Message-Id: <m0xbQz9-000S26C@inti.gov.ar>
Comments: Authenticated sender is <salvador AT natacha DOT inti DOT gov DOT ar>
From: "Salvador Eduardo Tropea (SET)" <salvador AT inti DOT edu DOT ar>
Organization: INTI
To: baldo AT chasque DOT apc DOT org, djgpp AT delorie DOT com
Date: Fri, 28 Nov 1997 17:06:50 +0000
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Subject: Re: BinUtils 2.8.1 speed.
In-reply-to: <3.0.1.32.19971128012725.00690fbc@chasque.apc.org>
Precedence: bulk

baldo AT chasque DOT apc DOT org wrote:

> 	Hello, I have tested the test program of Alan M. Doerhoefer in my machine.
> I have a Amd 5x86, 133mhz, 20mb-70ns RAM and 256kb Write-Back cache. I have
> DJDEV201, GCC2721, GPP2721, LGP271, and BinUtils 2.8.1. I dont have
> BinUtils 2.7 because I have downloaded 2.8.1 and deleted 2.7.
> 	In plain DOS this was the results:
> 
> -O3 -m486 -ffast-math -s:
> 14.72528 seconds, 1340 clocks.
> 
> -O3 -m386 -ffast-math -s:
> 14.72528, 1340 (yes, the same as 486).
> 
> And to my surprise, with -O3 -m486 -ffast-math -fomit-frame-pointer -s:
> 14.94506, 1360
> 
> Does not -fomit-frame-pointer supposed to produce better code?
You are using -O3 in a non-pentium, that will bloat your cache! Try -O2.
(Nabo! ;-)))
 
> 	Dah, I want a pentium!!!!
> 	Ahh! Also, what about K5? Someone can test that code with a K5 please?
> Please dont start a Pentium-K5 war! Only show the results and remember that
> this is the DJGPP mailing list/ng.
Mandamelo y trato.

SET 
------------------------------------ 0 --------------------------------
Visit my home page: http://www.geocities.com/SiliconValley/Vista/6552/
Salvador Eduardo Tropea (SET). (Electronics Engineer)
Alternative e-mail: set-sot AT usa DOT net - ICQ: 2951574
Address: Curapaligue 2124, Caseros, 3 de Febrero
Buenos Aires, (1678), ARGENTINA
TE: +(541) 759 0013