[PATCH 1/6] ARM: mx5: use config to define boot related addresses

Richard Zhao richard.zhao at freescale.com
Fri Dec 10 00:21:44 EST 2010


On Thu, Dec 09, 2010 at 09:00:52AM +0000, Russell King - ARM Linux wrote:
> On Thu, Dec 09, 2010 at 09:52:55AM +0100, Uwe Kleine-König wrote:
> > Hell Russell,
> > 
> > On Thu, Dec 09, 2010 at 08:44:30AM +0000, Russell King - ARM Linux wrote:
> > > On Thu, Dec 09, 2010 at 08:04:11AM +0100, Uwe Kleine-König wrote:
> > > > > diff --git a/arch/arm/mach-mx5/Makefile.boot b/arch/arm/mach-mx5/Makefile.boot
> > > > > index 9939a19..388d2e8 100644
> > > > > --- a/arch/arm/mach-mx5/Makefile.boot
> > > > > +++ b/arch/arm/mach-mx5/Makefile.boot
> > > > > @@ -1,3 +1,3 @@
> > > > > -   zreladdr-y	:= 0x90008000
> > > > > -params_phys-y	:= 0x90000100
> > > > > -initrd_phys-y	:= 0x90800000
> > > > > +   zreladdr-$(CONFIG_SOC_IMX51)	:= 0x90008000
> > > > > +params_phys-$(CONFIG_SOC_IMX51)	:= 0x90000100
> > > > > +initrd_phys-$(CONFIG_SOC_IMX51)	:= 0x90800000
> > > > That is one of the places that is not multi-soc capable.  You can make
> > > > it 66% less worse by just removing params_phys and initrd_phys.
> > > 
> > > Which then destroys the ability to use the bootp veneer which combines
> > > a kernel and initrd.
> > I am aware of that.  I think it's OK to add the assignments when they
> > are needed the first time which I don't expect to happen soon if at all.
> 
> That depends who is trying to use it.  If it's someone who isn't the
> original platform developer, they may give up with it if the definitions
> aren't provided.
Hi uwe,

IMO, removing params_phys and initrd_phys is just a work around. It's not 
the final way to fix multi-SoC support. Do you insist on removing it?

Thanks
Richard




More information about the linux-arm-kernel mailing list