[PATCH 2/3] PWM: vt8500: Update vt8500 PWM driver support

[Russell King - ARM Linux]

On Tue, Oct 23, 2012 at 11:22:47AM +0200, Thierry Reding wrote:
> On Tue, Oct 23, 2012 at 09:41:46PM +1300, Tony Prisk wrote:
> > Further to the discussion, my preference is still for of_clk_get()
> > (although I've changed the patch anyway as you saw because it makes no
> > difference in this case) :)
> > 
> > clk_get(x, NULL) and devm_clk_get(x, NULL) both seems like 'hacks' to
> > allow platforms to convert to DT without having to update all their
> > drivers first. It only allows the first (default) clock, as your pointed
> > out. Getting a 2nd... clock relies on an optional property in DT (which
> > again, seems like it is there to support 'old' drivers) which allows you
> > to request clocks by name.
> > 
> > of_clk_get() on the other hand seems like a properly native DT function.
> > You don't need to know anything about the clock, as long as the correct
> > clock is specified in the correct order as documented by the binding.
> > Relying on 'pre-OF' code for a OF-only driver also seems
> > counter-intuitive.
> 
> I do agree with those arguments. What I was saying is that for drivers
> which aren't DT only, of_clk_get() is not an option and that maybe
> others would be encouraged by the example to not use the generic APIs
> even if their driver could be used in non-DT setups. But maybe I'm
> worrying needlessly.
> 
> That said, maybe somebody with a broader view of things like Arnd
> (Cc'ed) could share his thoughts.

As I have already said, the way the DT bindings were done for the clk
stuff was wrong.  A little thought put into it would've come up with
a much better solution which wouldn't have needed of_clk_get() at all.

How?

The arguments for clk_get() are:
1. the struct device, which you can get the OF-node from.
2. a _device_ _specific_ _clock_ _input_ _name_ (or NULL if there's only
   one.)

So, we have something that defines a hardware clock input name, which
can be used to generate a property name for OF.  So, what _could_ have
been done is this:

	clock-<input-name> = <&provider-node clk-output-index>;

where the property name is generated by:

	snprintf(prop, sizeof(prop), "clk-%s", name ? name : "default");

So I continue to assert that our current design is wrong - and it will
cause driver authors to pointlessly have to make a choice at every stage
between DT and non-DT based systems.