Re: [PATCH/RFC] Add cmd_wrap_dts rules to arch/powerpc/boot/Makefile

[David Gibson <david@gibson.dropbear.id.au>]

On Wed, Jan 10, 2007 at 05:40:48PM -0700, Mark A. Greer wrote:
> On Thu, Oct 19, 2006 at 12:51:56PM -0700, Mark A. Greer wrote:
> > On Thu, Oct 19, 2006 at 09:03:18AM -0500, Jon Loeliger wrote:
> > > So, like, the other day "Mark A. Greer" mumbled:
> > > > 
> > > > This would be used by the sandpoint, say, by adding the following rules
> > > > to the Makefile in the appropriate places:
> > > > 
> > > > 	$(obj)/zImage.sandpoint: vmlinux $(wrapperbits)
> > > > 		$(call cmd,wrap_dts,sandpoint)
> > > > 
> > > > 	$(obj)/zImage.initrd.sandpoint: vmlinux $(wrapperbits)
> > > > 		$(call cmd,wrap_dts_initrd,sandpoint)
> > > > 
> > > > and:
> > > > 
> > > > 	image-$(CONFIG_SANDPOINT)               += zImage.sandpoint
> > > > 
> > > > Comments?
> > > 
> > > Isn't there a way to do this without hardcoding the
> > > platform name in the makefiles directly?  I fear for
> > > the one-per-platform effect that is yet to come with
> > > the conversion of many of the legacy boards.
> 
> Okay, its been a while so let's see if anyone remembers and/or cares
> anymore.  :)
> 
> How about something like the patch below?
> 
> Basically, if you want a dtb wrapped up in your zImage, set
> CONFIG_DEFAULT_DTB=y and also define CONFIG_DTS to point where your dts
> file is.
> 
> Does that meet everyones requirements?

Erm.. actually from what I could tell, the way the wrapper script
invokes dtc will make a mess if you're building with a different
object tree from the source tree.  I think the other suggested patches
which have a cmd_wrap_dtb and a spearate make rule to invoke dtc are a
better shot, and it might be a good idea to rip the dtc invocation out
of wrapper entirely.

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson