Re: [PATCH v4 6/8] ASoC: dt-bindings: qcom: Add Hawi LPASS macro codecs

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] dt-bindings: qcom,lpass-tx/va-macro: contradictory lpr clock substitution
--

commit 9ee86509828b9690a8234588233f08d334886d9b
Author: Prasad Kumpatla <prasad.kumpatla@oss.qualcomm.com>

ASoC: dt-bindings: qcom: Add Hawi LPASS macro codecs

This commit documents compatibles for the RX, TX, and VA macro codecs found
in the Low Power Audio SubSystem of the Qualcomm Hawi SoC. It specifies
standalone compatibles with their own clock constraints for all three macros.

> the RX macro uses the 'lpr' clock instead of the 'macro'
> clock, the TX macro is based on Bolero v4.0, and the VA macro takes an
> additional "heartbeatpulse" clock sourced from the LPASS VA CSR rate
> generator.

[ ... ]

> diff --git a/Documentation/devicetree/bindings/sound/qcom,lpass-tx-macro.yaml b/Documentation/devicetree/bindings/sound/qcom,lpass-tx-macro.yaml
> --- a/Documentation/devicetree/bindings/sound/qcom,lpass-tx-macro.yaml
> +++ b/Documentation/devicetree/bindings/sound/qcom,lpass-tx-macro.yaml
[ ... ]
> @@ -142,6 +143,24 @@ allOf:
>              - const: dcodec
>              - const: fsgen
>  
> +  - if:
> +      properties:
> +        compatible:
> +          contains:
> +            enum:
> +              - qcom,hawi-lpass-tx-macro
> +    then:
> +      properties:
> +        clocks:
> +          minItems: 4
> +          maxItems: 4
> +        clock-names:
> +          items:
> +            - const: mclk
> +            - const: lpr

[Severity: Medium]
The commit message states that only the RX macro uses the lpr clock instead
of the macro clock. Does the TX macro hardware also use the lpr clock, or
is this substitution an accidental copy from the RX macro block?

Since the TX driver still explicitly attempts to retrieve the macro clock via
devm_clk_get_optional, will this schema definition cause a regression by
contradicting driver expectations?

> +            - const: dcodec
> +            - const: fsgen
> +

[ ... ]

> diff --git a/Documentation/devicetree/bindings/sound/qcom,lpass-va-macro.yaml b/Documentation/devicetree/bindings/sound/qcom,lpass-va-macro.yaml
> --- a/Documentation/devicetree/bindings/sound/qcom,lpass-va-macro.yaml
> +++ b/Documentation/devicetree/bindings/sound/qcom,lpass-va-macro.yaml
[ ... ]
> @@ -185,6 +186,24 @@ allOf:
>              - const: mclk
>              - const: npl
>  
> +  - if:
> +      properties:
> +        compatible:
> +          contains:
> +            enum:
> +              - qcom,hawi-lpass-va-macro
> +    then:
> +      properties:
> +        clocks:
> +          minItems: 4
> +          maxItems: 4
> +        clock-names:
> +          items:
> +            - const: mclk
> +            - const: lpr

[Severity: Medium]
Similarly, does the VA macro actually replace the macro clock with the lpr
clock here? The commit message implies this substitution is unique to the RX
macro, while the VA macro only adds the heartbeatpulse clock.

> +            - const: dcodec
> +            - const: heartbeatpulse
> +

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260922-add_shikra_support-v1-0-6c6cb6444caf@oss.qualcomm.com?part=6