MDEV-443: Galera: Server crashes on flushing tables for SST if started with character_set_server utf16 or utf32 or ucs2, and with wsrep_sst_method=rsync

Analysis: In SST Galera directly calls parser using current client character
set. Similarly in BF Galera uses client character set to apply. However,
there are character sets that are not currently supported by the parser.

Fix: If currenct client character set is one of those that is not supported
by the parser, temporally set character set to latin1 before we enter
parser and restore it after we have parsed.

held on lock_rec_other_has_conflicting that will also call (eventually)
wsrep_innobase_kill_one_trx. Added a new parameter have_kernel_mutex
to mark do we already own kernel mutex or not.

Analysis: Currently you can set binlog_format to STATEMENT or MIXED 
even when they are not really supported by galera.

Fix: Produce an error message if binlog_format is set to STATEMENT
or MIXED and wsrep_on = ON. Added a test case for this.

In mysqld_safe script, during the processing of command line
options, the supplied wsrep_provider option was not getting
appended to the set of mysqld options. Hence, wsrep plugin
would never receive the path for wsrep provider.

Fixed by adding logic to append the wsrep_provider option
to the set of mysqld options.

           reachable" in wsrep_init_startup on MTR tests

The initialization & de-initialization of wsrep plugin is solely
controlled by wsrep_on system variable. Now, as wsrep_on is ON by
default, the plugin gets initialized during server start. However,
there would be a leakage in case wsrep_on is turned off before plugin's
deinitialization (while server is shutting down) takes place, in which
case the deinit routine will not be invoked.

In this particular case, wsrep_on is turned off as no wsrep provider
has been specified & thus the valgrind warnings.

Fixed by introducing a flag which tracks and now controls the
(de|)initialization of the wsrep plugin instead on wsrep_on.
Added necessary asserts.

main.connect                             [ pass ]  16534
valgrind_report                          [ pass ]

  (thd && thd->variables.wsrep_on)) && wsrep_emulate_bin_log) ||
  mysql_bin_log .is_open()' fails on SAVEPOINT with disabled
  wsrep_provider

As wsrep_on system variable is ON by default the binlog handlerton
is implicitly enabled during the server initialization phase. Later,
while wsrep plugin is getting initialized, wsrep_on in corrected &
turned OFF in case wsrep provider is not specified (leaving binlog
in ON state!). This leads to the reported assertion.

Fixed by adding another check for wsrep provider before binlog
is enabled.

Added a test case.

Analysis: There is two different THD's using the same table.

Fix: In galera brute force (BF) exclusive locks are cancelled. Thus other threads
could access the same table. Added additional check if this is the case, then we
should not assert here.

Analysis: Code acquired kernel_mutex when calling wsrep_innobase_kill_one_trx.
However, functions below also might take kernel mutex thus the assertion. 

Fix: Acquire kernel_mutex only for lock_cancel_waiting_and_release where
it is really needed.

Analysis: Assertion means that InnoDB transaction lock queue is not
correctly set. Reason seams to be the fact that wrong lock is cancelled.

Fix: We have selected c_lock to be cancelled. Thus we should cancel
and release lock that is waiting for this lock i.e. c_lock->trx->wait_lock.

only present in the Galera tree

Fix for some compiler warnings.

```---------------------------------------------------------
revno: 3942
committer: Seppo Jaakola <seppo.jaakola@codership.com>
branch nick: wsrep-5.5-23
timestamp: Thu 2013-11-07 17:37:10 +0200
message:
  References: lp:1248921 - checking if index is foreign earlier to adhere to lat
ch order protocol
```

---------------------------------------------------------
revno: 3941
fixes bug: https://launchpad.net/bugs/1248908
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Thu 2013-11-07 14:31:04 +0200
message:
  References lp:1248908 - Fixed access to trx sys header
    
  Pass trx sys header as argument for trx_sys_update_wsrep_checkpoint()
  and trx_sys_update_mysql_binlog_offset() to avoid successive calls
  to trx_sysf_get().
------------------------------------------------------------
revno: 3940
fixes bug: https://launchpad.net/bugs/1244100
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Tue 2013-11-05 15:31:12 +0200
message:
  References lp:1244100 - assign value for check_purge before return
------------------------------------------------------------
revno: 3939
fixes bug: https://launchpad.net/bugs/1247978
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Tue 2013-11-05 01:01:36 +0200
message:
  References lp:1247978 - force release transactional MDL locks if wsrep is on a
nd no active transaction at the end of mysql_execute_command()
------------------------------------------------------------
revno: 3938
committer: Vladislav Klyachin <vladislav.klyachin@codership.com>
branch nick: 5.5-23
timestamp: Sun 2013-11-03 20:48:06 +0400
message:
  References lp:1232789 - fix FLUSH STATUS zeroes up wsrep_cluster_size and wsre
p_local_index
------------------------------------------------------------
revno: 3937  5kB/s - 
committer: Vladislav Klyachin <vladislav.klyachin@codership.com>
branch nick: 5.5-23
timestamp: Sun 2013-11-03 17:06:31 +0400
message:
  References lp:587170 - reset auto_increment_* vars to defaults for TOI operati
ons
------------------------------------------------------------
revno: 3936
committer: Vladislav Klyachin <vladislav.klyachin@codership.com>
branch nick: 5.5-23
timestamp: Sun 2013-11-03 00:10:45 +0400
message:
  References lp:1072839 - fix for memory leaks with wsrep variables
------------------------------------------------------------
revno: 3935
committer: Seppo Jaakola <seppo.jaakola@codership.com>
branch nick: wsrep-5.5-23
timestamp: Wed 2013-10-30 14:34:32 +0200
message:
  References: lp:1246257 - skipping replication for CREATE TEMPORARY TABLE LIKE.
..  constructs
------------------------------------------------------------
revno: 3934  6kB/s \ 
fixes bug: https://launchpad.net/bugs/1241760
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Mon 2013-10-28 11:01:53 +0200
message:
  References lp:1241760 - save thd->db to wsrep_thd_shadow before replay
------------------------------------------------------------
revno: 3933
fixes bug: https://launchpad.net/bugs/1206129
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Sun 2013-10-27 18:15:12 +0200
message:
  References lp:1206129 - check binlog_hton->commit() return value, call wsrep_p
ost_commit() instead of wsrep_cleanup_transaction()
------------------------------------------------------------
revno: 3932
fixes bug: https://launchpad.net/bugs/1244661
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Sun 2013-10-27 13:14:02 +0200
message:
  References lp:1244661 - added wsrep_register_hton() to trans_rollback_implicit
()
------------------------------------------------------------
revno: 3931
fixes bug: https://launchpad.net/bugs/1244667
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Sun 2013-10-27 11:08:49 +0200
message:
  References lp:1244667 - restore thd->server_status after replay

```---------------------------------------------------------
revno: 3929 [merge]
fixes bug: https://launchpad.net/bugs/1243150
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Wed 2013-10-23 20:05:01 +0300
message:     8kB/s - 
  References lp:1243150 - initial wsrep hton cleanups
  
  * Removed wsrep_seqno_changed boolean
  * wsrep_cleanup_transaction() is now called explicitly whenever it is
    clear that transaction has come to an end
  * wsrep_trans_cache_is_empty() now checks from cache_mngr recardless of
    command type
  * Separated call to wsrep->post_commit() to own function, called from
    transaction.cc whenever appropriate
  * wsrep_thd_is_brute_force() now investigates only thd->wsrep_exec_mode
  * More comments and debug time assertions
  * Debug code to check that wsrep position stored in InnoDB is
    monotinically increasing. Enabled with UNIV_DEBUG
```

---------------------------------------------------------
revno: 3928
fixes bug: https://launchpad.net/bugs/1237889
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Tue 2013-10-22 22:01:20 +0300
message:
  References lp:1237889 - reverting fix in r3926, it broke crash recovery
------------------------------------------------------------
revno: 3927
fixes bug: https://launchpad.net/bugs/1240040
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Tue 2013-10-15 14:46:15 +0300
message:
  References lp:1240040 - added WSREP_MYSQL_DB as a key for DROP VIEW
------------------------------------------------------------
revno: 3926
fixes bug: https://launchpad.net/bugs/1237889
committer: Teemu Ollakka <teemu.ollakka@codership.com>
branch nick: 5.5-23
timestamp: Thu 2013-10-10 14:22:58 +0300
message:
  References lp:1237889 - register wsrep hton only if thd->wsrep_exec_mode == LO
CAL_STATE
------------------------------------------------------------
revno: 3925
fixes bug: https://launchpad.net/bugs/1235635
committer: Alexey Yurchenko <alexey.yurchenko@codership.com>
branch nick: 5.5-23
timestamp: Sat 2013-10-05 18:03:06 +0300
message:
  References lp:1235635 - fixed the warning by initializing c_lock to NULL.

This patch fixes both the reported issues :
(1) Included '-wsrep' to the server version information.

> select @@version;
+-----------------------------+
| @@version                   |
+-----------------------------+
| 5.5.33a-MariaDB-wsrep-debug |
+-----------------------------+

(2) The @@version_comment would no longer have XXXX
in the wsrep version information in case cmake couldn't
determine the current bzr revision.

            -DINSTALL_SYSCONFDIR/-DDEFAULT_SYSCONFDIR specified

The reason for a bug is in support-files/CMakeLists.txt that tries to
install files in INSTALL_SYSCONF2DIR, without checking whether this
variable is actually defined. 

Normally INSTALL_SYSCONF2DIR is always defined if INSTALL_SYSCONFDIR is
defined, and support-files/CMakeLists.txt assumes that. But when one
specifies INSTALL_SYSCONFDIR manually on the command line, this
assumption might be broken.

(some patches reverted, test case added)

sort xtradb status variables

MDEV-5069: Server crashes in SEL_ARG::increment_use_count with index_merge+index_merge_sort_union, FORCE INDEX
- Don't call incr_refs() is the merged SEL_ARG* is NULL.

Added test case for new system variable innodb_use_stacktrace and made sure that it can be used only on startup. Fixed compiler problems on solaris and other platforms that do not contain necessary headers and functions.

MDEV-5247: DB locked up at btr0cur.c line 568. Additional fixes to inconsistent usage of have_LRU_mutex and added additional debug assertions to guard incorrect usage of this mutex. Fixes issues found on additional testing and mysql test suite.

This bug was a consequence of the incorrect fix for bug mdev-5091.

Fix compiler error introduced on revision 3937, make sure that stackdump is compiled only on __linux__.

mysql_tzinfo_to_sql is not compiled under Windows in the pre-5.5 versions.
Do not set the MYSQL_TZINFO_TO_SQL variable in the version prior to 5.5
under Windows.

MDEV-5247: DB locked up at btr0cur.c line 568. There is inconsistent and non logical usage of have_LRU_mutex and incorrect value on ha_innodb.cc when buf_LRU_free_block is called. Additionally, for future long semaphore wait cases added a new configuration variable innodb_use_stacktrace. If this variable is true a signal handler for SIGUSR2 is installed when InnoDB server starts and when a long semaphore wait is detected at sync/sync0array.c we send SIGUSR2 signal to waiting thread and thread that has acuired RW-latch. For both threads a full stacktrace is produced as well as its is possible.

The function SELECT_LEX::update_used_tables should process all
ORDER BY lists in any subselect of a union.