From joe@tao.org.uk Wed Jun 30 14:47:38 1999
Return-Path: <joe@tao.org.uk>
Received: from florence.pavilion.net (florence.pavilion.net [194.242.128.25])
	by hub.freebsd.org (Postfix) with ESMTP id 4A60F1573C
	for <FreeBSD-gnats-submit@freebsd.org>; Wed, 30 Jun 1999 14:47:02 -0700 (PDT)
	(envelope-from joe@tao.org.uk)
Received: from flux.tao.org.uk (flux.tao.org.uk [194.242.131.254])
	by florence.pavilion.net (8.9.2/8.8.8) with ESMTP id WAA19915
	for <FreeBSD-gnats-submit@freebsd.org>; Wed, 30 Jun 1999 22:47:00 +0100 (BST)
	(envelope-from joe@tao.org.uk)
Received: by flux.tao.org.uk (Postfix, from userid 1000)
	id C502EBA20; Wed, 30 Jun 1999 21:46:26 +0000 (GMT)
Message-Id: <19990630214626.C502EBA20@flux.tao.org.uk>
Date: Wed, 30 Jun 1999 21:46:26 +0000 (GMT)
From: joe@tao.org.uk
Reply-To: joe@pavilion.net
To: FreeBSD-gnats-submit@freebsd.org
Subject: reboot causes dumps core after a 2.2.X to 3.2 upgrade
X-Send-Pr-Version: 3.2

>Number:         12468
>Category:       misc
>Synopsis:       reboot causes dumps core after a 2.2.X to 3.2 upgrade
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    freebsd-bugs
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          change-request
>Submitter-Id:   current-users
>Arrival-Date:   Wed Jun 30 14:50:00 PDT 1999
>Closed-Date:    Thu Jul 1 02:41:25 PDT 1999
>Last-Modified:  Thu Jul  1 02:44:07 PDT 1999
>Originator:     Josef Karthauser
>Release:        FreeBSD 3.1-STABLE i386
>Organization:
Tao Research
>Environment:
>Description:

	During the transition from a 2.2.X running operating
	system to 3.2 using 'cvsup' and 'make world', eventually
	one has to reboot the box onto the new kernel.  The
	traditional way of doing this is to use a 'reboot' command.
	Unfortunately it seems that the reboot functionality doesn't
	survive during a upgrade and instead the operator is rewarded
	with a 'dumped core' instead.  This isn't usually a problem
	if the server is in the physical locality of the operator, but
	is a 'right royal pain in the arse' if the server is located
	on a remote desert island ;)

>How-To-Repeat:

	Start with a 2.2.X machine.

	Cvsup the source tree to 3.2-RELEASE, or 3.2-STABLE.

	Make upgrade (or aout-to-elf-build, install, move, etc)

	Use mergemaster to rebuild the /etc/ config, etc.

	Rebuild the kernel and install.

	Disklabel -B primarydrive - to install new boot blocks for
	elf kernel.

	Reboot --- bus error!  Network is down.  Machine sits happily
	in single user mode.

>Fix:
	
	Try and find someone who's near your remote server and ask
	them to press the reset button.  Server then boots normally
	and arrive in 3.2 mode perfectly happily.


>Release-Note:
>Audit-Trail:

From: Ruslan Ermilov <ru@ucb.crimea.ua>
To: joe@pavilion.net
Cc: FreeBSD-gnats-submit@FreeBSD.ORG
Subject: Re: misc/12468: reboot causes dumps core after a 2.2.X to 3.2 upgrade
Date: Thu, 1 Jul 1999 10:27:25 +0300

 On Wed, Jun 30, 1999 at 09:46:26PM +0000, joe@tao.org.uk wrote:
 > 
 > >Description:
 > 
 > 	During the transition from a 2.2.X running operating
 > 	system to 3.2 using 'cvsup' and 'make world', eventually
 > 	one has to reboot the box onto the new kernel.  The
 > 	traditional way of doing this is to use a 'reboot' command.
 > 	Unfortunately it seems that the reboot functionality doesn't
 > 	survive during a upgrade and instead the operator is rewarded
 > 	with a 'dumped core' instead.  This isn't usually a problem
 > 	if the server is in the physical locality of the operator, but
 > 	is a 'right royal pain in the arse' if the server is located
 > 	on a remote desert island ;)
 > 
 > >How-To-Repeat:
 > 
 > 	Start with a 2.2.X machine.
 > 	Cvsup the source tree to 3.2-RELEASE, or 3.2-STABLE.
 > 	Make upgrade (or aout-to-elf-build, install, move, etc)
 > 	Use mergemaster to rebuild the /etc/ config, etc.
 > 	Rebuild the kernel and install.
 > 	Disklabel -B primarydrive - to install new boot blocks for elf kernel.
 > 	Reboot --- bus error!  Network is down.  Machine sits happily
 > 	in single user mode.
 
 As the first step of the install phase, ``upgrade'' saves the copies
 of /bin/sh and /sbin/reboot into /usr/obj, and displays the following
 message:
 +-------------------------------------------------------------+
 | Saving a copy of programs required to shut the system down. |
 +-------------------------------------------------------------+
 
 After it has finished populating your system with the new ELF stuff,
 ``upgrade'' will automatically install new boot blocks onto your root
 hard drive (you were asked for), build and install new ELF kernel,
 set the default obj format to ``ELF'' and ask you to reboot the system:
 
 +-------------------------------------------------------+
 | Your system has now been fully updated to elf!        |
 |                                                       |
 | It's now time to reboot from your new ELF kernel.     |
 | You can type Ctrl-C to abort this (at your own risk)  |
                                      ^^^^^^^^^^^^^^^^
 | or press return to reboot the system.                 |
 +-------------------------------------------------------+
 
 I assume that you pressed Ctrl-C at this point, so this is what
 "your own risk" mean -- you didn't knew how to reboot your system
 correctly.  The correct way is to execute /usr/obj/reboot.
 
 
 -- 
 Ruslan Ermilov		Sysadmin and DBA of the
 ru@ucb.crimea.ua	United Commercial Bank,
 ru@FreeBSD.org		FreeBSD committer,
 +380.652.247.647	Simferopol, Ukraine
 
 http://www.FreeBSD.org	The Power To Serve
 http://www.oracle.com	Enabling The Information Age
 

From: Josef Karthauser <joe@pavilion.net>
To: Ruslan Ermilov <ru@ucb.crimea.ua>
Cc: FreeBSD-gnats-submit@FreeBSD.ORG
Subject: Re: misc/12468: reboot causes dumps core after a 2.2.X to 3.2 upgrade
Date: Thu, 1 Jul 1999 09:01:48 +0100

 Excellent... I'm glad that that's be sorted.  Trust me not to notice
 the obvious.  I've been doing the aout-to-elf-install method by
 hand since since last Oct; it's a case of functional fixedness on
 my part - sorry.
 
 Thanks Ruslan,
 Joe
 
 On Thu, Jul 01, 1999 at 10:27:25AM +0300, Ruslan Ermilov wrote:
 > On Wed, Jun 30, 1999 at 09:46:26PM +0000, joe@tao.org.uk wrote:
 > > 
 > > >Description:
 > > 
 > > 	During the transition from a 2.2.X running operating
 > > 	system to 3.2 using 'cvsup' and 'make world', eventually
 > > 	one has to reboot the box onto the new kernel.  The
 > > 	traditional way of doing this is to use a 'reboot' command.
 > > 	Unfortunately it seems that the reboot functionality doesn't
 > > 	survive during a upgrade and instead the operator is rewarded
 > > 	with a 'dumped core' instead.  This isn't usually a problem
 > > 	if the server is in the physical locality of the operator, but
 > > 	is a 'right royal pain in the arse' if the server is located
 > > 	on a remote desert island ;)
 > > 
 > > >How-To-Repeat:
 > > 
 > > 	Start with a 2.2.X machine.
 > > 	Cvsup the source tree to 3.2-RELEASE, or 3.2-STABLE.
 > > 	Make upgrade (or aout-to-elf-build, install, move, etc)
 > > 	Use mergemaster to rebuild the /etc/ config, etc.
 > > 	Rebuild the kernel and install.
 > > 	Disklabel -B primarydrive - to install new boot blocks for elf kernel.
 > > 	Reboot --- bus error!  Network is down.  Machine sits happily
 > > 	in single user mode.
 > 
 > As the first step of the install phase, ``upgrade'' saves the copies
 > of /bin/sh and /sbin/reboot into /usr/obj, and displays the following
 > message:
 > +-------------------------------------------------------------+
 > | Saving a copy of programs required to shut the system down. |
 > +-------------------------------------------------------------+
 > 
 > After it has finished populating your system with the new ELF stuff,
 > ``upgrade'' will automatically install new boot blocks onto your root
 > hard drive (you were asked for), build and install new ELF kernel,
 > set the default obj format to ``ELF'' and ask you to reboot the system:
 > 
 > +-------------------------------------------------------+
 > | Your system has now been fully updated to elf!        |
 > |                                                       |
 > | It's now time to reboot from your new ELF kernel.     |
 > | You can type Ctrl-C to abort this (at your own risk)  |
 >                                      ^^^^^^^^^^^^^^^^
 > | or press return to reboot the system.                 |
 > +-------------------------------------------------------+
 > 
 > I assume that you pressed Ctrl-C at this point, so this is what
 > "your own risk" mean -- you didn't knew how to reboot your system
 > correctly.  The correct way is to execute /usr/obj/reboot.
 -- 
 Josef Karthauser	FreeBSD: How many times have you booted today?
 Technical Manager	Viagra for your server (http://www.uk.freebsd.org)
 Pavilion Internet plc.  [joe@pavilion.net, joe@uk.freebsd.org, joe@tao.org.uk]
 
State-Changed-From-To: open->closed 
State-Changed-By: ru 
State-Changed-When: Thu Jul 1 02:41:25 PDT 1999 
State-Changed-Why:  
Pilot error. 
>Unformatted:
 
