Re: [PATCH v3] dt-bindings: mfd: Convert ChromeOS EC bindings to json-schema

From: Enric Balletbo i Serra <enric.balletbo@collabora.com>
To: Rob Herring <robh@kernel.org>
Cc: Ikjoon Jang <ikjn@chromium.org>,
	devicetree@vger.kernel.org,
	Dmitry Torokhov <dmitry.torokhov@gmail.com>,
	Gwendal Grignou <gwendal@chromium.org>,
	Lee Jones <lee.jones@linaro.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Benson Leung <bleung@chromium.org>,
	Guenter Roeck <groeck@chromium.org>,
	Jiri Kosina <jikos@kernel.org>,
	Benjamin Tissoires <benjamin.tissoires@redhat.com>,
	Nicolas Boitchat <drinkcat@chromium.org>,
	linux-input@vger.kernel.org
Subject: Re: [PATCH v3] dt-bindings: mfd: Convert ChromeOS EC bindings to json-schema
Date: Mon, 27 Jan 2020 17:25:28 +0100	[thread overview]
Message-ID: <ef653218-4a59-73e1-186e-9b62b8ed6dae@collabora.com> (raw)
In-Reply-To: <20200127161206.GB30843@bogus>

Hi Rob,

On 27/1/20 17:12, Rob Herring wrote:
> On Mon, Jan 27, 2020 at 04:57:13PM +0100, Enric Balletbo i Serra wrote:
>> Hi Ikjoon,
>>
>> On 21/1/20 8:47, Ikjoon Jang wrote:
>>> Convert the ChromeOS EC bindings to json-schema.
>>>
>>> Signed-off-by: Ikjoon Jang <ikjn@chromium.org>
>>> ---
>>> v3: node name changed in rpmsg example
>>> v2: cleanup description, fix typos, remove LPC, add add RPMSG example
>>> ---
>>>  .../devicetree/bindings/mfd/cros-ec.txt       |  76 ------------
>>>  .../devicetree/bindings/mfd/cros-ec.yaml      | 111 ++++++++++++++++++
>>>  2 files changed, 111 insertions(+), 76 deletions(-)
>>>  delete mode 100644 Documentation/devicetree/bindings/mfd/cros-ec.txt
>>>  create mode 100644 Documentation/devicetree/bindings/mfd/cros-ec.yaml
>>>
>>
>> Still in mfd directory, this is not a MFD driver, please move to chrome/cros-ec.yaml
> 
> Please don't. Looks to me like the chrome EC has multiple functions and 
> is a device.
> 

This binding is part of the Chrome EC core and we explicitly moved all this out
from mfd subsytem in the past because wasn't really fit as mfd driver.

We still have one device driver (cros_ec_dev) in mfd, this is the one that
instantiates all the subdevices but the above binding is not related to this
one. All the core/bus part is in platform/chrome, looks to me more coherent have
the binding accordingly.

Anyway I'll let you the final decision, nothing that worries me too much.

Thanks,
 Enric

> Bindings directories are often 1:1 with kernel driver directories, but 
> not always.
> 
> Rob
> 