From owner-freebsd-current  Sun Aug  6 02:01:42 1995
Return-Path: current-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.11/8.6.6) id CAA01231
          for current-outgoing; Sun, 6 Aug 1995 02:01:42 -0700
Received: from asstdc.scgt.oz.au (asstdc.scgt.oz.au [202.14.234.65])
          by freefall.cdrom.com (8.6.11/8.6.6) with ESMTP id CAA01221
          for <current@freebsd.org>; Sun, 6 Aug 1995 02:01:37 -0700
Received: (from imb@localhost)
	by asstdc.scgt.oz.au (8.6.12/BSD4.4)
	id SAA10918; Sun, 6 Aug 1995 18:56:04 +1000
From: michael butler <imb@scgt.oz.au>
Message-Id: <199508060856.SAA10918@asstdc.scgt.oz.au>
Subject: Re: Buslogic BT542-B problems (Try to abort ...)
To: joerg_wunsch@uriah.heep.sax.de
Date: Sun, 6 Aug 1995 18:56:01 +1000 (EST)
Cc: msmith@atrad.adelaide.edu.au, current@freebsd.org
In-Reply-To: <199508041656.SAA27450@uriah.heep.sax.de> from "J Wunsch" at Aug 4, 95 06:56:39 pm
X-Mailer: ELM [version 2.4 PL24beta]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1258      
Sender: current-owner@freebsd.org
Precedence: bulk

> > Recent experiments (and despite source-code comments) have led me to
> > conclude that the BT742 driver does NOT co-operate with the
> > bounce-buffer scheme in -current on an ISA machine
 
> Err... the 742 is EISA, and if my memories are correct, the 542 is
> VLB.  Why the heck do you need this driver for ISA?
 
It seemed like a good idea at the time ... :-)

I've been running a BT542B in AHA-compatible mode for ages but Michael Smith
was kind enough to ship me some new firware so that the BT driver would
actually recognise it as a BT card (which I thought was a Good Thing).

I also thought that the different/extended mailbox stuff (as compared to the
AHA driver) might buy some better on-disk performance (which is what I was
really after). It wasn't to be :-(.

After some drama with microcode revisions that my card simply won't run
with, I found that I can't run the bt driver with 20 meg anyway. When I took
the extra 4 meg out .. it did run faster but not long enough to identify if
this was the driver or just the effect of not using bounce buffers - shortly
afterward, I ran into VM panics with the load that this machine carries ..

Now I'm back where I started, running in AHA mode <sigh> but with (slightly)
newer firmware,

	michael