When the thread executing a DDL was killed after finished its
execution but before writing the binlog event, the error code in
the binlog event could be set wrongly to ER_SERVER_SHUTDOWN or
ER_QUERY_INTERRUPTED.
This patch fixed the problem by ignoring the kill status when
constructing the event for DDL statements.
This patch also included the following changes in order to
provide the test case.
1) modified mysqltest to support variable for connection command
2) modified mysql-test-run.pl, add new variable MYSQL_SLAVE to
run mysql client against the slave mysqld.
When do 'insert delayed' operation, the time_zone info doesn't be keeped in the row info.
So when we do insert sometime later, time_zone didn't write into binlog.
This will cause wrong result for timestamp column in slave.
Our solution is that adding time_zone info with the delayed-row and
restoring time_zone from row-info when execute that row in the furture by another thread.
So we can write correct time_zone info into binlog and got correct result in slave.
mysql-test/r/rpl_timezone.result:
Test result
mysql-test/t/rpl_timezone.test:
Add test for bug#41719
sql/sql_insert.cc:
Add time_zone info in the delayed-row and restore time_zone when execute the row in the furture by another thread.
- Remove bothersome warning messages. This change focuses on the warnings
that are covered by the ignore file: support-files/compiler_warnings.supp.
- Strings are guaranteed to be max uint in length
If delayed insert fails to upgrade the lock it was not
freeing the temporary memory storage used to keep
newly constructed blob values in memory.
Fixed by iterating over the remaining rows in the delayed
insert rowset and freeing the blob storage for each row.
No test suite because it involves concurrent delayed inserts
on a table and cannot easily be made deterministic.
Added a correct valgrind suppression for Fedora 9.
mysql-test/valgrind.supp:
Added a vagrind suppression for Fedora 9
sql/sql_insert.cc:
Bug #38693: free the blobs temp storage on error.
in open_table()
Problem: repeating "CREATE... ( AUTOINCREMENT) ... SELECT" may lead to
an assertion failure.
Fix: reset table->auto_increment_field_not_null after each record
writing.
mysql-test/r/create.result:
Fix for bug#38821: Assert table->auto_increment_field_not_null failed
in open_table()
- test result.
mysql-test/t/create.test:
Fix for bug#38821: Assert table->auto_increment_field_not_null failed
in open_table()
- test case.
sql/sql_insert.cc:
Fix for bug#38821: Assert table->auto_increment_field_not_null failed
in open_table()
- reset table->auto_increment_field_not_null after writing a record
for "{CREATE, INSERT}..SELECT".
Bug#21413
"Engine table handler used by multiple threads in REPLACE DELAYED"
When executing a REPLACE DELAYED statement, the storage engine
::extra() method was invoked by a different thread than the thread
which has acquired the handler instance.
This did not cause problems within the current server and with
the current storage engines.
But it has the potential to confuse future storage engines.
Added code to avoid surplus calls to extra() method in case of DELAYED
which avoids calling storage engine from a different thread than
expected.
No test case.
This change does not change behavior in conjunction with current
storage engines. So it cannot be tested by the regression test suite.
sql/sql_insert.cc:
Bug#21413
When performing DELAYED operations, we should not call
the storage engine methods from the 'wrong' context.
Ensure that the calls to the methods are appropiately guarded
by either moving calls to sections where the lock_type is checked.
No need to copy all elements of TABLE_LIST when opening table
using open_ltable().
sporadically
Under some circumstances, the mysql_insert_id() value after SELECT ...
INSERT could return a wrong value. This could happen when the last
SELECT ... INSERT did not involve an AUTO_INCREMENT column, but the
value of mysql_insert_id() was changed by some previous statements.
Fixed by checking the value of thd->insert_id_used in
select_insert::send_eof() and returning 0 for mysql_insert_id() if it
is not set.
sql/sql_insert.cc:
Do not return thd->last_insert_id unconditionally in
select_insert::send_eof(). First check if thd->insert_id_used is
non-zero, and return 0 otherwise.
tests/mysql_client_test.c:
Added a test case for bug #34889.
value when inserting into a view.
The mysql_prepare_insert function checks all fields of the target table that
directly or indirectly (through a view) are specified in the INSERT
statement to have a default value. This check can be skipped if the INSERT
statement doesn't mention any insert fields. In case of a view this allows
fields that aren't mentioned in the view to bypass the check.
Now fields of the target table are always checked to have a default value
when insert goes into a view.
mysql-test/t/view.test:
Added a test case for the bug#29477: Not all fields of the target table were
checked to have a default value when inserting into a view.
mysql-test/r/view.result:
Added a test case for the bug#29477: Not all fields of the target table were
checked to have a default value when inserting into a view.
sql/sql_insert.cc:
Bug#29477: Not all fields of the target table were checked to have a default
value when inserting into a view.
Now fields of the target table are always checked to have a default value
when insert goes into a view.
into mysql.com:/home/gluh/MySQL/Merge/5.0-opt
client/mysql.cc:
Auto merged
client/mysqltest.c:
Auto merged
include/mysql_com.h:
Auto merged
libmysql/CMakeLists.txt:
Auto merged
myisam/mi_check.c:
Auto merged
mysql-test/r/delayed.result:
Auto merged
mysql-test/r/innodb.result:
Auto merged
mysql-test/r/myisam.result:
Auto merged
mysql-test/r/ps.result:
Auto merged
mysql-test/t/merge.test:
Auto merged
mysql-test/t/myisam.test:
Auto merged
mysql-test/t/subselect.test:
Auto merged
mysql-test/t/type_datetime.test:
Auto merged
mysql-test/t/variables.test:
Auto merged
sql/field.cc:
Auto merged
sql/ha_myisam.cc:
Auto merged
sql/item.cc:
Auto merged
sql/item.h:
Auto merged
sql/item_cmpfunc.cc:
Auto merged
sql/item_func.cc:
Auto merged
sql/item_func.h:
Auto merged
sql/mysql_priv.h:
Auto merged
sql/mysqld.cc:
Auto merged
sql/opt_range.cc:
Auto merged
sql/set_var.cc:
Auto merged
sql/set_var.h:
Auto merged
sql/sql_class.h:
Auto merged
sql/sql_select.cc:
Auto merged
sql/sql_table.cc:
Auto merged
sql/sql_yacc.yy:
Auto merged
mysql-test/r/func_misc.result:
manual merge
mysql-test/r/innodb_mysql.result:
manual merge
mysql-test/t/func_misc.test:
manual merge
mysql-test/t/innodb_mysql.test:
manual merge
sql/sql_insert.cc:
manual merge
into mysql.com:/home/ram/work/b32676/b32676.5.0
mysql-test/r/delayed.result:
Auto merged
mysql-test/t/delayed.test:
Auto merged
sql/sql_insert.cc:
Auto merged
insert ... select.
The 5.0 manual page for mysql_insert_id() does not mention anything
about INSERT ... SELECT, though its current behavior is incosistent
with what the manual says about the plain INSERT.
Fixed by changing the AUTO_INCREMENT and mysql_insert_id() handling
logic in INSERT ... SELECT to be consistent with the INSERT behavior,
the manual, and the changes in 5.1 introduced by WL3146:
- mysql_insert_id() now returns the first automatically generated
AUTO_INCREMENT value that was successfully inserted by INSERT ... SELECT
- if an INSERT ... SELECT statement is executed, and no automatically
generated value is successfully inserted, mysql_insert_id() now returns
the ID of the last inserted row.
sql/sql_class.h:
Replaced last_insert_id with autoinc_value_of_last_inserted_row to be
consistent with 5.1 code.
sql/sql_insert.cc:
Revised the AUTO_INCREMENT and mysql_insert_id() handling logic in
INSERT ... SELECT to be consistent with INSERT behavior, the manual, and
changes in 5.1 introduced by WL3146:
- mysql_insert_id() now returns the first automatically generated
AUTO_INCREMENT value that was successfully inserted;
- if an INSERT ... SELECT statement is executed, and no automatically
generated value is successfully inserted, mysql_insert_id() now returns
the ID of the last inserted row.
tests/mysql_client_test.c:
Backported the test cases related to INSERT ... SELECT and
mysql_insert_id() from WL3146 patch to 5.0.
Problem: using wrong local lock type value in the mysql_insert() results in a crash.
Fix: use a proper value.
mysql-test/r/delayed.result:
Fix for bug #32676: insert delayed crash with wrong column and function specified
- test result.
mysql-test/t/delayed.test:
Fix for bug #32676: insert delayed crash with wrong column and function specified
- test case.
sql/sql_insert.cc:
Fix for bug #32676: insert delayed crash with wrong column and function specified
- the local lock_type var assigment displaced just after the line
where the table_list->lock_type is filnally defined in the mysql_insert()
to avoid using its old value.
led to creating corrupted index.
Corrected fix. The new method called prepare2 is added to the select_create
class. As all preparations are done by the select_create::prepare function
it doesn't do anything. Slightly changed algorithm of calling the
start_bulk_insert function. Now it's called from the select_insert::prepare2
function when the SQL_BUFFER_RESULT flags is set.
The is_bulk_insert_mode flag is removed as it is not needed anymore.
sql/sql_class.h:
Bug#30384: Having SQL_BUFFER_RESULT option in the CREATE .. KEY(..) .. SELECT
led to creating corrupted index.
Corrected fix. The new method called prepare2 is added to the select_create
class. As all preparetions are done by the select_create::prepare function
it doesn't do anything.
The is_bulk_insert_mode flag is removed as it is not needed anymore.
sql/sql_insert.cc:
Bug#30384: Having SQL_BUFFER_RESULT option in the CREATE .. KEY(..) .. SELECT
led to creating corrupted index. Slightly changed algorithm of calling the
start_bulk_insert function. Now it's called from the select_insert::prepare2
function when the SQL_BUFFER_RESULT flags is set.
Corrected fix. The is_bulk_insert_mode flag is removed as it is not needed
anymore.
into mysql.com:/home/bar/mysql-work/mysql-5.0-rpl-merge
mysql-test/r/ctype_utf8.result:
Auto merged
mysql-test/r/func_regexp.result:
Auto merged
mysql-test/t/ctype_utf8.test:
Auto merged
mysql-test/t/func_regexp.test:
Auto merged
sql/item_cmpfunc.cc:
Auto merged
sql/item_cmpfunc.h:
Auto merged
sql/sql_class.h:
Auto merged
sql/sql_insert.cc:
Auto merged
sql/sql_parse.cc:
Auto merged
sql/sql_update.cc:
Auto merged
mysql-test/r/ctype_ucs.result:
After merge fix
mysql-test/t/ctype_ucs.test:
After merge fix
Query_log_event::error_code
A query can perform completely having the local var error of mysql_$query
zero, where $query in insert, update, delete, load,
and be binlogged with error_code e.g KILLED_QUERY while there is no
reason do to so.
That can happen because Query_log_event consults thd->killed flag to
evaluate error_code.
Fixed with implementing a scheme suggested and partly implemented at
time of bug@22725 work-on. error_status is cached immediatly after the
control leaves the main rows-loop and that instance always corresponds
to `error' the local of mysql_$query functions. The cached value
is passed to Query_log_event constructor, not the default thd->killed
which can be changed in between of the caching and the constructing.
mysql-test/r/binlog_killed.result:
results changed
mysql-test/t/binlog_killed.test:
Demonstrating that effective killing during rows-loop execution leads to the speficied actions:
binlogging with the error for a query modified a not-transactional table or
rolling back effects for transactional table;
fixing possible non-determinism with ID when query_log_enabled;
leave commented out tests for multi-update,delete due to another bug;
removing an obsolete tests template;
changing system rm to --remove_file.
sql/log_event.cc:
adding killed status arg
sql/log_event.h:
added killed status arg
sql/sql_delete.cc:
deploying the update part patch for delete, multi-delete
sql/sql_insert.cc:
deploying the update-part patch for insert..select
sql/sql_load.cc:
deploying the update-part patch for load data.
simulation added.
sql/sql_update.cc:
Impementing the fix as described in the comments left by bug@22725.
Also simulation of killing after the loop that would affect binlogging in the old code.
mysql-test/t/binlog_killed_bug27571-master.opt:
post rows-loop killing simulation's options
mysql-test/t/binlog_killed_bug27571.test:
Checking that if killing happens inbetween of the end of rows loop and
recording into binlog that will not lead to recording any error incl
the killed error.
mysql-test/t/binlog_killed_simulate-master.opt:
simulation options
mysql-test/t/binlog_killed_simulate.test:
tests for
a query (update is choosen) being killed after the row-loop;
load data killed within the loop - effective killed error in the event is gained.
When expanding a * in a USING/NATURAL join the check for table access
for both tables in the join was done using the grant information of the
first one.
Fixed by getting the grant information for the current table while
iterating through the columns of the join.
mysql-test/r/grant2.result:
Bug #30468: test case
mysql-test/t/grant2.test:
Bug #30468: test case
sql/sql_acl.cc:
Bug #30468: correctly check column grants
sql/sql_acl.h:
Bug #30468: correctly check column grants
sql/sql_base.cc:
Bug #30468: correctly check column grants
sql/sql_insert.cc:
Bug #30468: correctly check column grants
type of the result.
There are several functions that accept parameters of different types.
The result field type of such functions was determined based on
the aggregated result type of its arguments. As the DATE and the DATETIME
types are represented by the STRING type, the result field type
of the affected functions was always STRING for DATE/DATETIME arguments.
The affected functions are COALESCE, IF, IFNULL, CASE, LEAST/GREATEST, CASE.
Now the affected functions aggregate the field types of their arguments rather
than their result types and return the result of aggregation as their result
field type.
The cached_field_type member variable is added to the number of classes to
hold the aggregated result field type.
The str_to_date() function's result field type now defaults to the
MYSQL_TYPE_DATETIME.
The agg_field_type() function is added. It aggregates field types with help
of the Field::field_type_merge() function.
The create_table_from_items() function now uses the
item->tmp_table_field_from_field_type() function to get the proper field
when the item is a function with a STRING result type.
libmysql/libmysql.c:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The client library now accepts the NEWDATE type as a string.
sql/item_cmpfunc.cc:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
Now the fix_length_and_dec functions of the
Item_func_ifnull,Item_func_if, Item_func_case, Item_func_coalesce
classes are use agg_field_type function to find out the type of their
result field.
sql/item_cmpfunc.h:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The cached_field type is added to the Item_func_case, Item_func_if and
Item_func_coalesce classes.
The field_type function is added to the Item_func_ifnull, Item_func_if,
Item_func_coalesce, Item_func_case classes.
sql/item_func.cc:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The Item_func_min_max::fix_length_and_dec function now uses the agg_field_type
function to find out the correct type of the result field.
sql/item_func.h:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The cached_field_type variable is added to the Item_func_min_max class.
sql/item_timefunc.cc:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The result of the str_to_date function now defaults to the
DATETIME type.
sql/mysql_priv.h:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The prototype of the agg_field_type function is added.
sql/protocol.cc:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The Protocol_simple::store function is now used to store fields of NEWDATE type.
sql/sql_insert.cc:
Bug#27216: functions with parameters of different date types may return wrong
type of the result.
The create_table_from_items function now uses the
tmp_table_field_from_field_type function to get field for items with the
STRING result type.
mysql-test/r/date_formats.result:
A test case result corrected after fixing bug#27216.
mysql-test/r/type_datetime.result:
Added a test case for the bug#27216: functions with parameters of different
date types may return wrong type of the result.
mysql-test/t/type_datetime.test:
Added a test case for the bug#27216: functions with parameters of different
date types may return wrong type of the result.
led to creating corrupted index.
While execution of the CREATE .. SELECT SQL_BUFFER_RESULT statement the
engine->start_bulk_insert function was called twice. On the first call
On the first call MyISAM disabled all non-unique indexes and on the second
call it decides to not re-enable them because all indexes was disabled.
Due to this no indexes was actually created during CREATE TABLE thus
producing crashed table.
Now the select_inset class has is_bulk_insert_mode flag which prevents
calling the start_bulk_insert function twice.
The flag is set in the select_create::prepare, select_insert::prepare2
functions and the select_insert class constructor.
The flag is reset in the select_insert::send_eof function.
mysql-test/t/insert_select.test:
A test case is added for the bug#30384: Having SQL_BUFFER_RESULT option in the
CREATE .. KEY(..) .. SELECT led to creating corrupted index.
mysql-test/r/insert_select.result:
A test case is added for the bug#30384: Having SQL_BUFFER_RESULT option in the
CREATE .. KEY(..) .. SELECT led to creating corrupted index.
sql/sql_class.h:
Bug#30384: Having SQL_BUFFER_RESULT option in the CREATE .. KEY(..) .. SELECT
led to creating corrupted index.
The is_bulk_insert_mode flag is added to the select_insert class.
sql/sql_insert.cc:
Bug#30384: Having SQL_BUFFER_RESULT option in the CREATE .. KEY(..) .. SELECT
led to creating corrupted index.
The is_bulk_insert_mode is set in the select_create::prepare, select_insert::prepare2
functions and the select_insert class constructor.
The flag is reset in the select_insert::send_eof function.
The problem is that a SELECT on one thread is blocked by INSERT ... ON
DUPLICATE KEY UPDATE on another thread even when low_priority_updates is
activated.
The solution is to possibly downgrade the lock type to the setting of
low_priority_updates if the INSERT cannot be concurrent.
sql/sql_insert.cc:
Possibly downgrade lock type to the the setting of low_priority_updates if
if the INSERT cannot be concurrent.
Binlogging of the statement with a side effect like a modified non-trans table did not happen.
The artifact involved all binloggable dml queries.
Fixed with changing the binlogging conditions all over the code to exploit thd->transaction.stmt.modified_non_trans_table
introduced by the patch for bug@27417.
Multi-delete case has own specific addressed by another bug@29136. Multi-update case has been addressed by bug#27716 and
patch and will need merging.
mysql-test/r/mix_innodb_myisam_binlog.result:
results changed
mysql-test/r/sp_trans_log.result:
results changed
mysql-test/t/mix_innodb_myisam_binlog.test:
specific to the bug tests added
mysql-test/t/sp_trans_log.test:
refining of the proof of that there is an event in binlog
sql/sql_delete.cc:
deploying the binlogging check with thd->transaction.stmt.modified_non_trans_table
sql/sql_insert.cc:
binlogging when thd->transaction.stmt.modified_non_trans_table is TRUE. Merge with Bug#29571.
sql/sql_load.cc:
binlogging when thd->transaction.stmt.modified_non_trans_table is true
sql/sql_update.cc:
binlogging when thd->transaction.stmt.modified_non_trans_table is true
INSERT DELAYED on a replication slave was converted to regular INSERT,
whereas it should try concurrent INSERT first.
With this patch we try to convert delayed insert to concurrent insert on
a replication slave. If it is impossible for some reason, we fall back to
regular insert.
No test case for this fix. I do not see anything indicating this is
regression - we behave this way since Nov 2000.
sql/sql_insert.cc:
If we're executing INSERT DELAYED on a replication slave, we're upgrading
lock type to TL_WRITE as we need to ensure serial execution of queries on
the slave.
OTOH if we're executing regular INSERT on a replication slave, we're
trying TL_WRITE_CONCURRENT_INSERT first, and if we may not use it, we
fall back to TL_WRITE.
Fixed INSERT DELAYED on a replication slave to behave the same way as
regular INSERT, that is to try TL_WRITE_CONCURRENT_INSERT first.
Bug #27417 thd->no_trans_update.stmt lost value inside of SF-exec-stack
Once had been set the flag might later got reset inside of a stored routine
execution stack.
The reason was in that there was no check if a new statement started at time
of resetting.
The artifact affects most of binlogable DML queries. Notice, that multi-update
is wrapped up within
bug@27716 fix, multi-delete bug@29136.
Fixed with saving parent's statement flag of whether the statement modified
non-transactional table, and unioning (merging) the value with that was gained
in mysql_execute_command.
Resettling thd->no_trans_update members into thd->transaction.`member`;
Asserting code;
Effectively the following properties are held.
1. At the end of a substatement thd->transaction.stmt.modified_non_trans_table
reflects the fact if such a table got modified by the substatement.
That also respects THD::really_abort_on_warnin() requirements.
2. Eventually thd->transaction.stmt.modified_non_trans_table will be computed as
the union of the values of all invoked sub-statements.
That fixes this bug#27417;
Computing of thd->transaction.all.modified_non_trans_table is refined to base to
the stmt's value for all the case including insert .. select statement which
before the patch had an extra issue bug@28960.
Minor issues are covered with mysql_load, mysql_delete, and binloggin of insert in
to temp_table select.
The supplied test verifies limitely, mostly asserts. The ultimate testing is defered
for bug@13270, bug@23333.
mysql-test/r/mix_innodb_myisam_binlog.result:
results changed
mysql-test/t/mix_innodb_myisam_binlog.test:
regression test incl the related bug#28960.
sql/ha_ndbcluster.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/handler.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/handler.h:
new member added
sql/log.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/set_var.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/sp_head.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
and saving and merging stmt's flag at the end of a substatement.
sql/sql_class.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/sql_class.h:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/sql_delete.cc:
correcting basic delete incl truncate branch and multi-delete queries to set
stmt.modified_non_trans_table;
optimization to set the flag at the end of per-row loop;
multi-delete still has an extra issue similar to bug#27716 of multi-update
- to be address with bug_29136 fix.
sql/sql_insert.cc:
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/sql_load.cc:
eliminating a separate issue where the stmt flag was saved and re-stored after
write_record that actually could change it and the change would be lost but
should remain permanent;
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
sql/sql_parse.cc:
initialization to transaction.stmt.modified_non_trans_table at the common part
of all types of statements processing - mysql_execute_command().
sql/sql_table.cc:
moving the reset up to the mysql_execute_command() caller
sql/sql_update.cc:
correcting update query case (multi-update part of the issues covered by other
bug#27716 fix)
thd->transaction.{all,stmt}.modified_non_trans_table
instead of
thd->no_trans_update.{all,stmt}
the master but on the slave
MySQL can decide to "downgrade" a INSERT DELAYED statement
to normal insert in certain situations.
One such situation is when the slave is replaying a
replication feed.
However INSERT DELAYED is logged even if there're no updates
whereas the NORMAL INSERT is not logged in such cases.
Fixed by always logging a "downgraded" INSERT DELAYED: even
if there were no updates.
mysql-test/r/rpl_insert_delayed.result:
Bug #29571: test case
mysql-test/t/rpl_insert_delayed.test:
Bug #29571: test case
sql/sql_insert.cc:
Bug #29571: log INSERT DELAYED even if it was
"downgraded" to INSERT (and there were no updates)
No test case, since the bug requires a stress case with 30 INSERT DELAYED
threads and 1 killer thread to repeat. The patch is verified
manually.
Review fixes.
The server that is running DELAYED inserts would deadlock itself
or crash under high load if some of the delayed threads were KILLed
in the meanwhile.
The fix is to change internal lock acquisition order of delayed inserts
subsystem and to ensure that
Delayed_insert::table_list::db does not point to volatile memory in some
cases.
For details, please see a comment for sql_insert.cc.
sql/sql_insert.cc:
A fix for Bug#29431 killing an insert delayed thread causes crash
1) The deadlock was caused by different lock acquisition order
between delayed_get_table and handle_delayed_insert.
delayed_get_table would:
- acquire LOCK_delayed_create
- create a new Delayed_insert instance
- acquire instance mutex (di->mutex)
- "lock the instance in memory" by
increasing di->locked_in_memory variable
- start the delayed thread
- release di->mutex
- let the delayed thread open the delayed table
- discover that the delayed thread was killed
- try to unlock() the delayed insert instance in memory
- in Delayed_insert::unlock() do
* lock LOCK_delayed_insert
* decrease locks_in_memory and discover it's 0
* attempt to lock di->mutex to broadcast di->cond condition <-- deadlock
here
Meanwhile, the delayed thread would
* lock di->mutex
* open the table
* get killed
* not notice that and attempt to lock LOCK_delayed_insert
to register itself in the delayed insert list <-- deadlock here.
Simply put, delayed_get_table used to lock LOCK_delayed_insert and then
di->mutex, and handle_delayed_insert would lock di->mutex and then
LOCK_delayed_insert.
Fixed by moving registration in the list of delayed insert threads from
handle_delayed_insert to delayed_get_table - so that now
handle_delayed_insert doesn't need to acquire LOCK_delayed_insert mutex.
2) di->table_list.db was copied by-pointer from table_list.db of the first
producer -- the one who initiated creation of the delayed insert thread.
This producer might be long gone when the member is needed
(handle_delayed_insert:open_ltable),
e.g. by having been killed with KILL CONNECTION statement.
Fixed by using a pointer to the consumer's deep copy of THD::db.
3) In find_handler, we shouldn't assume that Delayed_insert object
already (or still) has a table opened all the time it is
present in the delayed insert list. E.g. it's not the case
when Delayed_insert decided to terminate, closed the table, but haven't
yet unregistered from the list (see the end of handle_delayed_insert).
"Federated INSERT failures"
Federated does not correctly handle "INSERT...ON DUPLICATE KEY UPDATE"
However, implementing such support is not reasonably possible without
increasing complexity of the storage engine: checking that constraints
on remote server match local server and parsing error messages.
This patch causes 'ON DUPLICATE KEY' to fail with ER_DUP_KEY message
if a conflict occurs and not to fail silently.
include/my_base.h:
bug25511
new storage engine hint: HA_EXTRA_INSERT_WITH_UPDATE
mysql-test/r/federated.result:
test for bug25511
mysql-test/t/federated.test:
test for bug25511
sql/ha_federated.cc:
bug25511
implement support for handling HA_EXTRA_INSERT_WITH_UPDATE hint
sql/ha_federated.h:
bug25511
new property: insert_dup_update
sql/sql_insert.cc:
bug25511
implement support for HA_EXTRA_INSERT_WITH_UPDATE
When checking duplicates flag, if it is DUP_UPDATE, send hint
to the storage engine.
The method select_insert::send_error does two things, it rolls back a statement
being executed and outputs an error message. But when a
nonexistent column is referenced, an error message has been published already and
there is no need to publish another.
Fixed by moving all functionality beyond publishing an error message into
select_insert::abort() and calling only that function.
mysql-test/r/errors.result:
Bug#28677: test result
mysql-test/t/errors.test:
Bug#28677: test case
sql/sql_class.h:
Bug#28677: overriding abort()
sql/sql_insert.cc:
Bug#28677:
- moved everything beyond producing an error message out of select_insert::send_error
and into new override select_insert::abort()
- made corresponding move of code from select_create::send_error to select_create::abort
sql/sql_select.cc:
Bug#28677: No need to pusblish an error here
When the INSERT .. ON DUPLICATE KEY UPDATE has to update a matched row but
the new data is the same as in the record then it returns as if
no rows were inserted or updated. Nevertheless the row is silently
updated. This leads to a situation when zero updated rows are reported
in the case when data has actually been changed.
Now the write_record function updates a row only if new data differs from
that in the record.
sql/sql_insert.cc:
Bug#28904: INSERT .. ON DUPLICATE was silently updating rows when it shouldn't.
Now the write_record function updates a row only if new data differs from
that in the record.
mysql-test/r/insert_update.result:
Added a test case for the bug#28904: INSERT .. ON DUPLICATE was silently
updating rows when it shouldn't.
mysql-test/t/insert_update.test:
Added a test case for the bug#28904: INSERT .. ON DUPLICATE was silently
updating rows when it shouldn't.
flag is set.
When the CLIENT_FOUND_ROWS flag is set then the server should return
found number of rows independently whether they were updated or not.
But this wasn't the case for the INSERT statement which always returned
number of rows that were actually changed thus providing wrong info to
the user.
Now the select_insert::send_eof method and the mysql_insert function
are sending the number of touched rows if the CLIENT_FOUND_ROWS flag is set.
tests/mysql_client_test.c:
Added a test case for the bug#28505: mysql_affected_rows() may return wrong result
if CLIENT_FOUND_ROWS flag is set.
sql/sql_insert.cc:
Bug#28505: mysql_affected_rows() may return wrong result if CLIENT_FOUND_ROWS
flag is set.
Now the select_insert::send_eof method and the mysql_insert function
are sending the number of touched rows if the CLIENT_FOUND_ROWS flag is set.
Refining the tests since pb revealed the older version's fragality - the error from SF() due to killed
may be different on different env:s.
DBUG_ASSERT instead of assert.
mysql-test/r/binlog_killed.result:
new result file
mysql-test/t/binlog_killed.test:
regression for bug#22725 simplified. tests for bug27563, BUG#27565 made inactive.
sql/sql_insert.cc:
DBUG_ASSERT
The reason for the bug was that replaying of a query on slave could not be possible since its event
was recorded with the killed error. Due to the specific of handling INSERT, which per-row-while-loop is
unbreakable to killing, the query on transactional table should have not appeared in binlog unless
there was a call to a stored routine that got interrupted with killing (and then there must be an error
returned out of the loop).
The offered solution added the following rule for binlogging of INSERT that accounts the above
specifics:
For INSERT on transactional-table if the error was not set the only raised flag
is harmless and is ignored via masking out on time of creation of binlog event.
For both table types the combination of raised error and KILLED flag indicates that there
was potentially partial execution on master and consistency is under the question.
In that case the code continues to binlog an event with an appropriate killed error.
The fix relies on the specified behaviour of stored routine that must propagate the error
to the top level query handling if the thd->killed flag was raised in the routine execution.
The patch adds an arg with the default killed-status-unset value to Query_log_event::Query_log_event.
sql/log_event.cc:
killed_status as the value of thd->killed can be passed as an arg to the constructor.
if the value is different from the default the arg is set to the current thd->killed value.
A caller might need to masquerade thd->killed with THD::NOT_KILLED.
So far only mysql_insert() uses such explicit way to tell the constructor about killing status.
sql/log_event.h:
default arg to the constructor with meaning of killed status of the query.
if the arg is not explicitly provided the status of thd->killed will be snapshot
inside of the constuctor, which is potentially incorrect (see bug#27571)
sql/sql_class.h:
extending killed_state with no-state member.
sql/sql_insert.cc:
ignore the KILLED flag incl KILL_BAD_DATA when the INSERT query event
is created without an `error';
sql/sql_update.cc:
Suggestion how to fix bug#27571 as comments.
mysql-test/r/binlog_killed.result:
new result file
mysql-test/t/binlog_killed.test:
regression tests also apply for bug27563, BUG#27565
into vajra.(none):/opt/local/work/mysql-5.0-21483
sql/sp_head.cc:
Auto merged
sql/sql_base.cc:
Auto merged
sql/sql_insert.cc:
Auto merged
sql/sql_lex.h:
Auto merged
Bug#21483 "Server abort or deadlock on INSERT DELAYED with another
implicit insert"
Also fixes and adds test cases for bugs:
20497 "Trigger with INSERT DELAYED causes Error 1165"
21714 "Wrong NEW.value and server abort on INSERT DELAYED to a
table with a trigger".
Post-review fixes.
Problem:
In MySQL INSERT DELAYED is a way to pipe all inserts into a
given table through a dedicated thread. This is necessary for
simplistic storage engines like MyISAM, which do not have internal
concurrency control or threading and thus can not
achieve efficient INSERT throughput without support from SQL layer.
DELAYED INSERT works as follows:
For every distinct table, which can accept DELAYED inserts and has
pending data to insert, a dedicated thread is created to write data
to disk. All user connection threads that attempt to
delayed-insert into this table interact with the dedicated thread in
producer/consumer fashion: all records to-be inserted are pushed
into a queue of the dedicated thread, which fetches the records and
writes them.
In this design, client connection threads never open or lock
the delayed insert table.
This functionality was introduced in version 3.23 and does not take
into account existence of triggers, views, or pre-locking.
E.g. if INSERT DELAYED is called from a stored function, which,
in turn, is called from another stored function that uses the delayed
table, a deadlock can occur, because delayed locking by-passes
pre-locking. Besides:
* the delayed thread works directly with the subject table through
the storage engine API and does not invoke triggers
* even if it was patched to invoke triggers, if triggers,
in turn, used other tables, the delayed thread would
have to open and lock involved tables (use pre-locking).
* even if it was patched to use pre-locking, without deadlock
detection the delayed thread could easily lock out user
connection threads in case when the same table is used both
in a trigger and on the right side of the insert query:
the delayed thread would not release locks until all inserts
are complete, and user connection can not complete inserts
without having locks on the tables used on the right side of the
query.
Solution:
These considerations suggest two general alternatives for the
future of INSERT DELAYED:
* it is considered a full-fledged alternative to normal INSERT
* it is regarded as an optimisation that is only relevant
for simplistic engines.
Since we missed our chance to provide complete support of new
features when 5.0 was in development, the first alternative
currently renders infeasible.
However, even the second alternative, which is to detect
new features and convert DELAYED insert into a normal insert,
is not easy to implement.
The catch-22 is that we don't know if the subject table has triggers
or is a view before we open it, and we only open it in the
delayed thread. We don't know if the query involves pre-locking
until we have opened all tables, and we always first create
the delayed thread, and only then open the remaining tables.
This patch detects the problematic scenarios and converts
DELAYED INSERT to a normal INSERT using the following approach:
* if the statement is executed under pre-locking (e.g. from
within a stored function or trigger) or the right
side may require pre-locking, we detect the situation
before creating a delayed insert thread and convert the statement
to a conventional INSERT.
* if the subject table is a view or has triggers, we shutdown
the delayed thread and convert the statement to a conventional
INSERT.
mysql-test/r/insert.result:
Update test results.
mysql-test/t/insert.test:
Add a test case for Bug#21483, Bug#20497, Bug#21714 (INSERT DELAYED
and stored routines, triggers).
sql/sp_head.cc:
Upgrade lock type to TL_WRITE when computing the pre-locking set.
sql/sql_base.cc:
Use a new method.
sql/sql_insert.cc:
INSERT DELAYED and pre-locking:
- if under pre-locking, upgrade the lock type to TL_WRITE
and proceed as a normal write
- if DELAYED table has triggers, also request a lock upgrade.
- make sure errors in the delayed thread are propagated
correctly
sql/sql_lex.h:
Add a method to check if a parsed tree refers to stored
routines.
into vajra.(none):/opt/local/work/mysql-5.0-runtime
sql/item.cc:
Auto merged
sql/item_func.cc:
Auto merged
sql/mysql_priv.h:
Auto merged
sql/set_var.cc:
Auto merged
sql/sql_class.h:
Auto merged
sql/sql_insert.cc:
Auto merged
sql/sql_parse.cc:
Auto merged
sql/sql_prepare.cc:
Auto merged
sql/sql_yacc.yy:
Auto merged
into mockturtle.local:/home/dlenev/src/mysql-5.0-cts-3
sql/mysql_priv.h:
Auto merged
sql/sql_insert.cc:
Auto merged
sql/sql_prepare.cc:
Auto merged
sql/sql_yacc.yy:
Auto merged
