A pre-requisite patch for the fix for Bug#52044.

This patch also fixes Bug#55452 "SET PASSWORD is
replicated twice in RBR mode".

The goal of this patch is to remove the release of 
metadata locks from close_thread_tables().
This is necessary to not mistakenly release
the locks in the course of a multi-step
operation that involves multiple close_thread_tables()
or close_tables_for_reopen().

On the same token, move statement commit outside 
close_thread_tables().

Other cleanups:
Cleanup COM_FIELD_LIST.
Don't call close_thread_tables() in COM_SHUTDOWN -- there
are no open tables there that can be closed (we leave
the locked tables mode in THD destructor, and this
close_thread_tables() won't leave it anyway).

Make open_and_lock_tables() and open_and_lock_tables_derived()
call close_thread_tables() upon failure.
Remove the calls to close_thread_tables() that are now
unnecessary.

Simplify the back off condition in Open_table_context.

Streamline metadata lock handling in LOCK TABLES 
implementation.

Add asserts to ensure correct life cycle of 
statement transaction in a session.

Remove a piece of dead code that has also become redundant
after the fix for Bug 37521.

sql/sql_priv.h

@@ -135,6 +135,16 @@ extern char err_shared_dir[];
   Type of locks to be acquired is specified directly.
 */
 #define SELECT_HIGH_PRIORITY            (1ULL << 34)     // SELECT, user
+/**
+  Is set in slave SQL thread when there was an
+  error on master, which, when is not reproducible
+  on slave (i.e. the query succeeds on slave),
+  is not terminal to the state of repliation,
+  and should be ignored. The slave SQL thread,
+  however, needs to rollback the effects of the
+  succeeded statement to keep replication consistent.
+*/
+#define OPTION_MASTER_SQL_ERROR (1ULL << 35)
 
 
 /* The rest of the file is included in the server only */