Message-ID: <b1d541c1-296a-4b56-bea3-52e5becadf0e@kernel.org>
Date: Tue, 24 Dec 2024 10:00:17 +0100
From: Krzysztof Kozlowski <krzk@...nel.org>
To: Mathieu Dubois-Briand <mathieu.dubois-briand@...tlin.com>,
 Lee Jones <lee@...nel.org>, Rob Herring <robh@...nel.org>,
 Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley
 <conor+dt@...nel.org>, Kamel Bouhara <kamel.bouhara@...tlin.com>,
 Linus Walleij <linus.walleij@...aro.org>, Bartosz Golaszewski
 <brgl@...ev.pl>, Dmitry Torokhov <dmitry.torokhov@...il.com>,
 Uwe Kleine-König <ukleinek@...nel.org>
Cc: devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
 linux-gpio@...r.kernel.org, linux-input@...r.kernel.org,
 linux-pwm@...r.kernel.org, Grégory Clement
 <gregory.clement@...tlin.com>,
 Thomas Petazzoni <thomas.petazzoni@...tlin.com>
Subject: Re: [PATCH 2/8] dt-bindings: Add MAX7360 subdevices

On 23/12/2024 16:20, Mathieu Dubois-Briand wrote:
> On Sat Dec 21, 2024 at 9:34 PM CET, Krzysztof Kozlowski wrote:
>> On 19/12/2024 17:21, Mathieu Dubois-Briand wrote:
>>> ---
>>>  .../devicetree/bindings/gpio/max7360-gpio.yaml     | 96 ++++++++++++++++++++++
>>>  .../devicetree/bindings/input/max7360-keypad.yaml  | 67 +++++++++++++++
>>>  .../devicetree/bindings/input/max7360-rotary.yaml  | 52 ++++++++++++
>>>  .../devicetree/bindings/pwm/max7360-pwm.yaml       | 35 ++++++++
>>>  4 files changed, 250 insertions(+)
>>
>>
>> I don't understand how this patchset was split. MFD binding cannot be
>> empty and cannot be before child devices.
>>
> 
> Ok, my bad. So I believe squashing both dt-bindings commit should fix
> this.
> 
>>> diff --git a/Documentation/devicetree/bindings/gpio/max7360-gpio.yaml b/Documentation/devicetree/bindings/gpio/max7360-gpio.yaml
>>> new file mode 100644
>>> index 000000000000..3c006dc0380b
>>> --- /dev/null
>>> +++ b/Documentation/devicetree/bindings/gpio/max7360-gpio.yaml
>>> @@ -0,0 +1,96 @@
>>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>>> +%YAML 1.2
>>> +---
>>> +$id: http://devicetree.org/schemas/gpio/max7360-gpio.yaml#
>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>> +
>>> +title: Maxim MAX7360 GPIO controller
>>> +
>>> +maintainers:
>>> +  - Kamel Bouhara <kamel.bouhara@...tlin.com>
>>> +  - Mathieu Dubois-Briand <mathieu.dubois-briand@...tlin.com>
>>> +
>>> +description: |
>>> +  Maxim MAX7360 GPIO controller, in MAX7360 MFD
>>> +  https://www.analog.com/en/products/max7360.html
>>> +
>>> +properties:
>>> +  compatible:
>>> +    enum:
>>> +      - maxim,max7360-gpio
>>> +      - maxim,max7360-gpo
>>
>> Why? What are the differences?
>>
> 
> Ok, so maybe my approach here is completely wrong. I'm not sure what
> would be the best way to describe the device here, if you have any
> suggestion I would be happy to use it. Let me try to summarize the GPIO
> setup of the chip.
> 
> First we have two series of GPIOs on the chips, which I tend to think
> about as two separate "banks". Thus two separate subnodes of the max7360
> node.

First, splitting MFD device into multiple children is pretty often wrong
approach because it tries to mimic Linux driver design.

Such split in DT makes sense if these are really separate blocks, e.g.
separate I2C addresses, re-usable on different designs.

In this case Functional Block Diagram shows separate blocks, but still
the same I2C block. This can be one device. This can be also two devices
if that's easier to represent in DT.

But in any case binding description should explain this.

> 
> - On one side we have what I refer to as GPIOs, here with
>   maxim,max7360-gpio:
>   - PORT0 to PORT7 pins of the chip.
>   - Shared with PWM and rotary encoder functionalities. Functionality
>     selection can be made independently for each pin. This selection is
>     not described here. Runtime will have to ensure the same pin is not
>     used by two drivers at the same time. E.g. we cannot have at the
>     same time GPIO4 and PWM4.
>   - Supports input and interrupts.
>   - Outputs may be configured as constant current.
>   - 8 GPIOS supported, so ngpios maximum is 8. Thinking about it now, we
>     should probably also set minimum to 8, I don't see any reason to
>     have ngpios set to something less.
> 
> On the other side, we have what I refer to as GPOs, here with
> maxim,max7360-gpo compatible:
>   - COL2 to COL7 pins of the chip.
>   - Shared with the keypad functionality. Selections is made by
>     partitioning the pins: first pins for keypad columns, last pins for
>     GPOs. Partition is described here by ngpios and on keypad node by
>     keypad,num-columns. Runtime will have to ensure values are coherent
>     and configure the chip accordingly.
>   - Only support outputs.
>   - No support for constant current mode.
>   - Supports 0 to 6 GPOs, so ngpios maximum is 6.
> 
>>> +
>>> +  gpio-controller: true
>>> +
>>> +  "#gpio-cells":
>>> +    const: 2
>>> +
>>> +  ngpios:
>>> +    minimum: 0
>>> +    maximum: 8
>>
>> Why this is flexible?
>>
> 
> I believe this makes sense, as this keypad/gpos partition really changes
> the actual number of GPIOS. Yet we could argue that this is just runtime
> configuration. Tell me what you think about it, if you think this should
> be a fixed value, I will find a way.

Depends whether this is actual runtime configuration. If you configure
keypad in DT, then the pins go away from GPIOs (especially considering
that board might have these pins really connected to keypad). Anyway,
explain this briefly in binding description.

> 
Best regards,
Krzysztof

Powered by blists - more mailing lists