From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263021AbUDASQu (ORCPT ); Thu, 1 Apr 2004 13:16:50 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263028AbUDASQu (ORCPT ); Thu, 1 Apr 2004 13:16:50 -0500 Received: from parcelfarce.linux.theplanet.co.uk ([195.92.249.252]:28102 "EHLO www.linux.org.uk") by vger.kernel.org with ESMTP id S263021AbUDASQg (ORCPT ); Thu, 1 Apr 2004 13:16:36 -0500 Message-ID: <406C5C77.6020300@pobox.com> Date: Thu, 01 Apr 2004 13:16:23 -0500 From: Jeff Garzik User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030703 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Carl-Daniel Hailfinger CC: Linux Kernel Mailing List Subject: Re: libata problems on Promise SX4000 controller References: <406BE23F.3000802@gmx.net> In-Reply-To: <406BE23F.3000802@gmx.net> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Carl-Daniel Hailfinger wrote: > Hi, > > I had problems getting my Promise SX4000 controller to run with libata. > First, I noticed libata refused to drive it (no associated PCI id). After > patching that (including the handler for the right chipset), a > "modprobe sata_promise" hung for about 15 minutes while initializing the > ECC RAM. It then recognized the one attached harddisk, but disk access was > not really possible (I gave up after waiting for half an hour or so). > Let me apologize for testing an older version of libata. At the time I had > the problems, it was the newest available version. I enabled debugging and > post the logs here in the hope that they might help supporting this > controller in the future. > > A note about the controller itself: It has 4 Parallel ATA ports and no > Serial ATA ports. The controller sports the usual SoftRAID from Promise > (0,1,0+1 and 5). The RAID5 seems to be a new feature. I do not care about > the RAID (well, I do, I'm writing a dm configuration helper for it right > now), but I would like to access the drive. Well, the main problem is that you are trying to drive a PATA controller with a SATA driver :) If you wanted to list this as a feature request on http://bugzilla.kernel.org/ that might be reasonable... I have no idea how different the PATA controllers are from the SATA one I tested. Although it does detect your device, I don't know what differences exist... Jeff