From: "Andrew Crabtree" Newsgroups: comp.os.msdos.djgpp Subject: Re: djgpp and optimizations Date: Tue, 15 Sep 1998 16:53:58 -0700 Organization: Hewlett Packard Lines: 36 Message-ID: <6tmumk$e5c$1@rosenews.rose.hp.com> References: <6tkaat$7f6$1 AT nnrp1 DOT dejanews DOT com> NNTP-Posting-Host: ros51675cra.rose.hp.com To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Precedence: bulk >I heard that gcc 2.8.1 (djgpp) supports the -mpentium command, if so, what >benifits does it give? Well, theoretically gcc should be able to provide code that runs better on a pentium than it would otherwise when given that option. Gcc has a cost table for various operations (add, shift, mul, div, ...) and it selects one based on what cpu you tell it to use. It helps it make decisions when you write something like x*320. It can determine whether it would be faster to do some lea and adds and shifts or to just do a mul. There is also a -march command (which more the pentium is a noop) but if you ask for 486 you will enable gcc to use bswap, and if you do ppro it will be able to use the conditional move instructions. This will generate code that doesn't run on an earlier version of the cpu. >Also when is djgpp going to officially support optimizations for the pentium? Doubtful ever. I think cygnus is adding some ppro specific stuff, which may benefit pentium also. > PGCC is cool, but doesn't support djgpp (using gcc 2.8.1) PGCC is its own complete compiler environment. It will never 'support' gcc 2.8.1 or any other version of gcc. It is equivalent to it, but with the pentium optimizations. , so what's a >person to do? > > >Also is -fstrength-reduce stable yet? I heard it is, and it isn't. What's >up? > >tom > >-----== Posted via Deja News, The Leader in Internet Discussion ==----- >http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum