Re: [PATCH v2 2/2] dt-bindings: usb: richtek,rt1711h: add missing 'port' property

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The commit modifies the device tree schema to legitimize a non-standard and semantically incorrect devicetree layout present in `hi3660-hikey960.dts`, instead of fixing the buggy devicetree file.
--

commit eb001738f71c4e41754a143f37ed24b7c078d74d
Author: Akash Sukhavasi <akash.sukhavasi@gmail.com>

dt-bindings: usb: richtek,rt1711h: add missing 'port' property

This commit adds a top-level port property to the richtek,rt1711h binding
to resolve a dtbs_check warning for hi3660-hikey960.dts. It also updates
the example to show the role-switch endpoint connection.

> Add the missing property using /schemas/graph.yaml#/properties/port,
> following the pattern used by realtek,rts5411.yaml and other USB
> controller bindings.

Is this comparison to realtek,rts5411.yaml accurate? The RTS5411 is a USB
hub that physically routes data, while the RT1711H is a Type-C Port
Controller (TCPC).

> diff --git a/Documentation/devicetree/bindings/usb/richtek,rt1711h.yaml b/Documentation/devicetree/bindings/usb/richtek,rt1711h.yaml
> index 7ded363845184..05ffc33a24fca 100644
> --- a/Documentation/devicetree/bindings/usb/richtek,rt1711h.yaml
> +++ b/Documentation/devicetree/bindings/usb/richtek,rt1711h.yaml
> @@ -52,6 +52,13 @@ properties:
>      description:
>        Properties for usb c connector.
>  
> +  port:
> +    $ref: /schemas/graph.yaml#/properties/port
> +    description:
> +      OF graph port describing the connection to the USB role switch
> +      consumer (e.g., a dwc3 controller with usb-role-switch), used to
> +      convey type-C data-role changes signaled by the TCPC.

Since the RT1711H is a TCPC, should this role switch endpoint logically belong
to the physical port inside the connector child node instead?

According to the standard bindings in connector/usb-connector.yaml, the
connection to the USB role switch is typically represented inside the
usb-c-connector child node.

The Linux TCPM core also appears to expect the standard layout, as tcpci.c
initializes tcpc->fwnode to the connector child node, and tcpm.c attempts
fwnode_usb_role_switch_get(tcpc->fwnode) first.

Would it be more appropriate to update hi3660-hikey960.dts to move the port
into the connector subnode rather than changing the schema to allow it at
the top level?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260518162014.18251-1-akash.sukhavasi@gmail.com?part=2