From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S266597AbUGUSuk (ORCPT ); Wed, 21 Jul 2004 14:50:40 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266599AbUGUSuj (ORCPT ); Wed, 21 Jul 2004 14:50:39 -0400 Received: from ebiederm.dsl.xmission.com ([166.70.28.69]:32671 "EHLO ebiederm.dsl.xmission.com") by vger.kernel.org with ESMTP id S266597AbUGUSuh (ORCPT ); Wed, 21 Jul 2004 14:50:37 -0400 To: bruce@it.usyd.edu.au (Bruce Janson) Cc: rddunlap@osdl.org, Subject: Re: kexec -l ... --ramdisk= References: <200407210653.i6L6rXsd022758@nlp0.cs.usyd.edu.au> From: ebiederm@xmission.com (Eric W. Biederman) Date: 21 Jul 2004 12:49:39 -0600 In-Reply-To: <200407210653.i6L6rXsd022758@nlp0.cs.usyd.edu.au> Message-ID: User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org bruce@it.usyd.edu.au (Bruce Janson) writes: > i Eric, Randy, > Thanks for kexec. It seems to mostly work well, and is an > improvement on the other contenders (2-kernel monte, bootimg and lobos). Thanks. I am bouncing this to lkml since this at the moment does not appear to be a kexec issue. I just checked and initial ramdisks work. The kernel messages I see look like: > checking if image is initramfs...it isn't (no cpio magic); looks like an initrd > Freeing initrd memory: 1503k freed > I am trying to use kexec's "--ramdisk=..." facility. Both > loader and loadee kernels are Linux 2.6.7. The new kernel loads and > runs successfully until it tries to mount its root file system. > Then it fails with this: > > RAMDISK: Couldn't find valid RAM disk image starting at 0. That message is because of your root=/dev/ram command line, and really has nothing to do with your initial ramdisk. > I believe that the ramdisk image that I supply (to kexec) > is a well-formed ext2 file system image. Just for grins I also tried > a copy of a RedHat 9.0 /boot/initrd* from one of our servers but it > produced the same failure and message. Perhaps you did not compile in initrd support? > The failure message (from .../linux/init/do_mounts_rd.c:130) > suggests that the booted kernel has not found a valid ext2 superblock > at the expected ramdisk location. As I trust the format of the > ramdisk image I now suspect that the ramdisk location may be wrong. Unless you have gotten quite creative I doubt it. Looking at all of the kernel messages would have been more interesting. > Any suggestions that you can offer would be appreciated. See above. Eric