Re: [PATCH 02/10] dt-bindings: interrupt-controller: apple,aic2: Add apple,t8132 compatible
Yureka Lilian <[email protected]>
| Newsgroups | org.kernel.vger.linux-watchdog,dev.linux.lists.sashiko-reviews,org.kernel.vger.linux-devicetree |
|---|---|
| Message-ID | <[email protected]> |
On 7/5/26 15:24, [email protected] wrote: > Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: > - [Low] The newly added compatible string 'apple,t8132-aic3' (and the pre-existing 'apple,t6030-aic3') are missing from the `if:properties:compatible:contains:enum:` conditional block that enforces `#interrupt-cells` constraints. > -- > > commit 4be0e67c14a594eec020638f7ba5290c4bf05496 > Author: Yureka Lilian <[email protected]> > > dt-bindings: interrupt-controller: apple,aic2: Add apple,t8132 compatible > > The Apple t8132 (M4) SoC uses an AIC3 as interrupt controller, same as > the M3 predecessor. > >> diff --git a/Documentation/devicetree/bindings/interrupt-controller/apple,aic2.yaml b/Documentation/devicetree/bindings/interrupt-controller/apple,aic2.yaml >> index d0d9a90e96e7b..7171c8c09feaf 100644 >> --- a/Documentation/devicetree/bindings/interrupt-controller/apple,aic2.yaml >> +++ b/Documentation/devicetree/bindings/interrupt-controller/apple,aic2.yaml >> @@ -44,6 +44,7 @@ properties: >> - items: >> - enum: >> - apple,t6030-aic3 >> + - apple,t8132-aic3 > [Severity: Low] > Does this new compatible string also need to be added to the allOf conditional > block further down in the schema? No, this is correct, as t8132 is merely another alias always to be used in combination with the existing apple,t8122-aic3 base compatible which is already in the list. > > Looking at the conditional block that enforces the constraints later in this > file: > > allOf: > - $ref: /schemas/interrupt-controller.yaml# > - if: > properties: > compatible: > contains: > enum: > - apple,t8112-aic > - apple,t8122-aic3 > then: > > It appears "apple,t8132-aic3" is omitted from this enum list. > > If a device tree node is written with compatible = "apple,t8132-aic3" and omits > the fallback string, the top-level schema will flag the missing fallback, but > this conditional block will silently fall through to the else branch and > incorrectly permit #interrupt-cells: 4 instead of enforcing the strict > #interrupt-cells: 3 constraint. > > This is a pre-existing issue, but it looks like the "apple,t6030-aic3" string > is also missing from this conditional block. > >> - const: apple,t8122-aic3 >> - const: apple,t8122-aic3 >>