Re: [PATCH] of: spi: Export single device registration method and accessors (v2)

[Pantelis Antoniou <panto-wVdstyuyKrO8r51toPun2/C9HSW9iNxf@public.gmane.org>]

Hi

> On Oct 29, 2014, at 12:14 , Mark Brown <broonie-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org> wrote:
> 
> On Wed, Oct 29, 2014 at 10:40:37AM +0200, Pantelis Antoniou wrote:
>> Dynamically inserting spi device nodes requires the use of a single
>> device registration method. Rework and export it.
>> 
>> Methods to lookup a device/master using a device node are added
>> as well, of_find_spi_master_by_node() & of_find_spi_device_by_node().
> 
> Why do we need to do this - I would expect that adding nodes would
> trigger parsing in the same way we normally do it.  Where is the user
> and how does it know that it's handling a SPI node?
> 
> This feels like there is an abstraction problem somewhere, whatever code
> is supposed to use this is going to need to be taught about each
> individual bus which is going to be tedious, I would expect that we'd
> have something like the bus being able to provide a callback which will
> get invoked whenever a new node appears on the parent node for the bus.
> 

There’s a whole patchset that does exactly this. 
Look at "OF: spi: Add OF notifier handler” and you’ll where this is used.

>> Changes since v1:
>> * Brown paper bug with parameter on of_register_spi_device().
> 
> Don't include noise like this in the changelog, put it after --- like
> SubmittingPatches says.  Please also try to keep your CC list sane,
> CCing random people just means that you're increasing the volume of mail
> they have to process.  I'm surprised kernel.org accepts so many CCs.
> 
> I have to say I don't recall ever seeing v1...

All of them are in the CC list for a reason. 

Regards

— Pantelis