mariadb/sql/event_data_objects.cc
Dmitry Lenev 6bf6272fda 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.
2010-11-11 20:11:05 +03:00

1535 lines
42 KiB
C++

/* Copyright (C) 2004-2006 MySQL AB, 2008-2009 Sun Microsystems, Inc
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
#define MYSQL_LEX 1
#include "my_global.h" /* NO_EMBEDDED_ACCESS_CHECKS */
#include "sql_priv.h"
#include "unireg.h"
#include "sql_parse.h" // parse_sql
#include "strfunc.h" // find_string_in_array
#include "sql_db.h" // get_default_db_collation
#include "sql_time.h" // interval_type_to_name,
// date_add_interval,
// calc_time_diff
#include "tztime.h" // my_tz_find, my_tz_OFFSET0, struct Time_zone
#include "sql_acl.h" // EVENT_ACL, SUPER_ACL
#include "sp.h" // load_charset, load_collation
#include "events.h"
#include "event_data_objects.h"
#include "event_db_repository.h"
#include "sp_head.h"
#include "sql_show.h" // append_definer, append_identifier
/**
@addtogroup Event_Scheduler
@{
*/
/*************************************************************************/
/**
Event_creation_ctx -- creation context of events.
*/
class Event_creation_ctx :public Stored_program_creation_ctx,
public Sql_alloc
{
public:
static bool load_from_db(THD *thd,
MEM_ROOT *event_mem_root,
const char *db_name,
const char *event_name,
TABLE *event_tbl,
Stored_program_creation_ctx **ctx);
public:
virtual Stored_program_creation_ctx *clone(MEM_ROOT *mem_root)
{
return new (mem_root)
Event_creation_ctx(m_client_cs, m_connection_cl, m_db_cl);
}
protected:
virtual Object_creation_ctx *create_backup_ctx(THD *thd) const
{
/*
We can avoid usual backup/restore employed in stored programs since we
know that this is a top level statement and the worker thread is
allocated exclusively to execute this event.
*/
return NULL;
}
private:
Event_creation_ctx(CHARSET_INFO *client_cs,
CHARSET_INFO *connection_cl,
CHARSET_INFO *db_cl)
: Stored_program_creation_ctx(client_cs, connection_cl, db_cl)
{ }
};
/**************************************************************************
Event_creation_ctx implementation.
**************************************************************************/
bool
Event_creation_ctx::load_from_db(THD *thd,
MEM_ROOT *event_mem_root,
const char *db_name,
const char *event_name,
TABLE *event_tbl,
Stored_program_creation_ctx **ctx)
{
/* Load character set/collation attributes. */
CHARSET_INFO *client_cs;
CHARSET_INFO *connection_cl;
CHARSET_INFO *db_cl;
bool invalid_creation_ctx= FALSE;
if (load_charset(event_mem_root,
event_tbl->field[ET_FIELD_CHARACTER_SET_CLIENT],
thd->variables.character_set_client,
&client_cs))
{
sql_print_warning("Event '%s'.'%s': invalid value "
"in column mysql.event.character_set_client.",
(const char *) db_name,
(const char *) event_name);
invalid_creation_ctx= TRUE;
}
if (load_collation(event_mem_root,
event_tbl->field[ET_FIELD_COLLATION_CONNECTION],
thd->variables.collation_connection,
&connection_cl))
{
sql_print_warning("Event '%s'.'%s': invalid value "
"in column mysql.event.collation_connection.",
(const char *) db_name,
(const char *) event_name);
invalid_creation_ctx= TRUE;
}
if (load_collation(event_mem_root,
event_tbl->field[ET_FIELD_DB_COLLATION],
NULL,
&db_cl))
{
sql_print_warning("Event '%s'.'%s': invalid value "
"in column mysql.event.db_collation.",
(const char *) db_name,
(const char *) event_name);
invalid_creation_ctx= TRUE;
}
/*
If we failed to resolve the database collation, load the default one
from the disk.
*/
if (!db_cl)
db_cl= get_default_db_collation(thd, db_name);
/* Create the context. */
*ctx= new Event_creation_ctx(client_cs, connection_cl, db_cl);
return invalid_creation_ctx;
}
/*************************************************************************/
/*
Initiliazes dbname and name of an Event_queue_element_for_exec
object
SYNOPSIS
Event_queue_element_for_exec::init()
RETURN VALUE
FALSE OK
TRUE Error (OOM)
*/
bool
Event_queue_element_for_exec::init(LEX_STRING db, LEX_STRING n)
{
if (!(dbname.str= my_strndup(db.str, dbname.length= db.length, MYF(MY_WME))))
return TRUE;
if (!(name.str= my_strndup(n.str, name.length= n.length, MYF(MY_WME))))
{
my_free(dbname.str);
return TRUE;
}
return FALSE;
}
/*
Destructor
SYNOPSIS
Event_queue_element_for_exec::~Event_queue_element_for_exec()
*/
Event_queue_element_for_exec::~Event_queue_element_for_exec()
{
my_free(dbname.str);
my_free(name.str);
}
/*
Constructor
SYNOPSIS
Event_basic::Event_basic()
*/
Event_basic::Event_basic()
{
DBUG_ENTER("Event_basic::Event_basic");
/* init memory root */
init_sql_alloc(&mem_root, 256, 512);
dbname.str= name.str= NULL;
dbname.length= name.length= 0;
time_zone= NULL;
DBUG_VOID_RETURN;
}
/*
Destructor
SYNOPSIS
Event_basic::Event_basic()
*/
Event_basic::~Event_basic()
{
DBUG_ENTER("Event_basic::~Event_basic");
free_root(&mem_root, MYF(0));
DBUG_VOID_RETURN;
}
/*
Short function to load a char column into a LEX_STRING
SYNOPSIS
Event_basic::load_string_field()
field_name The field( enum_events_table_field is not actually used
because it's unknown in event_data_objects.h)
fields The Field array
field_value The value
*/
bool
Event_basic::load_string_fields(Field **fields, ...)
{
bool ret= FALSE;
va_list args;
enum enum_events_table_field field_name;
LEX_STRING *field_value;
DBUG_ENTER("Event_basic::load_string_fields");
va_start(args, fields);
field_name= (enum enum_events_table_field) va_arg(args, int);
while (field_name < ET_FIELD_COUNT)
{
field_value= va_arg(args, LEX_STRING *);
if ((field_value->str= get_field(&mem_root, fields[field_name])) == NullS)
{
ret= TRUE;
break;
}
field_value->length= strlen(field_value->str);
field_name= (enum enum_events_table_field) va_arg(args, int);
}
va_end(args);
DBUG_RETURN(ret);
}
bool
Event_basic::load_time_zone(THD *thd, const LEX_STRING tz_name)
{
String str(tz_name.str, &my_charset_latin1);
time_zone= my_tz_find(thd, &str);
return (time_zone == NULL);
}
/*
Constructor
SYNOPSIS
Event_queue_element::Event_queue_element()
*/
Event_queue_element::Event_queue_element():
on_completion(Event_parse_data::ON_COMPLETION_DROP),
status(Event_parse_data::ENABLED), expression(0), dropped(FALSE),
execution_count(0)
{
DBUG_ENTER("Event_queue_element::Event_queue_element");
starts= ends= execute_at= last_executed= 0;
starts_null= ends_null= execute_at_null= TRUE;
DBUG_VOID_RETURN;
}
/*
Destructor
SYNOPSIS
Event_queue_element::Event_queue_element()
*/
Event_queue_element::~Event_queue_element()
{
}
/*
Constructor
SYNOPSIS
Event_timed::Event_timed()
*/
Event_timed::Event_timed():
created(0), modified(0), sql_mode(0)
{
DBUG_ENTER("Event_timed::Event_timed");
init();
DBUG_VOID_RETURN;
}
/*
Destructor
SYNOPSIS
Event_timed::~Event_timed()
*/
Event_timed::~Event_timed()
{
}
/*
Constructor
SYNOPSIS
Event_job_data::Event_job_data()
*/
Event_job_data::Event_job_data()
:sql_mode(0)
{
}
/*
Init all member variables
SYNOPSIS
Event_timed::init()
*/
void
Event_timed::init()
{
DBUG_ENTER("Event_timed::init");
definer_user.str= definer_host.str= body.str= comment.str= NULL;
definer_user.length= definer_host.length= body.length= comment.length= 0;
sql_mode= 0;
DBUG_VOID_RETURN;
}
/**
Load an event's body from a row from mysql.event.
@details This method is silent on errors and should behave like that.
Callers should handle throwing of error messages. The reason is that the
class should not know about how to deal with communication.
@return Operation status
@retval FALSE OK
@retval TRUE Error
*/
bool
Event_job_data::load_from_row(THD *thd, TABLE *table)
{
char *ptr;
size_t len;
LEX_STRING tz_name;
DBUG_ENTER("Event_job_data::load_from_row");
if (!table)
DBUG_RETURN(TRUE);
if (table->s->fields < ET_FIELD_COUNT)
DBUG_RETURN(TRUE);
if (load_string_fields(table->field,
ET_FIELD_DB, &dbname,
ET_FIELD_NAME, &name,
ET_FIELD_BODY, &body,
ET_FIELD_DEFINER, &definer,
ET_FIELD_TIME_ZONE, &tz_name,
ET_FIELD_COUNT))
DBUG_RETURN(TRUE);
if (load_time_zone(thd, tz_name))
DBUG_RETURN(TRUE);
Event_creation_ctx::load_from_db(thd, &mem_root, dbname.str, name.str, table,
&creation_ctx);
ptr= strchr(definer.str, '@');
if (! ptr)
ptr= definer.str;
len= ptr - definer.str;
definer_user.str= strmake_root(&mem_root, definer.str, len);
definer_user.length= len;
len= definer.length - len - 1;
/* 1:because of @ */
definer_host.str= strmake_root(&mem_root, ptr + 1, len);
definer_host.length= len;
sql_mode= (ulong) table->field[ET_FIELD_SQL_MODE]->val_int();
DBUG_RETURN(FALSE);
}
/**
Load an event's body from a row from mysql.event.
@details This method is silent on errors and should behave like that.
Callers should handle throwing of error messages. The reason is that the
class should not know about how to deal with communication.
@return Operation status
@retval FALSE OK
@retval TRUE Error
*/
bool
Event_queue_element::load_from_row(THD *thd, TABLE *table)
{
char *ptr;
MYSQL_TIME time;
LEX_STRING tz_name;
DBUG_ENTER("Event_queue_element::load_from_row");
if (!table)
DBUG_RETURN(TRUE);
if (table->s->fields < ET_FIELD_COUNT)
DBUG_RETURN(TRUE);
if (load_string_fields(table->field,
ET_FIELD_DB, &dbname,
ET_FIELD_NAME, &name,
ET_FIELD_DEFINER, &definer,
ET_FIELD_TIME_ZONE, &tz_name,
ET_FIELD_COUNT))
DBUG_RETURN(TRUE);
if (load_time_zone(thd, tz_name))
DBUG_RETURN(TRUE);
starts_null= table->field[ET_FIELD_STARTS]->is_null();
my_bool not_used= FALSE;
if (!starts_null)
{
table->field[ET_FIELD_STARTS]->get_date(&time, TIME_NO_ZERO_DATE);
starts= my_tz_OFFSET0->TIME_to_gmt_sec(&time,&not_used);
}
ends_null= table->field[ET_FIELD_ENDS]->is_null();
if (!ends_null)
{
table->field[ET_FIELD_ENDS]->get_date(&time, TIME_NO_ZERO_DATE);
ends= my_tz_OFFSET0->TIME_to_gmt_sec(&time,&not_used);
}
if (!table->field[ET_FIELD_INTERVAL_EXPR]->is_null())
expression= table->field[ET_FIELD_INTERVAL_EXPR]->val_int();
else
expression= 0;
/*
If neigher STARTS and ENDS is set, then both fields are empty.
Hence, if ET_FIELD_EXECUTE_AT is empty there is an error.
*/
execute_at_null= table->field[ET_FIELD_EXECUTE_AT]->is_null();
DBUG_ASSERT(!(starts_null && ends_null && !expression && execute_at_null));
if (!expression && !execute_at_null)
{
if (table->field[ET_FIELD_EXECUTE_AT]->get_date(&time,
TIME_NO_ZERO_DATE))
DBUG_RETURN(TRUE);
execute_at= my_tz_OFFSET0->TIME_to_gmt_sec(&time,&not_used);
}
/*
We load the interval type from disk as string and then map it to
an integer. This decouples the values of enum interval_type
and values actually stored on disk. Therefore the type can be
reordered without risking incompatibilities of data between versions.
*/
if (!table->field[ET_FIELD_TRANSIENT_INTERVAL]->is_null())
{
int i;
char buff[MAX_FIELD_WIDTH];
String str(buff, sizeof(buff), &my_charset_bin);
LEX_STRING tmp;
table->field[ET_FIELD_TRANSIENT_INTERVAL]->val_str(&str);
if (!(tmp.length= str.length()))
DBUG_RETURN(TRUE);
tmp.str= str.c_ptr_safe();
i= find_string_in_array(interval_type_to_name, &tmp, system_charset_info);
if (i < 0)
DBUG_RETURN(TRUE);
interval= (interval_type) i;
}
if (!table->field[ET_FIELD_LAST_EXECUTED]->is_null())
{
table->field[ET_FIELD_LAST_EXECUTED]->get_date(&time,
TIME_NO_ZERO_DATE);
last_executed= my_tz_OFFSET0->TIME_to_gmt_sec(&time,&not_used);
}
if ((ptr= get_field(&mem_root, table->field[ET_FIELD_STATUS])) == NullS)
DBUG_RETURN(TRUE);
DBUG_PRINT("load_from_row", ("Event [%s] is [%s]", name.str, ptr));
/* Set event status (ENABLED | SLAVESIDE_DISABLED | DISABLED) */
switch (ptr[0])
{
case 'E' :
status = Event_parse_data::ENABLED;
break;
case 'S' :
status = Event_parse_data::SLAVESIDE_DISABLED;
break;
case 'D' :
default:
status = Event_parse_data::DISABLED;
break;
}
if ((ptr= get_field(&mem_root, table->field[ET_FIELD_ORIGINATOR])) == NullS)
DBUG_RETURN(TRUE);
originator = table->field[ET_FIELD_ORIGINATOR]->val_int();
/* ToDo : Andrey . Find a way not to allocate ptr on event_mem_root */
if ((ptr= get_field(&mem_root,
table->field[ET_FIELD_ON_COMPLETION])) == NullS)
DBUG_RETURN(TRUE);
on_completion= (ptr[0]=='D'? Event_parse_data::ON_COMPLETION_DROP:
Event_parse_data::ON_COMPLETION_PRESERVE);
DBUG_RETURN(FALSE);
}
/**
Load an event's body from a row from mysql.event.
@details This method is silent on errors and should behave like that.
Callers should handle throwing of error messages. The reason is that the
class should not know about how to deal with communication.
@return Operation status
@retval FALSE OK
@retval TRUE Error
*/
bool
Event_timed::load_from_row(THD *thd, TABLE *table)
{
char *ptr;
size_t len;
DBUG_ENTER("Event_timed::load_from_row");
if (Event_queue_element::load_from_row(thd, table))
DBUG_RETURN(TRUE);
if (load_string_fields(table->field,
ET_FIELD_BODY, &body,
ET_FIELD_BODY_UTF8, &body_utf8,
ET_FIELD_COUNT))
DBUG_RETURN(TRUE);
if (Event_creation_ctx::load_from_db(thd, &mem_root, dbname.str, name.str,
table, &creation_ctx))
{
push_warning_printf(thd,
MYSQL_ERROR::WARN_LEVEL_WARN,
ER_EVENT_INVALID_CREATION_CTX,
ER(ER_EVENT_INVALID_CREATION_CTX),
(const char *) dbname.str,
(const char *) name.str);
}
ptr= strchr(definer.str, '@');
if (! ptr)
ptr= definer.str;
len= ptr - definer.str;
definer_user.str= strmake_root(&mem_root, definer.str, len);
definer_user.length= len;
len= definer.length - len - 1;
/* 1:because of @ */
definer_host.str= strmake_root(&mem_root, ptr + 1, len);
definer_host.length= len;
created= table->field[ET_FIELD_CREATED]->val_int();
modified= table->field[ET_FIELD_MODIFIED]->val_int();
comment.str= get_field(&mem_root, table->field[ET_FIELD_COMMENT]);
if (comment.str != NullS)
comment.length= strlen(comment.str);
else
comment.length= 0;
sql_mode= (ulong) table->field[ET_FIELD_SQL_MODE]->val_int();
DBUG_RETURN(FALSE);
}
/*
add_interval() adds a specified interval to time 'ltime' in time
zone 'time_zone', and returns the result converted to the number of
seconds since epoch (aka Unix time; in UTC time zone). Zero result
means an error.
*/
static
my_time_t
add_interval(MYSQL_TIME *ltime, const Time_zone *time_zone,
interval_type scale, INTERVAL interval)
{
if (date_add_interval(ltime, scale, interval))
return 0;
my_bool not_used;
return time_zone->TIME_to_gmt_sec(ltime, &not_used);
}
/*
Computes the sum of a timestamp plus interval.
SYNOPSIS
get_next_time()
time_zone event time zone
next the sum
start add interval_value to this time
time_now current time
i_value quantity of time type interval to add
i_type type of interval to add (SECOND, MINUTE, HOUR, WEEK ...)
RETURN VALUE
0 OK
1 Error
NOTES
1) If the interval is conversible to SECOND, like MINUTE, HOUR, DAY, WEEK.
Then we use TIMEDIFF()'s implementation as underlying and number of
seconds as resolution for computation.
2) In all other cases - MONTH, QUARTER, YEAR we use MONTH as resolution
and PERIOD_DIFF()'s implementation
*/
static
bool get_next_time(const Time_zone *time_zone, my_time_t *next,
my_time_t start, my_time_t time_now,
int i_value, interval_type i_type)
{
DBUG_ENTER("get_next_time");
DBUG_PRINT("enter", ("start: %lu now: %lu", (long) start, (long) time_now));
DBUG_ASSERT(start <= time_now);
longlong months=0, seconds=0;
switch (i_type) {
case INTERVAL_YEAR:
months= i_value*12;
break;
case INTERVAL_QUARTER:
/* Has already been converted to months */
case INTERVAL_YEAR_MONTH:
case INTERVAL_MONTH:
months= i_value;
break;
case INTERVAL_WEEK:
/* WEEK has already been converted to days */
case INTERVAL_DAY:
seconds= i_value*24*3600;
break;
case INTERVAL_DAY_HOUR:
case INTERVAL_HOUR:
seconds= i_value*3600;
break;
case INTERVAL_DAY_MINUTE:
case INTERVAL_HOUR_MINUTE:
case INTERVAL_MINUTE:
seconds= i_value*60;
break;
case INTERVAL_DAY_SECOND:
case INTERVAL_HOUR_SECOND:
case INTERVAL_MINUTE_SECOND:
case INTERVAL_SECOND:
seconds= i_value;
break;
case INTERVAL_DAY_MICROSECOND:
case INTERVAL_HOUR_MICROSECOND:
case INTERVAL_MINUTE_MICROSECOND:
case INTERVAL_SECOND_MICROSECOND:
case INTERVAL_MICROSECOND:
/*
We should return an error here so SHOW EVENTS/ SELECT FROM I_S.EVENTS
would give an error then.
*/
DBUG_RETURN(1);
break;
case INTERVAL_LAST:
DBUG_ASSERT(0);
}
DBUG_PRINT("info", ("seconds: %ld months: %ld", (long) seconds, (long) months));
MYSQL_TIME local_start;
MYSQL_TIME local_now;
/* Convert times from UTC to local. */
{
time_zone->gmt_sec_to_TIME(&local_start, start);
time_zone->gmt_sec_to_TIME(&local_now, time_now);
}
INTERVAL interval;
bzero(&interval, sizeof(interval));
my_time_t next_time= 0;
if (seconds)
{
longlong seconds_diff;
long microsec_diff;
bool negative= calc_time_diff(&local_now, &local_start, 1,
&seconds_diff, &microsec_diff);
if (!negative)
{
/*
The formula below returns the interval that, when added to
local_start, will always give the time in the future.
*/
interval.second= seconds_diff - seconds_diff % seconds + seconds;
next_time= add_interval(&local_start, time_zone,
INTERVAL_SECOND, interval);
if (next_time == 0)
goto done;
}
if (next_time <= time_now)
{
/*
If 'negative' is true above, then 'next_time == 0', and
'next_time <= time_now' is also true. If negative is false,
then next_time was set, but perhaps to the value that is less
then time_now. See below for elaboration.
*/
DBUG_ASSERT(negative || next_time > 0);
/*
If local_now < local_start, i.e. STARTS time is in the future
according to the local time (it always in the past according
to UTC---this is a prerequisite of this function), then
STARTS is almost always in the past according to the local
time too. However, in the time zone that has backward
Daylight Saving Time shift, the following may happen: suppose
we have a backward DST shift at certain date after 2:59:59,
i.e. local time goes 1:59:59, 2:00:00, ... , 2:59:59, (shift
here) 2:00:00 (again), ... , 2:59:59 (again), 3:00:00, ... .
Now suppose the time has passed the first 2:59:59, has been
shifted backward, and now is (the second) 2:20:00. The user
does CREATE EVENT with STARTS 'current-date 2:40:00'. Local
time 2:40:00 from create statement is treated by time
functions as the first such time, so according to UTC it comes
before the second 2:20:00. But according to local time it is
obviously in the future, so we end up in this branch.
Since we are in the second pass through 2:00:00--2:59:59, and
any local time form this interval is treated by system
functions as the time from the first pass, we have to find the
time for the next execution that is past the DST-affected
interval (past the second 2:59:59 for our example,
i.e. starting from 3:00:00). We do this in the loop until the
local time is mapped onto future UTC time. 'start' time is in
the past, so we may use 'do { } while' here, and add the first
interval right away.
Alternatively, it could be that local_now >= local_start. Now
for the example above imagine we do CREATE EVENT with STARTS
'current-date 2:10:00'. Local start 2:10 is in the past (now
is local 2:20), so we add an interval, and get next execution
time, say, 2:40. It is in the future according to local time,
but, again, since we are in the second pass through
2:00:00--2:59:59, 2:40 will be converted into UTC time in the
past. So we will end up in this branch again, and may add
intervals in a 'do { } while' loop.
Note that for any given event we may end up here only if event
next execution time will map to the time interval that is
passed twice, and only if the server was started during the
second pass, or the event is being created during the second
pass. After that, we never will get here (unless we again
start the server during the second pass). In other words,
such a condition is extremely rare.
*/
interval.second= seconds;
do
{
next_time= add_interval(&local_start, time_zone,
INTERVAL_SECOND, interval);
if (next_time == 0)
goto done;
}
while (next_time <= time_now);
}
}
else
{
long diff_months= ((long) local_now.year - (long) local_start.year)*12 +
((long) local_now.month - (long) local_start.month);
/*
Unlike for seconds above, the formula below returns the interval
that, when added to the local_start, will give the time in the
past, or somewhere in the current month. We are interested in
the latter case, to see if this time has already passed, or is
yet to come this month.
Note that the time is guaranteed to be in the past unless
(diff_months % months == 0), but no good optimization is
possible here, because (diff_months % months == 0) is what will
happen most of the time, as get_next_time() will be called right
after the execution of the event. We could pass last_executed
time to this function, and see if the execution has already
happened this month, but for that we will have to convert
last_executed from seconds since epoch to local broken-down
time, and this will greatly reduce the effect of the
optimization. So instead we keep the code simple and clean.
*/
interval.month= (ulong) (diff_months - diff_months % months);
next_time= add_interval(&local_start, time_zone,
INTERVAL_MONTH, interval);
if (next_time == 0)
goto done;
if (next_time <= time_now)
{
interval.month= (ulong) months;
next_time= add_interval(&local_start, time_zone,
INTERVAL_MONTH, interval);
if (next_time == 0)
goto done;
}
}
DBUG_ASSERT(time_now < next_time);
*next= next_time;
done:
DBUG_PRINT("info", ("next_time: %ld", (long) next_time));
DBUG_RETURN(next_time == 0);
}
/*
Computes next execution time.
SYNOPSIS
Event_queue_element::compute_next_execution_time()
RETURN VALUE
FALSE OK
TRUE Error
NOTES
The time is set in execute_at, if no more executions the latter is
set to 0.
*/
bool
Event_queue_element::compute_next_execution_time()
{
my_time_t time_now;
DBUG_ENTER("Event_queue_element::compute_next_execution_time");
DBUG_PRINT("enter", ("starts: %lu ends: %lu last_executed: %lu this: 0x%lx",
(long) starts, (long) ends, (long) last_executed,
(long) this));
if (status != Event_parse_data::ENABLED)
{
DBUG_PRINT("compute_next_execution_time",
("Event %s is DISABLED", name.str));
goto ret;
}
/* If one-time, no need to do computation */
if (!expression)
{
/* Let's check whether it was executed */
if (last_executed)
{
DBUG_PRINT("info",("One-time event %s.%s of was already executed",
dbname.str, name.str));
dropped= (on_completion == Event_parse_data::ON_COMPLETION_DROP);
DBUG_PRINT("info",("One-time event will be dropped: %d.", dropped));
status= Event_parse_data::DISABLED;
}
goto ret;
}
time_now= (my_time_t) current_thd->query_start();
DBUG_PRINT("info",("NOW: [%lu]", (ulong) time_now));
/* if time_now is after ends don't execute anymore */
if (!ends_null && ends < time_now)
{
DBUG_PRINT("info", ("NOW after ENDS, don't execute anymore"));
/* time_now is after ends. don't execute anymore */
execute_at= 0;
execute_at_null= TRUE;
if (on_completion == Event_parse_data::ON_COMPLETION_DROP)
dropped= TRUE;
DBUG_PRINT("info", ("Dropped: %d", dropped));
status= Event_parse_data::DISABLED;
goto ret;
}
/*
Here time_now is before or equals ends if the latter is set.
Let's check whether time_now is before starts.
If so schedule for starts.
*/
if (!starts_null && time_now <= starts)
{
if (time_now == starts && starts == last_executed)
{
/*
do nothing or we will schedule for second time execution at starts.
*/
}
else
{
DBUG_PRINT("info", ("STARTS is future, NOW <= STARTS,sched for STARTS"));
/*
starts is in the future
time_now before starts. Scheduling for starts
*/
execute_at= starts;
execute_at_null= FALSE;
goto ret;
}
}
if (!starts_null && !ends_null)
{
/*
Both starts and m_ends are set and time_now is between them (incl.)
If last_executed is set then increase with m_expression. The new MYSQL_TIME is
after m_ends set execute_at to 0. And check for on_completion
If not set then schedule for now.
*/
DBUG_PRINT("info", ("Both STARTS & ENDS are set"));
if (!last_executed)
{
DBUG_PRINT("info", ("Not executed so far."));
}
{
my_time_t next_exec;
if (get_next_time(time_zone, &next_exec, starts, time_now,
(int) expression, interval))
goto err;
/* There was previous execution */
if (ends < next_exec)
{
DBUG_PRINT("info", ("Next execution of %s after ENDS. Stop executing.",
name.str));
/* Next execution after ends. No more executions */
execute_at= 0;
execute_at_null= TRUE;
if (on_completion == Event_parse_data::ON_COMPLETION_DROP)
dropped= TRUE;
status= Event_parse_data::DISABLED;
}
else
{
DBUG_PRINT("info",("Next[%lu]", (ulong) next_exec));
execute_at= next_exec;
execute_at_null= FALSE;
}
}
goto ret;
}
else if (starts_null && ends_null)
{
/* starts is always set, so this is a dead branch !! */
DBUG_PRINT("info", ("Neither STARTS nor ENDS are set"));
/*
Both starts and m_ends are not set, so we schedule for the next
based on last_executed.
*/
if (last_executed)
{
my_time_t next_exec;
if (get_next_time(time_zone, &next_exec, starts, time_now,
(int) expression, interval))
goto err;
execute_at= next_exec;
DBUG_PRINT("info",("Next[%lu]", (ulong) next_exec));
}
else
{
/* last_executed not set. Schedule the event for now */
DBUG_PRINT("info", ("Execute NOW"));
execute_at= time_now;
}
execute_at_null= FALSE;
}
else
{
/* either starts or m_ends is set */
if (!starts_null)
{
DBUG_PRINT("info", ("STARTS is set"));
/*
- starts is set.
- starts is not in the future according to check made before
Hence schedule for starts + m_expression in case last_executed
is not set, otherwise to last_executed + m_expression
*/
if (!last_executed)
{
DBUG_PRINT("info", ("Not executed so far."));
}
{
my_time_t next_exec;
if (get_next_time(time_zone, &next_exec, starts, time_now,
(int) expression, interval))
goto err;
execute_at= next_exec;
DBUG_PRINT("info",("Next[%lu]", (ulong) next_exec));
}
execute_at_null= FALSE;
}
else
{
/* this is a dead branch, because starts is always set !!! */
DBUG_PRINT("info", ("STARTS is not set. ENDS is set"));
/*
- m_ends is set
- m_ends is after time_now or is equal
Hence check for m_last_execute and increment with m_expression.
If last_executed is not set then schedule for now
*/
if (!last_executed)
execute_at= time_now;
else
{
my_time_t next_exec;
if (get_next_time(time_zone, &next_exec, starts, time_now,
(int) expression, interval))
goto err;
if (ends < next_exec)
{
DBUG_PRINT("info", ("Next execution after ENDS. Stop executing."));
execute_at= 0;
execute_at_null= TRUE;
status= Event_parse_data::DISABLED;
if (on_completion == Event_parse_data::ON_COMPLETION_DROP)
dropped= TRUE;
}
else
{
DBUG_PRINT("info", ("Next[%lu]", (ulong) next_exec));
execute_at= next_exec;
execute_at_null= FALSE;
}
}
}
goto ret;
}
ret:
DBUG_PRINT("info", ("ret: 0 execute_at: %lu", (long) execute_at));
DBUG_RETURN(FALSE);
err:
DBUG_PRINT("info", ("ret=1"));
DBUG_RETURN(TRUE);
}
/*
Set the internal last_executed MYSQL_TIME struct to now. NOW is the
time according to thd->query_start(), so the THD's clock.
SYNOPSIS
Event_queue_element::mark_last_executed()
thd thread context
*/
void
Event_queue_element::mark_last_executed(THD *thd)
{
last_executed= (my_time_t) thd->query_start();
execution_count++;
}
static
void
append_datetime(String *buf, Time_zone *time_zone, my_time_t secs,
const char *name, uint len)
{
char dtime_buff[20*2+32];/* +32 to make my_snprintf_{8bit|ucs2} happy */
buf->append(STRING_WITH_LEN(" "));
buf->append(name, len);
buf->append(STRING_WITH_LEN(" '"));
/*
Pass the buffer and the second param tells fills the buffer and
returns the number of chars to copy.
*/
MYSQL_TIME time;
time_zone->gmt_sec_to_TIME(&time, secs);
buf->append(dtime_buff, my_datetime_to_str(&time, dtime_buff));
buf->append(STRING_WITH_LEN("'"));
}
/*
Get SHOW CREATE EVENT as string
SYNOPSIS
Event_timed::get_create_event(THD *thd, String *buf)
thd Thread
buf String*, should be already allocated. CREATE EVENT goes inside.
RETURN VALUE
0 OK
EVEX_MICROSECOND_UNSUP Error (for now if mysql.event has been
tampered and MICROSECONDS interval or
derivative has been put there.
*/
int
Event_timed::get_create_event(THD *thd, String *buf)
{
char tmp_buf[2 * STRING_BUFFER_USUAL_SIZE];
String expr_buf(tmp_buf, sizeof(tmp_buf), system_charset_info);
expr_buf.length(0);
DBUG_ENTER("get_create_event");
DBUG_PRINT("ret_info",("body_len=[%d]body=[%s]",
(int) body.length, body.str));
if (expression && Events::reconstruct_interval_expression(&expr_buf, interval,
expression))
DBUG_RETURN(EVEX_MICROSECOND_UNSUP);
buf->append(STRING_WITH_LEN("CREATE "));
append_definer(thd, buf, &definer_user, &definer_host);
buf->append(STRING_WITH_LEN("EVENT "));
append_identifier(thd, buf, name.str, name.length);
if (expression)
{
buf->append(STRING_WITH_LEN(" ON SCHEDULE EVERY "));
buf->append(expr_buf);
buf->append(' ');
LEX_STRING *ival= &interval_type_to_name[interval];
buf->append(ival->str, ival->length);
if (!starts_null)
append_datetime(buf, time_zone, starts, STRING_WITH_LEN("STARTS"));
if (!ends_null)
append_datetime(buf, time_zone, ends, STRING_WITH_LEN("ENDS"));
}
else
{
append_datetime(buf, time_zone, execute_at,
STRING_WITH_LEN("ON SCHEDULE AT"));
}
if (on_completion == Event_parse_data::ON_COMPLETION_DROP)
buf->append(STRING_WITH_LEN(" ON COMPLETION NOT PRESERVE "));
else
buf->append(STRING_WITH_LEN(" ON COMPLETION PRESERVE "));
if (status == Event_parse_data::ENABLED)
buf->append(STRING_WITH_LEN("ENABLE"));
else if (status == Event_parse_data::SLAVESIDE_DISABLED)
buf->append(STRING_WITH_LEN("DISABLE ON SLAVE"));
else
buf->append(STRING_WITH_LEN("DISABLE"));
if (comment.length)
{
buf->append(STRING_WITH_LEN(" COMMENT "));
append_unescaped(buf, comment.str, comment.length);
}
buf->append(STRING_WITH_LEN(" DO "));
buf->append(body.str, body.length);
DBUG_RETURN(0);
}
/**
Get an artificial stored procedure to parse as an event definition.
*/
bool
Event_job_data::construct_sp_sql(THD *thd, String *sp_sql)
{
LEX_STRING buffer;
const uint STATIC_SQL_LENGTH= 44;
DBUG_ENTER("Event_job_data::construct_sp_sql");
/*
Allocate a large enough buffer on the thread execution memory
root to avoid multiple [re]allocations on system heap
*/
buffer.length= STATIC_SQL_LENGTH + name.length + body.length;
if (! (buffer.str= (char*) thd->alloc(buffer.length)))
DBUG_RETURN(TRUE);
sp_sql->set(buffer.str, buffer.length, system_charset_info);
sp_sql->length(0);
sp_sql->append(C_STRING_WITH_LEN("CREATE "));
sp_sql->append(C_STRING_WITH_LEN("PROCEDURE "));
/*
Let's use the same name as the event name to perhaps produce a
better error message in case it is a part of some parse error.
We're using append_identifier here to successfully parse
events with reserved names.
*/
append_identifier(thd, sp_sql, name.str, name.length);
/*
The default SQL security of a stored procedure is DEFINER. We
have already activated the security context of the event, so
let's execute the procedure with the invoker rights to save on
resets of security contexts.
*/
sp_sql->append(C_STRING_WITH_LEN("() SQL SECURITY INVOKER "));
sp_sql->append(body.str, body.length);
DBUG_RETURN(thd->is_fatal_error);
}
/**
Get DROP EVENT statement to binlog the drop of ON COMPLETION NOT
PRESERVE event.
*/
bool
Event_job_data::construct_drop_event_sql(THD *thd, String *sp_sql)
{
LEX_STRING buffer;
const uint STATIC_SQL_LENGTH= 14;
DBUG_ENTER("Event_job_data::construct_drop_event_sql");
buffer.length= STATIC_SQL_LENGTH + name.length*2 + dbname.length*2;
if (! (buffer.str= (char*) thd->alloc(buffer.length)))
DBUG_RETURN(TRUE);
sp_sql->set(buffer.str, buffer.length, system_charset_info);
sp_sql->length(0);
sp_sql->append(C_STRING_WITH_LEN("DROP EVENT "));
append_identifier(thd, sp_sql, dbname.str, dbname.length);
sp_sql->append('.');
append_identifier(thd, sp_sql, name.str, name.length);
DBUG_RETURN(thd->is_fatal_error);
}
/**
Compiles and executes the event (the underlying sp_head object)
@retval TRUE error (reported to the error log)
@retval FALSE success
*/
bool
Event_job_data::execute(THD *thd, bool drop)
{
String sp_sql;
#ifndef NO_EMBEDDED_ACCESS_CHECKS
Security_context event_sctx, *save_sctx= NULL;
#endif
List<Item> empty_item_list;
bool ret= TRUE;
DBUG_ENTER("Event_job_data::execute");
mysql_reset_thd_for_next_command(thd);
/*
MySQL parser currently assumes that current database is either
present in THD or all names in all statements are fully specified.
And yet not fully specified names inside stored programs must be
be supported, even if the current database is not set:
CREATE PROCEDURE db1.p1() BEGIN CREATE TABLE t1; END//
-- in this example t1 should be always created in db1 and the statement
must parse even if there is no current database.
To support this feature and still address the parser limitation,
we need to set the current database here.
We don't have to call mysql_change_db, since the checks performed
in it are unnecessary for the purpose of parsing, and
mysql_change_db will be invoked anyway later, to activate the
procedure database before it's executed.
*/
thd->set_db(dbname.str, dbname.length);
lex_start(thd);
#ifndef NO_EMBEDDED_ACCESS_CHECKS
if (event_sctx.change_security_context(thd,
&definer_user, &definer_host,
&dbname, &save_sctx))
{
sql_print_error("Event Scheduler: "
"[%s].[%s.%s] execution failed, "
"failed to authenticate the user.",
definer.str, dbname.str, name.str);
goto end;
}
#endif
if (check_access(thd, EVENT_ACL, dbname.str, NULL, NULL, 0, 0))
{
/*
This aspect of behavior is defined in the worklog,
and this is how triggers work too: if TRIGGER
privilege is revoked from trigger definer,
triggers are not executed.
*/
sql_print_error("Event Scheduler: "
"[%s].[%s.%s] execution failed, "
"user no longer has EVENT privilege.",
definer.str, dbname.str, name.str);
goto end;
}
if (construct_sp_sql(thd, &sp_sql))
goto end;
/*
Set up global thread attributes to reflect the properties of
this Event. We can simply reset these instead of usual
backup/restore employed in stored programs since we know that
this is a top level statement and the worker thread is
allocated exclusively to execute this event.
*/
thd->variables.sql_mode= sql_mode;
thd->variables.time_zone= time_zone;
thd->set_query(sp_sql.c_ptr_safe(), sp_sql.length());
{
Parser_state parser_state;
if (parser_state.init(thd, thd->query(), thd->query_length()))
goto end;
if (parse_sql(thd, & parser_state, creation_ctx))
{
sql_print_error("Event Scheduler: "
"%serror during compilation of %s.%s",
thd->is_fatal_error ? "fatal " : "",
(const char *) dbname.str, (const char *) name.str);
goto end;
}
}
{
sp_head *sphead= thd->lex->sphead;
DBUG_ASSERT(sphead);
if (thd->enable_slow_log)
sphead->m_flags|= sp_head::LOG_SLOW_STATEMENTS;
sphead->m_flags|= sp_head::LOG_GENERAL_LOG;
sphead->set_info(0, 0, &thd->lex->sp_chistics, sql_mode);
sphead->set_creation_ctx(creation_ctx);
sphead->optimize();
ret= sphead->execute_procedure(thd, &empty_item_list);
/*
There is no pre-locking and therefore there should be no
tables open and locked left after execute_procedure.
*/
}
end:
if (drop && !thd->is_fatal_error)
{
/*
We must do it here since here we're under the right authentication
ID of the event definer.
*/
sql_print_information("Event Scheduler: Dropping %s.%s",
(const char *) dbname.str, (const char *) name.str);
/*
Construct a query for the binary log, to ensure the event is dropped
on the slave
*/
if (construct_drop_event_sql(thd, &sp_sql))
ret= 1;
else
{
ulong saved_master_access;
thd->set_query(sp_sql.c_ptr_safe(), sp_sql.length());
/*
NOTE: even if we run in read-only mode, we should be able to lock
the mysql.event table for writing. In order to achieve this, we
should call mysql_lock_tables() under the super-user.
*/
saved_master_access= thd->security_ctx->master_access;
thd->security_ctx->master_access |= SUPER_ACL;
ret= Events::drop_event(thd, dbname, name, FALSE);
thd->security_ctx->master_access= saved_master_access;
}
}
#ifndef NO_EMBEDDED_ACCESS_CHECKS
if (save_sctx)
event_sctx.restore_security_context(thd, save_sctx);
#endif
thd->lex->unit.cleanup();
thd->end_statement();
thd->cleanup_after_query();
/* Avoid races with SHOW PROCESSLIST */
thd->set_query(NULL, 0);
DBUG_PRINT("info", ("EXECUTED %s.%s ret: %d", dbname.str, name.str, ret));
DBUG_RETURN(ret);
}
/*
Checks whether two events are in the same schema
SYNOPSIS
event_basic_db_equal()
db Schema
et Compare et->dbname to `db`
RETURN VALUE
TRUE Equal
FALSE Not equal
*/
bool
event_basic_db_equal(LEX_STRING db, Event_basic *et)
{
return !sortcmp_lex_string(et->dbname, db, system_charset_info);
}
/*
Checks whether an event has equal `db` and `name`
SYNOPSIS
event_basic_identifier_equal()
db Schema
name Name
et The event object
RETURN VALUE
TRUE Equal
FALSE Not equal
*/
bool
event_basic_identifier_equal(LEX_STRING db, LEX_STRING name, Event_basic *b)
{
return !sortcmp_lex_string(name, b->name, system_charset_info) &&
!sortcmp_lex_string(db, b->dbname, system_charset_info);
}
/**
@} (End of group Event_Scheduler)
*/