From owner-freebsd-current  Sun Mar 10 10:54:16 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id KAA13720
          for current-outgoing; Sun, 10 Mar 1996 10:54:16 -0800 (PST)
Received: from uni-sb.de (uni-sb.de [134.96.7.230])
          by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id KAA13710
          for <current@freebsd.org>; Sun, 10 Mar 1996 10:54:14 -0800 (PST)
Received: by uni-sb.de with ESMTP; Sun, 10 Mar 1996 19:54:03 +0100 (MET)
Received: by cs.uni-sb.de with ESMTP; Sun, 10 Mar 1996 19:54:01 +0100 (MET)
Received: from doom.cs.uni-sb.de (doom [134.96.247.254])
          by wurzelausix.CS.Uni-SB.DE (8.7.5/wjp-SVR4/960127s) with SMTP
          id TAA16644; Sun, 10 Mar 1996 19:54:04 +0100 (MET)
Message-ID: <3143250D.3A88@wurzelausix.cs.uni-sb.de>
Date: Sun, 10 Mar 1996 19:53:01 +0100
From: Daniel Rock <rock@wurzelausix.cs.uni-sb.de>
Organization: Universität des Saarlandes
X-Mailer: Mozilla 2.0Ib6a [de] (Win95; I)
MIME-Version: 1.0
To: current@freebsd.org
Subject: Re: NCR disk controller, hp disk
References: <199603101457.AA04448@Sisyphos>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Stefan Esser wrote:
> 
> On Mar 8, 16:39, Jim Lowe wrote:
> } I have an NCR 825 PCI disk controller.  I just attached a second disk
> } to the controller and am trying to access it.  The second disk works
> } just fine under dos, but not freebsd-current.  I am getting the following
> } errors:
> }
> } sd1(ncr0:1:0): ABORTED COMMAND asc:4e,0 Overlapped commands attempted
> } , retries:4
> } sd1(ncr0:1:0): ABORTED COMMAND asc:4e,0 Overlapped commands attempted
> } , retries:3
> } sd1(ncr0:1:0): ABORTED COMMAND asc:4e,0 Overlapped commands attempted
> } , retries:2
> } sd1(ncr0:1:0): ABORTED COMMAND asc:4e,0 Overlapped commands attempted
> } , retries:1
> } sd1(ncr0:1:0): ABORTED COMMAND asc:4e,0 Overlapped commands attempted
> } , FAILURE
> } sd1: error reading primary partition table reading fsbn 0 (sd1 bn 0; cn 0 tn 0 s
> } n 0)
> 
> This particular HP drive (the C3724S) is known
> to cause problems with tagged commands. Please
> rebuild your kernel with
> 
> options "SCSI_NCR_DFLT_TAGS=0"
> 
> and everything will be fine again.
> 
> You can reenable tags for the first drive using
> 
> # ncrcontrol -s tags=0 -s tags=4
> 
> I'd be really interested to understand what is
> going wrong here, since this is the only drive
> known to cause that kind of problem. The driver
> sends the command and the drive its reply just
> fine, but this reply indicates that the drive
> doesn't accept a tagged command, though it did
> inidcate support for this feature before.

Not only FreeBSD has problems with HP drives. I had to
disable tagged-command-queueing on a SparcServer 600,
because the HP drive (a C3325A) showed up exactly the same
problems (the SCSI HBA is a FAS236).

Daniel