From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: Problem setting dma using pata_it821x driver in vanilla 2.6.23.12 and gentoo-sources 2.6.23-r3 Date: Wed, 09 Jan 2008 11:43:16 +0900 Message-ID: <478434C4.5000209@gmail.com> References: <200801051734.18088.marc@fallon.classyad.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from rv-out-0910.google.com ([209.85.198.190]:55245 "EHLO rv-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751893AbYAICnY (ORCPT ); Tue, 8 Jan 2008 21:43:24 -0500 Received: by rv-out-0910.google.com with SMTP id k20so59775rvb.1 for ; Tue, 08 Jan 2008 18:43:22 -0800 (PST) In-Reply-To: <200801051734.18088.marc@fallon.classyad.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Marc Zuckman Cc: linux-ide@vger.kernel.org Hello, Marc Zuckman wrote: > I recently compiled gentoo-sources-2.6.23-r3 and subsequently vanilla > 2.6.23.12 , but could not get hdparm v 7.7 (or earlier) to set my it8212f > controlled raid array to use dma. > > From past experience, I have learned that my raid 1 array (using the it8212f > raid controller present on my Gigabyte motherboard ) will not function > properly unless it is set to use dma with hdparm. As a result, I execute > hdparm -d1 /dev/hde in both the Gentoo boot and ordinary run levels. The > system has worked fine with multiple 2.6.x kernels. The last kernel that I > have successfully used is 2.6.20-gentoo-r7 (gentoo-sources). hdparm -d1 isn't supported (yet). pata_it821x will configure DMA automatically. > I have checked linux kernel archives several times and have posted this > problem on gentoo forums, but have not been able to find similar reports. > > Perhaps related, but I doubt it, are boot messages that I initially received > about disabling speeds > udma 33 due to failure to detect proper cabling on > the host side of several ide interfaces (more details unfortunately vague). > My cabling is definitely proper. I haven't noted the cabling messages with > recent boot attempts with the problem kernels and I think (not 100% sure) > they were abolished by using the "OBSOLETE" lilo boot loader append > commands "ide0=ata66, ide2=ata66" Can you post full boot log? -- tejun