From owner-freebsd-current@FreeBSD.ORG  Wed Apr 26 23:56:49 2006
Return-Path: <owner-freebsd-current@FreeBSD.ORG>
X-Original-To: current@freebsd.org
Delivered-To: freebsd-current@FreeBSD.ORG
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 73DC716A400;
	Wed, 26 Apr 2006 23:56:49 +0000 (UTC)
	(envelope-from jd@ugcs.caltech.edu)
Received: from barf.ugcs.caltech.edu (barf.ugcs.caltech.edu [131.215.176.99])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 3E86343D45;
	Wed, 26 Apr 2006 23:56:49 +0000 (GMT)
	(envelope-from jd@ugcs.caltech.edu)
Received: by barf.ugcs.caltech.edu (Postfix, from userid 3640)
	id 11AD41F00C2; Wed, 26 Apr 2006 16:56:49 -0700 (PDT)
Date: Wed, 26 Apr 2006 16:56:49 -0700
From: Paul Allen <nospam@ugcs.caltech.edu>
To: Jason Evans <jasone@freebsd.org>
Message-ID: <20060426235649.GB29737@barf.ugcs.caltech.edu>
References: <444F71F3.6030901@fsn.hu> <444F83AD.9040207@FreeBSD.org>
	<20060426182837.GA29737@barf.ugcs.caltech.edu>
	<444FD673.3070004@FreeBSD.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <444FD673.3070004@FreeBSD.org>
Sender: jd@ugcs.caltech.edu
Cc: current@freebsd.org
Subject: Re: malloc problems with MySQL?
X-BeenThere: freebsd-current@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Discussions about the use of FreeBSD-current
	<freebsd-current.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-current>, 
	<mailto:freebsd-current-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-current>
List-Post: <mailto:freebsd-current@freebsd.org>
List-Help: <mailto:freebsd-current-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-current>,
	<mailto:freebsd-current-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Wed, 26 Apr 2006 23:56:49 -0000

>From Jason Evans <jasone@freebsd.org>, Wed, Apr 26, 2006 at 01:22:11PM -0700:
> This code behaves as intended.  Note that in order to get chunk-aligned 
> memory via mmap(), it is necessary to over-allocate, then trim.  By 
> keeping a cache of unmapped chunks, we can avoid the extra system calls 
> most of the time, and we are also able to reduce memory fragmentation in 
> many cases. (Over-allocation and trimming fails to utilize chunk-sized 
> holes in the memory map.)
Yes, I see... perhaps mmap should be enhanced to take an optional alignment
parameter. :o)

                                   Paul