Re: [PATCH v4] dt-bindings: iio: proximity: move LIDAR-Lite out of trivial-devices

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] Making `vdd-supply` a strictly required property in the new schema breaks Device Tree validation for existing valid device trees.
- [Medium] The binding schema omits the previously allowed `interrupts` property and fails to document the mode-control capabilities, contradicting the commit message and breaking compatibility.
--

--- Patch [1]: [PATCH v4] dt-bindings: iio: proximity: move LIDAR-Lite out of trivial-devices ---
commit 19408504da4434cc55f5f0ccb25c557b08078221
Author: Rodrigo Gobbi <rodrigo.gobbi.7@gmail.com>

dt-bindings: iio: proximity: move LIDAR-Lite out of trivial-devices

The PulsedLight LIDAR-Lite v2 is moved out of trivial-devices into a
dedicated binding file. Support for LIDAR-Lite v3 is also documented as
a fallback compatible.

Link: https://lore.kernel.org/all/c39a2980-f5e5-44aa-9fd3-20e0658f62dc@gmail.com/

> diff --git a/Documentation/devicetree/bindings/iio/proximity/pulsedlight,lidar-lite-v2.yaml b/Documentation/devicetree/bindings/iio/proximity/pulsedlight,lidar-lite-v2.yaml
> new file mode 100644
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/iio/proximity/pulsedlight,lidar-lite-v2.yaml

[ ... ]

> +  powerdown-gpios:
> +    description: GPIO that can be driven low to shut off power to the device.
> +    maxItems: 1
> +
> +  vdd-supply: true
> +
> +required:
> +  - compatible
> +  - reg
> +  - vdd-supply

[Severity: Medium]
Since the device was previously defined in trivial-devices.yaml, which
explicitly disallowed vdd-supply, no existing valid device trees could have
included this property. Does making vdd-supply strictly required here
introduce a regression by breaking backward compatibility during
dt_binding_check for legacy device trees?

> +
> +additionalProperties: false

[Severity: Medium]
The commit message mentions a mode-control pin used for PWM distance output,
but there are no properties defined in the schema to describe it.

Additionally, trivial-devices.yaml previously permitted the interrupts
property, but it is omitted from this new schema. Because
additionalProperties is false, wouldn't legacy device trees specifying an
interrupt fail validation?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260714215433.41259-1-rodrigo.gobbi.7@gmail.com?part=1