> Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider: > - [High] The `reg` property check in the conditional `if` block uses an invalid JSON schema sequence that will always evaluate to false, causing valid device trees to fail validation. > - [High] The `ethernet-port@[0-5]` nodes model a `net_device` but strictly forbid standard network configuration properties like `mac-address` and `phy-handle`. > - [Low] Missing `#address-cells` and `#size-cells` in the `required` list for conditional child nodes. > -- > > commit 270a641148bef0e32520efd84b62bf6b97aa5a63 > Author: Lorenzo Bianconi > > dt-bindings: net: airoha: Add EN7581 ethernet-ports properties > > This commit introduces the ethernet-port property for EN7581 and AN7583 > SoCs to model multiple net_devices connected via an external arbiter to > GDM3 or GDM4 ports. > > > diff --git a/Documentation/devicetree/bindings/net/airoha,en7581-eth.yaml b/Documentation/devicetree/bindings/net/airoha,en7581-eth.yaml > > --- a/Documentation/devicetree/bindings/net/airoha,en7581-eth.yaml > > +++ b/Documentation/devicetree/bindings/net/airoha,en7581-eth.yaml > [ ... ] > > @@ -130,6 +130,38 @@ patternProperties: > > maximum: 4 > > description: GMAC port identifier > > > > + '#address-cells': > > + const: 1 > > + '#size-cells': > > + const: 0 > > + > > + allOf: > > + - if: > > + properties: > > + reg: > > + items: > > + - enum: > > + - 3 > > + - 4 > > Will this schema condition evaluate as expected? > > In dt-schema, reg properties are parsed into a matrix format, which is an > array of arrays of integers like [[3]]. This schema checks if the first > element of the outer array matches the scalar values 3 or 4, but since > the first element is an inner array, it won't match a scalar. > > Could this cause the if block to always evaluate to false, causing valid > ethernet-port nodes to be incorrectly rejected because the parent node > sets unevaluatedProperties: false? ack, I will fix it in v6. > > > + then: > > Since #address-cells and #size-cells are needed for the ethernet@[3-4] > variants to act as buses for the new ethernet-port children, should they > be included in a required list within this conditional block? > > Without them, an incomplete device tree could omit these properties and > silently pass schema validation. ack, I will fix it in v6. > > > + patternProperties: > > + "^ethernet-port@[0-5]$": > > + type: object > > + unevaluatedProperties: false > > Does setting unevaluatedProperties: false here without referencing > ethernet-controller.yaml# forbid standard network properties? > > Since this models a net_device, driver code often expects to parse standard > properties like mac-address or phy-handle. Device tree authors might not > be able to configure them without triggering schema validation errors. ack, I will fix it in v6. Regards, Lorenzo > > > + description: External ethernet port ID available on the GDM port > > + > > + properties: > > + compatible: > > + const: airoha,eth-port > > + > > + reg: > > + maxItems: 1 > > + description: External ethernet port identifier > > + > > + required: > > + - compatible > > + - reg > > -- > Sashiko AI review · https://sashiko.dev/#/patchset/20260509-airoha-eth-multi-serdes-v5-0-805e38edc2aa@kernel.org?part=1