From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bartlomiej Zolnierkiewicz Subject: Re: 2.6.15-rc1: IDE: fix potential data corruption with SL82C105 interfaces Date: Tue, 15 Nov 2005 14:27:08 +0100 Message-ID: <58cb370e0511150527r415a1916t4fbadfc654a2bd18@mail.gmail.com> References: <20051112165548.GB28987@flint.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT Return-path: Received: from nproxy.gmail.com ([64.233.182.200]:64386 "EHLO nproxy.gmail.com") by vger.kernel.org with ESMTP id S932521AbVKON1J convert rfc822-to-8bit (ORCPT ); Tue, 15 Nov 2005 08:27:09 -0500 Received: by nproxy.gmail.com with SMTP id o60so279591nfa for ; Tue, 15 Nov 2005 05:27:08 -0800 (PST) In-Reply-To: <20051112165548.GB28987@flint.arm.linux.org.uk> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org, Alan Cox On 11/12/05, Russell King wrote: > We must _never_ _ever_ on pain of death enable IDE DMA on SL82C105 > chipsets where the southbridge revision is <= 5, otherwise data > corruption will occur. > > Strangely this used to work, but something has changed in the upper > echelons of the IDE layer to break the hosts decision to deny DMA. > Let's make it crystal clear to the IDE layer that we know best. Has it changed recently? AFAICS this bug was introduced long time ago in the sl82c105 driver itself by setting hwif->autodma in init_hwif_sl82c105() without checking for bridge revision: http://linux.bkbits.net:8080/linux-2.6/patch@1.497.94.23?nav=index.html|src/|src/drivers|src/drivers/ide|src/drivers/ide/pci|related/drivers/ide/pci/sl82c105.c|cset@1.497.94.23 > Note: due to the urgency of this fix, I will be applying this to the > ARM tree. Any comments/criticisms can be dealt with further patches. Fine with me. Bartlomiej