Re: [PATCH v2] dt-bindings: ata: Add UniPhier controller binding

[Krzysztof Kozlowski]

On 16/02/2023 18:23, Kunihiko Hayashi wrote:
> On 2023/02/14 18:42, Krzysztof Kozlowski wrote:
>> On 14/02/2023 10:33, Kunihiko Hayashi wrote:
>>> Hi Krzysztof,
>>>
>>> On 2023/02/13 18:10, Krzysztof Kozlowski wrote:
>>>> On 13/02/2023 05:54, Kunihiko Hayashi wrote:
>>>>> Add UniPhier SATA controller compatible string to the platform binding.
>>>>> This controller needs maximum three reset controls.
>>>>>
>>>>> Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@xxxxxxxxxxxxx>
>>>>> ---
>>>>>    .../devicetree/bindings/ata/ahci-platform.yaml  | 17 +++++++++++++++++
>>>>>    1 file changed, 17 insertions(+)
>>>>>
>>>>> Changes since v1:
>>>>> - Restrict resets property changes with compatible strings
>>>>> - Fix maxItems from two to three
>>>>>
>>>>> diff --git a/Documentation/devicetree/bindings/ata/ahci-platform.yaml
>>>>> b/Documentation/devicetree/bindings/ata/ahci-platform.yaml
>>>>> index 7dc2a2e8f598..25dd5ffaa517 100644
>>>>> --- a/Documentation/devicetree/bindings/ata/ahci-platform.yaml
>>>>> +++ b/Documentation/devicetree/bindings/ata/ahci-platform.yaml
>>>>> @@ -45,6 +45,9 @@ properties:
>>>>>                  - marvell,armada-8k-ahci
>>>>>                  - marvell,berlin2-ahci
>>>>>                  - marvell,berlin2q-ahci
>>>>> +              - socionext,uniphier-pro4-ahci
>>>>> +              - socionext,uniphier-pxs2-ahci
>>>>> +              - socionext,uniphier-pxs3-ahci
>>>>>              - const: generic-ahci
>>>>>          - enum:
>>>>
>>>> Top level is saying reset=1, so did you test your bindings?
>>>
>>> Umm, I didn't see any errors on dt_binding_check, anyway I'll add
>>> initial minItems:1 and maxItems:3 on top level first.
>>
>> You need to test also all DTS using these bindings. Yours and others.
>> If you tested the DTS (with proper binding, not one which is basically
>> noop):
>>
>> uniphier-pro4-ace.dtb: sata@65600000: resets: [[27, 12], [27, 28], [37,
>> 3]] is too long
> 
> I've tried updating tools and doing dtbs_check, but I couldn't find this
> error. It seems that this error can't be detected unless there is the
> specified compatible in "select:".
> 
>> BTW, the patch has other errors - just look at the beginning of the
>> file. I cannot see it here in the diff, but when you open the file you
>> should notice it.
> 
> Sorry, but I cannot see anything wrong.
> I'll check the header or something...

If you open the file, you will notice the same compatibles in two
places. In select and in properties. You added your compatibles only to
one place, so not symmetrically.

Best regards,
Krzysztof