DISABLE_MYSQL_THREAD_H for strings
Also, enable unittests for perfschema

add WITH_MYSQLD_LIBS for extra linker flags for mysqld

Part V: performance schema implementation

Original revision:
------------------------------------------------------------
revno: 3817
revision-id: guilhem@mysql.com-20100108092756-k0zzf4kvx9b7bh38
parent: guilhem@mysql.com-20100107101133-hrrgcdqg508runuf
committer: Guilhem Bichot <guilhem@mysql.com>
branch nick: mysql-6.0-codebase-bugfixing
timestamp: Fri 2010-01-08 10:27:56 +0100
message:
  fix for BUG#50120 "Valgrind errors in any test, inside mysqltest"
  Problem was that as v->name[v->name_len] may be uninitialized (which is ok per se),
  it shouldn't be used in an if(). We remove this zero_the_char/restore_it logic by
  rather zero-terminating the v->name string when we create it in var_init().
------------------------------------------------------------

Original revision:

------------------------------------------------------------
revno: 3789.1.9
revision-id: serg@mysql.com-20091229134448-phe834ukzmi0k2e3
parent: serg@mysql.com-20091227081418-bgfg952gzumn1k3h
committer: Sergei Golubchik <serg@mysql.com>
branch nick: 6.0-codebase
timestamp: Tue 2009-12-29 14:44:48 +0100
message:
  better fix for Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds
  
  use setenv instead of putenv
------------------------------------------------------------

Add mysql_release.cmake intended for use by the build team (
compile options, build settings to do MySQL releases)

(handle dependent options, allow user with change WITH_READLINE to
WITH_LIBEDIT and other way around after the first compile)

Part IV: sql instrumentation

in mysqld--help test.

When mysql is compiled with different options,columns
might have different size (this depends on parameter 
names and some parameters might only be available 
when some option is switched on)

mysql-test/r/index_merge_myisam.result:
  @@optimizer_switch has a new flag
mysql-test/r/mysqld--help-notwin.result:
  @@optimizer_switch has a new flag
mysql-test/r/mysqld--help-win.result:
  @@optimizer_switch has a new flag
mysql-test/r/optimizer_switch_eng_cond_pushdown1.result:
  Check how --engine-condition-pushdown and --optimizer-switch influence each other when used together (last wins).
mysql-test/r/optimizer_switch_eng_cond_pushdown2.result:
  Check how --engine-condition-pushdown and --optimizer-switch influence each other when used together (last wins).
mysql-test/suite/ndb/r/ndb_condition_pushdown.result:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
mysql-test/suite/ndb/r/ndb_gis.result:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
mysql-test/suite/ndb/r/ndb_index_unique.result:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
mysql-test/suite/ndb/t/ndb_condition_pushdown.test:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
mysql-test/suite/ndb/t/ndb_gis.test:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
mysql-test/suite/ndb/t/ndb_index_unique.test:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
mysql-test/suite/sys_vars/r/engine_condition_pushdown_basic.result:
  Setting @@engine_condition_pushdown gives a deprecation warning now.
  We test that the engine_condition_pushdown flag of @@optimizer_switch, and @@engine_condition_pushdown
  influence each other (turning the flag on/off sets the variable on/off and vice-versa).
mysql-test/suite/sys_vars/r/optimizer_switch_basic.result:
  @@optimizer_switch has a new flag
mysql-test/suite/sys_vars/t/engine_condition_pushdown_basic.test:
  Setting @@engine_condition_pushdown gives a deprecation warning now.
  We test that the engine_condition_pushdown flag of @@optimizer_switch, and @@engine_condition_pushdown
  influence each other (turning the flag on/off sets the variable on/off and vice-versa).
mysql-test/t/optimizer_switch_eng_cond_pushdown1-master.opt:
  Check how --engine-condition-pushdown and --optimizer-switch influence each other when used together (last wins).
mysql-test/t/optimizer_switch_eng_cond_pushdown1.test:
  Check how --engine-condition-pushdown and --optimizer-switch influence each other when used together (last wins).
mysql-test/t/optimizer_switch_eng_cond_pushdown2-master.opt:
  Check how --engine-condition-pushdown and --optimizer-switch influence each other when used together (last wins).
mysql-test/t/optimizer_switch_eng_cond_pushdown2.test:
  Check how --engine-condition-pushdown and --optimizer-switch influence each other when used together (last wins).
sql/mysql_priv.h:
  new "engine_condition_pushdown" switch in @@optimizer_switch, on by default, like
  @@engine_condition_pushdown is on by default. Constants are ULL because optimizer_switch
  is stored in a ulonglong.
sql/mysqld.cc:
  Making --engine-condition-pushdown and --optimizer-switch (command-line options)
  influence each other (last wins)
sql/records.cc:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
sql/sql_select.cc:
  @@engine_condition_pushdown is deprecated, use @@optimizer_switch instead
sql/sys_vars.cc:
  Setting @@engine_condition_pushdown now issues a deprecation message. The version for removal
  is unknown at this point so I copied it from other deprecation warnings in this file.
  Turning on/off the engine_condition_pushdown flag of @@optimizer_switch (with SET) turns on/off the @@engine_condition_pushdown variable, and vice-versa, thanks to fix_* functions.

Subtraction of two unsigned months yielded a (very large) positive value.
Conversion of this to a signed value was not necessarily well defined.

Solution: do the subtraction on signed values.


mysql-test/r/events_scheduling.result:
  Add test case.
mysql-test/t/events_scheduling.test:
  Add test case.
sql/event_data_objects.cc:
  Convert month to signed before doing the subtraction.