@vger.kernel.org Message-ID: <1080745272.20830.13.camel@pc24.sr.bham.ac.uk> References: <1080677057.11947.83.camel@sage.kitchen> <63171.194.237.142.24.1080716884.squirrel@ncircle.nullnet.fi> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from smtp-out6.blueyonder.co.uk ([195.188.213.9]:39126 "EHLO smtp-out6.blueyonder.co.uk") by vger.kernel.org with ESMTP id S261988AbUCaPBW (ORCPT ); Wed, 31 Mar 2004 10:01:22 -0500 In-Reply-To: <63171.194.237.142.24.1080716884.squirrel@ncircle.nullnet.fi> List-Id: linux-ide@vger.kernel.org To: Tomi Orava Cc: linux-ide@vger.kernel.org, Marcelo Tosatti Dear Tomi, all, This patch appears to have cured the problems I was having. 5 loops over a 2GB file have shown no errors in copying or differing checksums. The only difference in testing is that I am logged in remotely, rather than from X on the server itself, though it will be a few more hours before I am in front of the machine again. Thanks for the patch Andre/Tomi. It has seemingly cured the silent disk corruption I was seeing. Marcelo - I'd certainly recommend this patch for 2.4.*. 2.4.26rc1 was giving random 4-byte corruptions for me without it. http://marc.theaimsgroup.com/?l=linux-ide&m=108071690505910&w=2 Mark On Wed, 2004-03-31 at 08:08, Tomi Orava wrote: > > Hi all, > > > > I've been having some trouble with an Abit IT7 machine. It is a > > pentium-4 machine, 1GB ram (passes days of all-test memtest86), 80GB > > seagate on the ICH4 as the system disk. 4 x 160GB seagates, one on each > > channel of the HPT374. > > It seems that 2.4.26rc1 doesn't still contain the HPT374 patch which > fixes at least some DMA-problems with that ide-chip. Perhaps you'd like > to retest after aplying the patch Andre Hedrick posted on this list a > short while ago ? I'm not saying it will fix your problems, but its worth > trying. > > Regards, > Tomi Orava