The iterators in table_data_locks::m_iterator[] where not being reset during ::reset_position(),
which caused them to report no more rows when reused subsequently.
Same problem was in table_data_lock_waits::reset_position().
Additionally, PFS_data_lock_wait_container::clear() didn't reset m_logical_row_index.
This patch introduces MTRs which expose these problems, fixes them,
and also changes some of integer types to size_t, to make number of warnings smaller.

Reviewed-by: Marc Alff <marc.alff@oracle.com>