Patch that refactors global read lock implementation and fixes

bug #57006 "Deadlock between HANDLER and FLUSH TABLES WITH READ
LOCK" and bug #54673 "It takes too long to get readlock for
'FLUSH TABLES WITH READ LOCK'".

The first bug manifested itself as a deadlock which occurred
when a connection, which had some table open through HANDLER
statement, tried to update some data through DML statement
while another connection tried to execute FLUSH TABLES WITH
READ LOCK concurrently.

What happened was that FTWRL in the second connection managed
to perform first step of GRL acquisition and thus blocked all
upcoming DML. After that it started to wait for table open
through HANDLER statement to be flushed. When the first connection
tried to execute DML it has started to wait for GRL/the second
connection creating deadlock.

The second bug manifested itself as starvation of FLUSH TABLES
WITH READ LOCK statements in cases when there was a constant
stream of concurrent DML statements (in two or more
connections).

This has happened because requests for protection against GRL
which were acquired by DML statements were ignoring presence of
pending GRL and thus the latter was starved.

This patch solves both these problems by re-implementing GRL
using metadata locks.

Similar to the old implementation acquisition of GRL in new
implementation is two-step. During the first step we block
all concurrent DML and DDL statements by acquiring global S
metadata lock (each DML and DDL statement acquires global IX
lock for its duration). During the second step we block commits
by acquiring global S lock in COMMIT namespace (commit code
acquires global IX lock in this namespace).

Note that unlike in old implementation acquisition of
protection against GRL in DML and DDL is semi-automatic.
We assume that any statement which should be blocked by GRL
will either open and acquires write-lock on tables or acquires
metadata locks on objects it is going to modify. For any such
statement global IX metadata lock is automatically acquired
for its duration.

The first problem is solved because waits for GRL become
visible to deadlock detector in metadata locking subsystem
and thus deadlocks like one in the first bug become impossible.

The second problem is solved because global S locks which
are used for GRL implementation are given preference over
IX locks which are acquired by concurrent DML (and we can
switch to fair scheduling in future if needed).

Important change:
FTWRL/GRL no longer blocks DML and DDL on temporary tables.
Before this patch behavior was not consistent in this respect:
in some cases DML/DDL statements on temporary tables were
blocked while in others they were not. Since the main use cases
for FTWRL are various forms of backups and temporary tables are
not preserved during backups we have opted for consistently
allowing DML/DDL on temporary tables during FTWRL/GRL.

Important change:
This patch changes thread state names which are used when
DML/DDL of FTWRL is waiting for global read lock. It is now
either "Waiting for global read lock" or "Waiting for commit
lock" depending on the stage on which FTWRL is.

Incompatible change:
To solve deadlock in events code which was exposed by this
patch we have to replace LOCK_event_metadata mutex with
metadata locks on events. As result we have to prohibit
DDL on events under LOCK TABLES.

This patch also adds extensive test coverage for interaction
of DML/DDL and FTWRL.

Performance of new and old global read lock implementations
in sysbench tests were compared. There were no significant
difference between new and old implementations.

mysql-test/include/check_ftwrl_compatible.inc:
  Added helper script which allows to check that a statement is
  compatible with FLUSH TABLES WITH READ LOCK.
mysql-test/include/check_ftwrl_incompatible.inc:
  Added helper script which allows to check that a statement is
  incompatible with FLUSH TABLES WITH READ LOCK.
mysql-test/include/handler.inc:
  Adjusted test case to the fact that now DROP TABLE closes
  open HANDLERs for the table to be dropped before checking
  if there active FTWRL in this connection.
mysql-test/include/wait_show_condition.inc:
  Fixed small error in the timeout message. The correct name
  of variable used as parameter for this script is "$condition"
  and not "$wait_condition".
mysql-test/r/delayed.result:
  Added test coverage for scenario which triggered assert in
  metadata locking subsystem.
mysql-test/r/events_2.result:
  Updated test results after prohibiting event DDL operations
  under LOCK TABLES.
mysql-test/r/flush.result:
  Added test coverage for bug #57006 "Deadlock between HANDLER
  and FLUSH TABLES WITH READ LOCK".
mysql-test/r/flush_read_lock.result:
  Added test coverage for various aspects of FLUSH TABLES WITH
  READ LOCK functionality.
mysql-test/r/flush_read_lock_kill.result:
  Adjusted test case after replacing custom global read lock
  implementation with one based on metadata locks. Use new
  debug_sync point. Do not disable concurrent inserts as now
  InnoDB we always use InnoDB table.
mysql-test/r/handler_innodb.result:
  Adjusted test case to the fact that now DROP TABLE closes
  open HANDLERs for the table to be dropped before checking
  if there active FTWRL in this connection.
mysql-test/r/handler_myisam.result:
  Adjusted test case to the fact that now DROP TABLE closes
  open HANDLERs for the table to be dropped before checking
  if there active FTWRL in this connection.
mysql-test/r/mdl_sync.result:
  Adjusted test case after replacing custom global read lock
  implementation with one based on metadata locks. Replaced
  usage of GRL-specific debug_sync's with appropriate sync
  points in MDL subsystem.
mysql-test/suite/perfschema/r/dml_setup_instruments.result:
  Updated test results after removing global
  COND_global_read_lock condition variable.
mysql-test/suite/perfschema/r/func_file_io.result:
  Ensure that this test doesn't affect subsequent tests.
  At the end of its execution enable back P_S instrumentation
  which this test disables at some point.
mysql-test/suite/perfschema/r/func_mutex.result:
  Ensure that this test doesn't affect subsequent tests.
  At the end of its execution enable back P_S instrumentation
  which this test disables at some point.
mysql-test/suite/perfschema/r/global_read_lock.result:
  Adjusted test case to take into account that new GRL
  implementation is based on MDL.
mysql-test/suite/perfschema/r/server_init.result:
  Adjusted test case after replacing custom global read
  lock implementation with one based on MDL and replacing
  LOCK_event_metadata mutex with metadata lock.
mysql-test/suite/perfschema/t/func_file_io.test:
  Ensure that this test doesn't affect subsequent tests.
  At the end of its execution enable back P_S instrumentation
  which this test disables at some point.
mysql-test/suite/perfschema/t/func_mutex.test:
  Ensure that this test doesn't affect subsequent tests.
  At the end of its execution enable back P_S instrumentation
  which this test disables at some point.
mysql-test/suite/perfschema/t/global_read_lock.test:
  Adjusted test case to take into account that new GRL
  implementation is based on MDL.
mysql-test/suite/perfschema/t/server_init.test:
  Adjusted test case after replacing custom global read
  lock implementation with one based on MDL and replacing
  LOCK_event_metadata mutex with metadata lock.
mysql-test/suite/rpl/r/rpl_tmp_table_and_DDL.result:
  Updated test results after prohibiting event DDL under
  LOCK TABLES.
mysql-test/t/delayed.test:
  Added test coverage for scenario which triggered assert in
  metadata locking subsystem.
mysql-test/t/events_2.test:
  Updated test case after prohibiting event DDL operations
  under LOCK TABLES.
mysql-test/t/flush.test:
  Added test coverage for bug #57006 "Deadlock between HANDLER
  and FLUSH TABLES WITH READ LOCK".
mysql-test/t/flush_block_commit.test:
  Adjusted test case after changing thread state name which
  is used when COMMIT waits for FLUSH TABLES WITH READ LOCK
  from "Waiting for release of readlock" to "Waiting for commit
  lock".
mysql-test/t/flush_block_commit_notembedded.test:
  Adjusted test case after changing thread state name which is
  used when DML waits for FLUSH TABLES WITH READ LOCK. Now we
  use "Waiting for global read lock" in this case.
mysql-test/t/flush_read_lock.test:
  Added test coverage for various aspects of FLUSH TABLES WITH
  READ LOCK functionality.
mysql-test/t/flush_read_lock_kill-master.opt:
  We no longer need to use make_global_read_lock_block_commit_loop
  debug tag in this test. Instead we rely on an appropriate
  debug_sync point in MDL code.
mysql-test/t/flush_read_lock_kill.test:
  Adjusted test case after replacing custom global read lock
  implementation with one based on metadata locks. Use new
  debug_sync point. Do not disable concurrent inserts as now
  InnoDB we always use InnoDB table.
mysql-test/t/lock_multi.test:
  Adjusted test case after changing thread state names which
  are used when DML or DDL waits for FLUSH TABLES WITH READ
  LOCK to "Waiting for global read lock".
mysql-test/t/mdl_sync.test:
  Adjusted test case after replacing custom global read lock
  implementation with one based on metadata locks. Replaced
  usage of GRL-specific debug_sync's with appropriate sync
  points in MDL subsystem. Updated thread state names which
  are used when DDL waits for FTWRL.
mysql-test/t/trigger_notembedded.test:
  Adjusted test case after changing thread state names which
  are used when DML or DDL waits for FLUSH TABLES WITH READ
  LOCK to "Waiting for global read lock".
sql/event_data_objects.cc:
  Removed Event_queue_element::status/last_executed_changed
  members and Event_queue_element::update_timing_fields()
  method. We no longer use this class for updating mysql.events
  once event is chosen for execution. Accesses to instances of
  this class in scheduler thread require protection by
  Event_queue::LOCK_event_queue mutex and we try to avoid
  updating table while holding this lock.
sql/event_data_objects.h:
  Removed Event_queue_element::status/last_executed_changed
  members and Event_queue_element::update_timing_fields()
  method. We no longer use this class for updating mysql.events
  once event is chosen for execution. Accesses to instances of
  this class in scheduler thread require protection by
  Event_queue::LOCK_event_queue mutex and we try to avoid
  updating table while holding this lock.
sql/event_db_repository.cc:
  - Changed Event_db_repository methods to not release all
    metadata locks once they are done updating mysql.events
    table. This allows to keep metadata lock protecting
    against GRL and lock protecting particular event around
    until corresponding DDL statement is written to the binary
    log.
  - Removed logic for conditional update of "status" and
    "last_executed" fields from update_timing_fields_for_event()
    method. In the only case when this method is called now
    "last_executed" is always modified and tracking change
    of "status" is too much hassle.
sql/event_db_repository.h:
  Removed logic for conditional update of "status" and
  "last_executed" fields from Event_db_repository::
  update_timing_fields_for_event() method.
  In the only case when this method is called now "last_executed"
  is always modified and tracking change of "status" field is
  too much hassle.
sql/event_queue.cc:
  Changed event scheduler code not to update mysql.events
  table while holding Event_queue::LOCK_event_queue mutex.
  Doing so led to a deadlock with a new GRL implementation.
  This deadlock didn't occur with old implementation due to
  fact that code acquiring protection against GRL ignored
  pending GRL requests (which lead to GRL starvation).
  One of goals of new implementation is to disallow GRL
  starvation and so we have to solve problem with this
  deadlock in a different way.
sql/events.cc:
  Changed methods of Events class to acquire protection
  against GRL while perfoming DDL statement and keep it
  until statement is written to the binary log.
  Unfortunately this step together with new GRL implementation
  exposed deadlock involving Events::LOCK_event_metadata
  and GRL. To solve it Events::LOCK_event_metadata mutex was
  replaced with a metadata lock on event. As a side-effect
  events DDL has to be prohibited under LOCK TABLES even in
  cases when mysql.events table was explicitly locked for
  write.
sql/events.h:
  Replaced Events::LOCK_event_metadata mutex with a metadata
  lock on event.
sql/ha_ndbcluster.cc:
  Updated code after replacing custom global read lock
  implementation with one based on MDL. Since MDL subsystem
  should now be able to detect deadlocks involving metadata
  locks and GRL there is no need for special handling of
  active GRL.
sql/handler.cc:
  Replaced custom implementation of global read lock with
  one based on metadata locks. Consequently when doing
  commit instead of calling method of Global_read_lock
  class to acquire protection against GRL we simply acquire
  IX in COMMIT namespace.
sql/lock.cc:
  Replaced custom implementation of global read lock with
  one based on metadata locks. This step allows to expose
  wait for GRL to deadlock detector of MDL subsystem and
  thus succesfully resolve deadlocks similar to one behind
  bug #57006 "Deadlock between HANDLER and FLUSH TABLES
  WITH READ LOCK". It also solves problem with GRL starvation
  described in bug #54673 "It takes too long to get readlock
  for 'FLUSH TABLES WITH READ LOCK'" since metadata locks used
  by GRL give preference to FTWRL statement instead of DML
  statements (if needed in future this can be changed to
  fair scheduling).
  
  Similar to old implementation of acquisition of GRL is
  two-step. During the first step we block all concurrent
  DML and DDL statements by acquiring global S metadata lock
  (each DML and DDL statement acquires global IX lock for
  its duration). During the second step we block commits by
  acquiring global S lock in COMMIT namespace (commit code
  acquires global IX lock in this namespace).
  
  Note that unlike in old implementation acquisition of
  protection against GRL in DML and DDL is semi-automatic.
  We assume that any statement which should be blocked by GRL
  will either open and acquires write-lock on tables or acquires
  metadata locks on objects it is going to modify. For any such
  statement global IX metadata lock is automatically acquired
  for its duration.
  
  To support this change:
  - Global_read_lock::lock/unlock_global_read_lock and
    make_global_read_lock_block_commit methods were changed
    accordingly.
  - Global_read_lock::wait_if_global_read_lock() and
    start_waiting_global_read_lock() methods were dropped.
    It is now responsibility of code acquiring metadata locks
    opening tables to acquire protection against GRL by
    explicitly taking global IX lock with statement duration.
  - Global variables, mutex and condition variable used by
    old implementation was removed.
  - lock_routine_name() was changed to use statement duration for
    its global IX lock. It was also renamed to lock_object_name()
    as it now also used to take metadata locks on events.
  - Global_read_lock::set_explicit_lock_duration() was added which
    allows not to release locks used for GRL when leaving prelocked
    mode.
sql/lock.h:
  - Renamed lock_routine_name() to lock_object_name() and changed
    its signature to allow its usage for events.
  - Removed broadcast_refresh() function. It is no longer needed
    with new GRL implementation.
sql/log_event.cc:
  Release metadata locks with statement duration at the end
  of processing legacy event for LOAD DATA. This ensures that
  replication thread processing such event properly releases
  its protection against global read lock.
sql/mdl.cc:
  Changed MDL subsystem to support new MDL-based implementation
  of global read lock.
  
  Added COMMIT and EVENTS namespaces for metadata locks. Changed
  thread state name for GLOBAL namespace to "Waiting for global
  read lock".
  
  Optimized MDL_map::find_or_insert() method to avoid taking
  m_mutex mutex when looking up MDL_lock objects for GLOBAL
  or COMMIT namespaces. We keep pre-created MDL_lock objects
  for these namespaces around and simply return pointers to
  these global objects when needed.
  
  Changed MDL_lock/MDL_scoped_lock to properly handle
  notification of insert delayed handler threads when FTWRL
  takes global S lock.
  
  Introduced concept of lock duration. In addition to locks with
  transaction duration which work in the way which is similar to
  how locks worked before (i.e. they are released at the end of
  transaction), locks with statement and explicit duration were
  introduced.
  Locks with statement duration are automatically released at the
  end of statement. Locks with explicit duration require explicit
  release and obsolete concept of transactional sentinel.
  
  * Changed MDL_request and MDL_ticket classes to support notion
    of duration.
  * Changed MDL_context to keep locks with different duration in
    different lists. Changed code handling ticket list to take
    this into account.
  * Changed methods responsible for releasing locks to take into
    account duration of tickets. Particularly public
    MDL_context::release_lock() method now only can release
    tickets with explicit duration (there is still internal
    method which allows to specify duration). To release locks
    with statement or transaction duration one have to use
    release_statement/transactional_locks() methods.
  * Concept of savepoint for MDL subsystem now has to take into
    account locks with statement duration. Consequently
    MDL_savepoint class was introduced and methods working with
    savepoints were updated accordingly.
  * Added methods which allow to set duration for one or all
    locks in the context.
sql/mdl.h:
  Changed MDL subsystem to support new MDL-based implementation
  of global read lock.
  
  Added COMMIT and EVENTS namespaces for metadata locks.
  
  Introduced concept of lock duration. In addition to locks with
  transaction duration which work in the way which is similar to
  how locks worked before (i.e. they are released at the end of
  transaction), locks with statement and explicit duration were
  introduced.
  Locks with statement duration are automatically released at the
  end of statement. Locks with explicit duration require explicit
  release and obsolete concept of transactional sentinel.
  
  * Changed MDL_request and MDL_ticket classes to support notion
    of duration.
  * Changed MDL_context to keep locks with different duration in
    different lists. Changed code handling ticket list to take
    this into account.
  * Changed methods responsible for releasing locks to take into
    account duration of tickets. Particularly public
    MDL_context::release_lock() method now only can release
    tickets with explicit duration (there is still internal
    method which allows to specify duration). To release locks
    with statement or transaction duration one have to use
    release_statement/transactional_locks() methods.
  * Concept of savepoint for MDL subsystem now has to take into
    account locks with statement duration. Consequently
    MDL_savepoint class was introduced and methods working with
    savepoints were updated accordingly.
  * Added methods which allow to set duration for one or all
    locks in the context.
sql/mysqld.cc:
  Removed global mutex and condition variables which were used
  by old implementation of GRL.
  Also we no longer need to initialize Events::LOCK_event_metadata
  mutex as it was replaced with metadata locks on events.
sql/mysqld.h:
  Removed global variable, mutex and condition variables which
  were used by old implementation of GRL.
sql/rpl_rli.cc:
  When slave thread closes tables which were open for handling
  of RBR events ensure that it releases global IX lock which
  was acquired as protection against GRL.
sql/sp.cc:
  Adjusted code to the new signature of lock_object/routine_name(),
  to the fact that one now needs specify duration of lock when
  initializing MDL_request and to the fact that savepoints for MDL
  subsystem are now represented by MDL_savepoint class.
sql/sp_head.cc:
  Ensure that statements in stored procedures release statement
  metadata locks and thus release their protectiong against GRL
  in proper moment in time.
  Adjusted code to the fact that one now needs specify duration
  of lock when initializing MDL_request.
sql/sql_admin.cc:
  Adjusted code to the fact that one now needs specify duration
  of lock when initializing MDL_request.
sql/sql_base.cc:
  - Implemented support for new approach to acquiring protection
    against global read lock. We no longer acquire such protection
    explicitly on the basis of statement flags. Instead we always
    rely on code which is responsible for acquiring metadata locks
    on object to be changed acquiring this protection. This is
    achieved by acquiring global IX metadata lock with statement
    duration. Code doing this also responsible for checking that
    current connection has no active GRL by calling an
    Global_read_lock::can_acquire_protection() method.
    Changed code in open_table() and lock_table_names()
    accordingly.
    Note that as result of this change DDL and DML on temporary
    tables is always compatible with GRL (before it was
    incompatible in some cases and compatible in other cases).
  - To speed-up code acquiring protection against GRL introduced
    m_has_protection_against_grl member in Open_table_context
    class. It indicates that protection was already acquired
    sometime during open_tables() execution and new attempts
    can be skipped.
  - Thanks to new GRL implementation calls to broadcast_refresh()
    became unnecessary and were removed.
  - Adjusted code to the fact that one now needs specify duration
    of lock when initializing MDL_request and to the fact that
    savepoints for MDL subsystem are now represented by
    MDL_savepoint class.
sql/sql_base.h:
  Adjusted code to the fact that savepoints for MDL subsystem are
  now represented by MDL_savepoint class.
  Also introduced Open_table_context::m_has_protection_against_grl
  member which allows to avoid acquiring protection against GRL
  while opening tables if such protection was already acquired.
sql/sql_class.cc:
  Changed THD::leave_locked_tables_mode() after transactional
  sentinel for metadata locks was obsoleted by introduction of
  locks with explicit duration.
sql/sql_class.h:
  - Adjusted code to the fact that savepoints for MDL subsystem
    are now represented by MDL_savepoint class.
  - Changed Global_read_lock class according to changes in
    global read lock implementation:
    * wait_if_global_read_lock and start_waiting_global_read_lock
      are now gone. Instead code needing protection against GRL
      has to acquire global IX metadata lock with statement
      duration itself. To help it new can_acquire_protection()
      was introduced. Also as result of the above change
      m_protection_count member is gone too.
    * Added m_mdl_blocks_commits_lock member to store metadata
      lock blocking commits.
    * Adjusted code to the fact that concept of transactional
      sentinel was obsoleted by concept of lock duration.
  - Removed CF_PROTECT_AGAINST_GRL flag as it is no longer
    necessary. New GRL implementation acquires protection
    against global read lock automagically when statement
    acquires metadata locks on tables or other objects it
    is going to change.
sql/sql_db.cc:
  Adjusted code to the fact that one now needs specify duration
  of lock when initializing MDL_request.
sql/sql_handler.cc:
  Removed call to broadcast_refresh() function. It is no longer
  needed with new GRL implementation.
  Adjusted code after introducing duration concept for metadata
  locks. Particularly to the fact transactional sentinel was
  replaced with explicit duration.
sql/sql_handler.h:
  Renamed mysql_ha_move_tickets_after_trans_sentinel() to
  mysql_ha_set_explicit_lock_duration() after transactional
  sentinel was obsoleted by locks with explicit duration.
sql/sql_insert.cc:
  Adjusted code handling delaying inserts after switching to
  new GRL implementation. Now connection thread initiating
  delayed insert has to acquire global IX lock in addition
  to metadata lock on table being inserted into. This IX lock
  protects against GRL and similarly to SW lock on table being
  inserted into has to be passed to handler thread in order to
  avoid deadlocks.
sql/sql_lex.cc:
  LEX::protect_against_global_read_lock member is no longer
  necessary since protection against GRL is automatically
  taken by code acquiring metadata locks/opening tables.
sql/sql_lex.h:
  LEX::protect_against_global_read_lock member is no longer
  necessary since protection against GRL is automatically
  taken by code acquiring metadata locks/opening tables.
sql/sql_parse.cc:
  - Implemented support for new approach to acquiring protection
    against global read lock. We no longer acquire such protection
    explicitly on the basis of statement flags. Instead we always
    rely on code which is responsible for acquiring metadata locks
    on object to be changed acquiring this protection. This is
    achieved by acquiring global IX metadata lock with statement
    duration. This lock is automatically released at the end of
    statement execution.
  - Changed implementation of CREATE/DROP PROCEDURE/FUNCTION not
    to release metadata locks and thus protection against of GRL
    in the middle of statement execution.
  - Adjusted code to the fact that one now needs specify duration
    of lock when initializing MDL_request and to the fact that
    savepoints for MDL subsystem are now represented by
    MDL_savepoint class.
sql/sql_prepare.cc:
  Adjusted code to the to the fact that savepoints for MDL
  subsystem are now represented by MDL_savepoint class.
sql/sql_rename.cc:
  With new GRL implementation there is no need to explicitly
  acquire protection against GRL before renaming tables.
  This happens automatically in code which acquires metadata
  locks on tables being renamed.
sql/sql_show.cc:
  Adjusted code to the fact that one now needs specify duration
  of lock when initializing MDL_request and to the fact that
  savepoints for MDL subsystem are now represented by
  MDL_savepoint class.
sql/sql_table.cc:
  - With new GRL implementation there is no need to explicitly
    acquire protection against GRL before dropping tables.
    This happens automatically in code which acquires metadata
    locks on tables being dropped.
  - Changed mysql_alter_table() not to release lock on new table
    name explicitly and to rely on automatic release of locks
    at the end of statement instead. This was necessary since
    now MDL_context::release_lock() is supported only for locks
    for explicit duration.
sql/sql_trigger.cc:
  With new GRL implementation there is no need to explicitly
  acquire protection against GRL before changing table triggers.
  This happens automatically in code which acquires metadata
  locks on tables which triggers are to be changed.
sql/sql_update.cc:
  Fix bug exposed by GRL testing. During prepare phase acquire
  only S metadata locks instead of SW locks to keep prepare of
  multi-UPDATE compatible with concurrent LOCK TABLES WRITE
  and global read lock.
sql/sql_view.cc:
  With new GRL implementation there is no need to explicitly
  acquire protection against GRL before creating view.
  This happens automatically in code which acquires metadata
  lock on view to be created.
sql/sql_yacc.yy:
  LEX::protect_against_global_read_lock member is no longer
  necessary since protection against GRL is automatically
  taken by code acquiring metadata locks/opening tables.
sql/table.cc:
  Adjusted code to the fact that one now needs specify duration
  of lock when initializing MDL_request.
sql/table.h:
  Adjusted code to the fact that one now needs specify duration
  of lock when initializing MDL_request.
sql/transaction.cc:
  Replaced custom implementation of global read lock with
  one based on metadata locks. Consequently when doing
  commit instead of calling method of Global_read_lock
  class to acquire protection against GRL we simply acquire
  IX in COMMIT namespace.
  Also adjusted code to the fact that MDL savepoint is now
  represented by MDL_savepoint class.

1 files changed, 1683 insertions, 0 deletions

diff --git a/mysql-test/r/flush_read_lock.result b/mysql-test/r/flush_read_lock.result
new file mode 100644
index 00000000000..2b1071a92b2
--- /dev/null
+++ b/mysql-test/r/flush_read_lock.result
@@ -0,0 +1,1683 @@
+# FTWRL takes two global metadata locks -- a global shared
+# metadata lock and the commit blocker lock.
+# The first lock prevents DDL from taking place.
+# Let's say that all DDL statements that take metadata
+# locks form class #1 -- incompatible with FTWRL because
+# take incompatible MDL table locks.
+# The first global lock doesn't, however, prevent standalone
+# COMMITs (or implicit COMMITs) from taking place, since a
+# COMMIT doesn't take table locks. It doesn't prevent
+# DDL on temporary tables either, since they don't
+# take any table locks either.
+# Most DDL statements do not perform an implicit commit
+# if operate on a temporary table. Examples are CREATE
+# TEMPORARY TABLE and DROP TEMPORARY TABLE. 
+# Thus, these DDL statements can go through in presence
+# of FTWRL. This is class #2 -- compatible because
+# do not take incompatible MDL locks and do not issue
+# implicit commit..
+# (Although these operations do not commit, their effects
+# cannot be rolled back either.)
+# ALTER TABLE, ANALYZE, OPTIMIZE and some others always
+# issue an implicit commit, even if its argument is a
+# temporary table. 
+# *Howewer* an implicit commit is a no-op if all engines
+# used since the start of transactiona are non-
+# transactional. Thus, for non-transactional engines,
+# these operations are not blocked by FTWRL.
+# This is class #3 -- compatible because do not take
+# MDL table locks and are non-transactional. 
+# On the contrary, for transactional engines, there
+# is always a commit, regardless of whether a table
+# is temporary or not. Thus, for example, ALTER TABLE
+# for a transactional engine will wait for FTWRL, 
+# even if the subject table is temporary.
+# Thus ALTER TABLE <temporary> is incompatible
+# with FTWRL. This is class #4 -- incompatible
+# becuase issue implicit COMMIT which is not a no-op.
+# Finally, there are administrative statements (such as 
+# RESET SLAVE) that do not take any locks and do not
+# issue COMMIT.
+# This is class #5.
+# The goal of this coverage is to test statements
+# of all classes.
+# @todo: documents the effects of @@autocommit,
+# DML and temporary transactional tables.
+# Use MyISAM engine for the most of the tables
+# used in this test in order to be able to 
+# check that DDL statements on temporary tables
+# are compatible with FTRWL. 
+drop tables if exists t1_base, t2_base, t3_trans;
+drop tables if exists tm_base, tm_base_temp;
+drop database if exists mysqltest1;
+# We're going to test ALTER DATABASE UPGRADE
+drop database if exists `#mysql50#mysqltest-2`;
+drop procedure if exists p1;
+drop function if exists f1;
+drop view if exists v1;
+drop procedure if exists p2;
+drop function if exists f2_base;
+drop function if exists f2_temp;
+drop event if exists e1;
+drop event if exists e2;
+create table t1_base(i int) engine=myisam;
+create table t2_base(j int) engine=myisam;
+create table t3_trans(i int) engine=innodb;
+create temporary table t1_temp(i int) engine=myisam;
+create temporary table t2_temp(j int) engine=myisam;
+create temporary table t3_temp_trans(i int) engine=innodb;
+create database mysqltest1;
+create database `#mysql50#mysqltest-2`;
+create procedure p1() begin end;
+create function f1() returns int return 0;
+create view v1 as select 1 as i;
+create procedure p2(i int) begin end;
+create function f2_base() returns int
+begin
+insert into t1_base values (1);
+return 0;
+end|
+create function f2_temp() returns int
+begin
+insert into t1_temp values (1);
+return 0;
+end|
+create event e1 on schedule every 1 minute do begin end;
+#
+# Test compatibility of FLUSH TABLES WITH READ LOCK 
+# with various statements.
+#
+# These tests don't cover some classes of statements:
+# - Replication-related - CHANGE MASTER TO, START/STOP SLAVE and etc
+#   (all compatible with FTWRL).
+# - Plugin-related - INSTALL/UNINSTALL (incompatible with FTWRL,
+#   require plugin support).
+#
+# 1) ALTER variants.
+#
+# 1.1) ALTER TABLE
+# 
+# 1.1.a) For base table should be incompatible with FTWRL.
+#
+Success: Was not able to run 'alter table t1_base add column c1 int' under FTWRL.
+Success: 'alter table t1_base add column c1 int' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter table t1_base add column c1 int' is active in another connection.
+#
+# 1.1.b) For a temporary table should be compatible with FTWRL.
+#
+Success: Was able to run 'alter table t1_temp add column c1 int' under FTWRL.
+Success: Was able to run 'alter table t1_temp add column c1 int' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'alter table t1_temp add column c1 int' was active in another connection.
+#
+# 1.2) ALTER DATABASE should be incompatible with FTWRL.
+#
+Success: Was not able to run 'alter database mysqltest1 default character set utf8' under FTWRL.
+Success: 'alter database mysqltest1 default character set utf8' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter database mysqltest1 default character set utf8' is active in another connection.
+#
+# 1.3) ALTER DATABASE UPGRADE DATA DIRECTORY NAME should be
+#      incompatible with FTWRL.
+#
+Success: Was not able to run 'alter database `#mysql50#mysqltest-2` upgrade data directory name' under FTWRL.
+Success: 'alter database `#mysql50#mysqltest-2` upgrade data directory name' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter database `#mysql50#mysqltest-2` upgrade data directory name' is active in another connection.
+#
+# 1.4) ALTER PROCEDURE should be incompatible with FTWRL.
+#
+Success: Was not able to run 'alter procedure p1 comment 'a'' under FTWRL.
+Success: 'alter procedure p1 comment 'a'' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter procedure p1 comment 'a'' is active in another connection.
+#
+# 1.5) ALTER FUNCTION should be incompatible with FTWRL.
+#
+Success: Was not able to run 'alter function f1 comment 'a'' under FTWRL.
+Success: 'alter function f1 comment 'a'' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter function f1 comment 'a'' is active in another connection.
+#
+# 1.6) ALTER VIEW should be incompatible with FTWRL.
+#
+Success: Was not able to run 'alter view v1 as select 2 as j' under FTWRL.
+Success: 'alter view v1 as select 2 as j' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter view v1 as select 2 as j' is active in another connection.
+#
+# 1.7) ALTER EVENT should be incompatible with FTWRL.
+#
+Success: Was not able to run 'alter event e1 comment 'test'' under FTWRL.
+Success: 'alter event e1 comment 'test'' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter event e1 comment 'test'' is active in another connection.
+#
+# 1.x) The rest of ALTER statements (ALTER TABLESPACE,
+#      ALTER LOGFILE GROUP and ALTER SERVER) are too
+#      special to be tested here.
+#
+#
+# 2) ANALYZE TABLE statement is compatible with FTWRL.
+# See Bug#43336 ANALYZE and OPTIMIZE do not honour
+#  --read-only for a discussion why.
+#
+Success: Was able to run 'analyze table t1_base' under FTWRL.
+Success: Was able to run 'analyze table t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'analyze table t1_base' was active in another connection.
+#
+# 3) BEGIN, ROLLBACK and COMMIT statements.
+#    BEGIN and ROLLBACK are compatible with FTWRL.
+#    COMMIT is not.
+# 
+# We need a special test for these statements as
+# FTWRL commits a transaction and because COMMIT
+# is handled in a special way.
+flush tables with read lock;
+begin;
+# ROLLBACK is allowed under FTWRL although there
+# no much sense in it. FTWRL commits any previous
+# changes and doesn't allows any DML after it.
+# So such a ROLLBACK is always a no-op.
+rollback;
+# Although COMMIT is incompatible with FTWRL in
+# other senses it is still allowed under FTWRL.
+# This fact relied upon by some versions of
+# innobackup tool.
+# Similarly to ROLLBACK it is a no-op in this situation.
+commit;
+unlock tables;
+# Check that BEGIN/ROLLBACK are not blocked and 
+# COMMIT is blocked by active FTWRL in another
+# connection.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+begin;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+# Do some work so ROLLBACK is not a no-op.
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+rollback;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+begin;
+# Do some work so COMMIT is not a no-op.
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+# Send:
+commit;
+# Switching to connection 'con1'.
+# Wait until COMMIT is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap COMMIT.
+delete from t3_trans;
+#
+# Check that COMMIT blocks FTWRL in another connection.
+begin;
+insert into t3_trans values (1);
+set debug_sync='RESET';
+set debug_sync='ha_commit_trans_after_acquire_commit_lock SIGNAL parked WAIT_FOR go';
+commit;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+# Switching to connection 'con2'.
+# Wait until FTWRL is blocked.
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap COMMIT.
+# Switching to connection 'con1'.
+# Reap FTWRL.
+unlock tables;
+# Switching to connection 'default'.
+delete from t3_trans;
+set debug_sync= "RESET";
+# We don't run similar test for BEGIN and ROLLBACK as
+# they release metadata locks in non-standard place.
+#
+# 4) BINLOG statement should be incompatible with FTWRL.
+#
+#
+# Provide format description BINLOG statement first.
+BINLOG '
+MfmqTA8BAAAAZwAAAGsAAAABAAQANS41LjctbTMtZGVidWctbG9nAAAAAAAAAAAAAAAAAAAAAAAA
+AAAAAAAAAAAAAAAAAAAx+apMEzgNAAgAEgAEBAQEEgAAVAAEGggAAAAICAgCAA==
+';
+# Now test compatibility for BINLOG statement which is 
+# equivalent to INSERT INTO t1_base VALUES (1).
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was not able to run 'BINLOG '
+MfmqTBMBAAAALgAAAN0AAAAAACgAAAAAAAEABHRlc3QAB3QxX2Jhc2UAAQMAAQ==
+MfmqTBcBAAAAIgAAAP8AAAAAACgAAAAAAAEAAf/+AQAAAA==
+'' under FTWRL.
+Success: 'BINLOG '
+MfmqTBMBAAAALgAAAN0AAAAAACgAAAAAAAEABHRlc3QAB3QxX2Jhc2UAAQMAAQ==
+MfmqTBcBAAAAIgAAAP8AAAAAACgAAAAAAAEAAf/+AQAAAA==
+'' is blocked by FTWRL active in another connection.
+#
+# 5) CALL statement. This statement uses resources in two
+#    ways: through expressions used as parameters and through
+#    sub-statements. This test covers only usage through
+#    parameters as sub-statements do locking individually.
+#
+# 5.a) In simple cases a parameter expression should be
+#      compatible with FTWRL.
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'call p2((select count(*) from t1_base))' under FTWRL.
+Success: Was able to run 'call p2((select count(*) from t1_base))' with FTWRL active in another connection.
+#
+# 5.b) In case when an expression uses function which updates
+#      base tables CALL should be incompatible with FTWRL.
+#
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was not able to run 'call p2(f2_base())' under FTWRL.
+Success: 'call p2(f2_base())' is blocked by FTWRL active in another connection.
+#
+# 5.c) If function used as argument updates temporary tables
+#      CALL statement should be compatible with FTWRL.
+#
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'call p2(f2_temp())' under FTWRL.
+Success: Was able to run 'call p2(f2_temp())' with FTWRL active in another connection.
+#
+# 6) CHECK TABLE statement is compatible with FTWRL.
+#
+Success: Was able to run 'check table t1_base' under FTWRL.
+Success: Was able to run 'check table t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'check table t1_base' was active in another connection.
+#
+# 7) CHECKSUM TABLE statement is compatible with FTWRL.
+#
+Success: Was able to run 'checksum table t1_base' under FTWRL.
+Success: Was able to run 'checksum table t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'checksum table t1_base' was active in another connection.
+#
+# 8) CREATE variants.
+#
+# 8.1) CREATE TABLE statement.
+#
+# 8.1.a) CREATE TABLE is incompatible with FTWRL when
+#        base table is created.
+Success: Was not able to run 'create table t3_base(i int)' under FTWRL.
+Success: 'create table t3_base(i int)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create table t3_base(i int)' is active in another connection.
+# 8.1.b) CREATE TABLE is compatible with FTWRL when
+#        temporary table is created.
+Success: Was able to run 'create temporary table t3_temp(i int)' under FTWRL.
+Success: Was able to run 'create temporary table t3_temp(i int)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'create temporary table t3_temp(i int)' was active in another connection.
+# 8.1.c) CREATE TABLE LIKE is incompatible with FTWRL when
+#        base table is created.
+Success: Was not able to run 'create table t3_base like t1_temp' under FTWRL.
+Success: 'create table t3_base like t1_temp' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create table t3_base like t1_temp' is active in another connection.
+# 8.1.d) CREATE TABLE LIKE is compatible with FTWRL when
+#        temporary table is created.
+Success: Was able to run 'create temporary table t3_temp like t1_base' under FTWRL.
+Success: Was able to run 'create temporary table t3_temp like t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'create temporary table t3_temp like t1_base' was active in another connection.
+# 8.1.e) CREATE TABLE SELECT is incompatible with FTWRL when
+#        base table is created.
+Success: Was not able to run 'create table t3_base select 1 as i' under FTWRL.
+Success: 'create table t3_base select 1 as i' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create table t3_base select 1 as i' is active in another connection.
+# 8.1.f) CREATE TABLE SELECT is compatible with FTWRL when
+#        temporary table is created.
+Success: Was able to run 'create temporary table t3_temp select 1 as i' under FTWRL.
+Success: Was able to run 'create temporary table t3_temp select 1 as i' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'create temporary table t3_temp select 1 as i' was active in another connection.
+# 8.2) CREATE INDEX statement.
+#
+# 8.2.a) CREATE INDEX is incompatible with FTWRL when
+#        applied to base table.
+Success: Was not able to run 'create index i on t1_base (i)' under FTWRL.
+Success: 'create index i on t1_base (i)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create index i on t1_base (i)' is active in another connection.
+# 8.2.b) CREATE INDEX is compatible with FTWRL when
+#        applied to temporary table.
+Success: Was able to run 'create index i on t1_temp (i)' under FTWRL.
+Success: Was able to run 'create index i on t1_temp (i)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'create index i on t1_temp (i)' was active in another connection.
+#
+# 8.3) CREATE DATABASE is incompatible with FTWRL.
+#
+Success: Was not able to run 'create database mysqltest2' under FTWRL.
+Success: 'create database mysqltest2' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create database mysqltest2' is active in another connection.
+#
+# 8.4) CREATE VIEW is incompatible with FTWRL.
+#
+Success: Was not able to run 'create view v2 as select 1 as j' under FTWRL.
+Success: 'create view v2 as select 1 as j' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create view v2 as select 1 as j' is active in another connection.
+#
+# 8.5) CREATE TRIGGER is incompatible with FTWRL.
+#
+Success: Was not able to run 'create trigger t1_bi before insert on t1_base for each row begin end' under FTWRL.
+Success: 'create trigger t1_bi before insert on t1_base for each row begin end' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create trigger t1_bi before insert on t1_base for each row begin end' is active in another connection.
+#
+# 8.6) CREATE FUNCTION is incompatible with FTWRL.
+#
+Success: Was not able to run 'create function f2() returns int return 0' under FTWRL.
+Success: 'create function f2() returns int return 0' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create function f2() returns int return 0' is active in another connection.
+#
+# 8.7) CREATE PROCEDURE is incompatible with FTWRL.
+#
+Success: Was not able to run 'create procedure p3() begin end' under FTWRL.
+Success: 'create procedure p3() begin end' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create procedure p3() begin end' is active in another connection.
+#
+# 8.8) CREATE EVENT should be incompatible with FTWRL.
+#
+Success: Was not able to run 'create event e2 on schedule every 1 minute do begin end' under FTWRL.
+Success: 'create event e2 on schedule every 1 minute do begin end' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create event e2 on schedule every 1 minute do begin end' is active in another connection.
+#
+# 8.9) CREATE USER should be incompatible with FTWRL.
+#
+Success: Was not able to run 'create user mysqltest_u1' under FTWRL.
+Success: 'create user mysqltest_u1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'create user mysqltest_u1' is active in another connection.
+#
+# 8.x) The rest of CREATE variants (CREATE LOGFILE GROUP,
+#      CREATE TABLESPACE and CREATE SERVER) are too special
+#      to test here.
+#
+#
+# 9) PREPARE, EXECUTE and DEALLOCATE PREPARE statements.
+#
+# 9.1) PREPARE statement is compatible with FTWRL as it
+#      doesn't change any data.
+# 
+# 9.1.a) Prepare of simple INSERT statement.
+#
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'prepare stmt1 from 'insert into t1_base values (1)'' under FTWRL.
+Success: Was able to run 'prepare stmt1 from 'insert into t1_base values (1)'' with FTWRL active in another connection.
+#
+# 9.1.b) Prepare of multi-UPDATE. At some point such statements
+#        tried to acquire thr_lock.c locks during prepare phase.
+#        This no longer happens and thus it is compatible with
+#        FTWRL.
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'prepare stmt1 from 'update t1_base, t2_base set t1_base.i= 1 where t1_base.i = t2_base.j'' under FTWRL.
+Success: Was able to run 'prepare stmt1 from 'update t1_base, t2_base set t1_base.i= 1 where t1_base.i = t2_base.j'' with FTWRL active in another connection.
+#
+# 9.1.c) Prepare of multi-DELETE. Again PREPARE of such
+#        statement should be compatible with FTWRL.
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'prepare stmt1 from 'delete t1_base from t1_base, t2_base where t1_base.i = t2_base.j'' under FTWRL.
+Success: Was able to run 'prepare stmt1 from 'delete t1_base from t1_base, t2_base where t1_base.i = t2_base.j'' with FTWRL active in another connection.
+#
+# 9.2) Compatibility of EXECUTE statement depends on statement
+#     to be executed.
+#
+# 9.2.a) EXECUTE for statement which is itself compatible with
+#       FTWRL should be compatible.
+prepare stmt1 from 'select * from t1_base';
+Success: Was able to run 'execute stmt1' under FTWRL.
+Success: Was able to run 'execute stmt1' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'execute stmt1' was active in another connection.
+deallocate prepare stmt1;
+#
+# 9.2.b) EXECUTE for statement which is incompatible with FTWRL
+#        should be also incompatible.
+#
+# Check that EXECUTE is not allowed under FTWRL.
+prepare stmt1 from 'insert into t1_base values (1)';
+flush tables with read lock;
+execute stmt1;
+ERROR HY000: Can't execute the query because you have a conflicting read lock
+unlock tables;
+# Check that active FTWRL in another connection
+# blocks EXECUTE which changes data.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+execute stmt1 ;
+# Switching to connection 'con1'.
+# Check that EXECUTE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap EXECUTE.
+set debug_sync='RESET';
+set debug_sync='execute_command_after_close_tables SIGNAL parked WAIT_FOR go';
+execute stmt1; ;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+# Switching to connection 'con2'.
+# Wait until FTWRL is blocked.
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap EXECUTE.
+# Switching to connection 'con1'.
+# Reap FTWRL.
+unlock tables;
+# Switching to connection 'default'.
+set debug_sync= "RESET";
+delete from t1_base;
+deallocate prepare stmt1;
+#
+# 9.3) DEALLOCATE PREPARE is compatible with FTWRL.
+#
+prepare stmt1 from 'insert into t1_base values (1)';
+Success: Was able to run 'deallocate prepare stmt1' under FTWRL.
+Success: Was able to run 'deallocate prepare stmt1' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'deallocate prepare stmt1' was active in another connection.
+deallocate prepare stmt1;
+#
+# 10) DELETE variations.
+#
+# 10.1) Simple DELETE.
+# 
+# 10.1.a) Simple DELETE on base table is incompatible with FTWRL.
+Success: Was not able to run 'delete from t1_base' under FTWRL.
+Success: 'delete from t1_base' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'delete from t1_base' is active in another connection.
+# 
+# 10.1.b) Simple DELETE on temporary table is compatible with FTWRL.
+Success: Was able to run 'delete from t1_temp' under FTWRL.
+Success: Was able to run 'delete from t1_temp' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'delete from t1_temp' was active in another connection.
+#
+# 10.2) Multi DELETE.
+# 
+# 10.2.a) Multi DELETE on base tables is incompatible with FTWRL.
+Success: Was not able to run 'delete t1_base from t1_base, t2_base where t1_base.i = t2_base.j' under FTWRL.
+Success: 'delete t1_base from t1_base, t2_base where t1_base.i = t2_base.j' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'delete t1_base from t1_base, t2_base where t1_base.i = t2_base.j' is active in another connection.
+# 
+# 10.2.b) Multi DELETE on temporary tables is compatible with FTWRL.
+Success: Was able to run 'delete t1_temp from t1_temp, t2_temp where t1_temp.i = t2_temp.j' under FTWRL.
+Success: Was able to run 'delete t1_temp from t1_temp, t2_temp where t1_temp.i = t2_temp.j' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'delete t1_temp from t1_temp, t2_temp where t1_temp.i = t2_temp.j' was active in another connection.
+#
+# 11) DESCRIBE should be compatible with FTWRL.
+#
+Success: Was able to run 'describe t1_base' under FTWRL.
+Success: Was able to run 'describe t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'describe t1_base' was active in another connection.
+#
+# 12) Compatibility of DO statement with FTWRL depends on its
+#     expression.
+#
+# 12.a) DO with expression which does not change base table
+#       should be compatible with FTWRL.
+Success: Was able to run 'do (select count(*) from t1_base)' under FTWRL.
+Success: Was able to run 'do (select count(*) from t1_base)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'do (select count(*) from t1_base)' was active in another connection.
+#
+# 12.b) DO which calls SF updating base table should be
+#       incompatible with FTWRL.
+Success: Was not able to run 'do f2_base()' under FTWRL.
+Success: 'do f2_base()' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'do f2_base()' is active in another connection.
+#
+# 12.c) DO which calls SF updating temporary table should be
+#       compatible with FTWRL.
+Success: Was able to run 'do f2_temp()' under FTWRL.
+Success: Was able to run 'do f2_temp()' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'do f2_temp()' was active in another connection.
+#
+# 13) DROP variants.
+#
+# 13.1) DROP TABLES.
+#
+# 13.1.a) DROP TABLES which affects base tables is incompatible
+#         with FTWRL.
+Success: Was not able to run 'drop table t2_base' under FTWRL.
+Success: 'drop table t2_base' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop table t2_base' is active in another connection.
+# 13.1.b) DROP TABLES which affects only temporary tables
+#         in theory can be compatible with FTWRL.
+#         In practice it is not yet.
+Success: Was not able to run 'drop table t2_temp' under FTWRL.
+Success: 'drop table t2_temp' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop table t2_temp' is active in another connection.
+#
+# 13.1.c) DROP TEMPORARY TABLES should be compatible with FTWRL.
+Success: Was able to run 'drop temporary table t2_temp' under FTWRL.
+Success: Was able to run 'drop temporary table t2_temp' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'drop temporary table t2_temp' was active in another connection.
+#
+# 13.2) DROP INDEX.
+#
+# 13.2.a) DROP INDEX on a base table is incompatible with FTWRL.
+create index i on t1_base (i);
+Success: Was not able to run 'drop index i on t1_base' under FTWRL.
+Success: 'drop index i on t1_base' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop index i on t1_base' is active in another connection.
+drop index i on t1_base;
+#
+# 13.2.b) DROP INDEX on a temporary table is compatible with FTWRL.
+create index i on t1_temp (i);
+Success: Was able to run 'drop index i on t1_temp' under FTWRL.
+Success: Was able to run 'drop index i on t1_temp' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'drop index i on t1_temp' was active in another connection.
+drop index i on t1_temp;
+#
+# 13.3) DROP DATABASE is incompatible with FTWRL
+#
+Success: Was not able to run 'drop database mysqltest1' under FTWRL.
+Success: 'drop database mysqltest1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop database mysqltest1' is active in another connection.
+#
+# 13.4) DROP FUNCTION is incompatible with FTWRL.
+#
+Success: Was not able to run 'drop function f1' under FTWRL.
+Success: 'drop function f1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop function f1' is active in another connection.
+#
+# 13.5) DROP PROCEDURE is incompatible with FTWRL.
+#
+Success: Was not able to run 'drop procedure p1' under FTWRL.
+Success: 'drop procedure p1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop procedure p1' is active in another connection.
+#
+# 13.6) DROP USER should be incompatible with FTWRL.
+#
+create user mysqltest_u1;
+Success: Was not able to run 'drop user mysqltest_u1' under FTWRL.
+Success: 'drop user mysqltest_u1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop user mysqltest_u1' is active in another connection.
+drop user mysqltest_u1;
+#
+# 13.7) DROP VIEW should be incompatible with FTWRL.
+#
+Success: Was not able to run 'drop view v1' under FTWRL.
+Success: 'drop view v1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop view v1' is active in another connection.
+#
+# 13.8) DROP EVENT should be incompatible with FTWRL.
+#
+Success: Was not able to run 'drop event e1' under FTWRL.
+Success: 'drop event e1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop event e1' is active in another connection.
+#
+# 13.9) DROP TRIGGER is incompatible with FTWRL.
+#
+create trigger t1_bi before insert on t1_base for each row begin end;
+Success: Was not able to run 'drop trigger t1_bi' under FTWRL.
+Success: 'drop trigger t1_bi' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'drop trigger t1_bi' is active in another connection.
+drop trigger t1_bi;
+#
+# 13.x) The rest of DROP variants (DROP TABLESPACE, DROP LOGFILE
+#       GROUP and DROP SERVER) are too special to test here.
+#
+#
+# 14) FLUSH variants.
+#
+# Test compatibility of _some_ important FLUSH variants with FTWRL.
+#
+# 14.1) FLUSH TABLES WITH READ LOCK is compatible with itself.
+# 
+# Check that FTWRL statements can be run while FTWRL
+# is active in another connection.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# The second FTWRL in a row is allowed at the moment.
+# It does not make much sense as it does only flush.
+flush tables with read lock;
+unlock tables;
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+flush tables with read lock;
+unlock tables;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+#
+# 14.2) FLUSH TABLES <list> WITH READ LOCK is not blocked by
+#       active FTWRL. But since the latter keeps tables open
+#       FTWRL is blocked by FLUSH TABLES <list> WITH READ LOCK.
+flush tables with read lock;
+# FT <list> WRL is allowed under FTWRL at the moment.
+# It does not make much sense though.
+flush tables t1_base, t2_base with read lock;
+unlock tables;
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+flush tables t1_base, t2_base with read lock;
+unlock tables;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+flush tables t1_base, t2_base with read lock;
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'con2'.
+# Wait until FTWRL is blocked.
+# Switching to connection 'default'.
+unlock tables;
+# Switching to connection 'con1'.
+# Reap FTWRL.
+unlock tables;
+# Switching to connection 'default'.
+#
+# 14.3) FLUSH TABLES is compatible with FTWRL.
+Success: Was able to run 'flush tables' under FTWRL.
+Success: Was able to run 'flush tables' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'flush tables' was active in another connection.
+#
+# 14.4) FLUSH TABLES <list> is compatible with FTWRL.
+Success: Was able to run 'flush table t1_base, t2_base' under FTWRL.
+Success: Was able to run 'flush table t1_base, t2_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'flush table t1_base, t2_base' was active in another connection.
+#
+# 14.5) FLUSH PRIVILEGES is compatible with FTWRL.
+Success: Was able to run 'flush privileges' under FTWRL.
+Success: Was able to run 'flush privileges' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'flush privileges' was active in another connection.
+#
+# 15) GRANT statement should be incompatible with FTWRL.
+#
+Success: Was not able to run 'grant all privileges on t1_base to mysqltest_u1' under FTWRL.
+Success: 'grant all privileges on t1_base to mysqltest_u1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'grant all privileges on t1_base to mysqltest_u1' is active in another connection.
+drop user mysqltest_u1;
+#
+# 16) All HANDLER variants are half-compatible with FTWRL.
+#     I.e. they are not blocked by active FTWRL. But since open
+#     HANDLER means open table instance FTWRL is blocked while
+#     HANDLER is not closed.
+#
+# Check that HANDLER statements succeed under FTWRL.
+flush tables with read lock;
+handler t1_base open;
+handler t1_base read first;
+i
+handler t1_base close;
+unlock tables;
+# Check that HANDLER statements can be run while FTWRL
+# is active in another connection.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+handler t1_base open;
+handler t1_base read first;
+i
+handler t1_base close;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+#
+# 17) HELP statement is compatible with FTWRL.
+#
+Success: Was able to run 'help no_such_topic' under FTWRL.
+Success: Was able to run 'help no_such_topic' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'help no_such_topic' was active in another connection.
+#
+# 18) INSERT statement.
+# 
+# 18.a) Ordinary INSERT into base table is incompatible with FTWRL.
+Success: Was not able to run 'insert into t1_base values (1)' under FTWRL.
+Success: 'insert into t1_base values (1)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'insert into t1_base values (1)' is active in another connection.
+#
+# 18.b) Ordinary INSERT into temp table is compatible with FTWRL.
+Success: Was able to run 'insert into t1_temp values (1)' under FTWRL.
+Success: Was able to run 'insert into t1_temp values (1)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'insert into t1_temp values (1)' was active in another connection.
+#
+# 18.c) INSERT DELAYED is incompatible with FTWRL.
+Success: Was not able to run 'insert delayed into t1_base values (1)' under FTWRL.
+Success: 'insert delayed into t1_base values (1)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'insert delayed into t1_base values (1)' is active in another connection.
+delete from t1_base;
+# 
+# 18.d) INSERT SELECT into base table is incompatible with FTWRL.
+Success: Was not able to run 'insert into t1_base select * from t1_temp' under FTWRL.
+Success: 'insert into t1_base select * from t1_temp' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'insert into t1_base select * from t1_temp' is active in another connection.
+#
+# 18.e) INSERT SELECT into temp table is compatible with FTWRL.
+Success: Was able to run 'insert into t1_temp select * from t1_base' under FTWRL.
+Success: Was able to run 'insert into t1_temp select * from t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'insert into t1_temp select * from t1_base' was active in another connection.
+#
+# 19) KILL statement is compatible with FTWRL.
+#
+# Check that KILL can be run under FTWRL.
+flush tables with read lock;
+set @id:= connection_id();
+kill query @id;
+ERROR 70100: Query execution was interrupted
+unlock tables;
+# Check that KILL statements can be run while FTWRL
+# is active in another connection.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+kill query @id;
+ERROR 70100: Query execution was interrupted
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+# Finally check that KILL doesn't block FTWRL
+set debug_sync='RESET';
+set debug_sync='execute_command_after_close_tables SIGNAL parked WAIT_FOR go';
+kill query @id;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+unlock tables;
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap KILL. 
+ERROR 70100: Query execution was interrupted
+set debug_sync='RESET';
+#
+# 20) LOAD DATA statement.
+# 
+# 20.a) LOAD DATA into base table is incompatible with FTWRL.
+Success: Was not able to run 'load data infile '../../std_data/rpl_loaddata.dat' into table t1_base (@dummy, i)' under FTWRL.
+Success: 'load data infile '../../std_data/rpl_loaddata.dat' into table t1_base (@dummy, i)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'load data infile '../../std_data/rpl_loaddata.dat' into table t1_base (@dummy, i)' is active in another connection.
+# 
+# 20.b) LOAD DATA into temporary table is compatible with FTWRL.
+Success: Was able to run 'load data infile '../../std_data/rpl_loaddata.dat' into table t1_temp (@dummy, i)' under FTWRL.
+Success: Was able to run 'load data infile '../../std_data/rpl_loaddata.dat' into table t1_temp (@dummy, i)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'load data infile '../../std_data/rpl_loaddata.dat' into table t1_temp (@dummy, i)' was active in another connection.
+#
+# 21) LOCK/UNLOCK TABLES statements.
+#
+# LOCK TABLES statement always (almost) blocks FTWRL as it
+# keeps tables open until UNLOCK TABLES.
+# Active FTWRL on the other hand blocks only those
+# LOCK TABLES which allow updating of base tables.
+#
+# 21.a) LOCK TABLES READ is allowed under FTWRL and
+#       is not blocked by active FTWRL.
+flush tables with read lock;
+lock tables t1_base read;
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+lock tables t1_base read;
+unlock tables;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+#
+# 21.b) LOCK TABLES WRITE on a base table is disallowed
+#       under FTWRL and should be blocked by active FTWRL.
+flush tables with read lock;
+lock tables t1_base write;
+ERROR HY000: Can't execute the query because you have a conflicting read lock
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+lock tables t1_base write ;
+# Switching to connection 'con1'.
+# Check that LOCK TABLES WRITE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap LOCK TABLES WRITE
+unlock tables;
+#
+# 21.c) LOCK TABLES WRITE on temporary table doesn't
+#       make much sense but is allowed under FTWRL
+#       and should not be blocked by active FTWRL.
+flush tables with read lock;
+lock tables t1_temp write;
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+lock tables t1_temp write;
+unlock tables;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+#
+# 22) OPTIMIZE TABLE statement.
+#
+# 22.a) OPTIMIZE TABLE of base table is incompatible with FTWRL.
+flush tables with read lock;
+# OPTIMIZE statement returns errors as part of result-set.
+optimize table t1_base;
+Table	Op	Msg_type	Msg_text
+test.t1_base	optimize	Error	Can't execute the query because you have a conflicting read lock
+test.t1_base	optimize	error	Corrupt
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+optimize table t1_base;
+# Switching to connection 'con1'.
+# Check that OPTIMIZE TABLE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap OPTIMIZE TABLE
+Table	Op	Msg_type	Msg_text
+test.t1_base	optimize	status	OK
+# We don't check that active OPTIMIZE TABLE blocks
+# FTWRL as this one of statements releasing metadata
+# locks in non-standard place.
+#
+# 22.b) OPTIMIZE TABLE of temporary table is compatible with FTWRL.
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'optimize table t1_temp' under FTWRL.
+Success: Was able to run 'optimize table t1_temp' with FTWRL active in another connection.
+#
+# 23) CACHE statement is compatible with FTWRL.
+#
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'cache index t1_base in default' under FTWRL.
+Success: Was able to run 'cache index t1_base in default' with FTWRL active in another connection.
+#
+# 24) LOAD INDEX statement is compatible with FTWRL.
+#
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'load index into cache t1_base' under FTWRL.
+Success: Was able to run 'load index into cache t1_base' with FTWRL active in another connection.
+#
+# 25) SAVEPOINT/RELEASE SAVEPOINT/ROLLBACK TO SAVEPOINT are
+#     compatible with FTWRL.
+#
+# Since manipulations on savepoint have to be done
+# inside transaction and FTWRL commits transaction we
+# need a special test for these statements.
+flush tables with read lock;
+begin;
+savepoint sv1;
+rollback to savepoint sv1;
+release savepoint sv1;
+unlock tables;
+commit;
+# Check that these statements are not blocked by
+# active FTWRL in another connection.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+begin;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+# Do some changes to avoid SAVEPOINT and friends
+# being almost no-ops.
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+savepoint sv1;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+insert into t3_trans values (2);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+rollback to savepoint sv1;
+release savepoint sv1;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+rollback;
+# Check that these statements don't block FTWRL in
+# another connection.
+begin;
+# Do some changes to avoid SAVEPOINT and friends
+# being almost no-ops.
+insert into t3_trans values (1);
+set debug_sync='RESET';
+set debug_sync='execute_command_after_close_tables SIGNAL parked WAIT_FOR go';
+savepoint sv1;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+unlock tables;
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap SAVEPOINT
+insert into t3_trans values (2);
+set debug_sync='execute_command_after_close_tables SIGNAL parked WAIT_FOR go';
+rollback to savepoint sv1;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+unlock tables;
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap ROLLBACK TO SAVEPOINT
+set debug_sync='execute_command_after_close_tables SIGNAL parked WAIT_FOR go';
+release savepoint sv1;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+unlock tables;
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap RELEASE SAVEPOINT
+rollback;
+set debug_sync= "RESET";
+#
+# 26) RENAME variants.
+#
+# 26.1) RENAME TABLES is incompatible with FTWRL.
+Success: Was not able to run 'rename table t1_base to t3_base' under FTWRL.
+Success: 'rename table t1_base to t3_base' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'rename table t1_base to t3_base' is active in another connection.
+#
+# 26.2) RENAME USER is incompatible with FTWRL.
+create user mysqltest_u1;
+Success: Was not able to run 'rename user mysqltest_u1 to mysqltest_u2' under FTWRL.
+Success: 'rename user mysqltest_u1 to mysqltest_u2' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'rename user mysqltest_u1 to mysqltest_u2' is active in another connection.
+drop user mysqltest_u1;
+#
+# 27) REPAIR TABLE statement.
+#
+# 27.a) REPAIR TABLE of base table is incompatible with FTWRL.
+flush tables with read lock;
+# REPAIR statement returns errors as part of result-set.
+repair table t1_base;
+Table	Op	Msg_type	Msg_text
+test.t1_base	repair	Error	Can't execute the query because you have a conflicting read lock
+test.t1_base	repair	error	Corrupt
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+repair table t1_base;
+# Switching to connection 'con1'.
+# Check that REPAIR TABLE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap REPAIR TABLE
+Table	Op	Msg_type	Msg_text
+test.t1_base	repair	status	OK
+# We don't check that active REPAIR TABLE blocks
+# FTWRL as this one of statements releasing metadata
+# locks in non-standard place.
+#
+# 27.b) REPAIR TABLE of temporary table is compatible with FTWRL.
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'repair table t1_temp' under FTWRL.
+Success: Was able to run 'repair table t1_temp' with FTWRL active in another connection.
+#
+# 28) REPLACE statement.
+# 
+# 28.a) Ordinary REPLACE into base table is incompatible with FTWRL.
+Success: Was not able to run 'replace into t1_base values (1)' under FTWRL.
+Success: 'replace into t1_base values (1)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'replace into t1_base values (1)' is active in another connection.
+#
+# 28.b) Ordinary REPLACE into temp table is compatible with FTWRL.
+Success: Was able to run 'replace into t1_temp values (1)' under FTWRL.
+Success: Was able to run 'replace into t1_temp values (1)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'replace into t1_temp values (1)' was active in another connection.
+# 
+# 28.c) REPLACE SELECT into base table is incompatible with FTWRL.
+Success: Was not able to run 'replace into t1_base select * from t1_temp' under FTWRL.
+Success: 'replace into t1_base select * from t1_temp' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'replace into t1_base select * from t1_temp' is active in another connection.
+#
+# 28.d) REPLACE SELECT into temp table is compatible with FTWRL.
+Success: Was able to run 'replace into t1_temp select * from t1_base' under FTWRL.
+Success: Was able to run 'replace into t1_temp select * from t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'replace into t1_temp select * from t1_base' was active in another connection.
+#
+# 29) REVOKE variants.
+# 
+# 29.1) REVOKE privileges is incompatible with FTWRL. 
+grant all privileges on t1_base to mysqltest_u1;
+Success: Was not able to run 'revoke all privileges on t1_base from mysqltest_u1' under FTWRL.
+Success: 'revoke all privileges on t1_base from mysqltest_u1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'revoke all privileges on t1_base from mysqltest_u1' is active in another connection.
+# 
+# 29.2) REVOKE ALL PRIVILEGES, GRANT OPTION is incompatible with FTWRL.
+Success: Was not able to run 'revoke all privileges, grant option from mysqltest_u1' under FTWRL.
+Success: 'revoke all privileges, grant option from mysqltest_u1' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'revoke all privileges, grant option from mysqltest_u1' is active in another connection.
+drop user mysqltest_u1;
+#
+# 30) Compatibility of SELECT statement with FTWRL depends on
+#     locking mode used and on functions being invoked by it.
+#
+# 30.a) Simple SELECT which does not change tables should be
+#       compatible with FTWRL.
+Success: Was able to run 'select count(*) from t1_base' under FTWRL.
+Success: Was able to run 'select count(*) from t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'select count(*) from t1_base' was active in another connection.
+# 30.b) SELECT ... FOR UPDATE is incompatible with FTWRL.
+Success: Was not able to run 'select count(*) from t1_base for update' under FTWRL.
+Success: 'select count(*) from t1_base for update' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'select count(*) from t1_base for update' is active in another connection.
+# 30.c) SELECT ... LOCK IN SHARE MODE is compatible with FTWRL.
+Success: Was able to run 'select count(*) from t1_base lock in share mode' under FTWRL.
+Success: Was able to run 'select count(*) from t1_base lock in share mode' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'select count(*) from t1_base lock in share mode' was active in another connection.
+#
+# 30.d) SELECT which calls SF updating base table should be
+#       incompatible with FTWRL.
+Success: Was not able to run 'select f2_base()' under FTWRL.
+Success: 'select f2_base()' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'select f2_base()' is active in another connection.
+#
+# 30.e) SELECT which calls SF updating temporary table should be
+#       compatible with FTWRL.
+Success: Was able to run 'select f2_temp()' under FTWRL.
+Success: Was able to run 'select f2_temp()' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'select f2_temp()' was active in another connection.
+#
+# 31) Compatibility of SET statement with FTWRL depends on its
+#     expression and on whether it is a special SET statement.
+#
+# 31.a) Ordinary SET with expression which does not
+#       changes base table should be compatible with FTWRL.
+# Skip last part of compatibility testing as our helper debug
+# sync-point doesn't work for SET statements.
+Success: Was able to run 'set @a:= (select count(*) from t1_base)' under FTWRL.
+Success: Was able to run 'set @a:= (select count(*) from t1_base)' with FTWRL active in another connection.
+#
+# 31.b) Ordinary SET which calls SF updating base table should
+#       be incompatible with FTWRL.
+# Skip last part of compatibility testing as our helper debug
+# sync-point doesn't work for SET statements.
+Success: Was not able to run 'set @a:= f2_base()' under FTWRL.
+Success: 'set @a:= f2_base()' is blocked by FTWRL active in another connection.
+#
+# 31.c) Ordinary SET which calls SF updating temporary table
+#       should be compatible with FTWRL.
+# Skip last part of compatibility testing as our helper debug
+# sync-point doesn't work for SET statements.
+Success: Was able to run 'set @a:= f2_temp()' under FTWRL.
+Success: Was able to run 'set @a:= f2_temp()' with FTWRL active in another connection.
+#
+# 31.d) Special SET variants have different compatibility with FTWRL.
+#
+# 31.d.I) SET PASSWORD is incompatible with FTWRL as it changes data.
+create user mysqltest_u1;
+# Skip last part of compatibility testing as our helper debug
+# sync-point doesn't work for SET statements.
+Success: Was not able to run 'set password for 'mysqltest_u1' = password('')' under FTWRL.
+Success: 'set password for 'mysqltest_u1' = password('')' is blocked by FTWRL active in another connection.
+drop user mysqltest_u1;
+#
+# 31.d.II) SET READ_ONLY is compatible with FTWRL (but has no
+#          effect when executed under it).
+# Skip last part of compatibility testing as our helper debug
+# sync-point doesn't work for SET statements.
+Success: Was able to run 'set global read_only= 1' under FTWRL.
+Success: Was able to run 'set global read_only= 1' with FTWRL active in another connection.
+#
+# 31.d.III) Situation with SET AUTOCOMMIT is complex.
+#           Turning auto-commit off is always compatible with FTWRL.
+#           Turning auto-commit on causes implicit commit and so
+#           is incompatible with FTWRL if there are changes to be
+#           committed.
+flush tables with read lock;
+set autocommit= 0;
+# Turning auto-commit on causes implicit commit so can
+# be incompatible with FTWRL if there is something to
+# commit. But since even in this case we allow commits
+# under active FTWRL such statement should always succeed.
+insert into t3_temp_trans values (1);
+set autocommit= 1;
+unlock tables;
+delete from t3_temp_trans;
+# Check that SET AUTOCOMMIT=0 is not blocked and 
+# SET AUTOCOMMIT=1 is blocked by active FTWRL in
+# another connection.
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+set autocommit= 0;
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+# Do some work so implicit commit in SET AUTOCOMMIT=1 
+# is not a no-op.
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+# Send:
+set autocommit= 1;
+# Switching to connection 'con1'.
+# Wait until SET AUTOCOMMIT=1 is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap SET AUTOCOMMIT=1.
+delete from t3_trans;
+#
+# Check that SET AUTOCOMMIT=1 blocks FTWRL in another connection.
+set autocommit= 0;
+insert into t3_trans values (1);
+set debug_sync='RESET';
+set debug_sync='ha_commit_trans_after_acquire_commit_lock SIGNAL parked WAIT_FOR go';
+set autocommit= 1;
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+# Switching to connection 'con2'.
+# Wait until FTWRL is blocked.
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap SET AUTOCOMMIT=1.
+# Switching to connection 'con1'.
+# Reap FTWRL.
+unlock tables;
+# Switching to connection 'default'.
+delete from t3_trans;
+set debug_sync= "RESET";
+#
+# 32) SHOW statements are compatible with FTWRL.
+#     Let us test _some_ of them.
+#
+# 32.1) SHOW TABLES.
+Success: Was able to run 'show tables from test' under FTWRL.
+Success: Was able to run 'show tables from test' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'show tables from test' was active in another connection.
+#
+# 32.1) SHOW TABLES.
+Success: Was able to run 'show tables from test' under FTWRL.
+Success: Was able to run 'show tables from test' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'show tables from test' was active in another connection.
+#
+# 32.2) SHOW EVENTS.
+Success: Was able to run 'show events from test' under FTWRL.
+Success: Was able to run 'show events from test' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'show events from test' was active in another connection.
+#
+# 32.3) SHOW GRANTS.
+create user mysqltest_u1;
+Success: Was able to run 'show grants for mysqltest_u1' under FTWRL.
+Success: Was able to run 'show grants for mysqltest_u1' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'show grants for mysqltest_u1' was active in another connection.
+drop user mysqltest_u1;
+#
+# 32.4) SHOW CREATE TABLE.
+Success: Was able to run 'show create table t1_base' under FTWRL.
+Success: Was able to run 'show create table t1_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'show create table t1_base' was active in another connection.
+#
+# 32.5) SHOW CREATE FUNCTION.
+Success: Was able to run 'show create function f1' under FTWRL.
+Success: Was able to run 'show create function f1' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'show create function f1' was active in another connection.
+#
+# 33) SIGNAL statement is compatible with FTWRL.
+#
+# Note that we don't cover RESIGNAL as it requires
+# active handler context. 
+Success: Was able to run 'signal sqlstate '01000'' under FTWRL.
+Success: Was able to run 'signal sqlstate '01000'' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'signal sqlstate '01000'' was active in another connection.
+#
+# 34) TRUNCATE TABLE statement.
+# 
+# 34.a) TRUNCATE of base table is incompatible with FTWRL.
+Success: Was not able to run 'truncate table t1_base' under FTWRL.
+Success: 'truncate table t1_base' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'truncate table t1_base' is active in another connection.
+# 
+# 34.b) TRUNCATE of temporary table is compatible with FTWRL.
+Success: Was able to run 'truncate table t1_temp' under FTWRL.
+Success: Was able to run 'truncate table t1_temp' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'truncate table t1_temp' was active in another connection.
+#
+# 35) UPDATE variants.
+#
+# 35.1) Simple UPDATE.
+# 
+# 35.1.a) Simple UPDATE on base table is incompatible with FTWRL.
+Success: Was not able to run 'update t1_base set i= 1 where i = 0' under FTWRL.
+Success: 'update t1_base set i= 1 where i = 0' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'update t1_base set i= 1 where i = 0' is active in another connection.
+# 
+# 35.1.b) Simple UPDATE on temporary table is compatible with FTWRL.
+Success: Was able to run 'update t1_temp set i= 1 where i = 0' under FTWRL.
+Success: Was able to run 'update t1_temp set i= 1 where i = 0' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'update t1_temp set i= 1 where i = 0' was active in another connection.
+#
+# 35.2) Multi UPDATE.
+# 
+# 35.2.a) Multi UPDATE on base tables is incompatible with FTWRL.
+Success: Was not able to run 'update t1_base, t2_base set t1_base.i= 1 where t1_base.i = t2_base.j' under FTWRL.
+Success: 'update t1_base, t2_base set t1_base.i= 1 where t1_base.i = t2_base.j' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'update t1_base, t2_base set t1_base.i= 1 where t1_base.i = t2_base.j' is active in another connection.
+# 
+# 35.2.b) Multi UPDATE on temporary tables is compatible with FTWRL.
+Success: Was able to run 'update t1_temp, t2_temp set t1_temp.i= 1 where t1_temp.i = t2_temp.j' under FTWRL.
+Success: Was able to run 'update t1_temp, t2_temp set t1_temp.i= 1 where t1_temp.i = t2_temp.j' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'update t1_temp, t2_temp set t1_temp.i= 1 where t1_temp.i = t2_temp.j' was active in another connection.
+#
+# 36) USE statement is compatible with FTWRL.
+#
+Success: Was able to run 'use mysqltest1' under FTWRL.
+Success: Was able to run 'use mysqltest1' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'use mysqltest1' was active in another connection.
+#
+# 37) XA statements.
+#     
+# XA statements are similar to BEGIN/COMMIT/ROLLBACK.
+#
+# XA BEGIN, END, PREPARE, ROLLBACK and RECOVER are compatible
+# with FTWRL. XA COMMIT is not.
+flush tables with read lock;
+# Although all below statements are allowed under FTWRL they
+# are almost no-ops as FTWRL does commit and does not allows
+# any non-temporary DML under it.
+xa start 'test1';
+xa end 'test1';
+xa prepare 'test1';
+xa rollback 'test1';
+xa start 'test1';
+xa end 'test1';
+xa prepare 'test1';
+xa commit 'test1';
+xa recover;
+unlock tables;
+# Check that XA non-COMMIT statements are not and COMMIT is
+# blocked by active FTWRL in another connection
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+xa start 'test1';
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+xa end 'test1';
+xa prepare 'test1';
+xa rollback 'test1';
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+xa start 'test1';
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+xa end 'test1';
+xa prepare 'test1';
+# Send:
+xa commit 'test1';;
+# Switching to connection 'con1'.
+# Wait until XA COMMIT is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap XA COMMIT.
+delete from t3_trans;
+#
+# Check that XA COMMIT blocks FTWRL in another connection.
+xa start 'test1';
+insert into t3_trans values (1);
+xa end 'test1';
+xa prepare 'test1';
+set debug_sync='RESET';
+set debug_sync='trans_xa_commit_after_acquire_commit_lock SIGNAL parked WAIT_FOR go';
+xa commit 'test1';
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+# Switching to connection 'con2'.
+# Wait until FTWRL is blocked.
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap XA COMMIT.
+# Switching to connection 'con1'.
+# Reap FTWRL.
+unlock tables;
+# Switching to connection 'default'.
+delete from t3_trans;
+set debug_sync= "RESET";
+#
+# 38) Test effect of auto-commit mode for DML on transactional
+#     temporary tables.
+# 
+# 38.1) When auto-commit is on each such a statement ends with commit
+#       of changes to temporary tables. But since transactions doing
+#       such changes are considered read only [sic!/QQ] this commit
+#       is compatible with FTWRL.
+#
+#       Let us demostrate this fact for some common DML statements.
+Success: Was able to run 'delete from t3_temp_trans' under FTWRL.
+Success: Was able to run 'delete from t3_temp_trans' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'delete from t3_temp_trans' was active in another connection.
+Success: Was able to run 'insert into t3_temp_trans values (1)' under FTWRL.
+Success: Was able to run 'insert into t3_temp_trans values (1)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'insert into t3_temp_trans values (1)' was active in another connection.
+Success: Was able to run 'update t3_temp_trans, t2_temp set t3_temp_trans.i= 1 where t3_temp_trans.i = t2_temp.j' under FTWRL.
+Success: Was able to run 'update t3_temp_trans, t2_temp set t3_temp_trans.i= 1 where t3_temp_trans.i = t2_temp.j' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'update t3_temp_trans, t2_temp set t3_temp_trans.i= 1 where t3_temp_trans.i = t2_temp.j' was active in another connection.
+# 
+# 38.2) When auto-commit is off DML on transaction temporary tables
+#       is compatible with FTWRL.
+#
+set autocommit= 0;
+Success: Was able to run 'delete from t3_temp_trans' under FTWRL.
+Success: Was able to run 'delete from t3_temp_trans' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'delete from t3_temp_trans' was active in another connection.
+Success: Was able to run 'insert into t3_temp_trans values (1)' under FTWRL.
+Success: Was able to run 'insert into t3_temp_trans values (1)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'insert into t3_temp_trans values (1)' was active in another connection.
+Success: Was able to run 'update t3_temp_trans, t2_temp set t3_temp_trans.i= 1 where t3_temp_trans.i = t2_temp.j' under FTWRL.
+Success: Was able to run 'update t3_temp_trans, t2_temp set t3_temp_trans.i= 1 where t3_temp_trans.i = t2_temp.j' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'update t3_temp_trans, t2_temp set t3_temp_trans.i= 1 where t3_temp_trans.i = t2_temp.j' was active in another connection.
+set autocommit= 1;
+#
+# 39) Test effect of DDL on transactional tables.
+#
+# 39.1) Due to implicit commit at the end of statement some of DDL
+#       statements which are compatible with FTWRL in non-transactional
+#       case are not compatible in case of transactional tables.
+#
+# 39.1.a) ANALYZE TABLE for transactional table is incompatible with
+#         FTWRL.
+flush tables with read lock;
+# Implicit commits are allowed under FTWRL.
+analyze table t3_trans;
+Table	Op	Msg_type	Msg_text
+test.t3_trans	analyze	status	OK
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+analyze table t3_trans;
+# Switching to connection 'con1'.
+# Check that ANALYZE TABLE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap ANALYZE TABLE
+Table	Op	Msg_type	Msg_text
+test.t3_trans	analyze	status	OK
+#
+# 39.1.b) CHECK TABLE for transactional table is compatible with FTWRL.
+#         Although it does implicit commit at the end of statement it
+#         is considered to be read-only operation.
+# Skip last part of compatibility testing as this statement
+# releases metadata locks in non-standard place.
+Success: Was able to run 'check table t3_trans' under FTWRL.
+Success: Was able to run 'check table t3_trans' with FTWRL active in another connection.
+#
+# 39.2) Situation with DDL on temporary transactional tables is
+#       complex.
+#
+# 39.2.a) Some statements compatible with FTWRL since they don't
+#         do implicit commit.
+#
+# For example, CREATE TEMPORARY TABLE:
+Success: Was able to run 'create temporary table t4_temp_trans(i int) engine=innodb' under FTWRL.
+Success: Was able to run 'create temporary table t4_temp_trans(i int) engine=innodb' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'create temporary table t4_temp_trans(i int) engine=innodb' was active in another connection.
+#
+# Or DROP TEMPORARY TABLE:
+Success: Was able to run 'drop temporary tables t3_temp_trans' under FTWRL.
+Success: Was able to run 'drop temporary tables t3_temp_trans' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'drop temporary tables t3_temp_trans' was active in another connection.
+#
+# 39.2.b) Some statements do implicit commit but are considered
+#         read-only and so are compatible with FTWRL.
+#
+# For example, REPAIR TABLE:
+Success: Was able to run 'repair table t3_temp_trans' under FTWRL.
+Success: Was able to run 'repair table t3_temp_trans' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'repair table t3_temp_trans' was active in another connection.
+#
+# And ANALYZE TABLE:
+Success: Was able to run 'analyze table t3_temp_trans' under FTWRL.
+Success: Was able to run 'analyze table t3_temp_trans' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'analyze table t3_temp_trans' was active in another connection.
+#
+# 39.2.c) Some statements do implicit commit and not
+#         considered read-only. As result they are 
+#         not compatible with FTWRL.
+#
+flush tables with read lock;
+# Implicit commits are allowed under FTWRL.
+alter table t3_temp_trans add column c1 int;
+unlock tables;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+alter table t3_temp_trans drop column c1;
+# Switching to connection 'con1'.
+# Check that ALTER TABLE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap ALTER TABLE
+#
+# 40) Test effect of implicit commit for DDL which is otherwise
+#     compatible with FTWRL. Implicit commit at the start of DDL
+#     statement can make it incompatible with FTWRL if there are
+#     some changes to be commited even in case when DDL statement
+#     itself is compatible with FTWRL.
+#
+# For example CHECK TABLE for base non-transactional tables and
+# ALTER TABLE for temporary non-transactional tables are affected.
+begin;
+insert into t3_trans values (1);
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+check table t1_base;
+# Switching to connection 'con1'.
+# Check that CHECK TABLE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap CHECK TABLE
+Table	Op	Msg_type	Msg_text
+test.t1_base	check	status	OK
+begin;
+delete from t3_trans;
+#
+# Switching to connection 'con1'.
+flush tables with read lock;
+# Switching to connection 'default'.
+alter table t1_temp add column c1 int;
+# Switching to connection 'con1'.
+# Check that ALTER TABLE is blocked.
+unlock tables;
+# Switching to connection 'default'.
+# Reap ALTER TABLE
+alter table t1_temp drop column c1;
+#
+# Check that FLUSH TABLES WITH READ LOCK is blocked by individual
+# statements and is not blocked in the presence of transaction which
+# has done some changes earlier but is idle now (or does only reads).
+# This allows to use this statement even on systems which has long
+# running transactions.
+#
+begin;
+insert into t1_base values (1);
+insert into t3_trans values (1);
+# Switching to connection 'con1'.
+# The below FTWRL should not be blocked by transaction in 'default'.
+flush tables with read lock;
+# Switching to connection 'default'.
+# Transaction still is able to read even with FTWRL active in another
+# connection.
+select * from t1_base;
+i
+1
+select * from t2_base;
+j
+select * from t3_trans;
+i
+1
+# Switching to connection 'con1'.
+unlock tables;
+# Switching to connection 'default'.
+commit;
+delete from t1_base;
+delete from t3_trans;
+#
+# Check that impending FTWRL blocks new DML statements and
+# so can't be starved by a constant flow of DML.
+# (a.k.a. test for bug #54673 "It takes too long to get
+# readlock for 'FLUSH TABLES WITH READ LOCK'").
+#
+set debug_sync='RESET';
+set debug_sync='execute_command_after_close_tables SIGNAL parked WAIT_FOR go';
+insert into t1_base values (1);
+# Switching to connection 'con1'.
+set debug_sync='now WAIT_FOR parked';
+flush tables with read lock;
+# Switching to connection 'con2'.
+# Wait until FTWRL is blocked.
+# Try to run another INSERT and see that it is blocked.
+insert into t2_base values (1);;
+# Switching to connection 'con3'.
+# Wait until new INSERT is blocked.
+# Unblock INSERT in the first connection.
+set debug_sync='now SIGNAL go';
+# Switching to connection 'default'.
+# Reap first INSERT.
+# Switching to connection 'con1'.
+# Reap FTWRL.
+unlock tables;
+# Switching to connection 'con2'.
+# Reap second INSERT.
+# Switching to connection 'default'.
+set debug_sync= "RESET";
+delete from t1_base;
+delete from t2_base;
+
+# Check that COMMIT thas is issued after
+# FLUSH TABLES WITH READ LOCK is not blocked by
+# FLUSH TABLES WITH READ LOCK from another connection.
+# This scenario is used in innobackup.pl. The COMMIT goes
+# through because the transaction started by FTWRL does
+# not modify any tables, and the commit blocker lock is
+# only taken when there were such modifications.
+
+flush tables with read lock;
+# Switching to connection 'con1'.
+# The below FTWRL should not be blocked by transaction in 'default'.
+flush tables with read lock;
+# Switching to connection 'default'.
+select * from t1_base;
+i
+select * from t3_trans;
+i
+commit;
+# Switching to connection 'con1'.
+select * from t1_base;
+i
+select * from t3_trans;
+i
+commit;
+unlock tables;
+# Switching to connection 'default'.
+unlock tables;
+#
+# Check how FLUSH TABLE WITH READ LOCK is handled for MERGE tables.
+# As usual there are tricky cases related to this type of tables.
+#
+#
+# 1) Most typical case - base MERGE table with base underlying tables.
+#    
+# 1.a) DML statements which change data should be incompatible with FTWRL.
+create table tm_base (i int) engine=merge union=(t1_base) insert_method=last;
+Success: Was not able to run 'insert into tm_base values (1)' under FTWRL.
+Success: 'insert into tm_base values (1)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'insert into tm_base values (1)' is active in another connection.
+#
+# 1.b) DDL statement on such table should be incompatible with FTWRL as well.
+Success: Was not able to run 'alter table tm_base insert_method=first' under FTWRL.
+Success: 'alter table tm_base insert_method=first' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter table tm_base insert_method=first' is active in another connection.
+drop table tm_base;
+#
+# 2) Temporary MERGE table with base underlying tables.
+#
+# 2.a) DML statements which change data should be incompatible with FTWRL
+#      as they affect base tables.
+create temporary table tm_temp_base (i int) engine=merge union=(t1_base) insert_method=last;
+Success: Was not able to run 'insert into tm_temp_base values (1)' under FTWRL.
+Success: 'insert into tm_temp_base values (1)' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'insert into tm_temp_base values (1)' is active in another connection.
+#
+# 2.b) Some of DDL statements on such table can be compatible with FTWRL
+#      as they don't affect base tables.
+Success: Was able to run 'drop temporary tables tm_temp_base' under FTWRL.
+Success: Was able to run 'drop temporary tables tm_temp_base' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'drop temporary tables tm_temp_base' was active in another connection.
+#
+# 2.c) ALTER statement is incompatible with FTWRL. Even though it does
+#      not change data in base table it still acquires strong metadata
+#      locks on them.
+Success: Was not able to run 'alter table tm_temp_base insert_method=first' under FTWRL.
+Success: 'alter table tm_temp_base insert_method=first' is blocked by FTWRL active in another connection.
+Success: FTWRL is blocked when 'alter table tm_temp_base insert_method=first' is active in another connection.
+drop table tm_temp_base;
+#
+# 3) Temporary MERGE table with temporary underlying tables.
+#
+# 3.a) DML statements should be compatible with FTWRL as
+#      no base table is going to be affected.
+create temporary table tm_temp_temp (i int) engine=merge union=(t1_temp) insert_method=last;
+Success: Was able to run 'insert into tm_temp_temp values (1)' under FTWRL.
+Success: Was able to run 'insert into tm_temp_temp values (1)' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'insert into tm_temp_temp values (1)' was active in another connection.
+#
+# 3.b) DDL statements should be compatible with FTWRL as well
+#      as no base table is going to be affected too.
+Success: Was able to run 'alter table tm_temp_temp union=(t1_temp) insert_method=first' under FTWRL.
+Success: Was able to run 'alter table tm_temp_temp union=(t1_temp) insert_method=first' with FTWRL active in another connection.
+Success: Was able to run FTWRL while 'alter table tm_temp_temp union=(t1_temp) insert_method=first' was active in another connection.
+drop table tm_temp_temp;
+#
+# 4) For the sake of completeness let us check that base MERGE tables
+#    with temporary underlying tables are not functional.
+create table tm_base_temp (i int) engine=merge union=(t1_temp) insert_method=last;
+select * from tm_base_temp;
+ERROR HY000: Unable to open underlying table which is differently defined or of non-MyISAM type or doesn't exist
+drop table tm_base_temp;
+#
+# Clean-up.
+#
+drop event e1;
+drop function f2_temp;
+drop function f2_base;
+drop procedure p2;
+drop view v1;
+drop function f1;
+drop procedure p1;
+drop database `#mysql50#mysqltest-2`;
+drop database mysqltest1;
+drop temporary tables t1_temp, t2_temp;
+drop tables t1_base, t2_base, t3_trans;