[ofw] Add Assert to debug BS while deleting WDF device

[Hefty, Sean]

> It's SM 3.3.6 if I understand correctly it uses winmad

Stan, can you confirm this?

There is an issue in the code that I can see.  If during D0Entry, winmad fails to create the ControlDevice for some reason, then it will crash during D0Exit.  Additional error handling or a couple of simple checks are needed to prevent this.

However, if opensm was using winmad at the time of the crash, then the ControlDevice must be valid.  If this is the case, then tracking down the problem is going to prove more difficult, and it may not be associated with the ControlDevice at all.  Possibly stack corruption, since WdfObjectDelete is called with a stack variable..?

- Sean