From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264890AbUEQEas (ORCPT ); Mon, 17 May 2004 00:30:48 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264888AbUEQEas (ORCPT ); Mon, 17 May 2004 00:30:48 -0400 Received: from willy.net1.nerim.net ([62.212.114.60]:7436 "EHLO willy.net1.nerim.net") by vger.kernel.org with ESMTP id S264890AbUEQEaL (ORCPT ); Mon, 17 May 2004 00:30:11 -0400 Date: Mon, 17 May 2004 06:29:04 +0200 From: Willy Tarreau To: Tomasz Chmielewski Cc: linux-kernel@vger.kernel.org Subject: Re: root fs on usb - is patching kernel still needed? Message-ID: <20040517042903.GB578@alpha.home.local> References: <40A7BDBC.9010209@interia.pl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <40A7BDBC.9010209@interia.pl> User-Agent: Mutt/1.4i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, On Sun, May 16, 2004 at 09:15:08PM +0200, Tomasz Chmielewski wrote: > Are there any solutions for this problem in stable or pre- kernel tree > yet (2.4 or 2.6)? If not, will such a solution be ever included in a > stable kernel? AFAICT, none has been merged (yet). > Or maybe are there any easier solutions for this (some lilo or grub > option?) An initrd might do the trick, although it's sometimes unconvenient to use : - lilo (or grub) loads the initrd into memory - lilo (or grub) loads the kernel and executes it - the kernel sees the initrd and mounts it and executes /linuxrc - linuxrc can load USB modules and wait the required amount of time - linuxrc then mounts the root fs and does the pivot_root() and exits - the kernel finally boots /sbin/init from your USB FS. Regards, Willy