From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264717AbUFSVej (ORCPT ); Sat, 19 Jun 2004 17:34:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264723AbUFSVej (ORCPT ); Sat, 19 Jun 2004 17:34:39 -0400 Received: from holomorphy.com ([207.189.100.168]:45711 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S264717AbUFSVeh (ORCPT ); Sat, 19 Jun 2004 17:34:37 -0400 Date: Sat, 19 Jun 2004 14:34:33 -0700 From: William Lee Irwin III To: Ingo Molnar Cc: linux-kernel@vger.kernel.org Subject: Re: [patch] flexible-mmap-2.6.7-D5 Message-ID: <20040619213433.GT1863@holomorphy.com> Mail-Followup-To: William Lee Irwin III , Ingo Molnar , linux-kernel@vger.kernel.org References: <20040618213814.GA589@elte.hu> <20040618231631.GO1863@holomorphy.com> <20040619074612.GB12020@elte.hu> <20040619083446.GP1863@holomorphy.com> <20040619113836.GA16197@elte.hu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040619113836.GA16197@elte.hu> User-Agent: Mutt/1.5.5.1+cvs20040105i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org * William Lee Irwin III wrote: >> Also, I suspect some more graceful fallback would make sense >> particularly for the case of RLIM_INFINITY, which would leave users >> that run with, say, all rlimits at RLIM_INFINITY in the interest of >> having full access to system resources with a mere 512MB of >> virtualspace for the heap, which IIRC glibc is intelligent enough to >> circumvent for malloc(), but not for mmap(NULL, ...). [...] On Sat, Jun 19, 2004 at 01:38:36PM +0200, Ingo Molnar wrote: > well, the 5/6=stack 1/6=malloc rule in the RLIM_INFINITY can be changed. > What would make the most sense - 1/2 for both? I had in mind fallback as opposed to a changed base, but a particular choice of the base may cover enough cases. The bugreport below seems to say there's no need for a change. * William Lee Irwin III wrote: >> If it's been in production that long, I find it hard to believe that's >> never been tripped over. [...] On Sat, Jun 19, 2004 at 01:38:36PM +0200, Ingo Molnar wrote: > it's been tripped over and the 5/6 rule was a fix for such a bugreport. > What happens more in practice frequently is that someone needs a big > stack and sets the stack ulimit to RLIM_INFINITY. This sounds like nothing is needed, then. * William Lee Irwin III wrote: >> [...] (also, that 128MB is currently wasted); [...] On Sat, Jun 19, 2004 at 01:38:36PM +0200, Ingo Molnar wrote: > the 128MB is 'wasted' to give some flexibility to the stack rlimits > changing runtime. But in practice it's far more important to have the > mmap()/malloc() space maximized and flexible than to give the stack > automatic flexibility. Fishing around down there to utilize it for mmap() placement should happen anyway if things fall back far enough in the top-down scheme. Answers like "I've thought about it" or "I've seen this and dealt with it" are good enough for me. There isn't much of a normative aspect to user virtualspace layout. Thanks. -- wli