Re: [PATCH] dt-bindings: interconnect: qcom,sa8775p-rpmh: Fix incorrectly added reg and clocks

[Odelu Kukatla <odelu.kukatla@oss.qualcomm.com>]

On 11/29/2025 3:16 PM, Krzysztof Kozlowski wrote:
> Commit 8a55fbe4c94d ("dt-bindings: interconnect: add reg and clocks
> properties to enable QoS on sa8775p") claims that all interconnects have
> clocks and MMIO address space, but that is just not true.  Only few
> have.  Bindings should restrict properties and should not allow
> specifying non-existing hardware description, so fix missing constraints
> for 'reg' and 'clocks'.
> 
> Fixes: 8a55fbe4c94d ("dt-bindings: interconnect: add reg and clocks properties to enable QoS on sa8775p")
> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
> ---
>  .../interconnect/qcom,sa8775p-rpmh.yaml       | 31 +++++++++++++++++++
>  1 file changed, 31 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/interconnect/qcom,sa8775p-rpmh.yaml b/Documentation/devicetree/bindings/interconnect/qcom,sa8775p-rpmh.yaml
> index 71428d2cce18..3dbe83e2de3d 100644
> --- a/Documentation/devicetree/bindings/interconnect/qcom,sa8775p-rpmh.yaml
> +++ b/Documentation/devicetree/bindings/interconnect/qcom,sa8775p-rpmh.yaml
> @@ -74,6 +74,37 @@ allOf:
>              - description: aggre UFS CARD AXI clock
>              - description: RPMH CC IPA clock
>  
> +  - if:
> +      properties:
> +        compatible:
> +          contains:
> +            enum:
> +              - qcom,sa8775p-config-noc
> +              - qcom,sa8775p-dc-noc
> +              - qcom,sa8775p-gem-noc
> +              - qcom,sa8775p-gpdsp-anoc
> +              - qcom,sa8775p-lpass-ag-noc
> +              - qcom,sa8775p-mmss-noc
> +              - qcom,sa8775p-nspa-noc
> +              - qcom,sa8775p-nspb-noc
> +              - qcom,sa8775p-pcie-anoc
> +              - qcom,sa8775p-system-noc
> +    then:
> +      properties:
> +        clocks: false
> +
> +  - if:
> +      properties:
> +        compatible:
> +          contains:
> +            enum:
> +              - qcom,sa8775p-clk-virt
> +              - qcom,sa8775p-mc-virt
> +    then:
> +      properties:
> +        reg: false
> +        clocks: false
> +
>  unevaluatedProperties: false
>  
>  examples:

Hi Krzysztof,
Thanks for fixing this, looks good to me.
Though I did not face any issue with dt_binding_check, agreed, the
earlier change was too permissive. Disallowing reg/clocks for the NOCs
without MMIO/clocks and the virtual providers is the right approach.

As you noted, the previous schema effectively allowed describing
non-existent resources, and this change ensures such issues are caught
early by dt_binding_check.

Thanks,
Odelu