Re: [PATCH 1/2] dt-bindings: mfd: Add ChromeOS fingerprint binding

[Stephen Boyd]

Quoting Lee Jones (2022-03-15 04:30:49)
> On Tue, 15 Mar 2022, Krzysztof Kozlowski wrote:
>
> > On 15/03/2022 12:10, Lee Jones wrote:
> > > On Tue, 15 Mar 2022, Krzysztof Kozlowski wrote:
> > >
> > >> On 15/03/2022 00:22, Stephen Boyd wrote:
> > >>> Add a binding to describe the fingerprint processor found on Chromeboks
> > >>> with a fingerprint sensor.
> > >>>
> > >>> Cc: Rob Herring <robh+dt@xxxxxxxxxx>
> > >>> Cc: <devicetree@xxxxxxxxxxxxxxx>
> > >>> Cc: Guenter Roeck <groeck@xxxxxxxxxxxx>
> > >>> Cc: Douglas Anderson <dianders@xxxxxxxxxxxx>
> > >>> Cc: Craig Hesling <hesling@xxxxxxxxxxxx>
> > >>> Cc: Tom Hughes <tomhughes@xxxxxxxxxxxx>
> > >>> Cc: Alexandru M Stan <amstan@xxxxxxxxxxxx>
> > >>> Signed-off-by: Stephen Boyd <swboyd@xxxxxxxxxxxx>
> > >>> ---
> > >>>  .../bindings/mfd/google,cros-ec-fp.yaml       | 89 +++++++++++++++++++
> > >>>  1 file changed, 89 insertions(+)
> > >>>  create mode 100644 Documentation/devicetree/bindings/mfd/google,cros-ec-fp.yaml
> > >>>
> > >>> diff --git a/Documentation/devicetree/bindings/mfd/google,cros-ec-fp.yaml b/Documentation/devicetree/bindings/mfd/google,cros-ec-fp.yaml
> > >>> new file mode 100644
> > >>> index 000000000000..05d2b2b9b713
> > >>> --- /dev/null
> > >>> +++ b/Documentation/devicetree/bindings/mfd/google,cros-ec-fp.yaml
> > >>> @@ -0,0 +1,89 @@
> > >>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> > >>> +%YAML 1.2
> > >>> +---
> > >>> +$id: http://devicetree.org/schemas/mfd/google,cros-ec-fp.yaml#
> > >>
> > >> Why is this in the MFD directory? Is it really a Multi Function Device?
> > >> Description is rather opposite. You also did not CC MFD maintainer.
> > >
> > > A lot of the ChromeOS Embedded Controller support used to be located
> > > in MFD.  There are still remnants, but most was moved to
> > > drivers/platform IIRC.
> > >
> > > Please see: drivers/mfd/cros_ec_dev.c
> >
> > Yes, I know, that part is a MFD. But why the fingerprint controller part
> > is MFD? To me it is closer to input device.
>
> It's tough to say from what I was sent above.
>
> But yes, sounds like it.
>
> We do not want any device 'functionality' in MFD ideally.
>

I put it next to the existing cros-ec binding. The existing binding is
there because of historical reasons as far as I know. Otherwise it
didn't seem MFD related so I didn't Cc mfd maintainer/list. New file
additions don't usually conflict with anything and this is in the
bindings directory so the driver side maintainer would be picking up the
binding.