From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00,FREEMAIL_FROM
	autolearn=ham autolearn_force=no version=3.4.4
X-Google-Thread: a07f3367d7,3a6a9f1d654285ba
X-Google-Attributes: gida07f3367d7,public,usenet
X-Google-NewGroupId: yes
X-Google-Language: ENGLISH,ASCII-7-bit
Path: 
 g2news2.google.com!postnews.google.com!k30g2000yqf.googlegroups.com!not-for-mail
From: jonathan <johnscpg@googlemail.com>
Newsgroups: comp.lang.ada
Subject: Re: Ada Shootout program for K-Nucleotide (patches)
Date: Fri, 14 Aug 2009 08:31:37 -0700 (PDT)
Organization: http://groups.google.com
Message-ID: 
 <071e37a6-f454-408c-ae2d-9aa44004fc92@k30g2000yqf.googlegroups.com>
References: <4a743343$0$32674$9b4e6d93@newsspool2.arcor-online.net>
	<wccab2gmyiy.fsf@shell01.TheWorld.com>
 <e4f6ff69-b128-4abc-92cf-e4c08dd77efb@y10g2000prf.googlegroups.com>
	<2bae762e-0d8a-4389-843a-466e87f59fd1@a37g2000prf.googlegroups.com>
	<4a786b15$0$30230$9b4e6d93@newsspool1.arcor-online.net>
 <67e76046-62d4-4c0e-bdd8-8d00cdf93bca@l35g2000pra.googlegroups.com>
	<4a79f712$0$31874$9b4e6d93@newsspool3.arcor-online.net>
 <7a5aea9c-1ade-4d73-98ef-08d226823161@z31g2000yqd.googlegroups.com>
	<4a7a0826$0$31862$9b4e6d93@newsspool3.arcor-online.net>
 <h5d5aa$qqg$1@news.tornevall.net>
	<4a7a96a2$0$31867$9b4e6d93@newsspool3.arcor-online.net>
 <4a7b4daa$0$31333$9b4e6d93@newsspool4.arcor-online.net>
	<211a9019-6b96-431d-a634-e399075110ec@t13g2000yqt.googlegroups.com>
	<4a848ec4$0$30231$9b4e6d93@newsspool1.arcor-online.net>
NNTP-Posting-Host: 143.117.23.233
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Trace: posting.google.com 1250263898 2839 127.0.0.1 (14 Aug 2009 15:31:38
 GMT)
X-Complaints-To: groups-abuse@google.com
NNTP-Posting-Date: Fri, 14 Aug 2009 15:31:38 +0000 (UTC)
Complaints-To: groups-abuse@google.com
Injection-Info: k30g2000yqf.googlegroups.com; posting-host=143.117.23.233;
	posting-account=Jzt5lQoAAAB4PhTgRLOPGuTLd_K1LY-C
User-Agent: G2/1.0
X-HTTP-UserAgent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.0.12)
	Gecko/2009072220 Iceweasel/3.0.6 (Debian-3.0.6-1),gzip(gfe),gzip(gfe)
Xref: g2news2.google.com comp.lang.ada:7794
Date: 2009-08-14T08:31:37-07:00
List-Id: <comp.lang.ada>


OK, I found a few more minutes to look at these
things.

Most important thing to remember: Whether you
should use  -funroll-loops, -funroll-all-loops,
-gnatn, -gnatN, or none-of-the-above
depends on the compiler.

After a quick test using GNAT 4.3.2,
none-of-the-above  seems an acceptible choice.

After another test using GNAT 4.3.4 (GPL 2009),
-gnatN -funroll-loops  seems quite a good choice.

On both  GNAT 4.3.2 and  GNAT 4.3.4,
-ffast-math  is near essential for top speed.
In fact it seemed even better with GNAT 4.3.2;
15% to 20% improvement.

Well, -ffast-math isn't usually that good, so I
took the trouble to find out what's going on.

One thing (in retrospect the only thing) that could
have shaved 15% off the running time was removal
of the floating point division by replacing
"x / constant" with "x * (1/constant)".

To verify this I modified the code by replacing
the "x / constant" with "x * (1/constant)" explicitly.
When I did this the  -ffast-math  switch became
unnecessary.

I didn't know -ffast-math could do that.

Jonathan