From owner-cvs-all Fri Oct 25 14:35:42 2002 Delivered-To: cvs-all@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6A39137B401; Fri, 25 Oct 2002 14:35:40 -0700 (PDT) Received: from thuvia.demon.co.uk (thuvia.demon.co.uk [193.237.34.248]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0E35543EAA; Fri, 25 Oct 2002 14:35:35 -0700 (PDT) (envelope-from mark@thuvia.demon.co.uk) Received: from dotar.thuvia.org (dotar.thuvia.org [10.0.0.4]) by phaidor.thuvia.org (8.12.3/8.12.3) with ESMTP id g9PLZAcF051159; Fri, 25 Oct 2002 22:35:10 +0100 (BST) (envelope-from mark@thuvia.demon.co.uk) Received: from dotar.thuvia.org (localhost [IPv6:::1]) by dotar.thuvia.org (8.12.6/8.12.6) with ESMTP id g9PLZ9H5082230; Fri, 25 Oct 2002 22:35:09 +0100 (BST) (envelope-from mark@dotar.thuvia.org) Received: (from mark@localhost) by dotar.thuvia.org (8.12.6/8.12.6/Submit) id g9PLZ8RA082229; Fri, 25 Oct 2002 22:35:08 +0100 (BST) Date: Fri, 25 Oct 2002 22:35:08 +0100 (BST) From: Mark Valentine Message-Id: <200210252135.g9PLZ8RA082229@dotar.thuvia.org> In-Reply-To: <20021025130324.A7985@Odin.AC.HMC.Edu> X-Mailer: Mail User's Shell (7.2.6 beta(5) 10/07/98) To: Brooks Davis Subject: Re: cvs commit: src/lib/libdisk Makefile chunk.c write_alpha_dis Cc: John Baldwin , Poul-Henning Kamp , "M. Warner Losh" , cvs-committers@freebsd.org, cvs-all@freebsd.org, Mikhail Teterin Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG > From: Brooks Davis > Date: Fri 25 Oct, 2002 > Subject: Re: cvs commit: src/lib/libdisk Makefile chunk.c write_alpha_dis > I don't understand how magicly ignoring the actual disk structure is > more reliable. It's not ignoring the disk structure, it's interpreting it the way it was meant to be interpreted (I make no claims as to the quality of the design of the DOS partition table, I simply describe how it works...). This makes my system boot correctly more often, and is a Good Thing. > For example, lets say you had a system configured like the one John > showed us with the first MBR partition being Windows and the second > being FreeBSD. Now let's say you decide the some new office suite is > sufficently good that you don't need Windows anymore so you stop using. > Now one day, you ralize that you're /usr is getting way too full so you > decided to nuke /dev/ad0s1 and install a BSD disk lable on it with some new > partitions for /usr/local and maybe one for your MP3 collection. Now, > if you were using /dev/ad0s2[a-h] entris in /etc/fstab you're not too > badly off, but if you were using /dev/ad0[a-h] entries, those now point > to /dev/ad0s1 not /dev/ad0s2 so init fails at boot. You're abusing the multiple BSD disklabel mis-feature. Everything works just fine with a single disklabel. The confusion you see is the equivalent of the DOS "shifting drive letter" problem, which should come as no surprise: DOS partition table entries are referenced by their (variable) type, not their (static) index. You've just demonstrated the obscure case where /dev/ad0[a-h] is ambiguous; it can be avoided by using a single disklabel. The more common case (in my experience) is when /dev/ad0s1[a-h] moves about; it can be avoided by using /dev/ad0[a-h]. Cheers, Mark. -- Mark Valentine, Thuvia Labs "Tigers will do ANYTHING for a tuna fish sandwich." Mark Valentine uses "We're kind of stupid that way." *munch* *munch* and endorses FreeBSD -- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message