From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Nieuwenhuizen Subject: bug#40839: Shepherd activation .GO files are not cross-compiled ... and the Hurd Date: Tue, 28 Apr 2020 19:20:36 +0200 Message-ID: <87o8rbcysb.fsf@gnu.org> References: <87imhnuc8a.fsf@gnu.org> <87sggru4oh.fsf@gmail.com> <871robtpjd.fsf@gnu.org> <87h7x5w1hi.fsf@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:470:142:3::10]:40650) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jTTvE-0007ZC-CK for bug-guix@gnu.org; Tue, 28 Apr 2020 13:22:05 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.90_1) (envelope-from ) id 1jTTuo-0004Pj-N6 for bug-guix@gnu.org; Tue, 28 Apr 2020 13:21:28 -0400 Received: from debbugs.gnu.org ([209.51.188.43]:57958) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1jTTuo-0004Pc-93 for bug-guix@gnu.org; Tue, 28 Apr 2020 13:21:02 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1jTTuo-0006wS-4w for bug-guix@gnu.org; Tue, 28 Apr 2020 13:21:02 -0400 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <87h7x5w1hi.fsf@gmail.com> (Mathieu Othacehe's message of "Mon, 27 Apr 2020 14:35:05 +0200") List-Id: Bug reports for GNU Guix List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: bug-guix-bounces+gcggb-bug-guix=m.gmane-mx.org@gnu.org Sender: "bug-Guix" To: Mathieu Othacehe Cc: 40839@debbugs.gnu.org Mathieu Othacehe writes: Hello Mathieu! > I had a look to (gnu system hurd), this is really nice! I think we could > try an explosive mixture of our two branches :) Sure, why not? ;-) I played a bit yesterday with wip-disk-image. Not that (gnu system hurd) already lives on core-updates; possibly we can start playing there? I tried rebasing wip-disk-image on core-updates and that was (almost?) painless. > More seriously, we could do something like: > > (define hurd-disk-image > (image > (format 'disk-image) > (partitions > (list > (partition > (size 'guess) > (label "Guix_image") > (file-system "ext2") > (flags '(boot)) > (initializer (gexp initialize-hurd-root-partition))))))) Sweet! > then we could have some mapping in guix/scripts/system.scm to > associate: > > * x86_64-linux -> efi-disk-image > * i586-pc-gnu -> hurd-disk-image > > and one could get a hurd disk-image by typing:=20 > > guix system disk-image --target=3Di586-pc-gnu my-hurd-os.scm Oh, that sounds real great. > One problem that can arise is the installation of grub. Currently > wip-disk-image does not support legacy Grub (MBR based) > installation. > > This is because running grub-install needs root permissions, to mess with > /dev/something in order to write the MBR I guess. Hmm...so we need to do some work, is that bad? > We could also create a Hurd ISO if grub-mkrescue (that is used to make > the ISO bootable), supports the Hurd. > > Adding Ludo that might have some insight here. Hopefully -- this is also pretty out of my comfort zone, otoh I am very motivated to get this going. :-) I have been wondering about the branch name in combination with its functionality: can/will/could "wip-disk-image" also be used for guix system init/reconfigure (we don't have qemu on the Hurd)? Greetings, janneke --=20 Jan Nieuwenhuizen | GNU LilyPond http://lilypond.org Freelance IT http://JoyofSource.com | Avatar=C2=AE http://AvatarAcademy.com