MDEV-6139: UPDATE w/ join against MRG_MyISAM table with read-only sub-table failsUPDATE w/ join against MRG_MyISAM table with read-only sub-table fails

The problem was that on opening all tables TL_WRITE, than local tables which is not updated set to TL_READ, but underlying tables of MyISAMmrg left untouched.

Prartition engine has not this problem.

All cases where lock_type assigned is not changed because call of virtual function is not cheap.

This is not yet a fix. This is change to print additional information at the point
when this assertion is going to happen. Print as much information about the pages
and index to find out why next page is not a compact format.

the part which was defined by MariaDB version and thus caused a wrong
check result

Added a 5.5-specific test which involves manual modification of system
tables. The problem itself was fixed in MDEV-6068

The test didn't clean debug_sync

Synchronization with master was missing

Fixed crashing bug for union queries where there was no real tables.


mysql-test/r/group_by.result:
  Added test case
mysql-test/t/group_by.test:
  Added test case
sql/db.opt:
  Removed genrated file
sql/item.cc:
  Handled case when table_list->pos_in_tables is not set. Can only happens when there is no real tables in query

MDEV-6134 SUBSTRING_INDEX returns wrong result for 8bit character sets when delimiter is not found

Recording the forgotten test results.

Both bugs are caused by the same problem: the function optimize_cond() should
update the value of *cond_equal rather than the value of join->cond_equal,
because it is called not only for the WHERE condition, but for the HAVING
condition as well.

   Some lines of code in file_logger.c were lost while moving to the
   general MariaDB tree. Adding them.

    The fill_schema_table() function used to call get_table_share() for a table name in WHERE
    then clear the error list. That way plugins receive the superfluous error notification if it
    happens in it. Also the problem was that error handler didn't prevent the suppressed
    error message from logging anyway as the logging happens in THD::raise_condition
    before the handler call.
    Trigger_error_handler is remade into Warnings_only_error_handler, so it stores the error
    message in all cases in the thd->stmt_da.
    Then later the stored error is raised.

     If a prepared statement calls an stored procedure,
     the thd->server_status out of the SP goes up
     to the PS and then to the client. So that the
     client gets the SERVER_STATUS_CURSOR_EXISTS status
     if the SP uses a cursor. Which makes the embedded
     server fail.
     Fixed by saving/restoring the upper-level server_status
     in sp_head::execute().

     Thread can be disconnected internally for example after COMMIT statements.
     So we should check this for the statement execution.

    mysqltest in the 'embedded-server' mode runs queries in a separate thread,
    but it didn't do so for the prepared statements - they were run in the main thread.
    That leads to inconsistencies.
    When a test sets SESSION 'dbug' variable like
            SET SESSION debug_dbug="+d,warn_during_ha_commit_trans";
    it is run as a plain query in that separate thread, so the main thread remains unaffected.
    After that the prepared statement run in the main thread doesn't produce expected 'dbug' errors,
    so the test fails.
    To fix that I made prepared statement to be run in that special thread along with the plain queries.
    That makes the environment consistent.

  As Davi added code like
          sav_protocol= thd->protocol
          thd->protocol= &thd->protocol_binary
          ...
          thd->protocol= sav_protocol
  the fucntions like emb_store_querycache_result() cannot determine
  the used protocol testing thd->protocol == &thd->protocol_binary.
  Fixed by additional check thd->command == COM_STMT_EXECUTE.

        The function Protocol::net_store_data(a, b, CHARSET_A, CHARSET_B) should
        be adapted to be working in the embedded server as it's done
        with the Protocol::net_store_data(a, b).
        That new function renamed as net_store_data_cs, so we can make it
        virtual.

Units of subqueroes from excluded expressions should be excluded from select_lex/select_unit tree.

(and don't append --user=root for --help now, when mysqld has a fix for that)

Add test case.

Revert the old patch revid:monty@askmonty.org-20100325133339-7mkel6valai0b4lb

This patch caused the InnoDB part of the transaction to not be marked
read-write in some cases, which messes up XA commit (and likely other
stuff as well).