Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 21 Jun 1995 15:00:56 -0500
From:      "Daniel M. Eischen" <deischen@iworks.InterWorks.org>
To:        gibbs@freebsd.org, mikebo@TELLABS.COM, questions@freebsd.org
Subject:   FBSD 2.xR: AIC7xxx driver vs. st() driver vs. tape drives
Message-ID:  <9506212000.AA13314@iworks.InterWorks.org>

next in thread | raw e-mail | index | archive | help

>Greetings -
>I sent e-mail on this problem to Justin, the author of the AIC7xxx
>driver. He suggested I post my problems to this list:
>
>I've been running an Adaptec 1542B controller under FBSD 2.0 in a 486/50
>EISA/VESA system for some time. I've never had a problem with my Archive
>Viper QIC150 and Exabyte 8200 tape drives before. Last week, I got an
>Adaptec 2742AT EISA dual-channel SCSI card, and have been unable to use
>either tape drive since.
>
>Thinking that the new and improved 274x driver code in v2.0.5R might
>solve the problem, I tried installing the new OS. The QIC150 tape install
>failed with the same error I get under 2.0R whenever I try to access the
>Archive Viper QIC150 drive.
>
>DEBUG: Notify: Attempting to extract from SCSI tape drive...
>DEBUG: Executing command 'cpio -iBduVm -H tar -I /dev/rst0' (Magic 0)
>st0(ahc1:2:0): ILLEGAL REQUEST
>st0: Cannot set selected mode
>st0: bad request, must be between 0 and 0
>cpio: read error: Input/output error
>DEBUG: Command 'cpio -iBduVm -H tar -I /dev/rst0' returns status of 1

How was the Q150 tape made and what block size did you use?  The
tape install failed for me on a HP 1533C DAT tape drive when I
made the tape with the default block size (for tar) of 20*512 bytes.
When I remade the tape with a blocking factor of 1, it worked.
I got an illegal block size message on the failed attempt, so
that's probably not your problem - I'm just throwing it out there.

>
>Another problem I'm having (under 2.0R) since installing the AHA2742AT
>card: Whenever I try to access the Exabyte 8200 8mm drive (even to do
>a "mt -f /dev/st1 rewind") I get a timeout message on the root disk
>drive, the drive select light stays lit and the entire system hangs.
>
>ahc0: target 0, lun 0 (sd0) timed out Abort unsupported!
>ahc0: target 0, lun 0 (sd0) timed out Abort unsupported!
>
>The root drive in this case is an Imprimis 94601-15 (aka. Seagate
>Wren-VII model ST41200N). This also worked fine on the AHA1542B,
>and began happening immediately after I installed the AHA2742AT.
>
>Can anyone answer:
>o  Are there known problems with the SCSI or ST drivers and/or Archive
>   Viper and/or Exabyte drives? Justin cannot confirm the Viper problem
>   as he doesn't own one of these drives. Recent posts to
>   comp.unix.bsd.freebsd.misc indicate I'm not the only one having the
>   "st0: bad request, must be between 0 and 0" problem.

At work we have an Archive 250ST (which is still slow as a dog), and
I installed FreeBSD (the Feb or March snapshot) on an EISA bus 486
with a 2742T (Rev C) using the Archive.  But, I installed by hand
from tape.  I didn't use the boot/cpio install.  I got through the
cpio installation, then escaped and installed manually.  I don't
remember, but I think the FreeBSD didn't support SCSI tape install
then.

We are planning on upgrading that system to 2.0.5 as soon as a
new disk drive arrives (in a week or two), so I could try the
tape install from the Archive.

>o  According to the 2.0.5R release notes, the AIC7xxx driver code is
>   much better than 2.0R, so is the problem in the SCSI or ST driver,
>   or is it the fault of my tape or disk drives?

Like I mentioned above, there does seem to be a bug with the installation
from SCSI tape for default tar block sizes of 20x512.  It's not a
SCSI tape driver bug, because I have no problem performing tar's to/from
tape drive outside of the installation program/script.

>o  Is there a workaround so I can get 2.0.5 running?

Try installing a minimum installation from floppies.  Then install
by hand from tape.

>o  I can't tell if Exabyte access will hang the OS under 2.0.5R since
>   I can't get it installed... but this is a major problem under 2.0R.
>   Does anyone know what the "Abort unsupported" message means and how
>   I can work around it?
>
>Both the Viper and Exabyte drives worked flawlessly with the 1542B, so
>I'm sorely tempted to go back to the ISA controller even tho I'm running
>an EISA system board. ACK!!!
>
>Any help or comments are much appreciated...
>- Mike
>--
>--------------------------------------------------------------------------
>Michael Borowiec        Network Operations      Tellabs Operations, Inc.
>mikebo@TELLABS.COM                              1000 Remington Blvd. MS109
>708-378-6007  FAX: 708-378-6714                 Bolingbrook, IL, USA 60440
>--------------------------------------------------------------------------

Dan Eischen
deischen@iworks.InterWorks.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9506212000.AA13314>