MySQL 5.1.9-beta がリリースされました。


以下チェンジログ
---------------------

■機能の追加と変更(5.1.9)
- SHOW PLUGIN was renamed to SHOW PLUGINS. SHOW PLUGIN now is deprecated and generates a warning. (Bug#17112: http://bugs.mysql.com/17112)

- mysqld_safe no longer checks for a mysqld-max binary. Instead, mysqld_safe nows checks only for the standard mysqld server unless another server binary is specified explicitly. (Bug#17861: http://bugs.mysql.com/17861)

- For partitioned tables, the output of SHOW TABLE STATUS now shows in the Engine column the name of the storage engine used by all partitions for the table; in the Create_options column, the output now shows "partitioned" for a partitioned table. This change also affects the values shown in the corresponding columns of the INFORMATION_SCHEMA.TABLES table. (Bug#17631: http://bugs.mysql.com/17631)

- The NDBCluster storage engine now supports CREATE TABLE statements of arbitrary length. (Previously, CREATE TABLE statements for MySQL Cluster tables could contain a maximum of 4096 characters only.) (Bug#17813: http://bugs.mysql.com/17813)

- Large file support was re-enabled for the MySQL server binary for the AIX 5.2 platform. (Bug#13571: http://bugs.mysql.com/13571)



■バグ修正

- NDB Cluster: BLOB columns did not work correctly with user-partitioned NDB tables. (Bug#16796: http://bugs.mysql.com/16796)

- mysql_config returned incorrect libraries on x86_64 systems. (Bug#13158: http://bugs.mysql.com/13158)

- mysql_reconnect() sent a SET NAMES statement to the server, even for pre-4.1 servers that do not understand the statement. (Bug#18830: http://bugs.mysql.com/18830)

- COUNT(*) on a MyISAM table could return different results for the base table and a view on the base table. (Bug#18237: http://bugs.mysql.com/18237)

- For mysql.server, if the basedir option was specified after datadir in an option file, the setting for datadir was ignored and assumed to be located under basedir. (Bug#16240: http://bugs.mysql.com/16240)

- For full-text searches in boolean mode, and when a full-text parser plugin was used, a MYSQL_FTPARSER_PARAM::ftparser_state could have been corrupted by recursive calls to the plugin. (Bug#18836: http://bugs.mysql.com/18836)

- EXTRACT(QUARTER FROM date) returned unexpected results. (Bug#18100: http://bugs.mysql.com/18100)

- TRUNCATE did not reset the AUTO_INCREMENT counter for MyISAM tables when issued inside a stored procedure. (Bug#14945: http://bugs.mysql.com/14945) Note: This bug did not affect InnoDB tables. Also, TRUNCATE does not reset the AUTO_INCREMENT counter for NDBCluster tables regardless of when it is called (see Bug#18864: http://bugs.mysql.com/18864).

- The server was always built as though --with-extra-charsets=complex had been specified. (Bug#12076: http://bugs.mysql.com/12076)

- Partition pruning did not work properly for some kinds of partitioning and subpartitioning, with certain WHERE clauses. (Partitions and subpartitions that should have been marked as used were not so marked.) The error could manifest as incorrect content in EXPLAIN PARTITIONS output as well as missing rows in the results of affected queries. (Bug#18558: http://bugs.mysql.com/18558)

- NDB Cluster: An unitialised internal variable could lead to unexpected results. (Bug#18831: http://bugs.mysql.com/18831)

- For tables created in a MySQL 4.1 installation upgraded to MySQL 5.0 and up, multiple-table updates could update only the first matching row. (Bug#16281: http://bugs.mysql.com/16281)

- Complex queries with nested joins could cause a server crash. (Bug#18279: http://bugs.mysql.com/18279)

- A query against a partitioned table using WHERE col IS NULL could produce incorrect results given the following conditions: + The table had partitions and subpartitions + The partitioning function depended on a single column col of one of the MySQL integer types + The partitioning function was not monotonically increasing The same issue could cause the server to crash when run in debug mode. (Bug#18659: http://bugs.mysql.com/18659)

- CAST(double AS SIGNED INT) for large double values outside the signed integer range truncates the result to be within range, but the result sometimes had the wrong sign, and no warning was generated. (Bug#15098: http://bugs.mysql.com/15098)

- MEDIUMINT columns were not handled in the same way as other column types by partition pruning. Partition pruning would sometimes use inappropriate columns in preforming queries. Both of these issues were rectified as part of the same bugfix. (Bug#18025: http://bugs.mysql.com/18025)

- Quoted values could not be used for partition option values. (Bug#13520: http://bugs.mysql.com/13520)

- Delimited identifiers could not be used in defining partitions. (Bug#13433: http://bugs.mysql.com/13433)

- Building the server using --with-example-storage-engine failed to enable the EXAMPLE storage engine in the server. (Bug#18464: http://bugs.mysql.com/18464)

- Triggers created in one version of the server could not be dropped after upgrading to a newer version. (Bug#15921: http://bugs.mysql.com/15921)

- Queries using WHERE ... IS NULL returned incorrect results from partitioned tables. (Bug#18070: http://bugs.mysql.com/18070)

- Partition pruning did not perform correctly with partitions on NULL, and could potentially crash the server. (Bug#18053: http://bugs.mysql.com/18053)

- If InnoDB encountered a HA_ERR_LOCK_TABLE_FULL error and rolled back a transaction, the transaction was still written to the binary log. (Bug#18283: http://bugs.mysql.com/18283)