*BSD News Article 30392


Return to BSD News archive

Newsgroups: comp.os.386bsd.misc
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!agate!howland.reston.ans.net!vixen.cso.uiuc.edu!milo.mcs.anl.gov!xray!winans
From: winans@xray.aps.anl.gov (John R. Winans)
Subject: Re: SCSI bus hangs -- Help needed
Message-ID: <Cpunx1.LpC@mcs.anl.gov>
Sender: usenet@mcs.anl.gov
Organization: Argonne National Laboratory, Chicago Illinois
References: <2qrksr$7vu@meatball.rwwa.com>
Date: Sun, 15 May 1994 15:19:48 GMT
Lines: 71

In article <2qrksr$7vu@meatball.rwwa.com> witr@rwwa.com writes:
>Am I the *only* person on the face of the earth with this problem?
>
>Config: 1.1-GAMMA, AHA1542B, 3 Maxtor LXT213S, 1 Sony CDU541,
>  1 Archive 2150ES. (The tape drive is external, and is sometimes
>  connected and sometimes not.  Termination is made ``correct''
>  either way.  Only two of the drives are in use.
>
>Problem: Random SCSI bus hangs.  Doesn't seem to correlate
>  with anything other than bus activity.  Doesn't depend on whether
>  tape drive is connected or not, or whether the CDROM is accessed
>  or not.  I get hangs both ways.
>
>When the bus hangs, all devices have their access lights off, the
>AHA his its light on.  I get console messages about timeouts and
>``adapter not taking commands --- frozen?''.
>
>When I push the reset button the AHA light remains on.  It *stays* on
>until the AHA bios starts sniffing the bus.  It seems very much like
>the SCSI half of the AHA card has crashed hard.

Yah, I get the exact same trouble with NETBSD 0.9 and the AHA1542C.  When 
all I had on it was a 60MB QIC it never happened, but when I added a 1GIG
SCSI2 <CONNER CFP1060S 1.05GB 1823>, I would get SCSI bus hangs once a
week or so (my system runs 24 hours/day.)  Since my root, /tmp, and swap all 
go to an IDE drive, I only use the SCSI drive as a regular file system.

I have also seen other postings about SCSI bus hangs posted with followups
that discuss cable quality and termination.  I suspect that could be the
cause of many problems that people are having, but doubt that it is mine, as
I have swapped things around and tried various options on the AHA board WRT
termination.


I get:

May  2 17:25:25 zone /netbsd: aha0: device 6 timed out 

messages on my console when ever my bus hangs... followed shortly by a total
system crash.  My LED on my AHA board has stayed on when ever I was around
to enjoy the disaster first hand, but my LED on the drive is sometimes
stuck on and sometimes not.  Again, when lit they stay on until the AHA
driver in netbsd inits again.

I have the sips on the drive installed and it is plugged into the DIP header
on the AHA board and I have a terminator on an external CDrom drive
installed.  I have the termination switch on the AHA board set to 'software
controlled' and have (in the past) selected the termination to be disabled
with their setup software.  I have also tried setting the termination switch
to 'on' and leaving the external cable unplugged.  Both configs seem to
work equally poorly.

Additional info... I keep the floppy and BIOS disabled on the board.  And in 
the no-CDrom config, I use a ribbion cable that is 18" long.

Before I had any SCSI stuff in my machine at all, it stayed up for 2 months.
(The only reason it came down was to add the SCSI drive.)
And yes, I use it just about every day to cross compile a development
operating system for a 68000 based controller.  I run the XFree stuff, and
slip into the internet almost daily.  So I would think that it is getting a 
reasonable workout.

Should you find any resolutions to this, I would be interested in hearing from
you.

--John
-- 
! John Winans                     Advanced Photon Source  (Controls)    !
! winans@phebos.aps.anl.gov       Argonne National Laboratory, Illinois !
!                                                                       !
!"The large print giveth, and the small print taketh away." - Tom Waits !