NetBSD Problem Report #20458

Received: (qmail 11936 invoked by uid 605); 21 Feb 2003 14:15:19 -0000
Message-Id: <200302211415.h1LEFGt09629@wheel.duzan.org>
Date: Fri, 21 Feb 2003 09:15:16 -0500 (EST)
From: Gary Duzan <gary@duzan.org>
Sender: gnats-bugs-owner@netbsd.org
Reply-To: gary@duzan.org
To: gnats-bugs@gnats.netbsd.org
Subject: cdhdtape image on 1.6 branch fails to load kernel
X-Send-Pr-Version: 3.95

>Number:         20458
>Category:       port-alpha
>Synopsis:       cdhdtape image on 1.6 branch fails to load kernel
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    port-alpha-maintainer
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Feb 21 14:16:00 +0000 2003
>Closed-Date:    
>Last-Modified:  Sun Apr 30 07:57:36 +0000 2006
>Originator:     Gary Duzan
>Release:        NetBSD 1.6 branch
>Organization:
	Ad Hoc
>Environment:
Architecture: alpha
Machine: alpha
>Description:
On my Jensen (DECpc AXP 150) I can't boot any of the 1.6 branch
cdhdtape images off CD. The initial bootstrap comes up fine, but
loading the kernel fails. The CDs appear to have been burned
properly, since I can read them and get the same image contents
back. A 1.6 ISO image loads the kernel fine, as does a 1.6L cdhdtape
image, so the problem seems to be confined to the 1.6 branch
cdhdtape images.

Here are the error messages for a number of versions:

1.6 ustar 1.3 (Sep 8 2002) - 8084288+227040@@@@@@@@lseek section headers: invalid file offset
1.6_STABLE ustar 1.3 (Nov 22 2002) - 8086928+227040@@@@@@@@@lseek section headers: invalid file offset
1.6.1_RC1 ustar 1.3 (Feb 14 2003) - 8081880@@@@@@@@@read text: Input/output error

>How-To-Repeat:
	Burn the respective cdhdtape images to CD and try booting them.
>Fix:
	Unknown.
>Release-Note:
>Audit-Trail:

From: "r.o.s.s" <ross@netbsd.org>
To: gary@duzan.org, gnats-bugs@gnats.netbsd.org
Cc:  
Subject: Re: port-alpha/20458: cdhdtape image on 1.6 branch fails to load kernel
Date: 22 Feb 2003 04:54:35 -0000

 I'll have to try this on a CD, but I just booted the most recent
 1.6.1 cdhdtape from releng.netbsd.org. (1.6.1_RC1 (INSTALL) #0:
 Mon Feb 10 07:34:26 UTC 2003) This was off a hard drive, tho.

 Perhaps it's a more subtle problem involving old SRM or something
 specific to a CD.

 //

From: Gary Duzan <gary@duzan.org>
To: "r.o.s.s" <ross@netbsd.org>
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-alpha/20458: cdhdtape image on 1.6 branch fails to load kernel 
Date: Sat, 22 Feb 2003 06:48:12 -0500

 In Message <20030222045435.14529.qmail@mail.netbsd.org> ,
    "r.o.s.s" <ross@netbsd.org> wrote:

 =>I'll have to try this on a CD, but I just booted the most recent
 =>1.6.1 cdhdtape from releng.netbsd.org. (1.6.1_RC1 (INSTALL) #0:
 =>Mon Feb 10 07:34:26 UTC 2003) This was off a hard drive, tho.

    I'll give it a try.

 =>Perhaps it's a more subtle problem involving old SRM or something
 =>specific to a CD.

    FWIW, I did finally manage to upgrade to the 2.2 firmware, but
 it hasn't changed anything, as far as I can see.

 					Gary Duzan



From: Gary Duzan <gary@duzan.org>
To: "r.o.s.s" <ross@netbsd.org>
Cc: gnats-bugs@gnats.netbsd.org
Subject: Re: port-alpha/20458: cdhdtape image on 1.6 branch fails to load kernel 
Date: Sat, 22 Feb 2003 07:44:09 -0500

 In Message <20030222045435.14529.qmail@mail.netbsd.org> ,
    "r.o.s.s" <ross@netbsd.org> wrote:

 =>I'll have to try this on a CD, but I just booted the most recent
 =>1.6.1 cdhdtape from releng.netbsd.org. (1.6.1_RC1 (INSTALL) #0:
 =>Mon Feb 10 07:34:26 UTC 2003) This was off a hard drive, tho.
 =>
 =>Perhaps it's a more subtle problem involving old SRM or something
 =>specific to a CD.

    Ok, I think I finally have this part sorted out. It turns out
 that adding the "-pad" option to cdrecord gives me a usable CD.
 I assume the loader attempts to read past the end of the data,
 which would succeed on a disk but fail on a CD. (I don't know about
 a tape.) It may be good enough to include a note in the README,
 but adding some padding to the end of the cdhdtape file or convincing
 the loader not to read beyond the end of the image would be nicer.
 I wonder if something like this was done to current already since
 a 1.6L cdhdtape worked for me, or if I was just lucky and it happened
 to end at the right place to avoid the extra read attempt.

 					Gary Duzan


>Unformatted:

NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.39 2013/11/01 18:47:49 spz Exp $
$NetBSD: gnats_config.sh,v 1.8 2006/05/07 09:23:38 tsutsui Exp $
Copyright © 1994-2007 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.