From mboxrd@z Thu Jan 1 00:00:00 1970 From: ludo@gnu.org (Ludovic =?UTF-8?Q?Court=C3=A8s?=) Subject: bug#28211: Grafting fails for latest Go release candidate Date: Thu, 24 Aug 2017 00:53:18 +0200 Message-ID: <87lgm9j3q9.fsf__27576.1277800941$1503528858$gmane$org@gnu.org> References: <20170731143925.GA19814@jasmine.lan> <87wp6kuzrp.fsf@fastmail.com> <87wp6fb0p7.fsf@fastmail.com> <87efsljb63.fsf@igalia.com> <87378lszri.fsf@gnu.org> <87h8x04qky.fsf@fastmail.com> <87wp5v2ylv.fsf@fastmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:40388) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dkeXD-0001tp-DQ for bug-guix@gnu.org; Wed, 23 Aug 2017 18:54:04 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dkeXC-0006RT-Fx for bug-guix@gnu.org; Wed, 23 Aug 2017 18:54:03 -0400 Received: from debbugs.gnu.org ([208.118.235.43]:43159) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1dkeXC-0006RP-CQ for bug-guix@gnu.org; Wed, 23 Aug 2017 18:54:02 -0400 Sender: "Debbugs-submit" Resent-Message-ID: In-Reply-To: <87wp5v2ylv.fsf@fastmail.com> (Marius Bakke's message of "Tue, 22 Aug 2017 21:26:04 +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: Marius Bakke Cc: Andy Wingo , guix-devel@gnu.org, 28211@debbugs.gnu.org Marius Bakke skribis: > Most likely just lucky, new grafting failure today. > > I've noticed they are more likely to occur when the host system is busy. > Grafting a large package such as 'texlive' is almost guaranteed to fail > on a fully-utilized machine. > > Maybe we should switch back to 2.0 for the time being? Yes, done: . I took a look: the code in (guix build graft) is multi-threaded but there=E2=80=99s no shared state among threads (except for the current output/error ports, which may partly explain the error messages that suggest memory corruption.) So it may be that the problem is in libguile. Ludo=E2=80=99.