Re: [PATCH 2/2] ARM: dt: tegra: cardhu: register core regulator tps65911

[Stephen Warren <swarren@wwwdotorg.org>]

On 06/01/2012 02:40 PM, Mark Brown wrote:
> On Fri, Jun 01, 2012 at 01:23:24PM -0600, Stephen Warren wrote:
> 
>> However, Mark warned that changing this would be a bit painful 
>> because there are already users of the existing scheme. It looks 
>> like that's only tps65910 (which we haven't started using yet), 
>> db8500, and ab8500, so probably not that big a deal.
> 
> No, there's a bunch of others - some queued for -next, others open
>  coding the same scheme.  Any device with more than one regulator
> in a node should be using the same scheme.
> 
>> We could either augment struct of_regulator_match with an
>> integer ID field for each regulator (which would perhaps make it
>> slightly painful to write the nodes and keep the IDs matched up),
>> or add a new property
> 
> No, that's awful.  How's anyone supposed to read stuff like that? 
> The interrupt bindings are a disaster, not a model.
> 
>> to each regulator provider node e.g. regulator-id which
>> contained the name that the regulator driver knows the regulator
>> as (which would match struct of_regulator_match.name), since the
>> existing regulator-name property is used for semantically
>> different purposes.
> 
> Oh, ick.  This isn't nice.  If anything I'd be more inclined to
> put a named property in there and have drivers look for its
> presence. The presence of multiple name properties isn't nice.

Could you expand on "named property" a bit; I'm not quite sure what
you're getting at - literally a property with name "named" (which
would be the same as regulator-id under just a different property
name), or ...?

>>> vdd1_reg: regulator@0 {
> 
> Can't we use the right hand side of this?  It appears to just be 
> syntactic sugar without any current meaning.

The stuff to the right of @ is the "unit address" and must match the
value in the reg property. Using that was the first proposal I had
above (which I also didn't like as much)