Re: [PATCH v3 1/3] arm64: dts: qcom: eliza: Describe the ADSP and USB related nodes

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Low] Redundant 'interrupt-parent' property in the 'aoss_qmp' node.
- [Low] Newly introduced compatible strings violate existing DT binding schemas, causing dtbs_check validation failures.
--

commit 56225b8dd9e95631f7a6297b203c68d9c7cdee8d
Author: Abel Vesa <abel.vesa@oss.qualcomm.com>

arm64: dts: qcom: eliza: Describe the ADSP and USB related nodes

This commit describes the ADSP remoteproc node along with its dependencies 
including the IPCC mailbox, AOSS QMP, and SMP2P links. It also describes 
the Eliza SoC's USB 3.1 Gen 2 controller, QMP combo PHY, and SNPS eUSB2 PHY.

> diff --git a/arch/arm64/boot/dts/qcom/eliza.dtsi b/arch/arm64/boot/dts/qcom/eliza.dtsi
> --- a/arch/arm64/boot/dts/qcom/eliza.dtsi
> +++ b/arch/arm64/boot/dts/qcom/eliza.dtsi
[ ... ]
> +		usb_hsphy: phy@88e3000 {
> +			compatible = "qcom,eliza-snps-eusb2-phy",
> +				     "qcom,sm8550-snps-eusb2-phy";

Are these compatible strings fully documented in the binding schemas?

It appears qcom,eliza-snps-eusb2-phy is missing from the allowed enum list in 
Documentation/devicetree/bindings/phy/qcom,snps-eusb2-phy.yaml, which will 
cause dtbs_check to fail validation.

[ ... ]
> +		usb_dp_qmpphy: phy@88e8000 {
> +			compatible = "qcom,eliza-qmp-usb3-dp-phy",
> +				     "qcom,sm8650-qmp-usb3-dp-phy";

Does this fallback compatible chain match the current schema constraints? 

The schema Documentation/devicetree/bindings/phy/qcom,sc8280xp-qmp-usb43dp-phy.yaml 
expects qcom,sm8650-qmp-usb3-dp-phy to be used alone without a fallback, which 
will also result in dtbs_check validation failures.

[ ... ]
> +		aoss_qmp: power-management@c300000 {
> +			compatible = "qcom,eliza-aoss-qmp", "qcom,aoss-qmp";

Similar to the PHY nodes, is qcom,eliza-aoss-qmp documented in the binding 
schema? It seems to be missing from the enum list in 
Documentation/devicetree/bindings/soc/qcom/qcom,aoss-qmp.yaml.

> +			reg = <0x0 0x0c300000 0x0 0x400>;
> +
> +			interrupt-parent = <&ipcc>;
> +			interrupts-extended = <&ipcc IPCC_CLIENT_AOP IPCC_MPROC_SIGNAL_GLINK_QMP
> +						     IRQ_TYPE_EDGE_RISING>;

Is the interrupt-parent property needed here? 

Since the node uses interrupts-extended which explicitly includes the interrupt 
controller phandle for each specifier, and there are no child nodes to inherit 
the interrupt parent, this property seems to be redundant.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260512-eliza-adsp-usb-v3-0-6420282841c2@oss.qualcomm.com?part=1