Re: [PATCH v3 2/3] dt-bindings: hwmon: pmbus: add MPS MPQ8646 binding

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] The DT binding introduces a custom property required for correct hardware operation, but no driver matches the compatible string to parse it.
- [Low] Device tree bindings must describe the hardware, not the OS driver.
--

--- Patch [2]: [PATCH v3 2/3] dt-bindings: hwmon: pmbus: add MPS MPQ8646 binding ---
commit eedf00342aacf25d6a0fb94f444aca85e5a1697f
Author: Vincent Jardin <vjardin@free.fr>

dt-bindings: hwmon: pmbus: add MPS MPQ8646 binding

This commit adds a device-tree binding for the Monolithic Power Systems
MPQ8646 step-down converter as a PMBus device.

> diff --git a/Documentation/devicetree/bindings/hwmon/pmbus/mps,mpq8646.yaml b/Documentation/devicetree/bindings/hwmon/pmbus/mps,mpq8646.yaml
> new file mode 100644
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/hwmon/pmbus/mps,mpq8646.yaml
> @@ -0,0 +1,50 @@
[ ... ]
> +description: |
> +  PMBus driver for the MPS MPQ8646 step-down converter.

[Severity: Low]
This isn't a bug, but the description describes an OS driver rather than
the hardware. Device tree bindings are meant to be OS-agnostic. Should
this be reworded to describe the hardware component itself, such as
MPS MPQ8646 PMBus step-down converter?

> +
> +properties:
> +  compatible:
> +    const: mps,mpq8646

[Severity: Medium]
Since this compatible string uses a const value without a fallback to an
existing driver like mps,mpq8785, is there a missing patch in this
series to add the corresponding driver match table entry?

> +
> +  reg:
> +    maxItems: 1
> +
> +  mps,vout-fb-divider-ratio-permille:
> +    description:
> +      The feedback resistor divider ratio, expressed in permille
> +      (Vfb / Vout * 1000). This value is written to the
> +      PMBUS_VOUT_SCALE_LOOP register and is required for correct output
> +      voltage presentation.

[Severity: Medium]
Since this custom property is required for correct hardware operation,
does the kernel currently have a driver capable of parsing it?

Because there is no driver support for the mps,mpq8646 compatible string
in this patch series, this property will be completely ignored by the
kernel. Was the corresponding driver implementation omitted from this
series?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260723-mpq8646_v0-v3-0-1af1f3d38848@free.fr?part=2