[PATCH v3 20/21] dt-bindings: pinctrl: mediatek: mt7620: split binding

[Arınç ÜNAL]

On 20.03.2023 18:00, AngeloGioacchino Del Regno wrote:
> Il 17/03/23 22:30, arinc9.unal at gmail.com ha scritto:
>> From: Arınç ÜNAL <arinc.unal at arinc9.com>
>>
>> The MT7628 and MT7688 SoCs contain different pin muxing information,
>> therefore, should be split. This can be done now that there are 
>> compatible
>> strings to distinguish them from other SoCs.
>>
>> Split the schema out to mediatek,mt76x8-pinctrl.yaml.
>>
>> Remove mediatek,mt76x8-pinctrl from mt7620.
>>
>> Signed-off-by: Arınç ÜNAL <arinc.unal at arinc9.com>
>> Reviewed-by: Rob Herring <robh at kernel.org>
>> ---
>>   .../pinctrl/mediatek,mt7620-pinctrl.yaml      | 379 +--------------
>>   .../pinctrl/mediatek,mt76x8-pinctrl.yaml      | 450 ++++++++++++++++++
> 
> AFAIK, wildcards aren't allowed, so the new file should instead be called
> "mediatek,mt7628-pinctrl.yaml", the compatible string 
> "ralink,mt76x8-pinctrl"
> should be deprecated (but still present as to not break ABI), and the 
> driver
> updated to accept the compatible string "ralink,mt7628-pinctrl".
> 
> At that point, you could update the devicetrees as well: for MT7628, you'd
> have a node using
> 
> compatible = "ralink,mt7628-pinctrl";
> 
> while for MT7688, you'd have
> 
> compatible = "ralink,mt7688-pinctrl", "ralink,mt7628-pinctrl";
> 
> If you don't want to go through those lengths, you could still do the
> cleanup that you wanted to perform, but with a filename containing no
> wildcards - anyway.

I was able to have it accepted by Rob, and the patches are already 
applied so I'm going to leave it as is for now.

Arınç