From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-89684-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 23722 invoked by alias); 21 Jan 2004 11:57:52 -0000
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc/>
List-Post: <mailto:gcc@gcc.gnu.org>
List-Help: <http://gcc.gnu.org/ml/>
Sender: gcc-owner@gcc.gnu.org
Received: (qmail 23707 invoked from network); 21 Jan 2004 11:57:50 -0000
Received: from unknown (HELO cam-admin0.cambridge.arm.com) (193.131.176.58)
  by sources.redhat.com with SMTP; 21 Jan 2004 11:57:50 -0000
Received: from pc960.cambridge.arm.com (pc960.cambridge.arm.com [10.1.205.4])
	by cam-admin0.cambridge.arm.com (8.12.10/8.12.10) with ESMTP id i0LBveAO007810;
	Wed, 21 Jan 2004 11:57:40 GMT
Received: from pc960.cambridge.arm.com (rearnsha@localhost)
	by pc960.cambridge.arm.com (8.11.6/8.9.3) with ESMTP id i0LBveM04894;
	Wed, 21 Jan 2004 11:57:40 GMT
Message-Id: <200401211157.i0LBveM04894@pc960.cambridge.arm.com>
X-Authentication-Warning: pc960.cambridge.arm.com: rearnsha owned process doing -bs
To: Geoffrey Keating <geoffk@apple.com>
cc: Alexandre Oliva <aoliva@redhat.com>,
        Scott Robert Ladd <coyote@coyotegulch.com>,
        Robert Dewar <dewar@gnat.com>, gcc@gcc.gnu.org,
        Nick Burrett <nick@dsvr.net>, Eric Botcazou <ebotcazou@libertysurf.fr>,
        Gabriel Dos Reis <gdr@integrable-solutions.net>,
        Marc Espie <espie@quatramaran.ens.fr>, Richard.Earnshaw@arm.com
Reply-To: Richard.Earnshaw@arm.com
Organization: ARM Ltd.
X-Telephone: +44 1223 400569 (direct+voicemail), +44 1223 400400 (switchbd)
X-Fax: +44 1223 400410
X-Address: ARM Ltd., 110 Fulbourn Road, Cherry Hinton, Cambridge CB1 9NJ.
X-Url: http://www.arm.com/
Subject: Re: gcc 3.5 integration branch proposal 
In-reply-to: Your message of "Tue, 20 Jan 2004 14:30:04 PST."
             <31870CEE-4B98-11D8-83EB-000A95B1F520@apple.com> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Wed, 21 Jan 2004 12:12:00 -0000
From: Richard Earnshaw <rearnsha@arm.com>
X-SW-Source: 2004-01/txt/msg01639.txt.bz2

> 
> Typically, on older, slower machines (eg. the mentioned 300Mhz 
> StrongARM), memory accesses are relatively cheap relative to 
> computation.  Thus, on newer machines it's useful to perform some 
> computation if this results in smaller final data structures, but this 
> is not true on older machines.
> 

But on the other hand, caches were smaller, not backed by L2, and were 
often allocate-on-write only (so zeroing a structure can be very 
expensive, especially if it is then re-initialized with other data).  All 
of these affect general approaches to what needs optimizing.

R.