Define USE_MARIA_FOR_TMP_TABLES preprocessor constant by default if Aria engine is compiled in.
Use CMake variable WITH_ARIA_TMP_TABLES to control the temp table engine setting.

Added option to aria_read_log to crash recovery at certain points in the recovery process.
Fixed bug that caused future recovery attempts to fail if we got a crash/got killed during closing of tables at end of recovery process.


mysql-test/mysql-test-run.pl:
  Don't abort if 'var' points to stale place; Just remove it.
mysql-test/suite/maria/r/maria.result:
  Fixed wrong indentation
mysql-test/t/partition_error.test:
  Fixed portability problem with partiton_error.test
storage/maria/ma_close.c:
  More DBUG_PRINT info
storage/maria/ma_pagecache.c:
  Copy flush_log_callback when writing to page cache. This fixes problem in recovery when switching from mode of file
storage/maria/ma_recovery.c:
  Added option to aria_read_log to crash recovery at certain points in the recovery process.
storage/maria/ma_recovery.h:
  Added option to aria_read_log to crash recovery at certain points in the recovery process.
storage/maria/maria_chk.c:
  Align aria_chk -d output
  Don't write warning Aria table '...' is usable but should be fixed if the table was before marked as crashed but now is ok
storage/maria/maria_read_log.c:
  Added option to aria_read_log to crash recovery at certain points in the recovery process.

Noted that there was no memory leak, just a lot of used partitioned tables.
Fixed old bug: 'show status' now shows memory usage when compiled with safemalloc.
Added option --flush to mysqlcheck.c to run a 'flush tables' between each check to keep down memory usage.
Changed '--safemalloc' options to mysqld so that one can use --safemalloc and --skip-safemalloc.
Now skip-safemalloc is default (ie, we only do checking of memory overrun during free()) to speed up tests.


client/client_priv.h:
  Added OPT_FLUSH_TABLES
client/mysqlcheck.c:
  Added option --flush to mysqlcheck.c to run a 'flush tables' between each check to keep down memory usage.
mysql-test/mysql-test-run.pl:
  Always run tests with --loose-skip-safemysqld for higher speed
sql/mysqld.cc:
  Changed '--safemalloc' options so that one can use --safemalloc and --skip-safemalloc.
  Now skip-safemalloc is default (ie, we only do checking of memory overrun during free()) to speed up tests
sql/sql_parse.cc:
  Fixed old bug: 'show status' now shows memory usage when compiled with safemalloc.
storage/archive/archive_reader.c:
  Changed all malloc() calls to use my_malloc()/my_free()
  Added checks of malloc() calls.
storage/archive/ha_archive.cc:
  Detect failure if init_archive_reader() and return errno. This fixed assert crash in my_seek().
  Changed all malloc() calls to use my_malloc()/my_free()

Fixed reference to not initialized memory detected by valgrind

sql/sql_select.cc:
  A bit better fix for tmp-table problem:
  Use only dynamic_record format for group by and distinct.
storage/maria/ma_create.c:
  DYNAMIC_RECORD format doesn't pack VARCHAR fields.
  This change fixes a non-fatal uninitialized memory copy.

The reason for this is that BLOCK_RECORD format is not good when there is a lot of duplicated keys as it first writes the data (to get the row position) and
then writes the key (and thus checks for duplicates).

Fixed problem that global status variables 'bytes_recieved' and 'binlog_bytes_written' where not correctly updated 

After talking to HeidiSQL people -  libmysql.dll that comes with their distribution in place, dont replace with our own.
It also will result in less HeidiSQL restarts during MariaDB upgrades (since libmysql.dll won't be replaced)

VS did not like to execute 2 commands in custom build step, workaround is 
to use single COMMAND instead of 2.

Fixed lock sorting and lock check issues with thr_lock that caused warnings when running test suite.
Safety check that could cause core dump when doing create table with virtual column.

mysql-test/mysql-test-run.pl:
  Show also warnings from thr_lock (which starts with just Warning, not Warning:)
mysql-test/r/lock.result:
  Added test that showed not relevant warning when using table locks.
mysql-test/t/lock.test:
  Added test that showed not relevant warning when using table locks.
mysys/thr_lock.c:
  Fixed sorting of locks.
  (Old sort code didn't handle case where TL_WRITE_CONCURRENT_INSERT must be sorted before TL_WRITE)
  Added more information to check_locks warning output.
  Fixed wrong testing of multiple different write locks for same table.
sql/item_cmpfunc.cc:
  Safety check that could cause core dump when doing create table with virtual column.

Added HeidiSQL as example, i.e cmake -DWITH_THIRD_PARTY=HeidiSQL  
and building MSI will bundle HeidiSQL.

CMake 2.8.4 crashed on this line
 IF(something AND IS_DIRECTORY(something_else))
when both  "something" and "something_else" were empty.

Changing the line slightly (using cascading "IF" instead)  solved the crash.

Disable call to setpriority() in pbxt. This caused mysqld to run with nice priority -19, which was far from optimal.

mysql-test/suite/innodb/r/innodb_bug60049.result:
  Updated results
mysql-test/suite/innodb/t/innodb_bug60049.test:
  Force global.innodb_fast_shutdown to 0 as test require it
mysql-test/suite/innodb_plugin/t/innodb_bug60049.test:
  Force global.innodb_fast_shutdown to 0 as test require it
storage/pbxt/src/pthread_xt.cc:
  Disable call to setpriority()

The error message is now based on GetLastError() rather than errno. 

Background: errno is C runtime specific and in many circumstances 
it is not set, e.g when using Win32 API or socket functions.

This simplifies postmortem analysis for crashes reported via Winqual.

HEAP is deprecated in favor of MEMORY.

Ensure we don't read a MariaDB 5.3 errmsg.sys file or new errmsg.sys file with holes for not used error messages

sql/derror.cc:
  Ensure we don't read a MariaDB 5.3 errmsg.sys file with moved error messages or a new errmsg.sys file with holes for not used error messages
  If error message file didn't exist and we have not read one in the past, don't continue.
  Give better error message if the errmsg.sys header has changed.

- Fix CRLF end of lines, use LF instead

The printing of include stack in the error case in mysqltest omitted the
bottom of the stack (the line number in original test case file), and instead
printed the top of the stack twice. Fix to print each element on the stack
once and only once.

- add version info for the client library, dynamic plugins and some utilities
- do not recompile client library sources 3 times (for mysqlclient , mysqlclient_notls and libmysql)
  One time is sufficient, so get rid of mysqlclient_notls, and link  static client library  to  the shared.
- remove incremental linking flag

 my_rwlock_destroy=>rwlock_destroy

- Fix  active lock in freed memory in ha_archive (share mutex was not released prior to free())
- Do not attempt vio_fastsend or vio_keepalive on named pipes and shared memory.

LPBUG#782223 : Memory released by Query_cache::resize() or Query_cache::free() contains active rwlocks.
The bug was found by application verifier. 
Fixed by destroying locks prior to free(),

LPBUG#782269 : critical sections are initialized twice in   xt_xn_init_db() 
LPBUG#782431: active lock in memory released by   xt_ind_exit()
LPBUG#782433 : xt_heap_release() does not release spinlock  hp->h_lock initialized in xt_heap_new().
LPBUG#782435: xt_exit_row_locks() tries to release unallocated locks