* Incompatible change: The Event Scheduler can now be in one of three states (on, off, or the new suspended state). In addition, due to the fact that SET GLOBAL event_scheduler; now acts in a synchronous rather than asynchronous manner, the Event Scheduler thread can be no longer be activated or deactivated at run time.
For more information regarding these changes, see Section 22.1, "Event Scheduler Overview."
* Previously, to build MySQL from source with SSL support enabled, you would invoke configure with either the –with-openssl or –with-yassl option. Those options both have been replaced by the –with-ssl option. By default, –with-ssl causes the bundled yaSSL library to be used. To select OpenSSL instead, give the option as –with-ssl=path, where path is the directory where the OpenSSL header files and libraries are
* Added the –ssl-verify-server-cert option to MySQL client programs. This option causes the server’s Common Name value in its certificate to be verified against the hostname used when connecting to the server, and the connection is rejected if there is a mismatch. Added MYSQL_OPT_SSL_VERIFY_SERVER_CERT option for the mysql_options() C API function to enable this verification. This feature can be used to prevent man-in-the-middle attacks. Verification is disabled by default.
* Added the ssl_ca, ssl_capath, ssl_cert, ssl_cipher, and ssl_key system variables, which display the values given via the corresponding command options. See Section 184.108.40.206, "SSL Command Options."
* NDB Cluster: The limit of 2048 ordered indexes per cluster has been lifted. There is now no upper limit on the number of ordered indexes (including AUTO_INCREMENT columns) that may be used.
* Added the log_queries_not_using_indexes system variable.
* Added the –angel-pid-file option to mysqlmanager for specifying the file in which the angel process records its process ID when mysqlmanager runs in daemon mode.
* To prevent inadvertant loss of data, ALTER TABLE can no longer be used to change the storage engine of a table to MERGE or BLACKHOLE. (Bug#10952:http://bugs.mysql.com/10952)
* The ENABLE KEYS and DISABLE KEYS clauses for the ALTER TABLE statement are now supported for partitioned tables.
* It is now possible to use NEW.var_name values within triggers as INOUT parameters to stored procedures.
* Security fix: An SQL-injection security hole has been found in multi-byte encoding processing. The bug was in the server, incorrectly parsing the string escaped with the mysql_real_escape_string() C API function.
This vulnerability was discovered and reported by Josh Berkus<email@example.com> and Tom Lane <firstname.lastname@example.org> as part of the inter-project security collaboration of the OSDB consortium. For more information about SQL injection, please see the text at the end of this announcement.
* File size specifications for InnoDB data files were case sensitive.
* InnoDB did not increment the handler_read_prev counter.
* A CREATE TABLE statement that created a table from a materialized view did not inherit default values from the underlying table.
* NDB Cluster: mysqld processes did not always detect cluster shutdown, leading to issues with CLuster replication and schema distribution.
* NDB Cluster: SELECT MIN(unique_column) from a Cluster table with user-defined partitioning crashed the server.
* If there is a global read lock, CREATE DATABASE, RENAME DATABASE, and DROP DATABASE could deadlock.
* Premature optimization of nested subqueries in the FROM clause that refer to aggregate functions could lead to incorrect results.
* For dates with 4-digit year parts less than 200, an implicit conversion to add a century was applied for date arithmetic performed with DATE_ADD(), DATE_SUB(), + INTERVAL, and – INTERVAL. (For example, DATE_ADD(‘0050-01-01 00:00:00’, INTERVAL 0 SECOND) became ‘2050-01-01 00:00:00’.) Now these operations return NULL rather than an incorrect non-NULL value.
* BLOB or TEXT arguments to or values returned from stored functions were not copied properly if too long and could become garbled.
* Simultaneous scheduled events whose actions conflicted with one another could crash the server.
* In was not possible to invoke a stored routine containing dynamic SQL from a scheduled event.
* NDB Cluster: Running ALL START in the NDB management client or restarting multiple nodes simultaneously could under some circumstances cause the cluster to crash.
* The result from CONV() is a string, but was not always treated the same way as a string when converted to a real value for an arithmetic operation.
* CREATE TABLE … SELECT … statements that used a stored function explicitly or implicitly (through a view) resulted in a Table not locked error.
* Within a trigger, SET used the SQL mode of the invoking statement, not the mode in effect at trigger creation time.
* The server no longer uses a signal handler for signal 0 because it could cause a crash on some platforms.
* The embedded server crashed with row-based replication enabled.
* Display better error message for ALTER TABLE operations that will result in duplicate keys due to AUTO_INCREMENT resequencing.
* The Data_free column in the output of SHOW TABLE STATUS always displayed 0 for partitioned tables.
* Adding an index to a table created using partitioning by KEY and the MEMORY storage engine caused the server to crash.
* When creating a table using CREATE TABLE … PARTITION BY … SELECT …, the partitioning clause was ignored.
* ALTER TABLE ENGINE=… failed when used to change a MySQL Cluster table having no explicit primary key to use a different storage engine.
Note: As a consequence of this fix, SHOW CREATE TABLE no longer displays auto-partitioning informaiton for NDBCluster tables.
* NDB Cluster (NDBAPI): On big-endian platforms, NdbOperation::write_attr() did not update 32-bit fields correctly.
* NDB Cluster: Using "stale" mysqld .FRM files could cause a newly-restored cluster to fail. This situation could arise when restarting a MySQL Cluster using the –intial option while leaving connected mysqld processes running.
* NDB Cluster (Replication): Memory was not freed after some ALTER TABLE operations, which could cause mysqld processes to crash.
* NDB Cluster (NDBAPI): The Ndb::dropEventOperation() method failed to clean up all objects used, which could cause memory leaks to occur.
* NDB Cluster: Data node failures could cause excessive CPU usage by ndb_mgmd.
* NDB Cluster: TRUNCATE failed on tables having BLOB or TEXT columns with the error Lock wait timeout exceeded. This affected both in-memory and Disk Data tables.
* Revised memory allocation for local objects within stored functions and triggers to avoid memory leak for repeated function or trigger invocation.
* EXPLAIN … SELECT INTO caused the client to hang.
* Symlinking .mysql_history to /dev/null to suppress statement history saving by mysql did not work. (mysql deleted the symlink and recreated .mysql_history as a regular file, and then wrote history to it.)
* The basedir and tmpdir system variables could not be accessed via @@var_name syntax.
* Corrected several problems with the treatment of the –log-error option by mysqld_safe. These problems were manifest as differences from mysqld in error log handling.
+ If a filename was given for –log-error, mysqld_safe ignored it and did not pass it to mysqld, which then wrote error information to stderr and resulted in incorrect log rotation when FLUSH LOGS was used.
+ mysql_safe now adds .err to the end of the filename if no extension is present (the same as mysqld).
+ mysqld_safe treated a relative pathname as relative to its own current working directory. Now it treats a relative pathname as relative to the data directory (the same as mysqld).
In addition, some argument quoting problems were corrected.
* Returning the value of a system variable from a stored function caused a server crash.
* Use of uninitialized user variables in a subquery in the FROM clause resulted in bad entries in the binary log.
* IS_USED_LOCK() could return an incorrect connection identifier.
* Concurrent reading and writing of privilege structures could crash the server.
Additional information on "SQL injection":
Discussion: An SQL-injection security hole has been found in multi-byte encoding processing. An SQL-injection security hole can include a situation whereby when a user supplied data to be inserted into a database, the user might inject SQL statements into the data that the server will execute. With regards to this vulnerability, when character set unaware-escaping is used (for example, addslashes() in PHP), it is possible to bypass the escaping in some multi-byte character sets (for example, SJIS, BIG5 and GBK). As a result, a function such as addslashes() is not able to prevent SQL-injection attacks. It is impossible to fix this on the server side. The best solution is for applications to use character set-aware escaping offered by a function such mysql_real_escape_string().
However, a bug was detected in how the MySQL server parses the output of mysql_real_escape_string(). As a result, even when the character set-aware function mysql_real_escape_string() was used, SQL injection was possible. This bug has been fixed.
Workarounds: If you are unable to upgrade MySQL to a version that includes the fix for the bug in mysql_real_escape_string() parsing, but run MySQL 5.0.1 or higher, you can use the NO_BACKSLASH_ESCAPES SQL mode as a workaround. (This mode was introduced in MySQL 5.0.1.) NO_BACKSLASH_ESCAPES enables an SQL standard compatibility mode, where backslash is not considered a special character.
The result will be that queries will fail.
To set this mode for the current connection, enter the following SQL statement:
You can also set the mode globally for all clients:
SET GLOBAL sql_mode=’NO_BACKSLASH_ESCAPES’;
This SQL mode also can be enabled automatically when the server starts by using the command-line option –sql-mode=NO_BACKSLASH_ESCAPES or by setting sql-mode=NO_BACKSLASH_ESCAPES in the server option file (for example, my.cnf or my.ini, depending on your system).