From mboxrd@z Thu Jan  1 00:00:00 1970
From: =?utf-8?Q?Ludovic_Court=C3=A8s?= <ludo@gnu.org>
Subject: Re: Guile 3 migration plan
Date: Thu, 23 Jan 2020 17:39:02 +0100
Message-ID: <87tv4mp1p5.fsf@gnu.org>
References: <874kwvemsw.fsf@inria.fr> <871rrxycs5.fsf@gnu.org>
 <87tv4rqgnu.fsf@gnu.org> <87o8uyf3xy.fsf@gnu.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Return-path: <guix-devel-bounces+gcggd-guix-devel=m.gmane-mx.org@gnu.org>
Received: from eggs.gnu.org ([2001:470:142:3::10]:38316)
 by lists.gnu.org with esmtp (Exim 4.90_1)
 (envelope-from <ludo@gnu.org>) id 1iufVZ-0002Jh-GT
 for guix-devel@gnu.org; Thu, 23 Jan 2020 11:39:06 -0500
Received: from fencepost.gnu.org ([2001:470:142:3::e]:32914)
 by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from <ludo@gnu.org>)
 id 1iufVZ-0008UK-Ax
 for guix-devel@gnu.org; Thu, 23 Jan 2020 11:39:05 -0500
Received: from [2001:660:6102:320:e120:2c8f:8909:cdfe] (port=36750 helo=ribbon)
 by fencepost.gnu.org with esmtpsa (TLS1.2:RSA_AES_256_CBC_SHA1:256)
 (Exim 4.82) (envelope-from <ludo@gnu.org>) id 1iufVY-00059F-GH
 for guix-devel@gnu.org; Thu, 23 Jan 2020 11:39:04 -0500
In-Reply-To: <87o8uyf3xy.fsf@gnu.org> ("Ludovic
 \=\?utf-8\?Q\?Court\=C3\=A8s\=22'\?\=
 \=\?utf-8\?Q\?s\?\= message of "Mon, 20 Jan 2020 18:11:37 +0100")
List-Id: "Development of GNU Guix and the GNU System distribution."
 <guix-devel.gnu.org>
List-Unsubscribe: <https://lists.gnu.org/mailman/options/guix-devel>,
 <mailto:guix-devel-request@gnu.org?subject=unsubscribe>
List-Archive: <https://lists.gnu.org/archive/html/guix-devel>
List-Post: <mailto:guix-devel@gnu.org>
List-Help: <mailto:guix-devel-request@gnu.org?subject=help>
List-Subscribe: <https://lists.gnu.org/mailman/listinfo/guix-devel>,
 <mailto:guix-devel-request@gnu.org?subject=subscribe>
Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane-mx.org@gnu.org
Sender: "Guix-devel"
 <guix-devel-bounces+gcggd-guix-devel=m.gmane-mx.org@gnu.org>
To: Guix-devel <guix-devel@gnu.org>

Ludovic Court=C3=A8s <ludo@gnu.org> skribis:

> Ludovic Court=C3=A8s <ludo@gnu.org> skribis:
>
>> However, on ARMv7, Guile 3.0 segfaults as it=E2=80=99s building:
>>
>>   https://ci.guix.gnu.org/log/8b8c0hxvm9qa5kff168vdr3943cc2s61-guile-nex=
t-3.0.0
>
> It seems that the regression happened between 2.9.4 and 2.9.9=C2=B9:
>
>   https://ci.guix.gnu.org/build/1687242/details
>
> I=E2=80=99ve reported it here:
>
>   https://issues.guix.gnu.org/issue/39208

Commit 6e6aab0b6af0bf60b2c7276fa9ee6ca6da9aef88 disables JIT on
armhf-linux to work around this issue.

>From a quick test on a machine that wasn=E2=80=99t quite idle, so take with
quite a bit of salt, the non-JIT =E2=80=98guile3.0-guix=E2=80=99 is ~6% *sl=
ower* than
=E2=80=98guix=E2=80=99 on armhf-linux (wall-clock time of =E2=80=9Cguix bui=
ld libreoffice -nd=E2=80=9D).

Could someone check what they get?

That=E2=80=99s a bummer but it=E2=80=99s understandable because the instruc=
tion set of
the VM in Guile=C2=A03 is larger and lower-level:

  https://wingolog.org/archives/2018/01/17/instruction-explosion-in-guile

Regardless, I would go ahead and move (guix self) to Guile=C2=A03 now.

Thoughts?

Thanks,
Ludo=E2=80=99.