From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <oliver@first.in-berlin.de>
Delivered-To: caml-list@yquem.inria.fr
Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78])
	by yquem.inria.fr (Postfix) with ESMTP id ABA4BBC75
	for <caml-list@yquem.inria.fr>; Fri, 11 Feb 2005 22:07:37 +0100 (CET)
Received: from first.in-berlin.de (dialin-145-254-065-087.arcor-ip.net [145.254.65.87])
	by nez-perce.inria.fr (8.13.0/8.13.0) with ESMTP id j1BL7Xgv010575
	for <caml-list@yquem.inria.fr>; Fri, 11 Feb 2005 22:07:35 +0100
Received: by first.in-berlin.de (Postfix, from userid 501)
	id 4770FAACC6; Fri, 11 Feb 2005 22:07:30 +0100 (CET)
Date: Fri, 11 Feb 2005 22:07:29 +0100
From: Oliver Bandel <oliver@first.in-berlin.de>
To: caml-list@yquem.inria.fr
Subject: Re: [Caml-list] Memory allocation nano-benchmark.
Message-ID: <20050211210729.GA466@first.in-berlin.de>
References: <420B7A7E.90504@or.uni-bonn.de> <005101c50f7f$6db0e560$d54380d9@mshome.net> <1108048745.16698.101.camel@pelican.wigram> <014801c50f8e$a08e9a40$d54380d9@mshome.net> <015f01c50f99$e8a57e60$d54380d9@mshome.net> <420BBC7D.5070103@t-online.de> <002401c5101b$4130b640$a2f40b50@mshome.net> <1108127068.16698.320.camel@pelican.wigram>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <1108127068.16698.320.camel@pelican.wigram>
User-Agent: Mutt/1.5.6i
X-Miltered: at nez-perce with ID 420D1E95.000 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)!
X-Spam: no; 0.00; oliver:01 bandel:01 oliver:01 in-berlin:01 caml-list:01 wrote:01 gava:01 wrote:01 imho:01 'real':01 bigarray:01 ...:98 paging:98 frederic:03 perhaps:03 
X-Spam-Checker-Version: SpamAssassin 3.0.2 (2004-11-16) on yquem.inria.fr
X-Spam-Status: No, score=0.1 required=5.0 tests=FORGED_RCVD_HELO 
	autolearn=disabled version=3.0.2
X-Spam-Level: 

On Sat, Feb 12, 2005 at 12:04:32AM +1100, skaller wrote:
> On Fri, 2005-02-11 at 20:22, Frédéric Gava wrote:
> 
> > 
> > You could write;
> > for i=0 to tablesize -1 do
> >  let row1 = table.(i) in
> >  for j=0 to tablesize -1 do
> >   let row2 = row1.(j) do
> >   for k=0 to tablesize -1 do
> >    row2.(k) <- (i+1)*(j+1)*(k+1)
> >  done done done
> > 
> > and peraps you will have a faster code.
> 
> I found no difference, here are two runs:
> 
> [skaller@pelican] ~>time ./zmem 250
>  
> real    0m3.110s
> user    0m2.820s
> sys     0m0.240s
> 
> [skaller@pelican] ~>time ./zmem 250
>  
> real    0m27.732s
> user    0m2.750s
> sys     0m0.340s


Two runs, no difference?

Are you sure zmem and zmem are the same?


IMHO 

> real    0m3.110s
> user    0m2.820s
> sys     0m0.240s

and

> real    0m27.732s
> user    0m2.750s
> sys     0m0.340s

differ...




> 
> The huge 'real' time there is VM paging.
> 
> Perhaps Xavier will bless us with a comment as to
> whether invariant code motion optimisation is actually
> done in 
> 
>         table.(i).(j).(k) <- (i+1)*(j+1)*(k+1)
> 
> Using bigarray (c_layout):
>  
> real    0m27.948s
> user    0m0.770s
> sys     0m0.500s
> 
> .. 4 times faster.


?

What is faster than what?!

Is zmem zmem?
What is zmem?

Ciao,
   Oliver