From owner-freebsd-hackers  Fri Jun  2 22:50:39 1995
Return-Path: hackers-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id WAA01283
          for hackers-outgoing; Fri, 2 Jun 1995 22:50:39 -0700
Received: from gndrsh.aac.dev.com (gndrsh.aac.dev.com [198.145.92.241])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id WAA01269
          for <FreeBSD-hackers@FreeBSD.Org>; Fri, 2 Jun 1995 22:50:31 -0700
Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id WAA09513; Fri, 2 Jun 1995 22:50:12 -0700
From: "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com>
Message-Id: <199506030550.WAA09513@gndrsh.aac.dev.com>
Subject: Re: A performance mystery
To: henrich@crh.cl.msu.edu (Charles Henrich)
Date: Fri, 2 Jun 1995 22:50:11 -0700 (PDT)
Cc: FreeBSD-hackers@FreeBSD.Org (FreeBSD hackers)
In-Reply-To: <199506030540.WAA09476@gndrsh.aac.dev.com> from "Charles Henrich" at Jun 3, 95 01:40:40 am
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1153      
Sender: hackers-owner@FreeBSD.Org
Precedence: bulk

> 
> > Notice the fact these numbers are *IDENTICAL*, but yours are not,
> > again something does not add up :-(
> 
> I am as baffled as you are, I just do not understand how the Compaq with
> such miserable memory performance (how exactly do you build a computer
> so that reads are *slower* than writes?) and much slower CPU can do so
> well in the overall.

The only thing I have ever seen that makes writes faster than reads is
a system with the turbo mode turned off.  But that would make the
slow machine even FASTER :-(.

I am starting to wonder about seek times of the disks, this can be 
a major contributor to build time of a kernel.

If you could only make the memory in both machines the same size so
that we could pull the whole kernel sources into the buffer cache
and see what the difference was.

cd /sys
find . | xargs cat >/dev/null
cd compile/YOURKERNEL; time make all >&make.OUT

Please include full output results from time command, it may show
us some more things to look at.

-- 
Rod Grimes                                      rgrimes@gndrsh.aac.dev.com
Accurate Automation Company                   Custom computers for FreeBSD