<!-- doc/src/sgml/release-9.0.sgml -->
<!-- See header comment in release.sgml about typical markup -->
+ <sect1 id="release-9-0-19">
+ <title>Release 9.0.19</title>
+
+ <note>
+ <title>Release Date</title>
+ <simpara>2015-02-05</simpara>
+ </note>
+
+ <para>
+ This release contains a variety of fixes from 9.0.18.
+ For information about new features in the 9.0 major release, see
+ <xref linkend="release-9-0">.
+ </para>
+
+ <sect2>
+ <title>Migration to Version 9.0.19</title>
+
+ <para>
+ A dump/restore is not required for those running 9.0.X.
+ </para>
+
+ <para>
+ However, if you are upgrading from a version earlier than 9.0.18,
+ see <xref linkend="release-9-0-18">.
+ </para>
+
+ </sect2>
+
+ <sect2>
+ <title>Changes</title>
+
+ <itemizedlist>
+
+ <listitem>
+ <para>
+ Fix information leak via constraint-violation error messages
+ (Stephen Frost)
+ </para>
+
+ <para>
+ Some server error messages show the values of columns that violate
+ a constraint, such as a unique constraint. If the user does not have
+ <literal>SELECT</> privilege on all columns of the table, this could
+ mean exposing values that the user should not be able to see. Adjust
+ the code so that values are displayed only when they came from the SQL
+ command or could be selected by the user.
+ (CVE-2014-8161)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Lock down regression testing's temporary installations on Windows
+ (Noah Misch)
+ </para>
+
+ <para>
+ Use SSPI authentication to allow connections only from the OS user
+ who launched the test suite. This closes on Windows the same
+ vulnerability previously closed on other platforms, namely that other
+ users might be able to connect to the test postmaster.
+ (CVE-2014-0067)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid possible data corruption if <command>ALTER DATABASE SET
+ TABLESPACE</> is used to move a database to a new tablespace and then
+ shortly later move it back to its original tablespace (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid corrupting tables when <command>ANALYZE</> inside a transaction
+ is rolled back (Andres Freund, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ If the failing transaction had earlier removed the last index, rule, or
+ trigger from the table, the table would be left in a corrupted state
+ with the relevant <structname>pg_class</> flags not set though they
+ should be.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix use-of-already-freed-memory problem in EvalPlanQual processing
+ (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, queries that lock or update
+ recently-updated rows could crash as a result of this bug.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix planning of <command>SELECT FOR UPDATE</> when using a partial
+ index on a child table (Kyotaro Horiguchi)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, <command>SELECT FOR UPDATE</> must
+ also recheck the partial index's <literal>WHERE</> condition when
+ rechecking a recently-updated row to see if it still satisfies the
+ query's <literal>WHERE</> condition. This requirement was missed if the
+ index belonged to an inheritance child table, so that it was possible
+ to incorrectly return rows that no longer satisfy the query condition.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix corner case wherein <command>SELECT FOR UPDATE</> could return a row
+ twice, and possibly miss returning other rows (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, a <command>SELECT FOR UPDATE</>
+ that is scanning an inheritance tree could incorrectly return a row
+ from a prior child table instead of the one it should return from a
+ later child table.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reject duplicate column names in the referenced-columns list of
+ a <literal>FOREIGN KEY</> declaration (David Rowley)
+ </para>
+
+ <para>
+ This restriction is per SQL standard. Previously we did not reject
+ the case explicitly, but later on the code would fail with
+ bizarre-looking errors.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix bugs in raising a <type>numeric</> value to a large integral power
+ (Tom Lane)
+ </para>
+
+ <para>
+ The previous code could get a wrong answer, or consume excessive
+ amounts of time and memory before realizing that the answer must
+ overflow.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <function>numeric_recv()</>, truncate away any fractional digits
+ that would be hidden according to the value's <literal>dscale</> field
+ (Tom Lane)
+ </para>
+
+ <para>
+ A <type>numeric</> value's display scale (<literal>dscale</>) should
+ never be less than the number of nonzero fractional digits; but
+ apparently there's at least one broken client application that
+ transmits binary <type>numeric</> values in which that's true.
+ This leads to strange behavior since the extra digits are taken into
+ account by arithmetic operations even though they aren't printed.
+ The least risky fix seems to be to truncate away such <quote>hidden</>
+ digits on receipt, so that the value is indeed what it prints as.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reject out-of-range numeric timezone specifications (Tom Lane)
+ </para>
+
+ <para>
+ Simple numeric timezone specifications exceeding +/- 168 hours (one
+ week) would be accepted, but could then cause null-pointer dereference
+ crashes in certain operations. There's no use-case for such large UTC
+ offsets, so reject them.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix bugs in <type>tsquery</> <literal>@></> <type>tsquery</>
+ operator (Heikki Linnakangas)
+ </para>
+
+ <para>
+ Two different terms would be considered to match if they had the same
+ CRC. Also, if the second operand had more terms than the first, it
+ would be assumed not to be contained in the first; which is wrong
+ since it might contain duplicate terms.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve ispell dictionary's defenses against bad affix files (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Allow more than 64K phrases in a thesaurus dictionary (David Boutin)
+ </para>
+
+ <para>
+ The previous coding could crash on an oversize dictionary, so this was
+ deemed a back-patchable bug fix rather than a feature addition.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix namespace handling in <function>xpath()</> (Ali Akbar)
+ </para>
+
+ <para>
+ Previously, the <type>xml</> value resulting from
+ an <function>xpath()</> call would not have namespace declarations if
+ the namespace declarations were attached to an ancestor element in the
+ input <type>xml</> value, rather than to the specific element being
+ returned. Propagate the ancestral declaration so that the result is
+ correct when considered in isolation.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix planner problems with nested append relations, such as inherited
+ tables within <literal>UNION ALL</> subqueries (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fail cleanly when a GiST index tuple doesn't fit on a page, rather
+ than going into infinite recursion (Andrew Gierth)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Exempt tables that have per-table <varname>cost_limit</>
+ and/or <varname>cost_delay</> settings from autovacuum's global cost
+ balancing rules (Álvaro Herrera)
+ </para>
+
+ <para>
+ The previous behavior resulted in basically ignoring these per-table
+ settings, which was unintended. Now, a table having such settings
+ will be vacuumed using those settings, independently of what is going
+ on in other autovacuum workers. This may result in heavier total I/O
+ load than before, so such settings should be re-examined for sanity.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid wholesale autovacuuming when autovacuum is nominally off
+ (Tom Lane)
+ </para>
+
+ <para>
+ Even when autovacuum is nominally off, we will still launch autovacuum
+ worker processes to vacuum tables that are at risk of XID wraparound.
+ However, such a worker process then proceeded to vacuum all tables in
+ the target database, if they met the usual thresholds for
+ autovacuuming. This is at best pretty unexpected; at worst it delays
+ response to the wraparound threat. Fix it so that if autovacuum is
+ turned off, workers <emphasis>only</> do anti-wraparound vacuums and
+ not any other work.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix race condition between hot standby queries and replaying a
+ full-page image (Heikki Linnakangas)
+ </para>
+
+ <para>
+ This mistake could result in transient errors in queries being
+ executed in hot standby.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix several cases where recovery logic improperly ignored WAL records
+ for <literal>COMMIT/ABORT PREPARED</> (Heikki Linnakangas)
+ </para>
+
+ <para>
+ The most notable oversight was
+ that <varname>recovery_min_apply_delay</> failed to delay application
+ of a two-phase commit.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid creating unnecessary <filename>.ready</> marker files for
+ timeline history files (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix possible null pointer dereference when an empty prepared statement
+ is used and the <varname>log_statement</> setting is <literal>mod</>
+ or <literal>ddl</> (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Change <quote>pgstat wait timeout</> warning message to be LOG level,
+ and rephrase it to be more understandable (Tom Lane)
+ </para>
+
+ <para>
+ This message was originally thought to be essentially a can't-happen
+ case, but it occurs often enough on our slower buildfarm members to be
+ a nuisance. Reduce it to LOG level, and expend a bit more effort on
+ the wording: it now reads <quote>using stale statistics instead of
+ current ones because stats collector is not responding</>.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix SPARC spinlock implementation to ensure correctness if the CPU is
+ being run in a non-TSO coherency mode, as some non-Solaris kernels do
+ (Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Warn if OS X's <function>setlocale()</> starts an unwanted extra
+ thread inside the postmaster (Noah Misch)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix processing of repeated <literal>dbname</> parameters
+ in <function>PQconnectdbParams()</> (Alex Shulgin)
+ </para>
+
+ <para>
+ Unexpected behavior ensued if the first occurrence
+ of <literal>dbname</> contained a connection string or URI to be
+ expanded.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Ensure that <application>libpq</> reports a suitable error message on
+ unexpected socket EOF (Marko Tiikkaja, Tom Lane)
+ </para>
+
+ <para>
+ Depending on kernel behavior, <application>libpq</> might return an
+ empty error string rather than something useful when the server
+ unexpectedly closed the socket.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Clear any old error message during <function>PQreset()</>
+ (Heikki Linnakangas)
+ </para>
+
+ <para>
+ If <function>PQreset()</> is called repeatedly, and the connection
+ cannot be re-established, error messages from the failed connection
+ attempts kept accumulating in the <structname>PGconn</>'s error
+ string.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Properly handle out-of-memory conditions while parsing connection
+ options in <application>libpq</> (Alex Shulgin, Heikki Linnakangas)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix array overrun in <application>ecpg</>'s version
+ of <function>ParseDateTime()</> (Michael Paquier)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <application>initdb</>, give a clearer error message if a password
+ file is specified but is empty (Mats Erik Andersson)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s <command>\s</> command to work nicely with
+ libedit, and add pager support (Stepan Rutz, Tom Lane)
+ </para>
+
+ <para>
+ When using libedit rather than readline, <command>\s</> printed the
+ command history in a fairly unreadable encoded format, and on recent
+ libedit versions might fail altogether. Fix that by printing the
+ history ourselves rather than having the library do it. A pleasant
+ side-effect is that the pager is used if appropriate.
+ </para>
+
+ <para>
+ This patch also fixes a bug that caused newline encoding to be applied
+ inconsistently when saving the command history with libedit.
+ Multiline history entries written by older <application>psql</>
+ versions will be read cleanly with this patch, but perhaps not
+ vice versa, depending on the exact libedit versions involved.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve consistency of parsing of <application>psql</>'s special
+ variables (Tom Lane)
+ </para>
+
+ <para>
+ Allow variant spellings of <literal>on</> and <literal>off</> (such
+ as <literal>1</>/<literal>0</>) for <literal>ECHO_HIDDEN</>
+ and <literal>ON_ERROR_ROLLBACK</>. Report a warning for unrecognized
+ values for <literal>COMP_KEYWORD_CASE</>, <literal>ECHO</>,
+ <literal>ECHO_HIDDEN</>, <literal>HISTCONTROL</>,
+ <literal>ON_ERROR_ROLLBACK</>, and <literal>VERBOSITY</>. Recognize
+ all values for all these variables case-insensitively; previously
+ there was a mishmash of case-sensitive and case-insensitive behaviors.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s expanded-mode display to work
+ consistently when using <literal>border</> = 3
+ and <literal>linestyle</> = <literal>ascii</> or <literal>unicode</>
+ (Stephen Frost)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix possible deadlock during parallel restore of a schema-only dump
+ (Robert Haas, Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix core dump in <literal>pg_dump --binary-upgrade</> on zero-column
+ composite type (Rushabh Lathia)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix block number checking
+ in <filename>contrib/pageinspect</>'s <function>get_raw_page()</>
+ (Tom Lane)
+ </para>
+
+ <para>
+ The incorrect checking logic could prevent access to some pages in
+ non-main relation forks.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <filename>contrib/pgcrypto</>'s <function>pgp_sym_decrypt()</>
+ to not fail on messages whose length is 6 less than a power of 2
+ (Marko Tiikkaja)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Handle unexpected query results, especially NULLs, safely in
+ <filename>contrib/tablefunc</>'s <function>connectby()</>
+ (Michael Paquier)
+ </para>
+
+ <para>
+ <function>connectby()</> previously crashed if it encountered a NULL
+ key value. It now prints that row but doesn't recurse further.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid a possible crash in <filename>contrib/xml2</>'s
+ <function>xslt_process()</> (Mark Simonetti)
+ </para>
+
+ <para>
+ <application>libxslt</> seems to have an undocumented dependency on
+ the order in which resources are freed; reorder our calls to avoid a
+ crash.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Numerous cleanups of warnings from Coverity static code analyzer
+ (Andres Freund, Tatsuo Ishii, Marko Kreen, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ These changes are mostly cosmetic but in some cases fix corner-case
+ bugs, for example a crash rather than a proper error report after an
+ out-of-memory failure. None are believed to represent security
+ issues.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Detect incompatible OpenLDAP versions during build (Noah Misch)
+ </para>
+
+ <para>
+ With OpenLDAP versions 2.4.24 through 2.4.31,
+ inclusive, <productname>PostgreSQL</> backends can crash at exit.
+ Raise a warning during <application>configure</> based on the
+ compile-time OpenLDAP version number, and test the crashing scenario
+ in the <filename>contrib/dblink</> regression test.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In non-MSVC Windows builds, ensure <filename>libpq.dll</> is installed
+ with execute permissions (Noah Misch)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Make <application>pg_regress</> remove any temporary installation it
+ created upon successful exit (Tom Lane)
+ </para>
+
+ <para>
+ This results in a very substantial reduction in disk space usage
+ during <literal>make check-world</>, since that sequence involves
+ creation of numerous temporary installations.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Support time zone abbreviations that change UTC offset from time to
+ time (Tom Lane)
+ </para>
+
+ <para>
+ Previously, <productname>PostgreSQL</> assumed that the UTC offset
+ associated with a time zone abbreviation (such as <literal>EST</>)
+ never changes in the usage of any particular locale. However this
+ assumption fails in the real world, so introduce the ability for a
+ zone abbreviation to represent a UTC offset that sometimes changes.
+ Update the zone abbreviation definition files to make use of this
+ feature in timezone locales that have changed the UTC offset of their
+ abbreviations since 1970 (according to the IANA timezone database).
+ In such timezones, <productname>PostgreSQL</> will now associate the
+ correct UTC offset with the abbreviation depending on the given date.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Update time zone abbreviations lists (Tom Lane)
+ </para>
+
+ <para>
+ Add CST (China Standard Time) to our lists.
+ Remove references to ADT as <quote>Arabia Daylight Time</>, an
+ abbreviation that's been out of use since 2007; therefore, claiming
+ there is a conflict with <quote>Atlantic Daylight Time</> doesn't seem
+ especially helpful.
+ Fix entirely incorrect GMT offsets for CKT (Cook Islands), FJT, and FJST
+ (Fiji); we didn't even have them on the proper side of the date line.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Update time zone data files to <application>tzdata</> release 2015a.
+ </para>
+
+ <para>
+ The IANA timezone database has adopted abbreviations of the form
+ <literal>A<replaceable>x</>ST</literal>/<literal>A<replaceable>x</>DT</literal>
+ for all Australian time zones, reflecting what they believe to be
+ current majority practice Down Under. These names do not conflict
+ with usage elsewhere (other than ACST for Acre Summer Time, which has
+ been in disuse since 1994). Accordingly, adopt these names into
+ our <quote>Default</> timezone abbreviation set.
+ The <quote>Australia</> abbreviation set now contains only CST, EAST,
+ EST, SAST, SAT, and WST, all of which are thought to be mostly
+ historical usage. Note that SAST has also been changed to be South
+ Africa Standard Time in the <quote>Default</> abbreviation set.
+ </para>
+
+ <para>
+ Also, add zone abbreviations SRET (Asia/Srednekolymsk) and XJT
+ (Asia/Urumqi), and use WSST/WSDT for western Samoa. Also, there were
+ DST law changes in Chile, Mexico, the Turks & Caicos Islands
+ (America/Grand_Turk), and Fiji. There is a new zone
+ Pacific/Bougainville for portions of Papua New Guinea. Also, numerous
+ corrections for historical (pre-1970) time zone data.
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ </sect2>
+ </sect1>
+
<sect1 id="release-9-0-18">
<title>Release 9.0.18</title>
<!-- doc/src/sgml/release-9.1.sgml -->
<!-- See header comment in release.sgml about typical markup -->
+ <sect1 id="release-9-1-15">
+ <title>Release 9.1.15</title>
+
+ <note>
+ <title>Release Date</title>
+ <simpara>2015-02-05</simpara>
+ </note>
+
+ <para>
+ This release contains a variety of fixes from 9.1.14.
+ For information about new features in the 9.1 major release, see
+ <xref linkend="release-9-1">.
+ </para>
+
+ <sect2>
+ <title>Migration to Version 9.1.15</title>
+
+ <para>
+ A dump/restore is not required for those running 9.1.X.
+ </para>
+
+ <para>
+ However, if you are upgrading from a version earlier than 9.1.14,
+ see <xref linkend="release-9-1-14">.
+ </para>
+
+ </sect2>
+
+ <sect2>
+ <title>Changes</title>
+
+ <itemizedlist>
+
+ <listitem>
+ <para>
+ Fix information leak via constraint-violation error messages
+ (Stephen Frost)
+ </para>
+
+ <para>
+ Some server error messages show the values of columns that violate
+ a constraint, such as a unique constraint. If the user does not have
+ <literal>SELECT</> privilege on all columns of the table, this could
+ mean exposing values that the user should not be able to see. Adjust
+ the code so that values are displayed only when they came from the SQL
+ command or could be selected by the user.
+ (CVE-2014-8161)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Lock down regression testing's temporary installations on Windows
+ (Noah Misch)
+ </para>
+
+ <para>
+ Use SSPI authentication to allow connections only from the OS user
+ who launched the test suite. This closes on Windows the same
+ vulnerability previously closed on other platforms, namely that other
+ users might be able to connect to the test postmaster.
+ (CVE-2014-0067)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid possible data corruption if <command>ALTER DATABASE SET
+ TABLESPACE</> is used to move a database to a new tablespace and then
+ shortly later move it back to its original tablespace (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid corrupting tables when <command>ANALYZE</> inside a transaction
+ is rolled back (Andres Freund, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ If the failing transaction had earlier removed the last index, rule, or
+ trigger from the table, the table would be left in a corrupted state
+ with the relevant <structname>pg_class</> flags not set though they
+ should be.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Ensure that unlogged tables are copied correctly
+ during <command>CREATE DATABASE</> or <command>ALTER DATABASE SET
+ TABLESPACE</> (Pavan Deolasee, Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <command>DROP</>'s dependency searching to correctly handle the
+ case where a table column is recursively visited before its table
+ (Petr Jelinek, Tom Lane)
+ </para>
+
+ <para>
+ This case is only known to arise when an extension creates both a
+ datatype and a table using that datatype. The faulty code might
+ refuse a <command>DROP EXTENSION</> unless <literal>CASCADE</> is
+ specified, which should not be required.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix use-of-already-freed-memory problem in EvalPlanQual processing
+ (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, queries that lock or update
+ recently-updated rows could crash as a result of this bug.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix planning of <command>SELECT FOR UPDATE</> when using a partial
+ index on a child table (Kyotaro Horiguchi)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, <command>SELECT FOR UPDATE</> must
+ also recheck the partial index's <literal>WHERE</> condition when
+ rechecking a recently-updated row to see if it still satisfies the
+ query's <literal>WHERE</> condition. This requirement was missed if the
+ index belonged to an inheritance child table, so that it was possible
+ to incorrectly return rows that no longer satisfy the query condition.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix corner case wherein <command>SELECT FOR UPDATE</> could return a row
+ twice, and possibly miss returning other rows (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, a <command>SELECT FOR UPDATE</>
+ that is scanning an inheritance tree could incorrectly return a row
+ from a prior child table instead of the one it should return from a
+ later child table.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reject duplicate column names in the referenced-columns list of
+ a <literal>FOREIGN KEY</> declaration (David Rowley)
+ </para>
+
+ <para>
+ This restriction is per SQL standard. Previously we did not reject
+ the case explicitly, but later on the code would fail with
+ bizarre-looking errors.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix bugs in raising a <type>numeric</> value to a large integral power
+ (Tom Lane)
+ </para>
+
+ <para>
+ The previous code could get a wrong answer, or consume excessive
+ amounts of time and memory before realizing that the answer must
+ overflow.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <function>numeric_recv()</>, truncate away any fractional digits
+ that would be hidden according to the value's <literal>dscale</> field
+ (Tom Lane)
+ </para>
+
+ <para>
+ A <type>numeric</> value's display scale (<literal>dscale</>) should
+ never be less than the number of nonzero fractional digits; but
+ apparently there's at least one broken client application that
+ transmits binary <type>numeric</> values in which that's true.
+ This leads to strange behavior since the extra digits are taken into
+ account by arithmetic operations even though they aren't printed.
+ The least risky fix seems to be to truncate away such <quote>hidden</>
+ digits on receipt, so that the value is indeed what it prints as.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reject out-of-range numeric timezone specifications (Tom Lane)
+ </para>
+
+ <para>
+ Simple numeric timezone specifications exceeding +/- 168 hours (one
+ week) would be accepted, but could then cause null-pointer dereference
+ crashes in certain operations. There's no use-case for such large UTC
+ offsets, so reject them.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix bugs in <type>tsquery</> <literal>@></> <type>tsquery</>
+ operator (Heikki Linnakangas)
+ </para>
+
+ <para>
+ Two different terms would be considered to match if they had the same
+ CRC. Also, if the second operand had more terms than the first, it
+ would be assumed not to be contained in the first; which is wrong
+ since it might contain duplicate terms.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve ispell dictionary's defenses against bad affix files (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Allow more than 64K phrases in a thesaurus dictionary (David Boutin)
+ </para>
+
+ <para>
+ The previous coding could crash on an oversize dictionary, so this was
+ deemed a back-patchable bug fix rather than a feature addition.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix namespace handling in <function>xpath()</> (Ali Akbar)
+ </para>
+
+ <para>
+ Previously, the <type>xml</> value resulting from
+ an <function>xpath()</> call would not have namespace declarations if
+ the namespace declarations were attached to an ancestor element in the
+ input <type>xml</> value, rather than to the specific element being
+ returned. Propagate the ancestral declaration so that the result is
+ correct when considered in isolation.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix planner problems with nested append relations, such as inherited
+ tables within <literal>UNION ALL</> subqueries (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fail cleanly when a GiST index tuple doesn't fit on a page, rather
+ than going into infinite recursion (Andrew Gierth)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Exempt tables that have per-table <varname>cost_limit</>
+ and/or <varname>cost_delay</> settings from autovacuum's global cost
+ balancing rules (Álvaro Herrera)
+ </para>
+
+ <para>
+ The previous behavior resulted in basically ignoring these per-table
+ settings, which was unintended. Now, a table having such settings
+ will be vacuumed using those settings, independently of what is going
+ on in other autovacuum workers. This may result in heavier total I/O
+ load than before, so such settings should be re-examined for sanity.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid wholesale autovacuuming when autovacuum is nominally off
+ (Tom Lane)
+ </para>
+
+ <para>
+ Even when autovacuum is nominally off, we will still launch autovacuum
+ worker processes to vacuum tables that are at risk of XID wraparound.
+ However, such a worker process then proceeded to vacuum all tables in
+ the target database, if they met the usual thresholds for
+ autovacuuming. This is at best pretty unexpected; at worst it delays
+ response to the wraparound threat. Fix it so that if autovacuum is
+ turned off, workers <emphasis>only</> do anti-wraparound vacuums and
+ not any other work.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ During crash recovery, ensure that unlogged relations are rewritten as
+ empty and are synced to disk before recovery is considered complete
+ (Abhijit Menon-Sen, Andres Freund)
+ </para>
+
+ <para>
+ This prevents scenarios in which unlogged relations might contain
+ garbage data following database crash recovery.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix race condition between hot standby queries and replaying a
+ full-page image (Heikki Linnakangas)
+ </para>
+
+ <para>
+ This mistake could result in transient errors in queries being
+ executed in hot standby.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix several cases where recovery logic improperly ignored WAL records
+ for <literal>COMMIT/ABORT PREPARED</> (Heikki Linnakangas)
+ </para>
+
+ <para>
+ The most notable oversight was
+ that <varname>recovery_min_apply_delay</> failed to delay application
+ of a two-phase commit.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid creating unnecessary <filename>.ready</> marker files for
+ timeline history files (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix possible null pointer dereference when an empty prepared statement
+ is used and the <varname>log_statement</> setting is <literal>mod</>
+ or <literal>ddl</> (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Change <quote>pgstat wait timeout</> warning message to be LOG level,
+ and rephrase it to be more understandable (Tom Lane)
+ </para>
+
+ <para>
+ This message was originally thought to be essentially a can't-happen
+ case, but it occurs often enough on our slower buildfarm members to be
+ a nuisance. Reduce it to LOG level, and expend a bit more effort on
+ the wording: it now reads <quote>using stale statistics instead of
+ current ones because stats collector is not responding</>.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix SPARC spinlock implementation to ensure correctness if the CPU is
+ being run in a non-TSO coherency mode, as some non-Solaris kernels do
+ (Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Warn if OS X's <function>setlocale()</> starts an unwanted extra
+ thread inside the postmaster (Noah Misch)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix processing of repeated <literal>dbname</> parameters
+ in <function>PQconnectdbParams()</> (Alex Shulgin)
+ </para>
+
+ <para>
+ Unexpected behavior ensued if the first occurrence
+ of <literal>dbname</> contained a connection string or URI to be
+ expanded.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Ensure that <application>libpq</> reports a suitable error message on
+ unexpected socket EOF (Marko Tiikkaja, Tom Lane)
+ </para>
+
+ <para>
+ Depending on kernel behavior, <application>libpq</> might return an
+ empty error string rather than something useful when the server
+ unexpectedly closed the socket.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Clear any old error message during <function>PQreset()</>
+ (Heikki Linnakangas)
+ </para>
+
+ <para>
+ If <function>PQreset()</> is called repeatedly, and the connection
+ cannot be re-established, error messages from the failed connection
+ attempts kept accumulating in the <structname>PGconn</>'s error
+ string.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Properly handle out-of-memory conditions while parsing connection
+ options in <application>libpq</> (Alex Shulgin, Heikki Linnakangas)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix array overrun in <application>ecpg</>'s version
+ of <function>ParseDateTime()</> (Michael Paquier)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <application>initdb</>, give a clearer error message if a password
+ file is specified but is empty (Mats Erik Andersson)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s <command>\s</> command to work nicely with
+ libedit, and add pager support (Stepan Rutz, Tom Lane)
+ </para>
+
+ <para>
+ When using libedit rather than readline, <command>\s</> printed the
+ command history in a fairly unreadable encoded format, and on recent
+ libedit versions might fail altogether. Fix that by printing the
+ history ourselves rather than having the library do it. A pleasant
+ side-effect is that the pager is used if appropriate.
+ </para>
+
+ <para>
+ This patch also fixes a bug that caused newline encoding to be applied
+ inconsistently when saving the command history with libedit.
+ Multiline history entries written by older <application>psql</>
+ versions will be read cleanly with this patch, but perhaps not
+ vice versa, depending on the exact libedit versions involved.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve consistency of parsing of <application>psql</>'s special
+ variables (Tom Lane)
+ </para>
+
+ <para>
+ Allow variant spellings of <literal>on</> and <literal>off</> (such
+ as <literal>1</>/<literal>0</>) for <literal>ECHO_HIDDEN</>
+ and <literal>ON_ERROR_ROLLBACK</>. Report a warning for unrecognized
+ values for <literal>COMP_KEYWORD_CASE</>, <literal>ECHO</>,
+ <literal>ECHO_HIDDEN</>, <literal>HISTCONTROL</>,
+ <literal>ON_ERROR_ROLLBACK</>, and <literal>VERBOSITY</>. Recognize
+ all values for all these variables case-insensitively; previously
+ there was a mishmash of case-sensitive and case-insensitive behaviors.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s expanded-mode display to work
+ consistently when using <literal>border</> = 3
+ and <literal>linestyle</> = <literal>ascii</> or <literal>unicode</>
+ (Stephen Frost)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve performance of <application>pg_dump</> when the database
+ contains many instances of multiple dependency paths between the same
+ two objects (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix possible deadlock during parallel restore of a schema-only dump
+ (Robert Haas, Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix core dump in <literal>pg_dump --binary-upgrade</> on zero-column
+ composite type (Rushabh Lathia)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Prevent WAL files created by <literal>pg_basebackup -x/-X</> from
+ being archived again when the standby is promoted (Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix upgrade-from-unpackaged script for <filename>contrib/citext</>
+ (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix block number checking
+ in <filename>contrib/pageinspect</>'s <function>get_raw_page()</>
+ (Tom Lane)
+ </para>
+
+ <para>
+ The incorrect checking logic could prevent access to some pages in
+ non-main relation forks.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <filename>contrib/pgcrypto</>'s <function>pgp_sym_decrypt()</>
+ to not fail on messages whose length is 6 less than a power of 2
+ (Marko Tiikkaja)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix file descriptor leak in <filename>contrib/pg_test_fsync</>
+ (Jeff Janes)
+ </para>
+
+ <para>
+ This could cause failure to remove temporary files on Windows.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Handle unexpected query results, especially NULLs, safely in
+ <filename>contrib/tablefunc</>'s <function>connectby()</>
+ (Michael Paquier)
+ </para>
+
+ <para>
+ <function>connectby()</> previously crashed if it encountered a NULL
+ key value. It now prints that row but doesn't recurse further.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid a possible crash in <filename>contrib/xml2</>'s
+ <function>xslt_process()</> (Mark Simonetti)
+ </para>
+
+ <para>
+ <application>libxslt</> seems to have an undocumented dependency on
+ the order in which resources are freed; reorder our calls to avoid a
+ crash.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Mark some <filename>contrib</> I/O functions with correct volatility
+ properties (Tom Lane)
+ </para>
+
+ <para>
+ The previous over-conservative marking was immaterial in normal use,
+ but could cause optimization problems or rejection of valid index
+ expression definitions. Since the consequences are not large, we've
+ just adjusted the function definitions in the extension modules'
+ scripts, without changing version numbers.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Numerous cleanups of warnings from Coverity static code analyzer
+ (Andres Freund, Tatsuo Ishii, Marko Kreen, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ These changes are mostly cosmetic but in some cases fix corner-case
+ bugs, for example a crash rather than a proper error report after an
+ out-of-memory failure. None are believed to represent security
+ issues.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Detect incompatible OpenLDAP versions during build (Noah Misch)
+ </para>
+
+ <para>
+ With OpenLDAP versions 2.4.24 through 2.4.31,
+ inclusive, <productname>PostgreSQL</> backends can crash at exit.
+ Raise a warning during <application>configure</> based on the
+ compile-time OpenLDAP version number, and test the crashing scenario
+ in the <filename>contrib/dblink</> regression test.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In non-MSVC Windows builds, ensure <filename>libpq.dll</> is installed
+ with execute permissions (Noah Misch)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Make <application>pg_regress</> remove any temporary installation it
+ created upon successful exit (Tom Lane)
+ </para>
+
+ <para>
+ This results in a very substantial reduction in disk space usage
+ during <literal>make check-world</>, since that sequence involves
+ creation of numerous temporary installations.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Support time zone abbreviations that change UTC offset from time to
+ time (Tom Lane)
+ </para>
+
+ <para>
+ Previously, <productname>PostgreSQL</> assumed that the UTC offset
+ associated with a time zone abbreviation (such as <literal>EST</>)
+ never changes in the usage of any particular locale. However this
+ assumption fails in the real world, so introduce the ability for a
+ zone abbreviation to represent a UTC offset that sometimes changes.
+ Update the zone abbreviation definition files to make use of this
+ feature in timezone locales that have changed the UTC offset of their
+ abbreviations since 1970 (according to the IANA timezone database).
+ In such timezones, <productname>PostgreSQL</> will now associate the
+ correct UTC offset with the abbreviation depending on the given date.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Update time zone abbreviations lists (Tom Lane)
+ </para>
+
+ <para>
+ Add CST (China Standard Time) to our lists.
+ Remove references to ADT as <quote>Arabia Daylight Time</>, an
+ abbreviation that's been out of use since 2007; therefore, claiming
+ there is a conflict with <quote>Atlantic Daylight Time</> doesn't seem
+ especially helpful.
+ Fix entirely incorrect GMT offsets for CKT (Cook Islands), FJT, and FJST
+ (Fiji); we didn't even have them on the proper side of the date line.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Update time zone data files to <application>tzdata</> release 2015a.
+ </para>
+
+ <para>
+ The IANA timezone database has adopted abbreviations of the form
+ <literal>A<replaceable>x</>ST</literal>/<literal>A<replaceable>x</>DT</literal>
+ for all Australian time zones, reflecting what they believe to be
+ current majority practice Down Under. These names do not conflict
+ with usage elsewhere (other than ACST for Acre Summer Time, which has
+ been in disuse since 1994). Accordingly, adopt these names into
+ our <quote>Default</> timezone abbreviation set.
+ The <quote>Australia</> abbreviation set now contains only CST, EAST,
+ EST, SAST, SAT, and WST, all of which are thought to be mostly
+ historical usage. Note that SAST has also been changed to be South
+ Africa Standard Time in the <quote>Default</> abbreviation set.
+ </para>
+
+ <para>
+ Also, add zone abbreviations SRET (Asia/Srednekolymsk) and XJT
+ (Asia/Urumqi), and use WSST/WSDT for western Samoa. Also, there were
+ DST law changes in Chile, Mexico, the Turks & Caicos Islands
+ (America/Grand_Turk), and Fiji. There is a new zone
+ Pacific/Bougainville for portions of Papua New Guinea. Also, numerous
+ corrections for historical (pre-1970) time zone data.
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ </sect2>
+ </sect1>
+
<sect1 id="release-9-1-14">
<title>Release 9.1.14</title>
<!-- doc/src/sgml/release-9.2.sgml -->
<!-- See header comment in release.sgml about typical markup -->
+ <sect1 id="release-9-2-10">
+ <title>Release 9.2.10</title>
+
+ <note>
+ <title>Release Date</title>
+ <simpara>2015-02-05</simpara>
+ </note>
+
+ <para>
+ This release contains a variety of fixes from 9.2.9.
+ For information about new features in the 9.2 major release, see
+ <xref linkend="release-9-2">.
+ </para>
+
+ <sect2>
+ <title>Migration to Version 9.2.10</title>
+
+ <para>
+ A dump/restore is not required for those running 9.2.X.
+ </para>
+
+ <para>
+ However, if you are a Windows user and are using the <quote>Norwegian
+ (Bokmål)</> locale, manual action is needed after the upgrade to
+ replace any <quote>Norwegian (Bokmål)_Norway</> locale names stored
+ in <productname>PostgreSQL</> system catalogs with the plain-ASCII
+ alias <quote>Norwegian_Norway</>. For details see
+ <ulink url="http://wiki.postgresql.org/wiki/Changes_To_Norwegian_Locale"></>
+ </para>
+
+ <para>
+ Also, if you are upgrading from a version earlier than 9.2.9,
+ see <xref linkend="release-9-2-9">.
+ </para>
+
+ </sect2>
+
+ <sect2>
+ <title>Changes</title>
+
+ <itemizedlist>
+
+ <listitem>
+ <para>
+ Fix information leak via constraint-violation error messages
+ (Stephen Frost)
+ </para>
+
+ <para>
+ Some server error messages show the values of columns that violate
+ a constraint, such as a unique constraint. If the user does not have
+ <literal>SELECT</> privilege on all columns of the table, this could
+ mean exposing values that the user should not be able to see. Adjust
+ the code so that values are displayed only when they came from the SQL
+ command or could be selected by the user.
+ (CVE-2014-8161)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Lock down regression testing's temporary installations on Windows
+ (Noah Misch)
+ </para>
+
+ <para>
+ Use SSPI authentication to allow connections only from the OS user
+ who launched the test suite. This closes on Windows the same
+ vulnerability previously closed on other platforms, namely that other
+ users might be able to connect to the test postmaster.
+ (CVE-2014-0067)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Cope with the Windows locale named <quote>Norwegian (Bokmål)</>
+ (Heikki Linnakangas)
+ </para>
+
+ <para>
+ Non-ASCII locale names are problematic since it's not clear what
+ encoding they should be represented in. Map the troublesome locale
+ name to a plain-ASCII alias, <quote>Norwegian_Norway</>.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid possible data corruption if <command>ALTER DATABASE SET
+ TABLESPACE</> is used to move a database to a new tablespace and then
+ shortly later move it back to its original tablespace (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid corrupting tables when <command>ANALYZE</> inside a transaction
+ is rolled back (Andres Freund, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ If the failing transaction had earlier removed the last index, rule, or
+ trigger from the table, the table would be left in a corrupted state
+ with the relevant <structname>pg_class</> flags not set though they
+ should be.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Ensure that unlogged tables are copied correctly
+ during <command>CREATE DATABASE</> or <command>ALTER DATABASE SET
+ TABLESPACE</> (Pavan Deolasee, Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <command>DROP</>'s dependency searching to correctly handle the
+ case where a table column is recursively visited before its table
+ (Petr Jelinek, Tom Lane)
+ </para>
+
+ <para>
+ This case is only known to arise when an extension creates both a
+ datatype and a table using that datatype. The faulty code might
+ refuse a <command>DROP EXTENSION</> unless <literal>CASCADE</> is
+ specified, which should not be required.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix use-of-already-freed-memory problem in EvalPlanQual processing
+ (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, queries that lock or update
+ recently-updated rows could crash as a result of this bug.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix planning of <command>SELECT FOR UPDATE</> when using a partial
+ index on a child table (Kyotaro Horiguchi)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, <command>SELECT FOR UPDATE</> must
+ also recheck the partial index's <literal>WHERE</> condition when
+ rechecking a recently-updated row to see if it still satisfies the
+ query's <literal>WHERE</> condition. This requirement was missed if the
+ index belonged to an inheritance child table, so that it was possible
+ to incorrectly return rows that no longer satisfy the query condition.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix corner case wherein <command>SELECT FOR UPDATE</> could return a row
+ twice, and possibly miss returning other rows (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, a <command>SELECT FOR UPDATE</>
+ that is scanning an inheritance tree could incorrectly return a row
+ from a prior child table instead of the one it should return from a
+ later child table.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reject duplicate column names in the referenced-columns list of
+ a <literal>FOREIGN KEY</> declaration (David Rowley)
+ </para>
+
+ <para>
+ This restriction is per SQL standard. Previously we did not reject
+ the case explicitly, but later on the code would fail with
+ bizarre-looking errors.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Restore previous behavior of conversion of domains to JSON
+ (Tom Lane)
+ </para>
+
+ <para>
+ This change causes domains over numeric and boolean to be treated
+ like their base types for purposes of conversion to JSON. It worked
+ like that before 9.3.5 and 9.2.9, but was unintentionally changed
+ while fixing a related problem.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix bugs in raising a <type>numeric</> value to a large integral power
+ (Tom Lane)
+ </para>
+
+ <para>
+ The previous code could get a wrong answer, or consume excessive
+ amounts of time and memory before realizing that the answer must
+ overflow.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <function>numeric_recv()</>, truncate away any fractional digits
+ that would be hidden according to the value's <literal>dscale</> field
+ (Tom Lane)
+ </para>
+
+ <para>
+ A <type>numeric</> value's display scale (<literal>dscale</>) should
+ never be less than the number of nonzero fractional digits; but
+ apparently there's at least one broken client application that
+ transmits binary <type>numeric</> values in which that's true.
+ This leads to strange behavior since the extra digits are taken into
+ account by arithmetic operations even though they aren't printed.
+ The least risky fix seems to be to truncate away such <quote>hidden</>
+ digits on receipt, so that the value is indeed what it prints as.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix incorrect search for shortest-first regular expression matches
+ (Tom Lane)
+ </para>
+
+ <para>
+ Matching would often fail when the number of allowed iterations is
+ limited by a <literal>?</> quantifier or a bound expression.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reject out-of-range numeric timezone specifications (Tom Lane)
+ </para>
+
+ <para>
+ Simple numeric timezone specifications exceeding +/- 168 hours (one
+ week) would be accepted, but could then cause null-pointer dereference
+ crashes in certain operations. There's no use-case for such large UTC
+ offsets, so reject them.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix bugs in <type>tsquery</> <literal>@></> <type>tsquery</>
+ operator (Heikki Linnakangas)
+ </para>
+
+ <para>
+ Two different terms would be considered to match if they had the same
+ CRC. Also, if the second operand had more terms than the first, it
+ would be assumed not to be contained in the first; which is wrong
+ since it might contain duplicate terms.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve ispell dictionary's defenses against bad affix files (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Allow more than 64K phrases in a thesaurus dictionary (David Boutin)
+ </para>
+
+ <para>
+ The previous coding could crash on an oversize dictionary, so this was
+ deemed a back-patchable bug fix rather than a feature addition.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix namespace handling in <function>xpath()</> (Ali Akbar)
+ </para>
+
+ <para>
+ Previously, the <type>xml</> value resulting from
+ an <function>xpath()</> call would not have namespace declarations if
+ the namespace declarations were attached to an ancestor element in the
+ input <type>xml</> value, rather than to the specific element being
+ returned. Propagate the ancestral declaration so that the result is
+ correct when considered in isolation.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Ensure that whole-row variables expose nonempty column names
+ to functions that pay attention to column names within composite
+ arguments (Tom Lane)
+ </para>
+
+ <para>
+ In some contexts, constructs like <literal>row_to_json(tab.*)</> may
+ not produce the expected column names. This is fixed properly as of
+ 9.4; in older branches, just ensure that we produce some nonempty
+ name. (In some cases this will be the underlying table's column name
+ rather than the query-assigned alias that should theoretically be
+ visible.)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix mishandling of system columns,
+ particularly <structfield>tableoid</>, in FDW queries (Etsuro Fujita)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid doing <literal><replaceable>indexed_column</> = ANY
+ (<replaceable>array</>)</literal> as an index qualifier if that leads
+ to an inferior plan (Andrew Gierth)
+ </para>
+
+ <para>
+ In some cases, <literal>= ANY</> conditions applied to non-first index
+ columns would be done as index conditions even though it would be
+ better to use them as simple filter conditions.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix planner problems with nested append relations, such as inherited
+ tables within <literal>UNION ALL</> subqueries (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fail cleanly when a GiST index tuple doesn't fit on a page, rather
+ than going into infinite recursion (Andrew Gierth)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Exempt tables that have per-table <varname>cost_limit</>
+ and/or <varname>cost_delay</> settings from autovacuum's global cost
+ balancing rules (Álvaro Herrera)
+ </para>
+
+ <para>
+ The previous behavior resulted in basically ignoring these per-table
+ settings, which was unintended. Now, a table having such settings
+ will be vacuumed using those settings, independently of what is going
+ on in other autovacuum workers. This may result in heavier total I/O
+ load than before, so such settings should be re-examined for sanity.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid wholesale autovacuuming when autovacuum is nominally off
+ (Tom Lane)
+ </para>
+
+ <para>
+ Even when autovacuum is nominally off, we will still launch autovacuum
+ worker processes to vacuum tables that are at risk of XID wraparound.
+ However, such a worker process then proceeded to vacuum all tables in
+ the target database, if they met the usual thresholds for
+ autovacuuming. This is at best pretty unexpected; at worst it delays
+ response to the wraparound threat. Fix it so that if autovacuum is
+ turned off, workers <emphasis>only</> do anti-wraparound vacuums and
+ not any other work.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ During crash recovery, ensure that unlogged relations are rewritten as
+ empty and are synced to disk before recovery is considered complete
+ (Abhijit Menon-Sen, Andres Freund)
+ </para>
+
+ <para>
+ This prevents scenarios in which unlogged relations might contain
+ garbage data following database crash recovery.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix race condition between hot standby queries and replaying a
+ full-page image (Heikki Linnakangas)
+ </para>
+
+ <para>
+ This mistake could result in transient errors in queries being
+ executed in hot standby.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix several cases where recovery logic improperly ignored WAL records
+ for <literal>COMMIT/ABORT PREPARED</> (Heikki Linnakangas)
+ </para>
+
+ <para>
+ The most notable oversight was
+ that <varname>recovery_min_apply_delay</> failed to delay application
+ of a two-phase commit.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Prevent latest WAL file from being archived a second time at completion
+ of crash recovery (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid creating unnecessary <filename>.ready</> marker files for
+ timeline history files (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix possible null pointer dereference when an empty prepared statement
+ is used and the <varname>log_statement</> setting is <literal>mod</>
+ or <literal>ddl</> (Fujii Masao)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Change <quote>pgstat wait timeout</> warning message to be LOG level,
+ and rephrase it to be more understandable (Tom Lane)
+ </para>
+
+ <para>
+ This message was originally thought to be essentially a can't-happen
+ case, but it occurs often enough on our slower buildfarm members to be
+ a nuisance. Reduce it to LOG level, and expend a bit more effort on
+ the wording: it now reads <quote>using stale statistics instead of
+ current ones because stats collector is not responding</>.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix SPARC spinlock implementation to ensure correctness if the CPU is
+ being run in a non-TSO coherency mode, as some non-Solaris kernels do
+ (Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Warn if OS X's <function>setlocale()</> starts an unwanted extra
+ thread inside the postmaster (Noah Misch)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix processing of repeated <literal>dbname</> parameters
+ in <function>PQconnectdbParams()</> (Alex Shulgin)
+ </para>
+
+ <para>
+ Unexpected behavior ensued if the first occurrence
+ of <literal>dbname</> contained a connection string or URI to be
+ expanded.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Ensure that <application>libpq</> reports a suitable error message on
+ unexpected socket EOF (Marko Tiikkaja, Tom Lane)
+ </para>
+
+ <para>
+ Depending on kernel behavior, <application>libpq</> might return an
+ empty error string rather than something useful when the server
+ unexpectedly closed the socket.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Clear any old error message during <function>PQreset()</>
+ (Heikki Linnakangas)
+ </para>
+
+ <para>
+ If <function>PQreset()</> is called repeatedly, and the connection
+ cannot be re-established, error messages from the failed connection
+ attempts kept accumulating in the <structname>PGconn</>'s error
+ string.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Properly handle out-of-memory conditions while parsing connection
+ options in <application>libpq</> (Alex Shulgin, Heikki Linnakangas)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix array overrun in <application>ecpg</>'s version
+ of <function>ParseDateTime()</> (Michael Paquier)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <application>initdb</>, give a clearer error message if a password
+ file is specified but is empty (Mats Erik Andersson)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s <command>\s</> command to work nicely with
+ libedit, and add pager support (Stepan Rutz, Tom Lane)
+ </para>
+
+ <para>
+ When using libedit rather than readline, <command>\s</> printed the
+ command history in a fairly unreadable encoded format, and on recent
+ libedit versions might fail altogether. Fix that by printing the
+ history ourselves rather than having the library do it. A pleasant
+ side-effect is that the pager is used if appropriate.
+ </para>
+
+ <para>
+ This patch also fixes a bug that caused newline encoding to be applied
+ inconsistently when saving the command history with libedit.
+ Multiline history entries written by older <application>psql</>
+ versions will be read cleanly with this patch, but perhaps not
+ vice versa, depending on the exact libedit versions involved.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve consistency of parsing of <application>psql</>'s special
+ variables (Tom Lane)
+ </para>
+
+ <para>
+ Allow variant spellings of <literal>on</> and <literal>off</> (such
+ as <literal>1</>/<literal>0</>) for <literal>ECHO_HIDDEN</>
+ and <literal>ON_ERROR_ROLLBACK</>. Report a warning for unrecognized
+ values for <literal>COMP_KEYWORD_CASE</>, <literal>ECHO</>,
+ <literal>ECHO_HIDDEN</>, <literal>HISTCONTROL</>,
+ <literal>ON_ERROR_ROLLBACK</>, and <literal>VERBOSITY</>. Recognize
+ all values for all these variables case-insensitively; previously
+ there was a mishmash of case-sensitive and case-insensitive behaviors.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s expanded-mode display to work
+ consistently when using <literal>border</> = 3
+ and <literal>linestyle</> = <literal>ascii</> or <literal>unicode</>
+ (Stephen Frost)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve performance of <application>pg_dump</> when the database
+ contains many instances of multiple dependency paths between the same
+ two objects (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <application>pg_dumpall</> to restore its ability to dump from
+ pre-8.1 servers (Gilles Darold)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix possible deadlock during parallel restore of a schema-only dump
+ (Robert Haas, Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix core dump in <literal>pg_dump --binary-upgrade</> on zero-column
+ composite type (Rushabh Lathia)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Prevent WAL files created by <literal>pg_basebackup -x/-X</> from
+ being archived again when the standby is promoted (Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix failure of <filename>contrib/auto_explain</> to print per-node
+ timing information when doing <command>EXPLAIN ANALYZE</> (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix upgrade-from-unpackaged script for <filename>contrib/citext</>
+ (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix block number checking
+ in <filename>contrib/pageinspect</>'s <function>get_raw_page()</>
+ (Tom Lane)
+ </para>
+
+ <para>
+ The incorrect checking logic could prevent access to some pages in
+ non-main relation forks.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix <filename>contrib/pgcrypto</>'s <function>pgp_sym_decrypt()</>
+ to not fail on messages whose length is 6 less than a power of 2
+ (Marko Tiikkaja)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix file descriptor leak in <filename>contrib/pg_test_fsync</>
+ (Jeff Janes)
+ </para>
+
+ <para>
+ This could cause failure to remove temporary files on Windows.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Handle unexpected query results, especially NULLs, safely in
+ <filename>contrib/tablefunc</>'s <function>connectby()</>
+ (Michael Paquier)
+ </para>
+
+ <para>
+ <function>connectby()</> previously crashed if it encountered a NULL
+ key value. It now prints that row but doesn't recurse further.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Avoid a possible crash in <filename>contrib/xml2</>'s
+ <function>xslt_process()</> (Mark Simonetti)
+ </para>
+
+ <para>
+ <application>libxslt</> seems to have an undocumented dependency on
+ the order in which resources are freed; reorder our calls to avoid a
+ crash.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Mark some <filename>contrib</> I/O functions with correct volatility
+ properties (Tom Lane)
+ </para>
+
+ <para>
+ The previous over-conservative marking was immaterial in normal use,
+ but could cause optimization problems or rejection of valid index
+ expression definitions. Since the consequences are not large, we've
+ just adjusted the function definitions in the extension modules'
+ scripts, without changing version numbers.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Numerous cleanups of warnings from Coverity static code analyzer
+ (Andres Freund, Tatsuo Ishii, Marko Kreen, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ These changes are mostly cosmetic but in some cases fix corner-case
+ bugs, for example a crash rather than a proper error report after an
+ out-of-memory failure. None are believed to represent security
+ issues.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Detect incompatible OpenLDAP versions during build (Noah Misch)
+ </para>
+
+ <para>
+ With OpenLDAP versions 2.4.24 through 2.4.31,
+ inclusive, <productname>PostgreSQL</> backends can crash at exit.
+ Raise a warning during <application>configure</> based on the
+ compile-time OpenLDAP version number, and test the crashing scenario
+ in the <filename>contrib/dblink</> regression test.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In non-MSVC Windows builds, ensure <filename>libpq.dll</> is installed
+ with execute permissions (Noah Misch)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Make <application>pg_regress</> remove any temporary installation it
+ created upon successful exit (Tom Lane)
+ </para>
+
+ <para>
+ This results in a very substantial reduction in disk space usage
+ during <literal>make check-world</>, since that sequence involves
+ creation of numerous temporary installations.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Support time zone abbreviations that change UTC offset from time to
+ time (Tom Lane)
+ </para>
+
+ <para>
+ Previously, <productname>PostgreSQL</> assumed that the UTC offset
+ associated with a time zone abbreviation (such as <literal>EST</>)
+ never changes in the usage of any particular locale. However this
+ assumption fails in the real world, so introduce the ability for a
+ zone abbreviation to represent a UTC offset that sometimes changes.
+ Update the zone abbreviation definition files to make use of this
+ feature in timezone locales that have changed the UTC offset of their
+ abbreviations since 1970 (according to the IANA timezone database).
+ In such timezones, <productname>PostgreSQL</> will now associate the
+ correct UTC offset with the abbreviation depending on the given date.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Update time zone abbreviations lists (Tom Lane)
+ </para>
+
+ <para>
+ Add CST (China Standard Time) to our lists.
+ Remove references to ADT as <quote>Arabia Daylight Time</>, an
+ abbreviation that's been out of use since 2007; therefore, claiming
+ there is a conflict with <quote>Atlantic Daylight Time</> doesn't seem
+ especially helpful.
+ Fix entirely incorrect GMT offsets for CKT (Cook Islands), FJT, and FJST
+ (Fiji); we didn't even have them on the proper side of the date line.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Update time zone data files to <application>tzdata</> release 2015a.
+ </para>
+
+ <para>
+ The IANA timezone database has adopted abbreviations of the form
+ <literal>A<replaceable>x</>ST</literal>/<literal>A<replaceable>x</>DT</literal>
+ for all Australian time zones, reflecting what they believe to be
+ current majority practice Down Under. These names do not conflict
+ with usage elsewhere (other than ACST for Acre Summer Time, which has
+ been in disuse since 1994). Accordingly, adopt these names into
+ our <quote>Default</> timezone abbreviation set.
+ The <quote>Australia</> abbreviation set now contains only CST, EAST,
+ EST, SAST, SAT, and WST, all of which are thought to be mostly
+ historical usage. Note that SAST has also been changed to be South
+ Africa Standard Time in the <quote>Default</> abbreviation set.
+ </para>
+
+ <para>
+ Also, add zone abbreviations SRET (Asia/Srednekolymsk) and XJT
+ (Asia/Urumqi), and use WSST/WSDT for western Samoa. Also, there were
+ DST law changes in Chile, Mexico, the Turks & Caicos Islands
+ (America/Grand_Turk), and Fiji. There is a new zone
+ Pacific/Bougainville for portions of Papua New Guinea. Also, numerous
+ corrections for historical (pre-1970) time zone data.
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ </sect2>
+ </sect1>
+
<sect1 id="release-9-2-9">
<title>Release 9.2.9</title>
<!-- doc/src/sgml/release-9.3.sgml -->
<!-- See header comment in release.sgml about typical markup -->
+ <sect1 id="release-9-3-6">
+ <title>Release 9.3.6</title>
+
+ <note>
+ <title>Release Date</title>
+ <simpara>2015-02-05</simpara>
+ </note>
+
+ <para>
+ This release contains a variety of fixes from 9.3.5.
+ For information about new features in the 9.3 major release, see
+ <xref linkend="release-9-3">.
+ </para>
+
+ <sect2>
+ <title>Migration to Version 9.3.6</title>
+
+ <para>
+ A dump/restore is not required for those running 9.3.X.
+ </para>
+
+ <para>
+ However, if you are a Windows user and are using the <quote>Norwegian
+ (Bokmål)</> locale, manual action is needed after the upgrade to
+ replace any <quote>Norwegian (Bokmål)_Norway</> locale names stored
+ in <productname>PostgreSQL</> system catalogs with the plain-ASCII
+ alias <quote>Norwegian_Norway</>. For details see
+ <ulink url="http://wiki.postgresql.org/wiki/Changes_To_Norwegian_Locale"></>
+ </para>
+
+ <para>
+ Also, if you are upgrading from a version earlier than 9.3.5,
+ see <xref linkend="release-9-3-5">.
+ </para>
+
+ </sect2>
+
+ <sect2>
+ <title>Changes</title>
+
+ <itemizedlist>
+
+<!--
+Author: Stephen Frost <sfrost@snowman.net>
+Branch: master [804b6b6db] 2015-01-28 12:31:30 -0500
+Branch: REL9_4_STABLE [3cc74a3d6] 2015-01-28 12:32:06 -0500
+Branch: REL9_3_STABLE [4b9874216] 2015-01-28 12:32:39 -0500
+Branch: REL9_2_STABLE [d49f84b08] 2015-01-28 12:32:56 -0500
+Branch: REL9_1_STABLE [9406884af] 2015-01-28 12:33:15 -0500
+Branch: REL9_0_STABLE [3a2063369] 2015-01-28 12:33:29 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix information leak via constraint-violation error messages
+ (Stephen Frost)
+ </para>
+
+ <para>
+ Some server error messages show the values of columns that violate
+ a constraint, such as a unique constraint. If the user does not have
+ <literal>SELECT</> privilege on all columns of the table, this could
+ mean exposing values that the user should not be able to see. Adjust
+ the code so that values are displayed only when they came from the SQL
+ command or could be selected by the user.
+ (CVE-2014-8161)
+ </para>
+ </listitem>
+
+<!--
+Author: Noah Misch <noah@leadboat.com>
+Branch: master [f6dc6dd5b] 2014-12-17 22:48:40 -0500
+Branch: REL9_4_STABLE [6b87d423d] 2014-12-17 22:48:45 -0500
+Branch: REL9_3_STABLE [442dc2c35] 2014-12-17 22:48:46 -0500
+Branch: REL9_2_STABLE [0046f651d] 2014-12-17 22:48:47 -0500
+Branch: REL9_1_STABLE [6aa98e957] 2014-12-17 22:48:47 -0500
+Branch: REL9_0_STABLE [6d45ee572] 2014-12-17 22:48:48 -0500
+-->
+
+ <listitem>
+ <para>
+ Lock down regression testing's temporary installations on Windows
+ (Noah Misch)
+ </para>
+
+ <para>
+ Use SSPI authentication to allow connections only from the OS user
+ who launched the test suite. This closes on Windows the same
+ vulnerability previously closed on other platforms, namely that other
+ users might be able to connect to the test postmaster.
+ (CVE-2014-0067)
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [8f80dcf3c] 2014-10-24 19:59:49 +0300
+Branch: REL9_2_STABLE [d440c4b55] 2014-10-24 19:59:52 +0300
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [2a1b34959] 2014-10-24 19:36:28 +0300
+Branch: REL9_2_STABLE [737ae3fc7] 2014-10-24 19:53:27 +0300
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master [aa1d2fc5e] 2015-01-16 13:28:19 +0200
+Branch: REL9_4_STABLE [2049a7d82] 2015-01-16 13:10:06 +0200
+Branch: REL9_3_STABLE [1619442a1] 2015-01-16 13:10:15 +0200
+Branch: REL9_2_STABLE [6bf343c6e] 2015-01-16 13:10:23 +0200
+-->
+
+ <listitem>
+ <para>
+ Cope with the Windows locale named <quote>Norwegian (Bokmål)</>
+ (Heikki Linnakangas)
+ </para>
+
+ <para>
+ Non-ASCII locale names are problematic since it's not clear what
+ encoding they should be represented in. Map the troublesome locale
+ name to a plain-ASCII alias, <quote>Norwegian_Norway</>.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [f88300168] 2014-11-04 13:24:14 -0500
+Branch: REL9_2_STABLE [db72ad02e] 2014-11-04 13:24:17 -0500
+Branch: REL9_1_STABLE [7c6f55e9e] 2014-11-04 13:24:22 -0500
+Branch: REL9_0_STABLE [45a607d5c] 2014-11-04 13:24:26 -0500
+-->
+
+ <listitem>
+ <para>
+ Avoid possible data corruption if <command>ALTER DATABASE SET
+ TABLESPACE</> is used to move a database to a new tablespace and then
+ shortly later move it back to its original tablespace (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [81f0a5e38] 2014-10-29 18:12:08 -0400
+Branch: REL9_2_STABLE [40058fbce] 2014-10-29 18:12:11 -0400
+Branch: REL9_1_STABLE [6ec1c3ef8] 2014-10-29 18:12:17 -0400
+Branch: REL9_0_STABLE [9d06da58e] 2014-10-29 18:12:20 -0400
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [e65b550b3] 2014-10-30 13:03:28 -0400
+Branch: REL9_2_STABLE [38cb8687a] 2014-10-30 13:03:31 -0400
+Branch: REL9_1_STABLE [fcf0246b2] 2014-10-30 13:03:34 -0400
+Branch: REL9_0_STABLE [73f950fc8] 2014-10-30 13:03:39 -0400
+-->
+
+ <listitem>
+ <para>
+ Avoid corrupting tables when <command>ANALYZE</> inside a transaction
+ is rolled back (Andres Freund, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ If the failing transaction had earlier removed the last index, rule, or
+ trigger from the table, the table would be left in a corrupted state
+ with the relevant <structname>pg_class</> flags not set though they
+ should be.
+ </para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [d7624e562] 2014-10-20 23:45:31 +0200
+Branch: REL9_2_STABLE [fd29810d1] 2014-10-20 23:47:00 +0200
+Branch: REL9_1_STABLE [d5fef87e9] 2014-10-20 23:47:45 +0200
+-->
+
+ <listitem>
+ <para>
+ Ensure that unlogged tables are copied correctly
+ during <command>CREATE DATABASE</> or <command>ALTER DATABASE SET
+ TABLESPACE</> (Pavan Deolasee, Andres Freund)
+ </para>
+ </listitem>
+
+<!--
+Author: Robert Haas <rhaas@postgresql.org>
+Branch: REL9_3_STABLE [e35db342a] 2014-09-22 16:19:59 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix incorrect processing
+ of <structname>CreateEventTrigStmt</>.<structfield>eventname</> (Petr
+ Jelinek)
+ </para>
+
+ <para>
+ This could result in misbehavior if <command>CREATE EVENT TRIGGER</>
+ were executed as a prepared query, or via extended query protocol.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [2a83e0349] 2014-11-11 17:00:21 -0500
+Branch: REL9_2_STABLE [1c2f9a4f6] 2014-11-11 17:00:25 -0500
+Branch: REL9_1_STABLE [94d5d57d5] 2014-11-11 17:00:28 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix <command>DROP</>'s dependency searching to correctly handle the
+ case where a table column is recursively visited before its table
+ (Petr Jelinek, Tom Lane)
+ </para>
+
+ <para>
+ This case is only known to arise when an extension creates both a
+ datatype and a table using that datatype. The faulty code might
+ refuse a <command>DROP EXTENSION</> unless <literal>CASCADE</> is
+ specified, which should not be required.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [c480cb9d2] 2015-01-15 18:52:58 -0500
+Branch: REL9_4_STABLE [b75d18bd4] 2015-01-15 18:53:05 -0500
+Branch: REL9_3_STABLE [34668c8ec] 2015-01-15 18:52:28 -0500
+Branch: REL9_2_STABLE [0acb32efb] 2015-01-15 18:52:31 -0500
+Branch: REL9_1_STABLE [450530fce] 2015-01-15 18:52:34 -0500
+Branch: REL9_0_STABLE [5308e085b] 2015-01-15 18:52:38 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix use-of-already-freed-memory problem in EvalPlanQual processing
+ (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, queries that lock or update
+ recently-updated rows could crash as a result of this bug.
+ </para>
+ </listitem>
+
+<!--
+Author: Alvaro Herrera <alvherre@alvh.no-ip.org>
+Branch: master [0e5680f47] 2014-12-26 13:52:27 -0300
+Branch: REL9_4_STABLE [0e3a1f71d] 2014-12-26 13:52:27 -0300
+Branch: REL9_3_STABLE [048912386] 2014-12-26 13:52:27 -0300
+-->
+
+ <listitem>
+ <para>
+ Avoid possible deadlock while trying to acquire tuple locks
+ in EvalPlanQual processing (Álvaro Herrera, Mark Kirkwood)
+ </para>
+ </listitem>
+
+<!--
+Author: Alvaro Herrera <alvherre@alvh.no-ip.org>
+Branch: master [d5e3d1e96] 2015-01-04 15:48:29 -0300
+Branch: REL9_4_STABLE [51742063b] 2015-01-04 15:48:29 -0300
+Branch: REL9_3_STABLE [54a8abc2b] 2015-01-04 15:48:29 -0300
+-->
+
+ <listitem>
+ <para>
+ Fix failure to wait when a transaction tries to acquire a <literal>FOR
+ NO KEY EXCLUSIVE</> tuple lock, while multiple other transactions
+ currently hold <literal>FOR SHARE</> locks (Álvaro Herrera)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [2ae8a01ca] 2014-12-11 21:02:31 -0500
+Branch: REL9_2_STABLE [cd63c57e5] 2014-12-11 21:02:34 -0500
+Branch: REL9_1_STABLE [bca39b578] 2014-12-11 21:02:38 -0500
+Branch: REL9_0_STABLE [662eebdc6] 2014-12-11 21:02:41 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix planning of <command>SELECT FOR UPDATE</> when using a partial
+ index on a child table (Kyotaro Horiguchi)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, <command>SELECT FOR UPDATE</> must
+ also recheck the partial index's <literal>WHERE</> condition when
+ rechecking a recently-updated row to see if it still satisfies the
+ query's <literal>WHERE</> condition. This requirement was missed if the
+ index belonged to an inheritance child table, so that it was possible
+ to incorrectly return rows that no longer satisfy the query condition.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [f14196c35] 2014-12-11 19:37:07 -0500
+Branch: REL9_2_STABLE [deadbf4f3] 2014-12-11 19:37:10 -0500
+Branch: REL9_1_STABLE [21946ac9b] 2014-12-11 19:37:14 -0500
+Branch: REL9_0_STABLE [f5e4e92fb] 2014-12-11 19:37:17 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix corner case wherein <command>SELECT FOR UPDATE</> could return a row
+ twice, and possibly miss returning other rows (Tom Lane)
+ </para>
+
+ <para>
+ In <literal>READ COMMITTED</> mode, a <command>SELECT FOR UPDATE</>
+ that is scanning an inheritance tree could incorrectly return a row
+ from a prior child table instead of the one it should return from a
+ later child table.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [a5cd70dcb] 2015-01-15 13:18:12 -0500
+Branch: REL9_4_STABLE [d25192892] 2015-01-15 13:18:16 -0500
+Branch: REL9_3_STABLE [939f0fb67] 2015-01-15 13:18:19 -0500
+-->
+
+ <listitem>
+ <para>
+ Improve performance of <command>EXPLAIN</> with large range tables
+ (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [7a9c8cefb] 2014-08-09 13:46:42 -0400
+Branch: REL9_2_STABLE [b4dacab12] 2014-08-09 13:46:45 -0400
+Branch: REL9_1_STABLE [bbe826f21] 2014-08-09 13:46:48 -0400
+Branch: REL9_0_STABLE [4ff49746e] 2014-08-09 13:46:52 -0400
+-->
+
+ <listitem>
+ <para>
+ Reject duplicate column names in the referenced-columns list of
+ a <literal>FOREIGN KEY</> declaration (David Rowley)
+ </para>
+
+ <para>
+ This restriction is per SQL standard. Previously we did not reject
+ the case explicitly, but later on the code would fail with
+ bizarre-looking errors.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [6306d0712] 2014-07-22 13:30:14 -0400
+-->
+
+ <listitem>
+ <para>
+ Re-enable error for <literal>SELECT ... OFFSET -1</> (Tom Lane)
+ </para>
+
+ <para>
+ A negative offset value has been an error since 8.4, but an
+ optimization added in 9.3 accidentally turned the case into a no-op.
+ Restore the expected behavior.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [8b65e0a33] 2014-08-09 18:40:34 -0400
+Branch: REL9_2_STABLE [15026ab97] 2014-08-09 18:40:38 -0400
+-->
+
+ <listitem>
+ <para>
+ Restore previous behavior of conversion of domains to JSON
+ (Tom Lane)
+ </para>
+
+ <para>
+ This change causes domains over numeric and boolean to be treated
+ like their base types for purposes of conversion to JSON. It worked
+ like that before 9.3.5 and 9.2.9, but was unintentionally changed
+ while fixing a related problem.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [8571ecb24] 2014-12-02 15:02:43 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix <function>json_agg()</> to not return extra trailing right
+ brackets in its result (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [25bf13fe1] 2014-09-11 23:30:57 -0400
+Branch: REL9_2_STABLE [228ed4370] 2014-09-11 23:31:00 -0400
+Branch: REL9_1_STABLE [cf5c20b06] 2014-09-11 23:31:03 -0400
+Branch: REL9_0_STABLE [26f8a4691] 2014-09-11 23:31:06 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix bugs in raising a <type>numeric</> value to a large integral power
+ (Tom Lane)
+ </para>
+
+ <para>
+ The previous code could get a wrong answer, or consume excessive
+ amounts of time and memory before realizing that the answer must
+ overflow.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [10b81fbdc] 2014-12-01 15:25:08 -0500
+Branch: REL9_2_STABLE [e640042e9] 2014-12-01 15:25:12 -0500
+Branch: REL9_1_STABLE [2e3cc3955] 2014-12-01 15:25:15 -0500
+Branch: REL9_0_STABLE [e6550626c] 2014-12-01 15:25:18 -0500
+-->
+
+ <listitem>
+ <para>
+ In <function>numeric_recv()</>, truncate away any fractional digits
+ that would be hidden according to the value's <literal>dscale</> field
+ (Tom Lane)
+ </para>
+
+ <para>
+ A <type>numeric</> value's display scale (<literal>dscale</>) should
+ never be less than the number of nonzero fractional digits; but
+ apparently there's at least one broken client application that
+ transmits binary <type>numeric</> values in which that's true.
+ This leads to strange behavior since the extra digits are taken into
+ account by arithmetic operations even though they aren't printed.
+ The least risky fix seems to be to truncate away such <quote>hidden</>
+ digits on receipt, so that the value is indeed what it prints as.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [bbfdf5d75] 2014-09-23 20:25:36 -0400
+Branch: REL9_2_STABLE [3359a818c] 2014-09-23 20:25:39 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix incorrect search for shortest-first regular expression matches
+ (Tom Lane)
+ </para>
+
+ <para>
+ Matching would often fail when the number of allowed iterations is
+ limited by a <literal>?</> quantifier or a bound expression.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [7672bbca0] 2014-07-21 22:41:27 -0400
+Branch: REL9_2_STABLE [f54d97c5e] 2014-07-21 22:41:30 -0400
+Branch: REL9_1_STABLE [124331b61] 2014-07-21 22:41:33 -0400
+Branch: REL9_0_STABLE [6e5a39c9e] 2014-07-21 22:41:36 -0400
+-->
+
+ <listitem>
+ <para>
+ Reject out-of-range numeric timezone specifications (Tom Lane)
+ </para>
+
+ <para>
+ Simple numeric timezone specifications exceeding +/- 168 hours (one
+ week) would be accepted, but could then cause null-pointer dereference
+ crashes in certain operations. There's no use-case for such large UTC
+ offsets, so reject them.
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [1aa526f3f] 2014-10-27 10:51:29 +0200
+Branch: REL9_2_STABLE [604d94d4c] 2014-10-27 10:51:33 +0200
+Branch: REL9_1_STABLE [4d1c738d1] 2014-10-27 10:51:36 +0200
+Branch: REL9_0_STABLE [10059c2da] 2014-10-27 10:51:38 +0200
+-->
+
+ <listitem>
+ <para>
+ Fix bugs in <type>tsquery</> <literal>@></> <type>tsquery</>
+ operator (Heikki Linnakangas)
+ </para>
+
+ <para>
+ Two different terms would be considered to match if they had the same
+ CRC. Also, if the second operand had more terms than the first, it
+ would be assumed not to be contained in the first; which is wrong
+ since it might contain duplicate terms.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [385f0d98a] 2014-10-23 13:11:34 -0400
+Branch: REL9_2_STABLE [f09369da8] 2014-10-23 13:11:37 -0400
+Branch: REL9_1_STABLE [94de3a679] 2014-10-23 13:11:41 -0400
+Branch: REL9_0_STABLE [21fa26b65] 2014-10-23 13:11:45 -0400
+-->
+
+ <listitem>
+ <para>
+ Improve ispell dictionary's defenses against bad affix files (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [5a74ff373] 2014-11-06 20:52:52 -0500
+Branch: REL9_2_STABLE [0bb318595] 2014-11-06 20:52:57 -0500
+Branch: REL9_1_STABLE [076688084] 2014-11-06 20:53:02 -0500
+Branch: REL9_0_STABLE [39493e4d9] 2014-11-06 20:53:07 -0500
+-->
+
+ <listitem>
+ <para>
+ Allow more than 64K phrases in a thesaurus dictionary (David Boutin)
+ </para>
+
+ <para>
+ The previous coding could crash on an oversize dictionary, so this was
+ deemed a back-patchable bug fix rather than a feature addition.
+ </para>
+ </listitem>
+
+<!--
+Author: Peter Eisentraut <peter_e@gmx.net>
+Branch: master [79af9a1d2] 2015-01-06 23:06:13 -0500
+Branch: REL9_4_STABLE [6bbf75192] 2015-01-17 22:11:20 -0500
+Branch: REL9_3_STABLE [e32cb8d0e] 2015-01-17 22:13:27 -0500
+Branch: REL9_2_STABLE [c8ef5b1ac] 2015-01-17 22:14:21 -0500
+Branch: REL9_1_STABLE [c975fa471] 2015-01-17 22:37:07 -0500
+Branch: REL9_0_STABLE [cebb3f032] 2015-01-17 22:37:32 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix namespace handling in <function>xpath()</> (Ali Akbar)
+ </para>
+
+ <para>
+ Previously, the <type>xml</> value resulting from
+ an <function>xpath()</> call would not have namespace declarations if
+ the namespace declarations were attached to an ancestor element in the
+ input <type>xml</> value, rather than to the specific element being
+ returned. Propagate the ancestral declaration so that the result is
+ correct when considered in isolation.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [07ab4ec4c] 2014-11-10 15:21:20 -0500
+Branch: REL9_2_STABLE [19ccaf9d4] 2014-11-10 15:21:26 -0500
+-->
+
+ <listitem>
+ <para>
+ Ensure that whole-row variables expose nonempty column names
+ to functions that pay attention to column names within composite
+ arguments (Tom Lane)
+ </para>
+
+ <para>
+ In some contexts, constructs like <literal>row_to_json(tab.*)</> may
+ not produce the expected column names. This is fixed properly as of
+ 9.4; in older branches, just ensure that we produce some nonempty
+ name. (In some cases this will be the underlying table's column name
+ rather than the query-assigned alias that should theoretically be
+ visible.)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [c57cdc9c1] 2014-11-22 16:01:12 -0500
+Branch: REL9_2_STABLE [906599f65] 2014-11-22 16:01:15 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix mishandling of system columns,
+ particularly <structfield>tableoid</>, in FDW queries (Etsuro Fujita)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [3d660d33a] 2015-01-30 12:30:59 -0500
+Branch: REL9_4_STABLE [b6a164e5c] 2015-01-30 12:31:08 -0500
+Branch: REL9_3_STABLE [527ff8baf] 2015-01-30 12:30:43 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix assorted oversights in range-operator selectivity estimation
+ (Emre Hasegeli)
+ </para>
+
+ <para>
+ This patch fixes corner-case <quote>unexpected operator NNNN</> planner
+ errors, and improves the selectivity estimates for some other cases.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [f6abf8f08] 2014-10-26 16:12:29 -0400
+Branch: REL9_2_STABLE [4586572d7] 2014-10-26 16:12:32 -0400
+-->
+
+ <listitem>
+ <para>
+ Avoid doing <literal><replaceable>indexed_column</> = ANY
+ (<replaceable>array</>)</literal> as an index qualifier if that leads
+ to an inferior plan (Andrew Gierth)
+ </para>
+
+ <para>
+ In some cases, <literal>= ANY</> conditions applied to non-first index
+ columns would be done as index conditions even though it would be
+ better to use them as simple filter conditions.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [4e54685d0] 2014-10-20 12:23:48 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix <quote>variable not found in subplan target list</> planner
+ failure when an inline-able SQL function taking a composite argument
+ is used in a <literal>LATERAL</> subselect and the composite argument
+ is a lateral reference (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [b2b95de61] 2014-10-01 19:30:30 -0400
+Branch: REL9_2_STABLE [71b88cf52] 2014-10-01 19:30:34 -0400
+Branch: REL9_1_STABLE [477023e94] 2014-10-01 19:30:38 -0400
+Branch: REL9_0_STABLE [288f15b7c] 2014-10-01 19:30:41 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix planner problems with nested append relations, such as inherited
+ tables within <literal>UNION ALL</> subqueries (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [ef8ac584e] 2014-10-03 14:50:29 +0300
+Branch: REL9_2_STABLE [8e137b075] 2014-10-03 14:50:40 +0300
+Branch: REL9_1_STABLE [f270a1611] 2014-10-03 14:50:46 +0300
+Branch: REL9_0_STABLE [f04b112d5] 2014-10-03 14:50:58 +0300
+-->
+
+ <listitem>
+ <para>
+ Fail cleanly when a GiST index tuple doesn't fit on a page, rather
+ than going into infinite recursion (Andrew Gierth)
+ </para>
+ </listitem>
+
+<!--
+Author: Alvaro Herrera <alvherre@alvh.no-ip.org>
+Branch: REL9_3_STABLE [67ed9d531] 2014-10-03 13:01:27 -0300
+Branch: REL9_2_STABLE [769d6815e] 2014-10-03 13:01:27 -0300
+Branch: REL9_1_STABLE [06646f52e] 2014-10-03 13:01:27 -0300
+Branch: REL9_0_STABLE [50a757698] 2014-10-03 13:01:27 -0300
+-->
+
+ <listitem>
+ <para>
+ Exempt tables that have per-table <varname>cost_limit</>
+ and/or <varname>cost_delay</> settings from autovacuum's global cost
+ balancing rules (Álvaro Herrera)
+ </para>
+
+ <para>
+ The previous behavior resulted in basically ignoring these per-table
+ settings, which was unintended. Now, a table having such settings
+ will be vacuumed using those settings, independently of what is going
+ on in other autovacuum workers. This may result in heavier total I/O
+ load than before, so such settings should be re-examined for sanity.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [4cbecdaaa] 2014-07-30 14:41:49 -0400
+Branch: REL9_2_STABLE [c83aed34b] 2014-07-30 14:41:53 -0400
+Branch: REL9_1_STABLE [f21afe421] 2014-07-30 14:41:58 -0400
+Branch: REL9_0_STABLE [91b4a881c] 2014-07-30 14:42:12 -0400
+-->
+
+ <listitem>
+ <para>
+ Avoid wholesale autovacuuming when autovacuum is nominally off
+ (Tom Lane)
+ </para>
+
+ <para>
+ Even when autovacuum is nominally off, we will still launch autovacuum
+ worker processes to vacuum tables that are at risk of XID wraparound.
+ However, such a worker process then proceeded to vacuum all tables in
+ the target database, if they met the usual thresholds for
+ autovacuuming. This is at best pretty unexpected; at worst it delays
+ response to the wraparound threat. Fix it so that if autovacuum is
+ turned off, workers <emphasis>only</> do anti-wraparound vacuums and
+ not any other work.
+ </para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [2c3ebfd1a] 2014-11-15 01:21:07 +0100
+Branch: REL9_2_STABLE [ea8b3833e] 2014-11-15 01:21:45 +0100
+Branch: REL9_1_STABLE [d85b646d0] 2014-11-15 01:22:32 +0100
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [672b43e68] 2014-11-15 01:21:02 +0100
+Branch: REL9_2_STABLE [c5baa708f] 2014-11-15 01:21:40 +0100
+Branch: REL9_1_STABLE [fde9994bc] 2014-11-15 01:22:32 +0100
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [c7299d32f] 2014-11-15 01:20:29 +0100
+Branch: REL9_2_STABLE [86673a44a] 2014-11-15 01:21:30 +0100
+Branch: REL9_1_STABLE [b0a48e996] 2014-11-15 01:22:32 +0100
+-->
+
+ <listitem>
+ <para>
+ During crash recovery, ensure that unlogged relations are rewritten as
+ empty and are synced to disk before recovery is considered complete
+ (Abhijit Menon-Sen, Andres Freund)
+ </para>
+
+ <para>
+ This prevents scenarios in which unlogged relations might contain
+ garbage data following database crash recovery.
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [861d3aa43] 2014-11-13 20:01:55 +0200
+Branch: REL9_2_STABLE [7eab804c2] 2014-11-13 20:01:18 +0200
+Branch: REL9_1_STABLE [5f1d931cf] 2014-11-13 20:01:09 +0200
+Branch: REL9_0_STABLE [681dbe7d4] 2014-11-13 20:00:51 +0200
+-->
+
+ <listitem>
+ <para>
+ Fix race condition between hot standby queries and replaying a
+ full-page image (Heikki Linnakangas)
+ </para>
+
+ <para>
+ This mistake could result in transient errors in queries being
+ executed in hot standby.
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [a2a718b22] 2014-07-29 11:57:52 +0300
+Branch: REL9_2_STABLE [1578d13dc] 2014-07-29 11:58:01 +0300
+Branch: REL9_1_STABLE [de88ec6bf] 2014-07-29 11:58:09 +0300
+Branch: REL9_0_STABLE [804983961] 2014-07-29 11:58:17 +0300
+-->
+
+ <listitem>
+ <para>
+ Fix several cases where recovery logic improperly ignored WAL records
+ for <literal>COMMIT/ABORT PREPARED</> (Heikki Linnakangas)
+ </para>
+
+ <para>
+ The most notable oversight was
+ that <varname>recovery_min_apply_delay</> failed to delay application
+ of a two-phase commit.
+ </para>
+ </listitem>
+
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: REL9_3_STABLE [d45cd9e19] 2014-10-23 16:22:46 +0900
+Branch: REL9_2_STABLE [f904e9afd] 2014-10-23 16:31:44 +0900
+-->
+
+ <listitem>
+ <para>
+ Prevent latest WAL file from being archived a second time at completion
+ of crash recovery (Fujii Masao)
+ </para>
+ </listitem>
+
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: REL9_3_STABLE [3a3b7e316] 2014-11-06 21:25:18 +0900
+Branch: REL9_2_STABLE [38eb5d9e8] 2014-11-06 21:25:45 +0900
+Branch: REL9_1_STABLE [4e7468074] 2014-11-06 21:26:15 +0900
+Branch: REL9_0_STABLE [83c7bfb9a] 2014-11-06 21:26:21 +0900
+-->
+
+ <listitem>
+ <para>
+ Avoid creating unnecessary <filename>.ready</> marker files for
+ timeline history files (Fujii Masao)
+ </para>
+ </listitem>
+
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: REL9_3_STABLE [52eed3d42] 2014-09-05 02:19:29 +0900
+Branch: REL9_2_STABLE [78b1228ef] 2014-09-05 02:19:45 +0900
+Branch: REL9_1_STABLE [81bffe63c] 2014-09-05 02:19:50 +0900
+Branch: REL9_0_STABLE [857a5d6b5] 2014-09-05 02:19:57 +0900
+-->
+
+ <listitem>
+ <para>
+ Fix possible null pointer dereference when an empty prepared statement
+ is used and the <varname>log_statement</> setting is <literal>mod</>
+ or <literal>ddl</> (Fujii Masao)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [75b48e1ff] 2015-01-19 23:01:33 -0500
+Branch: REL9_4_STABLE [3387cbbcb] 2015-01-19 23:01:36 -0500
+Branch: REL9_3_STABLE [19794e997] 2015-01-19 23:01:39 -0500
+Branch: REL9_2_STABLE [33b723538] 2015-01-19 23:01:41 -0500
+Branch: REL9_1_STABLE [b87c1dcef] 2015-01-19 23:01:44 -0500
+Branch: REL9_0_STABLE [a1a8d0249] 2015-01-19 23:01:46 -0500
+-->
+
+ <listitem>
+ <para>
+ Change <quote>pgstat wait timeout</> warning message to be LOG level,
+ and rephrase it to be more understandable (Tom Lane)
+ </para>
+
+ <para>
+ This message was originally thought to be essentially a can't-happen
+ case, but it occurs often enough on our slower buildfarm members to be
+ a nuisance. Reduce it to LOG level, and expend a bit more effort on
+ the wording: it now reads <quote>using stale statistics instead of
+ current ones because stats collector is not responding</>.
+ </para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [cbd9619ac] 2014-10-01 14:34:06 +0200
+-->
+
+ <listitem>
+ <para>
+ Fix possible corruption of postmaster's list of dynamic background
+ workers (Andres Freund)
+ </para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [27ef6b653] 2014-09-09 23:37:33 +0200
+Branch: REL9_2_STABLE [d0b7ffc0f] 2014-09-09 23:37:50 +0200
+Branch: REL9_1_STABLE [5af508f66] 2014-09-09 23:45:07 +0200
+Branch: REL9_0_STABLE [f25e89601] 2014-09-09 23:48:03 +0200
+-->
+
+ <listitem>
+ <para>
+ Fix SPARC spinlock implementation to ensure correctness if the CPU is
+ being run in a non-TSO coherency mode, as some non-Solaris kernels do
+ (Andres Freund)
+ </para>
+ </listitem>
+
+<!--
+Author: Noah Misch <noah@leadboat.com>
+Branch: master [894459e59] 2015-01-07 22:35:44 -0500
+Branch: REL9_4_STABLE [83fb1ca5c] 2015-01-07 22:36:35 -0500
+Branch: REL9_3_STABLE [1a366d51e] 2015-01-07 22:40:40 -0500
+Branch: REL9_2_STABLE [5ca4e444c] 2015-01-07 22:41:49 -0500
+Branch: REL9_1_STABLE [8dc83104e] 2015-01-07 22:42:42 -0500
+Branch: REL9_0_STABLE [2e4946169] 2015-01-07 22:46:20 -0500
+-->
+
+ <listitem>
+ <para>
+ Warn if OS X's <function>setlocale()</> starts an unwanted extra
+ thread inside the postmaster (Noah Misch)
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [08cd4d9a6] 2014-11-25 17:39:00 +0200
+Branch: REL9_2_STABLE [9b468bcec] 2014-11-25 17:39:04 +0200
+Branch: REL9_1_STABLE [95be34362] 2014-11-25 17:39:07 +0200
+Branch: REL9_0_STABLE [9880fea4f] 2014-11-25 17:39:09 +0200
+-->
+
+ <listitem>
+ <para>
+ Fix processing of repeated <literal>dbname</> parameters
+ in <function>PQconnectdbParams()</> (Alex Shulgin)
+ </para>
+
+ <para>
+ Unexpected behavior ensued if the first occurrence
+ of <literal>dbname</> contained a connection string or URI to be
+ expanded.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [52ef33f72] 2014-10-22 18:41:51 -0400
+Branch: REL9_2_STABLE [6fa31d8d1] 2014-10-22 18:41:54 -0400
+Branch: REL9_1_STABLE [98144378c] 2014-10-22 18:41:57 -0400
+Branch: REL9_0_STABLE [ac6e87537] 2014-10-22 18:42:01 -0400
+-->
+
+ <listitem>
+ <para>
+ Ensure that <application>libpq</> reports a suitable error message on
+ unexpected socket EOF (Marko Tiikkaja, Tom Lane)
+ </para>
+
+ <para>
+ Depending on kernel behavior, <application>libpq</> might return an
+ empty error string rather than something useful when the server
+ unexpectedly closed the socket.
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [1325b239b] 2014-10-29 14:40:47 +0200
+Branch: REL9_2_STABLE [1a27fbd07] 2014-10-29 14:39:10 +0200
+Branch: REL9_1_STABLE [8f7bd8e91] 2014-10-29 14:36:19 +0200
+Branch: REL9_0_STABLE [49ef4eba2] 2014-10-29 14:35:39 +0200
+-->
+
+ <listitem>
+ <para>
+ Clear any old error message during <function>PQreset()</>
+ (Heikki Linnakangas)
+ </para>
+
+ <para>
+ If <function>PQreset()</> is called repeatedly, and the connection
+ cannot be re-established, error messages from the failed connection
+ attempts kept accumulating in the <structname>PGconn</>'s error
+ string.
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [d3b162a3d] 2014-11-25 14:10:29 +0200
+Branch: REL9_2_STABLE [abcab2a64] 2014-11-25 14:10:43 +0200
+Branch: REL9_1_STABLE [5053ad206] 2014-11-25 14:10:48 +0200
+Branch: REL9_0_STABLE [1f3517039] 2014-11-25 14:10:54 +0200
+-->
+
+ <listitem>
+ <para>
+ Properly handle out-of-memory conditions while parsing connection
+ options in <application>libpq</> (Alex Shulgin, Heikki Linnakangas)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [d3cfe20c6] 2014-10-06 21:23:35 -0400
+Branch: REL9_2_STABLE [b513205bc] 2014-10-06 21:23:39 -0400
+Branch: REL9_1_STABLE [037b912ec] 2014-10-06 21:23:45 -0400
+Branch: REL9_0_STABLE [d9a1e9de5] 2014-10-06 21:23:50 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix array overrun in <application>ecpg</>'s version
+ of <function>ParseDateTime()</> (Michael Paquier)
+ </para>
+ </listitem>
+
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: REL9_3_STABLE [2df66f01a] 2014-12-05 14:31:51 +0200
+Branch: REL9_2_STABLE [2b53d583d] 2014-12-05 14:31:45 +0200
+Branch: REL9_1_STABLE [729202754] 2014-12-05 14:31:37 +0200
+Branch: REL9_0_STABLE [d67be559e] 2014-12-05 14:30:55 +0200
+-->
+
+ <listitem>
+ <para>
+ In <application>initdb</>, give a clearer error message if a password
+ file is specified but is empty (Mats Erik Andersson)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [b0fd5c552] 2014-09-08 16:09:52 -0400
+Branch: REL9_2_STABLE [886b58b47] 2014-09-08 16:09:56 -0400
+Branch: REL9_1_STABLE [440fcc568] 2014-09-08 16:10:01 -0400
+Branch: REL9_0_STABLE [44c518328] 2014-09-08 16:10:05 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s <command>\s</> command to work nicely with
+ libedit, and add pager support (Stepan Rutz, Tom Lane)
+ </para>
+
+ <para>
+ When using libedit rather than readline, <command>\s</> printed the
+ command history in a fairly unreadable encoded format, and on recent
+ libedit versions might fail altogether. Fix that by printing the
+ history ourselves rather than having the library do it. A pleasant
+ side-effect is that the pager is used if appropriate.
+ </para>
+
+ <para>
+ This patch also fixes a bug that caused newline encoding to be applied
+ inconsistently when saving the command history with libedit.
+ Multiline history entries written by older <application>psql</>
+ versions will be read cleanly with this patch, but perhaps not
+ vice versa, depending on the exact libedit versions involved.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [28551797a] 2014-12-31 12:18:50 -0500
+Branch: REL9_4_STABLE [c35249939] 2014-12-31 12:16:57 -0500
+Branch: REL9_3_STABLE [7582cce56] 2014-12-31 12:17:00 -0500
+Branch: REL9_2_STABLE [64c506535] 2014-12-31 12:17:04 -0500
+Branch: REL9_1_STABLE [1773e0702] 2014-12-31 12:17:08 -0500
+Branch: REL9_0_STABLE [2600e4436] 2014-12-31 12:17:12 -0500
+-->
+
+ <listitem>
+ <para>
+ Improve consistency of parsing of <application>psql</>'s special
+ variables (Tom Lane)
+ </para>
+
+ <para>
+ Allow variant spellings of <literal>on</> and <literal>off</> (such
+ as <literal>1</>/<literal>0</>) for <literal>ECHO_HIDDEN</>
+ and <literal>ON_ERROR_ROLLBACK</>. Report a warning for unrecognized
+ values for <literal>COMP_KEYWORD_CASE</>, <literal>ECHO</>,
+ <literal>ECHO_HIDDEN</>, <literal>HISTCONTROL</>,
+ <literal>ON_ERROR_ROLLBACK</>, and <literal>VERBOSITY</>. Recognize
+ all values for all these variables case-insensitively; previously
+ there was a mishmash of case-sensitive and case-insensitive behaviors.
+ </para>
+ </listitem>
+
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: REL9_3_STABLE [4b1953079] 2014-11-28 02:44:40 +0900
+-->
+
+ <listitem>
+ <para>
+ Make <application>psql</>'s <command>\watch</> command display
+ nulls as specified by <command>\pset null</> (Fujii Masao)
+ </para>
+ </listitem>
+
+<!--
+Author: Stephen Frost <sfrost@snowman.net>
+Branch: REL9_3_STABLE [7ec399094] 2014-09-12 11:24:28 -0400
+Branch: REL9_2_STABLE [5d63f2159] 2014-09-12 11:24:32 -0400
+Branch: REL9_1_STABLE [4d96e93cb] 2014-09-12 11:24:36 -0400
+Branch: REL9_0_STABLE [1f89fc218] 2014-09-12 11:24:39 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix <application>psql</>'s expanded-mode display to work
+ consistently when using <literal>border</> = 3
+ and <literal>linestyle</> = <literal>ascii</> or <literal>unicode</>
+ (Stephen Frost)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [adfc157dd] 2015-01-05 19:27:04 -0500
+Branch: REL9_4_STABLE [c99e41f68] 2015-01-05 19:27:06 -0500
+Branch: REL9_3_STABLE [bb1e2426b] 2015-01-05 19:27:09 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix <application>pg_dump</> to handle comments on event triggers
+ without failing (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Kevin Grittner <kgrittn@postgresql.org>
+Branch: master [cff1bd2a3] 2015-01-30 08:57:24 -0600
+Branch: REL9_4_STABLE [cb0168528] 2015-01-30 08:57:53 -0600
+Branch: REL9_3_STABLE [cc609c46f] 2015-01-30 09:01:36 -0600
+-->
+
+ <listitem>
+ <para>
+ Allow parallel <application>pg_dump</> to
+ use <option>--serializable-deferrable</> (Kevin Grittner)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [51fc61334] 2014-07-25 19:48:48 -0400
+Branch: REL9_2_STABLE [a6c567437] 2014-07-25 19:48:51 -0400
+Branch: REL9_1_STABLE [40c333c39] 2014-07-25 19:48:54 -0400
+-->
+
+ <listitem>
+ <para>
+ Improve performance of <application>pg_dump</> when the database
+ contains many instances of multiple dependency paths between the same
+ two objects (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [9fc887121] 2014-11-13 18:19:32 -0500
+Branch: REL9_2_STABLE [3c5ce5102] 2014-11-13 18:19:35 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix <application>pg_dumpall</> to restore its ability to dump from
+ pre-8.1 servers (Gilles Darold)
+ </para>
+ </listitem>
+
+<!--
+Author: Robert Haas <rhaas@postgresql.org>
+Branch: REL9_3_STABLE [d72ecc91c] 2014-09-26 11:23:43 -0400
+Branch: REL9_2_STABLE [a97c8c393] 2014-09-26 11:25:57 -0400
+Branch: REL9_1_STABLE [d1844c21b] 2014-09-26 11:36:25 -0400
+Branch: REL9_0_STABLE [bbe3c069a] 2014-09-26 11:43:56 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix possible deadlock during parallel restore of a schema-only dump
+ (Robert Haas, Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [9a540c1ef] 2014-10-17 12:49:06 -0400
+Branch: REL9_2_STABLE [787e20bd4] 2014-10-17 12:49:09 -0400
+Branch: REL9_1_STABLE [0c9391e52] 2014-10-17 12:49:11 -0400
+Branch: REL9_0_STABLE [31021e7ba] 2014-10-17 12:49:15 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix core dump in <literal>pg_dump --binary-upgrade</> on zero-column
+ composite type (Rushabh Lathia)
+ </para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: REL9_3_STABLE [26a4e0ed7] 2014-11-15 01:21:11 +0100
+-->
+
+ <listitem>
+ <para>
+ Fix failure to fsync tables in nondefault tablespaces
+ during <application>pg_upgrade</> (Abhijit Menon-Sen, Andres Freund)
+ </para>
+
+ <para>
+ With an operating system crash and some bad luck, this could result in
+ data loss during an upgrade.
+ </para>
+ </listitem>
+
+<!--
+Author: Bruce Momjian <bruce@momjian.us>
+Branch: REL9_3_STABLE [fca9f349b] 2014-08-07 14:56:13 -0400
+-->
+
+ <listitem>
+ <para>
+ In <application>pg_upgrade</>, cope with cases where the new cluster
+ creates a TOAST table for a table that didn't previously have one
+ (Bruce Momjian)
+ </para>
+
+ <para>
+ Previously this could result in failures due to OID conflicts.
+ </para>
+ </listitem>
+
+<!--
+Author: Bruce Momjian <bruce@momjian.us>
+Branch: REL9_3_STABLE [24ae44914] 2014-08-04 11:45:45 -0400
+-->
+
+ <listitem>
+ <para>
+ In <application>pg_upgrade</>, don't try to
+ set <literal>autovacuum_multixact_freeze_max_age</> for the old cluster
+ (Bruce Momjian)
+ </para>
+
+ <para>
+ This could result in failure because not all 9.3.X versions have that
+ parameter. Fortunately, we don't actually need to set it at all.
+ </para>
+ </listitem>
+
+<!--
+Author: Bruce Momjian <bruce@momjian.us>
+Branch: REL9_3_STABLE [5724f491d] 2014-09-11 18:39:46 -0400
+-->
+
+ <listitem>
+ <para>
+ In <application>pg_upgrade</>, preserve the transaction ID epoch
+ (Bruce Momjian)
+ </para>
+
+ <para>
+ This oversight did not bother <productname>PostgreSQL</> proper,
+ but could confuse some external replication tools.
+ </para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [2c0a48589] 2015-01-03 20:54:12 +0100
+Branch: REL9_4_STABLE [90e4a2bf9] 2015-01-03 20:54:13 +0100
+Branch: REL9_3_STABLE [f6cea4502] 2015-01-03 20:54:13 +0100
+Branch: REL9_2_STABLE [f961ad479] 2015-01-03 20:54:13 +0100
+Branch: REL9_1_STABLE [2a0bfa4d6] 2015-01-03 20:54:13 +0100
+-->
+
+ <listitem>
+ <para>
+ Prevent WAL files created by <literal>pg_basebackup -x/-X</> from
+ being archived again when the standby is promoted (Andres Freund)
+ </para>
+ </listitem>
+
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: REL9_3_STABLE [9747a9898] 2014-08-02 15:19:45 +0900
+-->
+
+ <listitem>
+ <para>
+ Fix memory leak in <application>pg_receivexlog</> (Fujii Masao)
+ </para>
+ </listitem>
+
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: REL9_3_STABLE [39217ce41] 2014-08-02 14:59:10 +0900
+-->
+
+ <listitem>
+ <para>
+ Fix unintended suppression of <application>pg_receivexlog</> verbose
+ messages (Fujii Masao)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [9474c9d81] 2014-09-19 13:19:02 -0400
+Branch: REL9_2_STABLE [5ff8c2d7d] 2014-09-19 13:19:05 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix failure of <filename>contrib/auto_explain</> to print per-node
+ timing information when doing <command>EXPLAIN ANALYZE</> (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [0ad403c98] 2014-08-28 18:21:14 -0400
+Branch: REL9_2_STABLE [f3998521f] 2014-08-28 18:21:17 -0400
+Branch: REL9_1_STABLE [9807c8220] 2014-08-28 18:21:20 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix upgrade-from-unpackaged script for <filename>contrib/citext</>
+ (Tom Lane)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [f44290b7b] 2014-11-04 16:54:59 -0500
+-->
+
+ <listitem>
+ <para>
+ Avoid integer overflow and buffer overrun
+ in <filename>contrib/hstore</>'s <function>hstore_to_json()</>
+ (Heikki Linnakangas)
+ </para>
+ </listitem>
+
+<!--
+Author: Andrew Dunstan <andrew@dunslane.net>
+Branch: REL9_3_STABLE [55c880797] 2014-12-01 11:44:48 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix recognition of numbers in <function>hstore_to_json_loose()</>,
+ so that JSON numbers and strings are correctly distinguished
+ (Andrew Dunstan)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [f59c8eff7] 2014-07-22 11:45:53 -0400
+Branch: REL9_2_STABLE [810f0d2a2] 2014-07-22 11:45:57 -0400
+Branch: REL9_1_STABLE [3c5232ae8] 2014-07-22 11:46:00 -0400
+Branch: REL9_0_STABLE [9dc2a3fd0] 2014-07-22 11:46:04 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix block number checking
+ in <filename>contrib/pageinspect</>'s <function>get_raw_page()</>
+ (Tom Lane)
+ </para>
+
+ <para>
+ The incorrect checking logic could prevent access to some pages in
+ non-main relation forks.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [419de696a] 2014-11-11 17:22:38 -0500
+Branch: REL9_2_STABLE [4a9710e6e] 2014-11-11 17:22:44 -0500
+Branch: REL9_1_STABLE [4ddd9e72f] 2014-11-11 17:22:51 -0500
+Branch: REL9_0_STABLE [ef5a3b957] 2014-11-11 17:22:58 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix <filename>contrib/pgcrypto</>'s <function>pgp_sym_decrypt()</>
+ to not fail on messages whose length is 6 less than a power of 2
+ (Marko Tiikkaja)
+ </para>
+ </listitem>
+
+<!--
+Author: Robert Haas <rhaas@postgresql.org>
+Branch: REL9_3_STABLE [8cf825974] 2014-11-19 12:14:22 -0500
+Branch: REL9_2_STABLE [57ce74661] 2014-11-19 12:20:47 -0500
+Branch: REL9_1_STABLE [a855c90a7] 2014-11-19 12:26:06 -0500
+-->
+
+ <listitem>
+ <para>
+ Fix file descriptor leak in <filename>contrib/pg_test_fsync</>
+ (Jeff Janes)
+ </para>
+
+ <para>
+ This could cause failure to remove temporary files on Windows.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [37507962c] 2015-01-29 20:18:33 -0500
+Branch: REL9_4_STABLE [202621d04] 2015-01-29 20:18:37 -0500
+Branch: REL9_3_STABLE [53ae24692] 2015-01-29 20:18:40 -0500
+Branch: REL9_2_STABLE [66cc74680] 2015-01-29 20:18:42 -0500
+Branch: REL9_1_STABLE [290c2daad] 2015-01-29 20:18:44 -0500
+Branch: REL9_0_STABLE [dc9a506e6] 2015-01-29 20:18:46 -0500
+-->
+
+ <listitem>
+ <para>
+ Handle unexpected query results, especially NULLs, safely in
+ <filename>contrib/tablefunc</>'s <function>connectby()</>
+ (Michael Paquier)
+ </para>
+
+ <para>
+ <function>connectby()</> previously crashed if it encountered a NULL
+ key value. It now prints that row but doesn't recurse further.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [c393847a1] 2014-11-27 11:12:51 -0500
+Branch: REL9_2_STABLE [a1cd04cd0] 2014-11-27 11:12:55 -0500
+Branch: REL9_1_STABLE [168636a99] 2014-11-27 11:12:59 -0500
+Branch: REL9_0_STABLE [6a694bbab] 2014-11-27 11:13:03 -0500
+-->
+
+ <listitem>
+ <para>
+ Avoid a possible crash in <filename>contrib/xml2</>'s
+ <function>xslt_process()</> (Mark Simonetti)
+ </para>
+
+ <para>
+ <application>libxslt</> seems to have an undocumented dependency on
+ the order in which resources are freed; reorder our calls to avoid a
+ crash.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [0247935c7] 2014-11-05 11:34:19 -0500
+Branch: REL9_2_STABLE [0ef754cad] 2014-11-05 11:34:22 -0500
+Branch: REL9_1_STABLE [7225abf00] 2014-11-05 11:34:25 -0500
+-->
+
+ <listitem>
+ <para>
+ Mark some <filename>contrib</> I/O functions with correct volatility
+ properties (Tom Lane)
+ </para>
+
+ <para>
+ The previous over-conservative marking was immaterial in normal use,
+ but could cause optimization problems or rejection of valid index
+ expression definitions. Since the consequences are not large, we've
+ just adjusted the function definitions in the extension modules'
+ scripts, without changing version numbers.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [a59ee8819] 2015-01-30 13:05:30 -0500
+Branch: REL9_4_STABLE [70da7aeba] 2015-01-30 13:04:59 -0500
+Branch: REL9_3_STABLE [f08cf8ad9] 2015-01-30 13:05:01 -0500
+Branch: REL9_2_STABLE [a97dfdfd9] 2015-01-30 13:05:04 -0500
+Branch: REL9_1_STABLE [8f51c432c] 2015-01-30 13:05:07 -0500
+Branch: REL9_0_STABLE [7c41a32b3] 2015-01-30 13:05:09 -0500
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [8824bae87] 2014-11-18 13:28:13 -0500
+Author: Robert Haas <rhaas@postgresql.org>
+Branch: master [0b49642b9] 2015-01-15 09:26:03 -0500
+Branch: REL9_4_STABLE [7b65f194e] 2015-01-15 09:29:41 -0500
+Branch: REL9_3_STABLE [ebbef4f39] 2015-01-15 09:29:55 -0500
+Branch: REL9_2_STABLE [d452bfd1b] 2015-01-15 09:42:21 -0500
+Branch: REL9_1_STABLE [151fb75b0] 2015-01-15 09:42:33 -0500
+Branch: REL9_0_STABLE [0a67c0018] 2015-01-15 09:42:47 -0500
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master [e37d474f9] 2015-01-13 14:33:05 +0200
+Branch: REL9_4_STABLE [4ebb3494e] 2015-01-13 16:01:04 +0200
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [8cadeb792] 2015-01-04 15:44:49 +0100
+Branch: REL9_4_STABLE [7ced1b6c5] 2015-01-04 15:52:52 +0100
+Branch: REL9_3_STABLE [a68b8aec7] 2015-01-04 15:53:08 +0100
+Branch: REL9_2_STABLE [6f9b84a40] 2015-01-04 15:55:00 +0100
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [58bc4747b] 2015-01-04 15:35:46 +0100
+Branch: REL9_4_STABLE [2d8411a0a] 2015-01-04 15:35:46 +0100
+Branch: REL9_3_STABLE [d33f36f16] 2015-01-04 15:35:47 +0100
+Branch: REL9_2_STABLE [029e41afd] 2015-01-04 15:35:47 +0100
+Branch: REL9_1_STABLE [39cdf365a] 2015-01-04 15:35:47 +0100
+Branch: REL9_0_STABLE [17797e18d] 2015-01-04 15:35:48 +0100
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [0398ece4c] 2015-01-04 14:36:21 +0100
+Branch: REL9_4_STABLE [ff7d46b85] 2015-01-04 14:36:21 +0100
+Branch: REL9_3_STABLE [ec14f1601] 2015-01-04 14:36:22 +0100
+Branch: REL9_2_STABLE [f4060db11] 2015-01-04 14:36:22 +0100
+Author: Tatsuo Ishii <ishii@postgresql.org>
+Branch: master [3b5a89c48] 2014-12-30 20:33:01 +0900
+Branch: REL9_4_STABLE [458e8bc65] 2014-12-30 20:27:26 +0900
+Branch: REL9_3_STABLE [ed0e03283] 2014-12-30 20:20:56 +0900
+Branch: REL9_2_STABLE [4db7eaae0] 2014-12-30 19:59:26 +0900
+Branch: REL9_1_STABLE [4c136b0b6] 2014-12-30 19:48:53 +0900
+Branch: REL9_0_STABLE [9b74f3574] 2014-12-30 19:37:55 +0900
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [66709133c] 2014-12-16 15:35:33 -0500
+Branch: REL9_4_STABLE [383d224a0] 2014-12-16 15:35:36 -0500
+Branch: REL9_3_STABLE [53960e7eb] 2014-12-16 15:35:40 -0500
+Branch: REL9_2_STABLE [e92c67ddc] 2014-12-16 15:35:43 -0500
+Branch: REL9_1_STABLE [5c784d96a] 2014-12-16 15:35:46 -0500
+Branch: REL9_0_STABLE [a2969bd72] 2014-12-16 15:35:49 -0500
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [d38e8d30c] 2014-12-16 13:31:42 -0500
+Branch: REL9_4_STABLE [6c75384ee] 2014-12-16 13:31:57 -0500
+Branch: REL9_3_STABLE [3b750ec15] 2014-12-16 13:32:02 -0500
+Branch: REL9_2_STABLE [5b2c8f04a] 2014-12-16 13:32:15 -0500
+Branch: REL9_1_STABLE [926da211a] 2014-12-16 13:32:25 -0500
+Branch: REL9_0_STABLE [961df1853] 2014-12-16 13:32:38 -0500
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [586dd5d6a] 2015-01-24 13:05:42 -0500
+Branch: REL9_4_STABLE [d51d4ff31] 2015-01-24 13:05:45 -0500
+Branch: REL9_3_STABLE [7240f9200] 2015-01-24 13:05:49 -0500
+Branch: REL9_2_STABLE [502e5f9c3] 2015-01-24 13:05:53 -0500
+Branch: REL9_1_STABLE [b00a08859] 2015-01-24 13:05:56 -0500
+Branch: REL9_0_STABLE [3a3ee655c] 2015-01-24 13:05:58 -0500
+-->
+
+ <listitem>
+ <para>
+ Numerous cleanups of warnings from Coverity static code analyzer
+ (Andres Freund, Tatsuo Ishii, Marko Kreen, Tom Lane, Michael Paquier)
+ </para>
+
+ <para>
+ These changes are mostly cosmetic but in some cases fix corner-case
+ bugs, for example a crash rather than a proper error report after an
+ out-of-memory failure. None are believed to represent security
+ issues.
+ </para>
+ </listitem>
+
+<!--
+Author: Robert Haas <rhaas@postgresql.org>
+Branch: REL9_3_STABLE [05c0059b3] 2014-07-30 12:10:20 -0400
+-->
+
+ <listitem>
+ <para>
+ Fix setup of background workers in EXEC_BACKEND builds, eg Windows
+ (Robert Haas)
+ </para>
+ </listitem>
+
+<!--
+Author: Noah Misch <noah@leadboat.com>
+Branch: REL9_3_STABLE [07115248f] 2014-07-22 11:01:41 -0400
+Branch: REL9_2_STABLE [cec0c2182] 2014-07-22 11:01:51 -0400
+Branch: REL9_1_STABLE [81af4185a] 2014-07-22 11:02:00 -0400
+Branch: REL9_0_STABLE [4c6d0abde] 2014-07-22 11:02:25 -0400
+-->
+
+ <listitem>
+ <para>
+ Detect incompatible OpenLDAP versions during build (Noah Misch)
+ </para>
+
+ <para>
+ With OpenLDAP versions 2.4.24 through 2.4.31,
+ inclusive, <productname>PostgreSQL</> backends can crash at exit.
+ Raise a warning during <application>configure</> based on the
+ compile-time OpenLDAP version number, and test the crashing scenario
+ in the <filename>contrib/dblink</> regression test.
+ </para>
+ </listitem>
+
+<!--
+Author: Noah Misch <noah@leadboat.com>
+Branch: REL9_3_STABLE [318fe2321] 2014-08-18 23:01:04 -0400
+Branch: REL9_2_STABLE [ebd4d9cdd] 2014-08-18 23:01:09 -0400
+Branch: REL9_1_STABLE [ba72fc054] 2014-08-18 23:01:13 -0400
+Branch: REL9_0_STABLE [e6841c4d6] 2014-08-18 23:01:23 -0400
+-->
+
+ <listitem>
+ <para>
+ In non-MSVC Windows builds, ensure <filename>libpq.dll</> is installed
+ with execute permissions (Noah Misch)
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [aa719391d] 2015-01-19 23:44:19 -0500
+Branch: REL9_4_STABLE [3de9f22ac] 2015-01-19 23:44:22 -0500
+Branch: REL9_3_STABLE [1681e2f74] 2015-01-19 23:44:24 -0500
+Branch: REL9_2_STABLE [89b6a19e1] 2015-01-19 23:44:28 -0500
+Branch: REL9_1_STABLE [f4f522deb] 2015-01-19 23:44:30 -0500
+Branch: REL9_0_STABLE [338ff75fc] 2015-01-19 23:44:33 -0500
+-->
+
+ <listitem>
+ <para>
+ Make <application>pg_regress</> remove any temporary installation it
+ created upon successful exit (Tom Lane)
+ </para>
+
+ <para>
+ This results in a very substantial reduction in disk space usage
+ during <literal>make check-world</>, since that sequence involves
+ creation of numerous temporary installations.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [137e7c164] 2014-10-16 15:22:17 -0400
+Branch: REL9_2_STABLE [7c67b9365] 2014-10-16 15:22:20 -0400
+Branch: REL9_1_STABLE [2784b68b3] 2014-10-16 15:22:23 -0400
+Branch: REL9_0_STABLE [870a980aa] 2014-10-16 15:22:26 -0400
+-->
+
+ <listitem>
+ <para>
+ Support time zone abbreviations that change UTC offset from time to
+ time (Tom Lane)
+ </para>
+
+ <para>
+ Previously, <productname>PostgreSQL</> assumed that the UTC offset
+ associated with a time zone abbreviation (such as <literal>EST</>)
+ never changes in the usage of any particular locale. However this
+ assumption fails in the real world, so introduce the ability for a
+ zone abbreviation to represent a UTC offset that sometimes changes.
+ Update the zone abbreviation definition files to make use of this
+ feature in timezone locales that have changed the UTC offset of their
+ abbreviations since 1970 (according to the IANA timezone database).
+ In such timezones, <productname>PostgreSQL</> will now associate the
+ correct UTC offset with the abbreviation depending on the given date.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [9701f238b] 2014-10-03 17:44:56 -0400
+Branch: REL9_2_STABLE [d7d546bbc] 2014-10-03 17:44:59 -0400
+Branch: REL9_1_STABLE [252af79d9] 2014-10-03 17:45:03 -0400
+Branch: REL9_0_STABLE [cc7bad30c] 2014-10-03 17:45:07 -0400
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [0190f0a76] 2014-12-24 16:35:40 -0500
+Branch: REL9_2_STABLE [5c8665892] 2014-12-24 16:35:44 -0500
+Branch: REL9_1_STABLE [310597e31] 2014-12-24 16:35:48 -0500
+Branch: REL9_0_STABLE [8b70023af] 2014-12-24 16:35:54 -0500
+-->
+
+ <listitem>
+ <para>
+ Update time zone abbreviations lists (Tom Lane)
+ </para>
+
+ <para>
+ Add CST (China Standard Time) to our lists.
+ Remove references to ADT as <quote>Arabia Daylight Time</>, an
+ abbreviation that's been out of use since 2007; therefore, claiming
+ there is a conflict with <quote>Atlantic Daylight Time</> doesn't seem
+ especially helpful.
+ Fix entirely incorrect GMT offsets for CKT (Cook Islands), FJT, and FJST
+ (Fiji); we didn't even have them on the proper side of the date line.
+ </para>
+ </listitem>
+
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [8470cd473] 2015-01-30 22:46:05 -0500
+Branch: REL9_2_STABLE [c9048d353] 2015-01-30 22:46:12 -0500
+Branch: REL9_1_STABLE [cb24cd3f4] 2015-01-30 22:46:17 -0500
+Branch: REL9_0_STABLE [3553d9c6e] 2015-01-30 22:46:22 -0500
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [ab45d907b] 2014-11-17 12:08:25 -0500
+Branch: REL9_2_STABLE [b1e996035] 2014-11-17 12:08:32 -0500
+Branch: REL9_1_STABLE [b96c47a3d] 2014-11-17 12:08:39 -0500
+Branch: REL9_0_STABLE [92979576e] 2014-11-17 12:08:46 -0500
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: REL9_3_STABLE [c66199151] 2014-10-04 14:18:33 -0400
+Branch: REL9_2_STABLE [8f75d7a25] 2014-10-04 14:18:36 -0400
+Branch: REL9_1_STABLE [745723c9e] 2014-10-04 14:18:39 -0400
+Branch: REL9_0_STABLE [b6391f587] 2014-10-04 14:18:43 -0400
+-->
+
+ <listitem>
+ <para>
+ Update time zone data files to <application>tzdata</> release 2015a.
+ </para>
+
+ <para>
+ The IANA timezone database has adopted abbreviations of the form
+ <literal>A<replaceable>x</>ST</literal>/<literal>A<replaceable>x</>DT</literal>
+ for all Australian time zones, reflecting what they believe to be
+ current majority practice Down Under. These names do not conflict
+ with usage elsewhere (other than ACST for Acre Summer Time, which has
+ been in disuse since 1994). Accordingly, adopt these names into
+ our <quote>Default</> timezone abbreviation set.
+ The <quote>Australia</> abbreviation set now contains only CST, EAST,
+ EST, SAST, SAT, and WST, all of which are thought to be mostly
+ historical usage. Note that SAST has also been changed to be South
+ Africa Standard Time in the <quote>Default</> abbreviation set.
+ </para>
+
+ <para>
+ Also, add zone abbreviations SRET (Asia/Srednekolymsk) and XJT
+ (Asia/Urumqi), and use WSST/WSDT for western Samoa. Also, there were
+ DST law changes in Chile, Mexico, the Turks & Caicos Islands
+ (America/Grand_Turk), and Fiji. There is a new zone
+ Pacific/Bougainville for portions of Papua New Guinea. Also, numerous
+ corrections for historical (pre-1970) time zone data.
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ </sect2>
+ </sect1>
+
<sect1 id="release-9-3-5">
<title>Release 9.3.5</title>
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [f88300168] 2014-11-04 13:24:14 -0500
-Branch: REL9_2_STABLE [db72ad02e] 2014-11-04 13:24:17 -0500
-Branch: REL9_1_STABLE [7c6f55e9e] 2014-11-04 13:24:22 -0500
-Branch: REL9_0_STABLE [45a607d5c] 2014-11-04 13:24:26 -0500
--->
-
- <listitem>
- <para>
- Avoid possible data corruption if <command>ALTER DATABASE SET
- TABLESPACE</> is used to move a database to a new tablespace and then
- shortly later move it back to its original tablespace (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [81f0a5e38] 2014-10-29 18:12:08 -0400
-Branch: REL9_2_STABLE [40058fbce] 2014-10-29 18:12:11 -0400
-Branch: REL9_1_STABLE [6ec1c3ef8] 2014-10-29 18:12:17 -0400
-Branch: REL9_0_STABLE [9d06da58e] 2014-10-29 18:12:20 -0400
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [e65b550b3] 2014-10-30 13:03:28 -0400
-Branch: REL9_2_STABLE [38cb8687a] 2014-10-30 13:03:31 -0400
-Branch: REL9_1_STABLE [fcf0246b2] 2014-10-30 13:03:34 -0400
-Branch: REL9_0_STABLE [73f950fc8] 2014-10-30 13:03:39 -0400
--->
-
- <listitem>
- <para>
- Avoid corrupting tables when <command>ANALYZE</> inside a transaction
- is rolled back (Andres Freund, Tom Lane, Michael Paquier)
- </para>
-
- <para>
- If the transaction had earlier removed the last index, rule, or
- trigger from the table, the table would be left in a corrupted state
- with the relevant <structname>pg_class</> flags not set though they
- should be.
- </para>
- </listitem>
-
-<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [d7624e562] 2014-10-20 23:45:31 +0200
-Branch: REL9_2_STABLE [fd29810d1] 2014-10-20 23:47:00 +0200
-Branch: REL9_1_STABLE [d5fef87e9] 2014-10-20 23:47:45 +0200
--->
-
- <listitem>
- <para>
- Ensure that unlogged tables are copied correctly
- during <command>CREATE DATABASE</> or <command>ALTER DATABASE SET
- TABLESPACE</> (Pavan Deolasee, Andres Freund)
- </para>
- </listitem>
-
-<!--
-Author: Robert Haas <rhaas@postgresql.org>
-Branch: REL9_3_STABLE [e35db342a] 2014-09-22 16:19:59 -0400
--->
-
- <listitem>
- <para>
- Fix incorrect processing
- of <structname>CreateEventTrigStmt</>.<structfield>eventname</> (Petr
- Jelinek)
- </para>
-
- <para>
- This could result in misbehavior if <command>CREATE EVENT TRIGGER</>
- were executed as a prepared query, or via extended query protocol.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [2a83e0349] 2014-11-11 17:00:21 -0500
-Branch: REL9_2_STABLE [1c2f9a4f6] 2014-11-11 17:00:25 -0500
-Branch: REL9_1_STABLE [94d5d57d5] 2014-11-11 17:00:28 -0500
--->
-
- <listitem>
- <para>
- Fix <command>DROP</>'s dependency searching to correctly handle the
- case where a table column is recursively visited before its table
- (Petr Jelinek, Tom Lane)
- </para>
-
- <para>
- This case is only known to arise when an extension creates both a
- datatype and a table using that datatype. The faulty code might
- refuse a <command>DROP EXTENSION</> unless <literal>CASCADE</> is
- specified, which should not be required.
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [c480cb9d2] 2015-01-15 18:52:58 -0500
</para>
<para>
- This could manifest as <quote>ctid is NULL</> errors, or possibly
- worse things, in queries that update recently-updated rows.
+ In <literal>READ COMMITTED</> mode, queries that lock or update
+ recently-updated rows could crash as a result of this bug.
</para>
</listitem>
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [2ae8a01ca] 2014-12-11 21:02:31 -0500
-Branch: REL9_2_STABLE [cd63c57e5] 2014-12-11 21:02:34 -0500
-Branch: REL9_1_STABLE [bca39b578] 2014-12-11 21:02:38 -0500
-Branch: REL9_0_STABLE [662eebdc6] 2014-12-11 21:02:41 -0500
--->
-
- <listitem>
- <para>
- Fix planning of <command>SELECT FOR UPDATE</> when using a partial
- index on a child table (Kyotaro Horiguchi)
- </para>
-
- <para>
- In <literal>READ COMMITTED</> mode, <command>SELECT FOR UPDATE</> must
- also recheck the partial index's <literal>WHERE</> condition when
- rechecking a recently-updated row to see if it still satisfies the
- query's <literal>WHERE</> condition. This requirement was missed if the
- index belonged to an inheritance child table, so that it was possible
- to incorrectly return rows that no longer satisfy the query condition.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [f14196c35] 2014-12-11 19:37:07 -0500
-Branch: REL9_2_STABLE [deadbf4f3] 2014-12-11 19:37:10 -0500
-Branch: REL9_1_STABLE [21946ac9b] 2014-12-11 19:37:14 -0500
-Branch: REL9_0_STABLE [f5e4e92fb] 2014-12-11 19:37:17 -0500
--->
-
- <listitem>
- <para>
- Fix corner case wherein <command>SELECT FOR UPDATE</> could return a row
- twice, and possibly miss returning other rows (Tom Lane)
- </para>
-
- <para>
- In <literal>READ COMMITTED</> mode, a <command>SELECT FOR UPDATE</>
- that is scanning an inheritance tree could incorrectly return a row
- from a prior child table instead of the one it should return from a
- later child table.
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [a5cd70dcb] 2015-01-15 13:18:12 -0500
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [7a9c8cefb] 2014-08-09 13:46:42 -0400
-Branch: REL9_2_STABLE [b4dacab12] 2014-08-09 13:46:45 -0400
-Branch: REL9_1_STABLE [bbe826f21] 2014-08-09 13:46:48 -0400
-Branch: REL9_0_STABLE [4ff49746e] 2014-08-09 13:46:52 -0400
--->
-
- <listitem>
- <para>
- Reject duplicate column names in the referenced-columns list of
- a <literal>FOREIGN KEY</> declaration (David Rowley)
- </para>
-
- <para>
- This restriction is per SQL standard. Previously we did not reject
- the case explicitly, but instead the code would fail with
- bizarre-looking errors.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [6306d0712] 2014-07-22 13:30:14 -0400
--->
-
- <listitem>
- <para>
- Re-enable error for <literal>SELECT ... OFFSET -1</> (Tom Lane)
- </para>
-
- <para>
- A negative offset value has been an error since 8.4, but an
- optimization added in 9.3 accidentally turned the case into a no-op.
- Restore the expected behavior.
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [451d28081] 2015-01-30 14:44:56 -0500
</para>
<para>
- The <type>json</> type did not have the storage-ambiguity problem, but
- it did have the problem of inconsistent de-escaped textual output.
- Therefore <literal>\u0000</> will now also be rejected
- in <type>json</> values when conversion to de-escaped form is
- required. This change does not break the ability to
- store <literal>\u0000</> in <type>json</> columns so long as no
- processing is done on the values. This is exactly parallel to the
- cases in which non-ASCII Unicode escapes are allowed when the database
- encoding is not UTF8.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [8b65e0a33] 2014-08-09 18:40:34 -0400
-Branch: REL9_2_STABLE [15026ab97] 2014-08-09 18:40:38 -0400
--->
-
- <listitem>
- <para>
- Restore previous behavior of conversion of domains to JSON
- (Tom Lane)
- </para>
-
- <para>
- This change causes domains over numeric and boolean to be treated
- like their base types for purposes of conversion to JSON. It worked
- like that before 9.3.5 and 9.2.9, but was unintentionally changed
- while fixing a related problem.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [8571ecb24] 2014-12-02 15:02:43 -0500
--->
-
- <listitem>
- <para>
- Fix <function>json_agg()</> to not return extra trailing right
- brackets in its result (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [25bf13fe1] 2014-09-11 23:30:57 -0400
-Branch: REL9_2_STABLE [228ed4370] 2014-09-11 23:31:00 -0400
-Branch: REL9_1_STABLE [cf5c20b06] 2014-09-11 23:31:03 -0400
-Branch: REL9_0_STABLE [26f8a4691] 2014-09-11 23:31:06 -0400
--->
-
- <listitem>
- <para>
- Fix bugs in raising a <type>numeric</> value to a large integral power
- (Tom Lane)
- </para>
-
- <para>
- The previous code could get a wrong answer, or consume excessive
- amounts of time and memory before realizing that the answer must
- overflow.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [10b81fbdc] 2014-12-01 15:25:08 -0500
-Branch: REL9_2_STABLE [e640042e9] 2014-12-01 15:25:12 -0500
-Branch: REL9_1_STABLE [2e3cc3955] 2014-12-01 15:25:15 -0500
-Branch: REL9_0_STABLE [e6550626c] 2014-12-01 15:25:18 -0500
--->
-
- <listitem>
- <para>
- In <function>numeric_recv()</>, truncate away any fractional digits
- that would be hidden according to the value's <literal>dscale</> field
- (Tom Lane)
- </para>
-
- <para>
- A <type>numeric</> value's display scale (<literal>dscale</>) should
- never be less than the number of nonzero fractional digits; but
- apparently there's at least one broken client application that
- transmits binary <type>numeric</> values in which that's true.
- This leads to strange behavior since the extra digits are taken into
- account by arithmetic operations even though they aren't printed.
- The least risky fix seems to be to truncate away such <quote>hidden</>
- digits on receipt, so that the value is indeed what it prints as.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [bbfdf5d75] 2014-09-23 20:25:36 -0400
-Branch: REL9_2_STABLE [3359a818c] 2014-09-23 20:25:39 -0400
--->
-
- <listitem>
- <para>
- Fix incorrect search for shortest-first regular expression matches
- (Tom Lane)
- </para>
-
- <para>
- Matching would often fail when the number of allowed iterations is
- limited by a <literal>?</> quantifier or a bound expression.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [7672bbca0] 2014-07-21 22:41:27 -0400
-Branch: REL9_2_STABLE [f54d97c5e] 2014-07-21 22:41:30 -0400
-Branch: REL9_1_STABLE [124331b61] 2014-07-21 22:41:33 -0400
-Branch: REL9_0_STABLE [6e5a39c9e] 2014-07-21 22:41:36 -0400
--->
-
- <listitem>
- <para>
- Reject out-of-range numeric timezone specifications (Tom Lane)
- </para>
-
- <para>
- Simple numeric timezone specifications exceeding +/- 168 hours (one
- week) would be accepted, but could then cause null-pointer dereference
- crashes in certain operations. There's no use-case for such large UTC
- offsets, so reject them.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [1aa526f3f] 2014-10-27 10:51:29 +0200
-Branch: REL9_2_STABLE [604d94d4c] 2014-10-27 10:51:33 +0200
-Branch: REL9_1_STABLE [4d1c738d1] 2014-10-27 10:51:36 +0200
-Branch: REL9_0_STABLE [10059c2da] 2014-10-27 10:51:38 +0200
--->
-
- <listitem>
- <para>
- Fix bugs in <type>tsquery</> <literal>@></> <type>tsquery</>
- operator (Heikki Linnakangas)
- </para>
-
- <para>
- Two different terms would be considered to match if they had the same
- CRC. Also, if the second operand had more terms than the first, it
- would be assumed not to match; which is wrong since it might contain
- duplicate terms.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [385f0d98a] 2014-10-23 13:11:34 -0400
-Branch: REL9_2_STABLE [f09369da8] 2014-10-23 13:11:37 -0400
-Branch: REL9_1_STABLE [94de3a679] 2014-10-23 13:11:41 -0400
-Branch: REL9_0_STABLE [21fa26b65] 2014-10-23 13:11:45 -0400
--->
-
- <listitem>
- <para>
- Improve ispell dictionary's defenses against bad affix files (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [5a74ff373] 2014-11-06 20:52:52 -0500
-Branch: REL9_2_STABLE [0bb318595] 2014-11-06 20:52:57 -0500
-Branch: REL9_1_STABLE [076688084] 2014-11-06 20:53:02 -0500
-Branch: REL9_0_STABLE [39493e4d9] 2014-11-06 20:53:07 -0500
--->
-
- <listitem>
- <para>
- Allow more than 64K phrases in a thesaurus dictionary (David Boutin)
- </para>
-
- <para>
- The previous coding could crash on an oversize dictionary, so this was
- deemed a back-patchable bug fix rather than a feature addition.
- </para>
- </listitem>
-
-<!--
-Author: Peter Eisentraut <peter_e@gmx.net>
-Branch: master [79af9a1d2] 2015-01-06 23:06:13 -0500
-Branch: REL9_4_STABLE [6bbf75192] 2015-01-17 22:11:20 -0500
-Branch: REL9_3_STABLE [e32cb8d0e] 2015-01-17 22:13:27 -0500
-Branch: REL9_2_STABLE [c8ef5b1ac] 2015-01-17 22:14:21 -0500
-Branch: REL9_1_STABLE [c975fa471] 2015-01-17 22:37:07 -0500
-Branch: REL9_0_STABLE [cebb3f032] 2015-01-17 22:37:32 -0500
--->
-
- <listitem>
- <para>
- Fix namespace handling in <function>xpath()</> (Ali Akbar)
- </para>
-
- <para>
- Previously, the <type>xml</> value resulting from
- an <function>xpath()</> call would not have namespace declarations if
- the namespace declarations were attached to an ancestor element in the
- input <type>xml</> value, rather than to the specific element being
- returned. Propagate the ancestral declaration so that the result is
- correct by itself.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [07ab4ec4c] 2014-11-10 15:21:20 -0500
-Branch: REL9_2_STABLE [19ccaf9d4] 2014-11-10 15:21:26 -0500
--->
-
- <listitem>
- <para>
- Ensure that whole-row variables expose nonempty column names
- to functions that pay attention to column names within composite
- arguments (Tom Lane)
- </para>
-
- <para>
- In some contexts, constructs like <literal>row_to_json(tab.*)</> may
- not produce the expected column names. This is fixed properly as of
- 9.4; in older branches, just ensure that we produce some nonempty
- name. (In some cases this will be the underlying table's column name
- rather than the query-assigned alias that should theoretically be
- visible.)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [c57cdc9c1] 2014-11-22 16:01:12 -0500
-Branch: REL9_2_STABLE [906599f65] 2014-11-22 16:01:15 -0500
--->
-
- <listitem>
- <para>
- Fix mishandling of system columns,
- particularly <structfield>tableoid</>, in FDW queries (Etsuro Fujita)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: master [3d660d33a] 2015-01-30 12:30:59 -0500
-Branch: REL9_4_STABLE [b6a164e5c] 2015-01-30 12:31:08 -0500
-Branch: REL9_3_STABLE [527ff8baf] 2015-01-30 12:30:43 -0500
--->
-
- <listitem>
- <para>
- Fix assorted oversights in range-operator selectivity estimation
- (Emre Hasegeli)
- </para>
-
- <para>
- This patch fixes corner-case <quote>unexpected operator NNNN</> planner
- errors, and improves the selectivity estimates for some other cases.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [f6abf8f08] 2014-10-26 16:12:29 -0400
-Branch: REL9_2_STABLE [4586572d7] 2014-10-26 16:12:32 -0400
--->
-
- <listitem>
- <para>
- Avoid doing <literal><replaceable>indexed_column</> = ANY
- (<replaceable>array</>)</literal> as an index qualifier if that leads
- to an inferior plan (Andrew Gierth)
- </para>
-
- <para>
- In some cases, <literal>= ANY</> conditions applied to non-first index
- columns would be done as index conditions even though it would be
- better to use them as simple filter conditions.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [4e54685d0] 2014-10-20 12:23:48 -0400
--->
-
- <listitem>
- <para>
- Fix <quote>variable not found in subplan target list</> planner
- failure when an inline-able SQL function taking a composite argument
- is used in a <literal>LATERAL</> subselect and the composite argument
- is a lateral reference (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [b2b95de61] 2014-10-01 19:30:30 -0400
-Branch: REL9_2_STABLE [71b88cf52] 2014-10-01 19:30:34 -0400
-Branch: REL9_1_STABLE [477023e94] 2014-10-01 19:30:38 -0400
-Branch: REL9_0_STABLE [288f15b7c] 2014-10-01 19:30:41 -0400
--->
-
- <listitem>
- <para>
- Fix planner problems with nested append relations, such as inherited
- tables within <literal>UNION ALL</> subqueries (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: master [930fd6845] 2014-12-30 14:53:11 +0200
-Branch: REL9_4_STABLE [4e241f7cd] 2014-12-30 14:53:03 +0200
--->
-
- <listitem>
- <para>
- Revert unintended reduction in maximum size of a GIN index item
- (Heikki Linnakangas)
- </para>
-
- <para>
- 9.4.0 could fail with <quote>index row size exceeds maximum</> errors
- for data that previous versions would accept.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: master [68fa75f31] 2015-01-30 17:58:23 +0100
-Branch: REL9_4_STABLE [dc40ca696] 2015-01-30 17:59:17 +0100
--->
-
- <listitem>
- <para>
- Fix query-duration memory leak during repeated GIN index rescans
- (Heikki Linnakangas)
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: master [31ed42b9a] 2015-01-29 19:35:55 +0200
-Branch: REL9_4_STABLE [28a37deab] 2015-01-29 19:37:29 +0200
-Branch: REL9_3_STABLE [1c2774f37] 2015-01-29 19:37:27 +0200
-Branch: REL9_2_STABLE [61729e99d] 2015-01-29 19:37:25 +0200
-Branch: REL9_1_STABLE [37e0f13f2] 2015-01-29 19:37:22 +0200
--->
-
- <listitem>
- <para>
- Fix possible crash when using
- nonzero <varname>gin_fuzzy_search_limit</> (Heikki Linnakangas)
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [ef8ac584e] 2014-10-03 14:50:29 +0300
-Branch: REL9_2_STABLE [8e137b075] 2014-10-03 14:50:40 +0300
-Branch: REL9_1_STABLE [f270a1611] 2014-10-03 14:50:46 +0300
-Branch: REL9_0_STABLE [f04b112d5] 2014-10-03 14:50:58 +0300
--->
-
- <listitem>
- <para>
- Fail cleanly when a GiST index tuple doesn't fit on a page, rather
- than going into infinite recursion (Andrew Gierth)
- </para>
- </listitem>
-
-<!--
-Author: Alvaro Herrera <alvherre@alvh.no-ip.org>
-Branch: REL9_3_STABLE [67ed9d531] 2014-10-03 13:01:27 -0300
-Branch: REL9_2_STABLE [769d6815e] 2014-10-03 13:01:27 -0300
-Branch: REL9_1_STABLE [06646f52e] 2014-10-03 13:01:27 -0300
-Branch: REL9_0_STABLE [50a757698] 2014-10-03 13:01:27 -0300
--->
-
- <listitem>
- <para>
- Exempt tables that have per-table <varname>cost_limit</>
- and/or <varname>cost_delay</> settings from autovacuum's global cost
- balancing rules (Álvaro Herrera)
- </para>
-
- <para>
- The previous behavior resulted in basically ignoring these per-table
- settings, which was unintended. Now, a table having such settings
- will be vacuumed using those settings, independently of what is going
- on in other autovacuum workers. This may result in heavier total I/O
- load than before, so such settings should be re-examined for sanity.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [4cbecdaaa] 2014-07-30 14:41:49 -0400
-Branch: REL9_2_STABLE [c83aed34b] 2014-07-30 14:41:53 -0400
-Branch: REL9_1_STABLE [f21afe421] 2014-07-30 14:41:58 -0400
-Branch: REL9_0_STABLE [91b4a881c] 2014-07-30 14:42:12 -0400
--->
-
- <listitem>
- <para>
- Avoid wholesale autovacuuming when autovacuum is nominally off
- (Tom Lane)
- </para>
-
- <para>
- Even when autovacuum is nominally off, we will still launch autovacuum
- worker processes to vacuum tables that are at risk of XID wraparound.
- However, such a worker process then proceeded to vacuum all tables in
- the target database, if they met the usual thresholds for
- autovacuuming. This is at best pretty unexpected; at worst it delays
- response to the wraparound threat. Fix it so that if autovacuum is
- turned off, workers <emphasis>only</> do anti-wraparound vacuums and
- not any other work.
- </para>
- </listitem>
-
-<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: master [3fabed070] 2015-01-07 00:19:37 +0100
-Branch: REL9_4_STABLE [7da102154] 2015-01-07 00:24:58 +0100
-Author: Andres Freund <andres@anarazel.de>
-Branch: master [31912d01d] 2015-01-07 00:18:00 +0100
-Branch: REL9_4_STABLE [84911ff51] 2015-01-07 00:24:47 +0100
--->
-
- <listitem>
- <para>
- Assorted fixes for logical decoding (Andres Freund)
- </para>
- </listitem>
-
-<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [2c3ebfd1a] 2014-11-15 01:21:07 +0100
-Branch: REL9_2_STABLE [ea8b3833e] 2014-11-15 01:21:45 +0100
-Branch: REL9_1_STABLE [d85b646d0] 2014-11-15 01:22:32 +0100
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [672b43e68] 2014-11-15 01:21:02 +0100
-Branch: REL9_2_STABLE [c5baa708f] 2014-11-15 01:21:40 +0100
-Branch: REL9_1_STABLE [fde9994bc] 2014-11-15 01:22:32 +0100
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [c7299d32f] 2014-11-15 01:20:29 +0100
-Branch: REL9_2_STABLE [86673a44a] 2014-11-15 01:21:30 +0100
-Branch: REL9_1_STABLE [b0a48e996] 2014-11-15 01:22:32 +0100
--->
-
- <listitem>
- <para>
- During crash recovery, ensure that unlogged relations are rewritten as
- empty and are synced to disk before recovery is considered complete
- (Abhijit Menon-Sen, Andres Freund)
- </para>
-
- <para>
- This prevents scenarios in which unlogged relations might contain
- garbage data following database crash recovery.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [861d3aa43] 2014-11-13 20:01:55 +0200
-Branch: REL9_2_STABLE [7eab804c2] 2014-11-13 20:01:18 +0200
-Branch: REL9_1_STABLE [5f1d931cf] 2014-11-13 20:01:09 +0200
-Branch: REL9_0_STABLE [681dbe7d4] 2014-11-13 20:00:51 +0200
--->
-
- <listitem>
- <para>
- Fix race condition between hot standby queries and replaying a
- full-page image (Heikki Linnakangas)
- </para>
-
- <para>
- This mistake could result in transient errors in queries being
- executed in hot standby.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [a2a718b22] 2014-07-29 11:57:52 +0300
-Branch: REL9_2_STABLE [1578d13dc] 2014-07-29 11:58:01 +0300
-Branch: REL9_1_STABLE [de88ec6bf] 2014-07-29 11:58:09 +0300
-Branch: REL9_0_STABLE [804983961] 2014-07-29 11:58:17 +0300
--->
-
- <listitem>
- <para>
- Fix several cases where recovery logic improperly ignored WAL records
- for <literal>COMMIT/ABORT PREPARED</> (Heikki Linnakangas)
- </para>
-
- <para>
- The most notable oversight was
- that <varname>recovery_min_apply_delay</> failed to delay application
- of a two-phase commit.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: master [49b04188f] 2015-01-15 20:52:41 +0200
-Branch: REL9_4_STABLE [b337d9657] 2015-01-15 20:52:18 +0200
--->
-
- <listitem>
- <para>
- Fix incorrect replay of WAL parameter change records that report
- changes in the <varname>wal_log_hints</> setting (Petr Jalinek)
- </para>
- </listitem>
-
-<!--
-Author: Fujii Masao <fujii@postgresql.org>
-Branch: REL9_3_STABLE [d45cd9e19] 2014-10-23 16:22:46 +0900
-Branch: REL9_2_STABLE [f904e9afd] 2014-10-23 16:31:44 +0900
--->
-
- <listitem>
- <para>
- Prevent latest WAL file from being archived a second time at completion
- of crash recovery (Fujii Masao)
- </para>
- </listitem>
-
-<!--
-Author: Fujii Masao <fujii@postgresql.org>
-Branch: REL9_3_STABLE [3a3b7e316] 2014-11-06 21:25:18 +0900
-Branch: REL9_2_STABLE [38eb5d9e8] 2014-11-06 21:25:45 +0900
-Branch: REL9_1_STABLE [4e7468074] 2014-11-06 21:26:15 +0900
-Branch: REL9_0_STABLE [83c7bfb9a] 2014-11-06 21:26:21 +0900
--->
-
- <listitem>
- <para>
- Avoid creating unnecessary <filename>.ready</> marker files for
- timeline history files (Fujii Masao)
- </para>
- </listitem>
-
-<!--
-Author: Fujii Masao <fujii@postgresql.org>
-Branch: REL9_3_STABLE [52eed3d42] 2014-09-05 02:19:29 +0900
-Branch: REL9_2_STABLE [78b1228ef] 2014-09-05 02:19:45 +0900
-Branch: REL9_1_STABLE [81bffe63c] 2014-09-05 02:19:50 +0900
-Branch: REL9_0_STABLE [857a5d6b5] 2014-09-05 02:19:57 +0900
--->
-
- <listitem>
- <para>
- Fix possible null pointer dereference when an empty prepared statement
- is used and the <varname>log_statement</> setting is <literal>mod</>
- or <literal>ddl</> (Fujii Masao)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: master [75b48e1ff] 2015-01-19 23:01:33 -0500
-Branch: REL9_4_STABLE [3387cbbcb] 2015-01-19 23:01:36 -0500
-Branch: REL9_3_STABLE [19794e997] 2015-01-19 23:01:39 -0500
-Branch: REL9_2_STABLE [33b723538] 2015-01-19 23:01:41 -0500
-Branch: REL9_1_STABLE [b87c1dcef] 2015-01-19 23:01:44 -0500
-Branch: REL9_0_STABLE [a1a8d0249] 2015-01-19 23:01:46 -0500
--->
-
- <listitem>
- <para>
- Change <quote>pgstat wait timeout</> warning message to be LOG level,
- and rephrase it to be more understandable (Tom Lane)
- </para>
-
- <para>
- This message was originally thought to be essentially a can't-happen
- case, but it occurs often enough on our slower buildfarm members to be
- a nuisance. Reduce it to LOG level, and expend a bit more effort on
- the wording: it now reads <quote>using stale statistics instead of
- current ones because stats collector is not responding</>.
- </para>
- </listitem>
-
-<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [cbd9619ac] 2014-10-01 14:34:06 +0200
--->
-
- <listitem>
- <para>
- Fix possible corruption of postmaster's list of dynamic background
- workers (Andres Freund)
- </para>
- </listitem>
-
-<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [27ef6b653] 2014-09-09 23:37:33 +0200
-Branch: REL9_2_STABLE [d0b7ffc0f] 2014-09-09 23:37:50 +0200
-Branch: REL9_1_STABLE [5af508f66] 2014-09-09 23:45:07 +0200
-Branch: REL9_0_STABLE [f25e89601] 2014-09-09 23:48:03 +0200
--->
-
- <listitem>
- <para>
- Fix SPARC spinlock implementation to ensure correctness if the CPU is
- not being run in TSO coherency mode, as some non-Solaris kernels do
- (Andres Freund)
- </para>
- </listitem>
-
-<!--
-Author: Noah Misch <noah@leadboat.com>
-Branch: master [894459e59] 2015-01-07 22:35:44 -0500
-Branch: REL9_4_STABLE [83fb1ca5c] 2015-01-07 22:36:35 -0500
-Branch: REL9_3_STABLE [1a366d51e] 2015-01-07 22:40:40 -0500
-Branch: REL9_2_STABLE [5ca4e444c] 2015-01-07 22:41:49 -0500
-Branch: REL9_1_STABLE [8dc83104e] 2015-01-07 22:42:42 -0500
-Branch: REL9_0_STABLE [2e4946169] 2015-01-07 22:46:20 -0500
--->
-
- <listitem>
- <para>
- Warn if OS X's <function>setlocale()</> starts an unwanted extra
- thread inside the postmaster (Noah Misch)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: master [080eabe2e] 2015-01-11 12:35:44 -0500
-Branch: REL9_4_STABLE [733728ff3] 2015-01-11 12:35:47 -0500
--->
-
- <listitem>
- <para>
- Fix <application>libpq</>'s behavior when <filename>/etc/passwd</>
- isn't readable (Tom Lane)
- </para>
-
- <para>
- While doing <function>PQsetdbLogin()</>, <application>libpq</>
- attempts to ascertain the user's operating system name, which on most
- Unix platforms involves reading <filename>/etc/passwd</>. As of 9.4,
- failure to do that was treated as a hard error. Restore the previous
- behavior, which was to fail only if the application does not provide a
- database role name to connect as. This supports operation in chroot
- environments that lack an <filename>/etc/passwd</> file.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [08cd4d9a6] 2014-11-25 17:39:00 +0200
-Branch: REL9_2_STABLE [9b468bcec] 2014-11-25 17:39:04 +0200
-Branch: REL9_1_STABLE [95be34362] 2014-11-25 17:39:07 +0200
-Branch: REL9_0_STABLE [9880fea4f] 2014-11-25 17:39:09 +0200
--->
-
- <listitem>
- <para>
- Fix processing of repeated <literal>dbname</> parameters
- in <function>PQconnectdbParams()</> (Alex Shulgin)
- </para>
-
- <para>
- Unexpected behavior ensued if the first occurrence
- of <literal>dbname</> contained a connection string or URI to be
- expanded.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [52ef33f72] 2014-10-22 18:41:51 -0400
-Branch: REL9_2_STABLE [6fa31d8d1] 2014-10-22 18:41:54 -0400
-Branch: REL9_1_STABLE [98144378c] 2014-10-22 18:41:57 -0400
-Branch: REL9_0_STABLE [ac6e87537] 2014-10-22 18:42:01 -0400
--->
-
- <listitem>
- <para>
- Ensure that <application>libpq</> reports a suitable error message on
- unexpected socket EOF (Marko Tiikkaja, Tom Lane)
- </para>
-
- <para>
- Depending on kernel behavior, <application>libpq</> might return an
- empty error string rather than something useful when the server
- unexpectedly closed the socket.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [1325b239b] 2014-10-29 14:40:47 +0200
-Branch: REL9_2_STABLE [1a27fbd07] 2014-10-29 14:39:10 +0200
-Branch: REL9_1_STABLE [8f7bd8e91] 2014-10-29 14:36:19 +0200
-Branch: REL9_0_STABLE [49ef4eba2] 2014-10-29 14:35:39 +0200
--->
-
- <listitem>
- <para>
- Clear any old error message during <function>PQreset()</>
- (Heikki Linnakangas)
- </para>
-
- <para>
- If <function>PQreset()</> is called repeatedly, and the connection
- cannot be re-established, error messages from the failed connection
- attempts kept accumulating in the <structname>PGconn</>'s error
- string.
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [d3b162a3d] 2014-11-25 14:10:29 +0200
-Branch: REL9_2_STABLE [abcab2a64] 2014-11-25 14:10:43 +0200
-Branch: REL9_1_STABLE [5053ad206] 2014-11-25 14:10:48 +0200
-Branch: REL9_0_STABLE [1f3517039] 2014-11-25 14:10:54 +0200
--->
-
- <listitem>
- <para>
- Properly handle out-of-memory conditions while parsing connection
- options in <application>libpq</> (Alex Shulgin, Heikki Linnakangas)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [d3cfe20c6] 2014-10-06 21:23:35 -0400
-Branch: REL9_2_STABLE [b513205bc] 2014-10-06 21:23:39 -0400
-Branch: REL9_1_STABLE [037b912ec] 2014-10-06 21:23:45 -0400
-Branch: REL9_0_STABLE [d9a1e9de5] 2014-10-06 21:23:50 -0400
--->
-
- <listitem>
- <para>
- Fix array overrun in <application>ecpg</>'s version
- of <function>ParseDateTime()</> (Michael Paquier)
- </para>
- </listitem>
-
-<!--
-Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
-Branch: REL9_3_STABLE [2df66f01a] 2014-12-05 14:31:51 +0200
-Branch: REL9_2_STABLE [2b53d583d] 2014-12-05 14:31:45 +0200
-Branch: REL9_1_STABLE [729202754] 2014-12-05 14:31:37 +0200
-Branch: REL9_0_STABLE [d67be559e] 2014-12-05 14:30:55 +0200
--->
-
- <listitem>
- <para>
- In <application>initdb</>, give a clearer error message if a password
- file is specified but is empty (Mats Erik Andersson)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [b0fd5c552] 2014-09-08 16:09:52 -0400
-Branch: REL9_2_STABLE [886b58b47] 2014-09-08 16:09:56 -0400
-Branch: REL9_1_STABLE [440fcc568] 2014-09-08 16:10:01 -0400
-Branch: REL9_0_STABLE [44c518328] 2014-09-08 16:10:05 -0400
--->
-
- <listitem>
- <para>
- Fix <application>psql</>'s <command>\s</> command to work nicely with
- libedit, and add pager support (Stepan Rutz, Tom Lane)
- </para>
-
- <para>
- When using libedit rather than readline, <command>\s</> printed the
- command history in a fairly unreadable encoded format, and on recent
- libedit versions might fail altogether. Fix that by printing the
- history ourselves rather than having the library do it. A pleasant
- side-effect is that the pager is used if appropriate.
- </para>
-
- <para>
- This patch also fixes a bug that caused newline encoding to be applied
- inconsistently when saving the command history with libedit.
- Multiline history entries written by older <application>psql</>
- versions will be read cleanly with this patch, but perhaps not
- vice versa, depending on the exact libedit versions involved.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: master [28551797a] 2014-12-31 12:18:50 -0500
-Branch: REL9_4_STABLE [c35249939] 2014-12-31 12:16:57 -0500
-Branch: REL9_3_STABLE [7582cce56] 2014-12-31 12:17:00 -0500
-Branch: REL9_2_STABLE [64c506535] 2014-12-31 12:17:04 -0500
-Branch: REL9_1_STABLE [1773e0702] 2014-12-31 12:17:08 -0500
-Branch: REL9_0_STABLE [2600e4436] 2014-12-31 12:17:12 -0500
--->
-
- <listitem>
- <para>
- Improve consistency of parsing of <application>psql</>'s special
- variables (Tom Lane)
- </para>
-
- <para>
- Allow variant spellings of <literal>on</> and <literal>off</> (such
- as <literal>1</>/<literal>0</>) for <literal>ECHO_HIDDEN</>
- and <literal>ON_ERROR_ROLLBACK</>. Report a warning for unrecognized
- values for <literal>COMP_KEYWORD_CASE</>, <literal>ECHO</>,
- <literal>ECHO_HIDDEN</>, <literal>HISTCONTROL</>,
- <literal>ON_ERROR_ROLLBACK</>, and <literal>VERBOSITY</>. Recognize
- all values for all these variables case-insensitively; previously
- there was a mishmash of case-sensitive and case-insensitive behaviors.
- </para>
- </listitem>
-
-<!--
-Author: Fujii Masao <fujii@postgresql.org>
-Branch: REL9_3_STABLE [4b1953079] 2014-11-28 02:44:40 +0900
--->
-
- <listitem>
- <para>
- Make <application>psql</>'s <command>\watch</> command display
- nulls as specified by <command>\pset null</> (Fujii Masao)
- </para>
- </listitem>
-
-<!--
-Author: Stephen Frost <sfrost@snowman.net>
-Branch: REL9_3_STABLE [7ec399094] 2014-09-12 11:24:28 -0400
-Branch: REL9_2_STABLE [5d63f2159] 2014-09-12 11:24:32 -0400
-Branch: REL9_1_STABLE [4d96e93cb] 2014-09-12 11:24:36 -0400
-Branch: REL9_0_STABLE [1f89fc218] 2014-09-12 11:24:39 -0400
--->
-
- <listitem>
- <para>
- Fix <application>psql</>'s expanded-mode display to work
- consistently when using border = 3 and ascii or unicode linestyle
- (Stephen Frost)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: master [adfc157dd] 2015-01-05 19:27:04 -0500
-Branch: REL9_4_STABLE [c99e41f68] 2015-01-05 19:27:06 -0500
-Branch: REL9_3_STABLE [bb1e2426b] 2015-01-05 19:27:09 -0500
--->
-
- <listitem>
- <para>
- Fix <application>pg_dump</> to handle comments on event triggers
- without failing (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Kevin Grittner <kgrittn@postgresql.org>
-Branch: master [cff1bd2a3] 2015-01-30 08:57:24 -0600
-Branch: REL9_4_STABLE [cb0168528] 2015-01-30 08:57:53 -0600
-Branch: REL9_3_STABLE [cc609c46f] 2015-01-30 09:01:36 -0600
--->
-
- <listitem>
- <para>
- Allow parallel <application>pg_dump</> to
- use <option>--serializable-deferrable</> (Kevin Grittner)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [51fc61334] 2014-07-25 19:48:48 -0400
-Branch: REL9_2_STABLE [a6c567437] 2014-07-25 19:48:51 -0400
-Branch: REL9_1_STABLE [40c333c39] 2014-07-25 19:48:54 -0400
--->
-
- <listitem>
- <para>
- Improve performance of <application>pg_dump</> when the database
- contains many instances of multiple dependency paths between the same
- two objects (Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [9fc887121] 2014-11-13 18:19:32 -0500
-Branch: REL9_2_STABLE [3c5ce5102] 2014-11-13 18:19:35 -0500
--->
-
- <listitem>
- <para>
- Fix <application>pg_dumpall</> to restore its ability to dump from
- pre-8.1 servers (Gilles Darold)
- </para>
- </listitem>
-
-<!--
-Author: Robert Haas <rhaas@postgresql.org>
-Branch: REL9_3_STABLE [d72ecc91c] 2014-09-26 11:23:43 -0400
-Branch: REL9_2_STABLE [a97c8c393] 2014-09-26 11:25:57 -0400
-Branch: REL9_1_STABLE [d1844c21b] 2014-09-26 11:36:25 -0400
-Branch: REL9_0_STABLE [bbe3c069a] 2014-09-26 11:43:56 -0400
--->
-
- <listitem>
- <para>
- Fix possible deadlock during parallel restore of a schema-only dump
- (Robert Haas, Tom Lane)
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [9a540c1ef] 2014-10-17 12:49:06 -0400
-Branch: REL9_2_STABLE [787e20bd4] 2014-10-17 12:49:09 -0400
-Branch: REL9_1_STABLE [0c9391e52] 2014-10-17 12:49:11 -0400
-Branch: REL9_0_STABLE [31021e7ba] 2014-10-17 12:49:15 -0400
--->
-
- <listitem>
- <para>
- Fix core dump in <literal>pg_dump --binary-upgrade</> on zero-column
- composite type (Rushabh Lathia)
+ The <type>json</> type did not have the storage-ambiguity problem, but
+ it did have the problem of inconsistent de-escaped textual output.
+ Therefore <literal>\u0000</> will now also be rejected
+ in <type>json</> values when conversion to de-escaped form is
+ required. This change does not break the ability to
+ store <literal>\u0000</> in <type>json</> columns so long as no
+ processing is done on the values. This is exactly parallel to the
+ cases in which non-ASCII Unicode escapes are allowed when the database
+ encoding is not UTF8.
</para>
</listitem>
<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: REL9_3_STABLE [26a4e0ed7] 2014-11-15 01:21:11 +0100
+Author: Peter Eisentraut <peter_e@gmx.net>
+Branch: master [79af9a1d2] 2015-01-06 23:06:13 -0500
+Branch: REL9_4_STABLE [6bbf75192] 2015-01-17 22:11:20 -0500
+Branch: REL9_3_STABLE [e32cb8d0e] 2015-01-17 22:13:27 -0500
+Branch: REL9_2_STABLE [c8ef5b1ac] 2015-01-17 22:14:21 -0500
+Branch: REL9_1_STABLE [c975fa471] 2015-01-17 22:37:07 -0500
+Branch: REL9_0_STABLE [cebb3f032] 2015-01-17 22:37:32 -0500
-->
<listitem>
<para>
- Fix failure to fsync tables in nondefault tablespaces
- during <application>pg_upgrade</> (Abhijit Menon-Sen, Andres Freund)
+ Fix namespace handling in <function>xpath()</> (Ali Akbar)
</para>
<para>
- With an operating system crash and some bad luck, this could result in
- data loss during an upgrade.
+ Previously, the <type>xml</> value resulting from
+ an <function>xpath()</> call would not have namespace declarations if
+ the namespace declarations were attached to an ancestor element in the
+ input <type>xml</> value, rather than to the specific element being
+ returned. Propagate the ancestral declaration so that the result is
+ correct when considered in isolation.
</para>
</listitem>
<!--
-Author: Bruce Momjian <bruce@momjian.us>
-Branch: REL9_3_STABLE [fca9f349b] 2014-08-07 14:56:13 -0400
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [3d660d33a] 2015-01-30 12:30:59 -0500
+Branch: REL9_4_STABLE [b6a164e5c] 2015-01-30 12:31:08 -0500
+Branch: REL9_3_STABLE [527ff8baf] 2015-01-30 12:30:43 -0500
-->
<listitem>
<para>
- In <application>pg_upgrade</>, cope with cases where the new cluster
- creates a TOAST table for a table that didn't previously have one
- (Bruce Momjian)
+ Fix assorted oversights in range-operator selectivity estimation
+ (Emre Hasegeli)
</para>
<para>
- Previously this could result in failures due to OID conflicts.
+ This patch fixes corner-case <quote>unexpected operator NNNN</> planner
+ errors, and improves the selectivity estimates for some other cases.
</para>
</listitem>
<!--
-Author: Bruce Momjian <bruce@momjian.us>
-Branch: REL9_3_STABLE [24ae44914] 2014-08-04 11:45:45 -0400
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master [930fd6845] 2014-12-30 14:53:11 +0200
+Branch: REL9_4_STABLE [4e241f7cd] 2014-12-30 14:53:03 +0200
-->
<listitem>
<para>
- In <application>pg_upgrade</>, don't try to
- set <literal>autovacuum_multixact_freeze_max_age</> for the old cluster
- (Bruce Momjian)
+ Revert unintended reduction in maximum size of a GIN index item
+ (Heikki Linnakangas)
</para>
<para>
- This could result in failure because not all 9.3.X versions have that
- parameter. Fortunately, we don't actually need to set it anyway.
+ 9.4.0 could fail with <quote>index row size exceeds maximum</> errors
+ for data that previous versions would accept.
</para>
</listitem>
<!--
-Author: Bruce Momjian <bruce@momjian.us>
-Branch: REL9_3_STABLE [5724f491d] 2014-09-11 18:39:46 -0400
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master [68fa75f31] 2015-01-30 17:58:23 +0100
+Branch: REL9_4_STABLE [dc40ca696] 2015-01-30 17:59:17 +0100
-->
<listitem>
<para>
- In <application>pg_upgrade</>, preserve the transaction ID epoch
- (Bruce Momjian)
- </para>
-
- <para>
- This oversight did not bother <productname>PostgreSQL</> proper,
- but could confuse some external replication tools.
+ Fix query-duration memory leak during repeated GIN index rescans
+ (Heikki Linnakangas)
</para>
</listitem>
<!--
-Author: Andres Freund <andres@anarazel.de>
-Branch: master [2c0a48589] 2015-01-03 20:54:12 +0100
-Branch: REL9_4_STABLE [90e4a2bf9] 2015-01-03 20:54:13 +0100
-Branch: REL9_3_STABLE [f6cea4502] 2015-01-03 20:54:13 +0100
-Branch: REL9_2_STABLE [f961ad479] 2015-01-03 20:54:13 +0100
-Branch: REL9_1_STABLE [2a0bfa4d6] 2015-01-03 20:54:13 +0100
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master [31ed42b9a] 2015-01-29 19:35:55 +0200
+Branch: REL9_4_STABLE [28a37deab] 2015-01-29 19:37:29 +0200
+Branch: REL9_3_STABLE [1c2774f37] 2015-01-29 19:37:27 +0200
+Branch: REL9_2_STABLE [61729e99d] 2015-01-29 19:37:25 +0200
+Branch: REL9_1_STABLE [37e0f13f2] 2015-01-29 19:37:22 +0200
-->
<listitem>
<para>
- Prevent WAL files created by <literal>pg_basebackup -x/-X</> from
- being archived again when the standby is promoted (Andres Freund)
+ Fix possible crash when using
+ nonzero <varname>gin_fuzzy_search_limit</> (Heikki Linnakangas)
</para>
</listitem>
<!--
-Author: Fujii Masao <fujii@postgresql.org>
-Branch: REL9_3_STABLE [9747a9898] 2014-08-02 15:19:45 +0900
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [3fabed070] 2015-01-07 00:19:37 +0100
+Branch: REL9_4_STABLE [7da102154] 2015-01-07 00:24:58 +0100
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [31912d01d] 2015-01-07 00:18:00 +0100
+Branch: REL9_4_STABLE [84911ff51] 2015-01-07 00:24:47 +0100
-->
<listitem>
<para>
- Fix memory leak in <application>pg_receivexlog</> (Fujii Masao)
+ Assorted fixes for logical decoding (Andres Freund)
</para>
</listitem>
<!--
-Author: Fujii Masao <fujii@postgresql.org>
-Branch: REL9_3_STABLE [39217ce41] 2014-08-02 14:59:10 +0900
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master [49b04188f] 2015-01-15 20:52:41 +0200
+Branch: REL9_4_STABLE [b337d9657] 2015-01-15 20:52:18 +0200
-->
<listitem>
<para>
- Fix unintended suppression of <application>pg_receivexlog</> verbose
- messages (Fujii Masao)
+ Fix incorrect replay of WAL parameter change records that report
+ changes in the <varname>wal_log_hints</> setting (Petr Jalinek)
</para>
</listitem>
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [9474c9d81] 2014-09-19 13:19:02 -0400
-Branch: REL9_2_STABLE [5ff8c2d7d] 2014-09-19 13:19:05 -0400
+Branch: master [75b48e1ff] 2015-01-19 23:01:33 -0500
+Branch: REL9_4_STABLE [3387cbbcb] 2015-01-19 23:01:36 -0500
+Branch: REL9_3_STABLE [19794e997] 2015-01-19 23:01:39 -0500
+Branch: REL9_2_STABLE [33b723538] 2015-01-19 23:01:41 -0500
+Branch: REL9_1_STABLE [b87c1dcef] 2015-01-19 23:01:44 -0500
+Branch: REL9_0_STABLE [a1a8d0249] 2015-01-19 23:01:46 -0500
-->
<listitem>
<para>
- Fix failure of <filename>contrib/auto_explain</> to print per-node
- timing information when doing <command>EXPLAIN ANALYZE</> (Tom Lane)
+ Change <quote>pgstat wait timeout</> warning message to be LOG level,
+ and rephrase it to be more understandable (Tom Lane)
+ </para>
+
+ <para>
+ This message was originally thought to be essentially a can't-happen
+ case, but it occurs often enough on our slower buildfarm members to be
+ a nuisance. Reduce it to LOG level, and expend a bit more effort on
+ the wording: it now reads <quote>using stale statistics instead of
+ current ones because stats collector is not responding</>.
</para>
</listitem>
<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [0ad403c98] 2014-08-28 18:21:14 -0400
-Branch: REL9_2_STABLE [f3998521f] 2014-08-28 18:21:17 -0400
-Branch: REL9_1_STABLE [9807c8220] 2014-08-28 18:21:20 -0400
+Author: Noah Misch <noah@leadboat.com>
+Branch: master [894459e59] 2015-01-07 22:35:44 -0500
+Branch: REL9_4_STABLE [83fb1ca5c] 2015-01-07 22:36:35 -0500
+Branch: REL9_3_STABLE [1a366d51e] 2015-01-07 22:40:40 -0500
+Branch: REL9_2_STABLE [5ca4e444c] 2015-01-07 22:41:49 -0500
+Branch: REL9_1_STABLE [8dc83104e] 2015-01-07 22:42:42 -0500
+Branch: REL9_0_STABLE [2e4946169] 2015-01-07 22:46:20 -0500
-->
<listitem>
<para>
- Fix upgrade-from-unpackaged script for <filename>contrib/citext</>
- (Tom Lane)
+ Warn if OS X's <function>setlocale()</> starts an unwanted extra
+ thread inside the postmaster (Noah Misch)
</para>
</listitem>
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [f44290b7b] 2014-11-04 16:54:59 -0500
+Branch: master [080eabe2e] 2015-01-11 12:35:44 -0500
+Branch: REL9_4_STABLE [733728ff3] 2015-01-11 12:35:47 -0500
-->
<listitem>
<para>
- Avoid integer overflow and buffer overrun
- in <filename>contrib/hstore</>'s <function>hstore_to_json()</>
- (Heikki Linnakangas)
+ Fix <application>libpq</>'s behavior when <filename>/etc/passwd</>
+ isn't readable (Tom Lane)
</para>
- </listitem>
-
-<!--
-Author: Andrew Dunstan <andrew@dunslane.net>
-Branch: REL9_3_STABLE [55c880797] 2014-12-01 11:44:48 -0500
--->
- <listitem>
<para>
- Fix recognition of numbers in <function>hstore_to_json_loose()</>,
- so that JSON numbers and strings are correctly distinguished
- (Andrew Dunstan)
+ While doing <function>PQsetdbLogin()</>, <application>libpq</>
+ attempts to ascertain the user's operating system name, which on most
+ Unix platforms involves reading <filename>/etc/passwd</>. As of 9.4,
+ failure to do that was treated as a hard error. Restore the previous
+ behavior, which was to fail only if the application does not provide a
+ database role name to connect as. This supports operation in chroot
+ environments that lack an <filename>/etc/passwd</> file.
</para>
</listitem>
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [f59c8eff7] 2014-07-22 11:45:53 -0400
-Branch: REL9_2_STABLE [810f0d2a2] 2014-07-22 11:45:57 -0400
-Branch: REL9_1_STABLE [3c5232ae8] 2014-07-22 11:46:00 -0400
-Branch: REL9_0_STABLE [9dc2a3fd0] 2014-07-22 11:46:04 -0400
+Branch: master [28551797a] 2014-12-31 12:18:50 -0500
+Branch: REL9_4_STABLE [c35249939] 2014-12-31 12:16:57 -0500
+Branch: REL9_3_STABLE [7582cce56] 2014-12-31 12:17:00 -0500
+Branch: REL9_2_STABLE [64c506535] 2014-12-31 12:17:04 -0500
+Branch: REL9_1_STABLE [1773e0702] 2014-12-31 12:17:08 -0500
+Branch: REL9_0_STABLE [2600e4436] 2014-12-31 12:17:12 -0500
-->
<listitem>
<para>
- Fix block number checking
- in <filename>contrib/pageinspect</>'s <function>get_raw_page()</>
- (Tom Lane)
+ Improve consistency of parsing of <application>psql</>'s special
+ variables (Tom Lane)
</para>
<para>
- The incorrect checking logic could prevent access to some pages in
- non-main relation forks.
+ Allow variant spellings of <literal>on</> and <literal>off</> (such
+ as <literal>1</>/<literal>0</>) for <literal>ECHO_HIDDEN</>
+ and <literal>ON_ERROR_ROLLBACK</>. Report a warning for unrecognized
+ values for <literal>COMP_KEYWORD_CASE</>, <literal>ECHO</>,
+ <literal>ECHO_HIDDEN</>, <literal>HISTCONTROL</>,
+ <literal>ON_ERROR_ROLLBACK</>, and <literal>VERBOSITY</>. Recognize
+ all values for all these variables case-insensitively; previously
+ there was a mishmash of case-sensitive and case-insensitive behaviors.
</para>
</listitem>
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [419de696a] 2014-11-11 17:22:38 -0500
-Branch: REL9_2_STABLE [4a9710e6e] 2014-11-11 17:22:44 -0500
-Branch: REL9_1_STABLE [4ddd9e72f] 2014-11-11 17:22:51 -0500
-Branch: REL9_0_STABLE [ef5a3b957] 2014-11-11 17:22:58 -0500
+Branch: master [adfc157dd] 2015-01-05 19:27:04 -0500
+Branch: REL9_4_STABLE [c99e41f68] 2015-01-05 19:27:06 -0500
+Branch: REL9_3_STABLE [bb1e2426b] 2015-01-05 19:27:09 -0500
-->
<listitem>
<para>
- Fix <filename>contrib/pgcrypto</>'s <function>pgp_sym_decrypt()</>
- to not fail on messages whose length is 6 less than a power of 2
- (Marko Tiikkaja)
+ Fix <application>pg_dump</> to handle comments on event triggers
+ without failing (Tom Lane)
</para>
</listitem>
<!--
-Author: Robert Haas <rhaas@postgresql.org>
-Branch: REL9_3_STABLE [8cf825974] 2014-11-19 12:14:22 -0500
-Branch: REL9_2_STABLE [57ce74661] 2014-11-19 12:20:47 -0500
-Branch: REL9_1_STABLE [a855c90a7] 2014-11-19 12:26:06 -0500
+Author: Kevin Grittner <kgrittn@postgresql.org>
+Branch: master [cff1bd2a3] 2015-01-30 08:57:24 -0600
+Branch: REL9_4_STABLE [cb0168528] 2015-01-30 08:57:53 -0600
+Branch: REL9_3_STABLE [cc609c46f] 2015-01-30 09:01:36 -0600
-->
<listitem>
<para>
- Fix file descriptor leak in <filename>contrib/pg_test_fsync</>
- (Jeff Janes)
+ Allow parallel <application>pg_dump</> to
+ use <option>--serializable-deferrable</> (Kevin Grittner)
</para>
+ </listitem>
+
+<!--
+Author: Andres Freund <andres@anarazel.de>
+Branch: master [2c0a48589] 2015-01-03 20:54:12 +0100
+Branch: REL9_4_STABLE [90e4a2bf9] 2015-01-03 20:54:13 +0100
+Branch: REL9_3_STABLE [f6cea4502] 2015-01-03 20:54:13 +0100
+Branch: REL9_2_STABLE [f961ad479] 2015-01-03 20:54:13 +0100
+Branch: REL9_1_STABLE [2a0bfa4d6] 2015-01-03 20:54:13 +0100
+-->
+ <listitem>
<para>
- This could cause failure to remove temporary files on Windows.
+ Prevent WAL files created by <literal>pg_basebackup -x/-X</> from
+ being archived again when the standby is promoted (Andres Freund)
</para>
</listitem>
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [c393847a1] 2014-11-27 11:12:51 -0500
-Branch: REL9_2_STABLE [a1cd04cd0] 2014-11-27 11:12:55 -0500
-Branch: REL9_1_STABLE [168636a99] 2014-11-27 11:12:59 -0500
-Branch: REL9_0_STABLE [6a694bbab] 2014-11-27 11:13:03 -0500
--->
-
- <listitem>
- <para>
- Avoid a possible crash in <filename>contrib/xml2</>'s
- <function>xslt_process()</> (Mark Simonetti)
- </para>
-
- <para>
- <application>libxslt</> seems to have an undocumented dependency on
- the order in which resources are freed; reorder our calls to avoid a
- crash.
- </para>
- </listitem>
-
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [0247935c7] 2014-11-05 11:34:19 -0500
-Branch: REL9_2_STABLE [0ef754cad] 2014-11-05 11:34:22 -0500
-Branch: REL9_1_STABLE [7225abf00] 2014-11-05 11:34:25 -0500
--->
-
- <listitem>
- <para>
- Mark some <filename>contrib</> I/O functions with correct volatility
- properties (Tom Lane)
- </para>
-
- <para>
- The previous over-conservative marking was immaterial in normal use,
- but could cause optimization problems or rejection of valid index
- expression definitions. Since the consequences are not large, we've
- just adjusted the function definitions in the extension modules'
- scripts, without changing version numbers.
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [a59ee8819] 2015-01-30 13:05:30 -0500
</para>
</listitem>
-<!--
-Author: Robert Haas <rhaas@postgresql.org>
-Branch: REL9_3_STABLE [05c0059b3] 2014-07-30 12:10:20 -0400
--->
-
- <listitem>
- <para>
- Fix setup of background workers in EXEC_BACKEND builds, eg Windows
- (Robert Haas)
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [85a2a8903] 2015-01-14 11:08:13 -0500
<listitem>
<para>
Allow <varname>CFLAGS</> from <application>configure</>'s environment
- to override its automatically-supplied <varname>CFLAGS</> (Tom Lane)
+ to override automatically-supplied <varname>CFLAGS</> (Tom Lane)
</para>
<para>
</para>
</listitem>
-<!--
-Author: Noah Misch <noah@leadboat.com>
-Branch: REL9_3_STABLE [07115248f] 2014-07-22 11:01:41 -0400
-Branch: REL9_2_STABLE [cec0c2182] 2014-07-22 11:01:51 -0400
-Branch: REL9_1_STABLE [81af4185a] 2014-07-22 11:02:00 -0400
-Branch: REL9_0_STABLE [4c6d0abde] 2014-07-22 11:02:25 -0400
--->
-
- <listitem>
- <para>
- Detect incompatible OpenLDAP versions during build (Noah Misch)
- </para>
-
- <para>
- With OpenLDAP versions 2.4.24 through 2.4.31,
- inclusive, <productname>PostgreSQL</> backends can crash at exit.
- Raise a warning during <application>configure</> based on the
- compile-time OpenLDAP version number, and test the scenario in the
- dblink regression test.
- </para>
- </listitem>
-
-<!--
-Author: Noah Misch <noah@leadboat.com>
-Branch: REL9_3_STABLE [318fe2321] 2014-08-18 23:01:04 -0400
-Branch: REL9_2_STABLE [ebd4d9cdd] 2014-08-18 23:01:09 -0400
-Branch: REL9_1_STABLE [ba72fc054] 2014-08-18 23:01:13 -0400
-Branch: REL9_0_STABLE [e6841c4d6] 2014-08-18 23:01:23 -0400
--->
-
- <listitem>
- <para>
- In non-MSVC Windows builds, ensure <filename>libpq.dll</> is installed
- with execute permissions (Noah Misch)
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [aa719391d] 2015-01-19 23:44:19 -0500
<listitem>
<para>
- Make <application>pg_regress</> remove the temporary installation upon
- successful exit (Tom Lane)
+ Make <application>pg_regress</> remove any temporary installation it
+ created upon successful exit (Tom Lane)
</para>
<para>
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [137e7c164] 2014-10-16 15:22:17 -0400
-Branch: REL9_2_STABLE [7c67b9365] 2014-10-16 15:22:20 -0400
-Branch: REL9_1_STABLE [2784b68b3] 2014-10-16 15:22:23 -0400
-Branch: REL9_0_STABLE [870a980aa] 2014-10-16 15:22:26 -0400
--->
-
- <listitem>
- <para>
- Support time zone abbreviations that change UTC offset from time to
- time (Tom Lane)
- </para>
-
- <para>
- Previously, <productname>PostgreSQL</> assumed that the UTC offset
- associated with a time zone abbreviation (such as <literal>EST</>)
- never changes in the usage of any particular locale. However this
- assumption fails in the real world, so introduce the ability for a
- zone abbreviation to represent a UTC offset that sometimes changes.
- Update the zone abbreviation definition files to make use of this
- feature in timezone locales that have changed the UTC offset of their
- abbreviations since 1970 (according to the IANA timezone database).
- In such timezones, <productname>PostgreSQL</> will now associate the
- correct UTC offset with the abbreviation depending on the given date.
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [5b89473d8] 2014-12-24 16:35:23 -0500
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [9701f238b] 2014-10-03 17:44:56 -0400
-Branch: REL9_2_STABLE [d7d546bbc] 2014-10-03 17:44:59 -0400
-Branch: REL9_1_STABLE [252af79d9] 2014-10-03 17:45:03 -0400
-Branch: REL9_0_STABLE [cc7bad30c] 2014-10-03 17:45:07 -0400
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [0190f0a76] 2014-12-24 16:35:40 -0500
-Branch: REL9_2_STABLE [5c8665892] 2014-12-24 16:35:44 -0500
-Branch: REL9_1_STABLE [310597e31] 2014-12-24 16:35:48 -0500
-Branch: REL9_0_STABLE [8b70023af] 2014-12-24 16:35:54 -0500
--->
-
- <listitem>
- <para>
- Update time zone abbreviations lists (Tom Lane)
- </para>
-
- <para>
- Add CST (China Standard Time) to our lists.
- Remove references to ADT as <quote>Arabia Daylight Time</>, an
- abbreviation that's been out of use since 2007; therefore, claiming
- there is a conflict with <quote>Atlantic Daylight Time</> doesn't seem
- especially helpful.
- Fix entirely incorrect GMT offsets for CKT (Cook Islands), FJT, and FJST
- (Fiji); we didn't even have them on the proper side of the date line.
- </para>
- </listitem>
-
<!--
Author: Tom Lane <tgl@sss.pgh.pa.us>
Branch: master [08bd0c581] 2015-01-30 22:45:44 -0500
</para>
</listitem>
-<!--
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [8470cd473] 2015-01-30 22:46:05 -0500
-Branch: REL9_2_STABLE [c9048d353] 2015-01-30 22:46:12 -0500
-Branch: REL9_1_STABLE [cb24cd3f4] 2015-01-30 22:46:17 -0500
-Branch: REL9_0_STABLE [3553d9c6e] 2015-01-30 22:46:22 -0500
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [ab45d907b] 2014-11-17 12:08:25 -0500
-Branch: REL9_2_STABLE [b1e996035] 2014-11-17 12:08:32 -0500
-Branch: REL9_1_STABLE [b96c47a3d] 2014-11-17 12:08:39 -0500
-Branch: REL9_0_STABLE [92979576e] 2014-11-17 12:08:46 -0500
-Author: Tom Lane <tgl@sss.pgh.pa.us>
-Branch: REL9_3_STABLE [c66199151] 2014-10-04 14:18:33 -0400
-Branch: REL9_2_STABLE [8f75d7a25] 2014-10-04 14:18:36 -0400
-Branch: REL9_1_STABLE [745723c9e] 2014-10-04 14:18:39 -0400
-Branch: REL9_0_STABLE [b6391f587] 2014-10-04 14:18:43 -0400
--->
-
- <listitem>
- <para>
- Update time zone data files to <application>tzdata</> release 2015a.
- </para>
-
- <para>
- The IANA timezone database has adopted abbreviations of the form
- <literal>A<replaceable>x</>ST</literal>/<literal>A<replaceable>x</>DT</literal>
- for all Australian time zones, reflecting what they believe to be
- current majority practice Down Under. These names do not conflict
- with usage elsewhere (other than ACST for Acre Summer Time, which has
- been in disuse since 1994). Accordingly, adopt these names into
- our <quote>Default</> timezone abbreviation set.
- The <quote>Australia</> abbreviation set now contains only CST, EAST,
- EST, SAST, SAT, and WST, all of which are thought to be mostly
- historical usage. Note that SAST has also been changed to be South
- Africa Standard Time in the <quote>Default</> abbreviation set.
- </para>
-
- <para>
- Also, add zone abbreviations SRET (Asia/Srednekolymsk) and XJT
- (Asia/Urumqi), and use WSST/WSDT for western Samoa. Also, there were
- DST law changes in Chile, Mexico, the Turks & Caicos Islands
- (America/Grand_Turk), and Fiji. There is a new zone
- Pacific/Bougainville for portions of Papua New Guinea. Also, numerous
- corrections for historical (pre-1970) time zone data.
- </para>
- </listitem>
-
</itemizedlist>
</sect2>
projects / postgresql / commitdiff