From owner-freebsd-hardware  Sat Nov  9 16:41:37 1996
Return-Path: owner-hardware
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id QAA00852
          for hardware-outgoing; Sat, 9 Nov 1996 16:41:37 -0800 (PST)
Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id QAA00844
          for <hardware@FreeBSD.org>; Sat, 9 Nov 1996 16:41:34 -0800 (PST)
Received: from x14.mi.uni-koeln.de (annexr2-39.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA23780
  (5.67b/IDA-1.5 for <hardware@FreeBSD.org>); Sun, 10 Nov 1996 01:40:49 +0100
Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.2/8.6.9) id XAA00830; Sat, 9 Nov 1996 23:42:59 +0100 (MET)
Message-Id: <199611092242.XAA00830@x14.mi.uni-koeln.de>
Date: Sat, 9 Nov 1996 23:41:39 +0100
From: se@zpr.uni-koeln.de (Stefan Esser)
To: richard@pegasus.com (Richard Foulk)
Cc: se@zpr.uni-koeln.de (Stefan Esser), jacs@gnome.co.uk (Chris Stenton),
        hardware@FreeBSD.org
Subject: Re: ncro 810 oops error
In-Reply-To: <199611090134.PAA12889@pegasus.com>; from Richard Foulk on Nov 8, 1996 15:33:50 -1001
References: <199611090134.PAA12889@pegasus.com>
X-Mailer: Mutt 0.45
Mime-Version: 1.0
Sender: owner-hardware@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

Richard Foulk writes:
> } > I have recently upgraded to a Asus Pentium PRO Motherboard with their 
> } > NCR810 SCSI controller. While using the Amanda backup programme I am 
> } > getting a large number of errors from the controller. 
> } > 
> } > They are of the form.
> } > 
> } > Nov  5 11:51:35 hawk /kernel: ncr?: scatter/gather failed (residue=493925376).
> } 
> } A request to transfer some 500MB of data was issued to the driver.
> } The scatter/gather table covers only 64KB, your memory perhaps 32MB
> } or 64MB, so the call can't be satisfied :)
> 
> If the request is in error then the driver's scatter/gather mechanism
> shouldn't be noted as the source of the problem.

The driver in -current contains an additional test for this condition. 
An error message containing the memory buffer address, the requested 
transfer length and the SCSI cmd is printed in that case.

You are of course right, that the scatter/gather message does not give
a good idea of what went wrong, but only of the place in the code were
some invalid parameter finally caused a problem. I'll think about a 
message change for 2.1.6 which corrects this ...

Regards, STefan