From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 29 May 2002 07:42:05 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 29 May 2002 07:42:05 -0400 Received: from ns.suse.de ([213.95.15.193]:57614 "EHLO Cantor.suse.de") by vger.kernel.org with ESMTP id ; Wed, 29 May 2002 07:42:03 -0400 Date: Wed, 29 May 2002 13:42:02 +0200 From: Dave Jones To: Pavel Machek Cc: Ruth Ivimey-Cook , Luigi Genoni , "J.A. Magallon" , Luca Barbieri , Marcelo Tosatti , Linux-Kernel ML Subject: Re: [PATCH] [2.4] [2.5] [i386] Add support for GCC 3.1 -march=pentium{-mmx,3,4} Message-ID: <20020529134202.F27463@suse.de> Mail-Followup-To: Dave Jones , Pavel Machek , Ruth Ivimey-Cook , Luigi Genoni , "J.A. Magallon" , Luca Barbieri , Marcelo Tosatti , Linux-Kernel ML In-Reply-To: <20020526023009.G16102@suse.de> <20020527085301.A38@toy.ucw.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, May 27, 2002 at 08:53:02AM +0000, Pavel Machek wrote: > Hi! > > > I would be (pleasantly) surprised to see gcc turn a C memcpy into faster > > assembly than our current implementation. And I'll bet > > gcc has hand-coded assembly inside itself, if gcc compiled memcpy is slower > than hand-optimized one, you found a compiler bug. Not at all. gcc compiled memcpy just has no knowledge of things like non-temporal stores, and using mmx/sse to move 64 bits at a time instead of 32 bit registers. (It's only recently it got prefetch abilities too). -- | Dave Jones. http://www.codemonkey.org.uk | SuSE Labs