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 autolearn=ham
	autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,38fc011071df5a27
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 2003-06-06 23:32:03 PST
Path: 
 archiver1.google.com!news1.google.com!newsfeed.stanford.edu!logbridge.uoregon.edu!nntp-server.caltech.edu!attla2!ip.att.net!attbi_feed3!attbi.com!sccrnsc01.POSTED!not-for-mail
From: tmoran@acm.org
Newsgroups: comp.lang.ada
Subject: Re: Ideas for Ada 200X
References: <bebbba07.0306062150.23d77158@posting.google.com>
X-Newsreader: Tom's custom newsreader
Message-ID: <CFfEa.48068$d51.104155@sccrnsc01>
NNTP-Posting-Host: 12.234.13.56
X-Complaints-To: abuse@attbi.com
X-Trace: sccrnsc01 1054967522 12.234.13.56 (Sat, 07 Jun 2003 06:32:02 GMT)
NNTP-Posting-Date: Sat, 07 Jun 2003 06:32:02 GMT
Organization: AT&T Broadband
Date: Sat, 07 Jun 2003 06:32:02 GMT
Xref: archiver1.google.com comp.lang.ada:38780
Date: 2003-06-07T06:32:02+00:00
List-Id: <comp.lang.ada>

>Compare
>
>    A := B + C + D
>
>with
>
>    A := B
>    A += C
>    A += D
  A good Ada programmer would first write a function "+"(Left,Right:Matrices)
and write
  A := B + C + D
and make sure his algorithm worked.
If he then found there were performance problems, and they were in that
part of his code, he would then try implementing
>    A := B
>    A += C
>    A += D
[though he would write it as
 A := B;
 Add(A, C);
 Add(A, D);]
and he would see if that made a worthwhile speedup.  If there were
still problems, he could try
  for row in A'range(1) loop
    for col in A'range(2) loop
       A(row,col) := B(row,col) + C(row,col) + D(row,col);
    end loop;
  end loop;
or he could try using special hardware vector instructions, each time
running timing tests to see whether the new code was indeed usefully faster.