Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 6 Apr 2001 19:39:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 6 Apr 2001 19:39:32 -0400 Received: from nrcvicex1a.hia.nrc.ca ([204.174.103.8]:18437 "EHLO nrcvicex1.hia.nrc.ca") by vger.kernel.org with ESMTP id ; Fri, 6 Apr 2001 19:39:24 -0400 Message-ID: <3ACE5501.F5F551E4@nrc.ca> Date: Fri, 06 Apr 2001 16:45:05 -0700 From: Tony Hoffmann Reply-To: tony.hoffmann@nrc.ca Organization: National Research Council of Canada X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.4.1 i686) X-Accept-Language: en MIME-Version: 1.0 To: linux-kernel@vger.kernel.org Subject: mkinitrd and 2.4.x annoyances 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 While trying to upgrade a redhat 6.2 I've come across the following annoyance. Changes states that you need mkinitrd 2.9 or better. No problem. Go to redhat site and grab source rpm. In my case 3.0.5 was the one available. Build and install the package. Run it. Curse. Turns out that mkinitrd relies on mktemp accepting the -d which under 6.2 isn't the case. And remakeing mktemp isn't an option as I'm not about to upgrade to glibc 2.2 just for that. It was an easy enough fix to patch mkinitrd to not rely on the -d flag but still bloody annoying. I can see why they use mktemp -d as they do all the work in the /tmp dir so want to avoid potential races in creating the temp directories but why use /tmp at all? It's possible to do everything in the current working, No? - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/