From owner-freebsd-current  Tue Sep  1 16:05:54 1998
Return-Path: <owner-freebsd-current@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id QAA14113
          for freebsd-current-outgoing; Tue, 1 Sep 1998 16:05:54 -0700 (PDT)
          (envelope-from owner-freebsd-current@FreeBSD.ORG)
Received: from morgan.angis.su.OZ.AU (morgan.angis.su.OZ.AU [129.78.25.38])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id QAA14095
          for <current@freebsd.org>; Tue, 1 Sep 1998 16:05:49 -0700 (PDT)
          (envelope-from tonym@angis.usyd.edu.au)
Received: (from tonym@localhost) by morgan.angis.su.OZ.AU (8.8.7/8.6.6) id JAA21588; Wed, 2 Sep 1998 09:04:45 +1000 (EST)
Date: Wed, 2 Sep 1998 09:04:45 +1000 (EST)
From: Tony Maher <tonym@angis.usyd.edu.au>
Message-Id: <199809012304.JAA21588@morgan.angis.su.OZ.AU>
To: jdp@polstra.com
Subject: Re: Excellent Elf and others
Cc: current@FreeBSD.ORG
Sender: owner-freebsd-current@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

>> 3.0-beta-aout-98-08-23
>> Shell scripts (8 concurrent)  48.7 lpm   (60 secs, 3 samples)
>> 
>> 3.0-beta-elf-98-09-01
>> Shell scripts (8 concurrent) 104.0 lpm   (60 secs, 3 samples)

3.0-beta-elf-98-09-01
(using elf version of bytebench - not that I think it matters much)

Shell scripts (8 concurrent)    103.0 lpm   (60 secs, 3 samples)

> For those of us who don't know anything about bytebench, could you
> explain what these numbers mean?  What's an "lpm"?

Sorry I generally look at them as just a figure of merit to compare with
previous runs. I run bytebench after each make world or new hardware change
just to get a quick basic figure of performance.

Think its "lines per minute".

the script used is basically:

sort >sort.$$ <sort.src
od sort.$$ | sort -n +1 > od.$$
grep the sort.$$ | tee grep.$$ | wc > wc.$$
rm sort.$$ grep.$$ od.$$ wc.$$

 
> "Well, of course, that's obviously to be expected!  The clear and
> trivial reason is blah blah blah ..."  But first, we have to know
> whether it's saying that ELF is faster or slower than a.out. ;-)

Faster! Of course ;-)

And now after reading the script it is probably the softupdates (I did mention
that :-) that is giving the performance boost.
I should have run bytebench after switching on softupdates.

The other tests from the run are similar to previous values.

thanks
tonym

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message