From mboxrd@z Thu Jan  1 00:00:00 1970
From: <bogus@does.not.exist.com>
Date: Fri, 23 Jan 2009 10:48:37 -0000
Subject: No subject
Message-ID: <mailman.36.1247132760.8188.u-boot@lists.denx.de>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de


.section ".text.head","ax"
ENTRY(_stext)
	bras	1f	/* Jump over bootinfo version numbers */

	.long	BOOTINFOV_MAGIC
	.long	0


And from U-Boot itself:


-> md ${loadaddr}
40010000: 27051956 97342d7d 4a54afd5 00300000    '..V.4-}JT...0..
40010010: 40020000 40020000 8a006fc1 050c0200    @... at .....o.....
40010020: 4c696e75 78204b65 726e656c 20496d61    Linux Kernel Ima
40010030: 67650000 00000000 00000000 00000000    ge..............
40010040: 60084249 561a0000 00004ef9 40304000    `.BIV.....N. at 0@.


I suspect that the failure of starting Linux is caused by some mismatch
in data structures (or versions there of) shared/copied between Linux
and U-Boot and that this data has somehow changed when U-Boot is built
with CONFIG_WATCHDOG #defined. However, I can't find any evidence of
this in the U-Boot code base. The bras instruction above must be key and
I suspect that this is the area causing the problem.

Can anybody offer any advice / explanation of what exactly could be
going on and why? I wouldn't have expected that defining CONFIG_WATCHDOG
would change anything as fundamental as this.

Apologies for the cross-post (U-Boot and m68k kernel dev) but I hope to
get opinions/feedback/help from both communities.

Any assistance very much appreciated indeed.

Many thanks,
--  Matt