mariadb/mysql-test/suite/innodb/r/cursor-restore-locking.result
Vlad Lesin f5373db898 MDEV-33004 innodb.cursor-restore-locking test fails
THE FIX MUST NOT BE MERGED TO 10.6+, BECAUSE 10.6+ IS NOT AFFECTED!

The test is waiting for delete-marked record purging. But this does not
happen under the following conditions:

1. "START TRANSACTION WITH CONSISTENT SNAPSHOT" - is active, has not
been rolled back yet
2. "DELETE FROM t WHERE b = 20 # trx_1" - is committed
3. "INSERT INTO t VALUES(10, 20) # trx_2" - hanging on
"ib_after_row_insert" sync point, waiting for "first_ins_cont" signal
4. "DELETE FROM t WHERE b = 20 # trx_3" - blocked on delete-marked by
trx_1 record, waiting for trx_2
5. connection "default" is waiting on
'now WAIT_FOR row_purge_del_mark_finished'

purge_coordinator_callback_low() sets

purge_state.m_history_length= srv_do_purge(&n_total_purged);

even if nothing was purged, like in our case. Nothing was purged because
transaction with consistent snapshot was still alive during purging
procedure.

Then purge_coordinator_timer_callback() does not wake purge thread if
the following condition is true:

purge_state.m_history_length == trx_sys.rseg_history_len

The above condition is true for our case, because we are waiting for
delete-marked record purging, and trx_sys.rseg_history_len does not
grow.

Only 10.5 is affected, because there is no such condition in 10.6, i.e.
purge thread is woken up even if history size was not changed during
purge coordinator thread suspending.

The easiest way to fix it is just to remove the test from 10.5.
2024-02-07 12:35:18 +02:00

40 lines
1.6 KiB
Text

CREATE TABLE t (a int PRIMARY KEY, b int NOT NULL UNIQUE) engine = InnoDB, STATS_PERSISTENT=0;
InnoDB 0 transactions not purged
connect prevent_purge,localhost,root,,;
start transaction with consistent snapshot;
connect con_del_1,localhost,root,,;
INSERT INTO t VALUES (20,20);
SET DEBUG_SYNC = 'innodb_row_search_for_mysql_exit SIGNAL first_del_row_search_mvcc_finished WAIT_FOR first_del_cont';
DELETE FROM t WHERE b = 20 # trx_1;
connect con_ins_1,localhost,root,,;
SET DEBUG_SYNC = 'now WAIT_FOR first_del_row_search_mvcc_finished';
SET DEBUG_SYNC = 'lock_wait_start SIGNAL first_ins_locked';
SET DEBUG_SYNC = 'ib_after_row_insert SIGNAL first_ins_row_inserted WAIT_FOR first_ins_cont';
INSERT INTO t VALUES(10, 20) # trx_2;
connect con_del_2,localhost,root,,;
SET TRANSACTION ISOLATION LEVEL READ COMMITTED;
SET DEBUG_SYNC = 'now WAIT_FOR first_ins_locked';
SET DEBUG_SYNC = 'lock_wait_start SIGNAL second_del_locked';
DELETE FROM t WHERE b = 20 # trx_3;
connection default;
SET DEBUG_SYNC = 'now WAIT_FOR second_del_locked';
SET @saved_dbug = @@GLOBAL.debug_dbug;
SET @@GLOBAL.debug_dbug="d,enable_row_purge_del_mark_exit_sync_point";
SET DEBUG_SYNC = 'now SIGNAL first_del_cont';
SET DEBUG_SYNC = 'now WAIT_FOR first_ins_row_inserted';
connection con_del_1;
connection default;
disconnect prevent_purge;
SET DEBUG_SYNC = 'now WAIT_FOR row_purge_del_mark_finished';
SET @@GLOBAL.debug_dbug = @saved_dbug;
SET DEBUG_SYNC = 'now SIGNAL first_ins_cont';
connection con_del_2;
connection con_ins_1;
connection default;
INSERT INTO t VALUES(30, 20);
disconnect con_ins_1;
disconnect con_del_1;
disconnect con_del_2;
connection default;
SET DEBUG_SYNC = 'RESET';
DROP TABLE t;