On 17/10/2022 06:37, Vinod Koul wrote: > On 15-10-22, 10:04, Krzysztof Kozlowski wrote: >> Several devices like SM6350, SM8150 and SC7280 are actually compatible, >> so use one compatible fallback for all of them. >> >> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx> >> --- >> Documentation/devicetree/bindings/dma/qcom,gpi.yaml | 10 ++++++---- >> 1 file changed, 6 insertions(+), 4 deletions(-) >> >> diff --git a/Documentation/devicetree/bindings/dma/qcom,gpi.yaml b/Documentation/devicetree/bindings/dma/qcom,gpi.yaml >> index 750b40c32213..0c2894498845 100644 >> --- a/Documentation/devicetree/bindings/dma/qcom,gpi.yaml >> +++ b/Documentation/devicetree/bindings/dma/qcom,gpi.yaml >> @@ -20,12 +20,14 @@ properties: >> compatible: >> oneOf: >> - enum: >> - - qcom,sc7280-gpi-dma >> - qcom,sdm845-gpi-dma >> - qcom,sm6350-gpi-dma >> - - qcom,sm8350-gpi-dma >> - - qcom,sm8450-gpi-dma >> - >> + - items: >> + - enum: >> + - qcom,sc7280-gpi-dma >> + - qcom,sm8350-gpi-dma >> + - qcom,sm8450-gpi-dma >> + - const: qcom,sm6350-gpi-dma > > I think it makes sense but can we document this in binding as well that > why people should use these two compatibles. I am fine with this being > comments here.. It is kind of implied (and maybe obvious) from the bindings - a list of two items, one enum and one fallback compatible. We usually do not document such patterns in the bindings with comments for that reason. If you insist, I can add it. Best regards, Krzysztof
