Stoat fails to track virtual methods in multiple interitance cases involving templates about stoat HOT 2 CLOSED

Current Failure in calf seems to be in the template dealiasing phase as the system seems to think that calf_plugins::saturator_audio_module is a child of audio_module organ_metadata

from stoat.

ok, so with that fixed, the problem that I'm observing which is why the process calls are missed is that process_slice calls process via vtable entry 68. This entry is a pure virtual on most things, which is not a problem, but when the class hierarchy is applied they can bind to a different vtable entry, from which there does not exist a default thunk function. Thus stoat doesn't figure out that they're the same. I guess adding in the extra thunk makes sense, but figuring out how to do that is quite hard.

from stoat.