- bugfix: reset slave should not reset Gtid_state
 - bugfix: always register binlog handler when acquiring ownership of
   GTIDs, so that ownership can be released
 - bugfix: write empty GTID to binary log when releasing ownership of
   GTIDs, if GTID is not already in binary log
 - bugfix: don't acquire ownership for empty statements
 - feature: added @@global.gtid_owned and @@session.gtid_owned
 - cleanup: don't pass gtid_state, global_sid_lock, or global_sid_map as arguments
   in functions in Group_cache, and zgroup_execution.cc
 - cleanup: pass Gtid instead of (rpl_sidno, rpl_gno) to some
   functions
 - cleanup: moved Group_cache::update_gtid_state to Gtid_state::update
 - extra assertions
 - clean up some comments
 - disable use of gtid_next_list until we need it
 - clean up some tests