From mboxrd@z Thu Jan 1 00:00:00 1970 From: Maxim Cournoyer Subject: bug#27284: Memory leak in 'guix pull' or 'make' in guix source Date: Mon, 25 Sep 2017 17:00:42 -0400 Message-ID: <87bmlyijd1.fsf@gmail.com> References: <20170608083935.izw747zaetkaxv4o@abyayala> <87poamv2i7.fsf@gnu.org> <87y3paccu5.fsf@gmail.com> <878th9tx3k.fsf@gnu.org> <87h8vtgu4u.fsf@gmail.com> <874lrr51bu.fsf@gnu.org> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:37287) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dwaUz-00081o-HT for bug-guix@gnu.org; Mon, 25 Sep 2017 17:01:11 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dwaUw-0007D9-HR for bug-guix@gnu.org; Mon, 25 Sep 2017 17:01:05 -0400 Received: from debbugs.gnu.org ([208.118.235.43]:50360) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1dwaUw-0007D5-DS for bug-guix@gnu.org; Mon, 25 Sep 2017 17:01:02 -0400 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <874lrr51bu.fsf@gnu.org> ("Ludovic \=\?utf-8\?Q\?Court\=C3\=A8s\=22'\?\= \=\?utf-8\?Q\?s\?\= message of "Sun, 24 Sep 2017 21:44:05 +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.org@gnu.org Sender: "bug-Guix" To: Ludovic =?UTF-8?Q?Court=C3=A8s?= Cc: 27284@debbugs.gnu.org ludo@gnu.org (Ludovic Court=C3=A8s) writes: > taylanbayirli@gmail.com (Taylan Ulrich "Bay=C4=B1rl=C4=B1/Kammer") skribi= s: > >> ludo@gnu.org (Ludovic Court=C3=A8s) writes: >> >>> [...] >>> >>> There=E2=80=99s also the option of not compiling (gnu packages *) and i= nstead >>> evaluating them, but currently this is too costly in terms of memory >>> and CPU. >>> >>> [...] >> >> Can't we leave this to auto-compilation during normal use of guix? > > We=E2=80=99d have the same problem, only at a different point in time (it= might > even be worse because auto-compilation would use -O2 instead of -O0.) I've experimented a bit with --no-auto-compile (or lack of) and also realized that it doesn't bring much to the table. Evaluating (rather than auto-compiling) the modules is much faster and we need to properly compile those anyway if we are to properly ship a guix binary substitute. Maxim