From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25283 invoked by alias); 6 Jan 2003 05:18:55 -0000 Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-owner@gcc.gnu.org Received: (qmail 25273 invoked from network); 6 Jan 2003 05:18:51 -0000 Received: from unknown (HELO snipe.mail.pas.earthlink.net) (207.217.120.62) by 209.249.29.67 with SMTP; 6 Jan 2003 05:18:51 -0000 Received: from ilmasc01-52.midwest.net ([209.248.16.62] helo=there) by snipe.mail.pas.earthlink.net with smtp (Exim 3.33 #1) id 18VPel-0003RG-00; Sun, 05 Jan 2003 21:18:36 -0800 Content-Type: text/plain; charset="iso-8859-1" From: Andy Walker To: dewar@gnat.com (Robert Dewar), lord@emf.net Subject: Re: Sythetic registers: modrm/gas question. Date: Mon, 06 Jan 2003 05:36:00 -0000 Cc: gcc@gcc.gnu.org References: <20030105131257.941B4F2D5D@nile.gnat.com> In-Reply-To: <20030105131257.941B4F2D5D@nile.gnat.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-Id: X-SW-Source: 2003-01/txt/msg00257.txt.bz2 On Sunday 05 January 2003 07:12 am, Robert Dewar wrote: > Actually I think the idea that the issue is in any sense related to the > operation of complex caches is completely bogus. Bogus to you, speculative to me. > In practice in typical > x86 code, nearly all EBP references with small offsets (references to > arguments or locals in the current stack frame) are in L1 cache. All our > data shows that, I didn't know that. Again, thank you for the information. > so if the idea of the SR proposal is somehow to improve > cache performance, that is unlikely to work out in practice. > You really HAVE to look at specific x86 assembly language sequences to > see whether there is anything in this idea or not. I couldn't agree more. I will finish modifying gcc and we will try it and see. Andy