From owner-freebsd-hackers  Sat Jul  1 15:07:02 1995
Return-Path: hackers-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id PAA03835
          for hackers-outgoing; Sat, 1 Jul 1995 15:07:02 -0700
Received: from inet-gw-1.pa.dec.com (inet-gw-1.pa.dec.com [16.1.0.22])
          by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id PAA03829
          for <hackers@freebsd.org>; Sat, 1 Jul 1995 15:07:01 -0700
Received: from muggsy.lkg.dec.com by inet-gw-1.pa.dec.com (5.65/24Feb95)
	id AA12216; Sat, 1 Jul 95 15:02:57 -0700
Received:  from whydos.lkg.dec.com by muggsy.lkg.dec.com (5.65/DEC-Ultrix/4.3) with SMTP
	id AA27429; Sat, 1 Jul 1995 18:02:56 -0400
Received: from localhost (localhost [127.0.0.1]) by whydos.lkg.dec.com (8.6.11/8.6.9) with SMTP id SAA09932 for <hackers@freebsd.org>; Sat, 1 Jul 1995 18:03:48 GMT
Message-Id: <199507011803.SAA09932@whydos.lkg.dec.com>
X-Authentication-Warning: whydos.lkg.dec.com: Host localhost didn't use HELO protocol
To: hackers@freebsd.org
Subject: Best way to diagnose system lockups?
X-Mailer: exmh version 1.5omega 10/6/94
Date: Sat, 01 Jul 1995 18:03:48 +0000
From: Matt Thomas <matt@lkg.dec.com>
Sender: hackers-owner@freebsd.org
Precedence: bulk


I'm trying to debug why FreeBSD lockups when using a 4 port Ethernet
controller.  (it's a PCI board with 4 DC21040's sitting behind a DC21050
PCI-PCI bridge; it's UTP only).

Note that UnixWare 2.0 works fine (I transferred many gigabytes over
one of the bridged ports without a problem) with the board which leads
me to believe this may be a problem with FreeBSD and not with the
hardware.

DDB is built into the kernel.  I've modified kern_clock.c to add a 
watchdog into hardclock/softclock (if softclock is not called at least
every 60 seconds, hardclock will force a panic; this is a standard
technique that I've used over the years quite effectively).  

When the system locks up, neither DDB (via CTRL/ALT/ESC) nor hardclock
(no watchdog panic fires) is called.  The only way back is reset which
of course doesn't result in any useful information.  (damn it!  I like
machines with halt buttons (like a VAX or an Alpha).  Sigh)

(stepping through the system with DDB is not really effective which I
don't know where the system is hanging; I've tried the obvious places
in my driver).

Getting an ICE or a bus monitor is impossible.  The best guess is that
my driver is triggering a bug somewhere else in FreeBSD.  

Any ideas on how to approach this?

Matt Thomas                          Internet:   matt@lkg.dec.com
U*X Networking                       WWW URL:    http://ftp.dec.com/%7Ethomas/
Digital Equipment Corporation        Disclaimer: This message reflects my
Littleton, MA                                    own warped views, etc.