From owner-freebsd-advocacy  Wed Sep 27 14:19:26 2000
Delivered-To: freebsd-advocacy@freebsd.org
Received: from smtp04.primenet.com (smtp04.primenet.com [206.165.6.134])
	by hub.freebsd.org (Postfix) with ESMTP
	id 301E137B423; Wed, 27 Sep 2000 14:19:22 -0700 (PDT)
Received: (from daemon@localhost)
	by smtp04.primenet.com (8.9.3/8.9.3) id NAA06941;
	Wed, 27 Sep 2000 13:53:25 -0700 (MST)
Received: from usr02.primenet.com(206.165.6.202)
 via SMTP by smtp04.primenet.com, id smtpdAAAslaq6m; Wed Sep 27 13:52:45 2000
Received: (from tlambert@localhost)
	by usr02.primenet.com (8.8.5/8.8.5) id NAA25048;
	Wed, 27 Sep 2000 13:55:09 -0700 (MST)
From: Terry Lambert <tlambert@primenet.com>
Message-Id: <200009272055.NAA25048@usr02.primenet.com>
Subject: Re: Pulse poll at Borland
To: brett@lariat.org (Brett Glass)
Date: Wed, 27 Sep 2000 20:55:08 +0000 (GMT)
Cc: 1mazda1@phoenixdsl.com (1mazda1),
	bright@wintelcom.net (Alfred Perlstein), chat@FreeBSD.ORG,
	advocacy@FreeBSD.ORG
In-Reply-To: <4.3.2.7.2.20000927111427.048114f0@localhost> from "Brett Glass" at Sep 27, 2000 11:20:37 AM
X-Mailer: ELM [version 2.5 PL2]
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-freebsd-advocacy@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

> >> Yep. Believe it or not, I even have a BeBox in the lab. 
> >
> >> --Brett 
> >
> >WHY??????????? 
> 
> It holds the door open. ;-)
> 
> Seriously, it's actually a pretty spiffy machine, with
> its Wurlizer-style jukebox LEDs and all. I'd run NetBSD on
> it if they could get SMP working, but right now I'd waste
> half the CPU if I didn't keep BeOS on it.

The BeBox will probably never run an SMP BSD, because BSD
SMP work is tied too tightly to the MESI cache coherency
model.

The dual PPC 603 BeBox cheats.  Since the 603 does not
have the signal lines for an IPI based cache invalidation
or update, the MMU was omitted from the BeBox, and the
MMU control lines used to implement SMP capability.  In
practice, this means that the machine is somewhat slower
than it needs to be (by one bus wait cycle per invalidation
event), and that it uses an MEI model, not an MESI model.

BSD SMP does not have the synchronization points necessary
to cope with the additional explicit invalidations required
in MEI in order to maintain interprocessor state coherency.

PS: FWIW, I'm the one who lobbied JLG, successfully, to
release the BeBox programming specifications needed to get
a non-Be OS on the thing.  It's a damn shame that they quit
building the hardware.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.


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