Re: [PATCH RFC 2/3] DT: leds: Add binding for the ISSI IS31FL32xx family of LED drivers

["David Rivshin (Allworx)" <drivshin.allworx@gmail.com>]

On Tue, 23 Feb 2016 19:15:08 -0600
Rob Herring <robh@kernel.org> wrote:

> On Tue, Feb 23, 2016 at 01:17:24PM -0500, David Rivshin (Allworx) wrote:
> > From: David Rivshin <drivshin@allworx.com>
> > 
> > This adds a binding description for the is31fl3236/35/18/16 I2C LED
> > drivers.
> > 
> > Signed-off-by: David Rivshin <drivshin@allworx.com>
> > ---
> >  .../devicetree/bindings/leds/leds-is31fl32xx.txt   | 51 ++++++++++++++++++++++
> >  1 file changed, 51 insertions(+)
> >  create mode 100644 Documentation/devicetree/bindings/leds/leds-is31fl32xx.txt  
> 
> Acked-by: Rob Herring <robh@kernel.org>

Thanks, Rob. I just want to double check whether you noticed some
binding-related questions I had in the coverletter [1]. In hindsight
I should probably have included them in the patch comments as well so 
they'd show up in patchwork. For convenience, I'll repeat them here:

I choose to have 'reg' be 1-based in order to follow the hardware 
documentation. It seems 0-based is the normal choice, although HW
docs also usually use the 0-based numbering. Perhaps being consistent 
with other bindings is more important than being consistent with the 
datasheet's numbering?

I used the 'reg' property for the LED channel, as it seemed ePAPR 
required that name. I also considered naming the property 'chan', 
and not pretending that it represented a bus address at all (and
then removing the @n from the subnode names). That would solve the
'reg' question above as a side-effect, but would be inconstant with
other LED bindings (tca6507, pca963x, tlc59108, etc).

Note that the recently-added (and only in for-next) SN3218 driver 
uses a 0-based 'reg' property, and the SN3218 has the same HW doc
numbering as the IS31FL32xx family (indeed the IS31FL3218 appears 
to be a rebranded SN3218). So perhaps that sets the precedent if 
there was not one before?

Any guidance you could offer would be appreciated (especially since
this is my first from-scratch binding, and I'd rather not form any
bad habits).

[1] http://www.spinics.net/lists/linux-leds/msg05564.html
    http://thread.gmane.org/gmane.linux.leds/4530