From owner-freebsd-scsi  Sat Jan  4 04:34:50 1997
Return-Path: <owner-freebsd-scsi>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id EAA16219
          for freebsd-scsi-outgoing; Sat, 4 Jan 1997 04:34:50 -0800 (PST)
Received: from hydrogen.nike.efn.org (resnet.uoregon.edu [128.223.170.28])
          by freefall.freebsd.org (8.8.4/8.8.4) with ESMTP id EAA16203
          for <freebsd-scsi@FreeBSD.org>; Sat, 4 Jan 1997 04:34:39 -0800 (PST)
Received: from localhost (localhost [127.0.0.1])
          by hydrogen.nike.efn.org (8.8.4/8.8.4) with SMTP
	  id EAA03824; Sat, 4 Jan 1997 04:33:51 -0800 (PST)
Date: Sat, 4 Jan 1997 04:33:50 -0800 (PST)
From: John-Mark Gurney <jmg@nike.efn.org>
Reply-To: John-Mark Gurney <gurney_j@resnet.uoregon.edu>
To: Joerg Wunsch <joerg_wunsch@uriah.heep.sax.de>
cc: FreeBSD SCSI list <freebsd-scsi@FreeBSD.org>
Subject: Re: Ideas on CD changers sought
In-Reply-To: <Mutt.19970104124751.j@uriah.heep.sax.de>
Message-ID: <Pine.NEB.3.95.970104042832.301D-100000@hydrogen.nike.efn.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-freebsd-scsi@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

On Sat, 4 Jan 1997, J Wunsch wrote:

> As I wrote:
> 
> > > (bt0:3:1): "unknown unknown ????" type 0 fixed SCSI 0
> > > sd2(bt0:3:1): Direct-Access 
> > > sd2(bt0:3:1): ILLEGAL REQUEST asc:24,0 Invalid field in CDB
> > 
> > That's the consequence out of assuming `direct access' from the null
> > inquiry.  I think this should be killed, the device should be assigned
> > to the `uk' driver (unknown SCSI device) in case it failed to deliver
> > inquiry data.  It makes things only worse this way.
> > 
> > (The docking station of our Toshiba T5100 notebook contains an AIC7850
> > and an unknown CD drive.  Since it times out during inquiry, the
> > driver also assumes it to be an `sd0', which finally causes the kernel
> > to panic.)
> 
> Since you're just in experimenting :), please try this.  It should
> make the misprobed devices (LUNs in your case) going to the uk driver.
> If nobody objects, i would commit that change, including the 2.2
> branch.  I feel much safer with it than with the current scenario of
> assigning those devices to the `sd' driver (which is IMHO an artifact
> only of T_DIRECT having the value 0).

[patch deleted]

I'm assuming I should combine it with the previous patch, correct?

I'm ifdef'ing the code so my `production' kernel won't be affected...  so
I'm going to be able to turn these hacks on and off for testing...

John-Mark

gurney_j@efn.org
http://resnet.uoregon.edu/~gurney_j/
Modem/FAX: (541) 683-6954   (FreeBSD Box)

Live in Peace, destroy Micro$oft, support free software, run FreeBSD (unix)