[PATCH 1/3] net: smc1111: allow platform specific accessors

[Sascha Hauer]

On Thu, Jan 29, 2015 at 03:01:03PM +0100, robert.jarzmik at free.fr wrote:
> ----- Mail original -----
> De: "Sascha Hauer" <s.hauer at pengutronix.de>
> À: "Robert Jarzmik" <robert.jarzmik at free.fr>
> Cc: barebox at lists.infradead.org
> Envoyé: Jeudi 29 Janvier 2015 10:47:20
> Objet: Re: [PATCH 1/3] net: smc1111: allow platform specific accessors
> 
> > Yeah I know, connecting the 91c111 inspired the board designers to many
> > funny things. Passing function pointers in platform_data is not very
> > future proof since this doesn't work with device tree. I see that the
> > Linux driver has device tree support. Does this binding work for you?
> 
> Well, it can't, because lubbock doesn't support device-tree yet, it's on
> my todo list for the kernel (and barebox of course). That means there is
> no way to even build a device-tree enabled kernel for lubbock ... yet.
> 
> Now the binding would work with :
>  - the address shift being declared (as in kernel driver)
>  - a set of 16 bit accessors being declared in barebox's smcs9111.c
>    driver (and not in platform specific accessors)
> 
> Would you rather have me change the approach with a new set of accessor
> in smcs9111.c, and the shift as a parameter (both platform_data and device-
> tree) ?

That sounds better. Using accessors in platform_data really is more
flexible, but when switching to device tree we'll need a width/shift
approach anyway.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |