Changes in MySQL 5.6.2 (2011-04-11, Developer Milestone)

Note

This is a milestone release, for use at your own risk. Upgrades between milestone releases (or from a milestone release to a GA release) are not supported. Significant development changes take place in milestone releases and you may encounter compatibility issues, such as data format changes that require attention in addition to the usual procedure of running mysql_upgrade. For example, you may find it necessary to dump your data with mysqldump before the upgrade and reload it afterward. (Making a backup before the upgrade is a prudent precaution in any case.)

• INFORMATION_SCHEMA Tables for InnoDB Buffer Pool Information

• INFORMATION_SCHEMA Table for InnoDB Metrics

• INFORMATION_SCHEMA Tables for InnoDB Data Dictionary

• Persistent InnoDB Optimizer Statistics

• InnoDB Configurable Data Dictionary Cache

• Optimizer Notes

• Explicit Partition Selection

• Performance Schema Notes

• Pluggable Authentication

• Plugin Notes

• Crash-Safe Binary Log

• Slave Log Tables

• Row Image Control

• Functionality Added or Changed

• Bugs Fixed

INFORMATION_SCHEMA Tables for InnoDB Buffer Pool Information

• InnoDB: The new INFORMATION_SCHEMA tables INNODB_BUFFER_PAGE, INNODB_BUFFER_PAGE_LRU, and INNODB_BUFFER_POOL_STATS display InnoDB buffer pool information for tuning on large-memory or highly loaded systems.

INFORMATION_SCHEMA Table for InnoDB Metrics

• InnoDB: A new INFORMATION_SCHEMA table, INNODB_METRICS, lets you query low-level InnoDB performance information, getting cumulative counts, averages, and minimum/maximum values for internal aspects of the storage engine operation. You can start, stop, and reset the metrics counters using the innodb_monitor_enable, innodb_monitor_disable, innodb_monitor_reset, and innodb_monitor_reset_all system variables.

INFORMATION_SCHEMA Tables for InnoDB Data Dictionary

• InnoDB: The InnoDB data dictionary, containing metadata about InnoDB tables, columns, indexes, and foreign keys, is available for SQL queries through a set of INFORMATION_SCHEMA tables.

Persistent InnoDB Optimizer Statistics

• Performance; InnoDB: The optimizer statistics for InnoDB tables can now persist across server restarts, producing more stable query performance. You can also control the amount of sampling done to estimate cardinality for each index, resulting in more accurate optimizer statistics. This feature involves the configuration options innodb_analyze_is_persistent (later replaced by innodb_stats_persistent), innodb_stats_persistent_sample_pages, and innodb_stats_transient_sample_pages, and the ANALYZE TABLE statement. See Configuring Persistent Optimizer Statistics Parameters for details.

InnoDB Configurable Data Dictionary Cache

• InnoDB: To ease the memory load on systems with huge numbers of tables, InnoDB now frees up the memory associated with an opened table using an LRU algorithm to select tables that have gone the longest without being accessed. To reserve more memory to hold metadata for open InnoDB tables, increase the value of the table_definition_cache configuration option. InnoDB treats this value as a “soft limit” for the number of open table instances in the InnoDB data dictionary cache. The actual number of tables with cached metadata could be higher than the value specified for table_definition_cache, because metadata for InnoDB system tables, and parent and child tables in foreign key relationships, is never evicted from memory. For additional information, refer to the table_definition_cache documentation. (Bug #20877, Bug #11745884)

Optimizer Notes

• The optimizer now more efficiently handles queries (and subqueries) of the following form:

  SELECT ...
      FROM single_table ...
      ORDER BY non_index_column [DESC]
      LIMIT [M,]N;

  That type of query is common in web applications that display only a few rows from a larger result set. For example:

  SELECT col1, ... FROM t1 ... ORDER BY name LIMIT 10;
  SELECT col1, ... FROM t1 ... ORDER BY RAND() LIMIT 15;

  The sort buffer has a size of sort_buffer_size. If the sort elements for N rows are small enough to fit in the sort buffer (M+N rows if M was specified), the server can avoid using a merge file and perform the sort entirely in memory. For details, see LIMIT Query Optimization.

• The optimizer implements Disk-Sweep Multi-Range Read. Reading rows using a range scan on a secondary index can result in many random disk accesses to the base table when the table is large and not stored in the storage engine's cache. With the Disk-Sweep Multi-Range Read (MRR) optimization, MySQL tries to reduce the number of random disk access for range scans by first scanning the index only and collecting the keys for the relevant rows. Then the keys are sorted and finally the rows are retrieved from the base table using the order of the primary key. The motivation for Disk-sweep MRR is to reduce the number of random disk accesses and instead achieve a more sequential scan of the base table data. For more information, see Multi-Range Read Optimization.

• The optimizer implements Index Condition Pushdown (ICP), an optimization for the case where MySQL retrieves rows from a table using an index. Without ICP, the storage engine traverses the index to locate rows in the base table and returns them to the MySQL server which evaluates the WHERE condition for the rows. With ICP enabled, and if parts of the WHERE condition can be evaluated by using only fields from the index, the MySQL server pushes this part of the WHERE condition down to the storage engine. The storage engine then evaluates the pushed index condition by using the index entry and only if this is satisfied is base row be read. ICP can reduce the number of accesses the storage engine has to do against the base table and the number of accesses the MySQL server has to do against the storage engine. For more information, see Index Condition Pushdown Optimization.

Explicit Partition Selection

• Partitioning: It is now possible to select one or more partitions or subpartitions when querying a partitioned table. In addition, many data modification statements (DELETE, INSERT, REPLACE, UPDATE, LOAD DATA, and LOAD XML) that act on partitioned tables also now support explicit partition selection. For example, assume we have a table named t with some integer column named c, and t has 4 partitions named p0, p1, p2, and p3. Then the query SELECT * FROM t PARTITION (p0, p1) WHERE c < 5 returns rows only in partitions p0 and p1 that match the WHERE condition, whereas partitions p2 and p3 are not checked.

  For additional information and examples, see Partition Selection, as well as the descriptions of the statements just listed.

Performance Schema Notes

• The Performance Schema has these additions:

  • The Performance Schema now has tables that contain summaries for table and index I/O wait events, as generated by the wait/io/table/sql/handler instrument:

    • table_io_waits_summary_by_table: Aggregates table I/O wait events. The grouping is by table.

    • table_io_waits_summary_by_index_usage: Aggregates table index I/O wait events. The grouping is by table index.

    The information in these tables can be used to assess the impact of table I/O performed by applications. For example, it is possible to see which tables are used and which indexes are used (or not used), or to identify bottlenecks on a table when multiple applications access it. The results may be useful to change how applications issue queries against a database, to minimize application footprint on the server and to improve application performance and scalability.

    A change that accompanies the new tables is that the events_waits_current table now has an INDEX_NAME column to identify which index was used for the operation that generated the event. The same is true of the event-history tables, events_waits_history, and events_waits_history_long.

  • The Performance Schema now has an instrument named wait/lock/table/sql/handler in the setup_instruments table for instrumenting table lock wait events. It differs from wait/io/table/sql/handler, which instruments table I/O. This enables independent instrumentation of table I/O and table locks.

    Accompanying the new instrument, the Performance Schema has a table named table_lock_waits_summary_by_table that aggregates table lock wait events, as generated by the new instrument. The grouping is by table.

    The information in this table may be used to assess the impact of table locking performed by applications. The results may be useful to change how applications issue queries against the database and use table locks, to minimize the application footprint on the server and to improve application performance and scalability. For example, an application locking tables for a long time may negatively affect other applications; the instrumentation makes this visible.

  • To selectively control which tables to instrument for I/O and locking, use the setup_objects table. See Pre-Filtering by Object.

  If you upgrade to this MySQL release from an earlier version, you must run mysql_upgrade (and restart the server) to incorporate these changes into the performance_schema database.

  For more information, see MySQL Performance Schema.

Pluggable Authentication

• MySQL distributions now include auth_socket, a server-side authentication plugin that authenticates clients that connect from the local host through the Unix socket file. The plugin uses the SO_PEERCRED socket option to obtain information about the user running the client program (and thus can be built only on systems that support this option). For a connection to succeed, the plugin requires a match between the login name of the connecting client user and the MySQL user name presented by the client program. For more information, see Socket Peer-Credential Pluggable Authentication. (Bug #59017, Bug #11765993, Bug #9411, Bug #11745104)

• MySQL distributions now include mysql_clear_password, a client-side authentication plugin that sends the password to the server without hashing or encryption. Although this is insecure, and thus appropriate precautions should be taken (such as using an SSL connection), the plugin is useful in conjunction with server-side plugins that must have access to the original password in clear text. For more information, see Client-Side Cleartext Pluggable Authentication.

Plugin Notes

• A new plugin service, my_plugin_log_service, enables plugins to report errors and specify error messages. The server writes the messages to the error log. See MySQL Services for Plugins.

Crash-Safe Binary Log

• Replication: Support for checksums when writing and reading the binary log is added to the MySQL Server. Writing checksums into the binary log is disabled by default; it can be enabled by starting the server with the --binlog-checksum option. To cause the server to read checksums from the binary log, start the server with the --master-verify-checksum option. The --slave-sql-verify-checksum option causes the slave to read checksums from the relay log.

• Replication: The MySQL Server now records and reads back only complete events or transactions to and from the binary log. By default, the server now logs the length of the event as well as the event itself and uses this information to verify that the event was written correctly to the log. A master also uses by default this value to verify events when reading from the binary log.

  If you enable writing of checksums (using the binlog_checksum system variable), the master can use these instead by enabling the master_verify_checksum system variable. The slave I/O thread also verifies events received from the master. You can cause the slave SQL thread to use checksums (if available) as well, when reading from the relay log, by enabling the slave_sql_verify_checksum system variable on the slave.

Slave Log Tables

• Replication: It is now possible to write information about the slave connection to the master and about the slave's execution point within the relay log to tables rather than files. Logging of master connection information and of slave relay log information to tables can be done independently of one another; this is controlled by the --master-info-repository and --relay-log-info-repository server options. When --master-info-repository is set to TABLE, connection information is logged in the slave_master_info table in the mysql system database. When --relay-log-info-repository is set to TABLE, relay log information is logged to the slave_relay_log_info table, also in the mysql database.

Row Image Control

• Replication: Added the binlog_row_image server system variable, which can be used to enable row image control for row-based replication. This means that you can potentially save disk space, network resources, and memory usage by the MySQL Server by logging only those columns that are required for uniquely identifying rows, or which are actually changed on each row, as opposed to logging all columns for each and every row change event. In addition, you can use a “noblob” mode where all columns, except for unneeded BLOB or TEXT columns, are logged.

  For more information, see System Variables Used with Binary Logging. (Bug #47200, Bug #47303, Bug #56917, Bug #11755426, Bug #11755513, Bug #11764116)

Functionality Added or Changed

• Incompatible Change: The following obsolete constructs have been removed. Where alternatives are shown, applications should be updated to use them.

  • The FLUSH MASTER and FLUSH SLAVE statements. Use the RESET MASTER and RESET SLAVE statements instead.

• Important Change; Replication: Added the binlog_rows_query_log_events system variable for mysqld. Enabling this variable causes a server logging in row-based mode to write informational rows query log events (SQL statements, for debugging and other purposes) to the binary log. MySQL server and MySQL programs from MySQL 5.6.2 and later normally ignore such events, so that they do not pose an issue when reading the binary log. mysqld and mysqlbinlog from previous MySQL releases cannot read such events in the binary log, and fail if they attempt to do so. For this reason, you should never prepare logs for a MySQL 5.6.1 or earlier replication slave server (or other reader such as mysqlbinlog) with this option enabled on the master. (Bug #50935, Bug #11758695)

• Performance; InnoDB: A separate InnoDB thread (page_cleaner) now handles the flushing of dirty pages that was formerly done by the InnoDB master thread. (Bug #11762412, Bug #55004)

• Performance; InnoDB: The innodb_purge_threads system variable can now be set to a value higher than 1.

• Performance; InnoDB: The InnoDB kernel mutex, which controls concurrent access to the InnoDB kernel, has been split into several mutexes and rw-locks, for improved concurrency.

• InnoDB: InnoDB can optionally log details about all deadlocks that occur, to assist with troubleshooting and diagnosis. This feature is controlled by the innodb_print_all_deadlocks system variable. (Bug #1784, Bug #17572)

• Replication: On MySQL replication slaves having multiple network interfaces, it is now possible to set which interface to use for connecting to the master using the MASTER_BIND='interface' option in a CHANGE MASTER TO statement.

  The value set by this option can be seen in the Master_Bind column of the output from SHOW SLAVE STATUS or the Bind column of the mysql.slave_master_info table. (Bug #25939, Bug #11746389)

• Replication: Added the log_bin_basename system variable, which contains the complete file name and path to the binary log file. (The log_bin system variable shows only whether or not binary logging is enabled; log_bin_basename, however, reflects the name set with the --log-bin server option.) Also added relay_log_basename system variable, which shows the file name and complete path to the relay log file.

  References: See also: Bug #19614, Bug #11745759.

• The mysql_upgrade, mysqlbinlog, mysqlcheck, mysqlimport, mysqlshow, and mysqlslap clients now have --default-auth and --plugin-dir options for specifying which authentication plugin and plugin directory to use. (Bug #58139)

• The server now includes the thread ID in rows written to the slow query log. In the slow query log file, the thread ID is the last value in the line. In the mysql.slow_log log table, there is a new thread_id column.

  To update the slow_log table if you are upgrading from an earlier release, run mysql_upgrade and restart the server. See mysql_upgrade — Check and Upgrade MySQL Tables. (Bug #53630, Bug #11761166)

• The server now writes thread shutdown messages to the error log during the shutdown procedure. (Bug #48388, Bug #11756464)

• If the init_file system variable is given, the server now writes messages indicating the beginning and end of file execution to the error log. (Bug #48387, Bug #11756463)

• Boolean system variables can be enabled at run time by setting them to the value ON or OFF, but previously this did not work at server startup. Now at startup such variables can be enabled by setting them to ON or TRUE, or disabled by setting them to OFF or FALSE. Any other nonnumeric value is invalid. (Bug #46393)

  References: See also: Bug #11754743, Bug #51631.

• MySQL distributions now include an INFO_SRC file that contains information about the source distribution, such as the MySQL version from which it was created. MySQL binary distributions additionally include an INFO_BIN file that contains information about how the distribution was built, such as compiler options and feature flags. In RPM packages, these files are located in the /usr/share/doc/packages/MySQL-server directory. In tar.gz and derived packages, they are located in the Docs directory under the location where the distribution is unpacked. (Bug #42969, Bug #11751935)

• Multi-read range access is now based on cost estimates and no longer used for simple queries for which it is not beneficial. (Bug #37576, Bug #11748865)

• Previously, for queries that were aborted due to a sort problem, the server wrote the message Sort aborted to the error log. Now the server writes more information to provide a more specific message, such as:

  Sort aborted: Out of memory (Needed 24 bytes)
  Out of sort memory, consider increasing server sort buffer size
  Sort aborted: Out of sort memory, consider increasing server sort
    buffer size
  Sort aborted: Incorrect number of arguments for FUNCTION test.f1;
    expected 0, got 1

  In addition, if the server was started with --log-warnings=2, the server writes information about the host, user, and query. (Bug #36022, Bug #11748358)

• Previously, for queries that were aborted due to a sort problem or terminated with KILL in the middle of a sort, the server wrote the message Sort aborted to the error log. Now the server writes more information about the cause of the error. These causes include:

  • Insufficient disk space in the temporary file directory prevented a temp file from being created

  • Insufficient memory for sort_buffer_size to be allocated

  • Somebody ran KILL id in the middle of a filesort operation

  • The server was shut down while some queries were sorting

  • A transaction was rolled back or aborted due to a lock wait timeout or deadlock

  • Unexpected errors, such as a source table or even temp table was corrupt

  • Processing of a subquery failed which was also sorting

  (Bug #30771, Bug #11747102)

• mysqldump --xml now displays comments from column definitions. (Bug #13618, Bug #11745324)

• Windows provides APIs based on UTF-16LE for reading from and writing to the console. MySQL now supports a utf16le character set for UTF-16LE, and the mysql client for Windows has been modified to provide improved Unicode support by using these APIs.

  To take advantage of this change, you must run mysql within a console that uses a compatible Unicode font and set the default character set to a Unicode character set that is supported for communication with the server. For instructions, see Unicode Support on Windows.

• The undocumented SHOW NEW MASTER statement has been removed, and the Com_show_new_master status variable along with it.

Bugs Fixed

• Security Fix: Pre-evaluation of LIKE predicates during view preparation could cause a server crash. (Bug #54568, Bug #11762026)

• Incompatible Change; Replication: It is no longer possible to issue a CREATE TABLE ... SELECT statement which changes any tables other than the table being created. Any such statement is not executed and instead fails with an error.

  One consequence of this change is that FOR UPDATE may no longer be used at all with the SELECT portion of a CREATE TABLE ... SELECT.

  This means that, prior to upgrading from a previous release, you should rewrite any CREATE TABLE ... SELECT statements that cause changes in other tables so that the statements no longer do so.

  This change also has implications for statement-based replication between a MySQL 5.6 (or later slave) and a master running a previous version of MySQL. In such a case, if a CREATE TABLE ... SELECT statement on the master that causes changes in other tables succeeds on the master, the statement nonetheless fails on the slave, causing replication to stop. To keep this from happening, you should either use row-based replication, or rewrite the offending statement before running it on the master. (Bug #11749792, Bug #11745361, Bug #39804, Bug #55876)

  References: See also: Bug #47899.

• Incompatible Change: When auto_increment_increment is greater than one, values generated by a bulk insert that reaches the maximum column value could wrap around rather producing an overflow error.

  As a consequence of the fix, it is no longer possible for an auto-generated value to be equal to the maximum BIGINT UNSIGNED value. It is still possible to store that value manually, if the column can accept it. (Bug #39828, Bug #11749800)

• Important Change; Partitioning: Date and time functions used as partitioning functions now have the types of their operands checked; use of a value of the wrong type is now disallowed in such cases. In addition, EXTRACT(WEEK FROM col_name), where col_name is a DATE or DATETIME column, is now disallowed altogether because its return value depends on the value of the default_week_format system variable. (Bug #54483, Bug #11761948)

  References: See also: Bug #57071, Bug #11764255.

• Important Change; Replication: The CHANGE MASTER TO statement required the value for RELAY_LOG_FILE to be an absolute path, whereas the MASTER_LOG_FILE path could be relative.

  The inconsistent behavior is resolved by permitting relative paths for RELAY_LOG_FILE, in which case the path is assumed to be relative to the slave's data directory. (Bug #12190, Bug #11745232)

• Performance; InnoDB: An UPDATE statement for an InnoDB table could be slower than necessary if it changed a column covered by a prefix index, but did not change the prefix portion of the value. The fix improves performance for InnoDB 1.1 in MySQL 5.5 and higher, and the InnoDB Plugin for MySQL 5.1. (Bug #58912, Bug #11765900)

• InnoDB; Partitioning: The partitioning handler did not pass locking information to a table's storage engine handler. This caused high contention and thus slower performance when working with partitioned InnoDB tables. (Bug #59013)

• InnoDB: This fix introduces a new configuration option, innodb_change_buffer_max_size, which defines the size of the change buffer as a percentage of the size of the buffer pool. Because the change buffer shares memory space with the buffer pool, a workload with a high rate of DML operations could cause pages accessed by queries to age out of the buffer pool sooner than desirable. This fix also devotes more I/O capacity to flushing entries from the change buffer when it exceeds 1/2 of its maximum size. (Bug #11766168, Bug #59214)

• InnoDB: The presence of a double quotation mark inside the COMMENT field for a column could prevent a foreign key constraint from being created properly. (Bug #59197, Bug #11766154)

• InnoDB: It was not possible to query the information_schema.INNODB_TRX table while other connections were running queries involving BLOB types. (Bug #55397, Bug #11762763)

• InnoDB: InnoDB returned values for “rows examined” in the query plan that were higher than expected. NULL values were treated in an inconsistent way. The inaccurate statistics could trigger “false positives” in combination with the max_join_size setting, because the queries did not really examine as many rows as reported.

  A new configuration option innodb_stats_method lets you specify how NULL values are treated when calculating index statistics. Allowed values are nulls_equal (the default), nulls_unequal and null_ignored. The meanings of these values are similar to those of the myisam_stats_method option. (Bug #30423)

• Partitioning: Failed ALTER TABLE ... PARTITION statements could cause memory leaks. (Bug #56380, Bug #11763641)

  References: See also: Bug #46949, Bug #11755209, Bug #56996, Bug #11764187.

• Replication: When using the statement-based logging format, INSERT ON DUPLICATE KEY UPDATE and INSERT IGNORE statements affecting transactional tables that did not fail were not written to the binary log if they did not insert any rows. (With statement-based logging, all successful statements should be logged, whether they do or do not cause any rows to be changed.) (Bug #59338, Bug #11766266)

• Replication: Formerly, STOP SLAVE stopped the slave I/O thread first and then stopped the slave SQL thread; thus, it was possible for the I/O thread to stop after replicating only part of a transaction which the SQL thread was executing, in which case—if the transaction could not be rolled back safely—the SQL thread could hang.

  Now, STOP SLAVE stops the slave SQL thread first and then stops the I/O thread; this guarantees that the I/O thread can fetch any remaining events in the transaction that the SQL thread is executing, so that the SQL thread can finish the transaction if it cannot be rolled back safely. (Bug #58546, Bug #11765563)

• Replication: mysqlbinlog printed USE statements to its output only when the default database changed between events. To illustrate how this could cause problems, suppose that a user issued the following sequence of statements:

  CREATE DATABASE mydb;
  USE mydb;
  CREATE TABLE mytable (column_definitions);
  DROP DATABASE mydb;
  CREATE DATABASE mydb;
  USE mydb;
  CREATE TABLE mytable (column_definitions);

  When played back using mysqlbinlog, the second CREATE TABLE statement failed with Error: No Database Selected because the second USE statement was not played back, due to the fact that a database other than mydb was never selected.

  This fix ensures that mysqlbinlog outputs a USE statement whenever it reads one from the binary log. (Bug #50914, Bug #11758677)

• Replication: The --help text for mysqlbinlog now indicates that the --verbose (-v) option outputs pseudo-SQL that is not necessarily valid SQL and cannot be guaranteed to work verbatim in MySQL clients. (Bug #47557, Bug #11755743)

• Microsoft Windows: On Windows, an object in thread local storage could be used before the object was created. (Bug #55730, Bug #11763065)

• Two unused test files in storage/ndb/test/sql contained incorrect versions of the GNU Lesser General Public License. The files and the directory containing them have been removed. (Bug #11810224)

  References: See also: Bug #11810156.

• Queries that used COALESCE() with cp1251 strings could result in an “illegal mix of collations” error. (Bug #60101, Bug #11766874)

• An assertion was raised if an XA COMMIT was issued when an XA transaction had already encountered an error (such as a deadlock) that required the transaction to be rolled back. (Bug #59986, Bug #11766788)

• On some systems, debug builds of comp_err could fail due to an uninitialized variable. (Bug #59906, Bug #11766729)

• Setting the optimizer_switch system variable to an invalid value caused a server crash. (Bug #59894, Bug #11766719)

• Attempting to create a spatial index on a CHAR column longer than 31 bytes led to an assertion failure if the server was compiled with safemutex support. (Bug #59888, Bug #11766714)

• Aggregation followed by a subquery could produce an incorrect result. (Bug #59839, Bug #11766675)

• The Performance Schema did not update status handler status variables, so SHOW STATUS LIKE '%handler%' produced undercounted values. (Bug #59799, Bug #11766645)

• Internally, XOR items partially behaved like functions and partially as conditions. This resulted in inconsistent handling and crashes. The issue is fixed by consistently treating XOR items as functions. (Bug #59793, Bug #11766642)

• An incorrect character set pointer passed to my_strtoll10_mb2() caused an assertion to be raised. (Bug #59648, Bug #11766519)

• DES_DECRYPT() could crash if the argument was not produced by DES_ENCRYPT(). (Bug #59632, Bug #11766505)

• The server and client did not always properly negotiate authentication plugin names. (Bug #59453, Bug #11766356)

• --autocommit=ON did not work (it set the global autocommit value to 0, not 1). (Bug #59432, Bug #11766339)

• FIND_IN_SET() could work differently in MySQL 5.5 than in 5.1. (Bug #59405, Bug #11766317)

• mysqldump did not quote database names in ALTER DATABASE statements in its output, which could cause an error at reload time for database names containing a dash. (Bug #59398, Bug #11766310)

• If filesort fell back to an ordinary sort/merge, it could fail to handle memory correctly. (Bug #59331, Bug #11766260)

• Comparisons of aggregate values with TIMESTAMP values were incorrect. (Bug #59330, Bug #11766259)

• The “greedy” query plan optimizer failed to consider the size of intermediate query results when calculating the cost of a query. This could result in slowly executing queries when there are much faster execution plans available. (Bug #59326, Bug #11766256)

• A query of the following form returned an incorrect result, where the values for col_name in the result set were entirely replaced with NULL values:

  SELECT DISTINCT col_name ... ORDER BY col_name DESC;

  (Bug #59308, Bug #11766241)

• The MYSQL_HOME environment variable was being ignored. (Bug #59280, Bug #11766219)

• SHOW PRIVILEGES did not display a row for the PROXY privilege. (Bug #59275, Bug #11766216)

• SHOW PROFILE could truncate source file names or fail to show function names. (Bug #59273, Bug #11766214)

• For DIV expressions, assignment of the result to multiple variables could cause a server crash. (Bug #59241, Bug #11766191)

  References: See also: Bug #8457.

• MIN(year_col) could return an incorrect result in some cases. (Bug #59211, Bug #11766165)

• With index condition pushdown enabled, a join could produce an extra row due to parts of the select condition for the second table in the join not being evaluated. (Bug #59186, Bug #11766144)

• DELETE or UPDATE statements could fail if they used DATE or DATETIME values with a year, month, or day part of zero. (Bug #59173)

• The ESCAPE clause for the LIKE operator permits only expressions that evaluate to a constant at execution time, but aggregate functions were not being rejected. (Bug #59149, Bug #11766110)

• Valgrind warnings about uninitialized variables were corrected. (Bug #59145, Bug #11766106)

• Memory leaks detected by Valgrind, some of which could cause incorrect query results, were corrected. (Bug #59110, Bug #11766075)

• mysqlslap failed to check for a NULL return from mysql_store_result() and crashed trying to process the result set. (Bug #59109, Bug #11766074)

• There was an erroneous restriction on file attributes for LOAD DATA. The requirement that a file be located in the database directory or world readable is now that the be located in the database directory or readable by the user account used to run the server. (Bug #59085, Bug #11766052)

• SHOW CREATE TRIGGER failed if there was a temporary table with the same name as the trigger subject table. (Bug #58996, Bug #11765972)

• The DEFAULT_CHARSET and DEFAULT_COLLATION CMake options did not work. (Bug #58991, Bug #11765967)

• In a subquery, a UNION with no referenced tables (or only a reference to the DUAL virtual table) did not permit an ORDER BY clause. (Bug #58970, Bug #11765950)

• OPTIMIZE TABLE for an InnoDB table could raise an assertion if the operation failed because it had been killed. (Bug #58933, Bug #11765920)

• If max_allowed_packet was set larger than 16MB, the server failed to reject too-large packets with “Packet too large” errors. (Bug #58887, Bug #11765878)

• With index condition pushdown enabled, incorrect results were returned for queries on MyISAM tables involving HAVING and LIMIT, when the column in the WHERE condition contained NULL. (Bug #58838, Bug #11765835)

• An uninitialized variable for the index condition pushdown access method could result in a server crash or Valgrind warnings. (Bug #58837, Bug #11765834)

• A NOT IN predicate with a subquery containing a HAVING clause could retrieve too many rows, when the subquery itself returned NULL. (Bug #58818, Bug #11765815)

• Running a query against an InnoDB table twice, first with index condition pushdown enabled and then with it disabled, could produce different results. (Bug #58816, Bug #11765813)

• An assertion was raised if a stored routine had a DELETE IGNORE statement that failed but due to the IGNORE had not reported any error. (Bug #58709, Bug #11765717)

• WHERE conditions of the following forms were evaluated incorrectly and could return incorrect results:

  WHERE null-valued-const-expression NOT IN (subquery)
  WHERE null-valued-const-expression IN (subquery) IS UNKNOWN

  (Bug #58628, Bug #11765642)

• Issuing EXPLAIN EXTENDED for a query that would use condition pushdown could cause mysqld to crash. (Bug #58553, Bug #11765570)

• An OUTER JOIN query using WHERE col_name IS NULL could return an incorrect result. (Bug #58490, Bug #11765513)

• Starting the server with the --defaults-file=file_name option, where the file name had no extension, caused a server crash. (Bug #58455, Bug #11765482)

• Outer joins with an empty table could produce incorrect results. (Bug #58422, Bug #11765451)

• In debug builds, SUBSTRING_INDEX(FORMAT(...), FORMAT(...)) could cause a server crash. (Bug #58371, Bug #11765406)

• When mysqladmin was run with the --sleep and --count options, it went into an infinite loop executing the specified command. (Bug #58221, Bug #11765270)

• Some string-manipulating SQL functions use a shared string object intended to contain an immutable empty string. This object was used by the SQL function SUBSTRING_INDEX() to return an empty string when one argument was of the wrong data type. If the string object was then modified by the SQL function INSERT(), undefined behavior ensued. (Bug #58165, Bug #11765225)

• Condition pushdown optimization could push down conditions with incorrect column references. (Bug #58134, Bug #11765196)

• injector::transaction did not have support for rollback. (Bug #58082, Bug #11765150)

• Parsing nested regular expressions could lead to recursion resulting in a stack overflow crash. (Bug #58026, Bug #11765099)

• The fix for Bug #25192 caused load_defaults() to add an argument separator to distinguish options loaded from option files from those provided on the command line, whether or not the application needed it. (Bug #57953, Bug #11765041)

  References: See also: Bug #25192, Bug #11746296.

• The mysql client went into an infinite loop if the standard input was a directory. (Bug #57450, Bug #11764598)

• If a multiple-table update updated a row through two aliases and the first update physically moved the row, the second update failed to locate the row. This resulted in different errors depending on the storage engine, although these errors did not accurately describe the problem:

  • MyISAM: Got error 134 from storage engine

  • InnoDB: Can't find record in 'tbl'

  For MyISAM, which is nontransactional, the update executed first was performed but the second was not. In addition, for two equal multiple-table update statements, one could succeed and the other fail depending on whether the record actually moved, which is inconsistent.

  Now such an update returns an error if it will update a table through multiple aliases, and perform an update that may physically move the row in at least one of these aliases. (Bug #57373, Bug #11764529, Bug #55385, Bug #11762751)

• SHOW WARNINGS output following EXPLAIN EXTENDED could include unprintable characters. (Bug #57341, Bug #11764503)

• Outer joins on a unique key could return incorrect results. (Bug #57034, Bug #11764219)

• For a query that used a subquery that included GROUP BY inside a < ANY() construct, no rows were returned when there should have been. (Bug #56690, Bug #11763918)

• Some RPM installation scripts used a hardcoded value for the data directory, which could result in a failed installation for users who have a nonstandard data directory location. The same was true for other configuration values such as the PID file name. (Bug #56581, Bug #11763817)

• On FreeBSD and OpenBSD, the server incorrectly checked the range of the system date, causing legal values to be rejected. (Bug #55755, Bug #11763089)

• If one connection locked the mysql.func table using either FLUSH TABLES WITH READ LOCK or LOCK TABLE mysql.func WRITE and a second connection tried to either create or drop a UDF function, a deadlock occurred when the first connection tried to use a UDF function. (Bug #53322, Bug #11760878)

• DISTINCT aggregates on DECIMAL UNSIGNED fields could trigger an assertion. (Bug #52171, Bug #11759827)

• On FreeBSD, if mysqld was killed with a SIGHUP signal, it could corrupt InnoDB .ibd files. (Bug #51023, Bug #11758773)

• An assertion could be raised if −1 was inserted into an AUTO_INCREMENT column by a statement writing more than one row. (Bug #50619, Bug #11758417)

• A query that contains an aggregate function but no GROUP BY clause is implicitly grouped. If such a query also contained an ORDER BY clause, the optimizer could choose to use a temporary table to perform the ordering. This is unnecessary because implicitly grouped queries return at most one row and need no ordering. (Bug #47853)

• The parser failed to initialize some internal objects properly, which could cause a server crash in the cleanup phase after statement execution. (Bug #47511, Bug #11755703)

• When CASE ... WHEN arguments had different character sets, 8-bit values could be referenced as utf16 or utf32 values, raising an assertion. (Bug #44793, Bug #11753363)

• When using ExtractValue() or UpdateXML(), if the XML to be read contained an incomplete XML comment, MySQL read beyond the end of the XML string when processing, leading to a crash of the server. (Bug #44332, Bug #11752979)

• Bitmap functions used in one thread could change bitmaps used by other threads, raising an assertion. (Bug #43152, Bug #11752069)

• DATE_ADD() and DATE_SUB() return a string if the first argument is a string, but incorrectly returned a binary string. Now they return a character string with a collation of connection_collation. (Bug #31384, Bug #11747221)