MySQL:68311: InnoDB: Failing assertion thr in row0upd.cc on concurrent ALTER, UPDATE, HANDLER

Description and RQG test are attached

They say it's a duplicate of internal bug#16063835.
So, duplicate or nor, if/when the fix appears without a test case, we can create one based on the provided stress test and the fix code, if we want to.

Update:

Bugfix in 5.6.11:

revno: 4672

revision-id: marko.makela@oracle.com-20130124151343-slg3a69eok4qu4r6

parent: jon.hauglid@oracle.com-20130124135614-vvkd88u27ntpshx2

committer: Marko Mäkelä <marko.makela@oracle.com>

branch nick: mysql-5.6

timestamp: Thu 2013-01-24 17:13:43 +0200

message:

  Bug#16063835 ASSERT THR ROW_UPD_CHANGES_ORD_FIELD_BINARY_FUNC,

  ROW_LOG_TABLE_APPLY_UPDATE

  btr_cur_pessimistic_update(), btr_cur_optimistic_update(),

  btr_cur_update_in_place(): Require thr!=NULL. Allow

  thr->graph->trx->id != trx_id during online table rebuild, as neither

  trx_id nor thr->graph->trx->id will be written to the record due to

  BTR_KEEP_SYS_FLAG.

  row_log_table_apply_update(): Pass thr to btr_cur_pessimistic_insert()

  in order to avoid the debug assertion failure.

  rb#1892

Came without a test case.

Cannot reproduce on the current 10.0 tree and on 10.0.10, apparently fixed by 5.6.14 or 5.6.15 merge

Bug #68694 - Server crashes in ptr_compare on NOW and CAST in ORDER BY
See also MDEV-4285

Reproducible on 5.6.10

Test case:

CREATE TABLE t1 (pk INT PRIMARY KEY) ENGINE=InnoDB;
INSERT INTO t1 VALUES (1),(2),(3),(4),(5),(6),(7),(8),(9),(10);
SELECT SQL_NO_CACHE * FROM t1 ORDER BY NOW(), CAST( pk AS CHAR(0) );

Update

Still not fixed in 10.0.14, raised the priority

Bug #68695 - Assertion `table_share->tmp_table != NO_TMP_TABLE || m_lock_type == 1' fails

Reproducible on 5.6.10-debug

Test case:
CREATE TABLE tmp SELECT 1 FROM DUAL WHERE EXP( 2000 );

Bug #68705 - Valgrind warnings (uninitialised value(s)) in Item_field::val on SELECT INTERVAL

Test case:

CREATE TABLE t1 (a INT);
SELECT SQL_NO_CACHE INTERVAL( 1, @var := a, 2, 3, 4, 5, 6, 7, 8 ) FROM t1;

http://bugs.mysql.com/bug.php?id=68759

DROP TABLE IF EXISTS t_inno;
CREATE TABLE t_inno (pk INT PRIMARY KEY, t_key TIME, t TIME, KEY(t_key)) ENGINE=InnoDB;
INSERT INTO t_inno VALUES (1,'12:12:12','12:12:12'),(2,'21:21:21','21:21:21');

SELECT * FROM t_inno WHERE t < 'z';
SELECT pk FROM t_inno WHERE t_key < 'z';
SELECT * FROM t_inno WHERE t_key < 'z';

MySQL [test]> DROP TABLE IF EXISTS t_inno;
Query OK, 0 rows affected (0.20 sec)

MySQL [test]> CREATE TABLE t_inno (pk INT PRIMARY KEY, t_key TIME, t TIME, KEY(t_key)) ENGINE=InnoDB;
Query OK, 0 rows affected (1.20 sec)

MySQL [test]> INSERT INTO t_inno VALUES (1,'12:12:12','12:12:12'),(2,'21:21:21','21:21:21');
Query OK, 2 rows affected (0.07 sec)
Records: 2 Duplicates: 0 Warnings: 0

MySQL [test]>
MySQL [test]> SELECT * FROM t_inno WHERE t < 'z';
--------------------

--------------------

--------------------
2 rows in set, 1 warning (0.00 sec)

MySQL [test]> SELECT pk FROM t_inno WHERE t_key < 'z';
----

----

----
2 rows in set, 2 warnings (0.00 sec)

MySQL [test]> SELECT * FROM t_inno WHERE t_key < 'z';
Empty set, 2 warnings (0.00 sec)

Possibly file later

MySQL [test]> select @@sql_mode;
------------------------

------------------------

------------------------
1 row in set (0.00 sec)

MySQL [test]> show create table t2;
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)

MySQL [test]> select * from t2;
--------------------

--------------------

--------------------
2 rows in set (0.00 sec)

MySQL [test]> update t2 set i = u-1;
ERROR 1690 (22003): BIGINT UNSIGNED value is out of range in '(`test`.`t2`.`u` - 1)'
MySQL [test]>

Filed bug http://bugs.mysql.com/bug.php?id=69203 (Valgrind warnings and crash on select from an InnoDB table), for 5.7 revno 5216.
Test case:
CREATE TABLE t1 (pk INT PRIMARY KEY) ENGINE=InnoDB;
INSERT INTO t1 VALUES (10),(11);

SELECT * FROM t1;

1. End of test case

When run with --valgrind-mysqld under MTR, causes a valgrind warnings and a crash on BUILD/compile-pentium-debug-max-no-ndb

MySQL (and MariaDB) prints garbage in warning records:

2013-08-09 23:46:52 7f4dafeec740 InnoDB: Warning: Using innodb_additional_mem_pool_size is DEPRECATED. This option may be removed in future releases, together with the option innodb_use_sys_malloc and with the InnoDB's internal memory allocator.
2013-08-09 23:46:52 7f4dafeec740 InnoDB: Warning: Using innodb_locks_unsafe_for_binlog is DEPRECATED. This option may be removed in future releases. Please use READ COMMITTED transaction isolation level instead, see http://dev.mysql.com/doc/refman/10.0/en/set-transaction.html.

I think there was an upstream bug about it, but can't find it now, need to check later.

https://github.com/MySQLOnRocksDB/mysql-5.6/issues/70