On Fri, 27 Nov 2015, Vladis Dronov wrote: > Hello, > > > > Hmm, I see quite a few drivers assuming that endpoint 0 will be present. > > > I wonder if that should not be solved at USB level. > > > > Every USB device always has endpoint 0. If one didn't, the kernel > > wouldn't be able to initialize and enumerate it. > > Yes for the normal USB device. This case is a weird USB device (with no > endpoints) specially designed to be weird. My point here is that even a > weird device probing should not crash a kernel by a NULL-deref. True. However, a weird device that didn't have any endpoint 0 would not crash the kernel, because the kernel wouldn't be able to communicate with it at all. > > > Alan, does it make sense to have drivers probe interface if it does not > > > have any endpoints? > > > > Yes; in theory an interface can do everything it needs using only > > endpoint 0. Driver shouldn't assume anything about the endpoints in > > the interfaces they problem. > > The current kernel code in drivers/usb/core/config.c accepts an interface > with no endpoints in one of its configurations, and I could not find a > direct ban for that in USB standard. So, I currently believe, it is a driver > job to check if the endpoint 0 exist, otherwise we must change the kernel > USB detection code. Drivers do not have to check whether endpoint 0 exists; it always does. But they do have to check any other endpoint they use. > btw, indeed, this is not the only driver with this problem, I've met 2 more. In fact, there's no way to check whether endpoint 0 exists. Where would you look to find out? There isn't any endpoint descriptor for it. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
