From owner-freebsd-questions  Thu May 11 11:59:05 1995
Return-Path: questions-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id LAA09795
          for questions-outgoing; Thu, 11 May 1995 11:59:05 -0700
Received: from hda.com (hda.com [199.232.40.182])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id LAA09789
          for <FreeBSD-Questions@FreeBSD.org>; Thu, 11 May 1995 11:59:03 -0700
Received: (dufault@localhost) by hda.com (8.6.9/8.3) id OAA06912; Thu, 11 May 1995 14:59:34 -0400
From: Peter Dufault <dufault@hda.com>
Message-Id: <199505111859.OAA06912@hda.com>
Subject: Re: Zip drives
To: fcawth@squid.umd.edu (Fred Cawthorne)
Date: Thu, 11 May 1995 14:59:34 -0400 (EDT)
Cc: chuckr@Glue.umd.edu, FreeBSD-Questions@FreeBSD.org
In-Reply-To: <9505111713.AA27882@squid.umd.edu> from "Fred Cawthorne" at May 11, 95 01:13:46 pm
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text
Content-Length: 1739      
Sender: questions-owner@FreeBSD.org
Precedence: bulk

Fred Cawthorne writes:
> 
> > 
> > Has anyone tried out the Iomega Zip drives?  I have a Adaptec 2842, will 
> > that drive one around the block?  Are there any gotcha's about mounting 
> > or umounting, other than what I'd use for, say, a floppy?
> >
> I have one on my FreeBSD 950322-SNAP machine here...
> Here's how it is probed:
> (ncr0:6:0): "IOMEGA ZIP 100 L.27" is a type 0 removable SCSI 2
> sd1(ncr0:6:0): Direct-Access 
> sd1(ncr0:6:0): ILLEGAL REQUEST asc:24,0Invalid field in CDB
> sd1 could not mode sense (4). Using ficticious geometry
> 96MB (196608 512 byte sectors)

I'm going to add something to print out what it doesn't like about the
CDB.

> have changed..."  and then gives me "device not configured" messages
> when I try to access any of the /dev/sd1* devices... (i.e. I can't mount
> it again until I reboot)
> Otherwise, it is nice.  Has anybody else seen this kind of problem with
> a removeable disk????   I get ~700 K/sec with my NCR 810 scsi controller
> with iozone creating and reading a 32 meg file...
> Any ideas???
> Is there a bug in the code that re-reads the disklabel when the medium
> is changed???  

It is supposed to lock out access until the disk is completely closed
and reopened.  If you unmounted the drive and nothing had the drive
open then a bug has snuck in and this code in sd.c:

>        /*
>         * If somebody still has it open, then forbid re-entry.
>         */
>        if (dsisopen(sd->dk_slices)) {
>            errcode = ENXIO;
>            goto bad;
>        }

is returning true when it shouldn't be.

-- 
Peter Dufault               Real Time Machine Control and Simulation
HD Associates, Inc.         Voice: 508 433 6936
dufault@hda.com             Fax:   508 433 5267