create a lock so tests don't have to worry about being async: false

[techgaun]

We need to have a process that controls mocking via lock. On every mock, the process will request for lock from this lock handling process. Once the test execution completes, it will release the lock & the next test can do the mocking. Would have to see how feasible it would be and if anything would actually block certain tests to the point it times out?

I am not sure if this will always work for consumers though. For example, anything not mocking will try to access the original module?