From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Lord Subject: Re: CompactFlash and HD unhappy together on the same IDE channel Date: Sun, 23 Sep 2007 12:26:37 -0400 Message-ID: <46F693BD.8070604@rtr.ca> References: <46C69913.1060505@g7.org> <46C799E8.8080407@rtr.ca> <20070819175632.349d11c5@the-village.bc.nu> <46EC70A0.1000800@g7.org> <20070922191248.5f2bc74a@the-village.bc.nu> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from rtr.ca ([76.10.145.34]:1584 "EHLO mail.rtr.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753204AbXIWQ0f (ORCPT ); Sun, 23 Sep 2007 12:26:35 -0400 In-Reply-To: <20070922191248.5f2bc74a@the-village.bc.nu> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Alan Cox Cc: Malcolm Gillies , linux-ide@vger.kernel.org Alan Cox wrote: >> I've acquired a new CF card that is capable of UDMA2, and with the same >> connectors, wiring and adaptor (i.e. I popped out the MWDMA card and >> inserted the UDMA one), I can now boot and read data on both devices >> without error at reasonable speeds. > > Fascinating. Does point more to a driver funny you are right. At the > moment I'm at a loss to guess what We seem to have had a lot of reports of CF cards failing with MDMA modes recently. Some of these reports are because of SATA<>PATA bridge chips that don't support MDMA, others are due to unknowns. The solution is beginning to look more and more obvious: default to not use MDMA modes for CF devices at boot. And eventually provide a sysfs or libata parameter (or support/snoop SET_XFER_MODE) to change the mode later. Cheers