Arnd Bergmann <arnd@arndb.de> writes:

> On Thursday 26 March 2015, Robert Jarzmik wrote:
>> 
>> Greg Kroah-Hartman <gregkh@linuxfoundation.org> writes:
>> 
>> > On Fri, Feb 20, 2015 at 05:02:57PM +0100, Robert Jarzmik wrote:
>> >> If there is no solution, I'll fallback through arch/arm/plat-pxa, not very nice,
>> >> but it has to land somewhere, I don't want lubbock to remain broken.
>> >
>> > drivers/platform/arm ?
>> Most certainly.
>> 
>> I'll submit that to drivers/platform/arm/pxa, and maintain that pxa tree. As for
>> drivers/platform/arm, do you want also maintainers to step up, or will you take
>> the review/merge burden ?
>> 
>
> I'd much prefer not to add drivers/platform/arm, which would make it too easy
> to add random stuff there. What is the problem with leaving it in mach-pxa?
Hi Arnd,

It's not as much a problem as a generic question : does a driver belong to
arch/* ?

Personaly it would have been far simpler for me to have it through the pxa tree,
but I want to be sure it's the right place. Others will follow, pxa mainstone is
such a candidate.

I was thinking so far that arch/arm/mach-* was for machine description,
ie. wirings, interconnections, initial setup etc ... The "driver" part, ie. code
really driving dynamics in IPs was as per my understanding in drivers/...

Now I can create arch/arm/mach-pxa/lubbock_cplds.c, that won't make any
difference to me, provided that it's the right thing to do.

Cheers.

-- 
Robert