Am Montag, 30. Januar 2012, 19:26:17 schrieb Bjørn Mork: > Oliver Neukum <oliver@xxxxxxxxxx> writes: > > > Bjørn, in this case it is best to provide two helping function > > pointers when you create the "subdevice" and realize the counters > > in the network driver. > > Do you mean to let the cdc-wdm subdriver call back the main driver when > it wants to set/clear the needs_remote_wakeup flag? Yes. > Yes, I can do that but I sort of feel it messes up the interface for > little gain. We'd need to store the callback pointers and modify > wdm_open and wdm_release to use them conditionally. Or provide local > callbacks for use as a normal driver. Not pretty at all. Providing local alternatives is pretty clean. probe() can pass them to create() > I am currently leaning towards just documenting that cdc-wdm will change > the needs_remote_wakeup flag and that any driver using cdc-wdm as a > subdriver therefore will have to prevent autosuspending while being > active. This can be done without involving neither usb core nor We want devices to be able to autosuspend as often as possible. > cdc-wdm. The main driver can keep its' private version of the > needs_remote_wakeup flag and consult that and device_can_wakeup() on > autosuspend. Ugh. Not good. Regards Oliver -- 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
