MDEV-324: SHOW EXPLAIN: Plan produced by SHOW EXPLAIN for a query with TEMPTABLE view loses 'DERIVED' line
- Make SHOW EXPLAIN code take into account that st_select_lex object without joins can be
  a full-featured SELECTs which were already executed and cleaned up.

   INDEX and SHOW EXPLAIN in parallel
- Rework locking code to use the LOCK_thd_data mutex for all synchronization. This also
  fixed MDEV-301.

- Make SHOW EXPLAIN ignore range accesses when printing "Range checked for each record" plans.

MDEV-305: SHOW EXPLAIN: ref returned by SHOW EXPLAIN is different from the normal EXPLAIN ('const' vs empty string)
- The problem was that create_ref_for_key() would act differently, depending on 
  whether we're running EXPLAIN or the actual query.
- As the first step, fixed the EXPLAIN printout not to depend on actions in create_ref_for_key().

MDEV-275: SHOW EXPLAIN: server crashes in JOIN::print_explain with IN subquery and aggregate function
- Don't try to produce plans after JOIN::cleanup() has been called, because:
   = JOIN::cleanup leaves data structures in partially-cleaned state
   = Walking them is hazardous (see this bug), and has funny effects
     (See previous commits, "Using join cache" may or may not be shown)
   = Changing data structures to be persisted may cause unwanted side effects
- The consequence is that SHOW EXPLAIN will show "Query plan already deleted" when e.g. 
  reading data after filesort.

- It turns out, there is a case where the join is degenerate, but 
  join->table_count!= && join->tables_list!=NULL. Need to also check 
  if join->zero_result_cause!=NULL, too.

- There is a slight problem: The code sets
    zero_result_cause= "no matching row in const table"
  when NOT running EXPLAIN. The result is that SHOW EXPLAIN will show this line while 
  regular EXPLAIN will not.

 select tables optimized away
- Take into account, that for some degenerate joins instead of "join->table_count=0"
  the code sets "join->tables_list=0".

- Make SHOW EXPLAIN code handle degenerate subquery cases (No tables, impossible where, etc)

- Fix the bug: SHOW EXPLAIN may hit a case where a join is partially 
  optimized.
- Change JOIN::optimized to use enum instead of numeric constants

- Make all functions that produce parts of EXPLAIN output take 
  explain_flags as parameter, instead of looking into thd->lex->describe

- Support SHOW EXPLAIN for selects that have "Using temporary; Using filesort".

Make SHOW EXPLAIN work for queries that do "Using temporary" and/or "Using filesort"
- Patch#1: Don't lose "Using temporary/filesort" in the SHOW EXPLAIN output.

The code to re-enable checkpointing after UNLOCK TABLES was lost in the 5.5
merge, so re-add it back in.

5.5 version. for cmake, not autotools.

mtr should do it as well, to avoid differences in test output. 

This fixes sys_vars.secure_file_priv on FreeBSD9.0 buildbot.

is gone.

The problem was increment of aborted_threads variable due to thd->killed which was set when threadpool connection was terminated .  The fix is not to set thd->killed anymore, there is no real reason for doing it..

Added a test that checks that status variable aborted_clients does not grow for ordinary disconnects, and that successful KILL increments this variable.

MDEV-214 lp:967242 Wrong result with JOIN, AND in ON condition, multi-part key, GROUP BY, subquery and OR in WHERE

The problem was in the code (update_const_equal_items()) which marked
index parts constant independently of the place where the equality was used.
In the test suite it marked t2_1.c part constant despite the fact that
it connected by OR with other expression.

Solution is to mark constant only top equalities connected with AND.

Fix is to set  maybe_null  flag for Item_func_last_day.

unnecessary date->string->date conversion