that is, don't create a "temporary" IO_CACHE with file==-1
and manually swap it later to the valid file descriptor.

remove some 14-year old code that added support for
LOAD DATA replication to IO_CACHE:
* three callbacks, of which only two were actually used and that
  were only needed for LOAD DATA replication but were
  tested in every IO_CACHE instance
* an additional opaque void * argument in IO_CACHE, also only
  used for LOAD DATA replication, but present everywhere
* the code to close IO_CACHE prematurely in LOAD DATA to have
  these callbacks called in the correct order and a long
  comment explaining what will happen if IO_CACHE is not
  closed prematurely
* a variable to track whether IO_CACHE was closed prematurely
  (to avoid double-closing it)

perfschema after 5.6 merge depends on sql_yacc.h

To make st_select_lex::master_unit() inlinable:
- moved it's definition to sql_lex.h
- removed base class virtual master_unit() declaration since this method is
  specific to st_select_lex

Overhead change:
st_select_lex::master_unit()      0.17% -> out of radar
execute_sqlcom_select()           0.13% -> 0.12%
JOIN::save_explain_data_intern()  0.27% -> 0.23%
JOIN::optimize_inner()            0.76% -> 0.72%
JOIN::exec_inner()                0.30% -> 0.24%
JOIN::prepare()                   0.30% -> 0.29%
JOIN::optimize()                  0.05% -> 0.05%

Moved THD::set_command() definition to sql_class.h so that it can be inlined.

Overhead change:
THD::set_command        0.05% -> out of radar

Moved THD::set_query() definition to sql_class.h so that it can be inlined.

Overhead change:
THD::set_query       0.07% -> out of radar

MDEV-8257 Erroneous "Impossible where" when mixing decimal comparison and LIKE

ulong is not portable - different width on different architectures

version_ssl_library should show the version
of openssl library that is being used, not the one
that server was compiled with.

don't run this test on 1.0.1c

Use quote_for_equal to correctly escape characters.

mysqldump now attempts to make use of the INFORMATION_SCHEMA tables.
If the table name is not found with a case sensitive search, it
fallbacks to a case insensitive search.

MDEV-8241: Debug build on Windows is broken: error LNK2019: unresolved external symbol pthread_detach referenced in function ma_checkpoint_init

pthread_detach() replaced with pthread_detach_this_thread()

pthread_detach_this_thread() definition fixed

(yes, we no longer add all variables to sys_vars suite!)

* mysqld_safe: Since wsrep_on variable is mandatory in 10.1, skip wsrep
position recovery if its OFF.
* mysqld: Remove "-wsrep" from server version
* mysqld: Remove wsrep patch version from @@version_comment
* mysqld: Introduce @@wsrep_patch_version

Do not reset timer inside stored procedures and functions.

that was apparently lost in 20c23048:

  commit 20c23048
  Author: Jan Lindström <jan.lindstrom@mariadb.com>
  Date:   Sun May 17 14:14:16 2015 +0300

      MDEV-8164: Server crashes in pfs_mutex_enter_func after fil_crypt_is_closing

This also reverts 8635c4b4:

  commit 8635c4b4
  Author: Jan Lindström <jan.lindstrom@mariadb.com>
  Date:   Thu May 21 11:02:03 2015 +0300

      Fix test failure.

as it's not used by the other Item_bool_func2 descendands.

Forgot that echo will also expand variables and path might be different.

Step #8: Adding get_mm_tree() in Item_func, Item_func_between,
Item_func_in, Item_equal. This removes one virtual call item->type()
in queries like:
  SELECT * FROM t1 WHERE c BETWEEN const1 AND const2;
  SELECT * FROM t1 WHERE c>const;
  SELECT * FROM t1 WHERE c IN (const_list);

Analysis: Problem was that encryption was skipped.

Fixed by making sure that tables with ENCRYPTED=YES are encrypted.

Analysis: Problem was that used compression method needs to be stored
to the page.

Fixed by storing compression method after key_version to the page.

Step #7 (mostly preparatory for the next step #8):
Splitting the function get_mm_parts() into a virtual method in Item.
This changes a virtual call for item->type() into a virtual call for item->get_mm_tree(),
but also *removes* one virtual call Item_cond::functype(), which used to distinguish
between COND_AND_FUNC vs COND_OR_FUNC.

MDEV-8213: encryption.encryption_force, encryption.encrypt_and_grep fail with valgrind warnings (Invalid read)

Analysis: Problem was that code used old pointer.

Fixed by using correct pointer.

When the slave processes the master restart format_description event,
parallel replication needs to complete any prior events before processing
the restart event (which closes temporary tables and such stuff).

This happens in wait_for_workers_idle(), however it was not waiting long
enough. The wait was using wait_for_prior_commit(), but at that points table
can still be open. This lead to assertion in this case.

So change wait_for_workers_idle() to wait until all worker threads have
reached finish_event_group(), at which point all tables should have been
closed.