From owner-freebsd-bugs  Mon Jul 31 10:42:26 1995
Return-Path: bugs-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.11/8.6.6) id KAA13938
          for bugs-outgoing; Mon, 31 Jul 1995 10:42:26 -0700
Received: from seagull.rtd.com (RTD.COM [192.195.240.142])
          by freefall.cdrom.com (8.6.11/8.6.6) with ESMTP id KAA13932
          for <freebsd-bugs@freebsd.org>; Mon, 31 Jul 1995 10:42:25 -0700
Received: (from tony@localhost) by seagull.rtd.com (8.6.12/8.6.9.1) id KAA06652 for freebsd-bugs@freebsd.org; Mon, 31 Jul 1995 10:42:22 -0700
Date: Mon, 31 Jul 1995 10:42:22 -0700
From: Tony <tony@rtd.com>
Message-Id: <199507311742.KAA06652@seagull.rtd.com>
To: freebsd-bugs@freebsd.org
Subject: Re: SCSI Tape bug.
Newsgroups: rtd.freebsd.bugs
Organization: RTD Internet Access
X-Newsreader: TIN [version 1.2 PL2]
Sender: bugs-owner@freebsd.org
Precedence: bulk

: Mike Pritchard
: My fix also corrected some problems with some of the "quirky" 525MB 1/4"
: drives, according to Joerg Wunsch, which I thought it might when
: I wrote the fix.  Since I don't have access to anything really
: fancy or state of the art (just an old Archive 2150s), I can't verify 
: if the above fix corrects some of the other "odd" SCSI tape bugs 
: that have been reported lately.

This is somewhat related (no complaints, just observations)

I was using a DEC TLZ07 for a while (2.0.5 FreeBSD)

The device probe for st0 would fail if a tape was present in the drive
at boot time - error "device offline".

Due to the POST memory checks and associated BIOS stuff, the tape was always 
fully ready by the time of the device probe.

I also had other problems reading tapes, sometimes it would read 100%, other 
times (with the same tape) abort with a kernel unknown blocksize message.

I also managed (by ejecting tapes at the wrong time, killing tape jobs
etc) to get the system into a state where any write I/O (i.e sync) would fail
with an 'I/O error'.

If I still had the drive, I'd try some debugging, but it had to go back to
work :-(

Also, I recall there being some inconsistencies in the st(4) man page,
such as the minor devno info for the control device.

tony