a2110bd3fc
These were found by buiding the unix coverage variant on macOS (so clang compiler). Mostly, these are fixing implicit cast of float/double to mp_float_t which is one of those two and one mp_int_t to size_t fix for good measure.