Download | Release Notes | Changelog | Overview of 5.5

Release date: 25 Feb 2014

For the highlights of this release, see the release notes.

The revision number links will take you to the revision's page on Launchpad. On Launchpad you can view more details of the revision and view diffs of the code modified in that revision.

  • Revision #4095 [merge] Sat 2014-02-22 22:51:20 +0100
    • 5.3 merge
    • Revision #2502.567.209 Thu 2014-02-20 21:27:33 -0800
      • After constant row substitution the optimizer should call the method update_used_tables for the the where condition to update cached indicators of constant subexpressions. It should be done before further possible simplification of the where condition.
    • Revision #2502.567.208 Wed 2014-02-19 18:34:12 +0400
      • Backport the following from 5.5 to 5.3: MDEV-4556 Server crashes in SEL_ARG::rb_insert with index_merge+index_merge_sort_union, FORCE INDEX - merge_same_index_scans() may put the same SEL_ARG tree in multiple result plans. make it call incr_refs() on the SEL_ARG trees that it does key_or() on, because key_or(sel_arg_tree_1, sel_arg_tree_2) call may invalidate SEL_ARG trees pointed by sel_arg_tree_1 and sel_arg_tree_2.
    • Revision #2502.567.207 Wed 2014-02-19 17:52:47 +0400
      • Fix compile failure:
    • Revision #2502.567.206 Wed 2014-02-19 17:47:02 +0400
      • Add a debugger helper function that does this:
    • Revision #2502.567.205 Wed 2014-02-19 17:45:33 +0400
      • MDEV-5600: Wrong result on 2nd execution of PS depending on the length of the query - Item_direct_view_ref didn't clear its pointer to item_equal in ::cleanup. - Some Item_direct_view_ref objects have statement lifetime (i.e. they survive across multiple EXECUTE commands). Item_equal objects live only for the duration of one EXECUTE. This caused Item_direct_view_ref to have a stale pointer, which could cause all sorts of effects. (In this bug's testcase it was pointing to the wrong Item_equal, causing wrong query result) - Fixed by doing what Item_field::cleanup() does - don't keep item_equal pointer value. - There is no testcase because the only testcase I've got is highly fragile (e.g. the bug will not show up if @@datadir is of the wrong length).
    • Revision #2502.567.204 Tue 2014-02-18 17:45:08 +0400
      • MDEV-5481 mysqldump fails to dump geometry types properly.
    • Revision #2502.567.203 Tue 2014-02-18 17:15:25 +0400
      • MDEV-5615 crash in Gcalc_function::add_operation.
    • Revision #2502.567.202 [merge] Sat 2014-02-15 01:26:53 +0400
      • Merge
      • Revision #2502.585.1 Sat 2014-02-15 01:21:46 +0400
        • MDEV-5581: Server crashes in in JOIN::prepare on 2nd execution of PS with materialization+semijoin - The problem was that JOIN::prepare() tried to set TABLE::maybe_null for a table in join. Non-merged semi-join tables 1) are present as join's base tables on second EXECUTE, but 2) do not yet have a TABLE object. Worked around the problem by putting mixed_implicit_grouping into JOIN object, and then passing it to JTBM tables in setup_jtbm_semi_joins().
  • Revision #4094 Fri 2014-02-21 23:42:29 +0100
    • Fix "cmake . && cmake -DWITHOUT_TOKUDB=1" to disable tokudb
  • Revision #4093 Fri 2014-02-21 00:53:02 +0100
    • MDEV-5624 mysqldump --dump-slave option does not restart the replication if the dump has failed
  • Revision #4092 Fri 2014-02-21 00:52:58 +0100
    • federatedx: avoid unnecessary bzero. improve dbug traces
  • Revision #4091 Fri 2014-02-21 00:52:50 +0100
    • MDEV-5698 Using ORDER BY in a FederatedX table is abnormally slow
  • Revision #4090 Wed 2014-02-19 21:40:34 +0100
  • Revision #4089 Wed 2014-02-19 21:40:25 +0100
    • MDEV-5390 doesn't install on fedora if mysql is installed, part 2
  • Revision #4088 Wed 2014-02-19 21:40:20 +0100
    • RPM: * readability fixes * CPackRPM wrapper to fix property leakage between components (cmake bug 13248)
  • Revision #4087 Wed 2014-02-19 21:40:15 +0100
    • increment and get the query_id atomically, otherwise two concurrent threads might end up having the same query id
  • Revision #4086 Wed 2014-02-19 21:39:16 +0100
    • MDEV-5529 Sync libmysqlclient.so symbol versioning across distributions
  • Revision #4085 Mon 2014-02-17 11:10:30 +0100
  • Revision #4084 Mon 2014-02-17 11:10:26 +0100
    • MDEV-5654 Server crashes on second installation of daemon_example plugin
  • Revision #4083 Mon 2014-02-17 11:10:22 +0100
    • At --log-warnings=9 or more, log at what address a dynamic plugin was loaded. It helps to interpret valgrind/safemalloc memory-related warnings that are printed when a plugin is unloaded (and thus cannot resolve addresses automatically)
  • Revision #4082 Mon 2014-02-17 11:10:18 +0100
    • mtr: smarter check for usable ipv6. Handles the case of sysctl net.ipv6.conf.all.disable_ipv6=1 net.ipv6.conf.default.disable_ipv6=1
  • Revision #4081 Mon 2014-02-17 11:10:14 +0100
    • errmsg.sys files are located in the builddir, not in the srcdir
  • Revision #4080 Mon 2014-02-17 11:10:03 +0100
    • MDEV-5579 rpm postun scriptlet leaks exit code to rpm
  • Revision #4079 Mon 2014-02-17 11:09:57 +0100
    • MDEV-5613 m_string.h exports generic function names without a namespace prefix, like str2int
  • Revision #4078 Mon 2014-02-17 11:09:50 +0100
    • RPM: fix alternative provides/obsoletes - put the correct architecture and version
  • Revision #4077 Mon 2014-02-17 11:09:45 +0100
    • MDEV-5436 mysql_config returns non-zero when running without parameters
  • Revision #4076 Tue 2014-02-18 10:54:05 +0400
    • send_eval may free evaluated query buffer before connection thread actually consumed it. With this patch evaluated query buffer is freed along with query buffer.
  • Revision #4075 Mon 2014-02-17 23:20:07 +0400
    • Post-merge fixes: merge MariaDB's fix for MDEV-5177 and MDEV-5555 with Oracle's fix for Bug#17588348 by reverting Oracle's fix.
  • Revision #4074 [merge] Mon 2014-02-17 18:53:54 +0400
    • Merge
    • Revision #4066.1.1 Mon 2014-02-17 18:50:54 +0400
      • MDEV-5177: ha_partition and innodb index intersection produce fewer rows (MySQL Bug#70703) MDEV-5555: Incorrect index_merge on BTREE indices - In ha_partition, make ordered index reads return rows in rowid order when index columns are the same.
  • Revision #4073 Mon 2014-02-17 11:09:40 +0100
    • don't open and fill all I_S tables for SELECT * FROM I_S.TRIGGERS
  • Revision #4072 Mon 2014-02-17 11:09:33 +0100
    • different fix for internal Oracle MySQL bug#16324629 that doesn't crash (simply, copied from FederatedX)
  • Revision #4071 Mon 2014-02-17 11:09:24 +0100
    • test case for MDEV-5689 ExtractValue(xml, 'substring(/x,/y)') crashes MySQL bug#12428404 MYSQLD.EXE CRASHES WHEN EXTRACTVALUE() IS CALLED WITH MALFORMED XPATH EXP
  • Revision #4070 [merge] Mon 2014-02-17 11:00:51 +0100
    • MySQL-5.5.36 merge (without few incorrect bugfixes and with 1250 files where only a copyright year was changed)
  • Revision #4069 Fri 2014-02-14 15:34:23 +0100
    • revert revno 4060:
  • Revision #4068 Fri 2014-02-14 15:16:23 +0100
    • fix SphinxSE to not leave Sphinx_error status variable uninitialized
  • Revision #4067 [merge] Fri 2014-02-14 14:09:29 +0100
  • Revision #4066 Thu 2014-02-13 16:41:08 +0400
    • MDEV-5616 - Deadlock between CREATE/DROP FUNCTION and SELECT from view
  • Revision #4065 Wed 2014-02-12 17:07:05 +0200
    • MDEV-5505: Assertion `! is_set()' fails on PREPARE SELECT with out of range in GROUP BY
  • Revision #4064 Tue 2014-02-11 21:43:08 +0200
    • Support 6 digit version numbers in executable comment syntax. This is needed to be able to ignore executable comments from version 10.0.
  • Revision #4063 Tue 2014-02-11 16:58:49 +0200
    • Fixed MDEV-5617: mysqld crashes when running a query with ONLY_FULL_GROUP_BY Problem was that we used cache_table in some cases where it was not initialized
  • Revision #4062 Tue 2014-02-11 16:57:28 +0200
    • Fixed failing test case
  • Revision #4061 Tue 2014-02-11 16:56:38 +0200
    • Set default progress report time to 5 seconds (Had accidently been set to 56 seconds in some merge)
  • Revision #4060 Tue 2014-02-11 14:03:13 +0200
    • Enable rpl_row_create_table (no reason to keep this disabled anymore)
  • Revision #4059 Tue 2014-02-11 14:02:42 +0200
    • Fix for MDEV-5629: Failing assertion: state == TRX_STATE_NOT_STARTED on concurrent CREATE OR REPLACE and transactional UPDATE
  • Revision #4058 Tue 2014-02-11 10:32:52 +0200
    • MDEV-5607: Query cache destroys uninitialized rwlock
  • Revision #4057 [merge] Mon 2014-02-10 20:34:52 -0800
    • Merge
    • Revision #4055.1.1 [merge] Mon 2014-02-10 17:00:51 -0800
      • Merge 5.3->5.5
      • Revision #2502.567.194 [merge] Fri 2014-02-07 16:55:25 -0800
        • Merge
        • Revision #2502.584.1 Fri 2014-02-07 15:50:36 -0800
          • Fixed bug MDEV-5611. The method Item_field::update_table_bitmaps() should not try to mark the bit for a self-referencing virtual column.
      • Revision #2502.567.193 [merge] Fri 2014-02-07 23:57:55 +0400
        • Merge
        • Revision #2502.583.1 Fri 2014-02-07 20:51:31 +0400
          • MDEV-5582: Plugin 'MEMORY' has ref_count=1 after shutdown with materialization+semijoin - Let cleanup_empty_jtbm_semi_joins() walk into semi-join nests.
      • Revision #2502.567.192 Wed 2014-02-05 17:47:38 -0800
        • Fixed bug MDEV-5468. The field JOIN::select_lex->where should be updated after the call of remove_eq_conds() in the function make_join_statistics(). This matters for subselects.
  • Revision #4056 Mon 2014-02-10 23:53:04 +0400
    • Do not include .result~ files and such into packages
  • Revision #4055 Wed 2014-02-05 14:25:37 +0400
    • unix_socket fails in some build environments when $USER variable appears to be unset, or when it contains 'root' even though the user does not have real root permissions
  • Revision #4054 Sat 2014-02-01 02:41:12 +0400
    • Increment the version number
  • Revision #4053 [merge] Wed 2014-01-29 00:19:53 +0200

Comments

Comments loading...