patch: regression: C++ overload

[Jimmy Guo]

>There weren't any regressions caused by the problem, you just get 2
>less unexpected passes, so i didn't notice any problems.

I should have caught that when I merged HP WDB's source up to 3/21's
yesterday.

Actually the testing caught this when I ran the tests at our HP aCC
pass, for which we are used to PASSs, but got XFAs for these test points
with odd number of arguments (and then appeared on and off at repeated
print commands, which points to some sort of UMR problem).

- Jimmy