On Thu, 9 Oct 2014, Robert Baldyga wrote: > In fact this patch solves in some way the problem 'who should handle > requests addressed to device?'. In theory, such requests should be handled by the top-level gadget driver (i.e., composite.c or equivalent). By definition, function drivers handle requests that affect only one function, not the entire device. In practice, there may be some requests that the top-level gadget driver doesn't know how to handle. What to do with them is unclear. If you know of cases where these requests are sent by a buggy host and they really should be adressed to a particular interface or endpoint, then you need to figure out which interface or endpoint and send the request on to the corresponding function driver. If you can't do that either, simply fail the request. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
