From owner-freebsd-current@FreeBSD.ORG Sun Oct 9 00:34:01 2011 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id DCC29106566B; Sun, 9 Oct 2011 00:34:00 +0000 (UTC) (envelope-from nwhitehorn@freebsd.org) Received: from adsum.doit.wisc.edu (adsum.doit.wisc.edu [144.92.197.210]) by mx1.freebsd.org (Postfix) with ESMTP id A959D8FC14; Sun, 9 Oct 2011 00:34:00 +0000 (UTC) MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-type: text/plain; CHARSET=US-ASCII; format=flowed Received: from avs-daemon.smtpauth1.wiscmail.wisc.edu by smtpauth1.wiscmail.wisc.edu (Sun Java(tm) System Messaging Server 7u2-7.05 32bit (built Jul 30 2009)) id <0LSR00500W8NF700@smtpauth1.wiscmail.wisc.edu>; Sat, 08 Oct 2011 19:33:59 -0500 (CDT) Received: from comporellon.tachypleus.net (adsl-76-208-71-153.dsl.mdsnwi.sbcglobal.net [76.208.71.153]) by smtpauth1.wiscmail.wisc.edu (Sun Java(tm) System Messaging Server 7u2-7.05 32bit (built Jul 30 2009)) with ESMTPSA id <0LSR00LUSW8M8O00@smtpauth1.wiscmail.wisc.edu>; Sat, 08 Oct 2011 19:33:59 -0500 (CDT) Date: Sat, 08 Oct 2011 19:33:57 -0500 From: Nathan Whitehorn In-reply-to: To: Matt Thyer Message-id: <4E90EBF5.4090101@freebsd.org> X-Spam-Report: AuthenticatedSender=yes, SenderIP=76.208.71.153 X-Spam-PmxInfo: Server=avs-11, Version=5.6.1.2065439, Antispam-Engine: 2.7.2.376379, Antispam-Data: 2011.10.9.2414, SenderIP=76.208.71.153 References: <20111008155252.GA24223.edited@vs2.gothschlampen.com> <4E90B4B1.7000009@FreeBSD.org> <4E90C801.4060108@FreeBSD.org> User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:7.0) Gecko/20110928 Thunderbird/7.0 Cc: Warren Block , Garrett Cooper , freebsd-current@freebsd.org, "Thomas K." , Glen Barber Subject: Re: Memstick image differences between 8.x and 9.x X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 09 Oct 2011 00:34:01 -0000 On 10/08/11 19:25, Matt Thyer wrote: > On Oct 9, 2011 8:52 AM, "Warren Block" wrote: >> On Sat, 8 Oct 2011, Glen Barber wrote: >> >>> On 10/8/11 5:40 PM, Warren Block wrote: >>>> On Sat, 8 Oct 2011, Glen Barber wrote: >>>> >>>>> On 10/8/11 2:21 PM, Garrett Cooper wrote: >>>>>>> Are there any general structural differences between FreeBSD 8 and 9 >>>>>>> memstick >>>>>>> images which could be at fault here? >>>>>>> >>>>>> The new memstick image uses GPT instead of MBR partitioning. >>>>> >>>>> GPT should have no impact on booting from the memory stick, as far as I >>>>> am aware. >>>> >>>> Memory stick should not be a problem, but some of the Lenovo notebooks >>>> hate GPT, even with a PMBR: >>>> http://forums.freebsd.org/showthread.php?t=26304 >>>> http://forums.freebsd.org/showthread.php?t=26759 >>>> >>> Ugh, that's annoying. I'm half-tempted to note this in the new >>> installer chapter, but I don't like the idea of such edge cases as these >>> to effectively turn that page into a pseudo-HCL. >> >> There are already a couple of notes about having to use MBR with XP and > other older operating systems. But instead of updating them, I'd rather see > somebody with one of the affected systems contact somebody with influence at > Lenovo and say "hey, the FreeBSD guys are talking about making your broken > GPT support famous" followed quickly by a BIOS update. > I believe this is actually a case of the memstick image being an improperly > formatted GPT as there is no backup partition table at the end of the > volume. > > The only sensible answer is to not use GPT for the memstick image. > > I not said this,loud enough yet but this is a show stopper for 9.0-RELEASE > and must be fixed. > > We can't have a major release that modern systems cannot install with one of > now most popular install methods. > > As a first step, Andriy Gapon has provided a quick patch for makefs(8) so it > can create filesystems with UFS labels (as bsdinstall relys on labels). > > If you want to fix your memstick, create a copy of the partition table at > the end of the volume and it should boot. It is being fixed, pending Andriy's change getting into the tree, which should be soon, and will end up being used for the next build (which I believe is RC1). There is also the interesting question of actually installing to GPT on the hard disk, which is the default in 9.0. Does this not work on some systems? If so, do we want to blacklist them and use a different default partition scheme? Can we identify systems that violate regular PC boot standards and reject GPT? Any data on any of these points would be appreciated. -Nathan