From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 28228 invoked by alias); 1 Aug 2003 08:58:35 -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 28214 invoked from network); 1 Aug 2003 08:58:34 -0000 Received: from unknown (HELO mx0.gmx.net) (213.165.64.100) by sources.redhat.com with SMTP; 1 Aug 2003 08:58:34 -0000 Received: (qmail 19948 invoked by uid 0); 1 Aug 2003 08:58:32 -0000 Date: Fri, 01 Aug 2003 10:20:00 -0000 From: Daniel Blueman To: gcc@gcc.gnu.org MIME-Version: 1.0 Subject: [IA32] -fschedule-insns and -fschedule-insns2... X-Priority: 3 (Normal) X-Authenticated-Sender: #0008973862@gmx.net X-Authenticated-IP: [194.202.174.101] Message-ID: <8716.1059728312@www63.gmx.net> X-Flags: 0001 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit X-SW-Source: 2003-08/txt/msg00016.txt.bz2 I have always been finding that adding the flag -fschedule-insns when building on IA32/i686 can buy as much as 5% performance improvement in code execution time. This page [http://www.gnu.org/software/gcc/projects/optimize.html] states that it is always disabled by the IA32 back-end - what is the current status of developers' views on it? I am aware that this an cause register allocation failures at compile-time, so is that why it is disabled? Also, this must imply that -fschedule-insns2 is disabled, no? Please CC me on any replies! Many thanks, Dan -- Daniel J Blueman COMPUTERBILD 15/03: Premium-e-mail-Dienste im Test -------------------------------------------------- 1. GMX TopMail - Platz 1 und Testsieger! 2. GMX ProMail - Platz 2 und Preis-Qualitätssieger! 3. Arcor - 4. web.de - 5. T-Online - 6. freenet.de - 7. daybyday - 8. e-Post