<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"><title>Appendix D. MySQL Community Server Enhancements and Release Notes</title><link rel="stylesheet" href="mysql-html.css" type="text/css"><meta name="generator" content="DocBook XSL Stylesheets V1.69.1"><link rel="start" href="index.html" title="MySQL 5.0 Reference Manual"><link rel="up" href="index.html" title="MySQL 5.0 Reference Manual"><link rel="prev" href="releasenotes-es.html" title="Appendix C. MySQL Enterprise Release Notes"><link rel="next" href="news.html" title="Appendix E. MySQL Change History"></head><body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="3" align="center">Appendix D. MySQL Community Server Enhancements and Release Notes</th></tr><tr><td width="20%" align="left"><a accesskey="p" href="releasenotes-es.html">Prev</a> </td><th width="60%" align="center"> </th><td width="20%" align="right"> <a accesskey="n" href="news.html">Next</a></td></tr></table><hr></div><div class="appendix" lang="en"><div class="titlepage"><div><div><h2 class="title"><a name="releasenotes-cs"></a>Appendix D. MySQL Community Server Enhancements and Release Notes</h2></div></div></div><div class="toc"><p><b>Table of Contents</b></p><dl><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0">D.1. MySQL Community Server 5.0 Enhancements and Release Notes</a></span></dt><dd><dl><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-51">D.1.1. Release Notes for MySQL Community Server 5.0.51 (15 November 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-45">D.1.2. Release Notes for MySQL Community Server 5.0.45 (04 July 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-41">D.1.3. Release Notes for MySQL Community Server 5.0.41 (01 May 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-37">D.1.4. Release Notes for MySQL Community Server 5.0.37 (27 February 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-33">D.1.5. Release Notes for MySQL Community Server 5.0.33 (09 January 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-27">D.1.6. Release Notes for MySQL Community Server 5.0.27 (21 October 2006)</a></span></dt></dl></dd></dl></div><a class="indexterm" name="id3236186"></a><a class="indexterm" name="id3236195"></a><a class="indexterm" name="id3236208"></a><a class="indexterm" name="id3236221"></a><a class="indexterm" name="id3236234"></a><a class="indexterm" name="id3236246"></a><a class="indexterm" name="id3236259"></a><p>
This appendix lists the enhancements and changes from version to
version in MySQL Community Server. This information is updated as
bugs are fixed and features are incorporated, so that everybody can
follow the development process.
</p><p>
Note that we tend to update the manual at the same time we make
changes to MySQL. If you find a recent version of MySQL listed here
that you can't find on our download page
(<a href="http://dev.mysql.com/downloads/" target="_top">http://dev.mysql.com/downloads/</a>), it means that the version has
not yet been released (and will normally be marked so in the
appropriate Release Note section).
</p><p>
The date mentioned with a release version is the date of the last
change done internally at MySQL AB (the BitKeeper ChangeSet) on
which the release was based, not the date when the packages were
made available. The binaries are usually made available a few days
after the date of the tagged ChangeSet, because building and testing
all packages takes some time.
</p><p>
For information on how to determine your current version and release
type, see <a href="installing.html#installation-version" title="2.2. Determining your current MySQL version">Section 2.2, “Determining your current MySQL version”</a>.
</p><div class="section" lang="en"><div class="titlepage"><div><div><h2 class="title" style="clear: both"><a name="releasenotes-cs-5-0"></a>D.1. MySQL Community Server 5.0 Enhancements and Release Notes</h2></div></div></div><div class="toc"><dl><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-51">D.1.1. Release Notes for MySQL Community Server 5.0.51 (15 November 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-45">D.1.2. Release Notes for MySQL Community Server 5.0.45 (04 July 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-41">D.1.3. Release Notes for MySQL Community Server 5.0.41 (01 May 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-37">D.1.4. Release Notes for MySQL Community Server 5.0.37 (27 February 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-33">D.1.5. Release Notes for MySQL Community Server 5.0.33 (09 January 2007)</a></span></dt><dt><span class="section"><a href="releasenotes-cs.html#releasenotes-cs-5-0-27">D.1.6. Release Notes for MySQL Community Server 5.0.27 (21 October 2006)</a></span></dt></dl></div><a class="indexterm" name="id3236350"></a><a class="indexterm" name="id3236363"></a><p>
This section documents all enhancements, changes, and bug fixes made
to MySQL Community Server from 5.0.27 on. For changes and bug fixes
to earlier versions, see <a href="news.html" title="Appendix E. MySQL Change History">Appendix E, <i>MySQL Change History</i></a>.
</p><div class="section" lang="en"><div class="titlepage"><div><div><h3 class="title"><a name="releasenotes-cs-5-0-51"></a>D.1.1. Release Notes for MySQL Community Server 5.0.51 (15 November 2007)</h3></div></div></div><p class="release-level">
This is a bugfix release for the current production release
family. It replaces MySQL 5.0.45.
</p><p>Functionality added or changed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
The parser accepted statements that contained <code class="literal">/* ...
*/</code> that were not properly closed with
<code class="literal">*/</code>, such as <code class="literal">SELECT 1 /* +
2</code>. Statements that contain unclosed
<code class="literal">/*</code>-comments now are rejected with a syntax
error.
</p><p>
This fix has the potential to cause incompatibilities. Because
of <a href="http://bugs.mysql.com/26302" target="_top">Bug#26302</a>, which caused the trailing <code class="literal">*/</code>
to be truncated from comments in views, stored routines,
triggers, and events, it is possible that objects of those types
may have been stored with definitions that now will be rejected
as syntactically invalid. Such objects should be dropped and
re-created so that their definitions do not contain truncated
comments. If a stored object definition contains only a single
statement (does not use a <code class="literal">BEGIN ... END</code>
block) and contains a comment within the statement, the comment
should be moved to follow the statement or the object should be
rewritten to use a <code class="literal">BEGIN ... END</code> block. For
example, this statement:
</p><pre class="programlisting">CREATE PROCEDURE p() SELECT 1 /* my comment */ ;
</pre><p>
Can be rewritten in either of these ways:
</p><pre class="programlisting">CREATE PROCEDURE p() SELECT 1; /* my comment */
CREATE PROCEDURE p() BEGIN SELECT 1 /* my comment */ ; END;
</pre><p>(<a href="http://bugs.mysql.com/28779" target="_top">Bug#28779</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Mapping of <code class="literal">NDB</code> error codes to MySQL storage
engine error codes has been improved.
(<a href="http://bugs.mysql.com/28423" target="_top">Bug#28423</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">auto_increment_increment</code> and
<code class="literal">auto_increment_offset</code> are now supported for
<code class="literal">NDB</code> tables.
(<a href="http://bugs.mysql.com/26342" target="_top">Bug#26342</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The output from the cluster management client showing the
progress of data node starts has been improved.
(<a href="http://bugs.mysql.com/23354" target="_top">Bug#23354</a>)</p></li><li><p>
Server parser performance was improved for expression parsing by
lowering the number of state transitions and reductions needed.
(<a href="http://bugs.mysql.com/30625" target="_top">Bug#30625</a>)</p></li><li><p>
Server parser performance was improved for boolean expressions.
(<a href="http://bugs.mysql.com/30237" target="_top">Bug#30237</a>)</p></li><li><p>
If a <code class="literal">MyISAM</code> table is created with no
<code class="literal">DATA DIRECTORY</code> option, the
<code class="filename">.MYD</code> file is created in the database
directory. By default, if <code class="literal">MyISAM</code> finds an
existing <code class="filename">.MYD</code> file in this case, it
overwrites it. The same applies to <code class="filename">.MYI</code>
files for tables created with no <code class="literal">INDEX
DIRECTORY</code> option. To suppress this behavior, start the
server with the new <code class="option">--keep_files_on_create</code>
option, in which case <code class="literal">MyISAM</code> will not
overwrite existing files and returns an error instead.
(<a href="http://bugs.mysql.com/29325" target="_top">Bug#29325</a>)</p></li><li><p>
If a <code class="literal">MERGE</code> table cannot be opened or used
because of a problem with an underlying table, <code class="literal">CHECK
TABLE</code> now displays information about which table
caused the problem.
(<a href="http://bugs.mysql.com/26976" target="_top">Bug#26976</a>)</p></li><li><p>
The <code class="literal">SQL_MODE</code>,
<code class="literal">FOREIGN_KEY_CHECKS</code>,
<code class="literal">UNIQUE_CHECKS</code>, character set/collations, and
<code class="literal">SQL_AUTO_IS_NULL</code> sesstion variables are
written to the binary log and honoured during replication. See
<a href="server-administration.html#binary-log" title="5.10.3. The Binary Log">Section 5.10.3, “The Binary Log”</a>.
</p></li><li><p>
The <code class="literal">EXAMPLE</code> storage engine is now enabled by
default.
</p></li></ul></div><p>Bugs fixed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Security Fix</strong></span>:
Using <code class="literal">RENAME TABLE</code> against a table with
explicit <code class="literal">DATA DIRECTORY</code> and <code class="literal">INDEX
DIRECTORY</code> options can be used to overwrite system
table information by replacing the symbolic link points. the
file to which the symlink points.
</p><p>
MySQL will now return an error when the file to which the
symlink points already exists.
(<a href="http://bugs.mysql.com/32111" target="_top">Bug#32111</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-5969" target="_top">CVE-2007-5969</a>)</p></li><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
The file <code class="filename">mysqld.exe</code> was mistakenly included
in binary distributions between MySQL 5.0.42 and 5.0.48. You
should use <code class="filename">mysqld-nt.exe</code>.
(<a href="http://bugs.mysql.com/32197" target="_top">Bug#32197</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>: <span class="bold"><strong>Packaging</strong></span>:
Some commercial MySQL Cluster RPM packages included support for
the <code class="literal">InnoDB</code> storage engine.
(<code class="literal">InnoDB</code> is not part of the standard
commercial MySQL Cluster offering.)
(<a href="http://bugs.mysql.com/31989" target="_top">Bug#31989</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Attempting to restore a backup made on a cluster host using one
endian to a machine using the other endian could cause the
cluster to fail.
(<a href="http://bugs.mysql.com/29674" target="_top">Bug#29674</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
When restarting a data node, queries could hang during that
node's start phase 5, and continue only after the node had
entered phase 6.
(<a href="http://bugs.mysql.com/29364" target="_top">Bug#29364</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Replica redo logs were inconsistently handled during a system
restart.
(<a href="http://bugs.mysql.com/29354" target="_top">Bug#29354</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Reads on <code class="literal">BLOB</code> columns were not locked when
they needed to be to guarantee consistency.
(<a href="http://bugs.mysql.com/29102" target="_top">Bug#29102</a>)</p><p>See also <a href="http://bugs.mysql.com/31482" target="_top">Bug#31482</a></p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A query using joins between several large tables and requiring
unique index lookups failed to complete, eventually returning
<span class="errortext">Uknown Error</span> after a very long period of
time. This occurred due to inadequate handling of instances
where the Transaction Coordinator ran out of
<code class="literal">TransactionBufferMemory</code>, when the cluster
should have returned NDB error code 4012 (<span class="errortext">Request
ndbd time-out</span>).
(<a href="http://bugs.mysql.com/28804" target="_top">Bug#28804</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The description of the <code class="option">--print</code> option provided
in the output from <span><strong class="command">ndb_restore <code class="option">--help</code>
</strong></span> was incorrect.
(<a href="http://bugs.mysql.com/27683" target="_top">Bug#27683</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The management client's response to <code class="literal">START BACKUP
WAIT COMPLETED</code> did not include the backup ID.
(<a href="http://bugs.mysql.com/27640" target="_top">Bug#27640</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
An invalid subselect on an <code class="literal">NDB</code> table could
cause <span><strong class="command">mysqld</strong></span> to crash.
(<a href="http://bugs.mysql.com/27494" target="_top">Bug#27494</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
An attempt to perform a <code class="literal">SELECT ... FROM
INFORMATION_SCHEMA.TABLES</code> whose result included
information about <code class="literal">NDB</code> tables for which the
user had no privileges crashed the MySQL Server on which the
query was performed.
(<a href="http://bugs.mysql.com/26793" target="_top">Bug#26793</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Warnings and errors generated by <span><strong class="command">ndb_config
<code class="option">--config-file=<em class="replaceable"><code>file</code></em></code></strong></span>
were sent to <code class="literal">stdout</code>, rather than to
<code class="literal">stderr</code>.
(<a href="http://bugs.mysql.com/25941" target="_top">Bug#25941</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Large file support did not work in AIX server binaries.
(<a href="http://bugs.mysql.com/10776" target="_top">Bug#10776</a>)</p></li><li><p>
When a <code class="literal">TIMESTAMP</code> with a non-zero time part
was converted to a <code class="literal">DATE</code> value, no warning was
generated. This caused index lookups to assume that this is a
valid conversion and was returning rows that match a comparison
between a <code class="literal">TIMESTAMP</code> value and a
<code class="literal">DATE</code> keypart. Now a warning is generated so
that <code class="literal">TIMESTAMP</code> with a non-zero time part will
not match <code class="literal">DATE</code> values.
(<a href="http://bugs.mysql.com/31221" target="_top">Bug#31221</a>)</p></li><li><p>
A server crash could occur when a
non-<code class="literal">DETERMINISTIC</code> stored function was used in
a <code class="literal">GROUP BY</code> clause.
(<a href="http://bugs.mysql.com/31035" target="_top">Bug#31035</a>)</p></li><li><p>
For an <code class="literal">InnoDB</code> table if a
<code class="literal">SELECT</code> was ordered by the primary key and
also had a <code class="literal">WHERE field = value</code> clause on a
different field that was indexed, a <code class="literal">DESC</code>
order instruction would be ignored.
(<a href="http://bugs.mysql.com/31001" target="_top">Bug#31001</a>)</p></li><li><p>
A failed <code class="literal">HANDLER ... READ</code> operation could
leave the table in a locked state.
(<a href="http://bugs.mysql.com/30632" target="_top">Bug#30632</a>)</p></li><li><p>
The optimization that uses a unique index to remove
<code class="literal">GROUP BY</code> did not ensure that the index was
actually used, thus violating the <code class="literal">ORDER BY</code>
that is implied by <code class="literal">GROUP BY</code>.
(<a href="http://bugs.mysql.com/30596" target="_top">Bug#30596</a>)</p></li><li><p>
<code class="literal">SHOW STATUS LIKE 'Ssl_cipher_list'</code> from a
MySQL client connected via SSL returned an empty string rather
than a list of available ciphers.
(<a href="http://bugs.mysql.com/30593" target="_top">Bug#30593</a>)</p></li><li><p>
Memory corruption occurred for some queries with a top-level
<code class="literal">OR</code> operation in the <code class="literal">WHERE</code>
condition if they contained equality predicates and other
sargable predicates in disjunctive parts of the condition.
(<a href="http://bugs.mysql.com/30396" target="_top">Bug#30396</a>)</p></li><li><p>
Issuing a <code class="literal">DELETE</code> statement having both an
<code class="literal">ORDER BY</code> clause and a
<code class="literal">LIMIT</code> clause could cause
<span><strong class="command">mysqld</strong></span> to crash.
(<a href="http://bugs.mysql.com/30385" target="_top">Bug#30385</a>)</p></li><li><p>
The <code class="literal">Last_query_cost</code> status variable value can
be computed accurately only for simple “<span class="quote">flat</span>”
queries, not complex queries such as those with subqueries or
<code class="literal">UNION</code>. However, the value was not
consistently being set to 0 for complex queries.
(<a href="http://bugs.mysql.com/30377" target="_top">Bug#30377</a>)</p></li><li><p>
Queries that had a <code class="literal">GROUP BY</code> clause and
selected <a href="functions.html#function_count"><code class="literal">COUNT(DISTINCT
<em class="replaceable"><code>bit_column</code></em>)</code></a> returned
incorrect results.
(<a href="http://bugs.mysql.com/30324" target="_top">Bug#30324</a>)</p></li><li><p>
The server created temporary tables for filesort operations in
the working directory, not in the directory specified by the
<code class="literal">tmpdir</code> system variable.
(<a href="http://bugs.mysql.com/30287" target="_top">Bug#30287</a>)</p></li><li><p>
The query cache does not support retrieval of statements for
which column level access control applies, but the server was
still caching such statements, thus wasting memory.
(<a href="http://bugs.mysql.com/30269" target="_top">Bug#30269</a>)</p></li><li><p>
Using <code class="literal">DISTINCT</code> or <code class="literal">GROUP BY</code>
on a <code class="literal">BIT</code> column in a
<code class="literal">SELECT</code> statement caused the column to be cast
internally as an integer, with incorrect results being returned
from the query.
(<a href="http://bugs.mysql.com/30245" target="_top">Bug#30245</a>)</p></li><li><p>
Multiple-table <code class="literal">DELETE</code> statements could delete
rows from the wrong table.
(<a href="http://bugs.mysql.com/30234" target="_top">Bug#30234</a>)</p></li><li><p>
<code class="literal">GROUP BY</code> on <code class="literal">BIT</code> columns
produced incorrect results.
(<a href="http://bugs.mysql.com/30219" target="_top">Bug#30219</a>)</p></li><li><p>
Using <code class="literal">KILL QUERY</code> or <code class="literal">KILL
CONNECTION</code> to kill a <code class="literal">SELECT</code>
statement caused a server crash if the query cache was enabled.
(<a href="http://bugs.mysql.com/30201" target="_top">Bug#30201</a>)</p></li><li><p>
Prepared statements containing
<a href="functions.html#function_connection-id"><code class="literal">CONNECTION_ID()</code></a> could be written
improperly to the binary log.
(<a href="http://bugs.mysql.com/30200" target="_top">Bug#30200</a>)</p></li><li><p>
When a thread executing a <code class="literal">DROP TABLE</code>
statement was killed, the table name locks that had been
acquired were not released.
(<a href="http://bugs.mysql.com/30193" target="_top">Bug#30193</a>)</p></li><li><p>
Short-format <span><strong class="command">mysql</strong></span> commands embedded within
<code class="literal">/*! ... */</code> comments were parsed incorrectly
by <span><strong class="command">mysql</strong></span>, which discarded the rest of the
comment including the terminating <code class="literal">*/</code>
characters. The result was a malformed (unclosed) comment. Now
<span><strong class="command">mysql</strong></span> does not discard the
<code class="literal">*/</code> characters.
(<a href="http://bugs.mysql.com/30164" target="_top">Bug#30164</a>)</p></li><li><p>
When <span><strong class="command">mysqldump</strong></span> wrote <code class="literal">DROP
DATABASE</code> statements within version-specific comments,
it included the terminating semicolon in the wrong place,
causing following statements to fail when the dump file was
reloaded.
(<a href="http://bugs.mysql.com/30126" target="_top">Bug#30126</a>)</p></li><li><p>
Use of local variables with non-ASCII names in stored procedures
crashed the server.
(<a href="http://bugs.mysql.com/30120" target="_top">Bug#30120</a>)</p></li><li><p>
On Windows, client libraries lacked symbols required for
linking.
(<a href="http://bugs.mysql.com/30118" target="_top">Bug#30118</a>)</p></li><li><p>
<code class="option">--myisam-recover=''</code> (empty option value) did
not disable <code class="literal">MyISAM</code> recovery.
(<a href="http://bugs.mysql.com/30088" target="_top">Bug#30088</a>)</p></li><li><p>
Statements within stored procedures ignored the value of the
<code class="literal">low_priority_updates</code> system variable.
(<a href="http://bugs.mysql.com/29963" target="_top">Bug#29963</a>)</p><p>See also <a href="http://bugs.mysql.com/26162" target="_top">Bug#26162</a></p></li><li><p>
For <code class="literal">MyISAM</code> tables on Windows,
<code class="literal">INSERT</code>, <code class="literal">DELETE</code>, or
<code class="literal">UPDATE</code> followed by <code class="literal">ALTER
TABLE</code> within <code class="literal">LOCK TABLES</code> could
cause table corruption.
(<a href="http://bugs.mysql.com/29957" target="_top">Bug#29957</a>)</p></li><li><p>
With auto-reconnect enabled, row fetching for a prepared
statement could crash after reconnect occurred because loss of
the statement handler was not accounted for.
(<a href="http://bugs.mysql.com/29948" target="_top">Bug#29948</a>)</p></li><li><p>
LOCK TABLES did not pre-lock tables used in triggers of the
locked tables. Unexpected locking behavior and statement
failures similar to <span class="errortext">failed: 1100: Table
'<em class="replaceable"><code>xx</code></em>' was not locked with LOCK
TABLES</span> could result.
(<a href="http://bugs.mysql.com/29929" target="_top">Bug#29929</a>)</p></li><li><p>
<code class="literal">INSERT ... VALUES(CONNECTION_ID(), ...)</code>
statements were written to the binary log in such a way that
they could not be properly restored.
(<a href="http://bugs.mysql.com/29928" target="_top">Bug#29928</a>)</p></li><li><p>
Adding <code class="literal">DISTINCT</code> could cause incorrect rows to
appear in a query result.
(<a href="http://bugs.mysql.com/29911" target="_top">Bug#29911</a>)</p></li><li><p>
Using the <a href="functions.html#function_date"><code class="literal">DATE()</code></a> function in a
<code class="literal">WHERE</code> clause did not return any records after
encountering <code class="literal">NULL</code>. However, using
<code class="literal">TRIM</code> or <code class="literal">CAST</code> produced the
correct results.
(<a href="http://bugs.mysql.com/29898" target="_top">Bug#29898</a>)</p></li><li><p>
Very long prepared statements in stored procedures could cause a
server crash.
(<a href="http://bugs.mysql.com/29856" target="_top">Bug#29856</a>)</p></li><li><p>
If query execution involved a temporary table,
<a href="functions.html#function_group-concat"><code class="literal">GROUP_CONCAT()</code></a> could return a
result with an incorrect character set.
(<a href="http://bugs.mysql.com/29850" target="_top">Bug#29850</a>)</p></li><li><p>
If one thread was performing concurrent inserts, other threads
reading from the same table using equality key searches could
see the index values for new rows before the data values had
been written, leading to reports of table corruption.
(<a href="http://bugs.mysql.com/29838" target="_top">Bug#29838</a>)</p></li><li><p>
Repeatedly accessing a view in a stored procedure (for example,
in a loop) caused a small amount of memory to be allocated per
access. Although this memory is deallocated on disconnect, it
could be a problem for a long running stored procedures that
make repeated access of views.
(<a href="http://bugs.mysql.com/29834" target="_top">Bug#29834</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> produced output that incorrectly
discarded the <code class="literal">NO_AUTO_VALUE_ON_ZERO</code> value of
the <code class="literal">SQL_MODE</code> variable after dumping triggers.
(<a href="http://bugs.mysql.com/29788" target="_top">Bug#29788</a>)</p></li><li><p>
An assertion failure occurred within yaSSL for very long keys.
(<a href="http://bugs.mysql.com/29784" target="_top">Bug#29784</a>)</p></li><li><p>
For <code class="literal">MEMORY</code> tables, the
<code class="literal">index_merge</code> union access method could return
incorrect results.
(<a href="http://bugs.mysql.com/29740" target="_top">Bug#29740</a>)</p></li><li><p>
Comparison of <code class="literal">TIME</code> values using the
<a href="functions.html#operator_between"><code class="literal">BETWEEN</code></a> operator led to string
comparison, producing incorrect results in some cases. Now the
values are compared as integers.
(<a href="http://bugs.mysql.com/29739" target="_top">Bug#29739</a>)</p></li><li><p>
The thread ID was not reset properly after execution of
<a href="apis.html#mysql-change-user" title="23.2.3.3. mysql_change_user()"><code class="literal">mysql_change_user()</code></a>, which
could cause replication failure when replicating temporary
tables.
(<a href="http://bugs.mysql.com/29734" target="_top">Bug#29734</a>)</p></li><li><p>
For a table with a <code class="literal">DATE</code> column
<em class="replaceable"><code>date_col</code></em> such that selecting rows
with <code class="literal">WHERE <em class="replaceable"><code>date_col</code></em> =
'<em class="replaceable"><code>date_val</code></em> 00:00:00'</code> yielded
a non-empty result, adding <code class="literal">GROUP BY
<em class="replaceable"><code>date_col</code></em> </code> caused the result
to be empty.
(<a href="http://bugs.mysql.com/29729" target="_top">Bug#29729</a>)</p></li><li><p>
In some cases, <code class="literal">INSERT INTO ... SELECT ... GROUP
BY</code> could insert rows even if the
<code class="literal">SELECT</code> by itself produced an empty result.
(<a href="http://bugs.mysql.com/29717" target="_top">Bug#29717</a>)</p></li><li><p>
For the embedded server, the
<a href="apis.html#mysql-stmt-store-result" title="23.2.7.27. mysql_stmt_store_result()"><code class="literal">mysql_stmt_store_result()</code></a> C API
function caused a memory leak for empty result sets.
(<a href="http://bugs.mysql.com/29687" target="_top">Bug#29687</a>)</p></li><li><p>
<code class="literal">EXPLAIN</code> produced <code class="literal">Impossible
where</code> for statements of the form <code class="literal">SELECT ...
FROM t WHERE c=0</code>, where <code class="literal">c</code> was an
<code class="literal">ENUM</code> column defined as a primary key.
(<a href="http://bugs.mysql.com/29661" target="_top">Bug#29661</a>)</p></li><li><p>
On Windows, <code class="literal">ALTER TABLE</code> hung if records were
locked in share mode by a long-running transaction.
(<a href="http://bugs.mysql.com/29644" target="_top">Bug#29644</a>)</p></li><li><p>
A left join between two views could produce incorrect results.
(<a href="http://bugs.mysql.com/29604" target="_top">Bug#29604</a>)</p></li><li><p>
Certain statements with unions, subqueries, and joins could
result in huge memory consumption.
(<a href="http://bugs.mysql.com/29582" target="_top">Bug#29582</a>)</p></li><li><p>
Clients using SSL could hang the server.
(<a href="http://bugs.mysql.com/29579" target="_top">Bug#29579</a>)</p></li><li><p>
A slave running with <code class="option">--log-slave-updates</code> would
fail to write <code class="literal">INSERT DELAY IGNORE</code> statements
to its binary log, resulting in different binary log contents on
the master and slave.
(<a href="http://bugs.mysql.com/29571" target="_top">Bug#29571</a>)</p></li><li><p>
An incorrect result was returned when comparing string values
that were converted to <code class="literal">TIME</code> values with
<a href="functions.html#function_cast"><code class="literal">CAST()</code></a>.
(<a href="http://bugs.mysql.com/29555" target="_top">Bug#29555</a>)</p></li><li><p>
<span><strong class="command">gcov</strong></span> coverage-testing information was not
written if the server crashed.
(<a href="http://bugs.mysql.com/29543" target="_top">Bug#29543</a>)</p></li><li><p>
Operations that used the time zone replicated the time zone only
for successful operations, but did not replicate the time zone
for errors that need to know it.
(<a href="http://bugs.mysql.com/29536" target="_top">Bug#29536</a>)</p></li><li><p>
Conversion of ASCII DEL (<code class="literal">0x7F</code>) to Unicode
incorrectly resulted in QUESTION MARK (<code class="literal">0x3F</code>)
rather than DEL.
(<a href="http://bugs.mysql.com/29499" target="_top">Bug#29499</a>)</p></li><li><p>
A field packet with <code class="literal">NULL</code> fields caused a
<code class="literal">libmysqlclient</code> crash.
(<a href="http://bugs.mysql.com/29494" target="_top">Bug#29494</a>)</p></li><li><p>
When using a combination of <code class="literal">HANDLER... READ</code>
and <code class="literal">DELETE</code> on a table, MySQL continued to
open new copies of the table every time, leading to an
exhaustion of file descriptors.
(<a href="http://bugs.mysql.com/29474" target="_top">Bug#29474</a>)</p><p>This regression was introduced by <a href="http://bugs.mysql.com/21587" target="_top">Bug#21587</a></p></li><li><p>
On Windows, the <span><strong class="command">mysql</strong></span> client died if the user
entered a statement and Return after entering Control-C.
(<a href="http://bugs.mysql.com/29469" target="_top">Bug#29469</a>)</p></li><li><p>
Failure to consider collation when comparing space characters
could lead to incorrect index entry order, making it impossible
to find some index values.
(<a href="http://bugs.mysql.com/29461" target="_top">Bug#29461</a>)</p></li><li><p>
Corrupt data resulted from use of <code class="literal">SELECT ... INTO
OUTFILE '<em class="replaceable"><code>file_name</code></em>' FIELDS ENCLOSED
BY '<em class="replaceable"><code>c</code></em>'</code>, where
<em class="replaceable"><code>c</code></em> is a digit or minus sign, followed
by <code class="literal">LOAD DATA INFILE
'<em class="replaceable"><code>file_name</code></em>' FIELDS ENCLOSED BY
'<em class="replaceable"><code>c</code></em>'</code>.
(<a href="http://bugs.mysql.com/29442" target="_top">Bug#29442</a>)</p></li><li><p>
Killing an <code class="literal">INSERT DELAYED</code> thread caused a
server crash.
(<a href="http://bugs.mysql.com/29431" target="_top">Bug#29431</a>)</p></li><li><p>
Use of <code class="literal">SHOW BINLOG EVENTS</code> for a non-existent
log file followed by <code class="literal">PURGE MASTER LOGS</code> caused
a server crash.
(<a href="http://bugs.mysql.com/29420" target="_top">Bug#29420</a>)</p></li><li><p>
Assertion failure could occur for grouping queries that employed
<code class="literal">DECIMAL</code> user variables with assignments to
them.
(<a href="http://bugs.mysql.com/29417" target="_top">Bug#29417</a>)</p></li><li><p>
For <a href="functions.html#function_cast"><code class="literal">CAST(<em class="replaceable"><code>expr</code></em> AS
DECIMAL(<em class="replaceable"><code>M</code></em>,<em class="replaceable"><code>D</code></em>))</code></a>,
the limits of 65 and 30 on the precision
(<em class="replaceable"><code>M</code></em>) and scale
(<em class="replaceable"><code>D</code></em>) were not enforced.
(<a href="http://bugs.mysql.com/29415" target="_top">Bug#29415</a>)</p></li><li><p>
If a view used a function in its <code class="literal">SELECT</code>
statement, the columns from the view were not inserted into the
<code class="literal">INFORMATION_SCHEMA.COLUMNS</code> table.
(<a href="http://bugs.mysql.com/29408" target="_top">Bug#29408</a>)</p></li><li><p>
Results for a select query that aliases the column names against
a view could duplicate one column while omitting another. This
bug could occur for a query over a multiple-table view that
includes an <code class="literal">ORDER BY</code> clause in its
definition.
(<a href="http://bugs.mysql.com/29392" target="_top">Bug#29392</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> created a stray file when a given a
too-long filename argument.
(<a href="http://bugs.mysql.com/29361" target="_top">Bug#29361</a>)</p></li><li><p>
The special “<span class="quote">zero</span>” <code class="literal">ENUM</code> value
was coerced to the normal empty string <code class="literal">ENUM</code>
value during a column-to-column copy. This affected
<code class="literal">CREATE ... SELECT</code> statements and
<code class="literal">SELECT</code> statements with aggregate functions on
<code class="literal">ENUM</code> columns in the <code class="literal">GROUP
BY</code> clause.
(<a href="http://bugs.mysql.com/29360" target="_top">Bug#29360</a>)</p></li><li><p>
Optimization of queries with <code class="literal">DETERMINISTIC</code>
stored functions in the <code class="literal">WHERE</code> clause was
ineffective: A sequential scan was always used.
(<a href="http://bugs.mysql.com/29338" target="_top">Bug#29338</a>)</p></li><li><p>
<code class="literal">MyISAM</code> corruption could occur with the
<code class="literal">cp932_japanese_ci</code> collation for the
<code class="literal">cp932</code> character set due to incorrect
comparison for trailing space.
(<a href="http://bugs.mysql.com/29333" target="_top">Bug#29333</a>)</p></li><li><p>
The <a href="apis.html#mysql-list-fields" title="23.2.3.42. mysql_list_fields()"><code class="literal">mysql_list_fields()</code></a> C API
function incorrectly set
<code class="literal">MYSQL_FIELD::decimals</code> for some view columns.
(<a href="http://bugs.mysql.com/29306" target="_top">Bug#29306</a>)</p></li><li><p>
<code class="literal">FULLTEXT</code> indexes could be corrupted by
certain <code class="literal">gbk</code> characters.
(<a href="http://bugs.mysql.com/29299" target="_top">Bug#29299</a>)</p></li><li><p>
<code class="literal">SELECT ... INTO OUTFILE</code> followed by
<code class="literal">LOAD DATA</code> could result in garbled characters
when the <code class="literal">FIELDS ENCLOSED BY</code> clause named a
delimiter of <code class="literal">'0'</code>, <code class="literal">'b'</code>,
<code class="literal">'n'</code>, <code class="literal">'r'</code>,
<code class="literal">'t'</code>, <code class="literal">'N'</code>, or
<code class="literal">'Z'</code> due to an interaction of character
encoding and doubling for data values containing the enclosed-by
character.
(<a href="http://bugs.mysql.com/29294" target="_top">Bug#29294</a>)</p></li><li><p>
Sort order of the collation wasn't used when comparing trailing
spaces. This could lead to incorrect comparison results,
incorrectly created indexes, or incorrect result set order for
queries that include an <code class="literal">ORDER BY</code> clause.
(<a href="http://bugs.mysql.com/29261" target="_top">Bug#29261</a>)</p></li><li><p>
If an <code class="literal">ENUM</code> column contained
<code class="literal">''</code> as one of its members (represented with
numeric value greater than 0), and the column contained error
values (represented as 0 and displayed as
<code class="literal">''</code>), using <code class="literal">ALTER TABLE</code> to
modify the column definition caused the 0 values to be given the
numeric value of the non-zero <code class="literal">''</code> member.
(<a href="http://bugs.mysql.com/29251" target="_top">Bug#29251</a>)</p></li><li><p>
Calling <a href="apis.html#mysql-options" title="23.2.3.49. mysql_options()"><code class="literal">mysql_options()</code></a> after
<a href="apis.html#mysql-real-connect" title="23.2.3.52. mysql_real_connect()"><code class="literal">mysql_real_connect()</code></a> could
cause clients to crash.
(<a href="http://bugs.mysql.com/29247" target="_top">Bug#29247</a>)</p></li><li><p>
<code class="literal">CHECK TABLE</code> for <code class="literal">ARCHIVE</code>
tables could falsely report table corruption or cause a server
crash.
(<a href="http://bugs.mysql.com/29207" target="_top">Bug#29207</a>)</p></li><li><p>
Mixing binary and <code class="literal">utf8</code> columns in a union
caused field lengths to be calculated incorrectly, resulting in
truncation.
(<a href="http://bugs.mysql.com/29205" target="_top">Bug#29205</a>)</p></li><li><p>
<code class="literal">AsText()</code> could fail with a buffer overrun.
(<a href="http://bugs.mysql.com/29166" target="_top">Bug#29166</a>)</p></li><li><p>
<code class="literal">InnoDB</code> refused to start on some versions of
FreeBSD with LinuxThreads. This is fixed by enabling file
locking on FreeBSD.
(<a href="http://bugs.mysql.com/29155" target="_top">Bug#29155</a>)</p></li><li><p>
<code class="literal">LOCK TABLES</code> was not atomic when more than one
<code class="literal">InnoDB</code> tables were locked.
(<a href="http://bugs.mysql.com/29154" target="_top">Bug#29154</a>)</p></li><li><p>
<code class="literal">INSERT DELAYED</code> statements on a master server
are replicated as non-<code class="literal">DELAYED</code> inserts on
slaves (which is normal, to preserve serialization), but the
inserts on the slave did not use concurrent inserts. Now
<code class="literal">INSERT DELAYED</code> on a slave is converted to a
concurrent insert when possible, and to a normal insert
otherwise.
(<a href="http://bugs.mysql.com/29152" target="_top">Bug#29152</a>)</p></li><li><p>
A network structure was initialized incorrectly, leading to
embedded server crashes.
(<a href="http://bugs.mysql.com/29117" target="_top">Bug#29117</a>)</p></li><li><p>
An assertion failure occurred if a query contained a conjunctive
predicate of the form <code class="literal">
<em class="replaceable"><code>view_column</code></em> = constant</code> in
the <code class="literal">WHERE</code> clause and the <code class="literal">GROUP
BY</code> clause contained a reference to a different view
column. The fix also enables application of an optimization that
was being skipped if a query contained a conjunctive predicate
of the form <code class="literal"> <em class="replaceable"><code>view_column</code></em> =
constant</code> in the <code class="literal">WHERE</code> clause and
the <code class="literal">GROUP BY</code> clause contained a reference to
the same view column.
(<a href="http://bugs.mysql.com/29104" target="_top">Bug#29104</a>)</p></li><li><p>
A maximum of 4TB <code class="literal">InnoDB</code> free space was
reported by <code class="literal">SHOW TABLE STATUS,</code> which is
incorrect on systems with more than 4TB space.
(<a href="http://bugs.mysql.com/29097" target="_top">Bug#29097</a>)</p></li><li><p>
If an <code class="literal">INSERT INTO ... SELECT</code> statement
inserted into the same table that the <code class="literal">SELECT</code>
retrieved from, and the <code class="literal">SELECT</code> included
<code class="literal">ORDER BY</code> and <code class="literal">LIMIT</code>
clauses, different data was inserted than the data produced by
the <code class="literal">SELECT</code> executed by itself.
(<a href="http://bugs.mysql.com/29095" target="_top">Bug#29095</a>)</p></li><li><p>
Queries that performed a lookup into a <code class="literal">BINARY</code>
index containing key values ending with spaces caused an
assertion failure for debug builds and incorrect results for
non-debug builds.
(<a href="http://bugs.mysql.com/29087" target="_top">Bug#29087</a>)</p></li><li><p>
The semantics of <code class="literal">BIGINT</code> depended on
platform-specific characteristics.
(<a href="http://bugs.mysql.com/29079" target="_top">Bug#29079</a>)</p></li><li><p>
A byte-order issue in writing a spatial index to disk caused bad
index files on some systems.
(<a href="http://bugs.mysql.com/29070" target="_top">Bug#29070</a>)</p></li><li><p>
If one of the queries in a <code class="literal">UNION</code> used the
<code class="literal">SQL_CACHE</code> option and another query in the
<code class="literal">UNION</code> contained a nondeterministic function,
the result was still cached. For example, this query was
incorrectly cached:
</p><pre class="programlisting">
SELECT NOW() FROM t1 UNION SELECT SQL_CACHE 1 FROM t1;
</pre><p>
(<a href="http://bugs.mysql.com/29053" target="_top">Bug#29053</a>)</p></li><li><p>
Creation of a legal stored procedure could fail if no default
database had been selected.
(<a href="http://bugs.mysql.com/29050" target="_top">Bug#29050</a>)</p></li><li><p>
<code class="literal">DROP USER</code> statements that named multiple
users, only some of which could be dropped, were replicated
incorrectly.
(<a href="http://bugs.mysql.com/29030" target="_top">Bug#29030</a>)</p></li><li><p>
<code class="literal">REPLACE</code>, <code class="literal">INSERT IGNORE</code>,
and <code class="literal">UPDATE IGNORE</code> did not work for
<code class="literal">FEDERATED</code> tables.
(<a href="http://bugs.mysql.com/29019" target="_top">Bug#29019</a>)</p></li><li><p>
Inserting into <code class="literal">InnoDB</code> tables and executing
<code class="literal">RESET MASTER</code> in multiple threads cause
assertion failure in debug server binaries.
(<a href="http://bugs.mysql.com/28983" target="_top">Bug#28983</a>)</p></li><li><p>
For a <code class="literal">ucs2</code> column,
<a href="functions.html#function_group-concat"><code class="literal">GROUP_CONCAT()</code></a> did not convert
separators to the result character set before inserting them,
producing a result containing a mixture of two different
character sets.
(<a href="http://bugs.mysql.com/28925" target="_top">Bug#28925</a>)</p></li><li><p>
Queries using UDFs or stored functions were cached.
(<a href="http://bugs.mysql.com/28921" target="_top">Bug#28921</a>)</p></li><li><p>
For a join with <code class="literal">GROUP BY</code> and/or
<code class="literal">ORDER BY</code> and a view reference in the
<code class="literal">FROM</code> list, the query metadata erroneously
showed empty table aliases and database names for the view
columns.
(<a href="http://bugs.mysql.com/28898" target="_top">Bug#28898</a>)</p></li><li><p>
Coercion of ASCII values to character sets that are a superset
of ASCII sometimes was not done, resulting in <span class="errortext">illegal
mix of collations</span> errors. These cases now are
resolved using repertoire, a new string expression attribute
(see <a href="internationalization-localization.html#charset-repertoire" title="9.1.6. String Repertoire">Section 9.1.6, “String Repertoire”</a>).
(<a href="http://bugs.mysql.com/28875" target="_top">Bug#28875</a>)</p></li><li><p>
Non-<code class="literal">utf8</code> characters could get mangled when
stored in <code class="literal">CSV</code> tables.
(<a href="http://bugs.mysql.com/28862" target="_top">Bug#28862</a>)</p></li><li><p>
<code class="literal">ALTER VIEW</code> is not supported as a prepared
statement but was not being rejected. <code class="literal">ALTER
VIEW</code> is now prohibited as a prepared statement or when
called within stored routines.
(<a href="http://bugs.mysql.com/28846" target="_top">Bug#28846</a>)</p></li><li><p>
In strict SQL mode, errors silently stopped the SQL thread even
for errors named using the <code class="option">--slave-skip-errors</code>
option.
(<a href="http://bugs.mysql.com/28839" target="_top">Bug#28839</a>)</p></li><li><p>
Fast <code class="literal">ALTER TABLE</code> (that works without
rebuilding the table) acquired duplicate locks in the storage
engine. In <code class="literal">MyISAM</code>, if <code class="literal">ALTER
TABLE</code> was issued under <code class="literal">LOCK TABLE</code>,
it caused all data inserted after <code class="literal">LOCK TABLE</code>
to disappear.
(<a href="http://bugs.mysql.com/28838" target="_top">Bug#28838</a>)</p></li><li><p>
Killing an SSL connection on platforms where MySQL is compiled
with <code class="literal">-DSIGNAL_WITH_VIO_CLOSE</code> (Windows, Mac OS
X, and some others) could crash the server.
(<a href="http://bugs.mysql.com/28812" target="_top">Bug#28812</a>)</p></li><li><p>
Runtime changes to the
<code class="literal">log_queries_not_using_indexes</code> system variable
were ignored.
(<a href="http://bugs.mysql.com/28808" target="_top">Bug#28808</a>)</p></li><li><p>
Tables using the <code class="literal">InnoDB</code> storage engine
incremented <code class="literal">AUTO_INCREMENT</code> values incorrectly
with <code class="literal">ON DUPLICATE KEY UPDATE</code>.
(<a href="http://bugs.mysql.com/28781" target="_top">Bug#28781</a>)</p></li><li><p>
Selecting a column not present in the selected-from table caused
an extra error to be produced by <code class="literal">SHOW ERRORS</code>.
(<a href="http://bugs.mysql.com/28677" target="_top">Bug#28677</a>)</p></li><li><p>
For a statement of the form <code class="literal">CREATE t1 SELECT
<em class="replaceable"><code>integer_constant</code></em> </code>, the
server created the column using the <code class="literal">DECIMAL</code>
data type for large negative values that are within the range of
<code class="literal">BIGINT</code>.
(<a href="http://bugs.mysql.com/28625" target="_top">Bug#28625</a>)</p></li><li><p>
For <code class="literal">InnoDB</code> tables, MySQL unnecessarily sorted
records in certain cases when the records were retrieved by
<code class="literal">InnoDB</code> in the proper order already.
(<a href="http://bugs.mysql.com/28591" target="_top">Bug#28591</a>)</p></li><li><p>
A <code class="literal">SELECT</code> in one connection could be blocked
by <code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code> in
another connection even when
<code class="literal">low_priority_updates</code> is set.
(<a href="http://bugs.mysql.com/28587" target="_top">Bug#28587</a>)</p></li><li><p>
<span><strong class="command">mysql_install_db</strong></span> could fail to find script
files that it needs.
(<a href="http://bugs.mysql.com/28585" target="_top">Bug#28585</a>)</p></li><li><p>
When one thread attempts to lock two (or more) tables and
another thread executes a statement that aborts these locks
(such as <code class="literal">REPAIR TABLE</code>, <code class="literal">OPTIMIZE
TABLE</code>, or <code class="literal">CHECK TABLE</code>), the thread
might get a table object with an incorrect lock type in the
table cache. The result is table corruption or a server crash.
(<a href="http://bugs.mysql.com/28574" target="_top">Bug#28574</a>)</p></li><li><p>
<span><strong class="command">mysql_upgrade</strong></span> could run binaries dynamically
linked against incorrect versions of shared libraries.
(<a href="http://bugs.mysql.com/28560" target="_top">Bug#28560</a>)</p></li><li><p>
If a stored procedure was created and invoked prior to selecting
a default database with <code class="literal">USE</code>, a <span class="errortext">No
database selected</span> error occurred.
(<a href="http://bugs.mysql.com/28551" target="_top">Bug#28551</a>)</p></li><li><p>
On Mac OS X, shared-library installation pathnames were
incorrect.
(<a href="http://bugs.mysql.com/28544" target="_top">Bug#28544</a>)</p></li><li><p>
Using the <code class="option">--skip-add-drop-table </code>option with
<span><strong class="command">mysqldump</strong></span> generated incorrect SQL if the
database included any views. The recreation of views requires
the creation and removal of temporary tables. This option
suppressed the removal of those temporary tables. The same
applied to <code class="option">--compact</code> since this option also
invokes <code class="option">--skip-add-drop-table</code>.
(<a href="http://bugs.mysql.com/28524" target="_top">Bug#28524</a>)</p></li><li><p>
<span><strong class="command">mysqlbinlog --hexdump</strong></span> generated incorrect
output due to omission of the “<span class="quote"> <code class="literal">#</code>
</span>” comment character for some comment lines.
(<a href="http://bugs.mysql.com/28293" target="_top">Bug#28293</a>)</p></li><li><p>
The <code class="literal">IS_UPDATABLE</code> column in the
<code class="literal">INFORMATION_SCHEMA.VIEWS</code> table was not always
set correctly.
(<a href="http://bugs.mysql.com/28266" target="_top">Bug#28266</a>)</p></li><li><p>
A race condition in the interaction between
<code class="literal">MyISAM</code> and the query cache code caused the
query cache not to invalidate itself for concurrently inserted
data.
(<a href="http://bugs.mysql.com/28249" target="_top">Bug#28249</a>)</p></li><li><p>
Indexing column prefixes in <code class="literal">InnoDB</code> tables
could cause table corruption.
(<a href="http://bugs.mysql.com/28138" target="_top">Bug#28138</a>)</p></li><li><p>
Index creation could fail due to truncation of key values to the
maximum key length rather than to a mulitiple of the maximum
character length.
(<a href="http://bugs.mysql.com/28125" target="_top">Bug#28125</a>)</p></li><li><p>
The <a href="functions.html#function_locate"><code class="literal">LOCATE()</code></a> function returned
<code class="literal">NULL</code> if any of its arguments evaluated to
<code class="literal">NULL</code>. Likewise, the predicate,
<a href="functions.html#function_locate"><code class="literal">LOCATE(<em class="replaceable"><code>str</code></em>,NULL)
IS NULL</code></a>, erroneously evaluated to
<code class="literal">FALSE</code>.
(<a href="http://bugs.mysql.com/27932" target="_top">Bug#27932</a>)</p></li><li><p>
<code class="literal">SHOW COLUMNS</code> returned <code class="literal">NULL</code>
instead of the empty string for the <code class="literal">Default</code>
value of columns that had no default specified.
(<a href="http://bugs.mysql.com/27747" target="_top">Bug#27747</a>)</p></li><li><p>
The modification of a table by a partially completed
multi-column update was not recorded in the binlog, rather than
being marked by an event and a corresponding error code.
(<a href="http://bugs.mysql.com/27716" target="_top">Bug#27716</a>)</p></li><li><p>
With recent versions of DBD::mysql,
<span><strong class="command">mysqlhotcopy</strong></span> generated table names that were
doubly qualified with the database name.
(<a href="http://bugs.mysql.com/27694" target="_top">Bug#27694</a>)</p></li><li><p>
The anonymous accounts were not being created during MySQL
installation.
(<a href="http://bugs.mysql.com/27692" target="_top">Bug#27692</a>)</p></li><li><p>
Some <code class="literal">SHOW</code> statements and
<code class="literal">INFORMATION_SCHEMA</code> queries could expose
information not allowed by the user's access privileges.
(<a href="http://bugs.mysql.com/27629" target="_top">Bug#27629</a>)</p></li><li><p>
A stack overrun could occur when storing
<code class="literal">DATETIME</code> values using repeated prepared
statements.
(<a href="http://bugs.mysql.com/27592" target="_top">Bug#27592</a>)</p></li><li><p>
Dropping a user-defined function could cause a server crash if
the function was still in use by another thread.
(<a href="http://bugs.mysql.com/27564" target="_top">Bug#27564</a>)</p></li><li><p>
Some character mappings in the <code class="filename">ascii.xml</code>
file were incorrect.
(<a href="http://bugs.mysql.com/27562" target="_top">Bug#27562</a>)</p></li><li><p>
The parser rules for the <code class="literal">SHOW PROFILE</code>
statement were revised to work with older versions of
<span><strong class="command">bison</strong></span>.
(<a href="http://bugs.mysql.com/27433" target="_top">Bug#27433</a>)</p></li><li><p>
An error that happened inside <code class="literal">INSERT</code>,
<code class="literal">UPDATE</code>, or <code class="literal">DELETE</code>
statements performed from within a stored function or trigger
could cause inconsistency between master and slave servers.
(<a href="http://bugs.mysql.com/27417" target="_top">Bug#27417</a>)</p></li><li><p>
Fixed a case of unsafe aliasing in the source that caused a
client library crash when compiled with <span><strong class="command">gcc</strong></span> 4
at high optimization levels.
(<a href="http://bugs.mysql.com/27383" target="_top">Bug#27383</a>)</p></li><li><p>
A <code class="literal">SELECT</code> with more than 31 nested dependent
subqueries returned an incorrect result.
(<a href="http://bugs.mysql.com/27352" target="_top">Bug#27352</a>)</p></li><li><p>
Index-based range reads could fail for comparisons that involved
contraction characters (such as <code class="literal">ch</code> in Czech
or <code class="literal">ll</code> in Spanish).
(<a href="http://bugs.mysql.com/27345" target="_top">Bug#27345</a>)</p></li><li><p>
Aggregations in subqueries that refer to outer query columns
were not always correctly referenced to the proper outer query.
(<a href="http://bugs.mysql.com/27333" target="_top">Bug#27333</a>)</p></li><li><p>
<code class="literal">INSERT INTO ... SELECT</code> caused a crash if
<code class="literal">innodb_locks_unsafe_for_binlog</code> was enabled.
(<a href="http://bugs.mysql.com/27294" target="_top">Bug#27294</a>)</p></li><li><p>
Error returns from the <code class="literal">time()</code> system call
were ignored.
(<a href="http://bugs.mysql.com/27198" target="_top">Bug#27198</a>)</p></li><li><p>
Phantom reads could occur under <code class="literal">InnoDB</code>
serializable isolation level.
(<a href="http://bugs.mysql.com/27197" target="_top">Bug#27197</a>)</p></li><li><p>
The <a href="functions.html#function_substring"><code class="literal">SUBSTRING()</code></a> function
returned the entire string instead of an empty string when it
was called from a stored procedure and when the length parameter
was specified by a variable with the value “<span class="quote">
<code class="literal">0</code> </span>”.
(<a href="http://bugs.mysql.com/27130" target="_top">Bug#27130</a>)</p></li><li><p>
<code class="literal">ALTER TABLE ... ENABLE KEYS</code> could cause
<span><strong class="command">mysqld</strong></span> to crash when executed on a table
containing on a <code class="literal">MyISAM</code> table containing
billions of rows.
(<a href="http://bugs.mysql.com/27029" target="_top">Bug#27029</a>)</p></li><li><p>
<code class="literal">FEDERATED</code> tables had an artificially low
maximum of key length.
(<a href="http://bugs.mysql.com/26909" target="_top">Bug#26909</a>)</p></li><li><p>
Binary content <code class="literal">0x00</code> in a
<code class="literal">BLOB</code> column sometimes became <code class="literal">0x5C
0x00</code> following a dump and reload, which could cause
problems with data using multi-byte character sets such as
<code class="literal">GBK</code> (Chinese). This was due to a problem with
<code class="literal">SELECT INTO OUTFILE</code> whereby <code class="literal">LOAD
DATA</code> later incorrectly interpreted
<code class="literal">0x5C</code> as the second byte of a multi-byte
sequence rather than as the <code class="literal">SOLIDUS</code>
(“<span class="quote">\</span>”) character, used by MySQL as the escape
character.
(<a href="http://bugs.mysql.com/26711" target="_top">Bug#26711</a>)</p></li><li><p>
Index creation could corrupt the table definition in the
<code class="filename">.frm</code> file: 1) A table with the maximum
number of key segments and maximum length key name would have a
corrupted <code class="filename">.frm</code> file, due to incorrect
calculation of the total key length. 2)
<code class="literal">MyISAM</code> would reject a table with the maximum
number of keys and the maximum number of key segments in all
keys. (It would allow one less than this total maximum.) Now
<code class="literal">MyISAM</code> accepts a table defined with the
maximum.
(<a href="http://bugs.mysql.com/26642" target="_top">Bug#26642</a>)</p></li><li><p>
After the first read of a <code class="literal">TEMPORARY</code> table,
<code class="literal">CHECK TABLE</code> could report the table as being
corrupt.
(<a href="http://bugs.mysql.com/26325" target="_top">Bug#26325</a>)</p></li><li><p>
If an operation had an <code class="literal">InnoDB</code> table, and two
triggers, <code class="literal">AFTER UPDATE</code> and <code class="literal">AFTER
INSERT</code>, competing for different resources (such as two
distinct <code class="literal">MyISAM</code> tables), the triggers were
unable to execute concurrently. In addition,
<code class="literal">INSERT</code> and <code class="literal">UPDATE</code>
statements for the <code class="literal">InnoDB</code> table were unable
to run concurrently.
(<a href="http://bugs.mysql.com/26141" target="_top">Bug#26141</a>)</p></li><li><p>
<code class="literal">ALTER DATABASE</code> did not require at least one
option.
(<a href="http://bugs.mysql.com/25859" target="_top">Bug#25859</a>)</p></li><li><p>
Using <code class="literal">HANDLER</code> to open a table having a
storage engine not supported by <code class="literal">HANDLER</code>
properly returned an error, but also improperly prevented the
table from being dropped by other connections.
(<a href="http://bugs.mysql.com/25856" target="_top">Bug#25856</a>)</p></li><li><p>
The index merge union access algorithm could produce incorrect
results with <code class="literal">InnoDB</code> tables. The problem could
also occur for queries that used <code class="literal">DISTINCT</code>.
(<a href="http://bugs.mysql.com/25798" target="_top">Bug#25798</a>)</p></li><li><p>
When using a <code class="literal">FEDERATED</code> table, the value of
<code class="literal">last_insert_id()</code> would not correctly update
the C API interface, which would affect the autogenerated ID
returned both through the C API and the MySQL protocol,
affecting Connectors that used the protocol and/or C API.
(<a href="http://bugs.mysql.com/25714" target="_top">Bug#25714</a>)</p></li><li><p>
The server was blocked from opening other tables while the
<code class="literal">FEDERATED</code> engine was attempting to open a
remote table. Now the server does not check the correctness of a
<code class="literal">FEDERATED</code> table at <code class="literal">CREATE
TABLE</code> time, but waits until the table actually is
accessed.
(<a href="http://bugs.mysql.com/25679" target="_top">Bug#25679</a>)</p></li><li><p>
Under ActiveState Perl, <code class="literal">mysql-test-run.pl</code>
could kill itself when attempting to kill other processes.
(<a href="http://bugs.mysql.com/25657" target="_top">Bug#25657</a>)</p></li><li><p>
Several <code class="literal">InnoDB</code> assertion failures were
corrected.
(<a href="http://bugs.mysql.com/25645" target="_top">Bug#25645</a>)</p></li><li><p>
A query with <code class="literal">DISTINCT</code> in the select list to
which the loose-scan optimization for grouping queries was
applied returned an incorrect result set when the query was used
with the <code class="literal">SQL_BIG_RESULT</code> option.
(<a href="http://bugs.mysql.com/25602" target="_top">Bug#25602</a>)</p></li><li><p>
For a multiple-row insert into a <code class="literal">FEDERATED</code>
table that refers to a remote transactional table, if the insert
failed for a row due to constraint failure, the remote table
would contain a partial commit (the rows preceding the failed
one) instead of rolling back the statement completely. This
occurred because the rows were treated as individual inserts.
</p><p>
Now <code class="literal">FEDERATED</code> performs bulk-insert handling
such that multiple rows are sent to the remote table in a batch.
This provides a performance improvement and enables the remote
table to perform statement rollback properly should an error
occur. This capability has the following limitations:
</p><div class="itemizedlist"><ul type="circle"><li><p>
The size of the insert cannot exceed the maximum packet size
between servers. If the insert exceeds this size, it is
broken into multiple packets and the rollback problem can
occur.
</p></li><li><p>
Bulk-insert handling does not occur for <code class="literal">INSERT ...
ON DUPLICATE KEY UPDATE</code>.
</p></li></ul></div><p>(<a href="http://bugs.mysql.com/25513" target="_top">Bug#25513</a>)</p></li><li><p>
The <code class="literal">FEDERATED</code> storage engine failed silently
for <code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code> if a
duplicate key violation occurred. <code class="literal">FEDERATED</code>
does not support <code class="literal">ON DUPLICATE KEY UPDATE</code>, so
now it correctly returns an <code class="literal">ER_DUP_KEY</code> error
if a duplicate key violation occurs.
(<a href="http://bugs.mysql.com/25511" target="_top">Bug#25511</a>)</p></li><li><p>
For <code class="literal">InnoDB</code> tables, <code class="literal">CREATE TABLE a AS
SELECT * FROM A</code> would fail.
(<a href="http://bugs.mysql.com/25164" target="_top">Bug#25164</a>)</p></li><li><p>
In a stored function or trigger, when <code class="literal">InnoDB</code>
detected deadlock, it attempted rollback and displayed an
incorrect error message (<span class="errortext">Explicit or implicit commit
is not allowed in stored function or trigger</span>). Now
<code class="literal">InnoDB</code> returns an error under these
conditions and does not attempt rollback. Rollback is handled
outside of <code class="literal">InnoDB</code> above the function/trigger
level.
(<a href="http://bugs.mysql.com/24989" target="_top">Bug#24989</a>)</p></li><li><p>
A too-long <code class="literal">shared-memory-base-name</code> value
could cause a buffer overflow and crash the server or clients.
(<a href="http://bugs.mysql.com/24924" target="_top">Bug#24924</a>)</p></li><li><p>
Dropping a temporary <code class="literal">InnoDB</code> table that had
been locked with <code class="literal">LOCK TABLES</code> caused a server
crash.
(<a href="http://bugs.mysql.com/24918" target="_top">Bug#24918</a>)</p></li><li><p>
On Windows, executables did not include Vista manifests.
(<a href="http://bugs.mysql.com/24732" target="_top">Bug#24732</a>)</p></li><li><p>
Slave servers could incorrectly interpret an out-of-memory error
from the master and reconnect using the wrong binary log
position.
(<a href="http://bugs.mysql.com/24192" target="_top">Bug#24192</a>)</p></li><li><p>
If MySQL/<code class="literal">InnoDB</code> crashed very quickly after
starting up, it would not force a checkpoint. In this case,
<code class="literal">InnoDB</code> would skip crash recovery at next
startup, and the database would become corrupt. Fix: If the redo
log scan at <code class="literal">InnoDB</code> startup goes past the last
checkpoint, force crash recovery.
(<a href="http://bugs.mysql.com/23710" target="_top">Bug#23710</a>)</p></li><li><p>
The server deducted some bytes from the
<code class="literal">key_cache_block_size</code> option value and reduced
it to the next lower 512 byte boundary. The resulting block size
was not a power of two. Setting the
<code class="literal">key_cache_block_size</code> system variable to a
value that is not a power of two resulted in
<code class="literal">MyISAM</code> table corruption.
(<a href="http://bugs.mysql.com/23068" target="_top">Bug#23068</a>, <a href="http://bugs.mysql.com/28478" target="_top">Bug#28478</a>, <a href="http://bugs.mysql.com/25853" target="_top">Bug#25853</a>)</p></li><li><p>
<code class="literal">SHOW INNODB STATUS</code> caused an assertion
failure under high load.
(<a href="http://bugs.mysql.com/22819" target="_top">Bug#22819</a>)</p></li><li><p>
<code class="literal">SHOW BINLOG EVENTS</code> displayed incorrect values
of <code class="literal">End_log_pos</code> for events associated with
transactional storage engines.
(<a href="http://bugs.mysql.com/22540" target="_top">Bug#22540</a>)</p></li><li><p>
A statement of the form <code class="literal">CREATE TABLE IF NOT EXISTS t1
SELECT f1() AS i</code> failed with a deadlock error if the
stored function <code class="literal">f1()</code> referred to a table with
the same name as the to-be-created table. Now it correctly
produces a message that the table already exists.
(<a href="http://bugs.mysql.com/22427" target="_top">Bug#22427</a>)</p></li><li><p>
Read lock requests that were blocked by a pending write lock
request were not allowed to proceed if the statement requesting
the write lock was killed.
(<a href="http://bugs.mysql.com/21281" target="_top">Bug#21281</a>)</p></li><li><p>
Under heavy load with a large query cache, invalidating part of
the cache could cause the server to freeze (that is, to be
unable to service other operations until the invalidation was
complete).
(<a href="http://bugs.mysql.com/21074" target="_top">Bug#21074</a>)</p></li><li><p>
<span><strong class="command">mysql-stress-test.pl</strong></span> and
<span><strong class="command">mysqld_multi.server.sh</strong></span> were missing from some
binary distributions.
(<a href="http://bugs.mysql.com/21023" target="_top">Bug#21023</a>, <a href="http://bugs.mysql.com/25486" target="_top">Bug#25486</a>)</p></li><li><p>
On Windows, the server used 10MB of memory for each connection
thread, resulting in memory exhaustion. Now each thread uses
1MB.
(<a href="http://bugs.mysql.com/20815" target="_top">Bug#20815</a>)</p></li><li><p>
Worked around an <span><strong class="command">icc</strong></span> problem with an
incorrect machine instruction being generated in the context of
software pre-fetching after a subroutine got in-lined.
(Upgrading to <span><strong class="command">icc</strong></span> 10.0.026 makes the
workaround unnecessary.)
(<a href="http://bugs.mysql.com/20803" target="_top">Bug#20803</a>)</p></li><li><p>
<code class="literal">InnoDB</code> produced an unnecessary (and harmless)
warning: <code class="literal">InnoDB: Error: trying to declare trx to enter
InnoDB, but InnoDB: it already is declared</code>.
(<a href="http://bugs.mysql.com/20090" target="_top">Bug#20090</a>)</p></li><li><p>
Under ActiveState Perl, <code class="literal">mysql-test-run.pl</code>
would not run.
(<a href="http://bugs.mysql.com/18415" target="_top">Bug#18415</a>)</p></li><li><p>
The server crashed when the size of an
<code class="literal">ARCHIVE</code> table grew larger than 2GB.
(<a href="http://bugs.mysql.com/15787" target="_top">Bug#15787</a>)</p></li><li><p>
<code class="literal">SQL_BIG_RESULT</code> had no effect for
<code class="literal">CREATE TABLE ... SELECT SQL_BIG_RESULT ...</code>
statements.
(<a href="http://bugs.mysql.com/15130" target="_top">Bug#15130</a>)</p></li><li><p>
On 64-bit Windows systems, the Config Wizard failed to complete
the setup because 64-bit Windows does not resolve dynamic
linking of the 64-bit <code class="filename">libmysql.dll</code> to a
32-bit application like the Config Wizard.
(<a href="http://bugs.mysql.com/14649" target="_top">Bug#14649</a>)</p></li><li><p>
<span><strong class="command">mysql_setpermission</strong></span> tried to grant
global-only privileges at the database level.
(<a href="http://bugs.mysql.com/14618" target="_top">Bug#14618</a>)</p></li><li><p>
Parameters of type <code class="literal">DATETIME</code> or
<code class="literal">DATE</code> in stored procedures were silently
converted to <code class="literal">VARBINARY</code>.
(<a href="http://bugs.mysql.com/13675" target="_top">Bug#13675</a>)</p></li><li><p>
For the general query log, logging of prepared statements
executed via the C API differed from logging of prepared
statements performed with <code class="literal">PREPARE</code> and
<code class="literal">EXECUTE</code>. Logging for the latter was missing
the <code class="literal">Prepare</code> and <code class="literal">Execute</code>
lines.
(<a href="http://bugs.mysql.com/13326" target="_top">Bug#13326</a>)</p></li><li><p>
The server returned data from <code class="literal">SHOW CREATE
TABLE</code> statement or a <code class="literal">SELECT</code>
statement on an INFORMATION_SCHEMA table using the
<code class="literal">binary</code> character set.
(<a href="http://bugs.mysql.com/10491" target="_top">Bug#10491</a>)</p></li><li><p>
Backup software can cause
<code class="literal">ERROR_SHARING_VIOLATION</code> or
<code class="literal">ERROR_LOCK_VIOLATION</code> conditions during file
operations. <code class="literal">InnoDB</code> now retries forever until
the condition goes away.
(<a href="http://bugs.mysql.com/9709" target="_top">Bug#9709</a>)</p></li><li><p>
Bulk-insert handling does not occur for <code class="literal">INSERT ... ON
DUPLICATE KEY UPDATE</code>.
</p></li><li><p>
The size of the insert cannot exceed the maximum packet size
between servers. If the insert exceeds this size, it is broken
into multiple packets and the rollback problem can occur.
</p></li></ul></div></div><div class="section" lang="en"><div class="titlepage"><div><div><h3 class="title"><a name="releasenotes-cs-5-0-45"></a>D.1.2. Release Notes for MySQL Community Server 5.0.45 (04 July 2007)</h3></div></div></div><p class="release-level">
This is a bugfix release for the current production release
family. It replaces MySQL 5.0.41.
</p><p>Functionality added or changed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
<code class="literal">INSERT DELAYED</code> is now downgraded to a normal
<code class="literal">INSERT</code> if the statement uses functions that
access tables or triggers, or that is called from a function or
a trigger.
</p><p>
This was done to resolve the following interrelated issues:
</p><div class="itemizedlist"><ul type="circle"><li><p>
The server could abort or deadlock for <code class="literal">INSERT
DELAYED</code> statements for which another insert was
performed implicitly (for example, via a stored function
that inserted a row).
</p></li><li><p>
A trigger using an <code class="literal">INSERT DELAYED</code>
caused the error <span class="errortext">INSERT DELAYED can't be used
with table ... because it is locked with LOCK
TABLES</span> although the target table was not
actually locked.
</p></li><li><p>
<code class="literal">INSERT DELAYED</code> into a table with a
<code class="literal">BEFORE INSERT</code> or <code class="literal">AFTER
INSERT</code> trigger gave an incorrect
<code class="literal">NEW</code> pseudocolumn value and caused the
server to deadlock or abort.
</p></li></ul></div><p>
(<a href="http://bugs.mysql.com/21483" target="_top">Bug#21483</a>)</p><p>See also <a href="http://bugs.mysql.com/20497" target="_top">Bug#20497</a>, <a href="http://bugs.mysql.com/21714" target="_top">Bug#21714</a></p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The server source tree now includes scripts to simplify building
MySQL with SCI support. For more information about SCI
interconnects and these build scripts, see
<a href="mysql-cluster.html#mysql-cluster-sci-sockets" title="16.9.1. Configuring MySQL Cluster to use SCI Sockets">Section 16.9.1, “Configuring MySQL Cluster to use SCI Sockets”</a>.
(<a href="http://bugs.mysql.com/25470" target="_top">Bug#25470</a>)</p></li><li><p>
Binaries for the Linux x86 statically linked
<code class="filename">tar.gz</code> Community package were linked
dynamically, not statically. Static linking has been re-enabled.
(<a href="http://bugs.mysql.com/29617" target="_top">Bug#29617</a>)</p></li><li><p>
Prior to this release, when <code class="literal">DATE</code> values were
compared with <code class="literal">DATETIME</code> values the time
portion of the <code class="literal">DATETIME</code> value was ignored.
Now a <code class="literal">DATE</code> value is coerced to the
<code class="literal">DATETIME</code> type by adding the time portion as
“<span class="quote">00:00:00</span>”. To mimic the old behavior use the
CAST() function in the following way: <code class="literal">SELECT
<em class="replaceable"><code>date_field</code></em> = CAST(NOW() as
DATE);</code>.
(<a href="http://bugs.mysql.com/28929" target="_top">Bug#28929</a>)</p></li><li><p>
<code class="literal">INSERT DELAYED</code> statements on
<code class="literal">BLACKHOLE</code> tables are now rejected, due to the
fact that the <code class="literal">BLACKHOLE</code> storage engine does
not support them.
(<a href="http://bugs.mysql.com/27998" target="_top">Bug#27998</a>)</p></li><li><p>
A new status variable, <code class="literal">Com_call_procedure</code>,
indicates the number of calls to stored procedures.
(<a href="http://bugs.mysql.com/27994" target="_top">Bug#27994</a>)</p></li><li><p>
Potential memory leaks in <code class="literal">SHOW PROFILE</code> were
eliminated.
(<a href="http://bugs.mysql.com/24795" target="_top">Bug#24795</a>)</p></li></ul></div><p>Bugs fixed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Security Fix</strong></span>:
A malformed password packet in the connection protocol could
cause the server to crash. Thanks for Dormando for reporting
this bug, and for providing details and a proof of concept.
(<a href="http://bugs.mysql.com/28984" target="_top">Bug#28984</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-3780" target="_top">CVE-2007-3780</a>)</p></li><li><p><span class="bold"><strong>Security Fix</strong></span>:
Use of a view could allow a user to gain update privileges for
tables in other databases.
(<a href="http://bugs.mysql.com/27878" target="_top">Bug#27878</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-3782" target="_top">CVE-2007-3782</a>)</p></li><li><p><span class="bold"><strong>Security Fix</strong></span>:
The requirement of the <code class="literal">DROP</code> privilege for
<code class="literal">RENAME TABLE</code> was not enforced.
(<a href="http://bugs.mysql.com/27515" target="_top">Bug#27515</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-2691" target="_top">CVE-2007-2691</a>)</p></li><li><p><span class="bold"><strong>Security Fix</strong></span>:
If a stored routine was declared using <code class="literal">SQL SECURITY
INVOKER</code>, a user who invoked the routine could gain
privileges.
(<a href="http://bugs.mysql.com/27337" target="_top">Bug#27337</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-2692" target="_top">CVE-2007-2692</a>)</p></li><li><p><span class="bold"><strong>Security Fix</strong></span>:
<code class="literal">CREATE TABLE LIKE</code> did not require any
privileges on the source table. Now it requires the
<code class="literal">SELECT</code> privilege.
</p><p>
In addition, <code class="literal">CREATE TABLE LIKE</code> was not
isolated from alteration by other connections, which resulted in
various errors and incorrect binary log order when trying to
execute concurrently a <code class="literal">CREATE TABLE LIKE</code>
statement and either DDL statements on the source table or DML
or DDL statements on the target table.
(<a href="http://bugs.mysql.com/23667" target="_top">Bug#23667</a>, <a href="http://bugs.mysql.com/25578" target="_top">Bug#25578</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-3781" target="_top">CVE-2007-3781</a>)</p></li><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
When <span><strong class="command">mysqldump</strong></span> was run with the
<code class="option">--delete-master-logs</code> option, binary log files
were deleted before it was known that the dump had succeeded,
not after. (The method for removing log files used
<code class="literal">RESET MASTER</code> prior to the dump. This also
reset the binary log sequence numbering to
<code class="literal">.000001</code>.) Now <span><strong class="command">mysqldump</strong></span>
flushes the logs (which creates a new binary log number with the
next sequence number), performs the dump, and then uses
<code class="literal">PURGE MASTER LOGS</code> to remove the log files
older than the new one. This also preserves log numbering
because the new log with the next number is generated and only
the preceding logs are removed. However, this may affect
applications if they rely on the log numbering sequence being
reset.
(<a href="http://bugs.mysql.com/24733" target="_top">Bug#24733</a>)</p></li><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
The use of an <code class="literal">ORDER BY</code> or
<code class="literal">DISTINCT</code> clause with a query containing a
call to the <a href="functions.html#function_group-concat"><code class="literal">GROUP_CONCAT()</code></a>
function caused results from previous queries to be redisplayed
in the current result. The fix for this includes replacing a
<code class="literal">BLOB</code> value used internally for sorting with a
<code class="literal">VARCHAR</code>. This means that for long results
(more than 65,535 bytes), it is possible for truncation to
occur; if so, an appropriate warning is issued.
(<a href="http://bugs.mysql.com/23856" target="_top">Bug#23856</a>, <a href="http://bugs.mysql.com/28273" target="_top">Bug#28273</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A corrupt schema file could cause a <span class="errortext">File already
open</span> error.
(<a href="http://bugs.mysql.com/28770" target="_top">Bug#28770</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Setting <code class="literal">InitialNoOpenFiles</code> equal to
<code class="literal">MaxNoOfOpenFiles</code> caused an error. This was
due to the fact that the actual value of
<code class="literal">MaxNoOfOpenFiles</code> as used by the cluster was
offset by 1 from the value set in
<code class="filename">config.ini</code>.
(<a href="http://bugs.mysql.com/28749" target="_top">Bug#28749</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">UPDATE IGNORE</code> statements involving the
primary keys of multiple tables could result in data corruption.
(<a href="http://bugs.mysql.com/28719" target="_top">Bug#28719</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A race condition could result when non-master nodes (in addition
to the master node) tried to update active status due to a local
checkpoint (that is, between <code class="literal">NODE_FAILREP</code> and
<code class="literal">COPY_GCIREQ</code> events). Now only the master
updates the active status.
(<a href="http://bugs.mysql.com/28717" target="_top">Bug#28717</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A fast global checkpoint under high load with high usage of the
redo buffer caused data nodes to fail.
(<a href="http://bugs.mysql.com/28653" target="_top">Bug#28653</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
When an API node sent more than 1024 signals in a single batch,
<code class="literal">NDB</code> would process only the first 1024 of
these, and then hang.
(<a href="http://bugs.mysql.com/28443" target="_top">Bug#28443</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A delay in obtaining <code class="literal">AUTO_INCREMENT</code> IDs could
lead to excess temporary errors.
(<a href="http://bugs.mysql.com/28410" target="_top">Bug#28410</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The cluster waited 30 seconds instead of 30 milliseconds before
reading table statistics.
(<a href="http://bugs.mysql.com/28093" target="_top">Bug#28093</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">INSERT IGNORE</code> wrongly ignored
<code class="literal">NULL</code> values in unique indexes.
(<a href="http://bugs.mysql.com/27980" target="_top">Bug#27980</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The name of the month “<span class="quote">March</span>” was given incorrectly
in the cluster error log.
(<a href="http://bugs.mysql.com/27926" target="_top">Bug#27926</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
It was not possible to add a unique index to an
<code class="literal">NDB</code> table while in single user mode.
(<a href="http://bugs.mysql.com/27710" target="_top">Bug#27710</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Repeated insertion of data generated by
<span><strong class="command">mysqldump</strong></span> into <code class="literal">NDB</code> tables
could eventually lead to failure of the cluster.
(<a href="http://bugs.mysql.com/27437" target="_top">Bug#27437</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">ndb_connectstring</code> did not appear in the
output of <code class="literal">SHOW VARIABLES</code>.
(<a href="http://bugs.mysql.com/26675" target="_top">Bug#26675</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A failure to release internal resources following an error could
lead to problems with single user mode.
(<a href="http://bugs.mysql.com/25818" target="_top">Bug#25818</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
For <code class="literal">BLOB</code> reads on operations with lock mode
<code class="literal">LM_CommittedRead</code>, the lock mode was not
upgraded to <code class="literal">LM_Read</code> before the state of the
<code class="literal">BLOB</code> had already been calculated. The
<code class="literal">NDB</code> API methods affected by this problem
included the following:
</p><div class="itemizedlist"><ul type="circle"><li><p>
<code class="literal">NdbOperation::readTuple()</code>
</p></li><li><p>
<code class="literal">NdbScanOperation::readTuples()</code>
</p></li><li><p>
<code class="literal">NdbIndexScanOperation::readTuples()</code>
</p></li></ul></div><p>
(<a href="http://bugs.mysql.com/27320" target="_top">Bug#27320</a>)</p></li><li><p>
On the IBM i5 platform, the installation script in the
<code class="filename">.savf</code> binaries unconditionally executed the
<span><strong class="command">mysql_install_db</strong></span> script. This problem was
fixed in a repackaged distribution numbered 5.0.45b.
(<a href="http://bugs.mysql.com/30084" target="_top">Bug#30084</a>)</p></li><li><p>
Long pathnames for internal temporary tables could cause stack
overflows.
(<a href="http://bugs.mysql.com/29015" target="_top">Bug#29015</a>)</p></li><li><p>
Using an <code class="literal">INTEGER</code> column from a table to
<a href="functions.html#function_round"><code class="literal">ROUND()</code></a> a number produced
different results than using a constant with the same value as
the <code class="literal">INTEGER</code> column.
(<a href="http://bugs.mysql.com/28980" target="_top">Bug#28980</a>)</p></li><li><p>
If a program binds a given number of parameters to a prepared
statement handle and then somehow changes
<code class="literal">stmt->param_count</code> to a different number,
<a href="apis.html#mysql-stmt-execute" title="23.2.7.10. mysql_stmt_execute()"><code class="literal">mysql_stmt_execute()</code></a> could
crash the client or server.
(<a href="http://bugs.mysql.com/28934" target="_top">Bug#28934</a>)</p></li><li><p>
<code class="literal">INSERT .. ON DUPLICATE KEY UPDATE</code> could under
some circumstances silently update rows when it should not have.
(<a href="http://bugs.mysql.com/28904" target="_top">Bug#28904</a>)</p></li><li><p>
Queries that used <a href="functions.html#function_uuid"><code class="literal">UUID()</code></a> were
incorrectly allowed into the query cache. (This should not
happen because <a href="functions.html#function_uuid"><code class="literal">UUID()</code></a> is
non-deterministic.)
(<a href="http://bugs.mysql.com/28897" target="_top">Bug#28897</a>)</p></li><li><p>
Using a <code class="literal">VIEW</code> created with a non-existing
<code class="literal">DEFINER</code> could lead to incorrect results under
some circumstances.
(<a href="http://bugs.mysql.com/28895" target="_top">Bug#28895</a>)</p></li><li><p>
On Windows, <code class="literal">USE_TLS</code> was not defined for
<code class="filename">mysqlclient.lib</code>.
(<a href="http://bugs.mysql.com/28860" target="_top">Bug#28860</a>)</p></li><li><p>
A subquery with <code class="literal">ORDER BY</code> and <code class="literal">LIMIT
1</code> could cause a server crash.
(<a href="http://bugs.mysql.com/28811" target="_top">Bug#28811</a>)</p></li><li><p>
Using <a href="functions.html#operator_between"><code class="literal">BETWEEN</code></a> with non-indexed
date columns and short formats of the date string could return
incorrect results.
(<a href="http://bugs.mysql.com/28778" target="_top">Bug#28778</a>)</p></li><li><p>
Selecting <code class="literal">GEOMETRY</code> columns in a
<code class="literal">UNION</code> caused a server crash.
(<a href="http://bugs.mysql.com/28763" target="_top">Bug#28763</a>)</p></li><li><p>
When constructing the path to the original
<code class="filename">.frm</code> file, <code class="literal">ALTER ..
RENAME</code> was unnecessarily (and incorrectly) lowercasing
the entire path when not on a case-insensitive filesystem,
causing the statement to fail.
(<a href="http://bugs.mysql.com/28754" target="_top">Bug#28754</a>)</p></li><li><p>
Searches on indexed and non-indexed <code class="literal">ENUM</code>
columns could return different results for empty strings.
(<a href="http://bugs.mysql.com/28729" target="_top">Bug#28729</a>)</p></li><li><p>
Executing <code class="literal">EXPLAIN EXTENDED</code> on a query using a
derived table over a grouping subselect could lead to a server
crash. This occurred only when materialization of the derived
tables required creation of an auxiliary temporary table, an
example being when a grouping operation was carried out with
usage of a temporary table.
(<a href="http://bugs.mysql.com/28728" target="_top">Bug#28728</a>)</p></li><li><p>
The result of evaluation for a view's <code class="literal">CHECK
OPTION</code> option over an updated record and records of
merged tables was arbitrary and dependant on the order of
records in the merged tables during the execution of the
<code class="literal">SELECT</code> statement.
(<a href="http://bugs.mysql.com/28716" target="_top">Bug#28716</a>)</p></li><li><p>
The “<span class="quote">manager thread</span>” of the LinuxThreads
implementation was unintentionally started before
<span><strong class="command">mysqld</strong></span> had dropped privileges (to run as an
unprivileged user). This caused signaling between threads in
<span><strong class="command">mysqld</strong></span> to fail when the privileges were
finally dropped.
(<a href="http://bugs.mysql.com/28690" target="_top">Bug#28690</a>)</p></li><li><p>
For debug builds, <code class="literal">ALTER TABLE</code> could trigger
an assertion failure due to occurrence of a deadlock when
committing changes.
(<a href="http://bugs.mysql.com/28652" target="_top">Bug#28652</a>)</p></li><li><p>
After an upgrade, the names of stored routines referenced by
views were no longer displayed by <code class="literal">SHOW CREATE
VIEW</code>.
(<a href="http://bugs.mysql.com/28605" target="_top">Bug#28605</a>)</p><p>This regression was introduced by <a href="http://bugs.mysql.com/23491" target="_top">Bug#23491</a></p></li><li><p>
Killing from one connection a long-running <code class="literal">EXPLAIN
QUERY</code> started from another connection caused
<span><strong class="command">mysqld</strong></span> to crash.
(<a href="http://bugs.mysql.com/28598" target="_top">Bug#28598</a>)</p></li><li><p>
Outer join queries with <code class="literal">ON</code> conditions over
constant outer tables did not return
<code class="literal">NULL</code>-complemented rows when conditions were
evaluated to <code class="literal">FALSE</code>.
(<a href="http://bugs.mysql.com/28571" target="_top">Bug#28571</a>)</p></li><li><p>
An update on a multiple-table view with the CHECK OPTION clause
and a subquery in the WHERE condition could cause an assertion
failure.
(<a href="http://bugs.mysql.com/28561" target="_top">Bug#28561</a>)</p></li><li><p>
<code class="literal">PURGE MASTER LOGS BEFORE
(<em class="replaceable"><code>subquery</code></em>)</code> caused a server
crash. Subqueries are forbidden in the <code class="literal">BEFORE</code>
clause now.
(<a href="http://bugs.mysql.com/28553" target="_top">Bug#28553</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> calculated the required memory for
a hex-blob string incorrectly causing a buffer overrun. This in
turn caused <span><strong class="command">mysqldump</strong></span> to crash silently and
produce incomplete output.
(<a href="http://bugs.mysql.com/28522" target="_top">Bug#28522</a>)</p></li><li><p>
Passing a <code class="literal">DECIMAL</code> value as a parameter of a
statement prepared with <code class="literal">PREPARE</code> resulted in
an error.
(<a href="http://bugs.mysql.com/28509" target="_top">Bug#28509</a>)</p></li><li><p>
<a href="apis.html#mysql-affected-rows" title="23.2.3.1. mysql_affected_rows()"><code class="literal">mysql_affected_rows()</code></a> could
return an incorrect result for <code class="literal">INSERT ... ON DUPLICATE
KEY UPDATE</code> if the <code class="literal">CLIENT_FOUND_ROWS</code>
flag was set.
(<a href="http://bugs.mysql.com/28505" target="_top">Bug#28505</a>)</p></li><li><p>
A query that grouped by the result of an expression returned a
different result when the expression was assigned to a user
variable.
(<a href="http://bugs.mysql.com/28494" target="_top">Bug#28494</a>)</p></li><li><p>
Subselects returning <code class="literal">LONG</code> values in MySQL
versions later than 5.0.24a returned <code class="literal">LONGLONG</code>
prior to this. The previous behavior was restored.
(<a href="http://bugs.mysql.com/28492" target="_top">Bug#28492</a>)</p><p>This regression was introduced by <a href="http://bugs.mysql.com/19714" target="_top">Bug#19714</a></p></li><li><p>
Forcing the use of an index on a <code class="literal">SELECT</code> query
when the index had been disabled would raise an error without
running the query. The query now executes, with a warning
generated noting that the use of a disabled index has been
ignored.
(<a href="http://bugs.mysql.com/28476" target="_top">Bug#28476</a>)</p></li><li><p>
The result of executing of a prepared statement created with
<code class="literal">PREPARE s FROM "SELECT 1 LIMIT ?"</code>
was not replicated correctly.
(<a href="http://bugs.mysql.com/28464" target="_top">Bug#28464</a>)</p></li><li><p>
The query <code class="literal">SELECT '2007-01-01' + INTERVAL
<em class="replaceable"><code>column_name</code></em> DAY FROM
<em class="replaceable"><code>table_name</code></em> </code> caused
<span><strong class="command">mysqld</strong></span> to fail.
(<a href="http://bugs.mysql.com/28450" target="_top">Bug#28450</a>)</p></li><li><p>
A server crash could happen under rare conditions such that a
temporary table outgrew heap memory reserved for it and the
remaining disk space was not big enough to store the table as a
<code class="literal">MyISAM</code> table.
(<a href="http://bugs.mysql.com/28449" target="_top">Bug#28449</a>)</p></li><li><p>
<span><strong class="command">mysql_upgrade</strong></span> failed if certain SQL modes
were set. Now it sets the mode itself to avoid this problem.
(<a href="http://bugs.mysql.com/28401" target="_top">Bug#28401</a>)</p></li><li><p>
A query with a <code class="literal">NOT IN</code> subquery predicate
could cause a crash when the left operand of the predicate
evaluated to <code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/28375" target="_top">Bug#28375</a>)</p></li><li><p>
The test case for <span><strong class="command">mysqldump</strong></span> failed with
<code class="literal">bin-log</code> disabled.
(<a href="http://bugs.mysql.com/28372" target="_top">Bug#28372</a>)</p></li><li><p>
Attempting to <code class="literal">LOAD_FILE</code> from an empty floppy
drive under Windows, caused the server to hang. For example, if
you opened a connection to the server and then issued the
command <span><strong class="command">SELECT LOAD_FILE('a:test');</strong></span>, with no
floppy in the drive, the server was inaccessible until the modal
pop-up dialog box was dismissed.
(<a href="http://bugs.mysql.com/28366" target="_top">Bug#28366</a>)</p></li><li><p>
A buffer overflow could occur when using
<code class="literal">DECIMAL</code> columns on Windows operating systems.
(<a href="http://bugs.mysql.com/28361" target="_top">Bug#28361</a>)</p></li><li><p>
<code class="literal">libmysql.dll</code> could not be dynamically loaded
on Windows.
(<a href="http://bugs.mysql.com/28358" target="_top">Bug#28358</a>)</p></li><li><p>
Grouping queries with correlated subqueries in
<code class="literal">WHERE</code> conditions could produce incorrect
results.
(<a href="http://bugs.mysql.com/28337" target="_top">Bug#28337</a>)</p></li><li><p>
<span><strong class="command">mysqltest</strong></span> used a too-large stack size on
PPC/Debian Linux, causing thread-creation failure for tests that
use many threads.
(<a href="http://bugs.mysql.com/28333" target="_top">Bug#28333</a>)</p></li><li><p>
<code class="literal">EXPLAIN</code> for a query on an empty table
immediately after its creation could result in a server crash.
(<a href="http://bugs.mysql.com/28272" target="_top">Bug#28272</a>)</p></li><li><p>
The <code class="literal">IS_UPDATABLE</code> column in the
<code class="literal">INFORMATION_SCHEMA.VIEWS</code> table was not always
set correctly.
(<a href="http://bugs.mysql.com/28266" target="_top">Bug#28266</a>)</p></li><li><p>
Comparing a <code class="literal">DATETIME</code> column value with a user
variable yielded incorrect results.
(<a href="http://bugs.mysql.com/28261" target="_top">Bug#28261</a>)</p></li><li><p>
For <a href="functions.html#function_cast"><code class="literal">CAST()</code></a> of a
<code class="literal">NULL</code> value with type
<code class="literal">DECIMAL</code>, the return value was incorrectly
initialized, producing a runtime error for binaries built using
Visual C++ 2005.
(<a href="http://bugs.mysql.com/28250" target="_top">Bug#28250</a>)</p></li><li><p>
Recreating a view that already exists on the master would cause
a replicating slave to terminate replication with a 'different
error message on slave and master' error.
(<a href="http://bugs.mysql.com/28244" target="_top">Bug#28244</a>)</p></li><li><p>
Portability problems caused by use of <code class="literal">isinf()</code>
were corrected.
(<a href="http://bugs.mysql.com/28240" target="_top">Bug#28240</a>)</p></li><li><p>
When dumping procedures, <span><strong class="command">mysqldump
<code class="option">--compact</code> </strong></span> generated output that
restored the session variable <code class="literal">SQL_MODE</code>
without first capturing it. When dumping routines,
<span><strong class="command">mysqldump <code class="option">--compact</code> </strong></span> neither
set nor retrieved the value of <code class="literal">SQL_MODE</code>.
(<a href="http://bugs.mysql.com/28223" target="_top">Bug#28223</a>)</p></li><li><p>
Comparison of the string value of a date showed as unequal to
<a href="functions.html#function_curtime"><code class="literal">CURTIME()</code></a>. Similar behavior was
exhibited for <code class="literal">DATETIME</code> values.
(<a href="http://bugs.mysql.com/28208" target="_top">Bug#28208</a>)</p></li><li><p>
For <code class="literal">InnoDB</code>, in some rare cases the optimizer
preferred a more expensive <code class="literal">ref</code> access to a
less expensive range access.
(<a href="http://bugs.mysql.com/28189" target="_top">Bug#28189</a>)</p></li><li><p>
A performance degradation was observed for outer join queries to
which a not-exists optimization was applied.
(<a href="http://bugs.mysql.com/28188" target="_top">Bug#28188</a>)</p></li><li><p>
<code class="literal">SELECT * INTO OUTFILE ... FROM
INFORMATION_SCHEMA.SCHEMATA</code> failed with an
<span class="errortext">Access denied</span> error, even for a user who
had the <code class="literal">FILE</code> privilege.
(<a href="http://bugs.mysql.com/28181" target="_top">Bug#28181</a>)</p></li><li><p>
The <code class="literal">Bytes_received</code> and
<code class="literal">Bytes_sent</code> status variables could hold only
32-bit values (not 64-bit values) on some platforms.
(<a href="http://bugs.mysql.com/28149" target="_top">Bug#28149</a>)</p></li><li><p>
Comparisons of <code class="literal">DATE</code> or
<code class="literal">DATETIME</code> values for the
<a href="functions.html#function_in"><code class="literal">IN()</code></a> function could yield
incorrect results.
(<a href="http://bugs.mysql.com/28133" target="_top">Bug#28133</a>)</p></li><li><p>
Storing a large number into a <code class="literal">FLOAT</code> or
<code class="literal">DOUBLE</code> column with a fixed length could
result in incorrect truncation of the number if the column's
length was greater than 31.
(<a href="http://bugs.mysql.com/28121" target="_top">Bug#28121</a>)</p></li><li><p>
The server could hang for <code class="literal">INSERT IGNORE ... ON
DUPLICATE KEY UPDATE</code> if an update failed.
(<a href="http://bugs.mysql.com/28000" target="_top">Bug#28000</a>)</p></li><li><p>
<code class="literal">DECIMAL</code> values beginning with nine
<code class="literal">9</code> digits could be incorrectly rounded.
(<a href="http://bugs.mysql.com/27984" target="_top">Bug#27984</a>)</p></li><li><p>
<a href="functions.html#function_cast"><code class="literal">CAST()</code></a> to
<code class="literal">DECIMAL</code> did not check for overflow.
(<a href="http://bugs.mysql.com/27957" target="_top">Bug#27957</a>)</p></li><li><p>
For <code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code>
statements that affected many rows, updates could be applied to
the wrong rows.
(<a href="http://bugs.mysql.com/27954" target="_top">Bug#27954</a>)</p></li><li><p>
Early <code class="literal">NULL</code>-filtering optimization did not
work for <code class="literal">eq_ref</code> table access.
(<a href="http://bugs.mysql.com/27939" target="_top">Bug#27939</a>)</p></li><li><p>
The second execution of a prepared statement from a
<code class="literal">UNION</code> query with <code class="literal">ORDER BY
RAND()</code> caused the server to crash. This problem could
also occur when invoking a stored procedure containing such a
query.
(<a href="http://bugs.mysql.com/27937" target="_top">Bug#27937</a>)</p></li><li><p>
Views ignored precision for
<a href="functions.html#function_cast"><code class="literal">CAST()</code></a> operations.
(<a href="http://bugs.mysql.com/27921" target="_top">Bug#27921</a>)</p></li><li><p>
For attempts to open a non-existent table, the server should
report <code class="literal">ER_NO_SUCH_TABLE</code> but sometimes
reported <code class="literal">ER_TABLE_NOT_LOCKED</code>.
(<a href="http://bugs.mysql.com/27907" target="_top">Bug#27907</a>)</p></li><li><p>
A stored program that uses a variable name containing multibyte
characters could fail to execute.
(<a href="http://bugs.mysql.com/27876" target="_top">Bug#27876</a>)</p></li><li><p>
Non-grouped columns were allowed by <code class="literal">*</code> in
<code class="literal">ONLY_FULL_GROUP_BY</code> SQL mode.
(<a href="http://bugs.mysql.com/27874" target="_top">Bug#27874</a>)</p></li><li><p>
<code class="literal">ON</code> conditions from <code class="literal">JOIN</code>
expressions were ignored when checking the <code class="literal">CHECK
OPTION</code> clause while updating a multiple-table view
that included such a clause.
(<a href="http://bugs.mysql.com/27827" target="_top">Bug#27827</a>)</p></li><li><p>
Debug builds on Windows generated false alarms about
uninitialized variables with some Visual Studio runtime
libraries.
(<a href="http://bugs.mysql.com/27811" target="_top">Bug#27811</a>)</p></li><li><p>
Certain queries that used uncorrelated scalar subqueries caused
<code class="literal">EXPLAIN</code> to crash.
(<a href="http://bugs.mysql.com/27807" target="_top">Bug#27807</a>)</p></li><li><p>
Changes to some system variables should invalidate statements in
the query cache, but invalidation did not happen.
(<a href="http://bugs.mysql.com/27792" target="_top">Bug#27792</a>)</p></li><li><p>
Performing a <code class="literal">UNION</code> on two views that had
<code class="literal">ORDER BY</code> clauses resulted in an
<code class="literal">Unknown column</code> error.
(<a href="http://bugs.mysql.com/27786" target="_top">Bug#27786</a>)</p></li><li><p>
<span><strong class="command">mysql_install_db</strong></span> is supposed to detect
existing system tables and create only those that do not exist.
Instead, it was exiting with an error if tables already existed.
(<a href="http://bugs.mysql.com/27783" target="_top">Bug#27783</a>)</p></li><li><p>
On some systems, <code class="filename">udf_example.c</code> returned an
incorrect result length. Also on some systems,
<span><strong class="command">mysql-test-run.pl</strong></span> could not find the shared
object built from <code class="filename">udf_example.c</code>.
(<a href="http://bugs.mysql.com/27741" target="_top">Bug#27741</a>)</p></li><li><p>
<span><strong class="command">mysqld</strong></span> did not check the length of option
values and could crash with a buffer overflow for long values.
(<a href="http://bugs.mysql.com/27715" target="_top">Bug#27715</a>)</p></li><li><p>
Comparisons using row constructors could fail for rows
containing <code class="literal">NULL</code> values.
(<a href="http://bugs.mysql.com/27704" target="_top">Bug#27704</a>)</p></li><li><p>
<code class="literal">LOAD DATA</code> did not use
<code class="literal">CURRENT_TIMESTAMP</code> as the default value for a
<code class="literal">TIMESTAMP</code> column for which no value was
provided.
(<a href="http://bugs.mysql.com/27670" target="_top">Bug#27670</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> could not connect using SSL.
(<a href="http://bugs.mysql.com/27669" target="_top">Bug#27669</a>)</p></li><li><p>
On Linux, the server could not create temporary tables if
<code class="literal">lower_case_table_names</code> was set to 1 and the
value of <code class="literal">tmpdir</code> was a directory name
containing any uppercase letters.
(<a href="http://bugs.mysql.com/27653" target="_top">Bug#27653</a>)</p></li><li><p>
For <code class="literal">InnoDB</code> tables, a multiple-row
<code class="literal">INSERT</code> of the form <code class="literal">INSERT INTO t
(id...) VALUES (NULL...) ON DUPLICATE KEY UPDATE
id=VALUES(id)</code>, where <code class="literal">id</code> is an
<code class="literal">AUTO_INCREMENT</code> column, could cause
<code class="literal">ERROR 1062 (23000): Duplicate entry...</code> errors
or lost rows.
(<a href="http://bugs.mysql.com/27650" target="_top">Bug#27650</a>)</p></li><li><p>
<code class="literal">HASH</code> indexes on <code class="literal">VARCHAR</code>
columns with binary collations did not ignore trailing spaces
from strings before comparisons. This could result in duplicate
records being successfully inserted into a
<code class="literal">MEMORY</code> table with unique key constraints. A
consequence was that internal <code class="literal">MEMORY</code> tables
used for <code class="literal">GROUP BY</code> calculation contained
duplicate rows that resulted in duplicate-key errors when
converting those temporary tables to <code class="literal">MyISAM</code>,
and that error was incorrectly reported as a <code class="literal">table is
full</code> error.
(<a href="http://bugs.mysql.com/27643" target="_top">Bug#27643</a>)</p></li><li><p>
The XML output representing an empty result was an empty string
rather than an empty <code class="literal"><resultset/></code>
element.
(<a href="http://bugs.mysql.com/27608" target="_top">Bug#27608</a>)</p></li><li><p>
An error occurred trying to connect to
<span><strong class="command">mysqld-debug.exe</strong></span>.
(<a href="http://bugs.mysql.com/27597" target="_top">Bug#27597</a>)</p></li><li><p>
Comparison of a <code class="literal">DATE</code> with a
<code class="literal">DATETIME</code> did not treat the
<code class="literal">DATE</code> as having a time part of
<code class="literal">00:00:00</code>.
(<a href="http://bugs.mysql.com/27590" target="_top">Bug#27590</a>)</p></li><li><p>
Selecting <a href="functions.html#function_min"><code class="literal">MIN()</code></a> on an indexed
column that contained only <code class="literal">NULL</code> values caused
<code class="literal">NULL</code> to be returned for other result columns.
(<a href="http://bugs.mysql.com/27573" target="_top">Bug#27573</a>)</p></li><li><p>
If a stored function or trigger was killed, it aborted but no
error was thrown, allowing the calling statement to continue
without noticing the problem. This could lead to incorrect
results.
(<a href="http://bugs.mysql.com/27563" target="_top">Bug#27563</a>)</p></li><li><p>
The fix for <a href="http://bugs.mysql.com/17212" target="_top">Bug#17212</a> provided correct sort order for
misordered output of certain queries, but caused significant
overall query performance degradation. (Results were correct
(good), but returned much more slowly (bad).) The fix also
affected performance of queries for which results were correct.
The performance degradation has been addressed.
(<a href="http://bugs.mysql.com/27531" target="_top">Bug#27531</a>)</p></li><li><p>
The <code class="literal">CRC32()</code> function returns an unsigned
integer, but the metadata was signed, which could cause certain
queries to return incorrect results. (For example, queries that
selected a <code class="literal">CRC32()</code> value and used that value
in the <code class="literal">GROUP BY</code> clause.)
(<a href="http://bugs.mysql.com/27530" target="_top">Bug#27530</a>)</p></li><li><p>
An interaction between <code class="literal">SHOW TABLE STATUS</code> and
other concurrent statements that modify the table could result
in a divide-by-zero error and a server crash.
(<a href="http://bugs.mysql.com/27516" target="_top">Bug#27516</a>)</p></li><li><p>
When <code class="literal">ALTER TABLE</code> was used to add a new
<code class="literal">DATE</code> column with no explicit default value,
<code class="literal">'0000-00-00'</code> was used as the default even if
the SQL mode included the <code class="literal">NO_ZERO_DATE</code> mode
to prohibit that value. A similar problem occurred for
<code class="literal">DATETIME</code> columns.
(<a href="http://bugs.mysql.com/27507" target="_top">Bug#27507</a>)</p></li><li><p>
A race condition between <code class="literal">DROP TABLE</code> and
<code class="literal">SHOW TABLE STATUS</code> could cause the latter to
display incorrect information.
(<a href="http://bugs.mysql.com/27499" target="_top">Bug#27499</a>)</p></li><li><p>
Using a <code class="literal">TEXT</code> local variable in a stored
routine in an expression such as <code class="literal">SET
<em class="replaceable"><code>var</code></em> =
SUBSTRING(<em class="replaceable"><code>var</code></em>, 3)</code> produced
an incorrect result.
(<a href="http://bugs.mysql.com/27415" target="_top">Bug#27415</a>)</p></li><li><p>
Nested aggregate functions could be improperly evaluated.
(<a href="http://bugs.mysql.com/27363" target="_top">Bug#27363</a>)</p></li><li><p>
A stored function invocation in the <code class="literal">WHERE</code>
clause was treated as a constant.
(<a href="http://bugs.mysql.com/27354" target="_top">Bug#27354</a>)</p></li><li><p>
Failure to allocate memory associated with
<code class="literal">transaction_prealloc_size</code> could cause a
server crash.
(<a href="http://bugs.mysql.com/27322" target="_top">Bug#27322</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> crashed if it got no data from
<code class="literal">SHOW CREATE PROCEDURE</code> (for example, when
trying to dump a routine defined by a different user and for
which the current user had no privileges). Now it prints a
comment to indicate the problem. It also returns an error, or
continues if the <code class="option">--force</code> option is given.
(<a href="http://bugs.mysql.com/27293" target="_top">Bug#27293</a>)</p></li><li><p>
The error message for error number <code class="literal">137</code> did
not report which database/table combination reported the
problem.
(<a href="http://bugs.mysql.com/27173" target="_top">Bug#27173</a>)</p></li><li><p>
<span><strong class="command">mysqlbinlog</strong></span> produced different output with
the <code class="option">-R</code> option than without it.
(<a href="http://bugs.mysql.com/27171" target="_top">Bug#27171</a>)</p></li><li><p>
A large filesort could result in a division by zero error and a
server crash.
(<a href="http://bugs.mysql.com/27119" target="_top">Bug#27119</a>)</p></li><li><p>
Times displayed by <code class="literal">SHOW PROFILE</code> were
incorrectly associated with the profile entry one later than the
corrrect one.
(<a href="http://bugs.mysql.com/27060" target="_top">Bug#27060</a>)</p></li><li><p>
Flow control optimization in stored routines could cause
exception handlers to never return or execute incorrect logic.
(<a href="http://bugs.mysql.com/26977" target="_top">Bug#26977</a>)</p></li><li><p>
<code class="literal">SHOW PROFILE</code> hung if executed before enabling
the <code class="literal">@@profiling</code> session variable.
(<a href="http://bugs.mysql.com/26938" target="_top">Bug#26938</a>)</p></li><li><p>
Binary logging of prepared statements could produce
syntactically incorrect queries in the binary log, replacing
some parameters with variable names rather than variable values.
This could lead to incorrect results on replication slaves.
(<a href="http://bugs.mysql.com/26842" target="_top">Bug#26842</a>, <a href="http://bugs.mysql.com/12826" target="_top">Bug#12826</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> would not dump a view for which the
<code class="literal">DEFINER</code> no longer exists.
(<a href="http://bugs.mysql.com/26817" target="_top">Bug#26817</a>)</p></li><li><p>
Connections from one <span><strong class="command">mysqld</strong></span> server to another
failed on Mac OS X, affecting replication and
<code class="literal">FEDERATED</code> tables.
(<a href="http://bugs.mysql.com/26664" target="_top">Bug#26664</a>)</p><p>See also <a href="http://bugs.mysql.com/29083" target="_top">Bug#29083</a></p></li><li><p>
Creating a temporary table with InnoDB when using the
one-file-per-table setting, and when the host filesystem for
temporary tables was <code class="literal">tmpfs</code>, would cause an
assertion within <code class="literal">mysqld</code>. This was due to the
use of <code class="literal">O_DIRECT</code> when opening the temporary
table file.
(<a href="http://bugs.mysql.com/26662" target="_top">Bug#26662</a>)</p></li><li><p>
<span><strong class="command">mysql_upgrade</strong></span> did not detect failure of
external commands that it runs.
(<a href="http://bugs.mysql.com/26639" target="_top">Bug#26639</a>)</p></li><li><p>
Some test suite files were missing from some MySQL-test
packages.
(<a href="http://bugs.mysql.com/26609" target="_top">Bug#26609</a>)</p></li><li><p>
Aborting a statement on the master that applied to a
non-transactional statement broke replication. The statement was
written to the binary log but not completely executed on the
master. Slaves receiving the statement executed it completely,
resulting in loss of data synchrony. Now an error code is
written to the error log so that the slaves stop without
executing the aborted statement. (That is, replication stops,
but synchrony to the point of the stop is preserved and you can
investigate the problem.)
(<a href="http://bugs.mysql.com/26551" target="_top">Bug#26551</a>)</p></li><li><p>
Statements within triggers ignored the value of the
<code class="literal">low_priority_updates</code> system variable.
(<a href="http://bugs.mysql.com/26162" target="_top">Bug#26162</a>)</p><p>See also <a href="http://bugs.mysql.com/29963" target="_top">Bug#29963</a></p></li><li><p>
Index hints (<code class="literal">USE INDEX</code>, <code class="literal">IGNORE
INDEX</code>, <code class="literal">FORCE INDEX</code>) cannot be used
with <code class="literal">FULLTEXT</code> indexes, but were not being
ignored.
(<a href="http://bugs.mysql.com/25951" target="_top">Bug#25951</a>)</p></li><li><p>
If <code class="literal">CREATE TABLE t1 LIKE t2</code> failed due to a
full disk, an empty <code class="filename">t2.frm</code> file could be
created but not removed. This file then caused subsequent
attempts to create a table named <code class="literal">t2</code> to fail.
This is easily corrected at the filesystem level by removing the
<code class="filename">t2.frm</code> file manually, but now the server
removes the file if the create operation does not complete
successfully.
(<a href="http://bugs.mysql.com/25761" target="_top">Bug#25761</a>)</p></li><li><p>
Running <code class="literal">CHECK TABLE</code> concurrently with a
<code class="literal">SELECT</code>, <code class="literal">INSERT</code> or other
statement on Windows could corrupt a MyISAM table.
(<a href="http://bugs.mysql.com/25712" target="_top">Bug#25712</a>)</p></li><li><p>
On Windows, connection handlers did not properly decrement the
server's thread count when exiting.
(<a href="http://bugs.mysql.com/25621" target="_top">Bug#25621</a>)</p></li><li><p>
<span><strong class="command">mysql_upgrade</strong></span> did not pass a password to
<span><strong class="command">mysqlcheck</strong></span> if one was given.
(<a href="http://bugs.mysql.com/25452" target="_top">Bug#25452</a>)</p></li><li><p>
On Windows, <span><strong class="command">mysql_upgrade</strong></span> was sensitive to
lettercase of the names of some required components.
(<a href="http://bugs.mysql.com/25405" target="_top">Bug#25405</a>)</p></li><li><p>
For storage engines that allow the current auto-increment value
to be set, using <code class="literal">ALTER TABLE ... ENGINE</code> to
convert a table from one such storage engine to another caused
loss of the current value. (For storage engines that do not
support setting the value, it cannot be retained anyway when
changing the storage engine.)
(<a href="http://bugs.mysql.com/25262" target="_top">Bug#25262</a>)</p></li><li><p>
Restoration of the default database after stored routine or
trigger execution on a slave could cause replication to stop if
the database no longer existed.
(<a href="http://bugs.mysql.com/25082" target="_top">Bug#25082</a>)</p></li><li><p>
Due to a race condition, executing <code class="literal">FLUSH
PRIVILEGES</code> in one thread could cause brief table
unavailability in other threads.
(<a href="http://bugs.mysql.com/24988" target="_top">Bug#24988</a>)</p></li><li><p>
Several math functions produced incorrect results for large
unsigned values. <a href="functions.html#function_round"><code class="literal">ROUND()</code></a>
produced incorrect results or a crash for a large
number-of-decimals argument.
(<a href="http://bugs.mysql.com/24912" target="_top">Bug#24912</a>)</p></li><li><p>
The result set of a query that used <code class="literal">WITH
ROLLUP</code> and <code class="literal">DISTINCT</code> could lack some
rollup rows (rows with <code class="literal">NULL</code> values for
grouping attributes) if the <code class="literal">GROUP BY</code> list
contained constant expressions.
(<a href="http://bugs.mysql.com/24856" target="_top">Bug#24856</a>)</p></li><li><p>
For queries that used <code class="literal">ORDER BY</code> with
<code class="literal">InnoDB</code> tables, if the optimizer chose an
index for accessing the table but found a covering index that
enabled the <code class="literal">ORDER BY</code> to be skipped, no
results were returned.
(<a href="http://bugs.mysql.com/24778" target="_top">Bug#24778</a>)</p></li><li><p>
Concurrent execution of <code class="literal">CREATE TABLE ...
SELECT</code> and other statements involving the target table
suffered from various race conditions, some of which might have
led to deadlocks.
(<a href="http://bugs.mysql.com/24738" target="_top">Bug#24738</a>)</p></li><li><p>
On some Linux distributions where LinuxThreads and NPTL
<code class="literal">glibc</code> versions both are available, statically
built binaries can crash because the linker defaults to
LinuxThreads when linking statically, but calls to external
libraries (such as <code class="literal">libnss</code>) are resolved to
NPTL versions. This cannot be worked around in the code, so
instead if a crash occurs on such a binary/OS combination, print
an error message that provides advice about how to fix the
problem.
(<a href="http://bugs.mysql.com/24611" target="_top">Bug#24611</a>)</p></li><li><p>
An attempt to execute <code class="literal">CREATE TABLE ... SELECT</code>
when a temporary table with the same name already existed led to
the insertion of data into the temporary table and creation of
an empty non-temporary table.
(<a href="http://bugs.mysql.com/24508" target="_top">Bug#24508</a>)</p></li><li><p>
The <code class="literal">MERGE</code> storage engine could return
incorrect results when several index values that compare
equality were present in an index (for example,
<code class="literal">'gross'</code> and <code class="literal">'gross '</code>,
which are considered equal but have different lengths).
(<a href="http://bugs.mysql.com/24342" target="_top">Bug#24342</a>)</p></li><li><p>
Some upgrade problems are detected and better error messages
suggesting that <span><strong class="command">mysql_upgrade</strong></span> be run are
produced.
(<a href="http://bugs.mysql.com/24248" target="_top">Bug#24248</a>)</p></li><li><p>
Some views could not be created even when the user had the
requisite privileges.
(<a href="http://bugs.mysql.com/24040" target="_top">Bug#24040</a>)</p></li><li><p>
Using <a href="functions.html#function_cast"><code class="literal">CAST()</code></a> to convert
<code class="literal">DATETIME</code> values to numeric values did not
work.
(<a href="http://bugs.mysql.com/23656" target="_top">Bug#23656</a>)</p></li><li><p>
The <code class="literal">AUTO_INCREMENT</code> value would not be
correctly reported for InnoDB tables when using <code class="literal">SHOW
CREATE TABLE</code> statement or <span><strong class="command">mysqldump</strong></span>
command.
(<a href="http://bugs.mysql.com/23313" target="_top">Bug#23313</a>)</p></li><li><p>
Implicit conversion of <code class="literal">9912101</code> to
<code class="literal">DATE</code> did not match
<a href="functions.html#function_cast"><code class="literal">CAST(9912101 AS DATE)</code></a>.
(<a href="http://bugs.mysql.com/23093" target="_top">Bug#23093</a>)</p></li><li><p>
Conversion errors could occur when constructing the condition
for an <code class="literal">IN</code> predicate. The predicate was
treated as if the affected column contains
<code class="literal">NULL</code>, but if the <code class="literal">IN</code>
predicate is inside <code class="literal">NOT</code>, incorrect results
could be returned.
(<a href="http://bugs.mysql.com/22855" target="_top">Bug#22855</a>)</p></li><li><p>
<code class="literal">SELECT COUNT(*)</code> from a table containing a
<code class="literal">DATETIME NOT NULL</code> column could produce
spurious warnings with the <code class="literal">NO_ZERO_DATE</code> SQL
mode enabled.
(<a href="http://bugs.mysql.com/22824" target="_top">Bug#22824</a>)</p></li><li><p>
When using transactions and replication, shutting down the
master in the middle of a transaction would cause all slaves to
stop replicating.
(<a href="http://bugs.mysql.com/22725" target="_top">Bug#22725</a>)</p></li><li><p>
Using <code class="literal">SET GLOBAL</code> to change the
<code class="literal">lc_time_names</code> system variable had no effect
on new connections.
(<a href="http://bugs.mysql.com/22648" target="_top">Bug#22648</a>)</p></li><li><p>
A multiple-table <code class="literal">UPDATE</code> could return an
incorrect rows-matched value if, during insertion of rows into a
temporary table, the table had to be converted from a
<code class="literal">MEMORY</code> table to a <code class="literal">MyISAM</code>
table.
(<a href="http://bugs.mysql.com/22364" target="_top">Bug#22364</a>)</p></li><li><p>
yaSSL crashed on pre-Pentium Intel CPUs.
(<a href="http://bugs.mysql.com/21765" target="_top">Bug#21765</a>)</p></li><li><p>
Linux binaries were unable to dump core after executing a
<code class="literal">setuid()</code> call.
(<a href="http://bugs.mysql.com/21723" target="_top">Bug#21723</a>)</p></li><li><p>
A slave that used <code class="option">--master-ssl-cipher</code> could not
connect to the master.
(<a href="http://bugs.mysql.com/21611" target="_top">Bug#21611</a>)</p></li><li><p>
Quoted labels in stored routines were mishandled, rendering the
routines unusable.
(<a href="http://bugs.mysql.com/21513" target="_top">Bug#21513</a>)</p></li><li><p>
Stack overflow caused server crashes.
(<a href="http://bugs.mysql.com/21476" target="_top">Bug#21476</a>)</p></li><li><p>
<a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> is less than
<a href="functions.html#function_now"><code class="literal">NOW()</code></a>, either when comparing
<a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> directly
(<code class="literal">CURDATE() < NOW()</code> is true) or when
casting <a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> to
<code class="literal">DATE</code> (<code class="literal">CAST(CURDATE() AS DATE) <
NOW()</code> is true). However, storing
<a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> in a
<code class="literal">DATE</code> column and comparing <code class="literal">
<em class="replaceable"><code>col_name</code></em> < NOW()</code>
incorrectly yielded false. This is fixed by comparing a
<code class="literal">DATE</code> column as <code class="literal">DATETIME</code>
for comparisons to a <code class="literal">DATETIME</code> constant.
(<a href="http://bugs.mysql.com/21103" target="_top">Bug#21103</a>)</p></li><li><p>
<code class="literal">CREATE TABLE IF NOT EXISTS ... SELECT</code> caused
a server crash if the target table already existed and had a
<code class="literal">BEFORE INSERT</code> trigger.
(<a href="http://bugs.mysql.com/20903" target="_top">Bug#20903</a>)</p></li><li><p>
Deadlock occurred for attempts to execute <code class="literal">CREATE TABLE
IF NOT EXISTS ... SELECT</code> when <code class="literal">LOCK
TABLES</code> had been used to acquire a read lock on the
target table.
(<a href="http://bugs.mysql.com/20662" target="_top">Bug#20662</a>, <a href="http://bugs.mysql.com/15522" target="_top">Bug#15522</a>)</p></li><li><p>
Changing a <code class="literal">utf8</code> column in an
<code class="literal">InnoDB</code> table to a shorter length did not
shorten the data values.
(<a href="http://bugs.mysql.com/20095" target="_top">Bug#20095</a>)</p></li><li><p>
For dates with 4-digit year parts less than 200, an incorrect
implicit conversion to add a century was applied for date
arithmetic performed with
<a href="functions.html#function_date-add"><code class="literal">DATE_ADD()</code></a>,
<a href="functions.html#function_date-sub"><code class="literal">DATE_SUB()</code></a>, <code class="literal">+
INTERVAL</code>, and <code class="literal">- INTERVAL</code>. (For
example, <a href="functions.html#function_date-add"><code class="literal">DATE_ADD('0050-01-01 00:00:00',
INTERVAL 0 SECOND)</code></a> became <code class="literal">'2050-01-01
00:00:00'</code>.)
(<a href="http://bugs.mysql.com/18997" target="_top">Bug#18997</a>)</p></li><li><p>
Using <code class="literal">CREATE TABLE LIKE ...</code> would raise an
assertion when replicated to a slave.
(<a href="http://bugs.mysql.com/18950" target="_top">Bug#18950</a>)</p></li><li><p>
Granting access privileges to an individual table where the
database or table name contained an underscore would fail.
(<a href="http://bugs.mysql.com/18660" target="_top">Bug#18660</a>)</p></li><li><p>
The <code class="literal">-lmtmalloc</code> library was removed from the
output of <span><strong class="command">mysql_config</strong></span> on Solaris, as it
caused problems when building <code class="literal">DBD::mysql</code> (and
possibly other applications) on that platform that tried to use
<span><strong class="command">dlopen()</strong></span> to access the client library.
(<a href="http://bugs.mysql.com/18322" target="_top">Bug#18322</a>)</p></li><li><p>
The <span><strong class="command">check-cpu</strong></span> script failed to detect AMD64
Turion processors correctly.
(<a href="http://bugs.mysql.com/17707" target="_top">Bug#17707</a>)</p></li><li><p>
Trying to shut down the server following a failed <code class="literal">LOAD
DATA INFILE</code> caused <span><strong class="command">mysqld</strong></span> to crash.
(<a href="http://bugs.mysql.com/17233" target="_top">Bug#17233</a>)</p></li><li><p>
The omission of leading zeros in dates could lead to erroneous
results when these were compared with the output of certain date
and time functions.
(<a href="http://bugs.mysql.com/16377" target="_top">Bug#16377</a>)</p></li><li><p>
<code class="literal">INSERT...ON DUPLICATE KEY UPDATE</code> could cause
<code class="literal">Error 1032: Can't find record in ...</code> for
inserts into an <code class="literal">InnoDB</code> table unique index
using key column prefixes with an underlying
<code class="literal">utf8</code> string column.
(<a href="http://bugs.mysql.com/13191" target="_top">Bug#13191</a>)</p></li><li><p>
Having the <code class="literal">EXECUTE</code> privilege for a routine in
a database should make it possible to <code class="literal">USE</code>
that database, but the server returned an error instead. This
has been corrected. As a result of the change, <code class="literal">SHOW
TABLES</code> for a database in which you have only the
<code class="literal">EXECUTE</code> privilege returns an empty set rather
than an error.
(<a href="http://bugs.mysql.com/9504" target="_top">Bug#9504</a>)</p></li></ul></div></div><div class="section" lang="en"><div class="titlepage"><div><div><h3 class="title"><a name="releasenotes-cs-5-0-41"></a>D.1.3. Release Notes for MySQL Community Server 5.0.41 (01 May 2007)</h3></div></div></div><p class="release-level">
This is a bugfix release for the current production release
family. It replaces MySQL 5.0.37.
</p><p>Functionality added or changed:</p><div class="itemizedlist"><ul type="disc"><li><p>
If a set function <em class="replaceable"><code>S</code></em> with an outer
reference <code class="literal">
<em class="replaceable"><code>S</code></em>(<em class="replaceable"><code>outer_ref</code></em>)</code>
cannot be aggregated in the outer query against which the outer
reference has been resolved, MySQL interprets <code class="literal">
<em class="replaceable"><code>S</code></em>(<em class="replaceable"><code>outer_ref</code></em>)</code>
the same way that it would interpret <code class="literal">
<em class="replaceable"><code>S</code></em>(<em class="replaceable"><code>const</code></em>)</code>.
However, standard SQL requires throwing an error in this
situation. An error now is thrown for such queries if the
<code class="literal">ANSI</code> SQL mode is enabled.
(<a href="http://bugs.mysql.com/27348" target="_top">Bug#27348</a>)</p></li><li><p>
Prefix lengths for columns in <code class="literal">SPATIAL</code> indexes
are no longer displayed in <code class="literal">SHOW CREATE TABLE</code>
output. <span><strong class="command">mysqldump</strong></span> uses that statement, so if
a table with <code class="literal">SPATIAL</code> indexes containing
prefixed columns is dumped and reloaded, the index is created
with no prefixes. (The full column width of each column is
indexed.)
(<a href="http://bugs.mysql.com/26794" target="_top">Bug#26794</a>)</p></li><li><p>
The output of <span><strong class="command">mysql <code class="option">--xml</code> </strong></span>
and <span><strong class="command">mysqldump <code class="option">--xml</code> </strong></span> now
includes a valid XML namespace.
(<a href="http://bugs.mysql.com/25946" target="_top">Bug#25946</a>)</p></li><li><p>
If you use SSL for a client connection, you can tell the client
not to authenticate the server certificate by specifying neither
<code class="option">--ssl-ca</code> nor <code class="option">--ssl-capath</code>. The
server still verifies the client according to any applicable
requirements established via <code class="literal">GRANT</code> statements
for the client, and it still uses any
<code class="option">--ssl-ca</code>/<code class="option">--ssl-capath</code> values
that were passed to server at startup time.
(<a href="http://bugs.mysql.com/25309" target="_top">Bug#25309</a>)</p></li><li><p>
The server now includes a timestamp in error messages that are
logged as a result of unhandled signals (such as <code class="literal">mysqld
got signal 11</code> messages).
(<a href="http://bugs.mysql.com/24878" target="_top">Bug#24878</a>)</p></li><li><p>
The syntax for index hints has been extended to enable explicit
specification that the hint applies only to join processing. See
<a href="sql-syntax.html#index-hints" title="12.2.7.2. Index Hint Syntax">Section 12.2.7.2, “Index Hint Syntax”</a>.
</p><p>
This is a new fix for this issue, and replaces the fix made in
MySQL 5.0.25 and reverted in 5.0.26.
(<a href="http://bugs.mysql.com/21174" target="_top">Bug#21174</a>)</p></li><li><p>
Added the <code class="option">--secure-file-priv</code> option for
<span><strong class="command">mysqld</strong></span>, which limits the effect of the
<a href="functions.html#function_load-file"><code class="literal">LOAD_FILE()</code></a> function and the
<code class="literal">LOAD DATA</code> and <code class="literal">SELECT ... INTO
OUTFILE</code> statements to work only with files in a given
directory.
(<a href="http://bugs.mysql.com/18628" target="_top">Bug#18628</a>)</p></li><li><p>
Binary distributions for some platforms did not include shared
libraries; now shared libraries are shipped for all platforms
except AIX 5.2 64-bit. <span class="emphasis"><em>Exception</em></span>: The
library for the <code class="literal">libmysqld</code> embedded server is
not shared except on Windows.
(<a href="http://bugs.mysql.com/16520" target="_top">Bug#16520</a>, <a href="http://bugs.mysql.com/26767" target="_top">Bug#26767</a>, <a href="http://bugs.mysql.com/13450" target="_top">Bug#13450</a>)</p></li><li><p>
Added the <code class="literal">hostname</code> system variable, which the
server sets at startup to the server hostname.
</p></li><li><p>
The <span><strong class="command">mysql_create_system_tables</strong></span> script was
removed because <span><strong class="command">mysql_install_db</strong></span> no longer
uses it in MySQL 5.0.
</p></li><li><p>
To satisfy different user requirements, we provide several
servers. <span><strong class="command">mysqld</strong></span> is an optimized server that
is a smaller, faster binary. Each package now also includes
<span><strong class="command">mysqld-debug</strong></span>, which is compiled with
debugging support but is otherwise configured identically to the
non-debug server.
</p></li></ul></div><p>Bugs fixed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
<code class="literal">INSERT DELAYED</code> statements are not supported
for <code class="literal">MERGE</code> tables, but the
<code class="literal">MERGE</code> storage engine was not rejecting such
statements, resulting in table corruption. Applications
previously using <code class="literal">INSERT DELAYED</code> into
<code class="literal">MERGE</code> table will break when upgrading to
versions with this fix. To avoid the problem, remove
<code class="literal">DELAYED</code> from such statements.
(<a href="http://bugs.mysql.com/26464" target="_top">Bug#26464</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">NDB</code> tables having <code class="literal">MEDIUMINT
AUTO_INCREMENT</code> columns were not restored correctly by
<span><strong class="command">ndb_restore</strong></span>, causing spurious duplicate key
errors. This issue did not affect <code class="literal">TINYINT</code>,
<code class="literal">INT</code>, or <code class="literal">BIGINT</code> columns
with <code class="literal">AUTO_INCREMENT</code>.
(<a href="http://bugs.mysql.com/27775" target="_top">Bug#27775</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">NDB</code> tables with indexes whose names contained
space characters were not restored correctly by
<span><strong class="command">ndb_restore</strong></span> (the index names were truncated).
(<a href="http://bugs.mysql.com/27758" target="_top">Bug#27758</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Under certain rare circumstances performing a <code class="literal">DROP
TABLE</code> or <code class="literal">TRUNCATE</code> on an
<code class="literal">NDB</code> table could cause a node failure or
forced cluster shutdown.
(<a href="http://bugs.mysql.com/27581" target="_top">Bug#27581</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Memory usage of a <span><strong class="command">mysqld</strong></span> process grew even
while idle.
(<a href="http://bugs.mysql.com/27560" target="_top">Bug#27560</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
It was not possible to set
<code class="literal">LockPagesInMainMemory</code> equal to
<code class="literal">0</code>.
(<a href="http://bugs.mysql.com/27291" target="_top">Bug#27291</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A race condition could sometimes occur if the node acting as
master failed while node IDs were still being allocated during
startup.
(<a href="http://bugs.mysql.com/27286" target="_top">Bug#27286</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
When a data node was taking over as the master node, a race
condition could sometimes occur as the node was assuming
responsibility for handling of global checkpoints.
(<a href="http://bugs.mysql.com/27283" target="_top">Bug#27283</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Error messages displayed when running in single user mode were
inconsistent.
(<a href="http://bugs.mysql.com/27021" target="_top">Bug#27021</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The failure of a data node while restarting could cause other
data nodes to hang or crash.
(<a href="http://bugs.mysql.com/27003" target="_top">Bug#27003</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
On Solaris, the value of an <code class="literal">NDB</code> table column
declared as <code class="literal">BIT(33)</code> was always displayed as
<code class="literal">0</code>.
(<a href="http://bugs.mysql.com/26986" target="_top">Bug#26986</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<span><strong class="command">mysqld</strong></span> processes would sometimes crash under
high load.
(<a href="http://bugs.mysql.com/26825" target="_top">Bug#26825</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The output from <span><strong class="command">ndb_restore
<code class="option">--print_data</code></strong></span> was incorrect for a
backup made of a database containing tables with
<code class="literal">TINYINT</code> or <code class="literal">SMALLINT</code>
columns.
(<a href="http://bugs.mysql.com/26740" target="_top">Bug#26740</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
An inadvertent use of unaligned data caused
<span><strong class="command">ndb_restore</strong></span> to fail on some 64-bit platforms,
including Sparc and Itanium-2.
(<a href="http://bugs.mysql.com/26739" target="_top">Bug#26739</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
An invalid pointer was returned following a
<code class="literal">FSCLOSECONF</code> signal when accessing the REDO
logs during a node restart or system restart.
(<a href="http://bugs.mysql.com/26515" target="_top">Bug#26515</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The failure of a data node when restarting it with
<code class="option">--initial</code> could lead to failures of subsequent
data node restarts.
(<a href="http://bugs.mysql.com/26481" target="_top">Bug#26481</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Takeover for local checkpointing due to multiple failures of
master nodes was sometimes incorrectly handled.
(<a href="http://bugs.mysql.com/26457" target="_top">Bug#26457</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <code class="literal">LockPagesInMemory</code> parameter was not read
until after distributed communication had already started
between cluster nodes. When the value of this parameter was
<code class="literal">1</code>, this could sometimes result in data node
failure due to missed heartbeats.
(<a href="http://bugs.mysql.com/26454" target="_top">Bug#26454</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Under some circumstances, following the restart of a management
node, all data nodes would connect to it normally, but some of
them subsequently failed to log any events to the management
node.
(<a href="http://bugs.mysql.com/26293" target="_top">Bug#26293</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
In some cases, <code class="literal">AFTER UPDATE</code> and
<code class="literal">AFTER DELETE</code> triggers on
<code class="literal">NDB</code> tables that referenced subject table did
not see the results of operation which caused invocation of the
trigger, but rather saw the row as it was prior to the update or
delete operation.
</p><p>
This was most noticeable when an update operation used a
subquery to obtain the rows to be updated. An example would be
<code class="literal">UPDATE tbl1 SET col2 = val1 WHERE tbl1.col1 IN (SELECT
col3 FROM tbl2 WHERE c4 = val2)</code> where there was an
<code class="literal">AFTER UPDATE</code> trigger on table
<code class="literal">tbl1</code>. In such cases, the trigger would fail
to execute.
</p><p>
The problem occurred because the actual update or delete
operations were deferred to be able to perform them later as one
batch. The fix for this bug solves the problem by disabling this
optimization for a given update or delete if the table has an
<code class="literal">AFTER</code> trigger defined for this operation.
(<a href="http://bugs.mysql.com/26242" target="_top">Bug#26242</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Condition pushdown did not work with prepared statements.
(<a href="http://bugs.mysql.com/26225" target="_top">Bug#26225</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Joins on multiple tables containing <code class="literal">BLOB</code>
columns could cause data nodes run out of memory, and to crash
with the error <span class="errortext">NdbObjectIdMap::expand unable to
expand</span>.
(<a href="http://bugs.mysql.com/26176" target="_top">Bug#26176</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
After entering single user mode it was not possible to alter
non-<code class="literal">NDB</code> tables on any SQL nodes other than
the one having sole access to the cluster.
(<a href="http://bugs.mysql.com/25275" target="_top">Bug#25275</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The management client command <code class="literal">
<em class="replaceable"><code>node_id</code></em> STATUS</code> displayed
the message <code class="literal">Node <em class="replaceable"><code>node_id</code></em>:
not connected</code> when <em class="replaceable"><code>node_id</code></em>
was not the node ID of a data node.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
The <code class="literal">ALL STATUS</code> command in the cluster
management client still displays status information for data
nodes only. This is by design. See
<a href="mysql-cluster.html#mysql-cluster-mgm-client-commands" title="16.6.2. Commands in the MySQL Cluster Management Client">Section 16.6.2, “Commands in the MySQL Cluster Management Client”</a>, for more
information.
</p></div><p>(<a href="http://bugs.mysql.com/21715" target="_top">Bug#21715</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The message <span class="errortext">Error 0 in readAutoIncrementValue(): no
Error</span> was written to the error log whenever
<code class="literal">SHOW TABLE STATUS</code> was performed on a Cluster
table that did not have an <code class="literal">AUTO_INCREMENT</code>
column.
(<a href="http://bugs.mysql.com/21033" target="_top">Bug#21033</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Some values of <code class="literal">MaxNoOfTables</code> caused the error
<span class="errortext">Job buffer congestion</span> to occur.
(<a href="http://bugs.mysql.com/19378" target="_top">Bug#19378</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
When trying to create tables on an SQL node not connected to the
cluster, a misleading error message <span class="errortext">Table
'<em class="replaceable"><code>tbl_name</code></em>' already exists</span>
was generated. The error now generated is <span class="errortext">Could not
connect to storage engine</span>.
(<a href="http://bugs.mysql.com/18676" target="_top">Bug#18676</a>)</p></li><li><p><span class="bold"><strong>Replication</strong></span>:
Some queries that updated multiple tables were not backed up
correctly.
(<a href="http://bugs.mysql.com/27748" target="_top">Bug#27748</a>)</p></li><li><p><span class="bold"><strong>Replication</strong></span>:
Out-of-memory errors were not reported. Now they are written to
the error log.
(<a href="http://bugs.mysql.com/26844" target="_top">Bug#26844</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
Using <code class="literal">NdbBlob::writeData()</code> to write data in
the middle of an existing blob value (that is, updating the
value) could overwrite some data past the end of the data to be
changed.
(<a href="http://bugs.mysql.com/27018" target="_top">Bug#27018</a>)</p></li><li><p>
Some equi-joins containing a <code class="literal">WHERE</code> clause
that included a <code class="literal">NOT IN</code> subquery caused a
server crash.
(<a href="http://bugs.mysql.com/27870" target="_top">Bug#27870</a>)</p></li><li><p>
<code class="literal">SELECT DISTINCT</code> could return incorrect
results if the select list contained duplicated columns.
(<a href="http://bugs.mysql.com/27659" target="_top">Bug#27659</a>)</p></li><li><p>
With <code class="literal">NO_AUTO_VALUE_ON_ZERO</code> SQL mode enabled,
<code class="literal">LOAD DATA</code> operations could assign incorrect
<code class="literal">AUTO_INCREMENT</code> values.
(<a href="http://bugs.mysql.com/27586" target="_top">Bug#27586</a>)</p></li><li><p>
Incorrect results could be returned for some queries that
contained a select list expression with <code class="literal">IN</code> or
<a href="functions.html#operator_between"><code class="literal">BETWEEN</code></a> together with an
<code class="literal">ORDER BY</code> or <code class="literal">GROUP BY</code> on
the same expression using <code class="literal">NOT IN</code> or
<code class="literal">NOT BETWEEN</code>.
(<a href="http://bugs.mysql.com/27532" target="_top">Bug#27532</a>)</p></li><li><p>
Evaluation of an <code class="literal">IN()</code> predicate containing a
decimal-valued argument caused a server crash.
(<a href="http://bugs.mysql.com/27513" target="_top">Bug#27513</a>, <a href="http://bugs.mysql.com/27362" target="_top">Bug#27362</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-2583" target="_top">CVE-2007-2583</a>)</p></li><li><p>
Profiling overhead was incurred even with profiling disabled.
(<a href="http://bugs.mysql.com/27501" target="_top">Bug#27501</a>)</p></li><li><p>
In out-of-memory conditions, the server might crash or otherwise
not report an error to the Windows event log.
(<a href="http://bugs.mysql.com/27490" target="_top">Bug#27490</a>)</p></li><li><p>
Passing nested row expressions with different structures to an
<code class="literal">IN</code> predicate caused a server crash.
(<a href="http://bugs.mysql.com/27484" target="_top">Bug#27484</a>)</p></li><li><p>
The <code class="filename">decimal.h</code> header file was incorrectly
omitted from binary distributions.
(<a href="http://bugs.mysql.com/27456" target="_top">Bug#27456</a>)</p></li><li><p>
With <code class="literal">innodb_file_per_table</code> enabled,
attempting to rename an <code class="literal">InnoDB</code> table to a
non-existent database caused the server to exit.
(<a href="http://bugs.mysql.com/27381" target="_top">Bug#27381</a>)</p></li><li><p>
A subquery could get incorrect values for references to outer
query columns when it contained aggregate functions that were
aggregated in outer context.
(<a href="http://bugs.mysql.com/27321" target="_top">Bug#27321</a>)</p></li><li><p>
In a view, a column that was defined using a
<code class="literal">GEOMETRY</code> function was treated as having the
<code class="literal">LONGBLOB</code> data type rather than the
<code class="literal">GEOMETRY</code> type.
(<a href="http://bugs.mysql.com/27300" target="_top">Bug#27300</a>)</p></li><li><p>
Queries containing subqueries with
<a href="functions.html#function_count"><code class="literal">COUNT(*)</code></a> aggregated in an outer
context returned incorrect results. This happened only if the
subquery did not contain any references to outer columns.
(<a href="http://bugs.mysql.com/27257" target="_top">Bug#27257</a>)</p></li><li><p>
<code class="literal">SELECT ... INTO OUTFILE</code> with a long
<code class="literal">FIELDS ENCLOSED BY</code> value could crash the
server.
(<a href="http://bugs.mysql.com/27231" target="_top">Bug#27231</a>)</p></li><li><p>
Use of an aggregate function from an outer context as an
argument to <a href="functions.html#function_group-concat"><code class="literal">GROUP_CONCAT()</code></a>
caused a server crash.
(<a href="http://bugs.mysql.com/27229" target="_top">Bug#27229</a>)</p></li><li><p>
String truncation upon insertion into an integer or year column
did not generate a warning (or an error in strict mode).
(<a href="http://bugs.mysql.com/27176" target="_top">Bug#27176</a>, <a href="http://bugs.mysql.com/26359" target="_top">Bug#26359</a>)</p></li><li><p>
Storing <code class="literal">NULL</code> values in spatial fields caused
excessive memory allocation and crashes on some systems.
(<a href="http://bugs.mysql.com/27164" target="_top">Bug#27164</a>)</p></li><li><p>
Row equalities in <code class="literal">WHERE</code> clauses could cause
memory corruption.
(<a href="http://bugs.mysql.com/27154" target="_top">Bug#27154</a>)</p></li><li><p>
<code class="literal">GROUP BY</code> on a <code class="literal">ucs2</code> column
caused a server crash when there was at least one empty string
in the column.
(<a href="http://bugs.mysql.com/27079" target="_top">Bug#27079</a>)</p></li><li><p>
Duplicate members in <code class="literal">SET</code> definitions were not
detected. Now they result in a warning; if strict SQL mode is
enabled, an error occurs instead.
(<a href="http://bugs.mysql.com/27069" target="_top">Bug#27069</a>)</p></li><li><p>
For <code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code>
statements on tables containing
<code class="literal">AUTO_INCREMENT</code> columns,
<a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a> was reset to 0
if no rows were successfully inserted or changed. “<span class="quote">Not
changed</span>” includes the case where a row was updated to its
current values, but in that case,
<a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a> should not be
reset to 0. Now <a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a>
is reset to 0 only if no rows were successfully inserted or
touched, whether or not touched rows were changed.
(<a href="http://bugs.mysql.com/27033" target="_top">Bug#27033</a>)</p><p>See also <a href="http://bugs.mysql.com/27210" target="_top">Bug#27210</a>, <a href="http://bugs.mysql.com/27006" target="_top">Bug#27006</a></p><p>This regression was introduced by <a href="http://bugs.mysql.com/19978" target="_top">Bug#19978</a></p></li><li><p>
<span><strong class="command">mysql_install_db</strong></span> could terminate with an
error after failing to determine that a system table already
existed.
(<a href="http://bugs.mysql.com/27022" target="_top">Bug#27022</a>)</p></li><li><p>
<code class="literal">AFTER UPDATE</code> triggers were not activated by
the update part of <code class="literal">INSERT ... ON DUPLICATE KEY
UPDATE</code> statements.
(<a href="http://bugs.mysql.com/27006" target="_top">Bug#27006</a>)</p><p>See also <a href="http://bugs.mysql.com/27033" target="_top">Bug#27033</a>, <a href="http://bugs.mysql.com/27210" target="_top">Bug#27210</a></p><p>This regression was introduced by <a href="http://bugs.mysql.com/19978" target="_top">Bug#19978</a></p></li><li><p>
In a <code class="literal">MEMORY</code> table, using a
<code class="literal">BTREE</code> index to scan for updatable rows could
lead to an infinite loop.
(<a href="http://bugs.mysql.com/26996" target="_top">Bug#26996</a>)</p></li><li><p>
Invalid optimization of pushdown conditions for queries where an
outer join was guaranteed to read only one row from the outer
table led to results with too few rows.
(<a href="http://bugs.mysql.com/26963" target="_top">Bug#26963</a>)</p></li><li><p>
Windows binaries contained no debug symbol file. Now
<code class="filename">.map</code> and <code class="literal">.pdb</code> files are
included in 32-bit builds for <span><strong class="command">mysqld-nt.exe</strong></span>,
<span><strong class="command">mysqld-debug.exe</strong></span>, and
<span><strong class="command">mysqlmanager.exe</strong></span>.
(<a href="http://bugs.mysql.com/26893" target="_top">Bug#26893</a>)</p></li><li><p>
For <code class="literal">MERGE</code> tables defined on underlying tables
that contained a short <code class="literal">VARCHAR</code> column
(shorter than four characters), using <code class="literal">ALTER
TABLE</code> on at least one but not all of the underlying
tables caused the table definitions to be considered different
from that of the <code class="literal">MERGE</code> table, even if the
<code class="literal">ALTER TABLE</code> did not change the definition.
(<a href="http://bugs.mysql.com/26881" target="_top">Bug#26881</a>)</p></li><li><p>
Improved out-of-memory detection when sending logs from a master
server to slaves, and log a message when allocation fails.
(<a href="http://bugs.mysql.com/26837" target="_top">Bug#26837</a>)</p></li><li><p>
For <code class="literal">InnoDB</code> tables having a clustered index
that began with a <code class="literal">CHAR</code> or
<code class="literal">VARCHAR</code> column, deleting a record and then
inserting another before the deleted record was purged could
result in table corruption.
(<a href="http://bugs.mysql.com/26835" target="_top">Bug#26835</a>)</p></li><li><p>
Use of a subquery containing <code class="literal">GROUP BY</code> and
<code class="literal">WITH ROLLUP</code> caused a server crash.
(<a href="http://bugs.mysql.com/26830" target="_top">Bug#26830</a>)</p></li><li><p>
Duplicates were not properly identified among (potentially) long
strings used as arguments for
<a href="functions.html#function_group-concat"><code class="literal">GROUP_CONCAT(DISTINCT)</code></a>.
(<a href="http://bugs.mysql.com/26815" target="_top">Bug#26815</a>)</p></li><li><p>
<code class="literal">ALTER VIEW</code> requires the <code class="literal">CREATE
VIEW</code> and <code class="literal">DROP</code> privileges for the
view. However, if the view was created by another user, the
server erroneously required the <code class="literal">SUPER</code>
privilege.
(<a href="http://bugs.mysql.com/26813" target="_top">Bug#26813</a>)</p></li><li><p>
Added support for <code class="option">--debugger=dbx</code> for
<span><strong class="command">mysql-test-run.pl</strong></span> and fixed support for
<code class="option">--debugger=devenv</code>,
<code class="option">--debugger=DevEnv</code>, and
<code class="option">--debugger=<em class="replaceable"><code>/path/to</code></em>/devenv</code>.
(<a href="http://bugs.mysql.com/26792" target="_top">Bug#26792</a>)</p></li><li><p>
A result set column formed by concatention of string literals
was incomplete when the column was produced by a subquery in the
<code class="literal">FROM</code> clause.
(<a href="http://bugs.mysql.com/26738" target="_top">Bug#26738</a>)</p></li><li><p>
SSL connections failed on Windows.
(<a href="http://bugs.mysql.com/26678" target="_top">Bug#26678</a>)</p></li><li><p>
When using the result of
<a href="functions.html#function_sec-to-time"><code class="literal">SEC_TO_TIME()</code></a> for time value
greater than 24 hours in an <code class="literal">ORDER BY</code> clause,
either directly or through a column alias, the rows were sorted
incorrectly as strings.
(<a href="http://bugs.mysql.com/26672" target="_top">Bug#26672</a>)</p></li><li><p>
Use of a subquery containing a <code class="literal">UNION</code> with an
invalid <code class="literal">ORDER BY</code> clause caused a server
crash.
(<a href="http://bugs.mysql.com/26661" target="_top">Bug#26661</a>)</p></li><li><p>
The range optimizer could cause the server to run out of memory.
(<a href="http://bugs.mysql.com/26625" target="_top">Bug#26625</a>)</p></li><li><p>
The range optimizer could consume a combinatorial amount of
memory for certain classes of <code class="literal">WHERE</code> clauses.
(<a href="http://bugs.mysql.com/26624" target="_top">Bug#26624</a>)</p></li><li><p>
In some error messages, inconsistent format specifiers were used
for the translations in different languages.
<span><strong class="command">comp_err</strong></span> (the error message compiler) now
checks for mismatches.
(<a href="http://bugs.mysql.com/26571" target="_top">Bug#26571</a>)</p></li><li><p>
Views that used a scalar correlated subquery returned incorrect
results.
(<a href="http://bugs.mysql.com/26560" target="_top">Bug#26560</a>)</p></li><li><p>
<code class="literal">UNHEX() IS NULL</code> comparisons failed when
<a href="functions.html#function_unhex"><code class="literal">UNHEX()</code></a> returned
<code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/26537" target="_top">Bug#26537</a>)</p></li><li><p>
On 64-bit Windows, large timestamp values could be handled
incorrectly.
(<a href="http://bugs.mysql.com/26536" target="_top">Bug#26536</a>)</p></li><li><p>
<code class="literal">mysqldump</code> could crash or exhibit incorrect
behavior when some options were given very long values, such as
<code class="option">--fields-terminated-by="<em class="replaceable"><code>some very long
string</code></em>"</code>. The code has been cleaned
up to remove a number of fixed-sized buffers and to be more
careful about error conditions in memory allocation.
(<a href="http://bugs.mysql.com/26346" target="_top">Bug#26346</a>)</p></li><li><p>
If the server was started with
<code class="option">--skip-grant-tables</code>, Selecting from
<code class="literal">INFORMATION_SCHEMA</code> tables causes a server
crash.
(<a href="http://bugs.mysql.com/26285" target="_top">Bug#26285</a>)</p></li><li><p>
For some values of the position argument, the
<a href="functions.html#function_insert"><code class="literal">INSERT()</code></a> function could insert a
NUL byte into the result.
(<a href="http://bugs.mysql.com/26281" target="_top">Bug#26281</a>)</p></li><li><p>
For an <code class="literal">INSERT</code> statement that should fail due
to a column with no default value not being assigned a value,
the statement succeeded with no error if the column was assigned
a value in an <code class="literal">ON DUPLICATE KEY UPDATE</code> clause,
even if that clause was not used.
(<a href="http://bugs.mysql.com/26261" target="_top">Bug#26261</a>)</p></li><li><p>
<code class="literal">INSERT DELAYED</code> statements inserted incorrect
values into <code class="literal">BIT</code> columns.
(<a href="http://bugs.mysql.com/26238" target="_top">Bug#26238</a>)</p></li><li><p>
The temporary file-creation code was cleaned up on Windows to
improve server stability.
(<a href="http://bugs.mysql.com/26233" target="_top">Bug#26233</a>)</p></li><li><p>
For <code class="literal">MyISAM</code> tables,
<a href="functions.html#function_count"><code class="literal">COUNT(*)</code></a> could return an
incorrect value if the <code class="literal">WHERE</code> clause compared
an indexed <code class="literal">TEXT</code> column to the empty string
(<code class="literal">''</code>). This happened if the column contained
empty strings and also strings starting with control characters
such as tab or newline.
(<a href="http://bugs.mysql.com/26231" target="_top">Bug#26231</a>)</p></li><li><p>
For <code class="literal">INSERT INTO ... SELECT</code> where index
searches used column prefixes, insert errors could occur when
key value type conversion was done.
(<a href="http://bugs.mysql.com/26207" target="_top">Bug#26207</a>)</p></li><li><p>
For <code class="literal">DELETE FROM <code class="literal">tbl_name</code> ORDER BY
<em class="replaceable"><code>col_name</code></em> </code> (with no
<code class="literal">WHERE</code> or <code class="literal">LIMIT</code> clause),
the server did not check whether
<em class="replaceable"><code>col_name</code></em> was a valid column in the
table.
(<a href="http://bugs.mysql.com/26186" target="_top">Bug#26186</a>)</p></li><li><p>
<code class="literal">REPAIR TABLE ... USE_FRM</code> with an
<code class="literal">ARCHIVE</code> table deleted all records from the
table.
(<a href="http://bugs.mysql.com/26138" target="_top">Bug#26138</a>)</p></li><li><p>
A multiple-row delayed insert with an auto-increment column
could cause duplicate entries to be created on the slave in a
replication environment.
(<a href="http://bugs.mysql.com/26116" target="_top">Bug#26116</a>, <a href="http://bugs.mysql.com/25507" target="_top">Bug#25507</a>)</p></li><li><p>
<a href="functions.html#function_benchmark"><code class="literal">BENCHMARK()</code></a> did not work
correctly for expressions that produced a
<code class="literal">DECIMAL</code> result.
(<a href="http://bugs.mysql.com/26093" target="_top">Bug#26093</a>)</p></li><li><p>
<code class="literal">LOAD DATA INFILE</code> sent an okay to the client
before writing the binary log and committing the changes to the
table had finished, thus violating ACID requirements.
(<a href="http://bugs.mysql.com/26050" target="_top">Bug#26050</a>)</p></li><li><p>
<code class="literal">X() IS NULL</code> and <code class="literal">Y() IS
NULL</code> comparisons failed when
<a href="spatial-extensions.html#function_x"><code class="literal">X()</code></a> and
<a href="spatial-extensions.html#function_y"><code class="literal">Y()</code></a> returned
<code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/26038" target="_top">Bug#26038</a>)</p></li><li><p>
<span><strong class="command">mysqldump</strong></span> crashed for
<code class="literal">MERGE</code> tables if the
<code class="option">--complete-insert</code> (<code class="option">-c</code>) option
was given.
(<a href="http://bugs.mysql.com/25993" target="_top">Bug#25993</a>)</p></li><li><p>
Indexes on <code class="literal">TEXT</code> columns were ignored when
<code class="literal">ref</code> accesses were evaluated.
(<a href="http://bugs.mysql.com/25971" target="_top">Bug#25971</a>)</p></li><li><p>
If a thread previously serviced a connection that was killed,
excessive memory and CPU use by the thread occurred if it later
serviced a connection that had to wait for a table lock.
(<a href="http://bugs.mysql.com/25966" target="_top">Bug#25966</a>)</p></li><li><p>
Setting a column to <code class="literal">NOT NULL</code> with an
<code class="literal">ON DELETE SET NULL</code> clause foreign key crashes
the server.
(<a href="http://bugs.mysql.com/25927" target="_top">Bug#25927</a>)</p></li><li><p>
<code class="literal">VIEW</code> restrictions were applied to
<code class="literal">SELECT</code> statements after a <code class="literal">CREATE
VIEW</code> statement failed, as though the
<code class="literal">CREATE</code> had succeeded.
(<a href="http://bugs.mysql.com/25897" target="_top">Bug#25897</a>)</p></li><li><p>
Several deficiencies in resolution of column names for
<code class="literal">INSERT ... SELECT</code> statements were corrected.
(<a href="http://bugs.mysql.com/25831" target="_top">Bug#25831</a>)</p></li><li><p>
Inserting <code class="literal">utf8</code> data into a
<code class="literal">TEXT</code> column that used a single-byte character
set could result in spurious warnings about truncated data.
(<a href="http://bugs.mysql.com/25815" target="_top">Bug#25815</a>)</p></li><li><p>
On Windows, debug builds of <span><strong class="command">mysqld</strong></span> could fail
with heap assertions.
(<a href="http://bugs.mysql.com/25765" target="_top">Bug#25765</a>)</p></li><li><p>
In certain situations, <code class="literal">MATCH ... AGAINST</code>
returned false hits for <code class="literal">NULL</code> values produced
by <code class="literal">LEFT JOIN</code> when no full-text index was
available.
(<a href="http://bugs.mysql.com/25729" target="_top">Bug#25729</a>)</p></li><li><p>
In certain cases it could happen that deleting a row corrupted
an <code class="literal">RTREE</code> index. This affected indexes on
spatial columns.
(<a href="http://bugs.mysql.com/25673" target="_top">Bug#25673</a>)</p></li><li><p>
When <a href="functions.html#function_rand"><code class="literal">RAND()</code></a> was called multiple
times inside a stored procedure, the server did not write the
correct random seed values to the binary log, resulting in
incorrect replication.
(<a href="http://bugs.mysql.com/25543" target="_top">Bug#25543</a>)</p></li><li><p>
<code class="literal">OPTIMIZE TABLE</code> might fail on Windows when it
attempts to rename a temporary file to the original name if the
original file had been opened, resulting in loss of the
<code class="filename">.MYD</code> file.
(<a href="http://bugs.mysql.com/25521" target="_top">Bug#25521</a>)</p></li><li><p>
For <code class="literal">SHOW ENGINE INNODB STATUS</code>, the
<code class="literal">LATEST DEADLOCK INFORMATION</code> was not always
cleared properly.
(<a href="http://bugs.mysql.com/25494" target="_top">Bug#25494</a>)</p></li><li><p>
<a href="apis.html#mysql-stmt-fetch" title="23.2.7.11. mysql_stmt_fetch()"><code class="literal">mysql_stmt_fetch()</code></a> did an
invalid memory deallocation when used with the embedded server.
(<a href="http://bugs.mysql.com/25492" target="_top">Bug#25492</a>)</p></li><li><p>
<code class="literal">GRANT</code> statements were not replicated if the
server was started with the
<code class="option">--replicate-ignore-table</code> or
<code class="option">--replicate-wild-ignore-table</code> option.
(<a href="http://bugs.mysql.com/25482" target="_top">Bug#25482</a>)</p></li><li><p>
Expressions involving <a href="functions.html#function_sum"><code class="literal">SUM()</code></a>,
when used in an <code class="literal">ORDER BY</code> clause, could lead
to out-of-order results.
(<a href="http://bugs.mysql.com/25376" target="_top">Bug#25376</a>)</p></li><li><p>
Use of a <code class="literal">GROUP BY</code> clause that referred to a
stored function result together with <code class="literal">WITH
ROLLUP</code> caused incorrect results.
(<a href="http://bugs.mysql.com/25373" target="_top">Bug#25373</a>)</p></li><li><p>
A stored procedure that made use of cursors failed when the
procedure was invoked from a stored function.
(<a href="http://bugs.mysql.com/25345" target="_top">Bug#25345</a>)</p></li><li><p>
Difficult repair or optimization operations could cause an
assertion failure, resulting in a server crash.
(<a href="http://bugs.mysql.com/25289" target="_top">Bug#25289</a>)</p></li><li><p>
On Windows, the server exhibited a file-handle leak after
reaching the limit on the number of open file descriptors.
(<a href="http://bugs.mysql.com/25222" target="_top">Bug#25222</a>)</p></li><li><p>
The <a href="functions.html#function_repeat"><code class="literal">REPEAT()</code></a> function did not
allow a column name as the <em class="replaceable"><code>count</code></em>
parameter.
(<a href="http://bugs.mysql.com/25197" target="_top">Bug#25197</a>)</p></li><li><p>
Duplicating the usage of a user variable in a stored procedure
or trigger would not be replicated correctly to the slave.
(<a href="http://bugs.mysql.com/25167" target="_top">Bug#25167</a>)</p></li><li><p>
A reference to a non-existent column in the <code class="literal">ORDER
BY</code> clause of an <code class="literal">UPDATE ... ORDER BY</code>
statement could cause a server crash.
(<a href="http://bugs.mysql.com/25126" target="_top">Bug#25126</a>)</p></li><li><p>
A view on a join is insertable for <code class="literal">INSERT</code>
statements that store values into only one table of the join.
However, inserts were being rejected if the inserted-into table
was used in a self-join because MySQL incorrectly was
considering the insert to modify multiple tables of the view.
(<a href="http://bugs.mysql.com/25122" target="_top">Bug#25122</a>)</p></li><li><p>
MySQL would not compile when configured using
<code class="option">--without-query-cache</code>.
(<a href="http://bugs.mysql.com/25075" target="_top">Bug#25075</a>)</p></li><li><p>
Duplicate entries were not assessed correctly in a
<code class="literal">MEMORY</code> table with a <code class="literal">BTREE</code>
primary key on a <code class="literal">utf8</code> <code class="literal">ENUM</code>
column.
(<a href="http://bugs.mysql.com/24985" target="_top">Bug#24985</a>)</p></li><li><p>
Selecting the result of <a href="functions.html#function_avg"><code class="literal">AVG()</code></a>
within a <code class="literal">UNION</code> could produce incorrect
values.
(<a href="http://bugs.mysql.com/24791" target="_top">Bug#24791</a>)</p></li><li><p>
<code class="literal">MBROverlaps()</code> returned incorrect values in
some cases.
(<a href="http://bugs.mysql.com/24563" target="_top">Bug#24563</a>)</p></li><li><p>
Increasing the width of a <code class="literal">DECIMAL</code> column
could cause column values to be changed.
(<a href="http://bugs.mysql.com/24558" target="_top">Bug#24558</a>)</p></li><li><p>
IF(<em class="replaceable"><code>expr</code></em>,
<em class="replaceable"><code>unsigned_expr</code></em>,
<em class="replaceable"><code>unsigned_expr</code></em>) was evaluated to a
signed result, not unsigned. This has been corrected. The fix
also affects constructs of the form <code class="literal">IS [NOT]
{TRUE|FALSE}</code>, which were transformed internally into
<a href="functions.html#function_if"><code class="literal">IF()</code></a> expressions that evaluated
to a signed result.
</p><p>
For existing views that were defined using <code class="literal">IS [NOT]
{TRUE|FALSE}</code> constructs, there is a related
implication. The definitions of such views were stored using the
<a href="functions.html#function_if"><code class="literal">IF()</code></a> expression, not the
original construct. This is manifest in that <code class="literal">SHOW
CREATE VIEW</code> shows the transformed
<a href="functions.html#function_if"><code class="literal">IF()</code></a> expression, not the
original one. Existing views will evaluate correctly after the
fix, but if you want <code class="literal">SHOW CREATE VIEW</code> to
display the original construct, you must drop the view and
re-create it using its original definition. New views will
retain the construct in their definition.
(<a href="http://bugs.mysql.com/24532" target="_top">Bug#24532</a>)</p></li><li><p>
A problem in handling of aggregate functions in subqueries
caused predicates containing aggregate functions to be ignored
during query execution.
(<a href="http://bugs.mysql.com/24484" target="_top">Bug#24484</a>)</p></li><li><p>
<code class="literal">DROP TRIGGER</code> statements would not be filtered
on the slave when using the
<code class="literal">replication-wild-do-table</code> option.
(<a href="http://bugs.mysql.com/24478" target="_top">Bug#24478</a>)</p></li><li><p>
For <code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code>
statements where some <code class="literal">AUTO_INCREMENT</code> values
were generated automatically for inserts and some rows were
updated, one auto-generated value was lost per updated row,
leading to faster exhaustion of the range of the
<code class="literal">AUTO_INCREMENT</code> column.
</p><p>
Because the original problem can affect replication (different
values on master and slave), it is recommended that the master
and its slaves be upgraded to the current version.
(<a href="http://bugs.mysql.com/24432" target="_top">Bug#24432</a>)</p></li><li><p>
The test for the
<code class="literal">MYSQL_OPT_SSL_VERIFY_SERVER_CERT</code> option for
<a href="apis.html#mysql-options" title="23.2.3.49. mysql_options()"><code class="literal">mysql_options()</code></a> was performed
incorrectly. Also changed as a result of this bugfix: The
<code class="literal">arg</code> option for the
<a href="apis.html#mysql-options" title="23.2.3.49. mysql_options()"><code class="literal">mysql_options()</code></a> C API function
was changed from <code class="literal">char *</code> to <code class="literal">void
*</code>.
(<a href="http://bugs.mysql.com/24121" target="_top">Bug#24121</a>)</p></li><li><p>
A user-defined variable could be assigned an incorrect value if
a temporary table was employed in obtaining the result of the
query used to determine its value.
(<a href="http://bugs.mysql.com/24010" target="_top">Bug#24010</a>)</p></li><li><p>
Queries that used a temporary table for the outer query when
evaluating a correlated subquery could return incorrect results.
(<a href="http://bugs.mysql.com/23800" target="_top">Bug#23800</a>)</p></li><li><p>
Replication between master and slave would infinitely retry
binary log transmission where the
<code class="option">max_allowed_packet</code> on the master was larger
than that on the slave if the size of the transfer was between
these two values.
(<a href="http://bugs.mysql.com/23775" target="_top">Bug#23775</a>)</p></li><li><p>
On Windows, debug builds of <span><strong class="command">mysqlbinlog</strong></span> could
fail with a memory error.
(<a href="http://bugs.mysql.com/23736" target="_top">Bug#23736</a>)</p></li><li><p>
When using certain server SQL modes, the
<code class="literal">mysql.proc</code> table was not created by
<span><strong class="command">mysql_install_db</strong></span>.
(<a href="http://bugs.mysql.com/23669" target="_top">Bug#23669</a>)</p></li><li><p>
The values displayed for the
<code class="literal">Innodb_row_lock_time</code>,
<code class="literal">Innodb_row_lock_time_avg</code>, and
<code class="literal">Innodb_row_lock_time_max</code> status variables
were incorrect.
(<a href="http://bugs.mysql.com/23666" target="_top">Bug#23666</a>)</p></li><li><p>
<code class="literal">DOUBLE</code> values such as
<code class="literal">20070202191048.000000</code> were being treated as
illegal arguments by <a href="functions.html#function_week"><code class="literal">WEEK()</code></a>.
(<a href="http://bugs.mysql.com/23616" target="_top">Bug#23616</a>)</p></li><li><p>
The server could crash if two or more threads initiated query
cache resize operation at moments very close in time.
(<a href="http://bugs.mysql.com/23527" target="_top">Bug#23527</a>)</p></li><li><p>
<code class="literal">SHOW CREATE VIEW</code> qualified references to
stored functions in the view definition with the function's
database name, even when the database was the default database.
This affected <span><strong class="command">mysqldump</strong></span> (which uses
<code class="literal">SHOW CREATE VIEW</code> to dump views) because the
resulting dump file could not be used to reload the database
into a different database. <code class="literal">SHOW CREATE VIEW</code>
now suppresses the database name for references to functions in
the default database.
(<a href="http://bugs.mysql.com/23491" target="_top">Bug#23491</a>)</p></li><li><p>
An <code class="literal">INTO OUTFILE</code> clause is allowed only for
the final <code class="literal">SELECT</code> of a
<code class="literal">UNION</code>, but this restriction was not being
enforced correctly.
(<a href="http://bugs.mysql.com/23345" target="_top">Bug#23345</a>)</p></li><li><p>
<a href="functions.html#function_now"><code class="literal">NOW()</code></a> returned the wrong value
in statements executed at server startup with the
<code class="option">--init-file</code> option.
(<a href="http://bugs.mysql.com/23240" target="_top">Bug#23240</a>)</p></li><li><p>
With the <code class="literal">NO_AUTO_VALUE_ON_ZERO</code> SQL mode
enabled, <a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a> could
return 0 after <code class="literal">INSERT ... ON DUPLICATE KEY
UPDATE</code>. Additionally, the next rows inserted (by the
same <code class="literal">INSERT</code>, or the following
<code class="literal">INSERT</code> with or without <code class="literal">ON DUPLICATE
KEY UPDATE</code>), would insert 0 for the auto-generated
value if the value for the <code class="literal">AUTO_INCREMENT</code>
column was <code class="literal">NULL</code> or missing.
(<a href="http://bugs.mysql.com/23233" target="_top">Bug#23233</a>)</p></li><li><p>
<a href="functions.html#function_soundex"><code class="literal">SOUNDEX()</code></a> returned an invalid
string for international characters in multi-byte character
sets.
(<a href="http://bugs.mysql.com/22638" target="_top">Bug#22638</a>)</p></li><li><p>
When nesting stored procedures within a trigger on a table, a
false dependency error was thrown when one of the nested
procedures contained a <code class="literal">DROP TABLE</code> statement.
(<a href="http://bugs.mysql.com/22580" target="_top">Bug#22580</a>)</p></li><li><p>
Instance Manager did not remove the angel PID file on a clean
shutdown.
(<a href="http://bugs.mysql.com/22511" target="_top">Bug#22511</a>)</p></li><li><p>
<code class="literal">EXPLAIN EXTENDED</code> did not show
<code class="literal">WHERE</code> conditions that were optimized away.
(<a href="http://bugs.mysql.com/22331" target="_top">Bug#22331</a>)</p></li><li><p>
<a href="functions.html#function_count"><code class="literal">COUNT(<em class="replaceable"><code>decimal_expr</code></em>)</code></a>
sometimes generated a spurious truncation warning.
(<a href="http://bugs.mysql.com/21976" target="_top">Bug#21976</a>)</p></li><li><p>
<code class="literal">IN ((<em class="replaceable"><code>subquery</code></em>))</code>,
<code class="literal">IN (((<em class="replaceable"><code>subquery</code></em>)))</code>,
and so forth, are equivalent to <code class="literal">IN
(<em class="replaceable"><code>subquery</code></em>)</code>, which is always
interpreted as a table subquery (so that it is allowed to return
more than one row). MySQL was treating the
“<span class="quote">over-parenthesized</span>” subquery as a single-row
subquery and rejecting it if it returned more than one row. This
bug primarily affected automatically generated code (such as
queries generated by Hibernate), because humans rarely write the
over-parenthesized forms.
(<a href="http://bugs.mysql.com/21904" target="_top">Bug#21904</a>)</p></li><li><p>
An <code class="literal">INSERT</code> trigger invoking a stored routine
that inserted into a table other than the one on which the
trigger was defined would fail with a <span class="errortext">Table '...'
doesn't exist</span> referring to the second table when
attempting to delete records from the first table.
(<a href="http://bugs.mysql.com/21825" target="_top">Bug#21825</a>)</p></li><li><p>
For <code class="literal">InnoDB</code>, fixed consistent-read behavior of
the first read statement, if the read was served from the query
cache, for the <code class="literal">READ COMMITTED</code> isolation
level.
(<a href="http://bugs.mysql.com/21409" target="_top">Bug#21409</a>)</p></li><li><p>
<a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> is less than
<a href="functions.html#function_now"><code class="literal">NOW()</code></a>, either when comparing
<a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> directly
(<code class="literal">CURDATE() < NOW()</code> is true) or when
casting <a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> to
<code class="literal">DATE</code> (<code class="literal">CAST(CURDATE() AS DATE) <
NOW()</code> is true). However, storing
<a href="functions.html#function_curdate"><code class="literal">CURDATE()</code></a> in a
<code class="literal">DATE</code> column and comparing <code class="literal">
<em class="replaceable"><code>col_name</code></em> < NOW()</code>
incorrectly yielded false. This is fixed by comparing a
<code class="literal">DATE</code> column as <code class="literal">DATETIME</code>
for comparisons to a <code class="literal">DATETIME</code> constant.
(<a href="http://bugs.mysql.com/21103" target="_top">Bug#21103</a>)</p></li><li><p>
When a stored routine attempted to execute a statement accessing
a nonexistent table, the error was not caught by the routine's
exception handler.
(<a href="http://bugs.mysql.com/20713" target="_top">Bug#20713</a>, <a href="http://bugs.mysql.com/8407" target="_top">Bug#8407</a>)</p></li><li><p>
For a stored procedure containing a <code class="literal">SELECT</code>
statement that used a complicated join with an
<code class="literal">ON</code> expression, the expression could be
ignored during re-execution of the procedure, yielding an
incorrect result.
(<a href="http://bugs.mysql.com/20492" target="_top">Bug#20492</a>)</p></li><li><p>
The conditions checked by the optimizer to allow use of indexes
in <code class="literal">IN</code> predicate calculations were
unnecessarily tight and were relaxed.
(<a href="http://bugs.mysql.com/20420" target="_top">Bug#20420</a>)</p></li><li><p>
When a <a href="functions.html#function_time-format"><code class="literal">TIME_FORMAT()</code></a> expression
was used as a column in a <code class="literal">GROUP BY</code> clause,
the expression result was truncated.
(<a href="http://bugs.mysql.com/20293" target="_top">Bug#20293</a>)</p></li><li><p>
The creation of MySQL system tables was not checked for by
<span><strong class="command">mysql-test-run.pl</strong></span>.
(<a href="http://bugs.mysql.com/20166" target="_top">Bug#20166</a>)</p></li><li><p>
For index reads, the <code class="literal">BLACKHOLE</code> engine did not
return end-of-file (which it must because
<code class="literal">BLACKHOLE</code> tables contain no rows), causing
some queries to crash.
(<a href="http://bugs.mysql.com/19717" target="_top">Bug#19717</a>)</p></li><li><p>
In some cases, the optimizer preferred a range or full index
scan access method over lookup access methods when the latter
were much cheaper.
(<a href="http://bugs.mysql.com/19372" target="_top">Bug#19372</a>)</p></li><li><p>
For <code class="literal"> <em class="replaceable"><code>expr</code></em>
IN(<em class="replaceable"><code>value_list</code></em>)</code>, the result
could be incorrect if <code class="literal">BIGINT UNSIGNED</code> values
were used for <em class="replaceable"><code>expr</code></em> or in the value
list.
(<a href="http://bugs.mysql.com/19342" target="_top">Bug#19342</a>)</p></li><li><p>
When attempting to call a stored procedure creating a table from
a trigger on a table <code class="literal">tbl</code> in a database
<code class="literal">db</code>, the trigger failed with <span class="errortext">ERROR
1146 (42S02): Table 'db.tbl' doesn't exist</span>. However,
the actual reason that such a trigger fails is due to the fact
that <code class="literal">CREATE TABLE</code> causes an implicit
<code class="literal">COMMIT</code>, and so a trigger cannot invoke a
stored routine containing this statement. A trigger which does
so now fails with <span class="errortext">ERROR 1422 (HY000): Explicit or
implicit commit is not allowed in stored function or
trigger</span>, which makes clear the reason for the
trigger's failure.
(<a href="http://bugs.mysql.com/18914" target="_top">Bug#18914</a>)</p></li><li><p>
The update columns for <code class="literal">INSERT ... SELECT ... ON
DUPLICATE KEY UPDATE</code> could be assigned incorrect
values if a temporary table was used to evaluate the
<code class="literal">SELECT</code>.
(<a href="http://bugs.mysql.com/16630" target="_top">Bug#16630</a>)</p></li><li><p>
Conversion of <code class="literal">DATETIME</code> values in numeric
contexts sometimes did not produce a double
(<code class="literal">YYYYMMDDHHMMSS.uuuuuu</code>) value.
(<a href="http://bugs.mysql.com/16546" target="_top">Bug#16546</a>)</p></li><li><p>
For <a href="functions.html#function_substring"><code class="literal">SUBSTRING()</code></a> evaluation using
a temporary table, when
<a href="functions.html#function_substring"><code class="literal">SUBSTRING()</code></a> was used on a
LONGTEXT column, the <code class="literal">max_length</code> metadata
value of the result was incorrectly calculated and set to 0.
Consequently, an empty string was returned instead of the
correct result.
(<a href="http://bugs.mysql.com/15757" target="_top">Bug#15757</a>)</p></li><li><p>
Loading data using <code class="literal">LOAD DATA INFILE</code> may not
replicate correctly (due to character set incompatibilities) if
the <code class="literal">character_set_database</code> variable is set
before the data is loaded.
(<a href="http://bugs.mysql.com/15126" target="_top">Bug#15126</a>)</p></li><li><p>
User defined variables used within stored procedures and
triggers are not replicated correctly when operating in
statement-based replication mode.
(<a href="http://bugs.mysql.com/14914" target="_top">Bug#14914</a>, <a href="http://bugs.mysql.com/20141" target="_top">Bug#20141</a>)</p></li><li><p>
Local variables in stored routines or triggers, when declared as
the <code class="literal">BIT</code> type, were interpreted as strings.
(<a href="http://bugs.mysql.com/12976" target="_top">Bug#12976</a>)</p></li><li><p>
<code class="literal">CONNECTION</code> is no longer treated as a reserved
word.
(<a href="http://bugs.mysql.com/12204" target="_top">Bug#12204</a>)</p></li></ul></div></div><div class="section" lang="en"><div class="titlepage"><div><div><h3 class="title"><a name="releasenotes-cs-5-0-37"></a>D.1.4. Release Notes for MySQL Community Server 5.0.37 (27 February 2007)</h3></div></div></div><p class="release-level">
This is a bugfix release for the current production release
family. It replaces MySQL 5.0.33.
</p><p>Functionality added or changed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Incompatible Change</strong></span>: <span class="bold"><strong>MySQL Cluster</strong></span>:
The <code class="literal">LockPagesInMainMemory</code> configuration
parameter has changed its type and possible values. For more
information, see
<code class="literal"><a href="mysql-cluster.html#mysql-cluster-param-ndbd-definition-lockpagesinmainmemory">LockPagesInMainMemory</a></code>.
</p><div class="important" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Important</h3><p>
The values <code class="literal">true</code> and
<code class="literal">false</code> are no longer accepted for this
parameter. If you were using this parameter and had it set to
<code class="literal">false</code> in a previous release, you must
change it to <code class="literal">0</code>. If you had this parameter
set to <code class="literal">true</code>, you should instead use
<code class="literal">1</code> to obtain the same behavior as
previously, or <code class="literal">2</code> to take advantage of new
functionality introduced with this release, as described in
the section cited above.
</p></div><p>(<a href="http://bugs.mysql.com/25686" target="_top">Bug#25686</a>)</p></li><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
Previously, the <a href="functions.html#function_date-format"><code class="literal">DATE_FORMAT()</code></a>
function returned a binary string. Now it returns a string with
a character set and collation given by
<code class="literal">character_set_connection</code> and
<code class="literal">collation_connection</code> so that it can return
month and weekday names containing non-ASCII characters.
(<a href="http://bugs.mysql.com/22646" target="_top">Bug#22646</a>)</p></li><li><p>
Added the <code class="literal">Uptime_since_flush_status</code> status
variable, which indicates the number of seconds since the most
recent <code class="literal">FLUSH STATUS</code> statement. (From Jeremy
Cole)
(<a href="http://bugs.mysql.com/24822" target="_top">Bug#24822</a>)</p></li><li><p>
Added the <code class="literal">SHOW PROFILES</code> and <code class="literal">SHOW
PROFILE</code> statements to display statement profile data,
and the accompanying
<code class="literal">INFORMATION_SCHEMA.PROFILING</code> table. Profiling
is controlled via the <code class="literal">profiling</code> and
<code class="literal">profiling_history_size</code> session variables. see
<a href="sql-syntax.html#show-profiles" title="12.5.4.22. SHOW PROFILES and SHOW PROFILE
Syntax">Section 12.5.4.22, “<code class="literal">SHOW PROFILES</code> and <code class="literal">SHOW PROFILE</code>
Syntax”</a>, and
<a href="information-schema.html#profiling-table" title="21.17. The INFORMATION_SCHEMA PROFILING Table">Section 21.17, “The <code class="literal">INFORMATION_SCHEMA PROFILING</code> Table”</a>. (From Jeremy Cole)
(<a href="http://bugs.mysql.com/24795" target="_top">Bug#24795</a>)</p></li><li><p>
The <code class="literal">localhost</code> anonymous user account created
during MySQL installation on Windows now has no global
privileges. Formerly this account had all global privileges. For
operations that require global privileges, the
<code class="literal">root</code> account can be used instead.
(<a href="http://bugs.mysql.com/24496" target="_top">Bug#24496</a>)</p></li><li><p>
The <code class="option">--skip-thread-priority</code> option now is
enabled by default for binary Mac OS X distributions. Use of
thread priorities degrades performance on Mac OS X.
(<a href="http://bugs.mysql.com/18526" target="_top">Bug#18526</a>)</p></li><li><div class="important" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Important</h3><p>
When using <code class="literal">MERGE</code> tables the definition of
the <code class="literal">MERGE</code> table and the
<code class="literal">MyISAM</code> tables are checked each time the
tables are opened for access (including any
<code class="literal">SELECT</code> or <code class="literal">INSERT</code>
statement. Each table is compared for column order, types,
sizes and associated. If there is a difference in any one of
the tables then the statement will fail.
</p></div></li><li><p>
This is the last version for which MySQL-Max RPM distributions
are available. (This change was already made for non-RPM binary
distributions in 5.0.27.)
</p></li><li><p>
The bundled yaSSL library was upgraded to version 1.5.8.
</p></li><li><p>
Added the <code class="option">--disable-grant-options</code> option to
<span><strong class="command">configure</strong></span>. If <span><strong class="command">configure</strong></span> is
run with this option, the <code class="option">--bootstrap</code>,
<code class="option">--skip-grant-tables</code>, and
<code class="option">--init-file</code> options for
<span><strong class="command">mysqld</strong></span> are disabled and cannot be used. For
Windows, the <span><strong class="command">configure.js</strong></span> script recognizes
the <code class="literal">DISABLE_GRANT_OPTIONS</code> flag, which has the
same effect.
</p></li></ul></div><p>Bugs fixed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Security Fix</strong></span>:
Using an <code class="literal">INFORMATION_SCHEMA</code> table with
<code class="literal">ORDER BY</code> in a subquery could cause a server
crash.
</p><p>
We would like to thank Oren Isacson of Flowgate Security
Consulting and Stefan Streichsbier of SEC Consult for informing
us of this problem.
(<a href="http://bugs.mysql.com/24630" target="_top">Bug#24630</a>, <a href="http://bugs.mysql.com/26556" target="_top">Bug#26556</a>, <a href="http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-1420" target="_top">CVE-2007-1420</a>)</p></li><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
For <code class="literal">ENUM</code> columns that had enumeration values
containing commas, the commas were mapped to
<code class="literal">0xff</code> internally. However, this rendered the
commas indistinguishable from true <code class="literal">0xff</code>
characters in the values. This no longer occurs. However, the
fix requires that you dump and reload any tables that have
<code class="literal">ENUM</code> columns containing any true
<code class="literal">0xff</code> values. Dump the tables using
<span><strong class="command">mysqldump</strong></span> with the current server before
upgrading from a version of MySQL 5.0 older than 5.0.36 to
version 5.0.36 or newer.
(<a href="http://bugs.mysql.com/24660" target="_top">Bug#24660</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A query with an <code class="literal">IN</code> clause against an
<code class="literal">NDB</code> table employing explicit user-defined
partitioning did not always return all matching rows.
(<a href="http://bugs.mysql.com/25821" target="_top">Bug#25821</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
It was not possible to create an <code class="literal">NDB</code> table
with a key on two <code class="literal">VARCHAR</code> columns where both
columns had a storage length in excess of 256.
(<a href="http://bugs.mysql.com/25746" target="_top">Bug#25746</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Hosts in clusters with large numbers of nodes could experience
excessive CPU usage while obtaining configuration data.
(<a href="http://bugs.mysql.com/25711" target="_top">Bug#25711</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
In some circumstances, shutting down the cluster could cause
connected <span><strong class="command">mysqld</strong></span> processes to crash.
(<a href="http://bugs.mysql.com/25668" target="_top">Bug#25668</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Memory allocations for <code class="literal">TEXT</code> columns were
calculated incorrectly, resulting in space being wasted and
other issues.
(<a href="http://bugs.mysql.com/25562" target="_top">Bug#25562</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The failure of a master node during a node restart could lead to
a resource leak, causing later node failures.
(<a href="http://bugs.mysql.com/25554" target="_top">Bug#25554</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
An <code class="literal">UPDATE</code> using an <code class="literal">IN</code>
clause on an <code class="literal">NDB</code> table on which there was a
trigger caused <span><strong class="command">mysqld</strong></span> to crash.
(<a href="http://bugs.mysql.com/25522" target="_top">Bug#25522</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A node shutdown occurred if the master failed during a commit.
(<a href="http://bugs.mysql.com/25364" target="_top">Bug#25364</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Creating a non-unique index with the <code class="literal">USING
HASH</code> clause silently created an ordered index instead
of issuing a warning.
(<a href="http://bugs.mysql.com/24820" target="_top">Bug#24820</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <code class="filename">ndb_size.tmpl</code> file (necessary for using
the <code class="filename">ndb_size.pl</code> script) was missing from
binary distributions.
(<a href="http://bugs.mysql.com/24191" target="_top">Bug#24191</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
When a data node was shut down using the management client
<code class="literal">STOP</code> command, a connection event
(<code class="literal">NDB_LE_Connected</code>) was logged instead of a
disconnection event (<code class="literal">NDB_LE_Disconnected</code>).
(<a href="http://bugs.mysql.com/22773" target="_top">Bug#22773</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The management server did not handle logging of node shutdown
events correctly in certain cases.
(<a href="http://bugs.mysql.com/22013" target="_top">Bug#22013</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">SELECT</code> statements with a
<code class="literal">BLOB</code> or <code class="literal">TEXT</code> column in the
selected column list and a <code class="literal">WHERE</code> condition
including a primary key lookup on a <code class="literal">VARCHAR</code>
primary key produced empty result sets.
(<a href="http://bugs.mysql.com/19956" target="_top">Bug#19956</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
Deletion of an <code class="literal">Ndb_cluster_connection</code> object
took a very long time.
(<a href="http://bugs.mysql.com/25487" target="_top">Bug#25487</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
Invoking the <code class="literal">NdbTransaction::execute()</code> method
using execution type <code class="literal">Commit</code> and abort option
<code class="literal">AO_IgnoreError</code> could lead to a crash of the
transaction coordinator (<code class="literal">DBTC</code>).
(<a href="http://bugs.mysql.com/25090" target="_top">Bug#25090</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
A unique index lookup on a non-existent tuple could lead to a
data node timeout (error 4012).
(<a href="http://bugs.mysql.com/25059" target="_top">Bug#25059</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
<code class="filename">libndbclient.so</code> was not versioned.
(<a href="http://bugs.mysql.com/13522" target="_top">Bug#13522</a>)</p></li><li><p>
Using <code class="literal">ORDER BY</code> or <code class="literal">GROUP BY</code>
could yield different results when selecting from a view and
selecting from the underlying table.
(<a href="http://bugs.mysql.com/26209" target="_top">Bug#26209</a>)</p></li><li><p>
<code class="literal">DISTINCT</code> queries that were executed using a
loose scan for an <code class="literal">InnoDB</code> table that had been
emptied caused a server crash.
(<a href="http://bugs.mysql.com/26159" target="_top">Bug#26159</a>)</p></li><li><p>
A <code class="literal">WHERE</code> clause that used
<a href="functions.html#operator_between"><code class="literal">BETWEEN</code></a> for
<code class="literal">DATETIME</code> values could be treated differently
for a <code class="literal">SELECT</code> and a view defined as that
<code class="literal">SELECT</code>.
(<a href="http://bugs.mysql.com/26124" target="_top">Bug#26124</a>)</p></li><li><p>
Collation for <code class="literal">LEFT JOIN</code> comparisons could be
evaluated incorrectly, leading to improper query results.
(<a href="http://bugs.mysql.com/26017" target="_top">Bug#26017</a>)</p></li><li><p>
The <code class="literal">WITH CHECK OPTION</code> clause for views was
ignored for updates of multiple-table views when the updates
could not be performed on fly and the rows to update had to be
put into temporary tables first.
(<a href="http://bugs.mysql.com/25931" target="_top">Bug#25931</a>)</p></li><li><p>
<code class="literal">LOAD DATA INFILE</code> did not work with pipes.
(<a href="http://bugs.mysql.com/25807" target="_top">Bug#25807</a>)</p></li><li><p>
The <a href="functions.html#function_sec-to-time"><code class="literal">SEC_TO_TIME()</code></a> and
<a href="functions.html#function_quarter"><code class="literal">QUARTER()</code></a> functions sometimes
did not handle <code class="literal">NULL</code> values correctly.
(<a href="http://bugs.mysql.com/25643" target="_top">Bug#25643</a>)</p></li><li><p>
The <code class="literal">InnoDB</code> parser sometimes did not account
for null bytes, causing spurious failure of some queries.
(<a href="http://bugs.mysql.com/25596" target="_top">Bug#25596</a>)</p></li><li><p>
View definitions that used the <code class="literal">!</code> operator
were treated as containing the <code class="literal">NOT</code> operator,
which has a different precedence and can produce different
results. .
(<a href="http://bugs.mysql.com/25580" target="_top">Bug#25580</a>)</p></li><li><p>
An error in the name resolution of nested <code class="literal">JOIN ...
USING</code> constructs was corrected.
(<a href="http://bugs.mysql.com/25575" target="_top">Bug#25575</a>)</p></li><li><p>
<code class="literal">GROUP BY</code> and <code class="literal">DISTINCT</code> did
not group <code class="literal">NULL</code> values for columns that have a
<code class="literal">UNIQUE</code> index. .
(<a href="http://bugs.mysql.com/25551" target="_top">Bug#25551</a>)</p></li><li><p>
The <code class="option">--with-readline</code> option for
<span><strong class="command">configure</strong></span> did not work for commercial source
packages, but no error message was printed to that effect. Now a
message is printed.
(<a href="http://bugs.mysql.com/25530" target="_top">Bug#25530</a>)</p></li><li><p>
<a href="apis.html#mysql-stmt-fetch" title="23.2.7.11. mysql_stmt_fetch()"><code class="literal">mysql_stmt_fetch()</code></a> did an
invalid memory deallocation when used with the embedded server.
(<a href="http://bugs.mysql.com/25492" target="_top">Bug#25492</a>)</p></li><li><p>
Referencing an ambiguous column alias in an expression in the
<code class="literal">ORDER BY</code> clause of a query caused the server
to crash.
(<a href="http://bugs.mysql.com/25427" target="_top">Bug#25427</a>)</p></li><li><p>
A yaSSL program named <span><strong class="command">test</strong></span> was installed,
causing conflicts with the <span><strong class="command">test</strong></span> system
utility. It is no longer installed.
(<a href="http://bugs.mysql.com/25417" target="_top">Bug#25417</a>)</p></li><li><p>
For a <code class="literal">UNIQUE</code> index containing many
<code class="literal">NULL</code> values, the optimizer would prefer the
index for <code class="literal"> <em class="replaceable"><code>col</code></em> IS
NULL</code> conditions over other more selective indexes. .
(<a href="http://bugs.mysql.com/25407" target="_top">Bug#25407</a>)</p></li><li><p>
An <code class="literal">AFTER UPDATE</code> trigger on an
<code class="literal">InnoDB</code> table with a composite primary key
caused the server to crash.
(<a href="http://bugs.mysql.com/25398" target="_top">Bug#25398</a>)</p></li><li><p>
Passing a <code class="literal">NULL</code> value to a user-defined
function from within a stored procedure crashes the server.
(<a href="http://bugs.mysql.com/25382" target="_top">Bug#25382</a>)</p></li><li><p>
<span><strong class="command">perror</strong></span> crashed on some platforms due to
failure to handle a <code class="literal">NULL</code> pointer.
(<a href="http://bugs.mysql.com/25344" target="_top">Bug#25344</a>)</p></li><li><p>
<span><strong class="command">mysql.server stop</strong></span> timed out too quickly (35
seconds) waiting for the server to exit. Now it waits up to 15
minutes, to ensure that the server exits.
(<a href="http://bugs.mysql.com/25341" target="_top">Bug#25341</a>)</p></li><li><p>
A query that contained an <code class="literal">EXIST</code> subquery with
a <code class="literal">UNION</code> over correlated and uncorrelated
<code class="literal">SELECT</code> queries could cause the server to
crash.
(<a href="http://bugs.mysql.com/25219" target="_top">Bug#25219</a>)</p></li><li><p>
<a href="apis.html#mysql-kill" title="23.2.3.38. mysql_kill()"><code class="literal">mysql_kill()</code></a> caused a server
crash when used on an SSL connection.
(<a href="http://bugs.mysql.com/25203" target="_top">Bug#25203</a>)</p></li><li><p>
yaSSL was sensitive to the presence of whitespace at the ends of
lines in PEM-encoded certificates, causing a server crash.
(<a href="http://bugs.mysql.com/25189" target="_top">Bug#25189</a>)</p></li><li><p>
A query with <code class="literal">ORDER BY</code> and <code class="literal">GROUP
BY</code> clauses where the <code class="literal">ORDER BY</code>
clause had more elements than the <code class="literal">GROUP BY</code>
clause caused a memory overrun leading to a crash of the server.
(<a href="http://bugs.mysql.com/25172" target="_top">Bug#25172</a>)</p></li><li><p>
Use of <code class="literal">ON DUPLICATE KEY UPDATE</code> defeated the
usual restriction against inserting into a join-based view
unless only one of the underlying tables is used.
(<a href="http://bugs.mysql.com/25123" target="_top">Bug#25123</a>)</p></li><li><p>
Using a view in combination with a <code class="literal">USING</code>
clause caused column aliases to be ignored.
(<a href="http://bugs.mysql.com/25106" target="_top">Bug#25106</a>)</p></li><li><p>
A multiple-table <code class="literal">DELETE QUICK</code> could sometimes
cause one of the affected tables to become corrupted.
(<a href="http://bugs.mysql.com/25048" target="_top">Bug#25048</a>)</p></li><li><p>
<code class="literal">ALTER TABLE ... ENABLE KEYS</code> acquired a global
lock, preventing concurrent execution of other statements that
use tables. .
(<a href="http://bugs.mysql.com/25044" target="_top">Bug#25044</a>)</p></li><li><p>
An assertion failed incorrectly for prepared statements that
contained a single-row uncorrelated subquery that was used as an
argument of the <a href="functions.html#operator_is-null"><code class="literal">IS NULL</code></a>
predicate.
(<a href="http://bugs.mysql.com/25027" target="_top">Bug#25027</a>)</p></li><li><p>
A return value of <code class="literal">-1</code> from user-defined
handlers was not handled well and could result in conflicts with
server code.
(<a href="http://bugs.mysql.com/24987" target="_top">Bug#24987</a>)</p></li><li><p>
Accessing a fixed record format table with a crashed key
definition results in server/<span><strong class="command">myisamchk</strong></span>
segmentation fault.
(<a href="http://bugs.mysql.com/24855" target="_top">Bug#24855</a>)</p></li><li><p>
<span><strong class="command">mysqld_multi</strong></span> and
<span><strong class="command">mysqlaccess</strong></span> looked for option files in
<code class="filename">/etc</code> even if the
<code class="option">--sysconfdir</code> option for
<span><strong class="command">configure</strong></span> had been given to specify a
different directory.
(<a href="http://bugs.mysql.com/24780" target="_top">Bug#24780</a>)</p></li><li><p>
If there was insufficient memory available to
<span><strong class="command">mysqld</strong></span>, this could sometimes cause the server
to hang during startup.
(<a href="http://bugs.mysql.com/24751" target="_top">Bug#24751</a>)</p></li><li><p>
Optimizations that are legal only for subqueries without tables
and <code class="literal">WHERE</code> conditions were applied for any
subquery without tables.
(<a href="http://bugs.mysql.com/24670" target="_top">Bug#24670</a>)</p></li><li><p>
If an <code class="literal">ORDER BY</code> or <code class="literal">GROUP BY</code>
list included a constant expression being optimized away and, at
the same time, containing single-row subselects that returned
more that one row, no error was reported. If a query required
sorting by expressions containing single-row subselects that
returned more than one row, execution of the query could cause a
server crash.
(<a href="http://bugs.mysql.com/24653" target="_top">Bug#24653</a>)</p></li><li><p>
For <code class="literal">ALTER TABLE</code>, using <code class="literal">ORDER BY
<em class="replaceable"><code>expression</code></em> </code> could cause a
server crash. Now the <code class="literal">ORDER BY</code> clause allows
only column names to be specified as sort criteria (which was
the only documented syntax, anyway).
(<a href="http://bugs.mysql.com/24562" target="_top">Bug#24562</a>)</p></li><li><p>
A workaround was implemented to avoid a race condition in the
NPTL <code class="literal">pthread_exit()</code> implementation.
(<a href="http://bugs.mysql.com/24507" target="_top">Bug#24507</a>)</p></li><li><p>
<span><strong class="command">mysqltest</strong></span> crashed with a stack overflow.
(<a href="http://bugs.mysql.com/24498" target="_top">Bug#24498</a>)</p></li><li><p>
Within stored routines or prepared statements, inconsistent
results occurred with multiple use of <code class="literal">INSERT ... SELECT
... ON DUPLICATE KEY UPDATE</code> when the <code class="literal">ON
DUPLICATE KEY UPDATE</code> clause erroneously tried to
assign a value to a column mentioned only in its
<code class="literal">SELECT</code> part.
(<a href="http://bugs.mysql.com/24491" target="_top">Bug#24491</a>)</p></li><li><p>
Expressions of the form <code class="literal">(a, b) IN (SELECT a, MIN(b)
FROM t GROUP BY a)</code> could produce incorrect results
when column <code class="literal">a</code> of table <code class="literal">t</code>
contained <code class="literal">NULL</code> values while column
<code class="literal">b</code> did not.
(<a href="http://bugs.mysql.com/24420" target="_top">Bug#24420</a>)</p></li><li><p>
If a prepared statement accessed a view, access to the tables
listed in the query after that view was checked in the security
context of the view.
(<a href="http://bugs.mysql.com/24404" target="_top">Bug#24404</a>)</p></li><li><p>
Attempts to access a <code class="literal">MyISAM</code> table with a
corrupt column definition caused a server crash.
(<a href="http://bugs.mysql.com/24401" target="_top">Bug#24401</a>)</p></li><li><p>
When opening a corrupted <code class="filename">.frm</code> file during a
query, the server crashes.
(<a href="http://bugs.mysql.com/24358" target="_top">Bug#24358</a>)</p></li><li><p>
Some joins in which one of the joined tables was a view could
return erroneous results or crash the server.
(<a href="http://bugs.mysql.com/24345" target="_top">Bug#24345</a>)</p></li><li><p>
A view was not handled correctly if the
<code class="literal">SELECT</code> part contained “<span class="quote">
<code class="literal">\Z</code> </span>”.
(<a href="http://bugs.mysql.com/24293" target="_top">Bug#24293</a>)</p></li><li><p>
A query using <code class="literal">WHERE
<em class="replaceable"><code>unsigned_column</code></em> NOT IN
('<em class="replaceable"><code>negative_value</code></em>')</code> could
cause the server to crash.
(<a href="http://bugs.mysql.com/24261" target="_top">Bug#24261</a>)</p></li><li><p>
When <code class="literal">SET PASSWORD</code> was written to the binary
log double quotes were included in the statement. If the slave
was running in with the server SQL mode set to
<code class="literal">ANSI_QUOTES</code>, then the event failed, which
halted the replication process.
(<a href="http://bugs.mysql.com/24158" target="_top">Bug#24158</a>)</p></li><li><p>
Expressions of the form <code class="literal">(a, b) IN (SELECT c, d
...)</code> could produce incorrect results if
<code class="literal">a</code>, <code class="literal">b</code>, or both were
<code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/24127" target="_top">Bug#24127</a>)</p></li><li><p>
A <code class="literal">FETCH</code> statement using a cursor on a table
which was not in the table cache could sometimes cause the
server to crash.
(<a href="http://bugs.mysql.com/24117" target="_top">Bug#24117</a>)</p></li><li><p>
Queries that evaluate <code class="literal">NULL IN (SELECT ... UNION SELECT
...)</code> could produce an incorrect result
(<code class="literal">FALSE</code> instead of <code class="literal">NULL</code>).
(<a href="http://bugs.mysql.com/24085" target="_top">Bug#24085</a>)</p></li><li><p>
Hebrew-to-Unicode conversion failed for some characters.
Definitions for the following Hebrew characters (as specified by
the ISO/IEC 8859-8:1999) were added: LEFT-TO-RIGHT MARK (LRM),
RIGHT-TO-LEFT MARK (RLM)
(<a href="http://bugs.mysql.com/24037" target="_top">Bug#24037</a>)</p></li><li><p>
Some <code class="literal">UPDATE</code> statements were slower than in
previous versions when the search key could not be converted to
a valid value for the type of the search column.
(<a href="http://bugs.mysql.com/24035" target="_top">Bug#24035</a>)</p></li><li><p>
<a href="functions.html#function_isnull"><code class="literal">ISNULL(DATE(NULL))</code></a> and
<a href="functions.html#function_isnull"><code class="literal">ISNULL(CAST(NULL AS DATE))</code></a>
erroneously returned false.
(<a href="http://bugs.mysql.com/23938" target="_top">Bug#23938</a>)</p></li><li><p>
Within a stored routine, accessing a declared routine variable
with <code class="literal">PROCEDURE ANALYSE()</code> caused a server
crash.
(<a href="http://bugs.mysql.com/23782" target="_top">Bug#23782</a>)</p></li><li><p>
When reading from the standard input on Windows,
<span><strong class="command">mysqlbinlog</strong></span> opened the input in text mode
rather than binary mode and consequently misinterpreted some
characters such as Control-Z.
(<a href="http://bugs.mysql.com/23735" target="_top">Bug#23735</a>)</p></li><li><p>
A stored procedure, executed from a connection using a binary
character set, and which wrote multibyte data, would write
incorrectly escaped entries to the binary log. This caused
syntax errors, and caused replication to fail.
(<a href="http://bugs.mysql.com/23619" target="_top">Bug#23619</a>, <a href="http://bugs.mysql.com/24492" target="_top">Bug#24492</a>)</p></li><li><p>
<code class="literal">OPTIMIZE TABLE</code> tried to sort R-tree indexes
such as spatial indexes, although this is not possible (see
<a href="sql-syntax.html#optimize-table" title="12.5.2.5. OPTIMIZE TABLE Syntax">Section 12.5.2.5, “<code class="literal">OPTIMIZE TABLE</code> Syntax”</a>).
(<a href="http://bugs.mysql.com/23578" target="_top">Bug#23578</a>)</p></li><li><p>
For an <code class="literal">InnoDB</code> table with any <code class="literal">ON
DELETE</code> trigger, <code class="literal">TRUNCATE TABLE</code>
mapped to <code class="literal">DELETE</code> and activated triggers. Now
a fast truncation occurs and triggers are not activated. .
(<a href="http://bugs.mysql.com/23556" target="_top">Bug#23556</a>)</p></li><li><p>
The row count for <code class="literal">MyISAM</code> tables was not
updated properly, causing <code class="literal">SHOW TABLE STATUS</code>
to report incorrect values.
(<a href="http://bugs.mysql.com/23526" target="_top">Bug#23526</a>)</p></li><li><p>
User-defined variables could consume excess memory, leading to a
crash caused by the exhaustion of resources available to the
<code class="literal">MEMORY</code> storage engine, due to the fact that
this engine is used by MySQL for variable storage and
intermediate results of <code class="literal">GROUP BY</code> queries.
Where <code class="literal">SET</code> had been used, such a condition
could instead give rise to the misleading error message
<span class="errortext">You may only use constant expressions with
SET</span>, rather than <span class="errortext">Out of memory (Needed
NNNNNN bytes)</span>.
(<a href="http://bugs.mysql.com/23443" target="_top">Bug#23443</a>)</p></li><li><p>
With <code class="literal">ONLY_FULL_GROUP_BY</code> enables, the server
was too strict: Some expressions involving only aggregate values
were rejected as non-aggregate (for example,
<a href="functions.html#function_max"><code class="literal">MAX(a)</code></a> –
<a href="functions.html#function_min"><code class="literal">MIN(a)</code></a>).
(<a href="http://bugs.mysql.com/23417" target="_top">Bug#23417</a>)</p></li><li><p>
The arguments to the <a href="functions.html#function_encode"><code class="literal">ENCODE()</code></a>
and the <a href="functions.html#function_decode"><code class="literal">DECODE()</code></a> functions were
not printed correctly, causing problems in the output of
<code class="literal">EXPLAIN EXTENDED</code> and in view definitions.
(<a href="http://bugs.mysql.com/23409" target="_top">Bug#23409</a>)</p></li><li><p>
A table created with the <code class="literal">ROW_FORMAT = FIXED</code>
table option lost the option if an index was added or dropped
with <code class="literal">CREATE INDEX</code> or <code class="literal">DROP
INDEX</code>.
(<a href="http://bugs.mysql.com/23404" target="_top">Bug#23404</a>)</p></li><li><p>
A deadlock could occur, with the server hanging on
<code class="literal">Closing tables</code>, with a sufficient number of
concurrent <code class="literal">INSERT DELAYED</code>, <code class="literal">FLUSH
TABLES</code>, and <code class="literal">ALTER TABLE</code> operations.
(<a href="http://bugs.mysql.com/23312" target="_top">Bug#23312</a>)</p></li><li><p>
Some queries against <code class="literal">INFORMATION_SCHEMA</code> that
used subqueries failed. .
(<a href="http://bugs.mysql.com/23299" target="_top">Bug#23299</a>)</p></li><li><p>
<code class="literal">readline</code> detection did not work correctly on
NetBSD.
(<a href="http://bugs.mysql.com/23293" target="_top">Bug#23293</a>)</p></li><li><p>
If there was insufficient memory to store or update a blob
record in a <code class="literal">MyISAM</code> table then the table will
marked as crashed.
(<a href="http://bugs.mysql.com/23196" target="_top">Bug#23196</a>)</p></li><li><p>
<a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a> was not reset
to 0 if <code class="literal">INSERT ... SELECT</code> inserted no rows.
(<a href="http://bugs.mysql.com/23170" target="_top">Bug#23170</a>)</p></li><li><p>
A compressed <code class="literal">MyISAM</code> table that became
corrupted could crash <span><strong class="command">myisamchk</strong></span> and possibly
the MySQL Server.
(<a href="http://bugs.mysql.com/23139" target="_top">Bug#23139</a>)</p></li><li><p>
The number of <code class="literal">setsockopt()</code> calls performed
for reads and writes to the network socket was reduced to
decrease system call overhead.
(<a href="http://bugs.mysql.com/22943" target="_top">Bug#22943</a>)</p></li><li><p>
<span><strong class="command">mysql_upgrade</strong></span> failed when called with a
<code class="literal">basedir</code> pathname containing spaces.
(<a href="http://bugs.mysql.com/22801" target="_top">Bug#22801</a>)</p></li><li><p>
<code class="literal">SET lc_time_names = <em class="replaceable"><code>value</code></em>
</code> allowed only exact literal values, not expression
values.
(<a href="http://bugs.mysql.com/22647" target="_top">Bug#22647</a>)</p></li><li><p>
Changes to the <code class="literal">lc_time_names</code> system variable
were not replicated.
(<a href="http://bugs.mysql.com/22645" target="_top">Bug#22645</a>)</p></li><li><p>
The <a href="functions.html#function_stddev"><code class="literal">STDDEV()</code></a> function returned a
positive value for data sets consisting of a single value.
(<a href="http://bugs.mysql.com/22555" target="_top">Bug#22555</a>)</p></li><li><p>
Storing values specified as hexadecimal values 64 or more bits
long in <code class="literal">BIT(64)</code>, <code class="literal">BIGINT</code>,
or <code class="literal">BIGINT UNSIGNED</code> columns did not raise any
warning or error if the value was out of range.
(<a href="http://bugs.mysql.com/22533" target="_top">Bug#22533</a>)</p></li><li><p>
<code class="literal">SHOW COLUMNS</code> reported some <code class="literal">NOT
NULL</code> columns as <code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/22377" target="_top">Bug#22377</a>)</p></li><li><p>
Type conversion errors during formation of index search
conditions were not correctly checked, leading to incorrect
query results.
(<a href="http://bugs.mysql.com/22344" target="_top">Bug#22344</a>)</p></li><li><p>
Changing the value of <code class="literal">MI_KEY_BLOCK_LENGTH</code> in
<code class="filename">myisam.h</code> and recompiling MySQL resulted in
a <span><strong class="command">myisamchk</strong></span> that saw existing
<code class="literal">MyISAM</code> tables as corrupt.
(<a href="http://bugs.mysql.com/22119" target="_top">Bug#22119</a>)</p></li><li><p>
A crash of the MySQL Server could occur when unpacking a
<code class="literal">BLOB</code> column from a row in a corrupted MyISAM
table. This could happen when trying to repair a table using
either <code class="literal">REPAIR TABLE</code> or
<span><strong class="command">myisamchk</strong></span>; it could also happen when trying
to access such a “<span class="quote">broken</span>” row using statements like
<code class="literal">SELECT</code> if the table was not marked as
crashed.
(<a href="http://bugs.mysql.com/22053" target="_top">Bug#22053</a>)</p></li><li><p>
The code for generating <code class="literal">USE</code> statements for
binary logging of <code class="literal">CREATE PROCEDURE</code> statements
resulted in confusing output from <span><strong class="command">mysqlbinlog</strong></span>
for <code class="literal">DROP PROCEDURE</code> statements.
(<a href="http://bugs.mysql.com/22043" target="_top">Bug#22043</a>)</p></li><li><p>
For the <a href="functions.html#function_if"><code class="literal">IF()</code></a> and
<a href="functions.html#function_coalesce"><code class="literal">COALESCE()</code></a> function and
<a href="functions.html#operator_case"><code class="literal">CASE</code></a> expressions, large
unsigned integer values could be mishandled and result in
warnings.
(<a href="http://bugs.mysql.com/22026" target="_top">Bug#22026</a>)</p></li><li><p>
SSL connections could hang at connection shutdown.
(<a href="http://bugs.mysql.com/21781" target="_top">Bug#21781</a>, <a href="http://bugs.mysql.com/24148" target="_top">Bug#24148</a>)</p></li><li><p>
The <code class="literal">FEDERATED</code> storage engine did not support
the <code class="literal">euckr</code> character set.
(<a href="http://bugs.mysql.com/21556" target="_top">Bug#21556</a>)</p></li><li><p>
When updating a table that used a <code class="literal">JOIN</code> of the
table itself (for example, when building trees) and the table
was modified on one side of the expression, the table would
either be reported as crashed or the wrong rows in the table
would be updated.
(<a href="http://bugs.mysql.com/21310" target="_top">Bug#21310</a>)</p></li><li><p>
<code class="filename">mysqld_error.h</code> was not installed when only
the client libraries were built.
(<a href="http://bugs.mysql.com/21265" target="_top">Bug#21265</a>)</p></li><li><p>
<code class="literal">InnoDB</code>: During a restart of the MySQL Server
that followed the creation of a temporary table using the
<code class="literal">InnoDB</code> storage engine, MySQL failed to clean
up in such a way that <code class="literal">InnoDB</code> still attempted
to find the files associated with such tables.
(<a href="http://bugs.mysql.com/20867" target="_top">Bug#20867</a>)</p></li><li><p>
Inserting <code class="literal">DEFAULT</code> into a column with no
default value could result in garbage in the column. Now the
same result occurs as when inserting <code class="literal">NULL</code>
into a <code class="literal">NOT NULL</code> column.
(<a href="http://bugs.mysql.com/20691" target="_top">Bug#20691</a>)</p></li><li><p>
A stored routine containing semicolon in its body could not be
reloaded from a dump of a binary log.
(<a href="http://bugs.mysql.com/20396" target="_top">Bug#20396</a>)</p></li><li><p>
<code class="literal">SELECT ... FOR UPDATE</code>, <code class="literal">SELECT ...
LOCK IN SHARE MODE</code>, <code class="literal">DELETE</code>, and
<code class="literal">UPDATE</code> statements executed using a full table
scan were not releasing locks on rows that did not satisfy the
<code class="literal">WHERE</code> condition.
(<a href="http://bugs.mysql.com/20390" target="_top">Bug#20390</a>)</p></li><li><p>
On Windows, if the server was installed as a service, it did not
auto-detect the location of the data directory.
(<a href="http://bugs.mysql.com/20376" target="_top">Bug#20376</a>)</p></li><li><p>
The <span><strong class="command">BUILD/check-cpu</strong></span> script did not recognize
Celeron processors.
(<a href="http://bugs.mysql.com/20061" target="_top">Bug#20061</a>)</p></li><li><p>
If a duplicate key value was present in the table,
<code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code> reported a
row count indicating that a record was updated, even when no
record actually changed due to the old and new values being the
same. Now it reports a row count of zero.
(<a href="http://bugs.mysql.com/19978" target="_top">Bug#19978</a>)</p><p>See also <a href="http://bugs.mysql.com/27006" target="_top">Bug#27006</a>, <a href="http://bugs.mysql.com/27033" target="_top">Bug#27033</a>, <a href="http://bugs.mysql.com/27210" target="_top">Bug#27210</a></p></li><li><p>
For <code class="literal">SET</code>, <code class="literal">SELECT</code>, and
<code class="literal">DO</code> statements that invoked a stored function
from a database other than the default database, the function
invocation could fail to be replicated.
(<a href="http://bugs.mysql.com/19725" target="_top">Bug#19725</a>)</p></li><li><p>
<code class="literal">ORDER BY</code> values of the
<code class="literal">DOUBLE</code> or <code class="literal">DECIMAL</code> types
could change the result returned by a query.
(<a href="http://bugs.mysql.com/19690" target="_top">Bug#19690</a>)</p></li><li><p>
The <code class="literal">readline</code> library wrote to uninitialized
memory, causing <span><strong class="command">mysql</strong></span> to crash.
(<a href="http://bugs.mysql.com/19474" target="_top">Bug#19474</a>)</p></li><li><p>
<span><strong class="command">mysqltest</strong></span> incorrectly tried to retrieve
result sets for some queries where no result set was available.
(<a href="http://bugs.mysql.com/19410" target="_top">Bug#19410</a>)</p></li><li><p>
Use of already freed memory caused SSL connections to hang
forever.
(<a href="http://bugs.mysql.com/19209" target="_top">Bug#19209</a>)</p></li><li><p>
Some <code class="literal">CASE</code> statements inside stored routines
could lead to excessive resource usage or a crash of the server.
(<a href="http://bugs.mysql.com/19194" target="_top">Bug#19194</a>, <a href="http://bugs.mysql.com/24854" target="_top">Bug#24854</a>)</p></li><li><p>
Instance Manager could crash during shutdown.
(<a href="http://bugs.mysql.com/19044" target="_top">Bug#19044</a>)</p></li><li><p>
The server might fail to use an appropriate index for
<code class="literal">DELETE</code> when <code class="literal">ORDER BY</code>,
<code class="literal">LIMIT</code>, and a non-restricting
<code class="literal">WHERE</code> are present.
(<a href="http://bugs.mysql.com/17711" target="_top">Bug#17711</a>)</p></li><li><p>
No warning was issued for use of the <code class="literal">DATA
DIRECTORY</code> or <code class="literal">INDEX DIRECTORY</code> table
options on a platform that does not support them.
(<a href="http://bugs.mysql.com/17498" target="_top">Bug#17498</a>)</p></li><li><p>
The <code class="literal">FEDERATED</code> storage engine did not support
the <code class="literal">utf8</code> character set.
(<a href="http://bugs.mysql.com/17044" target="_top">Bug#17044</a>)</p></li><li><p>
The optimizer removes expressions from <code class="literal">GROUP
BY</code> and <code class="literal">DISTINCT</code> clauses if they
happen to participate in <code class="literal">
<em class="replaceable"><code>expression</code></em> =
<em class="replaceable"><code>constant</code></em> </code> predicates of the
<code class="literal">WHERE</code> clause, the idea being that, if the
expression is equal to a constant, then it cannot take on
multiple values. However, for predicates where the expression
and the constant item are of different result types (for
example, when a string column is compared to 0), this is not
valid, and can lead to invalid results in such cases. The
optimizer now performs an additional check of the result types
of the expression and the constant; if their types differ, then
the expression is not removed from the <code class="literal">GROUP
BY</code> list.
(<a href="http://bugs.mysql.com/15881" target="_top">Bug#15881</a>)</p></li><li><p>
When a prepared statement failed during the prepare operation,
the error code was not cleared when it was reused, even if the
subsequent use was successful.
(<a href="http://bugs.mysql.com/15518" target="_top">Bug#15518</a>)</p></li><li><p>
Dropping a user-defined function sometimes did not remove the
UDF entry from the <code class="literal">mysql.proc</code> table.
(<a href="http://bugs.mysql.com/15439" target="_top">Bug#15439</a>)</p></li><li><p>
Inserting a row into a table without specifying a value for a
<code class="literal">BINARY(<em class="replaceable"><code>N</code></em>) NOT NULL</code>
column caused the column to be set to spaces, not zeroes.
(<a href="http://bugs.mysql.com/14171" target="_top">Bug#14171</a>)</p></li><li><p>
On Windows, the <a href="functions.html#function_sleep"><code class="literal">SLEEP()</code></a> function
could sleep too long, especially after a change to the system
clock.
(<a href="http://bugs.mysql.com/14094" target="_top">Bug#14094</a>, <a href="http://bugs.mysql.com/24686" target="_top">Bug#24686</a>, <a href="http://bugs.mysql.com/17635" target="_top">Bug#17635</a>)</p></li><li><p>
<span><strong class="command">mysqldump --order-by-primary</strong></span> failed if the
primary key name was an identifier that required quoting.
(<a href="http://bugs.mysql.com/13926" target="_top">Bug#13926</a>)</p></li><li><p>
To enable installation of MySQL RPMs on Linux systems running
RHEL 4 (which includes SE-Linux) additional information was
provided to specify some actions that are allowed to the MySQL
binaries.
(<a href="http://bugs.mysql.com/12676" target="_top">Bug#12676</a>)</p></li><li><p>
The presence of <code class="literal">ORDER BY</code> in a view definition
prevented the <code class="literal">MERGE</code> algorithm from being used
to resolve the view even if nothing else in the definition
required the <code class="literal">TEMPTABLE</code> algorithm.
(<a href="http://bugs.mysql.com/12122" target="_top">Bug#12122</a>)</p></li><li><p>
If a slave server closed its relay log (for example, due to an
error during log rotation), the I/O thread did not recognize
this and still tried to write to the log, causing a server
crash.
(<a href="http://bugs.mysql.com/10798" target="_top">Bug#10798</a>)</p></li><li><p>
The internal functions for table preparation, creation, and
alteration were not re-execution friendly, causing problems in
code that: repeatedly altered a table; repeatedly created and
dropped a table; opened and closed a cursor on a table, altered
the table, and then reopened the cursor; used <code class="literal">ALTER
TABLE</code> to change a table's current
<code class="literal">AUTO_INCREMENT</code> value; created indexes on
<code class="literal">utf8</code> columns.
</p><p>
Re-execution of <code class="literal">CREATE DATABASE</code>,
<code class="literal">CREATE TABLE</code>, and <code class="literal">ALTER
TABLE</code> statements in stored routines or as prepared
statements also caused incorrect results or crashes.
(<a href="http://bugs.mysql.com/4968" target="_top">Bug#4968</a>, <a href="http://bugs.mysql.com/6895" target="_top">Bug#6895</a>, <a href="http://bugs.mysql.com/19182" target="_top">Bug#19182</a>, <a href="http://bugs.mysql.com/19733" target="_top">Bug#19733</a>, <a href="http://bugs.mysql.com/22060" target="_top">Bug#22060</a>, <a href="http://bugs.mysql.com/24879" target="_top">Bug#24879</a>)</p></li></ul></div></div><div class="section" lang="en"><div class="titlepage"><div><div><h3 class="title"><a name="releasenotes-cs-5-0-33"></a>D.1.5. Release Notes for MySQL Community Server 5.0.33 (09 January 2007)</h3></div></div></div><p class="release-level">
This is a bugfix release for the current production release
family. It replaces MySQL 5.0.27.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
This version of MySQL Community Server has been released as a
source tarball only; there are no binaries built by MySQL.
</p></div><p>Functionality added or changed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
<code class="literal">InnoDB</code> rolls back only the last statement on
a transaction timeout. A new option,
<code class="option">--innodb_rollback_on_timeout</code>, causes
<code class="literal">InnoDB</code> to abort and roll back the entire
transaction if a transaction timeout occurs (the same behavior
as in MySQL 5.0.13 and earlier).
(<a href="http://bugs.mysql.com/24200" target="_top">Bug#24200</a>)</p></li><li><p><span class="bold"><strong>Incompatible Change</strong></span>:
The <code class="literal">prepared_stmt_count</code> system variable has
been converted to the <code class="literal">Prepared_stmt_count</code>
global status variable (viewable with the <code class="literal">SHOW GLOBAL
STATUS</code> statement).
(<a href="http://bugs.mysql.com/23159" target="_top">Bug#23159</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Setting the configuration parameter
<code class="literal">LockPagesInMainMemory</code> had no effect.
(<a href="http://bugs.mysql.com/24461" target="_top">Bug#24461</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <span><strong class="command">ndb_config</strong></span> utility now accepts
<code class="option">-c</code> as a short form of the
<code class="option">--ndb-connectstring</code> option.
(<a href="http://bugs.mysql.com/22295" target="_top">Bug#22295</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Added the --bind-address option for <span><strong class="command">ndbd</strong></span>.
This allows a data node process to be bound to a specific
network interface.
(<a href="http://bugs.mysql.com/22195" target="_top">Bug#22195</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
It is now possible to create a unique hashed index on a column
that is not defined as <code class="literal">NOT NULL</code>.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
This change applies only to tables using the
<code class="literal">NDB</code> storage engine.
</p></div><p>
Unique indexes on columns in <code class="literal">NDB</code> tables do
not store null values because they are mapped to primary keys in
an internal index table (and primary keys cannot contain nulls).
</p><p>
Normally, an additional ordered index is created when one
creates unique indexes on <code class="literal">NDB</code> table columns;
this can be used to search for <code class="literal">NULL</code> values.
However, if <code class="literal">USING HASH</code> is specified when such
an index is created, no ordered index is created.
</p><p>
The reason for permitting unique hash indexes with null values
is that, in some cases, the user wants to save space if a large
number of records are pre-allocated but not fully initialized.
This also assumes that the user will <span class="emphasis"><em>not</em></span>
try to search for null values. Since MySQL does not support
indexes that are not allowed to be searched in some cases, the
<code class="literal">NDB</code> storage engine uses a full table scan
with pushed conditions for the referenced index columns to
return the correct result.
</p><p>
A warning is returned if one creates a unique nullable hash
index, since the query optimizer should be provided a hint not
to use it with <code class="literal">NULL</code> values if this can be
avoided.
(<a href="http://bugs.mysql.com/21507" target="_top">Bug#21507</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <code class="literal">Ndb_number_of_storage_nodes</code> system
variable was renamed to
<code class="literal">Ndb_number_of_data_nodes</code>.
(<a href="http://bugs.mysql.com/20848" target="_top">Bug#20848</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <code class="literal">HELP</code> command in the Cluster management
client now provides command-specific help. For example,
<code class="literal">HELP RESTART</code> in <span><strong class="command">ndb_mgm</strong></span>
provides detailed information about the
<code class="literal">RESTART</code> command.
(<a href="http://bugs.mysql.com/19620" target="_top">Bug#19620</a>)</p></li><li><p>
<code class="literal">DROP TRIGGER</code> now supports an <code class="literal">IF
EXISTS</code> clause.
(<a href="http://bugs.mysql.com/23703" target="_top">Bug#23703</a>)</p></li><li><p>
The <code class="literal">Com_create_user</code> status variable was added
(for counting <code class="literal">CREATE USER</code> statements).
(<a href="http://bugs.mysql.com/22958" target="_top">Bug#22958</a>)</p></li><li><p>
The <code class="option">--memlock</code> option relies on system calls
that are unreliable on some operating systems. If a crash
occurs, the server now checks whether <code class="option">--memlock</code>
was specified and if so issues some information about possible
workarounds.
(<a href="http://bugs.mysql.com/22860" target="_top">Bug#22860</a>)</p></li><li><p>
If the user specified the server options
<code class="option">--max-connections=<em class="replaceable"><code>N</code></em> </code>
or <code class="option">--table-open-cache=<em class="replaceable"><code>M</code></em>
</code>, a warning would be given in some cases that some
values were recalculated, with the result that
<code class="option">--table-open-cache</code> could be assigned greater
value.
</p><p>
It should be noted that, in such cases, both the warning and the
increase in the <code class="option">--table-open-cache</code> value were
completely harmless. Note also that it is not possible for the
MySQL Server to predict or to control limitations on the maximum
number of open files, since this is determined by the operating
system.
</p><p>
The recalculation code has now been fixed to ensure that the
value of <code class="option">--table-open-cache</code> is no longer
increased automatically, and that a warning is now given only if
some values had to be decreased due to operating system limits.
(<a href="http://bugs.mysql.com/21915" target="_top">Bug#21915</a>)</p></li><li><p>
For the <code class="literal">CALL</code> statement, stored procedures
that take no arguments now can be invoked without parentheses.
That is, <code class="literal">CALL p()</code> and <code class="literal">CALL
p</code> are equivalent.
(<a href="http://bugs.mysql.com/21462" target="_top">Bug#21462</a>)</p></li><li><p>
<code class="literal">mysql_upgrade</code> now passes all the parameters
specified on the command line to both
<code class="literal">mysqlcheck</code> and <code class="literal">mysql</code> using
the <code class="filename">upgrade_defaults</code> file.
(<a href="http://bugs.mysql.com/20100" target="_top">Bug#20100</a>)</p></li><li><p>
<code class="literal">SHOW STATUS</code> is no longer logged to the slow
query log.
(<a href="http://bugs.mysql.com/19764" target="_top">Bug#19764</a>)</p></li><li><p>
<span><strong class="command">mysqldump --single-transaction</strong></span> now uses
<code class="literal">START TRANSACTION /*!40100 WITH CONSISTENT SNAPSHOT
*/</code> rather than <code class="literal">BEGIN</code> to start a
transaction, so that a consistent snapshot will be used on those
servers that support it.
(<a href="http://bugs.mysql.com/19660" target="_top">Bug#19660</a>)</p></li><li><p>
The bundled yaSSL library was upgraded to version 1.5.0.
</p></li></ul></div><p>Bugs fixed:</p><div class="itemizedlist"><ul type="disc"><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The failure of a data node failure during a schema operation
could lead to additional node failures.
(<a href="http://bugs.mysql.com/24752" target="_top">Bug#24752</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A committed read could be attempted before a data node had time
to connect, causing a timeout error.
(<a href="http://bugs.mysql.com/24717" target="_top">Bug#24717</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Sudden disconnection of an SQL or data node could lead to
shutdown of data nodes with the error <span class="errortext">failed
ndbrequire</span>.
(<a href="http://bugs.mysql.com/24447" target="_top">Bug#24447</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<span><strong class="command">ndb_config</strong></span> failed when trying to use 2
management servers and node IDs.
(<a href="http://bugs.mysql.com/23887" target="_top">Bug#23887</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Backup of a cluster failed if there were any tables with 128 or
more columns.
(<a href="http://bugs.mysql.com/23502" target="_top">Bug#23502</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Cluster backups failed when there were more than 2048 schema
objects in the cluster.
(<a href="http://bugs.mysql.com/23499" target="_top">Bug#23499</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The management client command <code class="literal">ALL DUMP 1000</code>
would cause the cluster to crash if data nodes were connected to
the cluster but not yret fully started.
(<a href="http://bugs.mysql.com/23203" target="_top">Bug#23203</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<code class="literal">INSERT ... ON DUPLICATE KEY UPDATE</code> on an
<code class="literal">NDB</code> table could lead to deadlocks and memory
leaks.
(<a href="http://bugs.mysql.com/23200" target="_top">Bug#23200</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
(NDB API): Inacivity timeouts for scans were not correctly
handled.
(<a href="http://bugs.mysql.com/23107" target="_top">Bug#23107</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
If a node restart could not be performed from the REDO log, no
node takeover took place. This could cause partitions to be left
empty during a system restart.
(<a href="http://bugs.mysql.com/22893" target="_top">Bug#22893</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Multiple node restarts in rapid succession could cause a system
restart to fail , or induce a race condition.
(<a href="http://bugs.mysql.com/22892" target="_top">Bug#22892</a>, <a href="http://bugs.mysql.com/23210" target="_top">Bug#23210</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
(NDB API): Attempting to read a nonexistent tuple using
<code class="literal">Commit</code> mode for
<code class="literal">NdbTransaction::execute()</code> caused node
failures.
(<a href="http://bugs.mysql.com/22672" target="_top">Bug#22672</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <code class="option">--help</code> output from <code class="literal">NDB</code>
binaries did not include file-related options.
(<a href="http://bugs.mysql.com/21994" target="_top">Bug#21994</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
(NDB API): Scans closed before being executed were still placed
in the send queue.
(<a href="http://bugs.mysql.com/21941" target="_top">Bug#21941</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A scan timeout returned Error 4028 (<span class="errortext">Node failure
caused abort of transaction</span>) instead of Error 4008
(<span class="errortext">Node failure caused abort of
transaction...</span>).
(<a href="http://bugs.mysql.com/21799" target="_top">Bug#21799</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The node recovery algorithm was missing a version check for
tables in the <code class="literal">ALTER_TABLE_COMMITTED</code> state (as
opposed to the <code class="literal">TABLE_ADD_COMMITTED</code> state,
which has the version check). This could cause inconsistent
schemas across nodes following node recovery.
(<a href="http://bugs.mysql.com/21756" target="_top">Bug#21756</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The output for the <code class="option">--help</code> option used with
<code class="literal">NDB</code> executable programs (such as
<span><strong class="command">ndbd</strong></span>, <span><strong class="command">ndb_mgm</strong></span>,
<span><strong class="command">ndb_restore</strong></span>, <span><strong class="command">ndb_config</strong></span>,
and others mentioned in
<a href="mysql-cluster.html#mysql-cluster-utilities" title="16.8. Cluster Utility Programs">Section 16.8, “Cluster Utility Programs”</a>) referred to the
<code class="filename">Ndb.cfg</code> file, instead of to
<code class="filename">my.cnf</code>.
(<a href="http://bugs.mysql.com/21585" target="_top">Bug#21585</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Partition distribution keys were updated only for the primary
and starting replicas during node recovery. This could lead to
node failure recovery for clusters having an odd number of
replicas.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
We recommend values for <code class="literal">NumberOfReplicas</code>
that are even powers of 2, for best results.
</p></div><p>(<a href="http://bugs.mysql.com/21535" target="_top">Bug#21535</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <span><strong class="command">ndb_mgm</strong></span> management client did not set the
exit status on errors, always returning 0 instead.
(<a href="http://bugs.mysql.com/21530" target="_top">Bug#21530</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Cluster logs were not rotated following the first rotation
cycle.
(<a href="http://bugs.mysql.com/21345" target="_top">Bug#21345</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
When inserting a row into an <code class="literal">NDB</code> table with a
duplicate value for a non-primary unique key, the error issued
would reference the wrong key.
(<a href="http://bugs.mysql.com/21072" target="_top">Bug#21072</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Condition pushdown did not work correctly with
<code class="literal">DATETIME</code> columns.
(<a href="http://bugs.mysql.com/21056" target="_top">Bug#21056</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Under some circumstances, local checkpointing would hang,
keeping any unstarted nodes from being started.
(<a href="http://bugs.mysql.com/20895" target="_top">Bug#20895</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Using an invalid node ID with the management client
<code class="literal">STOP</code> command could cause
<span><strong class="command">ndb_mgm</strong></span> to hang.
(<a href="http://bugs.mysql.com/20575" target="_top">Bug#20575</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Data nodes added while the cluster was running in single user
mode were all assigned node ID 0, which could later cause
multiple node failures. Adding nodes while in single user mode
is no longer possible.
(<a href="http://bugs.mysql.com/20395" target="_top">Bug#20395</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
In some cases where <code class="literal">SELECT COUNT(*)</code> from an
<code class="literal">NDB</code> table should have yielded an error,
<code class="literal">MAX_INT</code> was returned instead.
(<a href="http://bugs.mysql.com/19914" target="_top">Bug#19914</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Following the restart of a management node, the Cluster
management client did not automatically reconnect.
(<a href="http://bugs.mysql.com/19873" target="_top">Bug#19873</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
Error messages given when trying to make online changes
parameters such as <code class="literal">NoOfReplicas</code> thast can
only be changed via a complete shutdown and restart of the
cluster did not indicate the true nature of the problem.
(<a href="http://bugs.mysql.com/19787" target="_top">Bug#19787</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<span><strong class="command">ndb_restore</strong></span> did not always make clear that it
had recovered successfully from temporary errors while restoring
a cluster backup.
(<a href="http://bugs.mysql.com/19651" target="_top">Bug#19651</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
In rare situations with resource shortages, a crash could result
from insufficient <code class="literal">IndexScanOperations</code>.
(<a href="http://bugs.mysql.com/19198" target="_top">Bug#19198</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
<span><strong class="command">ndb_mgm -e show | head</strong></span> would hang after
displaying the first 10 lines of output.
(<a href="http://bugs.mysql.com/19047" target="_top">Bug#19047</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The error returned by the cluster when too many nodes were
defined did not make clear the nature of the problem.
(<a href="http://bugs.mysql.com/19045" target="_top">Bug#19045</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A unique constraint violation was not ignored by an
<code class="literal">UPDATE IGNORE</code> statement when the constraint
violation occurred on a non-primary key.
(<a href="http://bugs.mysql.com/18487" target="_top">Bug#18487</a>, <a href="http://bugs.mysql.com/24303" target="_top">Bug#24303</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
The <span><strong class="command">ndb_config</strong></span> utility did not perform host
lookups correctly when using the <code class="option">--host</code> option
(<a href="http://bugs.mysql.com/17582" target="_top">Bug#17582</a>)</p></li><li><p><span class="bold"><strong>MySQL Cluster</strong></span>:
A problem with takeover during a system restart caused ordered
indexes to be rebuilt incorrectly.
(<a href="http://bugs.mysql.com/15303" target="_top">Bug#15303</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
Using <code class="literal">BIT</code> values with any of the comparison
methods of the <code class="literal">NdbScanFilter</code> class caused
data nodes to fail.
(<a href="http://bugs.mysql.com/24503" target="_top">Bug#24503</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
Some MGM API function calls could yield incorrect return values
in certain cases where the cluster was operating under a very
high load, or experienced timeouts in inter-node communications.
(<a href="http://bugs.mysql.com/24011" target="_top">Bug#24011</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
The <code class="literal">NdbOperation::getBlobHandle()</code> method,
when called with the name of a nonexistent column, caused a
segmentation fault.
(<a href="http://bugs.mysql.com/21036" target="_top">Bug#21036</a>)</p></li><li><p><span class="bold"><strong>Cluster API</strong></span>:
When multiple processes or threads in parallel performed the
same ordered scan with exclusive lock and updated the retrieved
records, the scan could skip some records, which as a result
were not updated.
(<a href="http://bugs.mysql.com/20446" target="_top">Bug#20446</a>)</p></li><li><p>
The <a href="functions.html#function_repeat"><code class="literal">REPEAT()</code></a> function could
return <code class="literal">NULL</code> when passed a column for the
count argument.
(<a href="http://bugs.mysql.com/24947" target="_top">Bug#24947</a>)</p></li><li><p>
<span><strong class="command">mysql_upgrade</strong></span> failed if the
<code class="literal">--password</code> (or <code class="option">-p</code>) option
was given.
(<a href="http://bugs.mysql.com/24896" target="_top">Bug#24896</a>)</p></li><li><p>
With <code class="literal">innodb_file_per_table</code> enabled,
<code class="literal">InnoDB</code> displayed incorrect file times in the
output from <code class="literal">SHOW TABLE STATUS</code>.
(<a href="http://bugs.mysql.com/24712" target="_top">Bug#24712</a>)</p></li><li><p>
<code class="literal">ALTER ENABLE KEYS</code> or <code class="literal">ALTER TABLE
DISABLE KEYS</code> combined with another <code class="literal">ALTER
TABLE</code> option other than <code class="literal">RENAME TO</code>
did nothing. In addition, if ALTER TABLE was used on a table
having disabled keys, the keys of the resulting table were
enabled.
(<a href="http://bugs.mysql.com/24395" target="_top">Bug#24395</a>)</p></li><li><p>
The <code class="literal">InnoDB</code> mutex structure was simplified to
reduce memory load.
(<a href="http://bugs.mysql.com/24386" target="_top">Bug#24386</a>)</p></li><li><p>
The <code class="option">--extern</code> option for
<span><strong class="command">mysql-test-run.pl</strong></span> did not function correctly.
(<a href="http://bugs.mysql.com/24354" target="_top">Bug#24354</a>)</p></li><li><p>
Foreign key identifiers for <code class="literal">InnoDB</code> tables
could not contain certain characters.
(<a href="http://bugs.mysql.com/24299" target="_top">Bug#24299</a>)</p></li><li><p>
The <span><strong class="command">mysql.server</strong></span> script used the
<span><strong class="command">source</strong></span> command, which is less portable than
the <span><strong class="command">.</strong></span> command; it now uses
<span><strong class="command">.</strong></span> instead.
(<a href="http://bugs.mysql.com/24294" target="_top">Bug#24294</a>)</p></li><li><p>
<code class="literal">ALTER TABLE</code> statements that performed both
<code class="literal">RENAME TO</code> and <code class="literal">{ENABLE|DISABLE}
KEYS</code> operations caused a server crash.
(<a href="http://bugs.mysql.com/24219" target="_top">Bug#24219</a>)</p></li><li><p>
The loose index scan optimization for <code class="literal">GROUP
BY</code> with <code class="literal">MIN</code> or
<code class="literal">MAX</code> was not applied within other queries,
such as <code class="literal">CREATE TABLE ... SELECT ...</code>,
<code class="literal">INSERT ... SELECT ...</code>, or in the
<code class="literal">FROM</code> clauses of subqueries.
(<a href="http://bugs.mysql.com/24156" target="_top">Bug#24156</a>)</p></li><li><p>
There was a race condition in the <code class="literal">InnoDB</code>
<code class="literal">fil_flush_file_spaces()</code> function.
(<a href="http://bugs.mysql.com/24089" target="_top">Bug#24089</a>)</p><p>This regression was introduced by <a href="http://bugs.mysql.com/15653" target="_top">Bug#15653</a></p></li><li><p>
Subqueries for which a pushed-down condition did not produce
exactly one key field could cause a server crash.
(<a href="http://bugs.mysql.com/24056" target="_top">Bug#24056</a>)</p></li><li><p>
The size of <code class="literal">MEMORY</code> tables and internal
temporary tables was limited to 4GB on 64-bit Windows systems.
(<a href="http://bugs.mysql.com/24052" target="_top">Bug#24052</a>)</p></li><li><p>
Some yaSSL-related memory leaks detected by Valgrind were fixed.
(<a href="http://bugs.mysql.com/23981" target="_top">Bug#23981</a>)</p></li><li><p>
The internal SQL interpreter of <code class="literal">InnoDB</code> placed
an unnecessary lock on the supremum record when
<code class="literal">innodb_locks_unsafe_for_binlog=1</code>. This caused
an assertion failure when <code class="literal">InnoDB</code> was built
with debugging enabled.
(<a href="http://bugs.mysql.com/23769" target="_top">Bug#23769</a>)</p></li><li><p>
<a href="functions.html#function_row-count"><code class="literal">ROW_COUNT()</code></a> did not work
properly as an argument to a stored procedure.
(<a href="http://bugs.mysql.com/23760" target="_top">Bug#23760</a>)</p></li><li><p>
<a href="functions.html#function_last-day"><code class="literal">LAST_DAY('0000-00-00')</code></a> could
cause a server crash.
(<a href="http://bugs.mysql.com/23653" target="_top">Bug#23653</a>)</p></li><li><p>
A trigger that invoked a stored function could cause a server
crash when activated by different client connections.
(<a href="http://bugs.mysql.com/23651" target="_top">Bug#23651</a>)</p></li><li><p>
The stack size for NetWare binaries was increased to 128KB to
prevent problems caused by insufficient stack size.
(<a href="http://bugs.mysql.com/23504" target="_top">Bug#23504</a>)</p></li><li><p>
If elements in a non-top-level <code class="literal">IN</code> subquery
were accessed by an index and the subquery result set included a
<code class="literal">NULL</code> value, the quantified predicate that
contained the subquery was evaluated to <code class="literal">NULL</code>
when it should return a non-<code class="literal">NULL</code> value.
(<a href="http://bugs.mysql.com/23478" target="_top">Bug#23478</a>)</p></li><li><p>
When applying the <code class="literal">group_concat_max_len</code> limit,
<a href="functions.html#function_group-concat"><code class="literal">GROUP_CONCAT()</code></a> could truncate
multi-byte characters in the middle.
(<a href="http://bugs.mysql.com/23451" target="_top">Bug#23451</a>)</p></li><li><p>
MySQL 5.0.26 introduced an ABI incompatibility, which this
release reverts. Programs compiled against 5.0.26 are not
compatible with any other version and must be recompiled.
(<a href="http://bugs.mysql.com/23427" target="_top">Bug#23427</a>)</p></li><li><p>
<code class="literal"> <em class="replaceable"><code>M</code></em> % 0</code> returns
<code class="literal">NULL</code>, but (<code class="literal">
<em class="replaceable"><code>M</code></em> % 0) IS NULL</code> evaluated to
false.
(<a href="http://bugs.mysql.com/23411" target="_top">Bug#23411</a>)</p></li><li><p>
<a href="apis.html#mysql-affected-rows" title="23.2.3.1. mysql_affected_rows()"><code class="literal">mysql_affected_rows()</code></a> could
return values different from
<a href="apis.html#mysql-stmt-affected-rows" title="23.2.7.1. mysql_stmt_affected_rows()"><code class="literal">mysql_stmt_affected_rows()</code></a> for
the same sequence of statements.
(<a href="http://bugs.mysql.com/23383" target="_top">Bug#23383</a>)</p></li><li><p>
For not-yet-authenticated connections, the
<code class="literal">Time</code> column in <code class="literal">SHOW
PROCESSLIST</code> was a random value rather than
<code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/23379" target="_top">Bug#23379</a>)</p></li><li><p>
Accuracy was improved for comparisons between
<code class="literal">DECIMAL</code> columns and numbers represented as
strings.
(<a href="http://bugs.mysql.com/23260" target="_top">Bug#23260</a>)</p></li><li><p>
MySQL failed to build on Linux/Alpha.
(<a href="http://bugs.mysql.com/23256" target="_top">Bug#23256</a>)</p><p>This regression was introduced by <a href="http://bugs.mysql.com/21250" target="_top">Bug#21250</a></p></li><li><p>
If <a href="functions.html#function_compress"><code class="literal">COMPRESS()</code></a> returned
<code class="literal">NULL</code>, subsequent invocations of
<a href="functions.html#function_compress"><code class="literal">COMPRESS()</code></a> within a result set
or within a trigger also returned <code class="literal">NULL</code>.
(<a href="http://bugs.mysql.com/23254" target="_top">Bug#23254</a>)</p></li><li><p>
Calculation of <a href="functions.html#function_count"><code class="literal">COUNT(DISTINCT)</code></a>,
<a href="functions.html#function_avg"><code class="literal">AVG(DISTINCT)</code></a>, or
<a href="functions.html#function_sum"><code class="literal">SUM(DISTINCT)</code></a> when they are
referenced more than once in a single query with <code class="literal">GROUP
BY</code> could cause a server crash.
(<a href="http://bugs.mysql.com/23184" target="_top">Bug#23184</a>)</p></li><li><p>
Insufficient memory (<code class="literal">myisam_sort_buffer_size</code>)
could cause a server crash for several operations on
<code class="literal">MyISAM</code> tables: repair table, create index by
sort, repair by sort, parallel repair, bulk insert.
(<a href="http://bugs.mysql.com/23175" target="_top">Bug#23175</a>)</p></li><li><p>
The column default value in the output from <code class="literal">SHOW
COLUMNS</code> or <code class="literal">SELECT FROM
INFORMATION_SCHEMA.COLUMNS</code> was truncated to 64
characters.
(<a href="http://bugs.mysql.com/23037" target="_top">Bug#23037</a>)</p></li><li><p>
<span><strong class="command">mysql</strong></span> did not check for errors when fetching
data during result set printing.
(<a href="http://bugs.mysql.com/22913" target="_top">Bug#22913</a>)</p></li><li><p>
Changes to character set variables prior to an action on a
replication-ignored table were forgotten by slave servers.
(<a href="http://bugs.mysql.com/22877" target="_top">Bug#22877</a>)</p></li><li><p>
<code class="literal">InnoDB</code> exhibited thread thrashing with more
than 50 concurrent connections under an update-intensive
workload.
(<a href="http://bugs.mysql.com/22868" target="_top">Bug#22868</a>)</p></li><li><p>
The return value from <code class="literal">my_seek()</code> was ignored.
(<a href="http://bugs.mysql.com/22828" target="_top">Bug#22828</a>)</p></li><li><p>
The optimizer failed to use equality propagation for
<a href="functions.html#operator_between"><code class="literal">BETWEEN</code></a> and
<code class="literal">IN</code> predicates with string arguments.
(<a href="http://bugs.mysql.com/22753" target="_top">Bug#22753</a>)</p></li><li><p>
The <code class="literal">Handler_rollback</code> status variable
sometimes was incremented when no rollback had taken place.
(<a href="http://bugs.mysql.com/22728" target="_top">Bug#22728</a>)</p></li><li><p>
The <code class="literal">Host</code> column in <code class="literal">SHOW
PROCESSLIST</code> output was blank when the server was
started with the <code class="option">--skip-grant-tables</code> option.
(<a href="http://bugs.mysql.com/22723" target="_top">Bug#22723</a>)</p></li><li><p>
If a table contains an <code class="literal">AUTO_INCREMENT</code> column,
inserting into an insertable view on the table that does not
include the <code class="literal">AUTO_INCREMENT</code> column should not
change the value of
<a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a>, because the
side effects of inserting default values into columns not part
of the view should not be visible. MySQL was incorrectly setting
<a href="functions.html#function_last-insert-id"><code class="literal">LAST_INSERT_ID()</code></a> to zero.
(<a href="http://bugs.mysql.com/22584" target="_top">Bug#22584</a>)</p></li><li><p>
Queries using a column alias in an expression as part of an
<code class="literal">ORDER BY</code> clause failed, an example of such a
query being <code class="literal">SELECT mycol + 1 AS mynum FROM mytable
ORDER BY 30 - mynum</code>.
(<a href="http://bugs.mysql.com/22457" target="_top">Bug#22457</a>)</p></li><li><p>
Using <code class="literal">EXPLAIN</code> caused a server crash for
queries that selected from <code class="literal">INFORMATION_SCHEMA</code>
in a subquery in the <code class="literal">FROM</code> clause.
(<a href="http://bugs.mysql.com/22413" target="_top">Bug#22413</a>)</p></li><li><p>
Instance Manager had a race condition involving <span><strong class="command">
mysqld</strong></span> PID file removal.
(<a href="http://bugs.mysql.com/22379" target="_top">Bug#22379</a>)</p></li><li><p>
A server crash occurred when using <code class="literal">LOAD DATA</code>
to load a table containing a <code class="literal">NOT NULL</code> spatial
column, when the statement did not load the spatial column. Now
a <code class="literal">NULL supplied to NOT NULL column</code> error
occurs.
(<a href="http://bugs.mysql.com/22372" target="_top">Bug#22372</a>)</p></li><li><p>
The optimizer used the <code class="literal">ref</code> join type rather
than <code class="literal">eq_ref</code> for a simple join on strings.
(<a href="http://bugs.mysql.com/22367" target="_top">Bug#22367</a>)</p></li><li><p>
Some queries that used <a href="functions.html#function_max"><code class="literal">MAX()</code></a> and
<code class="literal">GROUP BY</code> could incorrectly return an empty
result.
(<a href="http://bugs.mysql.com/22342" target="_top">Bug#22342</a>)</p></li><li><p>
<a href="functions.html#function_date-add"><code class="literal">DATE_ADD()</code></a> requires complete
dates with no “<span class="quote">zero</span>” parts, but sometimes did not
return <code class="literal">NULL</code> when given such a date.
(<a href="http://bugs.mysql.com/22229" target="_top">Bug#22229</a>)</p></li><li><p>
If an <code class="literal">init_connect</code> SQL statement produced an
error, the connection was silently terminated with no error
message. Now the server writes a warning to the error log.
(<a href="http://bugs.mysql.com/22158" target="_top">Bug#22158</a>)</p></li><li><p>
Some small double precision numbers (such as
<code class="literal">1.00000001e-300</code>) that should have been
accepted were truncated to zero.
(<a href="http://bugs.mysql.com/22129" target="_top">Bug#22129</a>)</p></li><li><p>
For a nonexistent table, <code class="literal">DROP TEMPORARY TABLE</code>
failed with an incorrect error message if
<code class="literal">read_only</code> was enabled.
(<a href="http://bugs.mysql.com/22077" target="_top">Bug#22077</a>)</p></li><li><p>
Trailing spaces were not removed from Unicode
<code class="literal">CHAR</code> column values when used in indexes. This
resulted in excessive usage of storage space, and could affect
the results of some <code class="literal">ORDER BY</code> queries that
made use of such indexes.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
When upgrading, it is necessary to re-create any existing
indexes on Unicode <code class="literal">CHAR</code> columns in order to
take advantage of the fix. This can be done by using a
<code class="literal">REPAIR TABLE</code> statement on each affected
table.
</p></div><p>(<a href="http://bugs.mysql.com/22052" target="_top">Bug#22052</a>)</p></li><li><p>
The code for generating <code class="literal">USE</code> statements for
binary logging of <code class="literal">CREATE PROCEDURE</code> statements
resulted in confusing output from <span><strong class="command">mysqlbinlog</strong></span>
for <code class="literal">DROP PROCEDURE</code> statements.
(<a href="http://bugs.mysql.com/22043" target="_top">Bug#22043</a>)</p></li><li><p>
<a href="functions.html#function_str-to-date"><code class="literal">STR_TO_DATE()</code></a> returned
<code class="literal">NULL</code> if the format string contained a space
following a non-format character.
(<a href="http://bugs.mysql.com/22029" target="_top">Bug#22029</a>)</p></li><li><p>
Use of a DES-encrypted SSL certificate file caused a server
crash.
(<a href="http://bugs.mysql.com/21868" target="_top">Bug#21868</a>)</p></li><li><p>
Use of <code class="literal">PREPARE</code> with a <code class="literal">CREATE
PROCEDURE</code> statement that contained a syntax error
caused a server crash.
(<a href="http://bugs.mysql.com/21856" target="_top">Bug#21856</a>)</p></li><li><p>
Adding a day, month, or year interval to a
<code class="literal">DATE</code> value produced a
<code class="literal">DATE</code>, but adding a week interval produced a
<code class="literal">DATETIME</code> value. Now all produce a
<code class="literal">DATE</code> value.
(<a href="http://bugs.mysql.com/21811" target="_top">Bug#21811</a>)</p></li><li><p>
In some cases, the parser failed to distinguish a user-defined
function from a stored function.
(<a href="http://bugs.mysql.com/21809" target="_top">Bug#21809</a>)</p></li><li><p>
Use of a subquery that invoked a function in the column list of
the outer query resulted in a memory leak.
(<a href="http://bugs.mysql.com/21798" target="_top">Bug#21798</a>)</p></li><li><p>
Inserting a default or invalid value into a spatial column could
fail with <code class="literal">Unknown error</code> rather than a more
appropriate error.
(<a href="http://bugs.mysql.com/21790" target="_top">Bug#21790</a>)</p></li><li><p>
It was possible to use <code class="literal">DATETIME</code> values whose
year, month, and day parts were all zeroes but whose hour,
minute, and second parts contained nonzero values, an example of
such an illegal <code class="literal">DATETIME</code> being
<code class="literal">'0000-00-00 11:23:45'</code>.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
This fix was reverted in MySQL 5.0.40.
</p></div><p>(<a href="http://bugs.mysql.com/21789" target="_top">Bug#21789</a>)</p><p>See also <a href="http://bugs.mysql.com/25301" target="_top">Bug#25301</a></p></li><li><p>
yaSSL crashed on pre-Pentium Intel CPUs.
(<a href="http://bugs.mysql.com/21765" target="_top">Bug#21765</a>)</p></li><li><p>
Evaluation of subqueries that require the filesort algorithm
were allocating and freeing the
<code class="literal">sort_buffer_size</code> buffer many times, resulting
in slow performance. Now the buffer is allocated once and
reused.
(<a href="http://bugs.mysql.com/21727" target="_top">Bug#21727</a>)</p></li><li><p>
Through the C API, the member strings in
<code class="literal">MYSQL_FIELD</code> for a query that contains
expressions may return incorrect results.
(<a href="http://bugs.mysql.com/21635" target="_top">Bug#21635</a>)</p></li><li><p>
Selecting from a <code class="literal">MERGE</code> table could result in
a server crash if the underlying tables had fewer indexes than
the <code class="literal">MERGE</code> table itself.
(<a href="http://bugs.mysql.com/21617" target="_top">Bug#21617</a>, <a href="http://bugs.mysql.com/22937" target="_top">Bug#22937</a>)</p></li><li><p>
View columns were always handled as having implicit derivation,
leading to <code class="literal">illegal mix of collation errors</code>
for some views in <code class="literal">UNION</code> operations. Now view
column derivation comes from the original expression given in
the view definition.
(<a href="http://bugs.mysql.com/21505" target="_top">Bug#21505</a>)</p></li><li><p>
<code class="literal">InnoDB</code> crashed while performing XA recovery
of prepared transactions.
(<a href="http://bugs.mysql.com/21468" target="_top">Bug#21468</a>)</p></li><li><p>
<a href="functions.html#function_inet-aton"><code class="literal">INET_ATON()</code></a> returned a signed
<code class="literal">BIGINT</code> value, not an unsigned value.
(<a href="http://bugs.mysql.com/21466" target="_top">Bug#21466</a>)</p></li><li><p>
After <code class="literal">FLUSH TABLES WITH READ LOCK</code> followed by
<code class="literal">UNLOCK TABLES</code>, attempts to drop or alter a
stored routine failed with an error that the routine did not
exist, and attempts to execute the routine failed with a lock
conflict error.
(<a href="http://bugs.mysql.com/21414" target="_top">Bug#21414</a>)</p></li><li><p>
It was possible to set the backslash character (“<span class="quote">
<code class="literal">\</code> </span>”) as the delimiter character using
<code class="literal">DELIMITER</code>, but not actually possible to use
it as the delimiter.
(<a href="http://bugs.mysql.com/21412" target="_top">Bug#21412</a>)</p></li><li><p>
For multiple-table <code class="literal">UPDATE</code> statements, storage
engines were not notified of duplicate-key errors.
(<a href="http://bugs.mysql.com/21381" target="_top">Bug#21381</a>)</p></li><li><p>
Within a prepared statement, <code class="literal">SELECT (COUNT(*) =
1)</code> (or similar use of other aggregate functions) did
not return the correct result for statement re-execution.
(<a href="http://bugs.mysql.com/21354" target="_top">Bug#21354</a>)</p></li><li><p>
It was possible for a stored routine with a
non-<code class="literal">latin1</code> name to cause a stack overrun.
(<a href="http://bugs.mysql.com/21311" target="_top">Bug#21311</a>)</p></li><li><p>
Certain malformed <code class="literal">INSERT</code> statements could
crash the <span><strong class="command">mysql</strong></span> client.
(<a href="http://bugs.mysql.com/21142" target="_top">Bug#21142</a>)</p></li><li><p>
Creating a <code class="literal">TEMPORARY</code> table with the same name
as an existing table that was locked by another client could
result in a lock conflict for <code class="literal">DROP TEMPORARY
TABLE</code> because the server unnecessarily tried to
acquire a name lock.
(<a href="http://bugs.mysql.com/21096" target="_top">Bug#21096</a>)</p></li><li><p>
Incorrect results could be obtained from re-execution of a
parametrized prepared statement or a stored routine with a
<code class="literal">SELECT</code> that uses <code class="literal">LEFT JOIN</code>
with a second table having only one row.
(<a href="http://bugs.mysql.com/21081" target="_top">Bug#21081</a>)</p></li><li><p>
Within a stored routine, a view definition cannot refer to
routine parameters or local variables. However, an error did not
occur until the routine was called. Now it occurs during parsing
of the routine creation statement.
</p><div class="note" style="margin-left: 0.5in; margin-right: 0.5in;"><h3 class="title">Note</h3><p>
A side effect of this fix is that if you have already created
such routines, and error will occur if you execute
<code class="literal">SHOW CREATE PROCEDURE</code> or <code class="literal">SHOW
CREATE FUNCTION</code>. You should drop these routines
because they are erroneous.
</p></div><p>(<a href="http://bugs.mysql.com/20953" target="_top">Bug#20953</a>)</p></li><li><p>
In <span><strong class="command">mysql</strong></span>, invoking <code class="literal">connect</code>
or <code class="literal">\r</code> with very long
<em class="replaceable"><code>db_name</code></em> or
<em class="replaceable"><code>host_name</code></em> parameters caused buffer
overflow.
(<a href="http://bugs.mysql.com/20894" target="_top">Bug#20894</a>)</p></li><li><p>
<code class="literal">SHOW VARIABLES</code> truncated the
<code class="literal">Value</code> field to 256 characters.
(<a href="http://bugs.mysql.com/20862" target="_top">Bug#20862</a>)</p></li><li><p>
Selecting into variables sometimes returned incorrect wrong
results.
(<a href="http://bugs.mysql.com/20836" target="_top">Bug#20836</a>)</p></li><li><p>
<code class="literal">WITH ROLLUP</code> could group unequal values.
(<a href="http://bugs.mysql.com/20825" target="_top">Bug#20825</a>)</p></li><li><p>
Range searches on columns with an index prefix could miss
records.
(<a href="http://bugs.mysql.com/20732" target="_top">Bug#20732</a>)</p></li><li><p>
On slave servers, transactions that exceeded the lock wait
timeout failed to roll back properly.
(<a href="http://bugs.mysql.com/20697" target="_top">Bug#20697</a>)</p></li><li><p>
Inserting <code class="literal">DEFAULT</code> into a column with no
default value could result in garbage in the column. Now the
same result occurs as when inserting <code class="literal">NULL</code>
into a <code class="literal">NOT NULL</code> column.
(<a href="http://bugs.mysql.com/20691" target="_top">Bug#20691</a>)</p></li><li><p>
An <code class="literal">UPDATE</code> that referred to a key column in
the <code class="literal">WHERE</code> clause and activated a trigger that
modified the column resulted in a loop.
(<a href="http://bugs.mysql.com/20670" target="_top">Bug#20670</a>)</p></li><li><p>
<code class="literal">CONCURRENT</code> did not work correctly for
<code class="literal">LOAD DATA INFILE</code>.
(<a href="http://bugs.mysql.com/20637" target="_top">Bug#20637</a>)</p></li><li><p>
<code class="filename">mysql_fix_privilege_tables.sql</code> altered the
<code class="literal">table_privs.table_priv</code> column to contain too
few privileges, causing loss of the <code class="literal">CREATE
VIEW</code> and <code class="literal">SHOW VIEW</code> privileges.
(<a href="http://bugs.mysql.com/20589" target="_top">Bug#20589</a>)</p></li><li><p>
<code class="literal">LIKE</code> searches failed for indexed
<code class="literal">utf8</code> character columns.
(<a href="http://bugs.mysql.com/20471" target="_top">Bug#20471</a>)</p></li><li><p>
With <code class="literal">lower_case_table_names</code> set to 1,
<code class="literal">SHOW CREATE TABLE</code> printed incorrect output
for table names containing Turkish I (LATIN CAPITAL LETTER I
WITH DOT ABOVE).
(<a href="http://bugs.mysql.com/20404" target="_top">Bug#20404</a>)</p></li><li><p>
A query with a subquery that references columns of a view from
the outer <code class="literal">SELECT</code> could return an incorrect
result if used from a prepared statement.
(<a href="http://bugs.mysql.com/20327" target="_top">Bug#20327</a>)</p></li><li><p>
For queries that select from a view, the server was returning
<code class="literal">MYSQL_FIELD</code> metadata inconsistently for view
names and table names. For view columns, the server now returns
the view name in the <code class="literal">table</code> field and, if the
column selects from an underlying table, the table name in the
<code class="literal">org_table</code> field.
(<a href="http://bugs.mysql.com/20191" target="_top">Bug#20191</a>)</p></li><li><p>
Invalidating the query cache caused a server crash for
<code class="literal">INSERT INTO ... SELECT</code> statements that
selected from a view.
(<a href="http://bugs.mysql.com/20045" target="_top">Bug#20045</a>)</p></li><li><p>
With <code class="literal">SQL_MODE=TRADITIONAL</code>, MySQL incorrectly
aborted on warnings within stored routines and triggers.
(<a href="http://bugs.mysql.com/20028" target="_top">Bug#20028</a>)</p></li><li><p>
Unsigned <code class="literal">BIGINT</code> values treated as signed
values by the <a href="functions.html#function_mod"><code class="literal">MOD()</code></a> function.
(<a href="http://bugs.mysql.com/19955" target="_top">Bug#19955</a>)</p></li><li><p>
Compiling PHP 5.1 with the MySQL static libraries failed on some
versions of Linux.
(<a href="http://bugs.mysql.com/19817" target="_top">Bug#19817</a>)</p></li><li><p>
The <code class="literal">DELIMITER</code> statement did not work
correctly when used in an SQL file run using the
<code class="literal">SOURCE</code> statement.
(<a href="http://bugs.mysql.com/19799" target="_top">Bug#19799</a>)</p></li><li><p>
<span><strong class="command">mysqldump --xml</strong></span> produced invalid XML for
<code class="literal">BLOB</code> data.
(<a href="http://bugs.mysql.com/19745" target="_top">Bug#19745</a>)</p></li><li><p>
Column names were not quoted properly for replicated views.
(<a href="http://bugs.mysql.com/19736" target="_top">Bug#19736</a>)</p></li><li><p>
For a cast of a <code class="literal">DATETIME</code> value containing
microseconds to <code class="literal">DECIMAL</code>, the microseconds
part was truncated without generating a warning. Now the
microseconds part is preserved.
(<a href="http://bugs.mysql.com/19491" target="_top">Bug#19491</a>)</p></li><li><p>
<code class="literal">InnoDB</code>: Reduced optimization level for
Windows 64 builds to handle possible memory overrun.
(<a href="http://bugs.mysql.com/19424" target="_top">Bug#19424</a>)</p></li><li><p>
SQL statements close to the size of
<code class="literal">max_allowed_packet</code> could produce binary log
events larger than <code class="literal">max_allowed_packet</code> that
could not be read by slave servers.
(<a href="http://bugs.mysql.com/19402" target="_top">Bug#19402</a>)</p></li><li><p>
<code class="literal">VARBINARY</code> column values inserted on a MySQL
4.1 server had trailing zeroes following upgrade to MySQL 5.0 or
later.
(<a href="http://bugs.mysql.com/19371" target="_top">Bug#19371</a>)</p></li><li><p>
<code class="literal">FLUSH INSTANCES</code> in Instance Manager triggered
an assertion failure.
(<a href="http://bugs.mysql.com/19368" target="_top">Bug#19368</a>)</p></li><li><p>
For a debug server, a reference to an undefined user variable in
a prepared statment executed with <code class="literal">EXECUTE</code>
caused an assertion failure.
(<a href="http://bugs.mysql.com/19356" target="_top">Bug#19356</a>)</p></li><li><p>
The server could send incorrect column count information to the
client for queries that produce a larger number of columns than
can fit in a two-byte number.
(<a href="http://bugs.mysql.com/19216" target="_top">Bug#19216</a>)</p></li><li><p>
Within a trigger for a base table, selecting from a view on that
base table failed.
(<a href="http://bugs.mysql.com/19111" target="_top">Bug#19111</a>)</p></li><li><p>
The value of the <code class="literal">warning_count</code> system
variable was not being calculated correctly (also affecting
<code class="literal">SHOW COUNT(*) WARNINGS</code>).
(<a href="http://bugs.mysql.com/19024" target="_top">Bug#19024</a>)</p></li><li><p>
For some problems relating to character set conversion or
incorrect string values for <code class="literal">INSERT</code> or
<code class="literal">UPDATE</code>, the server was reporting truncation
or length errors instead.
(<a href="http://bugs.mysql.com/18908" target="_top">Bug#18908</a>)</p></li><li><p>
<code class="literal">DELETE IGNORE</code> could hang for foreign key
parent deletes.
(<a href="http://bugs.mysql.com/18819" target="_top">Bug#18819</a>)</p></li><li><p>
Constant expressions and some numeric constants used as input
parameters to user-defined functions were not treated as
constants.
(<a href="http://bugs.mysql.com/18761" target="_top">Bug#18761</a>)</p></li><li><p>
<code class="literal">InnoDB</code> used table locks (not row locks)
within stored functions.
(<a href="http://bugs.mysql.com/18077" target="_top">Bug#18077</a>)</p></li><li><p>
<span><strong class="command">myisampack</strong></span> wrote to unallocated memory,
causing a crash.
(<a href="http://bugs.mysql.com/17951" target="_top">Bug#17951</a>)</p></li><li><p>
<code class="literal">FLUSH LOGS</code> or <span><strong class="command">mysqladmin
flush-logs</strong></span> caused a server crash if the binary log was
not open.
(<a href="http://bugs.mysql.com/17733" target="_top">Bug#17733</a>)</p></li><li><p>
<span><strong class="command">mysql_fix_privilege_tables</strong></span> did not handle a
password containing embedded space or apostrophe characters.
(<a href="http://bugs.mysql.com/17700" target="_top">Bug#17700</a>)</p></li><li><p>
<span><strong class="command">mysql</strong></span> would lose its connection to the server
if its standard output was not writable.
(<a href="http://bugs.mysql.com/17583" target="_top">Bug#17583</a>)</p></li><li><p>
Attempting to use a view containing <code class="literal">DEFINER</code>
information for a non-existent user resulted in an error message
that revealed the definer account. Now the definer is revealed
only to superusers. Other users receive only an <code class="literal">access
denied</code> message.
(<a href="http://bugs.mysql.com/17254" target="_top">Bug#17254</a>)</p></li><li><p>
<span><strong class="command">mysql-test-run</strong></span> did not work correctly for
RPM-based installations.
(<a href="http://bugs.mysql.com/17194" target="_top">Bug#17194</a>)</p></li><li><p>
<a href="functions.html#function_in"><code class="literal">IN()</code></a> and
<a href="functions.html#function_char"><code class="literal">CHAR()</code></a> can return
<code class="literal">NULL</code>, but did not signal that to the query
processor, causing incorrect results for
<a href="functions.html#operator_is-null"><code class="literal">IS
NULL</code></a> operations.
(<a href="http://bugs.mysql.com/17047" target="_top">Bug#17047</a>)</p></li><li><p>
A client library crash was caused by executing a statement such
as <code class="literal">SELECT * FROM t1 PROCEDURE ANALYSE()</code> using
a server side cursor on a table <code class="literal">t1</code> that does
not have the same number of columns as the output from
<code class="literal">PROCEDURE ANALYSE()</code>.
(<a href="http://bugs.mysql.com/17039" target="_top">Bug#17039</a>)</p></li><li><p>
The <code class="literal">WITH CHECK OPTION</code> for a view failed to
prevent storing invalid column values for
<code class="literal">UPDATE</code> statements.
(<a href="http://bugs.mysql.com/16813" target="_top">Bug#16813</a>)</p></li><li><p>
Slave servers would retry the execution of a SQL statement an
infinite number of times, ignoring the value
<code class="literal">SLAVE_TRANSACTION_RETRIES</code> when using the NDB
engine.
(<a href="http://bugs.mysql.com/16228" target="_top">Bug#16228</a>)</p></li><li><p>
<code class="literal">InnoDB</code> showed substandard performance with
multiple queries running concurrently.
(<a href="http://bugs.mysql.com/15815" target="_top">Bug#15815</a>)</p></li><li><p>
<code class="literal">ALTER TABLE</code> was not able to rename a view.
(<a href="http://bugs.mysql.com/14959" target="_top">Bug#14959</a>)</p></li><li><p>
Statements such as <code class="literal">DROP PROCEDURE</code> and
<code class="literal">DROP VIEW</code> were written to the binary log too
late due to a race condition.
(<a href="http://bugs.mysql.com/14262" target="_top">Bug#14262</a>)</p></li><li><p>
A literal string in a <code class="literal">GROUP BY</code> clause could
be interpreted as a column name.
(<a href="http://bugs.mysql.com/14019" target="_top">Bug#14019</a>)</p></li><li><p>
Instance Manager didn't close the client socket file when
starting a new <span><strong class="command">mysqld</strong></span> instance.
<span><strong class="command">mysqld</strong></span> inherited the socket, causing clients
connected to Instance Manager to hang.
(<a href="http://bugs.mysql.com/12751" target="_top">Bug#12751</a>)</p></li><li><p>
Entries in the slow query log could have an incorrect
<code class="literal">Rows_examined</code> value.
(<a href="http://bugs.mysql.com/12240" target="_top">Bug#12240</a>)</p></li><li><p>
Warnings were generated when explicitly casting a character to a
number (for example, <a href="functions.html#function_cast"><code class="literal">CAST('x' AS
SIGNED)</code></a>), but not for implicit conversions in simple
arithmetic operations (such as <code class="literal">'x' + 0</code>). Now
warnings are generated in all cases.
(<a href="http://bugs.mysql.com/11927" target="_top">Bug#11927</a>)</p></li><li><p>
Lack of validation for input and output <code class="literal">TIME</code>
values resulted in several problems:
<a href="functions.html#function_sec-to-time"><code class="literal">SEC_TO_TIME()</code></a> within subqueries
incorrectly clipped large values;
<a href="functions.html#function_sec-to-time"><code class="literal">SEC_TO_TIME()</code></a> treated
<code class="literal">BIGINT UNSIGNED</code> values as signed; only
truncation warnings were produced when both truncation and
out-of-range <code class="literal">TIME</code> values occurred.
(<a href="http://bugs.mysql.com/11655" target="_top">Bug#11655</a>, <a href="http://bugs.mysql.com/20927" target="_top">Bug#20927</a>)</p></li><li><p>
Metadata for columns calculated from scalar subqueries was
limited to integer, double, or string, even if the actual type
of the column was different.
(<a href="http://bugs.mysql.com/11032" target="_top">Bug#11032</a>)</p></li><li><p>
Several string functions could return incorrect results when
given very large length arguments.
(<a href="http://bugs.mysql.com/10963" target="_top">Bug#10963</a>)</p></li><li><p>
<a href="functions.html#function_from-unixtime"><code class="literal">FROM_UNIXTIME()</code></a> did not accept
arguments up to <a href="functions.html#function_power"><code class="literal">POWER(2,31)-1</code></a>,
which it had previously.
(<a href="http://bugs.mysql.com/9191" target="_top">Bug#9191</a>)</p></li><li><p>
Subqueries of the form <code class="literal">NULL IN (SELECT ...)</code>
returned invalid results.
(<a href="http://bugs.mysql.com/8804" target="_top">Bug#8804</a>, <a href="http://bugs.mysql.com/23485" target="_top">Bug#23485</a>)</p></li><li><p>
<code class="literal">OPTIMIZE TABLE</code> with
<code class="literal">myisam_repair_threads</code> > 1 could result in
<code class="literal">MyISAM</code> table corruption.
(<a href="http://bugs.mysql.com/8283" target="_top">Bug#8283</a>)</p></li><li><p>
Transient errors in replication from master to slave may trigger
multiple <code class="literal">Got fatal error 1236: 'binlog truncated in the
middle of event'</code> errors on the slave.
(<a href="http://bugs.mysql.com/4053" target="_top">Bug#4053</a>)</p></li></ul></div></div><div class="section" lang="en"><div class="titlepage"><div><div><h3 class="title"><a name="releasenotes-cs-5-0-27"></a>D.1.6. Release Notes for MySQL Community Server 5.0.27 (21 October 2006)</h3></div></div></div><p class="release-level">
This is a bugfix release for the current production release
family. It replaces MySQL 5.0.26.
</p><p>Functionality added or changed:</p><div class="itemizedlist"><ul type="disc"><li><p>
This is the last version for which binary MySQL-Max
distributions are available, except for RPM distributions. (For
RPM distributions, the last version is 5.0.37.)
</p></li></ul></div><p>Bugs fixed:</p><div class="itemizedlist"><ul type="disc"><li><p>
MySQL 5.0.26 introduced an ABI incompatibility, which this
release reverts. Programs compiled against 5.0.26 are not
compatible with any other version and must be recompiled.
(<a href="http://bugs.mysql.com/23427" target="_top">Bug#23427</a>)</p></li></ul></div></div></div></div><div class="navfooter"><hr><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="releasenotes-es.html">Prev</a> </td><td width="20%" align="center"> </td><td width="40%" align="right"> <a accesskey="n" href="news.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">Appendix C. MySQL Enterprise Release Notes </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> Appendix E. MySQL Change History</td></tr></table></div></body></html>
distrib
>
Mandriva
>
2008.1
>
x86_64
>
media
>
main-testing
>
by-pkgid
>
b1e2421f2416edfc24c5845fbc1c5a2e
>
files
>
106
