Re: [PATCH net-next v4 3/6] dt-bindings: net: remove outdated hisilicon-femac

[Krzysztof Kozlowski]

On 27/02/2024 02:43, Yang Xiwen wrote:
> On 2/26/2024 3:50 PM, Krzysztof Kozlowski wrote:
>> On 22/02/2024 13:43, Yang Xiwen via B4 Relay wrote:
>>> From: Yang Xiwen <forbidden405@xxxxxxxxxxx>
>>>
>>> The user documented(Hi3516) is not found in current kernel anymore. And
>>> it seems this SoC has been EOL for a long time. While at it, Remove this
>>> binding entirely due to recent driver changes.
>>
>> This removes documented compatibles, which:
>> 1. should cause checkpatch issues,
>> 2. is not what we want.
>>
>> Driver is still there, so first the driver would have to be removed...
>> but then why do you add it immediately back? No, it does not work like that.
> 
> I has downloaded TRM for this old Hi3519 SoC. So i think with the new
> driver, it still works for Hi3519 FEMAC controller. So i'll rewrite
> these 3 patches and describe the whole thing in commit logs. i.e. the
> old binding is wrong, inconsistent with real hardware, etc..
> 
> So you mean conversion to YAML first, in a separate commit. And then do
> necessary edition to match the real hardware? I believe Hi3519 FEMAC and
> Hi3798MV200 FEMAC are compatible in hardware. So the old binding is
> wrong, missing critical hardware resources.

If you remove binding, then it has to be done in reverse order than
adding it (see DT submitting patches). However removing binding and
re-adding it, is not what you think. It's a obfuscated way to make a
change, so better just make a change and explicitly say why and what is
happening

Best regards,
Krzysztof