<?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.4. Release 11.2</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-11-3.html" title="E.3. Release 11.3" /><link rel="next" href="release-11-1.html" title="E.5. Release 11.1" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">E.4. Release 11.2</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="release-11-3.html" title="E.3. Release 11.3">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 11.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="release-11-1.html" title="E.5. Release 11.1">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="RELEASE-11-2"><div class="titlepage"><div><div><h2 class="title" style="clear: both">E.4. Release 11.2</h2></div></div></div><div class="toc"><dl class="toc"><dt><span class="sect2"><a href="release-11-2.html#id-1.11.6.8.4">E.4.1. Migration to Version 11.2</a></span></dt><dt><span class="sect2"><a href="release-11-2.html#id-1.11.6.8.5">E.4.2. Changes</a></span></dt></dl></div><p><strong>Release date: </strong>2019-02-14</p><p> This release contains a variety of fixes from 11.1. For information about new features in major release 11, see <a class="xref" href="release-11.html" title="E.6. Release 11">Section E.6</a>. </p><div class="sect2" id="id-1.11.6.8.4"><div class="titlepage"><div><div><h3 class="title">E.4.1. Migration to Version 11.2</h3></div></div></div><p> A dump/restore is not required for those running 11.X. </p><p> However, if you are upgrading from a version earlier than 11.1, see <a class="xref" href="release-11-1.html" title="E.5. Release 11.1">Section E.5</a>. </p></div><div class="sect2" id="id-1.11.6.8.5"><div class="titlepage"><div><div><h3 class="title">E.4.2. Changes</h3></div></div></div><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p> By default, panic instead of retrying after <code class="function">fsync()</code> failure, to avoid possible data corruption (Craig Ringer, Thomas Munro) </p><p> Some popular operating systems discard kernel data buffers when unable to write them out, reporting this as <code class="function">fsync()</code> failure. If we reissue the <code class="function">fsync()</code> request it will succeed, but in fact the data has been lost, so continuing risks database corruption. By raising a panic condition instead, we can replay from WAL, which may contain the only remaining copy of the data in such a situation. While this is surely ugly and inefficient, there are few alternatives, and fortunately the case happens very rarely. </p><p> A new server parameter <a class="xref" href="runtime-config-error-handling.html#GUC-DATA-SYNC-RETRY">data_sync_retry</a> has been added to control this; if you are certain that your kernel does not discard dirty data buffers in such scenarios, you can set <code class="varname">data_sync_retry</code> to <code class="literal">on</code> to restore the old behavior. </p></li><li class="listitem"><p> Include each major release branch's release notes in the documentation for only that branch, rather than that branch and all later ones (Tom Lane) </p><p> The duplication induced by the previous policy was getting out of hand. Our plan is to provide a full archive of release notes on the project's web site, but not duplicate it within each release. </p></li><li class="listitem"><p> Fix handling of unique indexes with <code class="literal">INCLUDE</code> columns on partitioned tables (Álvaro Herrera) </p><p> The uniqueness condition was not checked properly in such cases. </p></li><li class="listitem"><p> Ensure that <code class="literal">NOT NULL</code> constraints of a partitioned table are honored within its partitions (Álvaro Herrera, Amit Langote) </p></li><li class="listitem"><p> Update catalog state correctly for partition table constraints when detaching their partition (Amit Langote, Álvaro Herrera) </p><p> Previously, the <code class="structname">pg_constraint</code>.<code class="structfield">conislocal</code> field for such a constraint might improperly be left as <code class="literal">false</code>, rendering it undroppable. A dump/restore or pg_upgrade would cure the problem, but if necessary, the catalog field can be adjusted manually. </p></li><li class="listitem"><p> Create or delete foreign key enforcement triggers correctly when attaching or detaching a partition in a partitioned table that has a foreign-key constraint (Amit Langote, Álvaro Herrera) </p></li><li class="listitem"><p> Avoid useless creation of duplicate foreign key constraints in partitioned tables (Álvaro Herrera) </p></li><li class="listitem"><p> When an index is created on a partitioned table using <code class="literal">ONLY</code>, and there are no partitions yet, mark it valid immediately (Álvaro Herrera) </p><p> Otherwise there is no way to make it become valid. </p></li><li class="listitem"><p> Use a safe table lock level when detaching a partition (Álvaro Herrera) </p><p> The previous locking level was too weak and might allow concurrent DDL on the table, with bad results. </p></li><li class="listitem"><p> Fix problems with applying <code class="literal">ON COMMIT DROP</code> and <code class="literal">ON COMMIT DELETE ROWS</code> to partitioned tables and tables with inheritance children (Michael Paquier) </p></li><li class="listitem"><p> Disallow <code class="command">COPY FREEZE</code> on partitioned tables (David Rowley) </p><p> This should eventually be made to work, but it may require a patch that's too complicated to risk back-patching. </p></li><li class="listitem"><p> Fix possible index corruption when the indexed column has a <span class="quote">“<span class="quote">fast default</span>”</span> (that is, it was added by <code class="literal">ALTER TABLE ADD COLUMN</code> with a constant non-NULL default value specified, after the table already contained some rows) (Andres Freund) </p></li><li class="listitem"><p> Correctly adjust <span class="quote">“<span class="quote">fast default</span>”</span> values during <code class="command">ALTER TABLE ... ALTER COLUMN TYPE</code> (Andrew Dunstan) </p></li><li class="listitem"><p> Avoid possible deadlock when acquiring multiple buffer locks (Nishant Fnu) </p></li><li class="listitem"><p> Avoid deadlock between GIN vacuuming and concurrent index insertions (Alexander Korotkov, Andrey Borodin, Peter Geoghegan) </p><p> This change partially reverts a performance improvement, introduced in version 10.0, that attempted to reduce the number of index pages locked during deletion of a GIN posting tree page. That's now been found to lead to deadlocks, so we've removed it pending closer analysis. </p></li><li class="listitem"><p> Avoid deadlock between hot-standby queries and replay of GIN index page deletion (Alexander Korotkov) </p></li><li class="listitem"><p> Fix possible crashes in logical replication when index expressions or predicates are in use (Peter Eisentraut) </p></li><li class="listitem"><p> Avoid useless and expensive logical decoding of TOAST data during a table rewrite (Tomas Vondra) </p></li><li class="listitem"><p> Fix logic for stopping a subset of WAL senders when synchronous replication is enabled (Paul Guo, Michael Paquier) </p></li><li class="listitem"><p> Avoid possibly writing an incorrect replica identity field in a tuple deletion WAL record (Stas Kelvich) </p></li><li class="listitem"><p> Prevent incorrect use of WAL-skipping optimization during <code class="command">COPY</code> to a view or foreign table (Amit Langote, Michael Paquier) </p></li><li class="listitem"><p> Make the archiver prioritize WAL history files over WAL data files while choosing which file to archive next (David Steele) </p></li><li class="listitem"><p> Fix possible crash in <code class="command">UPDATE</code> with a multiple <code class="literal">SET</code> clause using a sub-<code class="literal">SELECT</code> as source (Tom Lane) </p></li><li class="listitem"><p> Fix crash when zero rows are fed to <code class="function">json[b]_populate_recordset()</code> or <code class="function">json[b]_to_recordset()</code> (Tom Lane) </p></li><li class="listitem"><p> Avoid crash if <span class="application">libxml2</span> returns a null error message (Sergio Conde Gómez) </p></li><li class="listitem"><p> Fix incorrect JIT tuple deforming code for tables with many columns (more than approximately 800) (Andres Freund) </p></li><li class="listitem"><p> Fix performance and memory leakage issues in hash-based grouping (Andres Freund) </p></li><li class="listitem"><p> Fix spurious grouping-related parser errors caused by inconsistent handling of collation assignment (Andrew Gierth) </p><p> In some cases, expressions that should be considered to match were not seen as matching, if they included operations on collatable data types. </p></li><li class="listitem"><p> Fix parsing of collation-sensitive expressions in the arguments of a <code class="command">CALL</code> statement (Peter Eisentraut) </p></li><li class="listitem"><p> Ensure proper cleanup after detecting an error in the argument list of a <code class="command">CALL</code> statement (Tom Lane) </p></li><li class="listitem"><p> Check whether the comparison function underlying <code class="function">LEAST()</code> or <code class="function">GREATEST()</code> is leakproof, rather than just assuming it is (Tom Lane) </p><p> Actual information leaks from btree comparison functions are typically hard to provoke, but in principle they could happen. </p></li><li class="listitem"><p> Fix incorrect planning of queries involving nested loops both above and below a Gather plan node (Tom Lane) </p><p> If both levels of nestloop needed to pass the same variable into their right-hand sides, an incorrect plan would be generated. </p></li><li class="listitem"><p> Fix incorrect planning of queries in which a lateral reference must be evaluated at a foreign table scan (Tom Lane) </p></li><li class="listitem"><p> Fix planner failure when the first column of a row comparison matches an index column, but later column(s) do not, and the index has included (non-key) columns (Tom Lane) </p></li><li class="listitem"><p> Fix corner-case underestimation of the cost of a merge join (Tom Lane) </p><p> The planner could prefer a merge join when the outer key range is much smaller than the inner key range, even if there are so many duplicate keys on the inner side that this is a poor choice. </p></li><li class="listitem"><p> Avoid O(N^2) planning time growth when a query contains many thousand indexable clauses (Tom Lane) </p></li><li class="listitem"><p> Improve planning speed for large inheritance or partitioning table groups (Amit Langote, Etsuro Fujita) </p></li><li class="listitem"><p> Improve <code class="command">ANALYZE</code>'s handling of concurrently-updated rows (Jeff Janes, Tom Lane) </p><p> Previously, rows deleted by an in-progress transaction were omitted from <code class="command">ANALYZE</code>'s sample, but this has been found to lead to more inconsistency than including them would do. In effect, the sample now corresponds to an MVCC snapshot as of <code class="command">ANALYZE</code>'s start time. </p></li><li class="listitem"><p> Make <code class="command">TRUNCATE</code> ignore inheritance child tables that are temporary tables of other sessions (Amit Langote, Michael Paquier) </p><p> This brings <code class="command">TRUNCATE</code> into line with the behavior of other commands. Previously, such cases usually ended in failure. </p></li><li class="listitem"><p> Fix <code class="command">TRUNCATE</code> to update the statistics counters for the right table (Tom Lane) </p><p> If the truncated table had a TOAST table, that table's counters were reset instead. </p></li><li class="listitem"><p> Process <code class="command">ALTER TABLE ONLY ADD COLUMN IF NOT EXISTS</code> correctly (Greg Stark) </p></li><li class="listitem"><p> Allow <code class="command">UNLISTEN</code> in hot-standby mode (Shay Rojansky) </p><p> This is necessarily a no-op, because <code class="command">LISTEN</code> isn't allowed in hot-standby mode; but allowing the dummy operation simplifies session-state-reset logic in clients. </p></li><li class="listitem"><p> Fix missing role dependencies in some schema and data type permissions lists (Tom Lane) </p><p> In some cases it was possible to drop a role to which permissions had been granted. This caused no immediate problem, but a subsequent dump/reload or upgrade would fail, with symptoms involving attempts to grant privileges to all-numeric role names. </p></li><li class="listitem"><p> Prevent use of a session's temporary schema within a two-phase transaction (Michael Paquier) </p><p> Accessing a temporary table within such a transaction has been forbidden for a long time, but it was still possible to cause problems with other operations on temporary objects. </p></li><li class="listitem"><p> Ensure relation caches are updated properly after adding or removing foreign key constraints (Álvaro Herrera) </p><p> This oversight could result in existing sessions failing to enforce a newly-created constraint, or continuing to enforce a dropped one. </p></li><li class="listitem"><p> Ensure relation caches are updated properly after renaming constraints (Amit Langote) </p></li><li class="listitem"><p> Fix replay of GiST index micro-vacuum operations so that concurrent hot-standby queries do not see inconsistent state (Alexander Korotkov) </p></li><li class="listitem"><p> Prevent empty GIN index pages from being reclaimed too quickly, causing failures of concurrent searches (Andrey Borodin, Alexander Korotkov) </p></li><li class="listitem"><p> Fix edge-case failures in float-to-integer coercions (Andrew Gierth, Tom Lane) </p><p> Values very slightly above the maximum valid integer value might not be rejected, and then would overflow, producing the minimum valid integer instead. Also, values that should round to the minimum or maximum integer value might be incorrectly rejected. </p></li><li class="listitem"><p> Fix parsing of space-separated lists of host names in the <code class="varname">ldapserver</code> parameter of LDAP authentication entries in <code class="filename">pg_hba.conf</code> (Thomas Munro) </p></li><li class="listitem"><p> When making a PAM authentication request, don't set the <code class="varname">PAM_RHOST</code> variable if the connection is via a Unix socket (Thomas Munro) </p><p> Previously that variable would be set to <code class="literal">[local]</code>, which is at best unhelpful, since it's supposed to be a host name. </p></li><li class="listitem"><p> Disallow setting <code class="varname">client_min_messages</code> higher than <code class="literal">ERROR</code> (Jonah Harris, Tom Lane) </p><p> Previously, it was possible to set this variable to <code class="literal">FATAL</code> or <code class="literal">PANIC</code>, which had the effect of suppressing transmission of ordinary error messages to the client. However, that's contrary to guarantees that are given in the <span class="productname">PostgreSQL</span> wire protocol specification, and it caused some clients to become very confused. In released branches, fix this by silently treating such settings as meaning <code class="literal">ERROR</code> instead. Version 12 and later will reject those alternatives altogether. </p></li><li class="listitem"><p> Fix <span class="application">ecpglib</span> to use <code class="function">uselocale()</code> or <code class="function">_configthreadlocale()</code> in preference to <code class="function">setlocale()</code> (Michael Meskes, Tom Lane) </p><p> Since <code class="function">setlocale()</code> is not thread-local, and might not even be thread-safe, the previous coding caused problems in multi-threaded <span class="application">ecpg</span> applications. </p></li><li class="listitem"><p> Fix incorrect results for numeric data passed through an <span class="application">ecpg</span> <acronym class="acronym">SQLDA</acronym> (SQL Descriptor Area) (Daisuke Higuchi) </p><p> Values with leading zeroes were not copied correctly. </p></li><li class="listitem"><p> Fix <span class="application">psql</span>'s <code class="command">\g</code> <em class="replaceable"><code>target</code></em> meta-command to work with <code class="command">COPY TO STDOUT</code> (Daniel Vérité) </p><p> Previously, the <em class="replaceable"><code>target</code></em> option was ignored, so that the copy data always went to the current query output target. </p></li><li class="listitem"><p> Make <span class="application">psql</span>'s LaTeX output formats render special characters properly (Tom Lane) </p><p> Backslash and some other ASCII punctuation characters were not rendered correctly, leading to document syntax errors or wrong characters in the output. </p></li><li class="listitem"><p> Make <span class="application">pgbench</span>'s random number generation fully deterministic and platform-independent when <code class="option">--random-seed=<em class="replaceable"><code>N</code></em></code> is specified (Fabien Coelho, Tom Lane) </p><p> On any specific platform, the sequence obtained with a particular value of <em class="replaceable"><code>N</code></em> will probably be different from what it was before this patch. </p></li><li class="listitem"><p> Fix <span class="application">pg_basebackup</span> and <span class="application">pg_verify_checksums</span> to ignore temporary files appropriately (Michael Banck, Michael Paquier) </p></li><li class="listitem"><p> Fix <span class="application">pg_dump</span>'s handling of materialized views with indirect dependencies on primary keys (Tom Lane) </p><p> This led to mis-labeling of such views' dump archive entries, causing harmless warnings about <span class="quote">“<span class="quote">archive items not in correct section order</span>”</span>; less harmlessly, selective-restore options depending on those labels, such as <code class="option">--section</code>, might misbehave. </p></li><li class="listitem"><p> Make <span class="application">pg_dump</span> include <code class="command">ALTER INDEX SET STATISTICS</code> commands (Michael Paquier) </p><p> When the ability to attach statistics targets to index expressions was added, we forgot to teach <span class="application">pg_dump</span> about it, so that such settings were lost in dump/reload. </p></li><li class="listitem"><p> Fix <span class="application">pg_dump</span>'s dumping of tables that have OIDs (Peter Eisentraut) </p><p> The <code class="literal">WITH OIDS</code> clause was omitted if it needed to be applied to the first table to be dumped. </p></li><li class="listitem"><p> Avoid null-pointer-dereference crash on some platforms when <span class="application">pg_dump</span> or <span class="application">pg_restore</span> tries to report an error (Tom Lane) </p></li><li class="listitem"><p> Prevent false index-corruption reports from <code class="filename">contrib/amcheck</code> caused by inline-compressed data (Peter Geoghegan) </p></li><li class="listitem"><p> Properly disregard <code class="literal">SIGPIPE</code> errors if <code class="command">COPY FROM PROGRAM</code> stops reading the program's output early (Tom Lane) </p><p> This case isn't actually reachable directly with <code class="command">COPY</code>, but it can happen when using <code class="filename">contrib/file_fdw</code>. </p></li><li class="listitem"><p> Fix <code class="filename">contrib/hstore</code> to calculate correct hash values for empty <code class="type">hstore</code> values that were created in version 8.4 or before (Andrew Gierth) </p><p> The previous coding did not give the same result as for an empty <code class="type">hstore</code> value created by a newer version, thus potentially causing wrong results in hash joins or hash aggregation. It is advisable to reindex any hash indexes built on <code class="type">hstore</code> columns, if the table might contain data that was originally stored as far back as 8.4 and was never dumped/reloaded since then. </p></li><li class="listitem"><p> Avoid crashes and excessive runtime with large inputs to <code class="filename">contrib/intarray</code>'s <code class="literal">gist__int_ops</code> index support (Andrew Gierth) </p></li><li class="listitem"><p> In <span class="application">configure</span>, look for <code class="command">python3</code> and then <code class="command">python2</code> if <code class="command">python</code> isn't found (Peter Eisentraut) </p><p> This allows PL/Python to be configured without explicitly specifying <code class="literal">PYTHON</code> on platforms that no longer provide an unversioned <code class="command">python</code> executable. </p></li><li class="listitem"><p> Include JIT-related headers in the installed set of header files (Donald Dong) </p></li><li class="listitem"><p> Support new Makefile variables <code class="literal">PG_CFLAGS</code>, <code class="literal">PG_CXXFLAGS</code>, and <code class="literal">PG_LDFLAGS</code> in <span class="application">pgxs</span> builds (Christoph Berg) </p><p> This simplifies customization of extension build processes. </p></li><li class="listitem"><p> Fix Perl-coded build scripts to not assume <span class="quote">“<span class="quote"><code class="literal">.</code></span>”</span> is in the search path, since recent Perl versions don't include that (Andrew Dunstan) </p></li><li class="listitem"><p> Fix server command-line option parsing problems on OpenBSD (Tom Lane) </p></li><li class="listitem"><p> Relocate call of <code class="literal">set_rel_pathlist_hook</code> so that extensions can use it to supply partial paths for parallel queries (KaiGai Kohei) </p><p> This is not expected to affect existing use-cases. </p></li><li class="listitem"><p> Update time zone data files to <span class="application">tzdata</span> release 2018i for DST law changes in Kazakhstan, Metlakatla, and Sao Tome and Principe. Kazakhstan's Qyzylorda zone is split in two, creating a new zone Asia/Qostanay, as some areas did not change UTC offset. Historical corrections for Hong Kong and numerous Pacific islands. </p></li></ul></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="release-11-3.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="release.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="release-11-1.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">E.3. Release 11.3 </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> E.5. Release 11.1</td></tr></table></div></body></html>