(fixing cc:) 2010/7/22 Michał Nazarewicz <m.nazarewicz@xxxxxxxxxxx>: > On Wed, 21 Jul 2010 23:44:25 +0200, Adam Kropelin > <akropel1@xxxxxxxxxxxxxxxx> wrote: >> In the broader case, to some extent I wonder if the logic in >> usb_choose_configuration() should be as simple as choosing the one and >> only config regardless of any heuristics in all cases where only a >> single config is present. Is there value in failing to select any >> configuration at all? > > I had an idea once to make the code choose configuration with the smallest > bConfigurationValue. This way, composite devices could provide RNDIS > configuration as the first one with bConifgurationValue == 2 and the other > configuration, as second one, with bConfigurationValue == 1. This is what > Ethernet gadget does for instance. A reasonable idea, but not the issue I'm trying to solve. I want a device with a single configuration to get configured even if it has an RNDIS interface when the kernel is not configured with RNDIS support. The heuristics to apply when choosing from multiple configs is a separate topic with many twisty passages. > Windows would still choose the first configuration so all would be grand. Actually, Windows' enumeration of composite devices is quite broken (ok, maybe "not ideal") and it will in fact only select a configuration if a composite device has a single config. If it has multiple configs Windows punts and makes you install an INF to force it to select a config (using a registry setting of course...ugh!). I've been trying to find a way to work around both the Linux RNDIS issue and various Windows issues from the device side since I have control of the device firmware. At this point I'm pretty much convinced it's not possible to make both Linux and Windows happy at the same time. I'd love to be wrong about that, though. --Adam -- 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
