Re: [PATCH v2 1/3] dt-bindings: net: qcom: ethernet: Add interconnect properties

[Sagar Cheluvegowda <quic_scheluve@quicinc.com>]

On 6/26/2024 7:53 AM, Andrew Halaney wrote:
> On Tue, Jun 25, 2024 at 04:49:28PM GMT, Sagar Cheluvegowda wrote:
>> Add documentation for the interconnect and interconnect-names
>> properties required when voting for AHB and AXI buses.
>>
>> Suggested-by: Andrew Halaney <ahalaney@redhat.com>
>> Signed-off-by: Sagar Cheluvegowda <quic_scheluve@quicinc.com>
>> ---
>>  Documentation/devicetree/bindings/net/qcom,ethqos.yaml | 8 ++++++++
>>  1 file changed, 8 insertions(+)
>>
>> diff --git a/Documentation/devicetree/bindings/net/qcom,ethqos.yaml b/Documentation/devicetree/bindings/net/qcom,ethqos.yaml
>> index 6672327358bc..b7e2644bfb18 100644
>> --- a/Documentation/devicetree/bindings/net/qcom,ethqos.yaml
>> +++ b/Documentation/devicetree/bindings/net/qcom,ethqos.yaml
>> @@ -63,6 +63,14 @@ properties:
>>  
> 
> Does it make sense to make these changes in snps,dwmac.yaml since you're
> trying to do this generically for stmmac? I don't poke bindings super
> often so might be a silly question, the inheritance of snps,dwmac.yaml
> into the various platform specific bindings (qcom,ethqos.yaml) would
> then let you define it once in the snps,dwmac.yaml right?
> 
>>    dma-coherent: true
>>  
>> +  interconnects:
>> +    maxItems: 2
>> +
>> +  interconnect-names:
>> +    items:
>> +      - const: axi
>> +      - const: ahb
> 
> Sorry to bikeshed, and with Krzysztof's review on this already its
> probably unnecessary, but would names like cpu-mac and mac-mem be
> more generic / appropriate? I see that sort of convention a lot in the
> other bindings, and to me those read really well and are understandable.

I agree with changing the names to "cpu-mac" and "mac-mem" in that
way the properties are more understandable.
@Krzysztof Kozlowski let me know your opinion on the same.