<!-- doc/src/sgml/release-15.sgml -->
<!-- See header comment in release.sgml about typical markup -->
+ <sect1 id="release-15-14">
+ <title>Release 15.14</title>
+
+ <formalpara>
+ <title>Release date:</title>
+ <para>2025-08-14</para>
+ </formalpara>
+
+ <para>
+ This release contains a variety of fixes from 15.13.
+ For information about new features in major release 15, see
+ <xref linkend="release-15"/>.
+ </para>
+
+ <sect2>
+ <title>Migration to Version 15.14</title>
+
+ <para>
+ A dump/restore is not required for those running 15.X.
+ </para>
+
+ <para>
+ However, if you have any
+ BRIN <literal>numeric_minmax_multi_ops</literal> indexes, it is
+ advisable to reindex them after updating. See the first changelog
+ entry below.
+ </para>
+
+ <para>
+ Also, if you are upgrading from a version earlier than 15.13,
+ see <xref linkend="release-15-13"/>.
+ </para>
+ </sect2>
+
+ <sect2>
+ <title>Changes</title>
+
+ <itemizedlist>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [80c758a2e] 2025-08-05 16:51:10 -0400
+Branch: REL_18_STABLE [9b681e239] 2025-08-05 16:51:10 -0400
+Branch: REL_17_STABLE [0b0d3c19b] 2025-08-05 16:51:10 -0400
+Branch: REL_16_STABLE [b9279058a] 2025-08-05 16:51:10 -0400
+Branch: REL_15_STABLE [835c9374d] 2025-08-05 16:51:10 -0400
+Branch: REL_14_STABLE [5b0c8e328] 2025-08-05 16:51:10 -0400
+-->
+ <para>
+ Fix incorrect distance calculation in
+ BRIN <literal>numeric_minmax_multi_ops</literal> support function
+ (Peter Eisentraut, Tom Lane)
+ <ulink url="&commit_baseurl;835c9374d">§</ulink>
+ </para>
+
+ <para>
+ The results were sometimes wrong on 64-bit platforms, and wildly
+ wrong on 32-bit platforms. This did not produce obvious failures
+ because the logic is only used to choose how to merge values into
+ ranges; at worst the index would become inefficient and bloated.
+ Nonetheless it's recommended to reindex any BRIN indexes that use
+ the <literal>numeric_minmax_multi_ops</literal> operator class.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [71c0921b6] 2025-07-28 16:50:41 -0400
+Branch: REL_18_STABLE [637ead2e1] 2025-07-28 16:50:41 -0400
+Branch: REL_17_STABLE [fd4ad33fe] 2025-07-28 16:50:41 -0400
+Branch: REL_16_STABLE [6d5e493b4] 2025-07-28 16:50:42 -0400
+Branch: REL_15_STABLE [0ffbd345e] 2025-07-28 16:50:42 -0400
+Branch: REL_14_STABLE [0395464af] 2025-07-28 16:50:42 -0400
+Branch: REL_13_STABLE [589d6e640] 2025-07-28 16:50:42 -0400
+Branch: master [902f92221] 2025-07-29 12:47:38 -0400
+Branch: REL_18_STABLE [d5f014d89] 2025-07-29 12:47:19 -0400
+Branch: REL_17_STABLE [7571e0f6e] 2025-07-29 12:47:19 -0400
+Branch: REL_16_STABLE [762c6d8d2] 2025-07-29 12:47:19 -0400
+Branch: REL_15_STABLE [0928e18eb] 2025-07-29 12:47:20 -0400
+Branch: REL_14_STABLE [cdcdabce5] 2025-07-29 12:47:20 -0400
+Branch: REL_13_STABLE [0ae824704] 2025-07-29 12:47:20 -0400
+-->
+ <para>
+ Avoid regression in the size of XML input that we will accept
+ (Michael Paquier, Erik Wienhold)
+ <ulink url="&commit_baseurl;0ffbd345e">§</ulink>
+ <ulink url="&commit_baseurl;0928e18eb">§</ulink>
+ </para>
+
+ <para>
+ Our workaround for a bug in early 2.13.x releases
+ of <application>libxml2</application> made use of a code path that
+ rejects text chunks exceeding 10MB, whereas the previous coding did
+ not. Those early releases are presumably extinct in the wild by
+ now, so revert to the previous coding.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Dean Rasheed <dean.a.rasheed@gmail.com>
+Branch: master Release: REL_18_BR [b006bcd53] 2025-05-31 12:12:58 +0100
+Branch: REL_17_STABLE [ab52f6b5b] 2025-05-31 12:17:30 +0100
+Branch: REL_16_STABLE [3611794af] 2025-05-31 12:19:37 +0100
+Branch: REL_15_STABLE [d6a3f3272] 2025-05-31 12:21:02 +0100
+-->
+ <para>
+ Fix <command>MERGE</command> into a plain-inheritance parent table
+ (Dean Rasheed)
+ <ulink url="&commit_baseurl;d6a3f3272">§</ulink>
+ </para>
+
+ <para>
+ Insertions into such a target table could crash or produce incorrect
+ query results due to failing to handle <literal>WITH CHECK
+ OPTION</literal> and <literal>RETURNING</literal> actions.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Etsuro Fujita <efujita@postgresql.org>
+Branch: master [9e63f83a7] 2025-08-08 17:35:00 +0900
+Branch: REL_18_STABLE [bba6a6faf] 2025-08-08 17:35:00 +0900
+Branch: REL_17_STABLE [e028ce911] 2025-08-08 17:35:01 +0900
+Branch: REL_16_STABLE [3863c6fb6] 2025-08-08 17:35:02 +0900
+Branch: REL_15_STABLE [f39a7f32a] 2025-08-08 17:35:02 +0900
+Branch: REL_14_STABLE [73eb6afa1] 2025-08-08 17:35:03 +0900
+Branch: REL_13_STABLE [afdb4cde1] 2025-08-08 17:35:04 +0900
+-->
+ <para>
+ Allow tables with statement-level triggers to become partitions or
+ inheritance children (Etsuro Fujita)
+ <ulink url="&commit_baseurl;f39a7f32a">§</ulink>
+ </para>
+
+ <para>
+ We do not allow partitions or inheritance child tables to have
+ row-level triggers with transition tables, because an operation on
+ the whole inheritance tree would need to maintain a separate
+ transition table for each such child table. But that problem does
+ not apply for statement-level triggers, because only the parent's
+ statement-level triggers will be fired. The code that checks
+ whether an existing table can become a partition or inheritance
+ child nonetheless rejected both kinds of trigger.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Etsuro Fujita <efujita@postgresql.org>
+Branch: master [62a1211d3] 2025-08-08 10:50:00 +0900
+Branch: REL_18_STABLE [ce8817022] 2025-08-08 10:50:01 +0900
+Branch: REL_17_STABLE [9048a83c7] 2025-08-08 10:50:02 +0900
+Branch: REL_16_STABLE [9cca445df] 2025-08-08 10:50:02 +0900
+Branch: REL_15_STABLE [d642d2306] 2025-08-08 10:50:03 +0900
+Branch: REL_14_STABLE [e94fc1a8a] 2025-08-08 10:50:04 +0900
+Branch: REL_13_STABLE [b6641f7b0] 2025-08-08 10:50:05 +0900
+-->
+ <para>
+ Disallow collecting transition tuples from child foreign tables
+ (Etsuro Fujita)
+ <ulink url="&commit_baseurl;d642d2306">§</ulink>
+ </para>
+
+ <para>
+ We do not support triggers with transition tables on foreign tables.
+ However, the case of a partition or inheritance child that is a
+ foreign table was overlooked. If the parent has such a trigger,
+ incorrect transition tuples were collected from the foreign child.
+ Instead throw an error, reporting that the case is not supported.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Nathan Bossart <nathan@postgresql.org>
+Branch: master [9eb6068fb] 2025-08-01 16:52:11 -0500
+Branch: REL_18_STABLE [7b9674a8b] 2025-08-01 16:52:11 -0500
+Branch: REL_17_STABLE [39ff05636] 2025-08-01 16:52:11 -0500
+Branch: REL_16_STABLE [b998ce327] 2025-08-01 16:52:11 -0500
+Branch: REL_15_STABLE [f79ca73d7] 2025-08-01 16:52:11 -0500
+-->
+ <para>
+ Allow resetting unknown custom parameters with reserved prefixes
+ (Nathan Bossart)
+ <ulink url="&commit_baseurl;f79ca73d7">§</ulink>
+ </para>
+
+ <para>
+ Previously, if a parameter setting had been stored
+ using <command>ALTER DATABASE/ROLE/SYSTEM</command>, the stored
+ setting could not be removed if the parameter was unknown but had a
+ reserved prefix. This case could arise if an extension used to have
+ a parameter, but that parameter had been removed in an upgrade.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Amit Kapila <akapila@postgresql.org>
+Branch: master [2ab2d6f97] 2025-08-01 07:58:48 +0000
+Branch: REL_18_STABLE [d9f01a287] 2025-08-01 07:46:22 +0000
+Branch: REL_17_STABLE [8c298324a] 2025-08-01 06:53:16 +0000
+Branch: REL_16_STABLE [adfd80219] 2025-08-01 06:40:06 +0000
+Branch: REL_15_STABLE [434d2d147] 2025-08-01 07:23:37 +0000
+Branch: REL_14_STABLE [41fb3f51c] 2025-08-01 07:16:30 +0000
+-->
+ <para>
+ Fix a potential deadlock during <command>ALTER SUBSCRIPTION ... DROP
+ PUBLICATION</command> (Ajin Cherian)
+ <ulink url="&commit_baseurl;434d2d147">§</ulink>
+ </para>
+
+ <para>
+ Ensure that server processes acquire catalog locks in a consistent
+ order during replication origin drops.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [5861b1f34] 2025-06-20 13:41:11 -0400
+Branch: REL_17_STABLE [fdd826922] 2025-06-20 13:41:11 -0400
+Branch: REL_16_STABLE [1e24ea160] 2025-06-20 13:41:11 -0400
+Branch: REL_15_STABLE [75b8982ea] 2025-06-20 13:41:11 -0400
+Branch: REL_14_STABLE [27af8b9be] 2025-06-20 13:41:11 -0400
+Branch: REL_13_STABLE [4b66cb188] 2025-06-20 13:41:11 -0400
+-->
+ <para>
+ Shorten the race condition window for creating indexes with
+ conflicting names (Tom Lane)
+ <ulink url="&commit_baseurl;75b8982ea">§</ulink>
+ </para>
+
+ <para>
+ When choosing an auto-generated name for an index, avoid conflicting
+ with not-yet-committed <structname>pg_class</structname> rows as
+ well as fully-valid ones. This avoids possibly choosing the same
+ name as some concurrent <command>CREATE INDEX</command> did,
+ when that command is still in process of filling its index, or is
+ done but is part of a not-yet-committed transaction. There's still
+ a window for trouble, but it's only as long as the time needed to
+ validate a new index's parameters and insert
+ its <structname>pg_class</structname> row.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master Release: REL_18_BR [661643ded] 2025-06-25 10:03:46 +0900
+Branch: REL_17_STABLE [2e0b5d252] 2025-06-25 10:03:50 +0900
+Branch: REL_16_STABLE [d187cabdd] 2025-06-25 10:03:52 +0900
+Branch: REL_15_STABLE [354944663] 2025-06-25 10:03:53 +0900
+Branch: REL_14_STABLE [c079ba3fc] 2025-06-25 10:03:54 +0900
+Branch: REL_13_STABLE [65c3223f9] 2025-06-25 10:03:56 +0900
+-->
+ <para>
+ Prevent usage of incorrect <command>VACUUM</command> options in some
+ cases where multiple tables are vacuumed in a single command (Nathan
+ Bossart, Michael Paquier)
+ <ulink url="&commit_baseurl;354944663">§</ulink>
+ </para>
+
+ <para>
+ The <literal>TRUNCATE</literal> and <literal>INDEX_CLEANUP</literal>
+ options of one table could be applied to others.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master Release: REL_18_BR [d46911e58] 2025-05-28 08:58:40 +0900
+Branch: REL_17_STABLE [e3ffc3e91] 2025-05-28 08:59:22 +0900
+Branch: REL_16_STABLE [e9e535d61] 2025-05-28 08:59:24 +0900
+Branch: REL_15_STABLE [b3e99115e] 2025-05-28 08:59:25 +0900
+Branch: REL_14_STABLE [1fe15d25e] 2025-05-28 08:59:27 +0900
+Branch: REL_13_STABLE [9481d1614] 2025-05-28 08:59:28 +0900
+Branch: master Release: REL_18_BR [4fbb46f61] 2025-05-28 09:43:31 +0900
+Branch: REL_17_STABLE [a3c6d92f3] 2025-05-28 09:43:45 +0900
+Branch: REL_16_STABLE [52d08620e] 2025-05-28 09:43:46 +0900
+Branch: REL_15_STABLE [4dc642e75] 2025-05-28 09:43:48 +0900
+Branch: REL_14_STABLE [0c09922c0] 2025-05-28 09:43:50 +0900
+Branch: REL_13_STABLE [31ee5ec69] 2025-05-28 09:43:51 +0900
+-->
+ <para>
+ Fix processing of character classes within <literal>SIMILAR
+ TO</literal> regular expressions (Laurenz Albe)
+ <ulink url="&commit_baseurl;b3e99115e">§</ulink>
+ <ulink url="&commit_baseurl;4dc642e75">§</ulink>
+ </para>
+
+ <para>
+ The code that translates <literal>SIMILAR TO</literal> pattern
+ matching expressions to POSIX-style regular expressions did not
+ consider that square brackets can be nested. For example, in a
+ pattern like <literal>[[:alpha:]%_]</literal>, the code treated
+ the <literal>%</literal> and <literal>_</literal> characters as
+ metacharacters when they should be literals.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Heikki Linnakangas <heikki.linnakangas@iki.fi>
+Branch: master Release: REL_18_BR [29f7ce6fe] 2025-05-19 18:50:26 +0300
+Branch: REL_17_STABLE [54c05292b] 2025-05-19 18:50:47 +0300
+Branch: REL_16_STABLE [92a9ba3b9] 2025-05-19 18:50:50 +0300
+Branch: REL_15_STABLE [72fe74ca5] 2025-05-19 18:50:52 +0300
+Branch: REL_14_STABLE [0420b24fe] 2025-05-19 18:50:54 +0300
+Branch: REL_13_STABLE [7ee00918f] 2025-05-19 18:49:34 +0300
+Branch: master Release: REL_18_BR [cbf53e2b8] 2025-05-20 10:39:14 +0300
+Branch: REL_17_STABLE [a4da7b0cf] 2025-05-20 10:41:20 +0300
+Branch: REL_16_STABLE [558ea446a] 2025-05-20 10:41:50 +0300
+-->
+ <para>
+ When deparsing queries, always add parentheses around the expression
+ in <literal>FETCH FIRST <replaceable>expression</replaceable> ROWS
+ WITH TIES</literal> clauses (Heikki Linnakangas)
+ <ulink url="&commit_baseurl;72fe74ca5">§</ulink>
+ </para>
+
+ <para>
+ This avoids some cases where the deparsed result wasn't
+ syntactically valid.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Alexander Korotkov <akorotkov@postgresql.org>
+Branch: REL_18_STABLE [bae507821] 2025-07-27 15:10:02 +0300
+Branch: REL_17_STABLE [13559de95] 2025-07-27 15:10:24 +0300
+Branch: REL_16_STABLE [f0cdc2afd] 2025-07-27 15:10:29 +0300
+Branch: REL_15_STABLE [b248a3ba4] 2025-07-27 15:10:29 +0300
+Branch: REL_14_STABLE [50026136c] 2025-07-27 15:10:31 +0300
+Branch: REL_13_STABLE [f32a47161] 2025-07-27 15:10:32 +0300
+Branch: master [466c5435f] 2025-08-07 14:29:02 +0300
+Branch: REL_18_STABLE [5cfbff48a] 2025-08-07 14:31:18 +0300
+Branch: REL_17_STABLE [605890034] 2025-08-07 14:59:54 +0300
+Branch: REL_16_STABLE [2ac50f118] 2025-08-07 14:31:23 +0300
+Branch: REL_15_STABLE [73f897ba5] 2025-08-07 14:31:24 +0300
+Branch: REL_14_STABLE [c5d66fc12] 2025-08-07 14:31:25 +0300
+Branch: REL_13_STABLE [7f872ae70] 2025-08-07 14:31:26 +0300
+-->
+ <para>
+ Limit the checkpointer process's fsync request queue size (Alexander
+ Korotkov, Xuneng Zhou)
+ <ulink url="&commit_baseurl;b248a3ba4">§</ulink>
+ <ulink url="&commit_baseurl;73f897ba5">§</ulink>
+ </para>
+
+ <para>
+ With very large <varname>shared_buffers</varname> settings, it was
+ possible for the checkpointer to attempt to allocate more than 1GB
+ for fsync requests, leading to failure and an infinite loop. Clamp
+ the queue size to prevent this scenario.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Alexander Korotkov <akorotkov@postgresql.org>
+Branch: master [d3917d8f1] 2025-07-19 13:45:51 +0300
+Branch: REL_18_STABLE [5449d5b7a] 2025-07-19 13:44:30 +0300
+Branch: REL_17_STABLE [c9f4e7520] 2025-07-20 01:29:14 +0300
+Branch: REL_16_STABLE [b485e1c89] 2025-07-19 13:46:02 +0300
+Branch: REL_15_STABLE [9f270f48f] 2025-07-19 13:46:03 +0300
+Branch: REL_14_STABLE [bedfdb85b] 2025-07-19 14:13:41 +0300
+Branch: REL_13_STABLE [762f352ca] 2025-07-19 14:13:58 +0300
+-->
+ <para>
+ Avoid infinite wait in logical decoding when reading a
+ partially-written WAL record (Vignesh C)
+ <ulink url="&commit_baseurl;9f270f48f">§</ulink>
+ </para>
+
+ <para>
+ If the server crashes after writing the first part of a WAL record
+ that would span multiple pages, subsequent logical decoding of the
+ WAL stream would wait for data to arrive on the next WAL page.
+ That might never happen if the server is now idle.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [64840e462] 2025-07-11 18:50:13 -0400
+Branch: REL_18_STABLE [ccacaf4fa] 2025-07-11 18:50:13 -0400
+Branch: REL_17_STABLE [50959f96e] 2025-07-11 18:50:13 -0400
+Branch: REL_16_STABLE [53a936b61] 2025-07-11 18:50:13 -0400
+Branch: REL_15_STABLE [de73cb3ed] 2025-07-11 18:50:13 -0400
+Branch: REL_14_STABLE [ac8cdb249] 2025-07-11 18:50:13 -0400
+Branch: REL_13_STABLE [093d3d745] 2025-07-11 18:50:13 -0400
+-->
+ <para>
+ Fix inconsistent quoting of role names in ACL strings (Tom Lane)
+ <ulink url="&commit_baseurl;de73cb3ed">§</ulink>
+ </para>
+
+ <para>
+ The previous quoting rule was locale-sensitive, which could lead to
+ portability problems when transferring <type>aclitem</type> values
+ across installations. (<application>pg_dump</application> does not
+ do that, but other tools might.) To ensure consistency, always quote
+ non-ASCII characters in <type>aclitem</type> output; but to preserve
+ backward compatibility, never require that they be quoted
+ during <type>aclitem</type> input.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [aa87f69c0] 2025-06-02 15:22:44 -0400
+Branch: REL_17_STABLE [d4046125d] 2025-06-02 15:22:44 -0400
+Branch: REL_16_STABLE [ab758ec4d] 2025-06-02 15:22:44 -0400
+Branch: REL_15_STABLE [e76097124] 2025-06-02 15:22:44 -0400
+Branch: REL_14_STABLE [eb4234647] 2025-06-02 15:22:45 -0400
+Branch: REL_13_STABLE [cd31eaaeb] 2025-06-02 15:22:45 -0400
+-->
+ <para>
+ Reject equal signs (<literal>=</literal>) in the names of relation
+ options and foreign-data options (Tom Lane)
+ <ulink url="&commit_baseurl;e76097124">§</ulink>
+ </para>
+
+ <para>
+ There's no evident use-case for option names like this, and allowing
+ them creates ambiguity in the stored representation.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master [3369a3b49] 2025-07-02 13:48:36 +0900
+Branch: REL_18_STABLE [d09d13793] 2025-07-02 13:48:41 +0900
+Branch: REL_17_STABLE [074003431] 2025-07-02 13:48:43 +0900
+Branch: REL_16_STABLE [5c639523f] 2025-07-02 13:48:45 +0900
+Branch: REL_15_STABLE [d44efe87e] 2025-07-02 13:48:48 +0900
+-->
+ <para>
+ Fix potentially-incorrect decompression of LZ4-compressed archive
+ data (Mikhail Gribkov)
+ <ulink url="&commit_baseurl;d44efe87e">§</ulink>
+ </para>
+
+ <para>
+ This error seems to manifest only with not-very-compressible input
+ data, which may explain why it escaped detection.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Peter Geoghegan <pg@bowt.ie>
+Branch: master Release: REL_18_BR [7c319f549] 2025-06-11 09:17:35 -0400
+Branch: REL_17_STABLE [40aa5ddea] 2025-06-11 09:17:33 -0400
+Branch: REL_16_STABLE [c7f25feb3] 2025-06-11 09:17:31 -0400
+Branch: REL_15_STABLE [d2ec67109] 2025-06-11 09:17:29 -0400
+Branch: REL_14_STABLE [7c7c0a77d] 2025-06-11 09:17:27 -0400
+Branch: REL_13_STABLE [38c8d2987] 2025-06-11 09:17:25 -0400
+-->
+ <para>
+ Avoid a rare scenario where a btree index scan could mark the wrong
+ index entries as dead (Peter Geoghegan)
+ <ulink url="&commit_baseurl;d2ec67109">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Masahiko Sawada <msawada@postgresql.org>
+Branch: master Release: REL_18_BR [d87d07b7a] 2025-06-16 17:36:01 -0700
+Branch: REL_17_STABLE [45c357e0e] 2025-06-16 17:35:58 -0700
+Branch: REL_16_STABLE [b2ae07720] 2025-06-16 17:35:55 -0700
+Branch: REL_15_STABLE [fc0fb77c5] 2025-06-16 17:35:53 -0700
+Branch: REL_14_STABLE [983b36362] 2025-06-16 17:35:50 -0700
+Branch: REL_13_STABLE [1230be12f] 2025-06-16 17:35:48 -0700
+Branch: REL_13_STABLE [87819f766] 2025-06-24 07:07:40 -0700
+-->
+ <para>
+ Avoid re-distributing cache invalidation messages from other
+ transactions during logical replication (vignesh C)
+ <ulink url="&commit_baseurl;fc0fb77c5">§</ulink>
+ </para>
+
+ <para>
+ Our previous round of minor releases included a bug fix to ensure
+ that replication receiver processes would respond to cross-process
+ cache invalidation messages, preventing them from using stale
+ catalog data while performing replication updates. However, the fix
+ unintentionally made them also redistribute those messages again,
+ leading to an exponential increase in the number of invalidation
+ messages, which would often end in a memory allocation failure.
+ Fix by not redistributing received messages.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Alexander Korotkov <akorotkov@postgresql.org>
+Branch: REL_17_STABLE [2090edc6f] 2025-06-14 03:52:45 +0300
+Branch: REL_16_STABLE [cea8f2c3e] 2025-06-14 03:53:18 +0300
+Branch: REL_15_STABLE [dd9bc1a17] 2025-06-14 04:15:04 +0300
+Branch: REL_14_STABLE [e2832bd96] 2025-06-14 04:15:24 +0300
+Branch: REL_13_STABLE [dd3df0b85] 2025-06-14 04:15:29 +0300
+-->
+ <para>
+ Avoid premature removal of old WAL during checkpoints (Vitaly Davydov)
+ <ulink url="&commit_baseurl;dd9bc1a17">§</ulink>
+ </para>
+
+ <para>
+ If a replication slot's restart point is advanced while a checkpoint
+ is in progress, no-longer-needed WAL segments could get removed too
+ soon, leading to recovery failure if the database crashes
+ immediately afterwards. Fix by keeping them for one additional
+ checkpoint cycle.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Amit Kapila <akapila@postgresql.org>
+Branch: master Release: REL_18_BR [ad5eaf390] 2025-05-19 12:13:06 +0530
+Branch: REL_17_STABLE [7318f241d] 2025-05-19 11:55:55 +0530
+Branch: REL_16_STABLE [c0f51fde5] 2025-05-19 11:41:22 +0530
+Branch: REL_15_STABLE [9d1a62359] 2025-05-19 11:28:19 +0530
+Branch: REL_14_STABLE [e68459489] 2025-05-19 11:15:09 +0530
+Branch: REL_13_STABLE [e323d9df0] 2025-05-19 11:04:39 +0530
+-->
+ <para>
+ Never move a replication slot's confirmed-flush position backwards
+ (Shveta Malik)
+ <ulink url="&commit_baseurl;9d1a62359">§</ulink>
+ </para>
+
+ <para>
+ In some cases a replication client could acknowledge an LSN that's
+ past what it has stored persistently, and then perhaps send an older
+ LSN after a restart. We consider this not-a-bug so long as the
+ client did not have anything it needed to do for the WAL between the
+ two points. However, we should not re-send that WAL for fear of
+ data duplication, so make sure we always believe the latest
+ confirmed LSN for a given slot.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: master Release: REL_18_BR [961553daf] 2025-05-31 00:08:40 +0900
+Branch: REL_17_STABLE [24c5ad5be] 2025-05-31 00:14:07 +0900
+Branch: REL_16_STABLE [63fa7caa9] 2025-05-31 00:14:14 +0900
+Branch: REL_15_STABLE [405cca9da] 2025-05-31 00:14:22 +0900
+Branch: REL_14_STABLE [9130d8eee] 2025-05-31 00:14:29 +0900
+Branch: REL_13_STABLE [706344f06] 2025-05-31 00:14:35 +0900
+-->
+ <para>
+ Allow waiting for a transaction on a standby server to be
+ interrupted (Kevin K Biju)
+ <ulink url="&commit_baseurl;405cca9da">§</ulink>
+ </para>
+
+ <para>
+ Creation of a replication slot on a standby server may require waiting
+ for some active transaction(s) to finish on the primary and then be
+ replayed on the standby. Since that could be an indefinite wait,
+ it's desirable to allow the operation to be cancelled, but there was
+ no check for query cancel in the loop.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [02502c1bc] 2025-05-23 14:43:43 -0400
+Branch: REL_17_STABLE [cd3064f98] 2025-05-23 14:43:43 -0400
+Branch: REL_16_STABLE [e087b5b79] 2025-05-23 14:43:44 -0400
+Branch: REL_15_STABLE [13d21b48a] 2025-05-23 14:43:44 -0400
+-->
+ <para>
+ Fix per-relation memory leakage in autovacuum (Tom Lane)
+ <ulink url="&commit_baseurl;13d21b48a">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Nathan Bossart <nathan@postgresql.org>
+Branch: master Release: REL_18_BR [706054b11] 2025-05-30 15:17:28 -0500
+Branch: REL_17_STABLE [fe8ea7a2a] 2025-05-30 15:17:28 -0500
+Branch: REL_16_STABLE [24135398f] 2025-05-30 15:17:28 -0500
+Branch: REL_15_STABLE [ddfcfb7ce] 2025-05-30 15:17:28 -0500
+Branch: REL_14_STABLE [b65be6ef0] 2025-05-30 15:17:28 -0500
+Branch: REL_13_STABLE [b7ba2c030] 2025-05-30 15:17:28 -0500
+-->
+ <para>
+ Fix some places that might try to fetch toasted fields of system
+ catalogs without any snapshot (Nathan Bossart)
+ <ulink url="&commit_baseurl;ddfcfb7ce">§</ulink>
+ </para>
+
+ <para>
+ This could result in an assertion failure or <quote>cannot fetch
+ toast data without an active snapshot</quote> error.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [8319e5cb5] 2025-06-29 13:56:03 -0400
+Branch: REL_17_STABLE [bbfcbc4cd] 2025-06-29 13:56:03 -0400
+Branch: REL_16_STABLE [c15798cf9] 2025-06-29 13:56:03 -0400
+Branch: REL_15_STABLE [614ffb26d] 2025-06-29 13:56:03 -0400
+Branch: REL_14_STABLE [25cab4473] 2025-06-29 13:56:03 -0400
+Branch: REL_13_STABLE [13f1e9f26] 2025-06-29 13:56:03 -0400
+Branch: master [a10f21e6c] 2025-07-03 13:46:07 -0400
+Branch: REL_18_STABLE [3d7a96871] 2025-07-03 13:46:07 -0400
+Branch: REL_17_STABLE [6d4395b40] 2025-07-03 13:46:07 -0400
+Branch: REL_16_STABLE [d36980b71] 2025-07-03 13:46:07 -0400
+Branch: REL_15_STABLE [e6dd6e6ee] 2025-07-03 13:46:07 -0400
+Branch: REL_14_STABLE [e902f8181] 2025-07-03 13:46:07 -0400
+Branch: REL_13_STABLE [f9ba071cc] 2025-07-03 13:46:07 -0400
+-->
+ <para>
+ Avoid assertion failure during cross-table constraint updates
+ (Tom Lane, Jian He)
+ <ulink url="&commit_baseurl;614ffb26d">§</ulink>
+ <ulink url="&commit_baseurl;e6dd6e6ee">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Álvaro Herrera <alvherre@kurilemu.de>
+Branch: master [b8926a5b4] 2025-07-17 17:40:22 +0200
+Branch: REL_18_STABLE [e0d3f3cfb] 2025-07-17 17:40:22 +0200
+Branch: REL_17_STABLE [0c466f5e0] 2025-07-17 17:40:22 +0200
+Branch: REL_16_STABLE [4871c1e9c] 2025-07-17 17:40:22 +0200
+Branch: REL_15_STABLE [c2720ac60] 2025-07-17 17:40:22 +0200
+Branch: REL_14_STABLE [b9a896828] 2025-07-17 17:40:22 +0200
+Branch: REL_13_STABLE [43cd85962] 2025-07-17 17:40:22 +0200
+-->
+ <para>
+ Remove faulty assertion that a command tag must have been determined
+ by the end of <function>PortalRunMulti()</function> (Álvaro Herrera)
+ <ulink url="&commit_baseurl;c2720ac60">§</ulink>
+ </para>
+
+ <para>
+ This failed in edge cases such as an empty prepared statement.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Richard Guo <rguo@postgresql.org>
+Branch: master Release: REL_18_BR [fe29b2a1d] 2025-05-15 17:09:04 +0900
+Branch: REL_17_STABLE [2f48b4f07] 2025-05-15 17:10:45 +0900
+Branch: REL_16_STABLE [d3716d4b1] 2025-05-15 17:21:15 +0900
+Branch: REL_15_STABLE [666103090] 2025-05-15 17:26:13 +0900
+-->
+ <para>
+ Fix assertion failure in <literal>XMLTABLE</literal> parsing
+ (Richard Guo)
+ <ulink url="&commit_baseurl;666103090">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [87b05fdc7] 2025-07-07 14:33:20 -0400
+Branch: REL_18_STABLE [440c5ee20] 2025-07-07 14:33:34 -0400
+Branch: REL_17_STABLE [a553a2289] 2025-07-07 14:33:47 -0400
+Branch: REL_16_STABLE [3bbc1c4a7] 2025-07-07 14:33:56 -0400
+Branch: REL_15_STABLE [c65c36ab5] 2025-07-07 14:34:04 -0400
+Branch: REL_14_STABLE [602c91cf2] 2025-07-07 14:34:12 -0400
+Branch: REL_13_STABLE [ae693c0bf] 2025-07-07 14:34:19 -0400
+-->
+ <para>
+ Restore the ability to run PL/pgSQL expressions in parallel
+ (Dipesh Dhameliya)
+ <ulink url="&commit_baseurl;c65c36ab5">§</ulink>
+ </para>
+
+ <para>
+ PL/pgSQL's notion of an <quote>expression</quote> is very broad,
+ encompassing any SQL <command>SELECT</command> query that returns a
+ single column and no more than one row. So there are cases, for
+ example evaluation of an aggregate function, where the query
+ involves significant work and it'd be useful to run it with parallel
+ workers. This used to be possible, but a previous bug fix
+ unintentionally disabled it.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [c6f7f11d8] 2025-06-01 14:48:35 -0400
+Branch: REL_17_STABLE [7559a16e2] 2025-06-01 14:48:35 -0400
+Branch: REL_16_STABLE [5c7fd5976] 2025-06-01 14:48:35 -0400
+Branch: REL_15_STABLE [b56a92651] 2025-06-01 14:48:35 -0400
+Branch: REL_14_STABLE [31a3a15fa] 2025-06-01 14:48:35 -0400
+Branch: REL_13_STABLE [1c78d5553] 2025-06-01 14:48:35 -0400
+Branch: master Release: REL_18_BR [4672b6223] 2025-06-01 14:55:24 -0400
+Branch: REL_17_STABLE [6f724fcf8] 2025-06-01 14:55:24 -0400
+Branch: REL_16_STABLE [ecc8fd2b7] 2025-06-01 14:55:24 -0400
+Branch: REL_15_STABLE [b898bb2a7] 2025-06-01 14:55:24 -0400
+Branch: REL_14_STABLE [d4556f592] 2025-06-01 14:55:24 -0400
+Branch: REL_13_STABLE [93aca1246] 2025-06-01 14:55:24 -0400
+-->
+ <para>
+ Fix edge-case resource leaks in PL/Python error reporting (Tom Lane)
+ <ulink url="&commit_baseurl;b56a92651">§</ulink>
+ <ulink url="&commit_baseurl;b898bb2a7">§</ulink>
+ </para>
+
+ <para>
+ An out-of-memory failure while reporting an error from Python could
+ result in failure to drop reference counts on Python objects,
+ leading to session-lifespan memory leakage.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [daf9bdc47] 2025-07-17 12:46:57 -0400
+Branch: REL_18_STABLE [bfa9b25c9] 2025-07-17 12:46:57 -0400
+Branch: REL_17_STABLE [3f10d2b66] 2025-07-17 12:46:58 -0400
+Branch: REL_16_STABLE [009c20a3d] 2025-07-17 12:46:58 -0400
+Branch: REL_15_STABLE [a372a64db] 2025-07-17 12:46:58 -0400
+Branch: REL_14_STABLE [d5cba7746] 2025-07-17 12:46:59 -0400
+Branch: REL_13_STABLE [9dcd1aa81] 2025-07-17 12:46:59 -0400
+-->
+ <para>
+ Fix <application>libpq</application>'s <function>PQport()</function>
+ function to never return NULL unless the passed connection is NULL
+ (Daniele Varrazzo)
+ <ulink url="&commit_baseurl;a372a64db">§</ulink>
+ </para>
+
+ <para>
+ This is the documented behavior, but
+ recent <application>libpq</application> versions would return NULL
+ in some cases where the user had not provided a port specification.
+ Revert to our historical behavior of returning an empty string in
+ such cases. (v18 and later will return the compiled-in default port
+ number, typically <literal>"5432"</literal>, instead.)
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [d98cefe11] 2025-05-30 12:55:15 -0400
+Branch: REL_17_STABLE [8b0aa7a6b] 2025-05-30 12:55:15 -0400
+Branch: REL_16_STABLE [ca70ee6ed] 2025-05-30 12:55:15 -0400
+Branch: REL_15_STABLE [39b1d1907] 2025-05-30 12:55:15 -0400
+Branch: REL_14_STABLE [a7da7914c] 2025-05-30 12:55:15 -0400
+Branch: REL_13_STABLE [c81cdffa1] 2025-05-30 12:55:15 -0400
+-->
+ <para>
+ Avoid failure when GSSAPI authentication requires packets larger
+ than 16kB (Jacob Champion, Tom Lane)
+ <ulink url="&commit_baseurl;39b1d1907">§</ulink>
+ </para>
+
+ <para>
+ Larger authentication packets are needed for Active Directory users
+ who belong to many AD groups. This limitation manifested in
+ connection failures with unintelligible error messages,
+ typically <quote>GSSAPI context establishment error: The routine
+ must be called again to complete its function: Unknown
+ error</quote>.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [137935bd1] 2025-06-10 18:39:34 -0400
+Branch: REL_17_STABLE [30e0d9ee9] 2025-06-10 18:39:34 -0400
+Branch: REL_16_STABLE [3f37400cf] 2025-06-10 18:39:34 -0400
+Branch: REL_15_STABLE [6a4d93eda] 2025-06-10 18:39:34 -0400
+Branch: REL_14_STABLE [0703c9385] 2025-06-10 18:39:34 -0400
+Branch: REL_13_STABLE [f09fea386] 2025-06-10 18:39:34 -0400
+-->
+ <para>
+ Fix timing-dependent failures in SSL and GSSAPI data transmission
+ (Tom Lane)
+ <ulink url="&commit_baseurl;6a4d93eda">§</ulink>
+ </para>
+
+ <para>
+ When using SSL or GSSAPI encryption in non-blocking
+ mode, <application>libpq</application> sometimes failed
+ with <quote>SSL error: bad length</quote> or <quote>GSSAPI caller
+ failed to retransmit all data needing to be retried</quote>.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master [1b8bbee05] 2025-07-22 14:00:00 +0900
+Branch: REL_18_STABLE [0ded7615d] 2025-07-22 14:00:04 +0900
+Branch: REL_17_STABLE [2805e1c1e] 2025-07-22 14:00:05 +0900
+Branch: REL_16_STABLE [313d3102f] 2025-07-22 14:00:07 +0900
+Branch: REL_15_STABLE [0123922f8] 2025-07-22 14:00:08 +0900
+Branch: REL_14_STABLE [408fe659a] 2025-07-22 14:00:10 +0900
+Branch: REL_13_STABLE [c934d5673] 2025-07-22 14:00:12 +0900
+-->
+ <para>
+ Avoid null-pointer dereference during connection lookup
+ in <application>ecpg</application> applications (Aleksander
+ Alekseev)
+ <ulink url="&commit_baseurl;0123922f8">§</ulink>
+ </para>
+
+ <para>
+ The case could occur only if the application has some connections
+ that are named and some that are not.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Masahiko Sawada <msawada@postgresql.org>
+Branch: master [f5a987c0e] 2025-07-09 05:45:34 -0700
+Branch: REL_18_STABLE [765a4c94c] 2025-07-09 05:45:31 -0700
+Branch: REL_17_STABLE [c1c6169eb] 2025-07-09 05:45:28 -0700
+Branch: REL_16_STABLE [d69836b13] 2025-07-09 05:45:26 -0700
+Branch: REL_15_STABLE [e3584e457] 2025-07-09 05:45:23 -0700
+Branch: REL_14_STABLE [0514616f0] 2025-07-09 05:45:20 -0700
+-->
+ <para>
+ Improve <application>psql</application>'s tab completion
+ for <command>COPY</command> and <command>\copy</command> options
+ (Atsushi Torikoshi)
+ <ulink url="&commit_baseurl;e3584e457">§</ulink>
+ </para>
+
+ <para>
+ The same completions were offered for both <command>COPY
+ FROM</command> and <command>COPY TO</command>, although some options
+ are only valid for one case or the other. Distinguish these cases
+ to provide more accurate suggestions.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: master [12efa4897] 2025-08-03 10:49:03 +0900
+Branch: REL_18_STABLE [fee46ab4f] 2025-08-03 10:49:54 +0900
+Branch: REL_17_STABLE [398e07162] 2025-08-03 10:50:01 +0900
+Branch: REL_16_STABLE [1d3ded521] 2025-08-03 10:50:22 +0900
+Branch: REL_15_STABLE [6914a330f] 2025-08-03 10:50:59 +0900
+-->
+ <para>
+ Avoid assertion failure in <application>pgbench</application> when
+ multiple pipeline sync messages are received (Fujii Masao)
+ <ulink url="&commit_baseurl;6914a330f">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Álvaro Herrera <alvherre@kurilemu.de>
+Branch: master [0858f0f96] 2025-07-16 19:22:53 +0200
+Branch: REL_18_STABLE [dca0e9693] 2025-07-16 19:22:53 +0200
+Branch: REL_17_STABLE [d07bc7c2b] 2025-07-16 19:22:53 +0200
+Branch: REL_16_STABLE [cef998ef8] 2025-07-16 19:22:53 +0200
+Branch: REL_15_STABLE [5a261c135] 2025-07-16 19:22:53 +0200
+Branch: REL_14_STABLE [e04aca1c4] 2025-07-16 19:22:53 +0200
+Branch: REL_13_STABLE [57949cea5] 2025-07-16 19:22:53 +0200
+-->
+ <para>
+ Ensure that <application>pg_dump</application> dumps comments on
+ domain constraints in a valid order (Jian He)
+ <ulink url="&commit_baseurl;5a261c135">§</ulink>
+ </para>
+
+ <para>
+ In some cases the comment command could appear before creation of
+ the constraint.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Noah Misch <noah@leadboat.com>
+Branch: master [0decd5e89] 2025-07-31 06:37:56 -0700
+Branch: REL_18_STABLE [c0ae03384] 2025-07-31 06:37:59 -0700
+Branch: REL_17_STABLE [1ca1889ea] 2025-07-31 06:38:00 -0700
+Branch: REL_16_STABLE [0ac1581c3] 2025-07-31 06:38:00 -0700
+Branch: REL_15_STABLE [22f126da6] 2025-07-31 06:38:01 -0700
+Branch: REL_14_STABLE [7ee7c1cd3] 2025-07-31 06:38:02 -0700
+Branch: REL_13_STABLE [04bc2c42f] 2025-07-31 06:38:03 -0700
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [350e6b8ea] 2024-11-04 13:31:12 -0500
+Branch: REL_17_STABLE [5dd4957b2] 2025-07-31 06:38:00 -0700
+Branch: REL_16_STABLE [9affed263] 2025-07-31 06:38:00 -0700
+Branch: REL_15_STABLE [e99010cbd] 2025-07-31 06:38:01 -0700
+Branch: REL_14_STABLE [25388fb2c] 2025-07-31 06:38:02 -0700
+Branch: REL_13_STABLE [cc9a62c51] 2025-07-31 06:38:03 -0700
+Author: Noah Misch <noah@leadboat.com>
+Branch: REL_18_STABLE [0d2734eac] 2025-08-10 13:05:13 -0700
+Branch: REL_17_STABLE [28e7252e4] 2025-08-10 13:05:16 -0700
+Branch: REL_16_STABLE [216683296] 2025-08-10 13:05:16 -0700
+Branch: REL_15_STABLE [70637d7ae] 2025-08-10 13:05:17 -0700
+Branch: REL_14_STABLE [7846f4709] 2025-08-10 13:05:17 -0700
+Branch: REL_13_STABLE [bc05590c7] 2025-08-10 13:05:17 -0700
+-->
+ <para>
+ Ensure stable sort ordering in <application>pg_dump</application>
+ for all types of database objects (Noah Misch, Andreas Karlsson)
+ <ulink url="&commit_baseurl;22f126da6">§</ulink>
+ <ulink url="&commit_baseurl;e99010cbd">§</ulink>
+ <ulink url="&commit_baseurl;70637d7ae">§</ulink>
+ </para>
+
+ <para>
+ <application>pg_dump</application> sorts objects by their logical
+ names before performing dependency-driven reordering. This sort did
+ not account for the full unique key identifying certain object types
+ such as rules and constraints, and thus it could produce dissimilar
+ sort orders for logically-identical databases. That made it
+ difficult to compare databases by
+ diff'ing <application>pg_dump</application> output, so improve the
+ logic to ensure stable sort ordering in all cases.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Álvaro Herrera <alvherre@kurilemu.de>
+Branch: master [f295494d3] 2025-07-04 18:05:43 +0200
+Branch: REL_18_STABLE [07da2985d] 2025-07-04 18:05:43 +0200
+Branch: REL_17_STABLE [b8b2e6052] 2025-07-04 18:05:43 +0200
+Branch: REL_16_STABLE [f63e408e8] 2025-07-04 18:05:43 +0200
+Branch: REL_15_STABLE [0f4958685] 2025-07-04 18:05:43 +0200
+Branch: REL_14_STABLE [ea3386cc7] 2025-07-04 18:05:43 +0200
+Branch: REL_13_STABLE [b0f0e2221] 2025-07-04 18:05:43 +0200
+Branch: master [144ad723a] 2025-07-04 21:30:05 +0200
+Branch: REL_18_STABLE [1e007722f] 2025-07-04 21:30:05 +0200
+Branch: REL_17_STABLE [bcb8d47cd] 2025-07-04 21:30:05 +0200
+Branch: REL_16_STABLE [f943e2339] 2025-07-04 21:30:05 +0200
+Branch: REL_15_STABLE [bd48455d8] 2025-07-04 21:30:05 +0200
+Branch: REL_14_STABLE [94dc2b37b] 2025-07-04 21:30:05 +0200
+Branch: REL_13_STABLE [708469281] 2025-07-04 21:30:05 +0200
+Author: Peter Eisentraut <peter@eisentraut.org>
+Branch: master [1beda2c3c] 2025-08-07 11:48:43 +0200
+Branch: REL_18_STABLE [1084e76f3] 2025-08-07 11:59:14 +0200
+Branch: REL_17_STABLE [930e1faec] 2025-08-07 11:59:20 +0200
+Branch: REL_16_STABLE [05b367bea] 2025-08-07 11:59:26 +0200
+Branch: REL_15_STABLE [63c79a6fc] 2025-08-07 11:59:32 +0200
+Branch: REL_14_STABLE [a00eb374c] 2025-08-07 11:59:38 +0200
+Branch: REL_13_STABLE [160131033] 2025-08-07 11:59:42 +0200
+Author: Peter Eisentraut <peter@eisentraut.org>
+Branch: REL_15_STABLE [22d783350] 2025-08-08 00:18:19 +0200
+Branch: REL_14_STABLE [f9f6595e3] 2025-08-08 00:18:15 +0200
+Branch: REL_13_STABLE [7406a7d82] 2025-08-08 00:18:11 +0200
+Author: Peter Eisentraut <peter@eisentraut.org>
+Branch: REL_16_STABLE [06f444816] 2025-08-08 00:27:14 +0200
+Branch: REL_15_STABLE [a8b31b160] 2025-08-08 00:28:58 +0200
+Branch: REL_14_STABLE [ed2bb0cdd] 2025-08-08 00:29:31 +0200
+Branch: REL_13_STABLE [de7fd83cd] 2025-08-08 00:29:34 +0200
+Author: Peter Eisentraut <peter@eisentraut.org>
+Branch: REL_15_STABLE [18d2d8ae4] 2025-08-08 08:47:10 +0200
+Branch: REL_14_STABLE [bef2c2a4e] 2025-08-08 08:48:28 +0200
+Branch: REL_13_STABLE [095e83d09] 2025-08-08 08:48:36 +0200
+-->
+ <para>
+ In <application>pg_upgrade</application>, check for inconsistent
+ inherited not-null constraints (Ali Akbar)
+ <ulink url="&commit_baseurl;0f4958685">§</ulink>
+ <ulink url="&commit_baseurl;bd48455d8">§</ulink>
+ <ulink url="&commit_baseurl;63c79a6fc">§</ulink>
+ <ulink url="&commit_baseurl;22d783350">§</ulink>
+ <ulink url="&commit_baseurl;a8b31b160">§</ulink>
+ <ulink url="&commit_baseurl;18d2d8ae4">§</ulink>
+ </para>
+
+ <para>
+ <productname>PostgreSQL</productname> versions before 18 allow an
+ inherited column not-null constraint to be dropped. However, this
+ results in a schema that cannot be restored, leading to failure
+ in <application>pg_upgrade</application>. Detect such cases
+ during <application>pg_upgrade</application>'s preflight checks to
+ allow users to fix them before initiating the upgrade.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master [5a6c39b6d] 2025-07-04 15:09:24 +0900
+Branch: REL_18_STABLE [29a4b63c6] 2025-07-04 15:10:17 +0900
+Branch: REL_17_STABLE [ae20c105f] 2025-07-04 15:10:19 +0900
+Branch: REL_16_STABLE [7e7059abf] 2025-07-04 15:10:21 +0900
+Branch: REL_15_STABLE [dcbbd4331] 2025-07-04 15:10:22 +0900
+Branch: REL_14_STABLE [b61ddcaf4] 2025-07-04 15:10:24 +0900
+Branch: REL_13_STABLE [8bca4476f] 2025-07-04 15:10:25 +0900
+-->
+ <para>
+ Avoid assertion failure if <varname>track_commit_timestamp</varname>
+ is enabled during <application>initdb</application> (Hayato Kuroda,
+ Andy Fan)
+ <ulink url="&commit_baseurl;dcbbd4331">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Fujii Masao <fujii@postgresql.org>
+Branch: master Release: REL_18_BR [0bd762e81] 2025-05-21 11:55:14 +0900
+Branch: REL_17_STABLE [11efaaffa] 2025-05-21 11:56:24 +0900
+Branch: REL_16_STABLE [0d2063585] 2025-05-21 11:56:32 +0900
+Branch: REL_15_STABLE [0e0174b49] 2025-05-21 11:56:39 +0900
+-->
+ <para>
+ Fix <application>pg_waldump</application> to show information about
+ dropped statistics in <literal>PREPARE TRANSACTION</literal> WAL
+ records (Daniil Davydov)
+ <ulink url="&commit_baseurl;0e0174b49">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [470273da0] 2025-05-29 10:39:55 -0400
+Branch: REL_17_STABLE [e20b3256a] 2025-05-29 10:39:55 -0400
+Branch: REL_16_STABLE [8eef55db1] 2025-05-29 10:39:55 -0400
+Branch: REL_15_STABLE [09c9ae8f6] 2025-05-29 10:39:55 -0400
+Branch: REL_14_STABLE [2cd2222ca] 2025-05-29 10:39:55 -0400
+Branch: REL_13_STABLE [e7d3d4ed4] 2025-05-29 10:39:55 -0400
+-->
+ <para>
+ Avoid possible leak of the open connection
+ during <filename>contrib/dblink</filename> connection establishment
+ (Tom Lane)
+ <ulink url="&commit_baseurl;09c9ae8f6">§</ulink>
+ </para>
+
+ <para>
+ In the rare scenario where we hit out-of-memory while inserting the
+ new connection object into dblink's hashtable, the open connection
+ would be leaked until end of session, leaving an idle session
+ sitting on the remote server.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Robert Haas <rhaas@postgresql.org>
+Branch: master Release: REL_18_BR [016e407f4] 2025-06-06 08:18:27 -0400
+Branch: REL_17_STABLE [e4b8f925a] 2025-06-06 08:18:26 -0400
+Branch: REL_16_STABLE [169429264] 2025-06-06 08:18:23 -0400
+Branch: REL_15_STABLE [d59ff3be2] 2025-06-06 08:18:22 -0400
+Branch: REL_14_STABLE [a4b9707c4] 2025-06-06 08:18:20 -0400
+Branch: REL_13_STABLE [4adbaa36c] 2025-06-06 08:18:15 -0400
+-->
+ <para>
+ Make <filename>contrib/pg_prewarm</filename> cope with very
+ large <varname>shared_buffers</varname> settings (Daria Shanina)
+ <ulink url="&commit_baseurl;d59ff3be2">§</ulink>
+ </para>
+
+ <para>
+ Autoprewarm failed with a memory allocation error
+ if <varname>shared_buffers</varname> was larger than about 50
+ million buffers (400GB).
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master Release: REL_18_BR [35a428f30] 2025-05-29 11:26:03 +0900
+Branch: REL_17_STABLE [290e8ab32] 2025-05-29 11:26:23 +0900
+Branch: REL_16_STABLE [7e8b44f4e] 2025-05-29 11:26:27 +0900
+Branch: REL_15_STABLE [130300a15] 2025-05-29 11:26:29 +0900
+Branch: REL_14_STABLE [8a1459f62] 2025-05-29 11:26:31 +0900
+Branch: REL_13_STABLE [3c03b8cd7] 2025-05-29 11:26:34 +0900
+-->
+ <para>
+ In <filename>contrib/pg_stat_statements</filename>, avoid leaving
+ gaps in the set of parameter numbers used in a normalized query
+ (Sami Imseih)
+ <ulink url="&commit_baseurl;130300a15">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [232d8caea] 2025-05-30 13:45:41 -0400
+Branch: REL_17_STABLE [9339c85af] 2025-05-30 13:45:41 -0400
+Branch: REL_16_STABLE [2b92dc4ee] 2025-05-30 13:45:41 -0400
+Branch: REL_15_STABLE [3c31594f5] 2025-05-30 13:45:41 -0400
+Branch: REL_14_STABLE [4a07c0961] 2025-05-30 13:45:41 -0400
+Branch: REL_13_STABLE [271cb7eaa] 2025-05-30 13:45:41 -0400
+-->
+ <para>
+ Fix memory leakage in <filename>contrib/postgres_fdw</filename>'s
+ DirectModify methods (Tom Lane)
+ <ulink url="&commit_baseurl;3c31594f5">§</ulink>
+ </para>
+
+ <para>
+ The <structname>PGresult</structname> holding the results of the
+ remote modify command would be leaked for the rest of the session if
+ the query fails between invocations of the DirectModify methods,
+ which could happen when there's <literal>RETURNING</literal> data to
+ process.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [4300d8b6a] 2025-07-29 15:17:40 -0400
+Branch: REL_18_STABLE [8e5e3ff55] 2025-07-29 15:17:40 -0400
+Branch: REL_17_STABLE [a644f5fc6] 2025-07-29 15:17:41 -0400
+Branch: REL_16_STABLE [bbc20c8a9] 2025-07-29 15:17:41 -0400
+Branch: REL_15_STABLE [19857437b] 2025-07-29 15:17:41 -0400
+Branch: REL_14_STABLE [2dee95bd0] 2025-07-29 15:17:41 -0400
+Branch: REL_13_STABLE [c5bd803e5] 2025-07-29 15:17:41 -0400
+-->
+ <para>
+ Ensure that directories listed
+ in <application>configure</application>'s
+ <option>--with-includes</option>
+ and <option>--with-libraries</option> options are searched before
+ system-supplied directories (Tom Lane)
+ <ulink url="&commit_baseurl;19857437b">§</ulink>
+ </para>
+
+ <para>
+ A common reason for using these options is to allow a user-built
+ version of some library to override the system-supplied version.
+ However, that failed to work in some environments because of
+ careless ordering of switches in the commands issued by the makefiles.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master [1a5212775] 2025-07-30 11:55:42 +0900
+Branch: REL_18_STABLE [cd2d52cc6] 2025-07-30 11:55:46 +0900
+Branch: REL_17_STABLE [8de56323c] 2025-07-30 11:55:47 +0900
+Branch: REL_16_STABLE [c1984be23] 2025-07-30 11:55:49 +0900
+Branch: REL_15_STABLE [d6ffc43f9] 2025-07-30 11:55:51 +0900
+Branch: REL_14_STABLE [60953d4cb] 2025-07-30 11:55:53 +0900
+Branch: REL_13_STABLE [612f5b806] 2025-07-30 11:55:54 +0900
+-->
+ <para>
+ Fix <application>configure</application>'s checks
+ for <function>__cpuid()</function>
+ and <function>__cpuidex()</function> (Lukas Fittl, Michael Paquier)
+ <ulink url="&commit_baseurl;d6ffc43f9">§</ulink>
+ </para>
+
+ <para>
+ <application>configure</application> failed to detect these
+ Windows-specific functions, so that they would not be used,
+ leading to slower-than-necessary CRC computations since the
+ availability of hardware instructions could not be verified.
+ The practical impact of this error was limited, because production
+ builds for Windows typically do not use the Autoconf toolchain.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [e6dfd068e] 2025-07-23 15:44:29 -0400
+Branch: REL_18_STABLE [3d039b53a] 2025-07-23 15:44:29 -0400
+Branch: REL_17_STABLE [635a85627] 2025-07-23 15:44:29 -0400
+Branch: REL_16_STABLE [e4d585455] 2025-07-23 15:44:29 -0400
+Branch: REL_15_STABLE [b252ce311] 2025-07-23 15:44:29 -0400
+Branch: REL_14_STABLE [868b39a54] 2025-07-23 15:44:29 -0400
+Branch: REL_13_STABLE [1ccb3851d] 2025-07-23 15:44:29 -0400
+-->
+ <para>
+ Fix build failure with <option>--with-pam</option> option on
+ Solaris-based platforms (Tom Lane)
+ <ulink url="&commit_baseurl;b252ce311">§</ulink>
+ </para>
+
+ <para>
+ Solaris is inconsistent with other Unix platforms about the API for
+ PAM authentication. This manifested as an <quote>inconsistent
+ pointer</quote> compiler warning, which we never did anything about.
+ But as of GCC 14 it's an error not warning by default, so fix it.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [1fd772d19] 2025-07-01 12:40:35 -0400
+Branch: REL_18_STABLE [581305a46] 2025-07-01 12:40:35 -0400
+Branch: REL_17_STABLE [0991249d7] 2025-07-01 12:40:35 -0400
+Branch: REL_16_STABLE [d25d392e8] 2025-07-01 12:40:35 -0400
+Branch: master [29213636e] 2025-07-01 12:08:20 -0400
+Branch: REL_18_STABLE [45c527662] 2025-07-01 12:08:20 -0400
+Branch: REL_17_STABLE [29c54ea7b] 2025-07-01 12:08:20 -0400
+Branch: REL_16_STABLE [3a2617e4f] 2025-07-01 12:08:20 -0400
+Branch: REL_15_STABLE [0fb496c70] 2025-07-01 12:08:57 -0400
+Branch: REL_14_STABLE [71d71ac4d] 2025-07-01 12:08:57 -0400
+Branch: REL_13_STABLE [d0a695cf4] 2025-07-01 12:08:58 -0400
+-->
+ <para>
+ Make our code portable to GNU Hurd (Michael Banck, Christoph Berg,
+ Samuel Thibault)
+ <ulink url="&commit_baseurl;0fb496c70">§</ulink>
+ </para>
+
+ <para>
+ Fix assumptions about <literal>IOV_MAX</literal>
+ and <literal>O_RDONLY</literal> that don't hold on Hurd.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master Release: REL_18_BR [12eee85e5] 2025-05-18 12:45:55 -0400
+Branch: REL_17_STABLE [5355a2400] 2025-05-18 12:45:55 -0400
+Branch: REL_16_STABLE [253cf661c] 2025-05-18 12:45:55 -0400
+Branch: REL_16_STABLE [4b53cb493] 2025-05-20 10:55:05 -0400
+Branch: REL_15_STABLE [00652b3c9] 2025-05-18 12:45:55 -0400
+Branch: REL_14_STABLE [a1e0e7076] 2025-05-18 12:45:55 -0400
+Branch: REL_13_STABLE [9bc2d37cc] 2025-05-18 12:45:55 -0400
+-->
+ <para>
+ Make our usage of <function>memset_s()</function> conform strictly
+ to the C11 standard (Tom Lane)
+ <ulink url="&commit_baseurl;00652b3c9">§</ulink>
+ </para>
+
+ <para>
+ This avoids compile failures on some platforms.
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Tom Lane <tgl@sss.pgh.pa.us>
+Branch: master [aad1617b7] 2025-07-15 18:11:18 -0400
+Branch: REL_18_STABLE [0b6dfce0c] 2025-07-15 18:11:18 -0400
+Branch: REL_17_STABLE [5a2139a90] 2025-07-15 18:11:18 -0400
+Branch: REL_16_STABLE [5db55e13f] 2025-07-15 18:11:18 -0400
+Branch: REL_15_STABLE [f32e45641] 2025-07-15 18:11:18 -0400
+Branch: REL_14_STABLE [8254b7e33] 2025-07-15 18:11:18 -0400
+Branch: REL_13_STABLE [6c93bf735] 2025-07-15 18:11:18 -0400
+-->
+ <para>
+ Prevent uninitialized-value compiler warnings in JSONB comparison
+ code (Tom Lane)
+ <ulink url="&commit_baseurl;f32e45641">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: Michael Paquier <michael@paquier.xyz>
+Branch: master [8aa54aa7e] 2025-07-07 08:53:57 +0900
+Branch: REL_18_STABLE [8d1071e7d] 2025-07-07 08:54:30 +0900
+Branch: REL_17_STABLE [c911e7802] 2025-07-07 08:54:37 +0900
+Branch: REL_16_STABLE [d24a96ce2] 2025-07-07 08:54:39 +0900
+Branch: REL_15_STABLE [75c1eb60b] 2025-07-07 08:54:41 +0900
+Branch: REL_14_STABLE [ac166b19a] 2025-07-07 08:54:43 +0900
+Branch: REL_13_STABLE [e688cfc2b] 2025-07-07 08:54:44 +0900
+-->
+ <para>
+ Avoid deprecation warnings when building
+ with <application>libxml2</application> 2.14 and later
+ (Michael Paquier)
+ <ulink url="&commit_baseurl;75c1eb60b">§</ulink>
+ </para>
+ </listitem>
+
+ <listitem>
+<!--
+Author: John Naylor <john.naylor@postgresql.org>
+Branch: master [ed26c4e25] 2025-07-09 14:20:22 +0700
+Branch: master [90bfae9f9] 2025-08-07 17:10:52 +0700
+Branch: REL_18_STABLE [dd2926207] 2025-08-07 17:12:44 +0700
+Branch: REL_17_STABLE [21ae8fc5f] 2025-08-07 17:13:55 +0700
+Branch: REL_16_STABLE [aae9aad19] 2025-08-07 17:14:39 +0700
+Branch: REL_15_STABLE [baacfb9e6] 2025-08-07 17:15:09 +0700
+Branch: REL_14_STABLE [a4f891b96] 2025-08-07 17:16:00 +0700
+Branch: REL_13_STABLE [9560c1ec1] 2025-08-07 17:16:33 +0700
+-->
+ <para>
+ Avoid problems when compiling <filename>pg_locale.h</filename> under
+ C++ (John Naylor)
+ <ulink url="&commit_baseurl;baacfb9e6">§</ulink>
+ </para>
+
+ <para>
+ <productname>PostgreSQL</productname> header files generally need to
+ be wrapped in <literal>extern "C" { ... }</literal> in order to be
+ included in extensions written in C++. This failed
+ for <filename>pg_locale.h</filename> because of its use
+ of <application>libicu</application> headers, but we can work around
+ that by suppressing C++-only declarations in those headers. C++
+ extensions that want to use <application>libicu</application>'s C++
+ APIs can do so by including the <application>libicu</application>
+ headers ahead of <filename>pg_locale.h</filename>.
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ </sect2>
+ </sect1>
+
<sect1 id="release-15-13">
<title>Release 15.13</title>
projects / thirdparty / postgresql.git / commitdiff
