When a direct call to a drwrap target is found, we now proactively add the post-call point to the to-be-instrumented table, potentially avoiding a flush if that PC were to be reached through some other method before we find it in our callee control point.
This also enables finding post-call points even if we limited dynamic return address investigation, which does add overhead. A future change will explore this further.
Issue: #4070 (closed)