From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262040AbTJSSxS (ORCPT ); Sun, 19 Oct 2003 14:53:18 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262048AbTJSSxS (ORCPT ); Sun, 19 Oct 2003 14:53:18 -0400 Received: from adsl-63-194-133-30.dsl.snfc21.pacbell.net ([63.194.133.30]:7301 "EHLO penngrove.fdns.net") by vger.kernel.org with ESMTP id S262040AbTJSSxR (ORCPT ); Sun, 19 Oct 2003 14:53:17 -0400 From: John Mock To: linux-kernel@vger.kernel.org Subject: Re: Mounting /dev/md0 as root in 2.6.0-test7 Message-Id: Date: Sun, 19 Oct 2003 11:53:25 -0700 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > First, is it possible to mount an md device as root (superblock is > present)? Yes, software RAID5 works for me on a PPC with 2.6.0-test7 (aside from other video problems which make it hard for me to test adequately). There is a HOW-TO which explains how to do this (you want the >= 0.9 version, not the 0.4 version which is not relevant for modern kernels). > If so, I can't get it to work :( > I pass root=/dev/md0 to the kernl, but I get the "Kernel panic: VFS: > Unable to mount root fs on md0" error. The basic problem here is that the kernel has no idea what disks comprise your RAID when you just say "root=/dev/md0", e.g., it has no idea where to start. Here's an extract from my PPC's .config file: CONFIG_CMDLINE="root=/dev/md0 md=0,/dev/sda7,/dev/sdb7,/dev/sdc7" Something like that could also be in /etc/lilo.conf if you use LILO (which a PowerPC does not). Thus the kernel has three disk to look at, to find out which partitions make up /dev/md0. Since i want the machine to work no matter which of my three disk might fail, i have arranged my disks to have the root partition on the same partition number on each drive (even though they're not the same capacity or even organized in the same way otherwise). If /dev/sda and /dev/sdb differ too much, then if /dev/sda fails sufficiently badly, then the drives get 'renumbered' (e.g. what was /dev/sdb because /dev/sda, /dev/sdc becomes /dev/sdb, etc.). Write privately if you want to discuss this further (albeit i am hardly an expert on this), as this mailing list is about maintaining the kernel rather than about helping people get started with new configurations. Yes, it can work, and if it does with one recent kernel, but not another, then that's a good hint that this may be an appropriate place to post (assuming a search of the archives doesn't already contain alot of discussion on your topic). Good luck! -- JM