summaryrefslogtreecommitdiffstats
path: root/doc/src/sgml/html/release-14-3.html
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--doc/src/sgml/html/release-14-3.html429
1 files changed, 429 insertions, 0 deletions
diff --git a/doc/src/sgml/html/release-14-3.html b/doc/src/sgml/html/release-14-3.html
new file mode 100644
index 0000000..a64dd99
--- /dev/null
+++ b/doc/src/sgml/html/release-14-3.html
@@ -0,0 +1,429 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>E.3. Release 14.3</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><link rel="prev" href="release-14-4.html" title="E.2. Release 14.4" /><link rel="next" href="release-14-2.html" title="E.4. Release 14.2" /></head><body id="docContent" class="container-fluid col-10"><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">E.3. Release 14.3</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="release-14-4.html" title="E.2. Release 14.4">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="release.html" title="Appendix E. Release Notes">Up</a></td><th width="60%" align="center">Appendix E. Release Notes</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 14.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="release-14-2.html" title="E.4. Release 14.2">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="RELEASE-14-3"><div class="titlepage"><div><div><h2 class="title" style="clear: both">E.3. Release 14.3</h2></div></div></div><div class="toc"><dl class="toc"><dt><span class="sect2"><a href="release-14-3.html#id-1.11.6.7.4">E.3.1. Migration to Version 14.3</a></span></dt><dt><span class="sect2"><a href="release-14-3.html#id-1.11.6.7.5">E.3.2. Changes</a></span></dt></dl></div><p><strong>Release date: </strong>2022-05-12</p><p>
+ This release contains a variety of fixes from 14.2.
+ For information about new features in major release 14, see
+ <a class="xref" href="release-14.html" title="E.6. Release 14">Section E.6</a>.
+ </p><div class="sect2" id="id-1.11.6.7.4"><div class="titlepage"><div><div><h3 class="title">E.3.1. Migration to Version 14.3</h3></div></div></div><p>
+ A dump/restore is not required for those running 14.X.
+ </p><p>
+ However, if you have any GiST indexes on columns of type
+ <code class="type">ltree</code> (supplied by the <code class="filename">contrib/ltree</code>
+ extension), you should re-index them after updating.
+ See the second changelog entry below.
+ </p><p>
+ Also, if you are upgrading from a version earlier than 14.2,
+ see <a class="xref" href="release-14-2.html" title="E.4. Release 14.2">Section E.4</a>.
+ </p></div><div class="sect2" id="id-1.11.6.7.5"><div class="titlepage"><div><div><h3 class="title">E.3.2. Changes</h3></div></div></div><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
+ Confine additional operations within <span class="quote">“<span class="quote">security restricted
+ operation</span>”</span> sandboxes (Sergey Shinderuk, Noah Misch)
+ </p><p>
+ Autovacuum, <code class="command">CLUSTER</code>, <code class="command">CREATE
+ INDEX</code>, <code class="command">REINDEX</code>, <code class="command">REFRESH
+ MATERIALIZED VIEW</code>,
+ and <span class="application">pg_amcheck</span> activated
+ the <span class="quote">“<span class="quote">security restricted operation</span>”</span> protection
+ mechanism too late, or even not at all in some code paths.
+ A user having permission to create non-temporary objects within a
+ database could define an object that would execute arbitrary SQL
+ code with superuser permissions the next time that autovacuum
+ processed the object, or that some superuser ran one of the affected
+ commands against it.
+ </p><p>
+ The <span class="productname">PostgreSQL</span> Project thanks
+ Alexander Lakhin for reporting this problem.
+ (CVE-2022-1552)
+ </p></li><li class="listitem"><p>
+ Fix default signature length for <code class="literal">gist_ltree_ops</code>
+ indexes (Tomas Vondra, Alexander Korotkov)
+ </p><p>
+ The default signature length (hash size) for GiST indexes
+ on <code class="type">ltree</code> columns was accidentally changed while
+ upgrading that operator class to support operator class parameters.
+ If any operations had been done on such an index without first
+ upgrading the <code class="literal">ltree</code> extension to version 1.2,
+ they were done assuming that the signature length was 28 bytes
+ rather than the intended 8. This means it is very likely that such
+ indexes are now corrupt. For safety we recommend re-indexing all
+ GiST indexes on <code class="type">ltree</code> columns after installing this
+ update. (Note that GiST indexes on <code class="type">ltree[]</code>
+ columns, that is arrays of <code class="type">ltree</code>, are not affected.)
+ </p></li><li class="listitem"><p>
+ Stop using query-provided column aliases for the columns of
+ whole-row variables that refer to plain tables (Tom Lane)
+ </p><p>
+ The column names in tuples produced by a whole-row variable (such
+ as <code class="literal">tbl.*</code> in contexts other than the top level of
+ a <code class="literal">SELECT</code> list) are now always those of the
+ associated named composite type, if there is one. We'd previously
+ attempted to make them track any column aliases that had been
+ applied to the <code class="literal">FROM</code> entry the variable refers to.
+ But that's semantically dubious, because really then the output of
+ the variable is not at all of the composite type it claims to be.
+ Previous attempts to deal with that inconsistency had bad results up
+ to and including storing unreadable data on disk, so just give up on
+ the whole idea.
+ </p><p>
+ In cases where it's important to be able to relabel such columns,
+ a workaround is to introduce an extra level of
+ sub-<code class="literal">SELECT</code>, so that the whole-row variable is
+ referring to the sub-<code class="literal">SELECT</code>'s output and not to a
+ plain table. Then the variable is of type <code class="type">record</code>
+ to begin with and there's no issue.
+ </p></li><li class="listitem"><p>
+ Fix incorrect roundoff when extracting epoch values from intervals
+ (Peter Eisentraut)
+ </p><p>
+ The new <code class="type">numeric</code>-based code
+ for <code class="function">EXTRACT()</code> failed to yield results
+ equivalent to the old <code class="type">float</code>-based code, as a result of
+ accidentally truncating the <code class="literal">DAYS_PER_YEAR</code> value
+ to an integer.
+ </p></li><li class="listitem"><p>
+ Defend against <code class="literal">pg_stat_get_replication_slot(NULL)</code>
+ (Andres Freund)
+ </p><p>
+ This function should be marked strict in the catalog data, but it
+ was not in v14, so add a run-time check instead.
+ </p></li><li class="listitem"><p>
+ Fix incorrect output for types <code class="type">timestamptz</code>
+ and <code class="type">timetz</code> in <code class="function">table_to_xmlschema()</code>
+ and allied functions (Renan Soares Lopes)
+ </p><p>
+ The xmlschema output for these types included a malformed regular
+ expression.
+ </p></li><li class="listitem"><p>
+ Avoid core dump in parser for a <code class="literal">VALUES</code> clause with
+ zero columns (Tom Lane)
+ </p></li><li class="listitem"><p>
+ Fix planner failure when a Result plan node appears immediately
+ underneath an Append node (Etsuro Fujita)
+ </p><p>
+ Recently-added code to support asynchronous remote queries failed to
+ handle this case, leading to crashes or errors about unrecognized
+ node types.
+ </p></li><li class="listitem"><p>
+ Fix planner failure if a query using <code class="literal">SEARCH</code>
+ or <code class="literal">CYCLE</code> features contains a duplicate CTE name
+ (Tom Lane, Kyotaro Horiguchi)
+ </p><p>
+ When the name of the recursive <code class="literal">WITH</code> query is
+ re-used within itself, the planner could crash or report odd errors
+ such as <span class="quote">“<span class="quote">could not find attribute 2 in subquery
+ targetlist</span>”</span>.
+ </p></li><li class="listitem"><p>
+ Fix planner errors for <code class="function">GROUPING()</code> constructs
+ that reference outer query levels (Richard Guo, Tom Lane)
+ </p></li><li class="listitem"><p>
+ Fix plan generation for index-only scans on indexes with
+ both returnable and non-returnable columns (Tom Lane)
+ </p><p>
+ The previous coding could try to read non-returnable columns
+ in addition to the returnable ones. This was fairly harmless
+ because it didn't actually do anything with the bogus values,
+ but it fell foul of a recently-added error check that rejected
+ such a plan.
+ </p></li><li class="listitem"><p>
+ Avoid accessing a no-longer-pinned shared buffer while attempting
+ to lock an outdated tuple during EvalPlanQual (Tom Lane)
+ </p><p>
+ The code would touch the buffer a couple more times after releasing
+ its pin. In theory another process could recycle the buffer (or
+ more likely, try to defragment its free space) as soon as the pin is
+ gone, probably leading to failure to find the newer version of the
+ tuple.
+ </p></li><li class="listitem"><p>
+ Fix query-lifespan memory leak in an IndexScan node that is
+ performing reordering (Aliaksandr Kalenik)
+ </p></li><li class="listitem"><p>
+ Fix <code class="command">ALTER FUNCTION</code> to support changing a
+ function's parallelism property and
+ its <code class="literal">SET</code>-variable list in the same command (Tom
+ Lane)
+ </p><p>
+ The parallelism property change was lost if the same command also
+ updated the function's <code class="literal">SET</code> clause.
+ </p></li><li class="listitem"><p>
+ Tighten lookup of the index <span class="quote">“<span class="quote">owned by</span>”</span> a constraint
+ (Tom Lane, Japin Li)
+ </p><p>
+ Some code paths mistook the index depended on by a foreign key
+ constraint for one owned by a unique or primary key constraint,
+ resulting in odd errors during certain <code class="command">ALTER
+ TABLE</code> operations on tables having foreign key constraints.
+ </p></li><li class="listitem"><p>
+ Fix bogus errors from attempts to alter system columns of tables
+ (Tom Lane)
+ </p><p>
+ The system should just tell you that you can't do it, but sometimes
+ it would report <span class="quote">“<span class="quote">no owned sequence found</span>”</span> instead.
+ </p></li><li class="listitem"><p>
+ Fix mis-sorting of table rows when <code class="command">CLUSTER</code>ing
+ using an index whose leading key is an expression (Peter Geoghegan,
+ Thomas Munro)
+ </p><p>
+ The table would be rebuilt with the correct data, but in an order
+ having little to do with the index order.
+ </p></li><li class="listitem"><p>
+ Prevent data loss if a system crash occurs shortly after a sorted
+ GiST index build (Heikki Linnakangas)
+ </p><p>
+ The code path for building GiST indexes using sorting neglected
+ to <code class="function">fsync</code> the file upon completion. This could
+ result in a corrupted index if the operating system crashed shortly
+ later.
+ </p></li><li class="listitem"><p>
+ Fix risk of deadlock failures while dropping a partitioned index
+ (Jimmy Yih, Gaurab Dey, Tom Lane)
+ </p><p>
+ Ensure that the required table and index locks are taken in the
+ standard order (parents before children, tables before indexes).
+ The previous coding for <code class="command">DROP INDEX</code> did it
+ differently, and so could deadlock against concurrent queries taking
+ these locks in the standard order.
+ </p></li><li class="listitem"><p>
+ Fix race condition between <code class="command">DROP TABLESPACE</code> and
+ checkpointing (Nathan Bossart)
+ </p><p>
+ The checkpoint forced by <code class="command">DROP TABLESPACE</code> could
+ sometimes fail to remove all dead files from the tablespace's
+ directory, leading to a bogus <span class="quote">“<span class="quote">tablespace is not empty</span>”</span>
+ error.
+ </p></li><li class="listitem"><p>
+ Fix possible trouble in crash recovery after
+ a <code class="command">TRUNCATE</code> command that overlaps a checkpoint
+ (Kyotaro Horiguchi, Heikki Linnakangas, Robert Haas)
+ </p><p>
+ <code class="command">TRUNCATE</code> must ensure that the table's disk file
+ is truncated before the checkpoint is allowed to complete.
+ Otherwise, replay starting from that checkpoint might find
+ unexpected data in the supposedly-removed pages, possibly causing
+ replay failure.
+ </p></li><li class="listitem"><p>
+ Fix unsafe toast-data accesses during temporary object cleanup
+ (Andres Freund)
+ </p><p>
+ Temporary-object deletion during server process exit could fail
+ with <span class="quote">“<span class="quote">FATAL: cannot fetch toast data without an active
+ snapshot</span>”</span>. This was usually harmless since the next use of
+ that temporary schema would clean up successfully.
+ </p></li><li class="listitem"><p>
+ Re-allow underscore as the first character in a custom parameter
+ name (Japin Li)
+ </p><p>
+ Such names were unintentionally disallowed in v14.
+ </p></li><li class="listitem"><p>
+ Add <code class="literal">regress</code> option for
+ the <code class="varname">compute_query_id</code> parameter (Michael Paquier)
+ </p><p>
+ This is intended to facilitate testing, by allowing query IDs to be
+ computed but not shown in <code class="command">EXPLAIN</code> output.
+ </p></li><li class="listitem"><p>
+ Improve wait logic in RegisterSyncRequest (Thomas Munro)
+ </p><p>
+ If we run out of space in the checkpointer sync request queue (which
+ is hopefully rare on real systems, but is common when testing with a
+ very small buffer pool), we wait for it to drain. While waiting, we
+ should report that as a wait event so that users know what is going
+ on, and also watch for postmaster death, since otherwise the loop might
+ never terminate if the checkpointer has already exited.
+ </p></li><li class="listitem"><p>
+ Wake up for latch events when the checkpointer is waiting between
+ writes (Thomas Munro)
+ </p><p>
+ This improves responsiveness to backends sending sync requests.
+ The change also creates a proper wait event class for these waits.
+ </p></li><li class="listitem"><p>
+ Fix <span class="quote">“<span class="quote">PANIC: xlog flush request is not satisfied</span>”</span>
+ failure during standby promotion when there is a missing WAL
+ continuation record (Sami Imseih)
+ </p></li><li class="listitem"><p>
+ Fix possibility of self-deadlock in hot standby conflict handling
+ (Andres Freund)
+ </p><p>
+ With unlucky timing, the WAL-applying process could get stuck
+ while waiting for some other process to release a buffer lock.
+ </p></li><li class="listitem"><p>
+ Fix possible mis-identification of the correct ancestor relation
+ to publish logical replication changes through (Tomas Vondra, Hou
+ zj, Amit Kapila)
+ </p><p>
+ If <code class="varname">publish_via_partition_root</code> is enabled, and
+ there are multiple publications naming different ancestors of the
+ currently-modified relation, the wrong ancestor might be chosen for
+ reporting the change.
+ </p></li><li class="listitem"><p>
+ Ensure that logical replication apply workers can be restarted even
+ when we're up against
+ the <code class="varname">max_sync_workers_per_subscription</code> limit
+ (Amit Kapila)
+ </p><p>
+ Faulty coding of the limit check caused a restarted worker to exit
+ immediately, leaving fewer workers than there should be.
+ </p></li><li class="listitem"><p>
+ Include unchanged replica identity key columns in the WAL log for an
+ update, if they are stored out-of-line (Dilip Kumar, Amit Kapila)
+ </p><p>
+ Otherwise subscribers cannot see the values and will fail to
+ replicate the update.
+ </p></li><li class="listitem"><p>
+ Cope correctly with platforms that have no support for altering the
+ server process's display in <span class="application">ps(1)</span> (Andrew
+ Dunstan)
+ </p><p>
+ Few platforms are like this (the only supported one is Cygwin), so
+ we'd managed not to notice that refactoring introduced a potential
+ memory clobber.
+ </p></li><li class="listitem"><p>
+ Make the server more robust against missed timer interrupts (Michael
+ Harris, Tom Lane)
+ </p><p>
+ An optimization added in v14 meant that if a server process somehow
+ missed a timer interrupt, it would never again ask the kernel for
+ another one, thus breaking timeout detection for the remainder of the
+ session. This seems unduly fragile, so add a recovery path.
+ </p></li><li class="listitem"><p>
+ Disallow execution of SPI functions during PL/Perl function
+ compilation (Tom Lane)
+ </p><p>
+ Perl can be convinced to execute user-defined code during compilation
+ of a PL/Perl function. However, it's not okay for such code to try
+ to invoke SQL operations via SPI. That results in a crash, and if
+ it didn't crash it would be a security hazard, because we really
+ don't want code execution during function validation. Put in a
+ check to give a friendlier error message instead.
+ </p></li><li class="listitem"><p>
+ Make <span class="application">libpq</span> accept root-owned SSL private
+ key files (David Steele)
+ </p><p>
+ This change synchronizes <span class="application">libpq</span>'s rules
+ for safe ownership and permissions of SSL key files with the rules
+ the server has used since release 9.6. Namely, in addition to the
+ current rules, allow the case where the key file is owned by root
+ and has permissions <code class="literal">rw-r-----</code> or less. This is
+ helpful for system-wide management of key files.
+ </p></li><li class="listitem"><p>
+ Fix behavior of <span class="application">libpq</span>'s
+ <code class="function">PQisBusy()</code> function after a connection failure
+ (Tom Lane)
+ </p><p>
+ If we'd detected a write failure, <code class="function">PQisBusy()</code>
+ would always return true, which is the wrong thing: we want input
+ processing to carry on normally until we've read whatever is
+ available from the server. The practical effect of this error is
+ that applications using <span class="application">libpq</span>'s
+ async-query API would typically detect connection loss only
+ when <code class="function">PQconsumeInput()</code> returns a hard failure.
+ With this fix, a connection loss will normally be reported via an
+ error <code class="structname">PGresult</code> object, which is a much
+ cleaner behavior for most applications.
+ </p></li><li class="listitem"><p>
+ Re-allow <em class="replaceable"><code>database</code></em>.<em class="replaceable"><code>schema</code></em>.<em class="replaceable"><code>table</code></em>
+ patterns
+ in <span class="application">psql</span>, <span class="application">pg_dump</span>,
+ and <span class="application">pg_amcheck</span> (Mark Dilger)
+ </p><p>
+ Versions before v14 silently ignored all but
+ the <em class="replaceable"><code>schema</code></em>
+ and <em class="replaceable"><code>table</code></em> fragments of a pattern
+ containing more than one dot. Refactoring in v14 accidentally broke
+ that use-case. Reinstate it, but now complain if the first fragment
+ is not the name of the current database.
+ </p></li><li class="listitem"><p>
+ Make <span class="application">pg_ctl</span> recheck postmaster aliveness
+ while waiting for stop/restart/promote actions (Tom Lane)
+ </p><p>
+ <span class="application">pg_ctl</span> would verify that the postmaster
+ is alive as a side-effect of sending the stop or promote signal, but
+ then it just naively waited to see the on-disk state change. If the
+ postmaster died uncleanly without having removed its PID file or
+ updated the control file, <span class="application">pg_ctl</span> would
+ wait until timeout. Instead make it recheck every so often that the
+ postmaster process is still there.
+ </p></li><li class="listitem"><p>
+ Fix error handling in <span class="application">pg_waldump</span> (Kyotaro
+ Horiguchi, Andres Freund)
+ </p><p>
+ While trying to read a WAL file to determine the WAL segment size,
+ <span class="application">pg_waldump</span> would report an incorrect
+ error for the case of a too-short file. In addition, the file name
+ reported in this and related error messages could be garbage.
+ </p></li><li class="listitem"><p>
+ Ensure that <code class="filename">contrib/pageinspect</code> functions cope
+ with all-zero pages (Michael Paquier)
+ </p><p>
+ This is a legitimate edge case, but the module was mostly unprepared
+ for it. Arrange to return nulls, or no rows, as appropriate; that
+ seems more useful than raising an error.
+ </p></li><li class="listitem"><p>
+ In <code class="filename">contrib/pageinspect</code>, add defenses against
+ incorrect page <span class="quote">“<span class="quote">special space</span>”</span> contents, tighten checks
+ for correct page size, and add some missing checks that an index is
+ of the expected type (Michael Paquier, Justin Pryzby, Julien
+ Rouhaud)
+ </p><p>
+ These changes make it less likely that the module will crash on bad
+ data.
+ </p></li><li class="listitem"><p>
+ In <code class="filename">contrib/postgres_fdw</code>, disable batch
+ insertion when <code class="literal">BEFORE INSERT ... FOR EACH ROW</code>
+ triggers exist on the foreign table (Etsuro Fujita)
+ </p><p>
+ Such a trigger might query the table it's on and expect to see
+ previously-inserted rows. With batch insertion, those rows might
+ not be visible yet, so disable the feature to avoid unexpected
+ behavior.
+ </p></li><li class="listitem"><p>
+ In <code class="filename">contrib/postgres_fdw</code>, verify
+ that <code class="literal">ORDER BY</code> clauses are safe to ship before
+ requesting a remotely-ordered query, and include
+ a <code class="literal">USING</code> clause if necessary (Ronan Dunklau)
+ </p><p>
+ This fix prevents situations where the remote server might sort in a
+ different order than we intend. While sometimes that would be only
+ cosmetic, it could produce thoroughly wrong results if the remote
+ data is used as input for a locally-performed merge join.
+ </p></li><li class="listitem"><p>
+ Fix <span class="application">configure</span> to handle platforms that
+ have <code class="filename">sys/epoll.h</code> but
+ not <code class="filename">sys/signalfd.h</code> (Tom Lane)
+ </p></li><li class="listitem"><p>
+ Update <acronym class="acronym">JIT</acronym> code to work with LLVM 14
+ (Thomas Munro)
+ </p></li><li class="listitem"><p>
+ Clean up assorted failures under <span class="productname">clang</span>'s
+ <code class="literal">-fsanitize=undefined</code> checks (Tom Lane, Andres
+ Freund, Zhihong Yu)
+ </p><p>
+ Most of these changes are just for pro-forma compliance with the
+ letter of the C and POSIX standards, and are unlikely to have any
+ effect on production builds.
+ </p></li><li class="listitem"><p>
+ Do not add OpenSSL dependencies to <span class="application">libpq</span>'s
+ <code class="filename">pkg-config</code> file when building without OpenSSL
+ (Fabrice Fontaine)
+ </p></li><li class="listitem"><p>
+ Fix PL/Perl so it builds on C compilers that don't support statements
+ nested within expressions (Tom Lane)
+ </p><p>
+ </p></li><li class="listitem"><p>
+ Fix possible build failure of <span class="application">pg_dumpall</span>
+ on Windows, when not using MSVC to build (Andres Freund)
+ </p></li><li class="listitem"><p>
+ In Windows builds, use <span class="application">gendef</span> instead
+ of <span class="application">pexports</span> to build DEF files (Andrew
+ Dunstan)
+ </p><p>
+ This adapts the build process to work on recent MSys tool chains.
+ </p></li><li class="listitem"><p>
+ Prevent extra expansion of shell wildcard patterns in programs built
+ under MinGW (Andrew Dunstan)
+ </p><p>
+ For some reason the C library provided by MinGW will expand shell
+ wildcard characters in a program's command-line arguments by
+ default. This is confusing, not least because it doesn't happen
+ under MSVC, so turn it off.
+ </p></li><li class="listitem"><p>
+ Update time zone data files to <span class="application">tzdata</span>
+ release 2022a for DST law changes in Palestine, plus historical
+ corrections for Chile and Ukraine.
+ </p></li></ul></div></div></div><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navfooter"><hr></hr><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="release-14-4.html" title="E.2. Release 14.4">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="release.html" title="Appendix E. Release Notes">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="release-14-2.html" title="E.4. Release 14.2">Next</a></td></tr><tr><td width="40%" align="left" valign="top">E.2. Release 14.4 </td><td width="20%" align="center"><a accesskey="h" href="index.html" title="PostgreSQL 14.5 Documentation">Home</a></td><td width="40%" align="right" valign="top"> E.4. Release 14.2</td></tr></table></div></body></html> \ No newline at end of file