MySQL 5.0.21がリリースされました。


最も普及しているオープンソースデータベース管理システムであるMySQL 5.0.21がリリースされました。コミュニティエディションは、とミラーサイトのダウンロード・ページから、ソースコード及び多くのプラットフォームのためのバイナリで現在利用可能です。







Functionality added or changed:

* Security enhancement: Added the global max_prepared_stmt_count
  system variable to limit the total number of prepared
  statements in the server. This limits the potential for
  denial-of-service attacks based on causing the server to run
  causing the server to run out of memory by preparing huge numbers
  of statements. The current number of prepared statements is
  available through the 'prepared_stmt_count' status variable.
* NDB Cluster: It is now possible to perform a partial start of
  a cluster. That is, it is now possible to bring up the cluster
  without running ndbd --initial on all configured data nodes
  first. (Bug#18606:
* NDB Cluster: It is now possible to install MySQL with Cluster
  support to a non-default location and change the search path
  for font description files using either the --basedir or
  --character-sets-dir options. (Previously in MySQL 5.0, ndbd
  searched only the default path for character sets.)
* In result set metadata, the MYSQL_FIELD.length value for BIT
  columns now is reported in number of bits. For example, the
  value for a BIT(9) column is 9. (Formerly, the value was
  related to number of bytes.)
* The default for the innodb_thread_concurrency system variable
  was changed to 8. (Bug#15868:

Bugs fixed:

* Security bugfix: A malicious client, using specially crafted
  invalid COM_TABLE_DUMP packets was able to trigger an exploitable
  buffer overflow on the server. Thanks to Stefano Di Paola
  <stefano.dipaola@stripped> for finding and reporting this bug.
* Security bugfix: A malicious client, using specially crafted
  invalid login or COM_TABLE_DUMP packets was able to read
  uninitialized memory, which potentially, though unlikely in MySQL,
  could lead to an information disclosure. Thanks to Stefano Di Paola
  <stefano.dipaola@stripped> for finding and reporting this bug.
* NDB Cluster: A simultaneous DROP TABLE and table update
  operation utilising a table scan could trigger a node failure.
* Conversion of a number to a CHAR UNICODE string returned an
  invalid result. (Bug#18691:
* DELETE and UPDATE statements that used large NOT IN
  (value_list) clauses could use large amounts of memory.
* Prevent recursive views caused by using RENAME TABLE on a view
  after creating it. (Bug#14308:
* A LOCK TABLES statement that failed could cause MyISAM not to
  update table statistics properly, causing a subsequent CHECK
  TABLE to report table corruption.
* For a reference to a non-existent stored function in a stored
  routine that had a CONTINUE handler, the server continued as
  though a useful result had been returned, possibly resulting
  in a server crash. (Bug#18787:
* InnoDB did not use a consistent read for CREATE ... SELECT
  when innodb_locks_unsafe_for_binlog was set.
* InnoDB could read a delete mark from its system tables
  incorrectly. (Bug#19217:
* Corrected a syntax error in
* A missing DBUG_RETURN() caused the server to emit a spurious
  error message: missing DBUG_RETURN or DBUG_VOID_RETURN macro
  in function "open_table".
* DROP DATABASE did not drop stored routines associated with the
  database if the database name was longer than 21 characters.
* Avoid trying to include <asm/atomic.h> when it doesn't work in
  C++ code. (Bug#13621:
* Executing SELECT on a large table that had been compressed
  within myisampack could cause a crash.
* NDB Cluster: When attempting to create an index on a BIT or
  BLOB column, Error 743: Unsupported character set in table or
  index was returned instead of Error 906: Unsupported attribute
  type in index.
* Within stored routines, usernames were parsed incorrectly if
  they were enclosed within quotes.
* Casting a string to DECIMAL worked, but casting a trimmed
  string (using LTRIM() or RTRIM()) resulted in loss of decimal
  digits. (Bug#17043:
* NDB Cluster: On slow networks or CPUs, the management client
  SHOW command could sometimes erroneously show all data nodes
  as being master nodes belonging to nodegroup 0.
* If the second or third argument to BETWEEN was a constant
  expression such as '2005-09-01 - INTERVAL 6 MONTH' and the
  other two arguments were columns, BETWEEN was evaluated
  incorrectly. (Bug#18618:
* If the first argument to BETWEEN was a DATE or TIME column of
  a view and the other arguments were constants, BETWEEN did not
  perform conversion of the constants to the appropriate
  temporary type, resulting in incorrect evaluation.
* Server and clients ignored the --sysconfdir option that was
  passed to configure. (Bug#15069:
* NDB Cluster: In a 2-node cluster with a node failure,
  restarting the node with a low value for StartPartialTimeout
  could cause the cluster to come up partitioned ("split-brain"
  issue). (Bug#16447:
  A similar issue could occur when the cluster was first started
  with a sufficiently low value for this parameter.
* NDB Cluster: On systems with multiple network interfaces, data
  nodes would get "stuck" in startup phase 2 if the interface
  connecting them to the management server was working on node
  startup while the interface interconnecting the data nodes
  experienced a temporary outage.
* NDB Cluster: Unused open handlers for tables in which the
  metadata had changed were not properly closed. This could
  result in stale results from Cluster tables following an ALTER
  TABLE. (Bug#13228:
* NDB Cluster: Uninitialised internal variables could lead to
  unexpected results. (Bug#11033:,
* For InnoDB tables, an expression of the form col_name BETWEEN
  col_name2 - INTERVAL x DAY AND col_name2 + INTERVAL x DAY when
  used in a join returned incorrect results.
* INSERT DELAYED into a view caused an infinite loop.
* Lettercase in database name qualifiers was not consistently
  handled properly in queries when lower_case_table_names was
  set to 1. (Bug#15917:
* The optimizer could cause a server crash or use a non-optimal
  subset of indexes when evaluating whether to use Index
  Merge/Intersection variant of index_merge optimization.
* The presence of multiple equalities in a condition after
  reading a constant table could cause the optimizer not to use
  an index. This resulted in certain queries being much slower
  than in MySQL 4.1. (Bug#16504:
* A recent change caused the mysql client not to display NULL
  values correctly and to display numeric columns left-justified
  rather than right-justified. The problems have been corrected.
* mysql_reconnect() sent a SET NAMES statement to the server,
  even for pre-4.1 servers that do not understand the statement.
* COUNT(*) on a MyISAM table could return different results for
  the base table and a view on the base table.
* DELETE with LEFT JOIN for InnoDB tables could crash the server
  if innodb_locks_unsafe_for_binlog was enabled.
* InnoDB failure to release an adaptive hash index latch could
  cause a server crash if the query cache was enabled.
* 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.
* The euro sign () was not stored correctly in columns using
  the latin1_german1_ci or latin1_general_ci collation.
* EXTRACT(QUARTER FROM date) returned unexpected results.
* TRUNCATE did not reset the AUTO_INCREMENT counter for MyISAM
  tables when issued inside a stored procedure.
  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
* The server was always built as though
  --with-extra-charsets=complex had been specified.
* A query using WHERE (column_1, column_2) IN ((value_1,
  value_2)[, (..., ...), ...]) would return incorrect results.
* Queries of the form SELECT DISTINCT timestamp_column WHERE
  date_function(timestamp_col) = constant did not return all
  matching rows. (Bug#16710:
* When running a query that contained a GROUP_CONCAT( SELECT
  GROUP_CONCAT(...) ), the result was NULL except in the ROLLUP
  part of the result, if there was one.
* 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:
* NDB Cluster: When multiple node restarts were attempted
  without allowing each restart to complete, the error message
  returned was Array index out of bounds rather than Too many
  crashed replicas. (Bug#18349:
* 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:
* Updating a field value when also requesting a lock with
  GET_LOCK() would cause slave servers in a replication
  environment to terminate.