From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.3 (2006-06-01) on yquem.inria.fr X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=disabled version=3.1.3 X-Original-To: caml-list@yquem.inria.fr Delivered-To: caml-list@yquem.inria.fr Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by yquem.inria.fr (Postfix) with ESMTP id B431CBC69 for ; Mon, 30 Oct 2006 08:50:22 +0100 (CET) Received: from smeltpunt.science.ru.nl (smeltpunt.science.ru.nl [131.174.16.145]) by concorde.inria.fr (8.13.6/8.13.6) with ESMTP id k9U7oMMh014979 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=FAIL) for ; Mon, 30 Oct 2006 08:50:22 +0100 Received: from tandem.cs.ru.nl [131.174.142.18] (helo=tandem.cs.ru.nl) by smeltpunt.science.ru.nl (8.13.7/5.10) with ESMTP id k9U7oHuW010996 for ; Mon, 30 Oct 2006 08:50:17 +0100 (MET) Received: from tews by tandem.cs.ru.nl with local (Exim 4.63) (envelope-from ) id 1GeRuT-0005FK-CB for caml-list@inria.fr; Mon, 30 Oct 2006 08:50:17 +0100 From: Hendrik Tews To: caml-list@inria.fr Subject: Re: [Caml-list] caml_oldify_local_roots takes 50% of the total runtime References: <17727.34685.561877.977822@tandem.cs.ru.nl> <1161802911.12050.5.camel@rosella.wigram> <1161870528.20369.27.camel@localhost.localdomain> <454206CF.1080306@janestcapital.com> Date: Mon, 30 Oct 2006 08:50:17 +0100 In-Reply-To: <454206CF.1080306@janestcapital.com> (Brian Hurt's message of "Fri, 27 Oct 2006 09:17:03 -0400") Message-ID: User-Agent: Gnus/5.110006 (No Gnus v0.6) Emacs/21.4 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Scanned-By: MIMEDefang 2.56 on 131.174.16.145 X-Miltered: at concorde with ID 4545AEBE.000 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)! X-Spam: no; 0.00; hendrik:01 tews:01 tews:01 oldify:01 runtime:01 hendrik:01 heap:01 heap:01 garbage:01 minor:01 minor:01 caml-list:01 writes:01 short:01 data:02 Brian Hurt writes: With respect to 4146 (minor heap size adjusts to memory size)- I'm not sure this is a good idea. I didn't think at all about cache size. For some applications it might certainly be a good idea to keep the minor heap close to the cache size. I was mainly concerned about applications that consume a lot of data, which is not going to fit into cache anyway. If, in addition, they are short running it is much better to have a huge minor heap and let the OS collect the garbage afterwards. Maybe it would be nice to let the application configure the cache policy: - eager: fill all free memory before starting GC - fit_cache: Keep the minor heap close to the cache size Hendrik