From mboxrd@z Thu Jan  1 00:00:00 1970
From: Albert ARIBAUD <albert.u.boot@aribaud.net>
Date: Wed, 11 Jun 2014 08:45:58 +0200
Subject: [U-Boot] [PATCH] arm: Allow u-boot to run from offset base
 address
In-Reply-To: <20140611044928.25167382C16@gemini.denx.de>
References: <1400105145-6628-1-git-send-email-drambo@broadcom.com>
	<1400107265.16796.8.camel@yellow> <5374CD55.3010703@broadcom.com>
	<20140515152119.9939F38047D@gemini.denx.de>
	<5374E64B.1060104@broadcom.com>
	<20140515191957.74A853804B6@gemini.denx.de>
	<E1WorXt-0005yz-Lp@janus> <538367B7.5090108@broadcom.com>
	<E1WrMdj-0001wW-DR@janus> <538D18C5.206@broadcom.com>
	<E1Wtwjm-00064W-6k@janus> <53961CFE.4090106@broadcom.com>
	<E1WuEQN-0002pK-El@janus> <539746C4.9040004@broadcom.com>
	<20140610181341.7E10138131B@gemini.denx.de>
	<53975EC2.1080209@broadcom.com>
	<20140610203500.0D37038131B@gemini.denx.de>
	<53979199.5010100@broadcom.com>
	<20140611044928.25167382C16@gemini.denx.de>
Message-ID: <E1WucIJ-0001Lv-5Q@janus>
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

Hi Wolfgang,

On Wed, 11 Jun 2014 06:49:28 +0200, Wolfgang Denk <wd@denx.de> wrote:

> Dear Steve,
> 
> In message <53979199.5010100@broadcom.com> you wrote:
> > 
> > OK - I think that one of the alternate proposals would be to 
> > conditionally reserve a "32 byte block" prior to the _start symbol (in 
> > "arch/arm/cpu/armv8/start.S") which would then be filled in by a 
> > post-processing step... This could be implemented by:
> 
> Yes, that illustrates the idea.  However, this implementation suffers
> from the use of an #ifdef where none is actually needed.  Instead, you
> can create your own source file which defines the header; this could
> be then even in it's own segment, say:
> 
> your_header.c:
> 
> struct your_header {
> 	u_int32[8];
> } your_header __attribute__ ((__section__ (".your_hdr")));
> 
> All that is needed then is to make the linker place this segment in
> front of the text segment.
> 
> This avoids an ugly #ifdef, and also modifications in the common code.

Agreed and seconded.

Plus, using a dedicated 'header' section and a separate C source file
for the header makes it automatic that if no input 'header' section
were provided then no output 'header' section would be emitted; IOW,
we would not need two different linker scripts, a single one would be
useable for both 'headerless' and 'headerful' image types.

Also, the alignment constraint should be configurable.

> Best regards,
> 
> Wolfgang Denk

Amicalement,
-- 
Albert.