Re: [PATCH v4 08/17] dt-bindings: fsi: ast2600-fsi-master: Convert to json-schema

[Krzysztof Kozlowski <krzk@xxxxxxxxxx>]

On 01/05/2024 18:12, Eddie James wrote:
> 
> On 4/30/24 02:04, Krzysztof Kozlowski wrote:
>> On 29/04/2024 23:01, Eddie James wrote:
>>> Convert to json-schema for the AST2600 FSI master documentation.
>> Please mention all the changes from pure conversion.
> 
> 
> Sure.
> 
> 
>>
>>> Signed-off-by: Eddie James <eajames@xxxxxxxxxxxxx>
>>> ---
>>> Changes since v3:
>>>   - Remove quotes around compatible strings
>>>   - Re-order allOf to below required
>>>   - Add child node in the example
>>>   - Change commit message to match similar commits
>>>
>>>   .../fsi/aspeed,ast2600-fsi-master.yaml        | 81 +++++++++++++++++++
>>>   .../bindings/fsi/fsi-master-aspeed.txt        | 36 ---------
>>>   2 files changed, 81 insertions(+), 36 deletions(-)
>>>   create mode 100644 Documentation/devicetree/bindings/fsi/aspeed,ast2600-fsi-master.yaml
>>>   delete mode 100644 Documentation/devicetree/bindings/fsi/fsi-master-aspeed.txt
>>>
>>> diff --git a/Documentation/devicetree/bindings/fsi/aspeed,ast2600-fsi-master.yaml b/Documentation/devicetree/bindings/fsi/aspeed,ast2600-fsi-master.yaml
>>> new file mode 100644
>>> index 000000000000..fcf7c4b93b78
>>> --- /dev/null
>>> +++ b/Documentation/devicetree/bindings/fsi/aspeed,ast2600-fsi-master.yaml
>>> @@ -0,0 +1,81 @@
>>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>>> +%YAML 1.2
>>> +---
>>> +$id: http://devicetree.org/schemas/fsi/aspeed,ast2600-fsi-master.yaml#
>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>> +
>>> +title: Aspeed FSI master
>>> +
>>> +maintainers:
>>> +  - Eddie James <eajames@xxxxxxxxxxxxx>
>>> +
>>> +description:
>>> +  The AST2600 and later contain two identical FSI masters. They share a
>>> +  clock and have a separate interrupt line and output pins.
>>> +
>>> +properties:
>>> +  compatible:
>>> +    enum:
>>> +      - aspeed,ast2600-fsi-master
>>> +      - aspeed,ast2700-fsi-master
>> There was no such compatible before.
>>
>> How does this even validate? Where is fsi-master? You dropped a
>> compatible without any explanation.
> 
> 
> I can make it a separate change to add ast2700.
> 
> 
> I suppose I don't understand having two compatibles... Aspeed master 
> shouldn't use "fsi-master" as that is too generic, right? Why wouldn't 

Not necessarily, depends. Dropping it silently is confusing. What about
other users? firmware, bootloaders, out-of-tree, other OS? Did you
investigate all of them?

> it validate? Devicetrees using "fsi-master" also use 
> "aspeed,ast2600-fsi-master" so they should be OK...

No, because the compatibles do not match. Run validation and you will
see the errors.

I am fine with dropping such compatible, which is not used by current
kernel ABI, but first DTS must be fixed and second some explanation and
justification is needed.

Best regards,
Krzysztof