[PATCH v2 4/8] dt-bindings: media: Add CAMSS device for Maili

[sashiko-bot at kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] dt-bindings: media: missing port constraints for maili-camss variant
- [Medium] dt-bindings: media: invalid superset fallback compatible for hardware subset
--

commit 5533f51204e6b2bb4713bbf162c5fc5e10ea6c2b
Author: Nihal Kumar Gupta <nihal.gupta at oss.qualcomm.com>

dt-bindings: media: Add CAMSS device for Maili

This commit updates the CAMSS device tree bindings to support the Maili
hardware variant. It adds qcom,maili-camss as a compatible string, falling
back to the existing qcom,hawi-camss compatible.

> diff --git a/Documentation/devicetree/bindings/media/qcom,hawi-camss.yaml b/Documentation/devicetree/bindings/media/qcom,hawi-camss.yaml
> index 5bf0dca3d2462..d1d53d6f024cb 100644
> --- a/Documentation/devicetree/bindings/media/qcom,hawi-camss.yaml
> +++ b/Documentation/devicetree/bindings/media/qcom,hawi-camss.yaml
[ ... ]
>  properties:
>    compatible:
> -    const: qcom,hawi-camss
> +    oneOf:
> +      - const: qcom,hawi-camss
> +      - items:
> +          - const: qcom,maili-camss
> +          - const: qcom,hawi-camss

[Severity: Medium]
Since Maili has 5 CSIPHYs while Hawi has 6, Maili is a hardware subset of Hawi.

Is it safe to claim compatibility with the Hawi superset?

If an older kernel that implements support for qcom,hawi-camss boots on
Maili hardware and binds using this fallback compatible, it might attempt
to access registers, clocks, or state machines for the 6th CSIPHY which
does not physically exist on the Maili hardware.

Could this fallback cause boot failures or undefined behavior?

[Severity: Medium]
Will this schema silently pass invalid device trees for Maili that define a
non-existent 6th CSIPHY port (port at 5)?

The base schema in Documentation/devicetree/bindings/media/qcom,hawi-camss.yaml
unconditionally defines:

    patternProperties:
      "^port@[0-5]$":

which allows up to 6 ports globally.

If a device tree is created for Maili using qcom,maili-camss and incorrectly
includes a port at 5 node, it will pass validation.

Since Maili has only 5 CSIPHY instances, should an allOf conditional block be
added to restrict the ports property (for example, setting port at 5: false or
lowering maxItems) when the compatible string contains qcom,maili-camss?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261013-hawi-maili-camss-v2-0-fc93222da89c@oss.qualcomm.com?part=4