From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 4 Sep 2001 11:03:01 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 4 Sep 2001 11:02:52 -0400 Received: from mailout00.sul.t-online.com ([194.25.134.16]:15372 "EHLO mailout00.sul.t-online.de") by vger.kernel.org with ESMTP id ; Tue, 4 Sep 2001 11:02:44 -0400 Message-ID: <3B94ECB2.6634FE9E@t-online.de> Date: Tue, 04 Sep 2001 17:01:06 +0200 From: SPATZ1@t-online.de (Frank Schneider) X-Mailer: Mozilla 4.76 [de] (X11; U; Linux 2.4.3-test i686) X-Accept-Language: en MIME-Version: 1.0 To: Patrick Allaire CC: "linux-kernel@vger.kernel.org" Subject: Re: Failure to mount root fs ... In-Reply-To: <9A1957CB9FC45A4FA6F35961093ABB84043F1D0E@srvmail-mtl.ubisoft.qc.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Patrick Allaire schrieb: > > Hi all, > > I am currently trying to boot for a DiscOnChip with mtd drivers. When I boot > the system stop with the following message : "Kernel panic: VFS: Unable to > mount root fs on 03:05" ! > > I know that o3:05 mean : /dev/hda5. HTis is where I did compile my kernel > ... but now the kernel is on /dev/nftla1 ... but the lilo.conf on this > device is configured to nftla1 and I did run lilo??? so I dont know where > the kernel take this information about boot ing on /dev/hda5 ??? I The Kernel gets this info at compile-time, it is only overridable by a kernel-argument ("root=..."), this is what lilo does. If you want to change this so called "root-device" inside the kernel, you should take a look at the tool "rdev" and its man-page. I had the same problem some time ago as i wanted to build a boot-diskette without any loader like lilo or loadlin, etc. As the kernel contains an i386 bootsector, it should be able to boot itself if it is copied on a diskette and the bios jumps to its begin...but then you cannot say the kernel, where the root-filesystem is, so you have to change the root-device inside the compiled kernel, and this does "rdev". Unfortunately, it did not work for me, so i switched back to loadlin...:-((...but i tried it only one time, perhaps *i* made a mistake and not the tool...:-) Solong.. Frank. -- Frank Schneider, . Microsoft isn't the answer. Microsoft is the question, and the answer is NO. ... -.-