From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christopher Lemmer Webber Subject: Re: Medium-term road map Date: Mon, 27 Apr 2020 09:06:45 -0400 Message-ID: <87wo61axi2.fsf@dustycloud.org> References: <87mu6zd6tz.fsf@gnu.org> <87pnbtwdh8.fsf@igalia.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]:42418) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jT3TF-0002Km-MX for guix-devel@gnu.org; Mon, 27 Apr 2020 09:06:49 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.90_1) (envelope-from ) id 1jT3TF-0000bu-8R for guix-devel@gnu.org; Mon, 27 Apr 2020 09:06:49 -0400 In-reply-to: <87pnbtwdh8.fsf@igalia.com> List-Id: "Development of GNU Guix and the GNU System distribution." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane-mx.org@gnu.org Sender: "Guix-devel" To: guix-devel@gnu.org Andy Wingo writes: > On Sat 25 Apr 2020 15:37, Ludovic Court=C3=A8s writes: > >> 2. Performance. There are many things we can improve there, first and >> foremost: the =E2=80=9CComputing derivation=E2=80=9D part of =E2=80= =98guix pull=E2=80=99, Guile=E2=80=99s >> compiler terrible time and space requirements > > I think I have a solution here, as discussed on IRC. Basic idea is to > make a direct compiler from Tree-IL to bytecode, skipping the CPS step. > The result won't be optimal in terms of generated code quality (I > estimate on average 2x slowdown) but it will be very cheap to produce (I > estimate 10x improvement in time, and going from O(n log n) space to > O(n), with better constant factors. We'd use this compiler for -O0 or > maybe even -O1 code. > > Got to spend a few days working it up tho! > > Andy That sound like a nice path to an improvement!