Hi I'm wondering how to handle the case where the outputs from a device that generates interrupts are not all wired to an interrupt controller, and also where different systems wire up different interrupts. E.g. a datasheet says device generates 3 interrupts A, B and C, but one system using the device doesn't connect up B and another doesn't connect up C. >From what I can see, device-tree bindings for a device will tend to list rigid length and order for the contents of 'interrupt' and 'interrupt- names' properties, but what happens if one of them isn't wired up in hardware? Is there such a thing as a null interrupt specifier to put in the list? One solution, for devices without bindings yet is to specify interrupts as optional but require interrupt-names for those that are present so OS drivers can match the correct interrupt functions to the interrupt specifiers. E.g for my two example systems interrupts = <10>, <11>; interrupt-names = "A", "C">; interrupts = <10>, <11>; interrupt-names = "A", "B"; Is such an approach acceptable? I've searched quite a bit existing practice in this area but have come up blank. Thanks for any help. -- Tixy -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
