Investigate using usbdk on windows

[mattico]

This was mentioned in #6 but not fixed by #21 so I wanted to open an issue to keep tracking the idea.

Using libusb on windows requires replacing a device's driver with a libusb one, e.g. using Zadig.

https://github.com/daynix/UsbDk could be used instead, which detaches the device from its drivers and Windows' PNP stack at runtime allowing the device to be controlled without needing to replace its driver.

[jonas-schievink]

I prototyped this on the context-usbdk branch a while back, and it seemed to work. I don't know if it'll still work with the WinUSB driver though.

[mattico]

I didn't know that libusb supported usbdk as a backend, I thought this was going to be a lot more work!

From my 10 minutes of research, usbdk should bypass whatever driver windows wants to use for the device but testing is required of course. Worst case we create 2 libusb contexts, set the option on one of them and enumerate both.

I should be able to work on this if I ever get probe-rs working with my micro.

[jonas-schievink]

Worst case we create 2 libusb contexts, set the option on one of them and enumerate both.

Yeah, that's a reasonable workaround I think.

[mattico]

I got this working: master...mattico:72d8ac9

Some notes to myself:

1. probe_run enumerates the device, closes it, then immediately tries to enumerate it again to open the device. UsbDk doesn't release the device fast enough (?) so the device doesn't show up in the second enumeration. A sleep_ms(1000) in new_from_selector "fixes" this.
2. When open_usb is called the second time handle.set_active_configuration doesn't actually work - active_configuration() keeps returning 0 so we get a "another application is accessing the device" error. If we ignore that error everything seems to work. We may be not de-initializing everything properly after the first open_usb, the libusb docs have many requirements, but rusb seems to be closing everything properly. Try to repro with libusb, might be a bug in that or usbdk.