Thank you to the MySQL Community, on behalf of the MySQL team @ Oracle. Your bug reports, testcases and patches have helped create a better MySQL 5.6.22.
In particular:
- Thank you to Ramesh Sivaraman for reporting that setting the innodb_limit_optimistic_insert_debug debug configuration option to 1 caused an infinite B-tree page split. Ramesh also provided a minimal testcase to reproduce. Bug #74605.
- Thank you to Jeremy Cole for reporting that InnoDB pages with a checksum value of zero were incorrectly treated as empty pages. Bug #73689.
- Thank you to Zhai Weixiang for reporting that the MySQL server would fail to shutdown when started with innodb-force-recovery = 6. Bug #73341.
- Thank you to Bart Butler for reporting that InnoDB would shutdown unclean on systems with lower_case_table_names=2 and containing tables using uppercase. Bug #72043.
- Thank you to Laurynas Biveinis for reporting two assertions that could occur in InnoDB using debug builds. Bug #71343, Bug #68116.
- Thank you to Stewart Smith for reporting that InnoDB was missing a memory barrier when setting up asynchronous I/O (AIO). Stewart was also kind enough to provide a patch addressing the issue. Bug #72809.
- Thank you to Jean-François Gagné for reporting an out of memory situation when using row-based replication and having different data types on the slave. Jean-François also provided an example shell script, which saved us a lot of time in reproducing the issue. Bug #72610.
- Thank you to Jean-François Gagné for reporting that relay log recovery did not correctly handle the case where Relay_Master_Log_File could be empty. Bug #73039.
- Thank you to Santosh Praneeth Banda for reporting that replication masters did not correctly handle the case when slaves requested GTIDs which had already been purged. Bug #73032.
- Thank you to Domas Mituzas for reporting that mysqlbinlog incurred significant kernel mutex contention by calling localtime() for every event read. We have since modified mysqlbinlog to use localtime_r instead. Bug #72701.
- Thank you to Jeremy Cole for reporting that the variable SQL_LOG_BIN having global scope is dangerous. We agree with Jeremy and have modified the behaviour so that it may only be set on a session basis. Bug #67433.
- Thank you to Jean-François Gagné for reporting that the mysqld startup script would fail on CentOS when the socket option was relative. Bug #74111.
- Thank you to Eugene Zheganin for reporting that MySQL would not compile with DTrace on Solaris 11.2. Bug #73826.
- Thank you to Nikola Glavinceski for reporting that certain queries for which subquery materialization or UNION DISTINCT was used together with a hash index on a temporary table could produce incorrect results or cause a server exit. Bug #73368.
- Thank you to Hartmut Holzgraefe for reporting that the ENABLED_LOCAL_INFILE cmake option was incorrectly enabled by default. Bug #72106.
- Thank you to Jason Rider for reporting that the use of ODBC-format date literals could produce incorrect query results. Bug #69233.
- Thank you to Nicholas Bamber for reporting that mysql_setpermission failed to properly quote user names in SQL statements that it generated. Bug #66317.
Thank you again to all the community contributors listed above. In addition, the MySQL team would like to thank the community for their feedback on Bug #67433.
If I missed a name here, please let me know!
– Morgan