RFE: Behavior for non-matching (NO_)TESTS filter

[Paul Hargrove]

When using make dev-tests I recently made a typo in my TESTS filter, which led to no tests being compiled. Since the final output was PASSED compiling 425 tests, I was not aware at the time that I had failed to test what I had intended.

I would argue that case of TESTS having no matches is deserving of a warning.
Similarly, the case that NO_TESTS is set to a non-default value but still fails to match anything may also deserve a warning.

In either case, the lack of matches could mean a typo and that the user has NOT tested what they had actually intended.

[Dan Bonachea]

Removing version: mk-develop branch (automated comment)