From owner-freebsd-hackers  Mon Apr 15 12:49:29 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id MAA28604
          for hackers-outgoing; Mon, 15 Apr 1996 12:49:29 -0700 (PDT)
Received: from homer31.u.washington.edu (junkins@homer31.u.washington.edu [140.142.78.17])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id MAA28599
          for <hackers@freebsd.org>; Mon, 15 Apr 1996 12:49:20 -0700 (PDT)
Received: from localhost by homer31.u.washington.edu
	(5.65+UW96.04/UW-NDC Revision: 2.33 ) id AA155914;
	Mon, 15 Apr 96 12:49:05 -0700
Date: Mon, 15 Apr 1996 12:49:04 -0700 (PDT)
From: "D. Junkins" <junkins@u.washington.edu>
To: hackers@freebsd.org
Subject: Re: Pentium fast copy? 
In-Reply-To: <199604140351.UAA02955@Root.COM>
Message-Id: <Pine.A32.3.92a.960415124112.96827C-100000@homer31.u.washington.edu>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

The Usenix '96 paper is by Kevin Lai and Mary Baker from Stanford's
Operating System and Network Group.  The paper, Usenix slides, and new
Linux NFS benchmark results are available at:

	http://plastique.stanford.edu/~laik/benchmarks/index.html

The paper is very interesting reading.  There is a link from the above
page to the benchmarks that were used, but I haven't looked at them yet.

- Doug
+------------------------------+------------------------------------------+
|  Doug Junkins                |  See my home page for my PGP Public Key  |
|  Network Engineer            +------------------------------------------+
|  Computers & Communications  |  junkins@u.washington.edu                |
|  University of Washington    |  http://weber.u.washington.edu/~junkins  |
+------------------------------+------------------------------------------+

On Sat, 13 Apr 1996, David Greenman wrote:

> >The following quote from a paper (sorry, I don't know what the paper is
> >offhand) appeared on the Firewalls mailing list. What's FreeBSD like on
> >this front?
> >
> >    Our results show that none of the systems adequately delivers the
> >    Pentium's memory write performance.  For example, the Pentium can
> >    copy data at over 160 megabytes/second using a prefetching copy
> >    routine, yet none of the systems we tested have implemented such a
> >    routine.  As described below, the prefetching routines address the
> >    fact that the Pentium does not have a write-allocate cache.
> >    Without this optimization, the same routines copy data at about 40
> >    megabytes/second.
>
>    This is a quote from the Usenix '95 paper comparing several PC operating
> systems, FreeBSD 2.0.5R being one of them. The bottom line is that we are
> working on improving this, but some of the "improvements" were recently found
> to be pessimizations in real-world situations. Not much more to say at this
> point.
>
> -DG
>
> David Greenman
> Core-team/Principal Architect, The FreeBSD Project
>