Document why the binding uses oneOf when specifying just one of the
interrupt properties is supposed to be enough.

dtschema's fixups.py has special treatment of the interrupts and
interrupts-extended properties, but that appears to work at the top
level only. Elsewhere, an explicit oneOf is required.

Signed-off-by: André Draszik <andre.draszik@linaro.org>
---
 Documentation/devicetree/bindings/mfd/samsung,s2mps11.yaml | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/Documentation/devicetree/bindings/mfd/samsung,s2mps11.yaml b/Documentation/devicetree/bindings/mfd/samsung,s2mps11.yaml
index d6b9e29147965b6d8eef786b0fb5b5f198ab69ab..31d544a9c05cad878d10a0ae9b99631f08eb04a8 100644
--- a/Documentation/devicetree/bindings/mfd/samsung,s2mps11.yaml
+++ b/Documentation/devicetree/bindings/mfd/samsung,s2mps11.yaml
@@ -81,6 +81,9 @@ allOf:
         samsung,s2mps11-acokb-ground: false
         samsung,s2mps11-wrstbi-ground: false
 
+      # oneOf is required, because dtschema's fixups.py doesn't handle this
+      # nesting here. Its special treatment to allow either interrupt property
+      # when only one is specified in the binding works at the top level only.
       oneOf:
         - required: [interrupts]
         - required: [interrupts-extended]

---
base-commit: 1b152eeca84a02bdb648f16b82ef3394007a9dcf
change-id: 20250627-s2mpg10-binding-comment-9b632a246529

Best regards,
-- 
André Draszik <andre.draszik@linaro.org>

On Fri, 27 Jun 2025 10:15:25 +0100, André Draszik wrote:
> Document why the binding uses oneOf when specifying just one of the
> interrupt properties is supposed to be enough.
> 
> dtschema's fixups.py has special treatment of the interrupts and
> interrupts-extended properties, but that appears to work at the top
> level only. Elsewhere, an explicit oneOf is required.
> 
> [...]

Applied, thanks!

[1/1] dt-bindings: mfd: samsung,s2mps11: add comment about interrupts properties
      commit: 37f716488780c4ce9e30a17f3c4ff9baee27c96d

--
Lee Jones [李琼斯]

On 27/06/2025 11:15, André Draszik wrote:
> Document why the binding uses oneOf when specifying just one of the
> interrupt properties is supposed to be enough.
> 
> dtschema's fixups.py has special treatment of the interrupts and
> interrupts-extended properties, but that appears to work at the top
> level only. Elsewhere, an explicit oneOf is required.
> 
> Signed-off-by: André Draszik <andre.draszik@linaro.org>
> ---

Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>

Best regards,
Krzysztof