From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ludovic =?UTF-8?Q?Court=C3=A8s?= Subject: bug#39941: Disk-image size increase on core-updates. Date: Wed, 11 Mar 2020 21:33:31 +0100 Message-ID: <87r1xyfx7o.fsf@gnu.org> References: <87sgimdjcj.fsf@gmail.com> <87eeu0h8d6.fsf@gmail.com> <87a74nhstf.fsf@gnu.org> <87zhcnympn.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]:43105) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jC83H-0003n3-Kn for bug-guix@gnu.org; Wed, 11 Mar 2020 16:34:04 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1jC83G-0001ra-Ja for bug-guix@gnu.org; Wed, 11 Mar 2020 16:34:03 -0400 Received: from debbugs.gnu.org ([209.51.188.43]:49649) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1jC83G-0001rU-Ga for bug-guix@gnu.org; Wed, 11 Mar 2020 16:34:02 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1jC83G-00038p-Cm for bug-guix@gnu.org; Wed, 11 Mar 2020 16:34:02 -0400 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <87zhcnympn.fsf@gmail.com> (Mathieu Othacehe's message of "Wed, 11 Mar 2020 15:45:24 +0100") 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: 39941@debbugs.gnu.org Hi! Mathieu Othacehe skribis: > Turns out on master, > > guix size `guix build guile|grep -v debug` > > > returns ~124 MiB, while, > > guix size `guix build --target=3Daarch64-linux-gnu guile|grep -v debug` > > returns ~523 MiB. > > When building the native guile, only the "lib" output of gcc is > kept. When cross-compiling, the whole cross-compiler is dragged down (as > there's only "out" defined for cross-gcc). Ah yes, we should try to introduce a =E2=80=9Clib=E2=80=9D output there as = well. I vaguely remember this was tricky but I=E2=80=99m not sure why; worth a try! > That does not explain the sudden increase, but this is still quite > unfortunate :p Yep, would also be nice to track down. Speaking of which, the Data Service has things like: http://data.guix.gnu.org/gnu/store/p8in2npgl5yhliy25ikz7shjbq0gii95-guile= -next-3.0.0 We could eventually use it to track the evolution of the size of individual packages, or that of their closure (even for cross-compiled packages). I remember a discussion with Chris Baines on this but I=E2=80= =99m not sure what the outcome was. Ludo=E2=80=99.