upstream/postgresql
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2026-04-27 17:16:43 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions
postgresql/doc/src/sgml/release-10.sgml
Tom Lane c1602c7a1b Doc: update v10 release notes through today. 
Also, another round of copy-editing.  I merged a few items that
didn't seem to be meaningfully different from a user's perspective.
2017-09-08 16:59:26 -04:00

3150 lines
90 KiB
Text
Raw Blame History

<!-- doc/src/sgml/release-10.sgml -->
<!-- See header comment in release.sgml about typical markup -->
<sect1 id="release-10">
<title>Release 10</title>
<formalpara>
<title>Release date:</title>
<para>2017-??-?? (current as of 2017-09-07, commit 08cb36417)</para>
</formalpara>
<sect2>
<title>Overview</title>
<para>
Major enhancements in <productname>PostgreSQL</> 10 include:
</para>
<!-- Items in this list summarize one or more items below -->
<itemizedlist>
<listitem><para>Logical replication using publish/subscribe</para></listitem>
<listitem><para>Declarative table partitioning</para></listitem>
<listitem><para>Improved query parallelism</para></listitem>
<listitem><para>Significant general performance improvements</para></listitem>
<listitem><para>Stronger password authentication based on SCRAM-SHA-256</para></listitem>
<listitem><para>Improved monitoring and control</para></listitem>
</itemizedlist>
<para>
The above items are explained in more detail in the sections below.
</para>
</sect2>
<sect2>
<title>Migration to Version 10</title>
<para>
A dump/restore using <xref linkend="app-pg-dumpall">, or use of <xref
linkend="pgupgrade">, is required for those wishing to migrate data
from any previous release.
</para>
<para>
Version 10 contains a number of changes that may affect compatibility
with previous releases. Observe the following incompatibilities:
</para>
<itemizedlist>
<listitem>
<!--
2017-04-03 [ea69a0dea] Expand hash indexes more gradually.
2017-05-19 [a95410e2e] pg_upgrade: Handle hash index upgrades more smoothly.
2017-08-04 [620b49a16] hash: Increase the number of possible overflow bitmaps b
-->
<para>
Hash indexes must be rebuilt after <application>pg_upgrade</>-ing
from any previous major <productname>PostgreSQL</> version (Mithun
Cy, Robert Haas, Amit Kapila)
</para>
<para>
Major hash index improvements necessitated this requirement.
<application>pg_upgrade</> will create a script to assist with this.
</para>
</listitem>
<listitem>
<!--
2016-10-20 [f82ec32ac] Rename "pg_xlog" directory to "pg_wal"
2017-03-17 [88e66d193] Rename "pg_clog" directory to "pg_xact".
-->
<para>
Rename write-ahead log directory <filename>pg_xlog</>
to <link linkend="wal"><filename>pg_wal</></>, and rename transaction
status directory <filename>pg_clog</> to <filename>pg_xact</>
(Michael Paquier)
</para>
<para>
Users have occasionally thought that these directories contained only
inessential log files, and proceeded to remove write-ahead log files
or transaction status files manually, causing irrecoverable data
loss. These name changes are intended to discourage such errors in
future.
</para>
</listitem>
<listitem>
<!--
2017-02-09 [806091c96] Remove all references to "xlog" from SQL-callable functi
2017-02-09 [85c11324c] Rename user-facing tools with "xlog" in the name to say
2017-02-09 [62e8b3875] Rename command line options for ongoing xlog -> wal conv
2017-02-15 [0dfa89ba2] Replace reference to "xlog-method" with "wal-method" in
-->
<para>
Rename <acronym>SQL</> functions, tools, and options that reference
<quote>xlog</> to <quote>wal</> (Robert Haas)
</para>
<para>
For example, <function>pg_switch_xlog()</> becomes
<function>pg_switch_wal()</>, <application>pg_receivexlog</>
becomes <application>pg_receivewal</>, and <option>--xlogdir</>
becomes <option>--waldir</>. This is for consistency with the
change of the <filename>pg_xlog</> directory name; in general,
the <quote>xlog</> terminology is no longer used in any user-facing
places.
</para>
</listitem>
<listitem>
<!--
2017-05-11 [d10c626de] Rename WAL-related functions and views to use "lsn" not
-->
<para>
Rename <acronym>WAL</>-related functions and views to use <literal>lsn</>
instead of <literal>location</> (David Rowley)
</para>
<para>
There was previously an inconsistent mixture of the two terminologies.
</para>
</listitem>
<listitem>
<!--
2017-01-16 [d43a619c6] Fix check_srf_call_placement() to handle VALUES cases co
2017-01-18 [69f4b9c85] Move targetlist SRF handling from expression evaluation
2017-01-18 [f13a1277a] Doc: improve documentation of new SRF-in-tlist behavior.
2017-06-13 [0436f6bde] Disallow set-returning functions inside CASE or COALESCE
-->
<para>
Change the implementation of set-returning functions appearing in
a query's <literal>SELECT</> list (Andres Freund)
</para>
<para>
Set-returning functions are now evaluated before evaluation of scalar
expressions in the <literal>SELECT</> list, much as though they had
been placed in a <literal>LATERAL FROM</>-clause item. This allows
saner semantics for cases where multiple set-returning functions are
present. If they return different numbers of rows, the shorter results
are extended to match the longest result by adding nulls. Previously
the results were cycled until they all terminated at the same time,
producing a number of rows equal to the least common multiple of the
functions' periods. In addition, set-returning functions are now
disallowed within <literal>CASE</> and <literal>COALESCE</> constructs.
For more information
see <xref linkend="xfunc-sql-functions-returning-set">.
</para>
</listitem>
<listitem>
<!--
2016-12-20 [1753b1b02] Add pg_sequence system catalog
2016-11-18 [67dc4ccbb] Add pg_sequences view
2017-05-15 [f8dc1985f] Fix ALTER SEQUENCE locking
2017-06-01 [3d79013b9] Make ALTER SEQUENCE, including RESTART, fully transactio
-->
<para>
Move sequences' metadata fields into a new <link
linkend="catalog-pg-sequence"><structname>pg_sequence</></>
system catalog (Peter Eisentraut)
</para>
<para>
A sequence relation now stores only the fields that can be modified
by <function>nextval()</>, that
is <structfield>last_value</>, <structfield>log_cnt</>,
and <structfield>is_called</>. Other sequence properties, such as
the starting value and increment, are kept in a corresponding row of
the <structname>pg_sequence</> catalog.
<command>ALTER SEQUENCE</> updates are now fully transactional,
implying that the sequence is locked until commit.
The <function>nextval()</> and <function>setval()</> functions
remain nontransactional.
</para>
<para>
The main incompatibility introduced by this change is that selecting
from a sequence relation now returns only the three fields named
above. To obtain the sequence's other properties, applications must
look into <structname>pg_sequence</>. The new system
view <link linkend="view-pg-sequences"><structname>pg_sequences</></>
can also be used for this purpose; it provides column names that are
more compatible with existing code.
</para>
</listitem>
<listitem>
<!--
2017-01-04 [9a4d51077] Make wal streaming the default mode for pg_basebackup
-->
<para>
Make <application><xref linkend="app-pgbasebackup"></> stream the
<acronym>WAL</> needed to restore the backup by default (Magnus
Hagander)
</para>
<para>
This changes <application>pg_basebackup</>'s
<option>-X</>/<option>--xlog-method</> default to <literal>stream</>.
An option value <literal>none</> has been added to reproduce the old
behavior. The <application>pg_basebackup</> option <option>-x</>
has been removed (instead, use <literal>-X fetch</>).
</para>
</listitem>
<listitem>
<!--
2017-02-13 [8df9bd0b4] Change logical replication pg_hba.conf use
-->
<para>
Change how logical replication
uses <link linkend="auth-pg-hba-conf"><filename>pg_hba.conf</filename></link>
(Peter Eisentraut)
</para>
<para>
In previous releases, a logical replication connection required
the <literal>replication</literal> keyword in the database column. As
of this release, logical replication matches a normal entry with a
database name or keywords such as <literal>all</literal>. Physical
replication continues to use the <literal>replication</literal> keyword.
Since built-in logical replication is new in this release, this
change only affects users of third-party logical replication plugins.
</para>
</listitem>
<listitem>
<!--
2017-01-14 [05cd12ed5] pg_ctl: Change default to wait for all actions
-->
<para>
Make all <application><xref linkend="app-pg-ctl"></> actions wait
for completion by default (Peter Eisentraut)
</para>
<para>
Previously some <application>pg_ctl</> actions didn't wait for
completion, and required the use of <option>-w</> to do so.
</para>
</listitem>
<listitem>
<!--
2017-05-08 [eb61136dc] Remove support for password_encryption='off' / 'plain'.
-->
<para>
Remove the ability to store unencrypted passwords on the server
(Heikki Linnakangas)
</para>
<para>
The <xref linkend="guc-password-encryption"> server parameter
no longer supports <literal>off</> or <literal>plain</>.
The <literal>UNENCRYPTED</> option is no longer supported in
<command>CREATE/ALTER USER ... PASSSWORD</>. Similarly, the
<option>--unencrypted</> option has been removed
from <application>createuser</>. Unencrypted passwords migrated from
older versions will be stored encrypted in this release. The default
setting for <varname>password_encryption</> is still
<literal>md5</>.
</para>
</listitem>
<listitem>
<!--
2016-10-26 [94aceed31] Support multi-dimensional arrays in PL/python.
2016-10-26 [cfd9c87a5] Only treat Python Lists as array dimensions.
-->
<para>
Allow multi-dimensional arrays to be passed into PL/Python functions,
and returned as nested Python lists (Alexey Grishchenko, Dave Cramer,
Heikki Linnakangas)
</para>
<para>
This feature requires a backwards-incompatible change to the handling
of arrays of composite types in PL/Python. Previously, you could
return an array of composite values by writing, e.g., <literal>[[col1,
col2], [col1, col2]]</>; but now that is interpreted as a
two-dimensional array. Composite types in arrays must now be written
as Python tuples, not lists, to resolve the ambiguity; that is,
write <literal>[(col1, col2), (col1, col2)]</> instead.
</para>
</listitem>
<listitem>
<!--
2017-02-27 [817f2a586] Remove PL/Tcl's "module" facility.
-->
<para>
Remove PL/Tcl's <quote>module</> auto-loading facility (Tom Lane)
</para>
<para>
This functionality has been replaced by new server
parameters <xref linkend="guc-pltcl-start-proc">
and <xref linkend="guc-pltclu-start-proc">, which are easier to use
and more similar to features available in other PLs.
</para>
</listitem>
<listitem>
<!--
2017-03-27 [3371e4d9b] Change default of log_directory to 'log'
-->
<para>
Change the default value of the <xref linkend="guc-log-directory">
server parameter from <filename>pg_log</> to <filename>log</>
(Andreas Karlsson)
</para>
</listitem>
<listitem>
<!--
2016-12-23 [e13486eba] Remove sql_inheritance GUC.
-->
<para>
Remove <varname>sql_inheritance</> server parameter (Robert Haas)
</para>
<para>
Changing this setting from the default value caused queries referencing
parent tables to not include child tables. The <acronym>SQL</>
standard requires them to be included, however, and this has been the
default since <productname>PostgreSQL</> 7.1.
</para>
</listitem>
<listitem>
<!--
2017-02-15 [51ee6f316] Replace min_parallel_relation_size with two new GUCs.
-->
<para>
Add <xref linkend="guc-min-parallel-table-scan-size">
and <xref linkend="guc-min-parallel-index-scan-size"> server
parameters to control parallel queries (Amit Kapila, Robert Haas)
</para>
<para>
These replace <varname>min_parallel_relation_size</>, which was
found to be too generic.
</para>
</listitem>
<listitem>
<!--
2017-06-20 [a69dfe5f4] Don't downcase entries within shared_preload_libraries e
-->
<para>
Don't downcase unquoted text
within <xref linkend="guc-shared-preload-libraries"> and related
server parameters (QL Zhuo)
</para>
<para>
These settings are really lists of file names, but they were
previously treated as lists of SQL identifiers, which have different
parsing rules.
</para>
</listitem>
<listitem>
<!--
2016-10-12 [64f3524e2] Remove pg_dump/pg_dumpall support for dumping from pre-8
-->
<para>
Remove <application>pg_dump</>/<application>pg_dumpall</> support
for dumping from pre-8.0 servers (Tom Lane)
</para>
<para>
Users needing to dump from pre-8.0 servers will need to use dump
programs from <productname>PostgreSQL</> 9.6 or earlier. The
resulting output should still load successfully into newer servers.
</para>
</listitem>
<listitem>
<!--
2017-02-23 [b6aa17e0a] De-support floating-point timestamps.
-->
<para>
Remove support for floating-point timestamps and intervals (Tom Lane)
</para>
<para>
This removes configure's <option>--disable-integer-datetimes</>
option. Floating-point timestamps have few advantages and have not
been the default since <productname>PostgreSQL</> 8.3.
</para>
</listitem>
<listitem>
<!--
2016-10-11 [2f1eaf87e] Drop server support for FE/BE protocol version 1.0.
-->
<para>
Remove server support for client/server protocol version 1.0 (Tom Lane)
</para>
<para>
This protocol hasn't had client support
since <productname>PostgreSQL</> 6.3.
</para>
</listitem>
<listitem>
<!--
2017-07-31 [c0a15e07c] Always use 2048 bit DH parameters for OpenSSL ephemeral
-->
<para>
Add configuration option <xref linkend="guc-ssl-dh-params-file"> to
specify file name for custom OpenSSL DH parameters (Heikki Linnakangas)
</para>
<para>
This replaces the hardcoded, undocumented file
name <filename>dh1024.pem</>. Note that <filename>dh1024.pem</> is
no longer examined by default; you must set this option if you want
to use custom DH parameters.
</para>
</listitem>
<listitem>
<!--
2017-07-31 [c0a15e07c] Always use 2048 bit DH parameters for OpenSSL ephemeral
-->
<para>
Increase the size of the default DH parameters used for OpenSSL
ephemeral DH ciphers to 2048 bits (Heikki Linnakangas)
</para>
<para>
The size of the compiled-in DH parameters has been increased from
1024 to 2048 bits, making DH key exchange more resistant to
brute-force attacks. However, some old SSL implementations, notably
some revisions of Java Runtime Environment version 6, will not accept
DH parameters longer than 1024 bits, and hence will not be able to
connect over SSL. If it's necessary to support such old clients, you
can use custom 1024-bit DH parameters instead of the compiled-in
defaults. See <xref linkend="guc-ssl-dh-params-file">.
</para>
</listitem>
<listitem>
<!--
2017-08-04 [c30f1770a] Apply ALTER ... SET NOT NULL recursively in ALTER ... AD
-->
<para>
When <command>ALTER TABLE ... ADD PRIMARY KEY</> marks
columns <literal>NOT NULL</>, that change now propagates to
inheritance child tables as well (Michael Paquier)
</para>
</listitem>
<listitem>
<!--
2017-02-13 [7ada2d31f] Remove contrib/tsearch2.
-->
<para>
Remove <filename>contrib/tsearch2</> module (Robert Haas)
</para>
<para>
This module provided compatibility with the version of full text
search that shipped in pre-8.3 <productname>PostgreSQL</> releases.
</para>
</listitem>
<listitem>
<!--
2017-03-23 [50c956add] Remove createlang and droplang
-->
<para>
Remove <application>createlang</> and <application>droplang</>
command-line applications (Peter Eisentraut)
</para>
<para>
These had been deprecated since <productname>PostgreSQL</> 9.1.
Instead, use <command>CREATE EXTENSION</> and <command>DROP
EXTENSION</> directly.
</para>
</listitem>
<listitem>
<!--
2017-03-30 [5ded4bd21] Remove support for version-0 calling conventions.
-->
<para>
Remove support for version-0 function calling conventions (Andres
Freund)
</para>
<para>
Extensions providing C-coded functions must now conform to version 1
calling conventions. Version 0 has been deprecated since 2001.
</para>
</listitem>
</itemizedlist>
</sect2>
<sect2>
<title>Changes</title>
<para>
Below you will find a detailed account of the changes between
<productname>PostgreSQL</productname> 10 and the previous major
release.
</para>
<sect3>
<title>Server</title>
<sect4>
<title>Parallel Queries</title>
<itemizedlist>
<listitem>
<!--
2017-02-15 [569174f1b] btree: Support parallel index scans.
2017-02-15 [5262f7a4f] Add optimizer and executor support for parallel index sc
2017-02-19 [0414b26ba] Add optimizer and executor support for parallel index-on
-->
<para>
Support parallel B-tree index scans (Rahila Syed, Amit Kapila,
Robert Haas, Rafia Sabih)
</para>
<para>
This change allows B-tree index pages to be searched by separate
parallel workers.
</para>
</listitem>
<listitem>
<!--
2017-03-08 [98e6e8904] tidbitmap: Support shared iteration.
2017-03-08 [f35742ccb] Support parallel bitmap heap scans.
-->
<para>
Support parallel bitmap heap scans (Dilip Kumar)
</para>
<para>
This allows a single index scan to dispatch parallel workers to
process different areas of the heap.
</para>
</listitem>
<listitem>
<!--
2017-03-07 [3bc7dafa9] Consider parallel merge joins.
-->
<para>
Allow merge joins to be performed in parallel (Dilip Kumar)
</para>
</listitem>
<listitem>
<!--
2017-02-14 [5e6d8d2bb] Allow parallel workers to execute subplans.
-->
<para>
Allow non-correlated subqueries to be run in parallel (Amit Kapila)
</para>
</listitem>
<listitem>
<!--
2017-03-09 [355d3993c] Add a Gather Merge executor node.
-->
<para>
Improve ability of parallel workers to return pre-sorted data
(Rushabh Lathia)
</para>
</listitem>
<listitem>
<!--
2017-03-24 [61c2e1a95] Improve access to parallel query from procedural languag
-->
<para>
Increase parallel query usage in procedural language functions
(Robert Haas, Rafia Sabih)
</para>
</listitem>
<listitem>
<!--
2016-12-02 [b460f5d66] Add max_parallel_workers GUC.
2016-12-05 [2b959d495] Reduce the default for max_worker_processes back to 8.
-->
<para>
Add <xref linkend="guc-max-parallel-workers"> server parameter
to limit the number of worker processes that can be used for
query parallelism (Julien Rouhaud)
</para>
<para>
This parameter can be set lower than <xref
linkend="guc-max-worker-processes"> to reserve worker processes
for purposes other than parallel queries.
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>Indexes</title>
<itemizedlist>
<listitem>
<!--
2017-02-27 [30df93f69] hash: Refactor overflow page allocation.
2017-03-14 [c11453ce0] hash: Add write-ahead logging support.
2017-02-27 [b0f18cb77] hash: Refactor bucket squeeze code.
-->
<para>
Add write-ahead logging support to hash indexes (Amit Kapila)
</para>
<para>
This makes hash indexes crash-safe and replicatable.
The former warning message about their use is removed.
</para>
</listitem>
<listitem>
<!--
2016-11-30 [6d46f4783] Improve hash index bucket split behavior.
2017-02-07 [293e24e50] Cache hash index's metapage in rel->rd_amcache.
2017-03-15 [6977b8b7f] Port single-page btree vacuum logic to hash indexes.
2017-04-03 [ea69a0dea] Expand hash indexes more gradually.
2017-08-04 [620b49a16] hash: Increase the number of possible overflow bitmaps b
-->
<para>
Improve hash index performance (Amit Kapila, Mithun Cy, Ashutosh
Sharma)
</para>
</listitem>
<listitem>
<!--
2016-08-23 [77e290682] Create an SP-GiST opclass for inet/cidr.
-->
<para>
Add <acronym>SP-GiST</> index support for <type>INET</> and
<type>CIDR</> data types (Emre Hasegeli)
</para>
</listitem>
<listitem>
<!--
2017-04-01 [7526e1022] BRIN auto-summarization
-->
<para>
Add option to allow <acronym>BRIN</> index summarization to happen
more aggressively (&Aacute;lvaro Herrera)
</para>
<para>
A new <link linkend="SQL-CREATEINDEX"><command>CREATE
INDEX</></> option enables auto-summarization of the
previous <acronym>BRIN</> page range when a new page
range is created.
</para>
</listitem>
<listitem>
<!--
2017-04-01 [c655899ba] BRIN de-summarization
-->
<para>
Add functions to remove and re-add <acronym>BRIN</>
summarization for <acronym>BRIN</> index ranges (&Aacute;lvaro
Herrera)
</para>
<para>
The new <acronym>SQL</> function <link
linkend="functions-admin-index-table"><function>brin_summarize_range()</></>
updates <acronym>BRIN</> index summarization for a specified
range and <function>brin_desummarize_range()</> removes it.
This is helpful to update summarization of a range that is now
smaller due to <command>UPDATE</>s and <command>DELETE</>s.
</para>
</listitem>
<listitem>
<!--
2017-04-06 [7e534adcd] Fix BRIN cost estimation
-->
<para>
Improve accuracy in determining if a <acronym>BRIN</> index scan
is beneficial (David Rowley, Emre Hasegeli)
</para>
</listitem>
<listitem>
<!--
2016-09-09 [b1328d78f] Invent PageIndexTupleOverwrite, and teach BRIN and GiST
-->
<para>
Allow faster <acronym>GiST</> inserts and updates by reusing
index space more efficiently (Andrey Borodin)
</para>
</listitem>
<listitem>
<!--
2017-03-23 [218f51584] Reduce page locking in GIN vacuum
-->
<para>
Reduce page locking during vacuuming of <acronym>GIN</> indexes
(Andrey Borodin)
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>Locking</title>
<itemizedlist>
<listitem>
<!--
2016-09-05 [15bc038f9] Relax transactional restrictions on ALTER TYPE ... ADD V
-->
<para>
Reduce locking required for adding values to enum types (Andrew
Dunstan, Tom Lane)
</para>
<para>
Previously it was impossible to run <command>ALTER TYPE ... ADD
VALUE</> in a transaction block unless the enum type was created
in the same block. Now, only references to uncommitted enum
values from other transactions are prohibited.
</para>
</listitem>
<listitem>
<!--
2017-03-06 [21d4e2e20] Reduce lock levels for table storage params related to p
2017-04-05 [68ea2b7f9] Reduce lock level for CREATE STATISTICS
-->
<para>
Reduce locking required to change table parameters (Simon Riggs,
Fabr&iacute;zio Mello)
</para>
<para>
For example, changing a table's <xref
linkend="guc-effective-io-concurrency"> setting can now be done
with a more lightweight lock.
</para>
</listitem>
<listitem>
<!--
2017-04-07 [c63172d60] Add GUCs for predicate lock promotion thresholds.
-->
<para>
Allow tuning of predicate lock promotion thresholds (Dagfinn
Ilmari Manns&aring;ker)
</para>
<para>
Lock promotion can now be controlled through two new server
parameters, <xref
linkend="guc-max-pred-locks-per-relation"> and
<xref linkend="guc-max-pred-locks-per-page">.
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>Optimizer</title>
<itemizedlist>
<listitem>
<!--
2017-03-24 [7b504eb28] Implement multivariate n-distinct coefficients
2017-04-05 [2686ee1b7] Collect and use multi-column dependency stats
2017-05-12 [bc085205c] Change CREATE STATISTICS syntax
-->
<para>
Add multi-column optimizer statistics to compute the correlation
ratio and number of distinct values (Tomas Vondra, David Rowley,
&Aacute;lvaro Herrera)
</para>
<para>
New commands are <link
linkend="SQL-CREATESTATISTICS"><command>CREATE STATISTICS</></>,
<link linkend="SQL-ALTERSTATISTICS"><command>ALTER STATISTICS</></>, and
<link linkend="SQL-DROPSTATISTICS"><command>DROP STATISTICS</></>.
This feature is helpful in estimating query memory usage and when
combining the statistics from individual columns.
</para>
</listitem>
<listitem>
<!--
2017-01-18 [215b43cdc] Improve RLS planning by marking individual quals with se
-->
<para>
Improve performance of queries affected by row-level security
restrictions (Tom Lane)
</para>
<para>
The optimizer now has more knowledge about where it can place RLS
filter conditions, allowing better plans to be generated while still
enforcing the RLS conditions safely.
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>General Performance</title>
<itemizedlist>
<listitem>
<!--
2016-09-02 [9cca11c91] Speed up SUM calculation in numeric aggregates.
-->
<para>
Speed up aggregate functions that calculate a running sum
using <type>numeric</>-type arithmetic, including some variants
of <function>SUM()</>, <function>AVG()</>,
and <function>STDDEV()</> (Heikki Linnakangas)
</para>
</listitem>
<listitem>
<!--
2017-03-13 [aeed17d00] Use radix tree for character encoding conversions.
-->
<para>
Improve performance of character encoding conversions by
using radix trees (Kyotaro Horiguchi, Heikki Linnakangas)
</para>
</listitem>
<listitem>
<!--
2017-03-25 [b8d7f053c] Faster expression evaluation and targetlist projection.
2017-07-30 [cc9f08b6b] Move ExecProcNode from dispatch to function pointer base
-->
<para>
Reduce expression evaluation overhead during query execution,
as well as plan node calling overhead (Andres Freund)
</para>
<para>
This is particularly helpful for queries that process many rows.
</para>
</listitem>
<listitem>
<!--
2017-03-27 [b5635948a] Support hashed aggregation with grouping sets.
-->
<para>
Allow hashed aggregation to be used with grouping sets (Andrew
Gierth)
</para>
</listitem>
<listitem>
<!--
2017-04-07 [9c7f5229a] Optimize joins when the inner relation can be proven uni
2017-06-03 [23886581b] Fix old corner-case logic error in final_cost_nestloop()
-->
<para>
Use uniqueness guarantees to optimize certain join types (David
Rowley)
</para>
</listitem>
<listitem>
<!--
2017-03-29 [f90d23d0c] Implement SortSupport for macaddr data type
-->
<para>
Improve sort performance of the <type>macaddr</type> data type (Brandur Leach)
</para>
</listitem>
<listitem>
<!--
2017-03-27 [090010f2e] Improve performance of find_tabstat_entry()/get_tabstat_
-->
<para>
Reduce statistics tracking overhead in sessions that reference
many thousands of relations (Aleksander Alekseev)
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>Monitoring</title>
<itemizedlist>
<listitem>
<!--
2017-03-08 [f9b1a0dd4] Expose explain's SUMMARY option
-->
<para>
Allow explicit control
over <link linkend="SQL-EXPLAIN"><command>EXPLAIN</></>'s display
of planning and execution time (Ashutosh Bapat)
</para>
<para>
By default planning and execution time are displayed by
<command>EXPLAIN ANALYZE</> and are not displayed in other cases.
The new <command>EXPLAIN</> option <literal>SUMMARY</> allows
explicit control of this.
</para>
</listitem>
<listitem>
<!--
2017-03-30 [25fff4079] Default monitoring roles
-->
<para>
Add default monitoring roles (Dave Page)
</para>
<para>
New roles <literal>pg_monitor</>, <literal>pg_read_all_settings</>,
<literal>pg_read_all_stats</>, and <literal>pg_stat_scan_tables</>
allow simplified permission configuration.
</para>
</listitem>
<listitem>
<!--
2017-03-18 [17f8ffa1e] Fix REFRESH MATERIALIZED VIEW to report activity to the
-->
<para>
Properly update the statistics collector during <link
linkend="SQL-REFRESHMATERIALIZEDVIEW"><command>REFRESH MATERIALIZED
VIEW</></> (Jim Mlodgenski)
</para>
</listitem>
</itemizedlist>
<sect5>
<title>Logging</title>
<itemizedlist>
<listitem>
<!--
2016-10-17 [7d3235ba4] By default, set log_line_prefix = '%m [%p] '.
-->
<para>
Change the default value of <xref linkend="guc-log-line-prefix">
to include current timestamp (with milliseconds) and the process ID
in each line of postmaster log output (Christoph Berg)
</para>
<para>
The previous default was an empty prefix.
</para>
</listitem>
<listitem>
<!--
2017-03-16 [befd73c50] Add pg_ls_logdir() and pg_ls_waldir() functions.
-->
<para>
Add functions to return the log and <acronym>WAL</> directory
contents (Dave Page)
</para>
<para>
The new functions
are <link linkend="functions-admin-genfile-table"><function>pg_ls_logdir()</></>
and <link linkend="functions-admin-genfile-table"><function>pg_ls_waldir()</></>
and can be executed by non-superusers with the proper
permissions.
</para>
</listitem>
<listitem>
<!--
2017-03-03 [19dc233c3] Add pg_current_logfile() function.
-->
<para>
Add function <link
linkend="functions-info-session-table"><function>pg_current_logfile()</></>
to read logging collector's current stderr and csvlog output file names
(Gilles Darold)
</para>
</listitem>
<listitem>
<!--
2017-03-10 [f9dfa5c97] Improve postmaster's logging of listen socket creation.
2017-03-14 [2b32ac2a5] Include port number when logging successful binding to a
-->
<para>
Report the address and port number of each listening socket
in the server log during postmaster startup (Tom Lane)
</para>
<para>
Also, when logging failure to bind a listening socket, include
the specific address we attempted to bind to.
</para>
</listitem>
<listitem>
<!--
2017-03-10 [6ec4c8584] Reduce log verbosity of startup/shutdown for launcher su
-->
<para>
Reduce log chatter about the starting and stopping of launcher
subprocesses (Tom Lane)
</para>
<para>
These are now <literal>DEBUG1</>-level messages.
</para>
</listitem>
<listitem>
<!--
2016-11-17 [a43f1939d] Remove or reduce verbosity of some debug messages.
-->
<para>
Reduce message verbosity of lower-numbered debug levels
controlled by
<xref linkend="guc-log-min-messages"> (Robert Haas)
</para>
<para>
This also changes the verbosity of <xref
linkend="guc-client-min-messages"> debug levels.
</para>
</listitem>
</itemizedlist>
</sect5>
<sect5>
<title><link linkend="pg-stat-activity-view"><structname>pg_stat_activity</></link></title>
<itemizedlist>
<listitem>
<!--
2016-10-04 [6f3bd98eb] Extend framework from commit 53be0b1ad to report latch w
2017-03-18 [249cf070e] Create and use wait events for read, write, and fsync op
-->
<para>
Add <structname>pg_stat_activity</> reporting of low-level wait
states (Michael Paquier, Robert Haas, Rushabh Lathia)
</para>
<para>
This change enables reporting of numerous low-level wait conditions,
including latch waits, file reads/writes/fsyncs, client reads/writes,
and synchronous replication.
</para>
</listitem>
<listitem>
<!--
2016-09-12 [fc3d4a44e] Identify walsenders in pg_stat_activity
2017-03-26 [fc70a4b0d] Show more processes in pg_stat_activity.
-->
<para>
Show auxiliary processes, background workers, and walsender
processes in <structname>pg_stat_activity</> (Kuntal Ghosh,
Michael Paquier)
</para>
<para>
This simplifies monitoring. A new
column <structfield>backend_type</> identifies the process type.
</para>
</listitem>
<listitem>
<!--
2017-02-22 [4c728f382] Pass the source text for a parallel query to the workers
-->
<para>
Allow <structname>pg_stat_activity</> to show the SQL query
being executed by parallel workers (Rafia Sabih)
</para>
</listitem>
<listitem>
<!--
2016-12-16 [3761fe3c2] Simplify LWLock tranche machinery by removing array_base
-->
<para>
Rename
<structname>pg_stat_activity</>.<structfield>wait_event_type</>
values <literal>LWLockTranche</> and
<literal>LWLockNamed</> to <literal>LWLock</> (Robert Haas)
</para>
<para>
This makes the output more consistent.
</para>
</listitem>
</itemizedlist>
</sect5>
</sect4>
<sect4>
<title><acronym>Authentication</></title>
<itemizedlist>
<listitem>
<!--
2017-03-07 [818fd4a67] Support SCRAM-SHA-256 authentication (RFC 5802 and 7677)
2017-03-24 [7ac955b34] Allow SCRAM authentication, when pg_hba.conf says 'md5'.
2017-04-07 [60f11b87a] Use SASLprep to normalize passwords for SCRAM authentica
2017-04-18 [c727f120f] Rename "scram" to "scram-sha-256" in pg_hba.conf and pas
-->
<para>
Add <link linkend="auth-pg-hba-conf"><literal>SCRAM-SHA-256</></>
support for password negotiation and storage (Michael Paquier,
Heikki Linnakangas)
</para>
<para>
This provides better security than the existing <literal>md5</>
negotiation and storage method.
</para>
</listitem>
<listitem>
<!--
2016-09-28 [babe05bc2] Turn password_encryption GUC into an enum.
-->
<para>
Change the <xref linkend="guc-password-encryption"> server parameter
from <type>boolean</> to <type>enum</> (Michael Paquier)
</para>
<para>
This was necessary to support additional password hashing options.
</para>
</listitem>
<listitem>
<!--
2017-01-30 [de16ab723] Invent pg_hba_file_rules view to show the content of pg_
-->
<para>
Add view <link
linkend="view-pg-hba-file-rules"><structname>pg_hba_file_rules</></>
to display the contents of <filename>pg_hba.conf</> (Haribabu
Kommi)
</para>
<para>
This shows the file contents, not the currently active settings.
</para>
</listitem>
<listitem>
<!--
2017-03-22 [6b76f1bb5] Support multiple RADIUS servers
-->
<para>
Support multiple <acronym>RADIUS</> servers (Magnus Hagander)
</para>
<para>
All the <acronym>RADIUS</> related parameters are now plural and
support a comma-separated list of servers.
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>Server Configuration</title>
<itemizedlist>
<listitem>
<!--
2017-01-02 [de41869b6] Allow SSL configuration to be updated at SIGHUP.
2017-01-03 [1e942c747] Disable prompting for passphrase while (re)loading SSL c
2017-01-04 [6667d9a6d] Re-allow SSL passphrase prompt at server start, but not
-->
<para>
Allow <acronym>SSL</> configuration to be updated during
configuration reload (Andreas Karlsson, Tom Lane)
</para>
<para>
This allows <acronym>SSL</> to be reconfigured without a server
restart, by using <command>pg_ctl reload</>, <command>SELECT
pg_reload_conf()</>, or sending a <literal>SIGHUP</> signal.
However, reloading the <acronym>SSL</> configuration does not work
if the server's <acronym>SSL</> key requires a passphrase, as there
is no way to re-prompt for the passphrase. The original
configuration will apply for the life of the postmaster in that
case.
</para>
</listitem>
<listitem>
<!--
2017-02-02 [14ca9abfb] Increase upper bound for bgwriter_lru_maxpages.
-->
<para>
Make the maximum value of <xref
linkend="guc-bgwriter-lru-maxpages"> effectively unlimited
(Jim Nasby)
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title>Reliability</title>
<itemizedlist>
<listitem>
<!--
2017-03-27 [1b02be21f] Fsync directory after creating or unlinking file.
-->
<para>
After creating or unlinking files, perform an fsync on their parent
directory (Michael Paquier)
</para>
<para>
This reduces the risk of data loss after a power failure.
</para>
</listitem>
</itemizedlist>
<sect5>
<title><link linkend="wal">Write-Ahead Log</> (<acronym>WAL</>)</title>
<itemizedlist>
<listitem>
<!--
2016-12-22 [6ef2eba3f] Skip checkpoints, archiving on idle systems.
-->
<para>
Prevent unnecessary checkpoints and <acronym>WAL</> archiving on
otherwise-idle systems (Michael Paquier)
</para>
</listitem>
<listitem>
<!--
2017-02-08 [a507b8690] Add WAL consistency checking facility.
2017-03-14 [bb4a39637] hash: Support WAL consistency checking.
-->
<para>
Add <xref linkend="guc-wal-consistency-checking"> server parameter
to add details to <acronym>WAL</> that can be sanity-checked on
the standby (Kuntal Ghosh, Robert Haas)
</para>
<para>
Any sanity-check failure generates a fatal error on the standby.
</para>
</listitem>
<listitem>
<!--
2017-04-05 [00b6b6feb] Allow -\-with-wal-segsize=n up to n=1024MB
-->
<para>
Increase the maximum configurable <acronym>WAL</> segment size
to one gigabyte (Beena Emerson)
</para>
<para>
A larger <acronym>WAL</> segment size allows for fewer
<xref linkend="guc-archive-command"> invocations and fewer
<acronym>WAL</> files to manage.
</para>
</listitem>
</itemizedlist>
</sect5>
</sect4>
</sect3>
<sect3>
<title>Replication and Recovery</title>
<itemizedlist>
<listitem>
<!--
2017-01-20 [665d1fad9] Logical replication
2017-03-23 [7c4f52409] Logical replication support for initial data copy
2017-04-12 [ff7bce174] Add max_sync_workers_per_subscription to postgresql.conf
-->
<para>
Add the ability to <link linkend="logical-replication">logically
replicate</> tables to standby servers (Petr Jelinek)
</para>
<para>
Logical replication allows more flexibility than physical
replication does, including replication between different major
versions of <productname>PostgreSQL</> and selective
replication.
</para>
</listitem>
<listitem>
<!--
2016-12-19 [3901fd70c] Support quorum-based synchronous replication.
-->
<para>
Allow waiting for commit acknowledgement from standby
servers irrespective of the order they appear in <xref
linkend="guc-synchronous-standby-names"> (Masahiko Sawada)
</para>
<para>
Previously the server always waited for the active standbys that
appeared first in <varname>synchronous_standby_names</>. The new
<varname>synchronous_standby_names</> keyword <literal>ANY</> allows
waiting for any number of standbys irrespective of their ordering.
This is known as quorum commit.
</para>
</listitem>
<listitem>
<!--
2017-01-14 [f6d6d2920] Change default values for backup and replication paramet
2017-05-02 [34fc61673] Change hot_standby default value to 'on'
-->
<para>
Reduce configuration changes necessary to perform streaming backup
and replication (Magnus Hagander, Dang Minh Huong)
</para>
<para>
Specifically, the defaults were changed for <xref
linkend="guc-wal-level">, <xref linkend="guc-max-wal-senders">,
<xref linkend="guc-max-replication-slots">, and <xref
linkend="guc-hot-standby"> to make them suitable for these usages
out-of-the-box.
</para>
</listitem>
<listitem>
<!--
2017-03-09 [be37c2120] Enable replication connections by default in pg_hba.conf
-->
<para>
Enable replication from localhost connections by default in
<link linkend="auth-pg-hba-conf"><filename>pg_hba.conf</></>
(Michael Paquier)
</para>
<para>
Previously <filename>pg_hba.conf</>'s replication connection
lines were commented out by default. This is particularly useful for
<application><xref linkend="app-pgbasebackup"></>.
</para>
</listitem>
<listitem>
<!--
2017-03-23 [6912acc04] Replication lag tracking for walsenders
-->
<para>
Add columns to <link
linkend="monitoring-stats-views-table"><structname>pg_stat_replication</></>
to report replication delay times (Thomas Munro)
</para>
<para>
The new columns are <structfield>write_lag</>,
<structfield>flush_lag</>, and <structfield>replay_lag</>.
</para>
</listitem>
<listitem>
<!--
2016-09-03 [35250b6ad] New recovery target recovery_target_lsn
-->
<para>
Allow specification of the recovery stopping point by Log Sequence
Number (<acronym>LSN</>) in
<link linkend="recovery-config"><filename>recovery.conf</></>
(Michael Paquier)
</para>
<para>
Previously the stopping point could only be selected by timestamp or
XID.
</para>
</listitem>
<listitem>
<!--
2017-03-22 [017e4f258] Expose waitforarchive option through pg_stop_backup()
2017-08-05 [52f8a59dd] Make pg_stop_backup's wait_for_archive flag work on stan
-->
<para>
Allow users to disable <link
linkend="functions-admin"><function>pg_stop_backup()</></>'s
waiting for all <acronym>WAL</> to be archived (David Steele)
</para>
<para>
An optional second argument to <function>pg_stop_backup()</>
controls that behavior.
</para>
</listitem>
<listitem>
<!--
2016-12-12 [a924c327e] Add support for temporary replication slots
-->
<para>
Allow creation of <link
linkend="functions-replication-table">temporary replication slots</>
(Petr Jelinek)
</para>
<para>
Temporary slots are automatically removed on session exit or error.
</para>
</listitem>
<listitem>
<!--
2017-03-22 [9b013dc23] Improve performance of replay of AccessExclusiveLocks
-->
<para>
Improve performance of hot standby replay with better tracking of
Access Exclusive locks (Simon Riggs, David Rowley)
</para>
</listitem>
<listitem>
<!--
2017-04-04 [728bd991c] Speedup 2PC recovery by skipping two phase state files i
-->
<para>
Speed up two-phase commit recovery performance (Stas Kelvich,
Nikhil Sontakke, Michael Paquier)
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Queries</title>
<itemizedlist>
<listitem>
<!--
2017-03-08 [fcec6caaf] Support XMLTABLE query expression
-->
<para>
Add <link
linkend="functions-xml-processing-xmltable"><function>XMLTABLE</></>
function that converts <type>XML</>-formatted data into a row set
(Pavel Stehule, &Aacute;lvaro Herrera)
</para>
</listitem>
<listitem>
<!--
2016-11-22 [906bfcad7] Improve handling of "UPDATE ... SET (column_list) = row_
-->
<para>
Allow standard row constructor syntax in <literal>UPDATE ... SET
(<replaceable>column_list</>) = <replaceable>row_constructor</></literal>
(Tom Lane)
</para>
<para>
The <replaceable>row_constructor</> can now begin with the
keyword <literal>ROW</>; previously that had to be omitted. Also,
an occurrence of <literal><replaceable>table_name</>.*</literal>
within the <replaceable>row_constructor</> is now expanded into
multiple columns, as in other uses
of <replaceable>row_constructor</>s.
</para>
</listitem>
<listitem>
<!--
2016-09-05 [c54159d44] Make locale-dependent regex character classes work for l
-->
<para>
Fix regular expressions' character class handling for large character
codes, particularly Unicode characters above <literal>U+7FF</>
(Tom Lane)
</para>
<para>
Previously, such characters were never recognized as belonging to
locale-dependent character classes such as <literal>[[:alpha:]]</>.
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Utility Commands</title>
<itemizedlist>
<listitem>
<!--
2016-12-07 [f0e44751d] Implement table partitioning.
-->
<para>
Add table <link linkend="SQL-CREATETABLE-PARTITION">partitioning
syntax</> that automatically creates partition constraints and
handles routing of tuple insertions and updates (Amit Langote)
</para>
<para>
The syntax supports range and list partitioning.
</para>
</listitem>
<listitem>
<!--
2016-11-04 [8c48375e5] Implement syntax for transition tables in AFTER triggers
2017-04-04 [5ebeb579b] Follow-on cleanup for the transition table patch.
2017-03-31 [597027163] Add transition table support to plpgsql.
-->
<para>
Add <link linkend="SQL-CREATETRIGGER"><literal>AFTER</> trigger</>
transition tables to record changed rows (Kevin Grittner)
</para>
<para>
Transition tables are accessible from triggers written in
server-side languages.
</para>
</listitem>
<listitem>
<!--
2016-12-05 [093129c9d] Add support for restrictive RLS policies
-->
<para>
Allow <link linkend="SQL-CREATEPOLICY">restrictive row-level
security policies</> (Stephen Frost)
</para>
<para>
Previously all security policies were permissive, meaning that any
matching policy allowed access. A restrictive policy must
match for access to be granted. These policy types can be combined.
</para>
</listitem>
<listitem>
<!--
2017-03-31 [64d4da511] For foreign keys, check REFERENCES privilege only on the
-->
<para>
When creating a foreign-key constraint, check
for <literal>REFERENCES</> permission on only the referenced table
(Tom Lane)
</para>
<para>
Previously <literal>REFERENCES</> permission on the referencing
table was also required. This appears to have stemmed from a
misreading of the SQL standard. Since creating a foreign key (or
any other type of) constraint requires ownership privilege on the
constrained table, additionally requiring <literal>REFERENCES</>
permission seems rather pointless.
</para>
</listitem>
<listitem>
<!--
2017-03-28 [ab89e465c] Altering default privileges on schemas
-->
<para>
Allow <link linkend="SQL-ALTERDEFAULTPRIVILEGES">default
permissions</> on schemas (Matheus Oliveira)
</para>
<para>
This is done using the <literal>ALTER DEFAULT PRIVILEGES</> command.
</para>
</listitem>
<listitem>
<!--
2017-02-10 [2ea5b06c7] Add CREATE SEQUENCE AS <data type> clause
-->
<para>
Add <link linkend="SQL-CREATESEQUENCE"><command>CREATE SEQUENCE
AS</></> command to create a sequence matching an integer data type
(Peter Eisentraut)
</para>
<para>
This simplifies the creation of sequences matching the range of
base columns.
</para>
</listitem>
<listitem>
<!--
2016-11-10 [279c439c7] Support "COPY view FROM" for views with INSTEAD OF INSER
-->
<para>
Allow <literal>COPY <replaceable>view</>
FROM <replaceable>source</></> on views with <literal>INSTEAD
INSERT</> triggers (Haribabu Kommi)
</para>
<para>
The triggers are fed the data rows read by <command>COPY</>.
</para>
</listitem>
<listitem>
<!--
2017-03-14 [aefeb6874] Allow referring to functions without arguments when uniq
-->
<para>
Allow the specification of a function name without arguments in
<acronym>DDL</> commands, if it is unique (Peter Eisentraut)
</para>
<para>
For example, allow <link linkend="SQL-DROPFUNCTION"><command>DROP
FUNCTION</></> on a function name without arguments if there
is only one function with that name. This behavior is required by the
<acronym>SQL</> standard.
</para>
</listitem>
<listitem>
<!--
2017-03-06 [583f6c414] Allow dropping multiple functions at once
-->
<para>
Allow multiple functions, operators, and aggregates to be dropped
with a single <command>DROP</> command (Peter Eisentraut)
</para>
</listitem>
<listitem>
<!--
2017-02-15 [6d16ecc64] Add CREATE COLLATION IF NOT EXISTS clause
2017-03-20 [b6fb534f1] Add IF NOT EXISTS for CREATE SERVER and CREATE USER MAPP
-->
<para>
Support <literal>IF NOT EXISTS</>
in <link linkend="SQL-CREATESERVER"><command>CREATE SERVER</></>,
<link linkend="SQL-CREATEUSERMAPPING"><command>CREATE USER MAPPING</></>,
and <link linkend="SQL-CREATECOLLATION"><command>CREATE COLLATION</></>
(Anastasia Lubennikova, Peter Eisentraut)
</para>
</listitem>
<listitem>
<!--
2017-03-25 [70adf2fbe] Make VACUUM VERBOSE report the number of skipped frozen
2017-03-03 [9eb344faf] Allow vacuums to report oldestxmin
-->
<para>
Make <link linkend="SQL-VACUUM"><command>VACUUM VERBOSE</></> report
the number of skipped frozen pages and oldest xmin (Masahiko
Sawada, Simon Riggs)
</para>
<para>
This information is also included in <xref
linkend="guc-log-autovacuum-min-duration"> output.
</para>
</listitem>
<listitem>
<!--
2017-01-23 [7e26e02ee] Prefetch blocks during lazy vacuum's truncation scan
-->
<para>
Improve speed of <command>VACUUM</>'s removal of trailing empty
heap pages (Claudio Freire, &Aacute;lvaro Herrera)
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Data Types</title>
<itemizedlist>
<listitem>
<!--
2017-03-31 [e306df7f9] Full Text Search support for JSON and JSONB
-->
<para>
Add full text search support for <type>JSON</> and <type>JSONB</>
(Dmitry Dolgov)
</para>
<para>
The functions <function>ts_headline()</> and
<function>to_tsvector()</> can now be used on these data types.
</para>
</listitem>
<listitem>
<!--
2017-03-15 [c7a9fa399] Add support for EUI-64 MAC addresses as macaddr8
-->
<para>
Add support for <acronym>EUI-64</> <acronym>MAC</> addresses, as a
new data type <link linkend="datatype-macaddr8"><type>macaddr8</></>
(Haribabu Kommi)
</para>
<para>
This complements the existing support
for <acronym>EUI-48</> <acronym>MAC</> addresses
(type <type>macaddr</>).
</para>
</listitem>
<listitem>
<!--
2017-04-06 [321732705] Identity columns
-->
<para>
Add <link linkend="SQL-CREATETABLE">identity columns</> for
assigning a numeric value to columns on insert (Peter Eisentraut)
</para>
<para>
These are similar to <type>SERIAL</> columns, but are
<acronym>SQL</> standard compliant.
</para>
</listitem>
<listitem>
<!--
2016-09-07 [0ab9c56d0] Support renaming an existing value of an enum type.
-->
<para>
Allow <link linkend="datatype-enum"><type>ENUM</></> values to be
renamed (Dagfinn Ilmari Manns&aring;ker)
</para>
<para>
This uses the syntax <link linkend="SQL-ALTERTYPE"><command>ALTER
TYPE ... RENAME VALUE</></>.
</para>
</listitem>
<listitem>
<!--
2017-02-22 [502a3832c] Correctly handle array pseudotypes in to_json and to_jso
-->
<para>
Properly treat array pseudotypes
(<type>anyarray</>) as arrays in <link
linkend="functions-json-creation-table"><function>to_json()</></>
and <function>to_jsonb()</> (Andrew Dunstan)
</para>
<para>
Previously columns declared as <type>anyarray</> (particularly those
in the <structname>pg_stats</> view) were converted to <type>JSON</>
strings rather than arrays.
</para>
</listitem>
<listitem>
<!--
2017-01-17 [323b96aa3] Register missing money operators in system catalogs
-->
<para>
Add operators for multiplication and division
of <link linkend="datatype-money"><type>money</></link> values
with <type>int8</> values (Peter Eisentraut)
</para>
<para>
Previously such cases would result in converting the <type>int8</>
values to <type>float8</> and then using
the <type>money</>-and-<type>float8</> operators. The new behavior
avoids possible precision loss. But note that division
of <type>money</> by <type>int8</> now truncates the quotient, like
other integer-division cases, while the previous behavior would have
rounded.
</para>
</listitem>
<listitem>
<!--
2016-09-14 [656df624c] Add overflow checks to money type input function
-->
<para>
Check for overflow in the <type>money</> type's input function
(Peter Eisentraut)
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Functions</title>
<itemizedlist>
<listitem>
<!--
2016-08-17 [cf9b0fea5] Implement regexp_match(), a simplified alternative to re
-->
<para>
Add simplified <link
linkend="functions-posix-regexp"><function>regexp_match()</></>
function (Emre Hasegeli)
</para>
<para>
This is similar to <function>regexp_matches()</>, but it only
returns results from the first match so it does not need to return a
set, making it easier to use for simple cases.
</para>
</listitem>
<listitem>
<!--
2017-01-18 [d00ca333c] Implement array version of jsonb_delete and operator
-->
<para>
Add a version of <type>jsonb</>'s <link
linkend="functions-jsonb-op-table">delete operator</> that takes
an array of keys to delete (Magnus Hagander)
</para>
</listitem>
<listitem>
<!--
2017-04-06 [cf35346e8] Make json_populate_record and friends operate recursivel
-->
<para>
Make <link linkend="functions-json-processing-table"><function>json_populate_record()</></>
and related functions process JSON arrays and objects recursively
(Nikita Glukhov)
</para>
<para>
With this change, array-type fields in the destination SQL type are
properly converted from JSON arrays, and composite-type fields are
properly converted from JSON objects. Previously, such cases would
fail because the text representation of the JSON value would be fed
to <function>array_in()</> or <function>record_in()</>, and its
syntax would not match what those input functions expect.
</para>
</listitem>
<listitem>
<!--
2016-08-23 [86f31695f] Add txid_current_ifassigned().
-->
<para>
Add function <link
linkend="functions-txid-snapshot"><function>txid_current_ifassigned()</></>
to return the current transaction ID or <literal>NULL</> if no
transaction ID has been assigned (Craig Ringer)
</para>
<para>
This is different from <link
linkend="functions-txid-snapshot"><function>txid_current()</></>,
which always returns a transaction ID, assigning one if necessary.
Unlike that function, this function can be run on standby servers.
</para>
</listitem>
<listitem>
<!--
2017-03-24 [857ee8e39] Add a txid_status function.
-->
<para>
Add function <link
linkend="functions-txid-snapshot"><function>txid_status()</></>
to check if a transaction was committed (Craig Ringer)
</para>
<para>
This is useful for checking after an abrupt disconnection whether
your previous transaction committed and you just didn't receive
the acknowledgement.
</para>
</listitem>
<listitem>
<!--
2017-01-19 [30bcebbdc] Allow negative years in make_date to represent BC years
-->
<para>
Allow <link
linkend="functions-formatting-table"><function>make_date()</></>
to interpret negative years as <acronym>BC</> years (&Aacute;lvaro
Herrera)
</para>
</listitem>
<listitem>
<!--
2016-09-28 [d3cd36a13] Make to_timestamp() and to_date() range-check fields of
-->
<para>
Make <function>to_timestamp()</> and <function>to_date()</> reject
out-of-range input fields (Artur Zakirov)
</para>
<para>
For example,
previously <literal>to_date('2009-06-40','YYYY-MM-DD')</> was
accepted and returned <literal>2009-07-10</>. It will now generate
an error.
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Server-Side Languages</title>
<itemizedlist>
<listitem>
<!--
2017-03-27 [70ec3f1f8] PL/Python: Add cursor and execute methods to plan object
-->
<para>
Allow PL/Python's <function>cursor()</> and <function>execute()</>
functions to be called as methods of their plan-object arguments
(Peter Eisentraut)
</para>
<para>
This allows a more object-oriented programming style.
</para>
</listitem>
<listitem>
<!--
2016-12-13 [55caaaeba] Improve handling of array elements as getdiag_targets an
-->
<para>
Allow PL/pgSQL's <command>GET DIAGNOSTICS</> statement to retrieve
values into array elements (Tom Lane)
</para>
<para>
Previously, a syntactic restriction prevented the target variable
from being an array element.
</para>
</listitem>
</itemizedlist>
<sect4>
<title><link linkend="pltcl">PL/Tcl</></title>
<itemizedlist>
<listitem>
<!--
2016-11-06 [26abb50c4] Support PL/Tcl functions that return composite types and
-->
<para>
Allow PL/Tcl functions to return composite types and sets
(Jim Nasby)
</para>
</listitem>
<listitem>
<!--
2017-03-11 [b58fd4a9c] Add a "subtransaction" command to PL/Tcl.
-->
<para>
Add a subtransaction command to PL/Tcl (Victor Wagner)
</para>
<para>
This allows PL/Tcl queries to fail without aborting the entire
function.
</para>
</listitem>
<listitem>
<!--
2017-03-07 [0d2b1f305] Invent start_proc parameters for PL/Tcl.
-->
<para>
Add server parameters <xref linkend="guc-pltcl-start-proc">
and <xref linkend="guc-pltclu-start-proc">, to allow initialization
functions to be called on PL/Tcl startup (Tom Lane)
</para>
</listitem>
</itemizedlist>
</sect4>
</sect3>
<sect3>
<title>Client Interfaces</title>
<itemizedlist>
<listitem>
<!--
2016-11-03 [274bb2b38] libpq: Allow connection strings and URIs to specify mult
2017-05-19 [5f374fe7a] libpq: Try next host if one of them times out.
2017-07-10 [7b02ba62e] Allow multiple hostaddrs to go with multiple hostnames.
-->
<para>
Allow specification of <link linkend="libpq-connect-host">multiple
host names or addresses</> in libpq connection strings and URIs
(Robert Haas, Heikki Linnakangas)
</para>
<para>
libpq will connect to the first responsive server in the list.
</para>
</listitem>
<listitem>
<!--
2016-11-29 [721f7bd3c] libpq: Add target_session_attrs parameter.
-->
<para>
Allow libpq connection strings and URIs to request a <link
linkend="libpq-connect-target-session-attrs">read/write host</>,
that is a master server rather than a standby server
(Victor Wagner, Mithun Cy)
</para>
<para>
This is useful when multiple host names are
specified. It is controlled by libpq connection parameter
<option>target_session_attrs</>.
</para>
</listitem>
<listitem>
<!--
2017-01-24 [ba005f193] Allow password file name to be specified as a libpq conn
-->
<para>
Allow the <link linkend="libpq-connect-passfile">password file name</>
to be specified as a libpq connection parameter (Julian Markwort)
</para>
<para>
Previously this could only be specified via an environment variable.
</para>
</listitem>
<listitem>
<!--
2017-05-03 [8f8b9be51] Add PQencryptPasswordConn function to libpq, use it in p
-->
<para>
Add function <link
linkend="libpq-pqencryptpasswordconn"><function>PQencryptPasswordConn()</></>
to allow creation of more types of encrypted passwords on the
client side (Michael Paquier, Heikki Linnakangas)
</para>
<para>
Previously only <literal>MD5</>-encrypted passwords could be created
using <link
linkend="libpq-pqencryptpassword"><function>PQencryptPassword()</></>.
This new function can also create <link
linkend="auth-pg-hba-conf"><literal>SCRAM-SHA-256</></>-encrypted
passwords.
</para>
</listitem>
<listitem>
<!--
2016-08-16 [a7b5573d6] Remove separate version numbering for ecpg preprocessor.
-->
<para>
Change <application>ecpg</> preprocessor version from 4.12 to 10
(Tom Lane)
</para>
<para>
Henceforth the <application>ecpg</> version will match
the <productname>PostgreSQL</> distribution version number.
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Client Applications</title>
<sect4>
<title><xref linkend="APP-PSQL"></title>
<itemizedlist>
<listitem>
<!--
2017-03-30 [e984ef586] Support \if ... \elif ... \else ... \endif in psql scrip
2017-04-02 [5dbc5da11] Fix behavior of psql's \p to agree with \g, \w, etc.
2017-04-02 [68dba97a4] Document psql's behavior of recalling the previously exe
-->
<para>
Add conditional branch support to <application>psql</> (Corey
Huinker)
</para>
<para>
This feature adds <application>psql</>
meta-commands <command>\if</>, <command>\elif</>, <command>\else</>,
and <command>\endif</>. This is primarily helpful for scripting.
</para>
</listitem>
<listitem>
<!--
2017-03-07 [b2678efd4] psql: Add \gx command
-->
<para>
Add <application>psql</> <command>\gx</> meta-command to execute
(<command>\g</>) a query in expanded mode (<command>\x</>)
(Christoph Berg)
</para>
</listitem>
<listitem>
<!--
2017-04-01 [f833c847b] Allow psql variable substitution to occur in backtick co
-->
<para>
Expand <application>psql</> variable references in
backtick-executed strings (Tom Lane)
</para>
<para>
This is particularly useful in the new <application>psql</>
conditional branch commands.
</para>
</listitem>
<listitem>
<!--
2017-01-30 [511ae628f] Make psql reject attempts to set special variables to in
2017-02-01 [86322dc7e] Improve psql's behavior for \set and \unset of its contr
2017-02-02 [fd6cd6980] Clean up psql's behavior for a few more control variable
-->
<para>
Prevent <application>psql</>'s special variables from being set to
invalid values (Daniel V&eacute;rit&eacute;, Tom Lane)
</para>
<para>
Previously, setting one of <application>psql</>'s special variables
to an invalid value silently resulted in the default behavior.
<command>\set</> on a special variable now fails if the proposed
new value is invalid. As a special exception, <command>\set</>
with an empty or omitted new value, on a boolean-valued special
variable, still has the effect of setting the variable
to <literal>on</>; but now it actually acquires that value rather
than an empty string. <command>\unset</> on a special variable now
explicitly sets the variable to its default value, which is also
the value it acquires at startup. In sum, a control variable now
always has a displayable value that reflects
what <application>psql</> is actually doing.
</para>
</listitem>
<listitem>
<!--
2017-09-06 [a6c678f01] Add psql variables showing server version and psql versi
-->
<para>
Add variables showing server version and <application>psql</> version
(Fabien Coelho)
</para>
</listitem>
<listitem>
<!--
2016-11-03 [a0f357e57] psql: Split up "Modifiers" column in \d and \dD
-->
<para>
Improve <application>psql</>'s <command>\d</> (display relation)
and <command>\dD</> (display domain) commands to show collation,
nullable, and default properties in separate columns (Peter
Eisentraut)
</para>
<para>
Previously they were shown in a single <quote>Modifiers</> column.
</para>
</listitem>
<listitem>
<!--
2017-07-27 [77cb4a1d6] Standardize describe.c's behavior for no-matching-object
-->
<para>
Make the various <command>\d</> commands handle no-matching-object
cases more consistently (Daniel Gustafsson)
</para>
<para>
They now all print the message about that to stderr, not stdout,
and the message wording is more consistent.
</para>
</listitem>
<listitem>
<!--
2016-08-18 [49917dbd7] Improve psql's tab completion for ALTER EXTENSION foo UP
2016-08-18 [8019b5a89] Improve psql's tab completion for \l.
2016-09-01 [76f9dd4fa] Improve tab completion for BEGIN & START|SET TRANSACTION
2016-09-11 [52803098a] psql tab completion for CREATE DATABASE ... TEMPLATE ...
2016-09-12 [63c1a8719] Fix recent commit for tab-completion of database templat
2016-11-03 [1d15d0db5] psql: Tab-complete LOCK [TABLE] ... IN {ACCESS|ROW|SHARE
2016-11-04 [927d7bb6b] Improve tab completion for CREATE TRIGGER.
2016-11-08 [577f0bdd2] psql: Tab completion for renaming enum values.
2017-03-01 [b5a388392] psql: Add tab completion for DEALLOCATE
2017-03-16 [d7d77f382] psql: Add completion for \help DROP|ALTER
-->
<para>
Improve <application>psql</>'s tab completion (Jeff Janes,
Ian Barwick, Andreas Karlsson, Sehrope Sarkuni, Thomas Munro,
Kevin Grittner, Dagfinn Ilmari Manns&aring;ker)
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title><xref linkend="pgbench"></title>
<itemizedlist>
<listitem>
<!--
2016-11-09 [41124a91e] pgbench: Allow the transaction log file prefix to be cha
-->
<para>
Add <application>pgbench</> option <option>--log-prefix</> to
control the log file prefix (Masahiko Sawada)
</para>
</listitem>
<listitem>
<!--
2017-01-20 [cdc2a7047] Allow backslash line continuations in pgbench's meta com
-->
<para>
Allow <application>pgbench</>'s meta-commands to span multiple
lines (Fabien Coelho)
</para>
<para>
A meta-command can now be continued onto the next line by writing
backslash-return.
</para>
</listitem>
<listitem>
<!--
2017-08-11 [796818442] Remove pgbench's restriction on placement of -M switch.
-->
<para>
Remove restriction on placement of <option>-M</> option relative to
other command line options (Tom Lane)
</para>
</listitem>
</itemizedlist>
</sect4>
</sect3>
<sect3>
<title>Server Applications</title>
<itemizedlist>
<listitem>
<!--
2017-01-17 [cada1af31] Add compression support to pg_receivexlog
-->
<para>
Add <link
linkend="app-pgreceivewal"><application>pg_receivewal</></>
option <option>-Z</>/<option>--compress</> to specify compression
(Michael Paquier)
</para>
</listitem>
<listitem>
<!--
2017-01-04 [7c030783a] Add pg_recvlogical -\-endpos=LSN
-->
<para>
Add <link
linkend="app-pgrecvlogical"><application>pg_recvlogical</></> option
<option>--endpos</> to specify the ending position (Craig Ringer)
</para>
<para>
This complements the existing <option>--startpos</> option.
</para>
</listitem>
<listitem>
<!--
2016-10-19 [5d58c07a4] initdb pg_basebackup: Rename -\-noxxx options to -\-no-x
-->
<para>
Rename <link linkend="APP-INITDB"><application>initdb</></>
options <option>--noclean</> and <option>--nosync</> to be spelled
<option>--no-clean</> and <option>--no-sync</> (Vik Fearing,
Peter Eisentraut)
</para>
<para>
The old spellings are still supported.
</para>
</listitem>
</itemizedlist>
<sect4>
<title><link linkend="APP-PGDUMP"><application>pg_dump</></>,
<link linkend="APP-PG-DUMPALL"><application>pg_dumpall</></>,
<link linkend="APP-PGRESTORE"><application>pg_restore</></></title>
<itemizedlist>
<listitem>
<!--
2016-09-20 [46b55e7f8] pg_restore: Add -N option to exclude schemas
-->
<para>
Allow <application>pg_restore</> to exclude schemas (Michael Banck)
</para>
<para>
This adds a new <option>-N</>/<option>--exclude-schema</> option.
</para>
</listitem>
<listitem>
<!--
2016-11-29 [4fafa579b] Add -\-no-blobs option to pg_dump
-->
<para>
Add <option>--no-blobs</> option to
<application>pg_dump</> (Guillaume Lelarge)
</para>
<para>
This suppresses dumping of large objects.
</para>
</listitem>
<listitem>
<!--
2017-03-07 [9a83d56b3] Allow pg_dumpall to dump roles w/o user passwords
-->
<para>
Add <application>pg_dumpall</> option
<option>--no-role-passwords</> to omit role passwords
(Robins Tharakan, Simon Riggs)
</para>
<para>
This allows use of <application>pg_dumpall</> by non-superusers;
without this option, it fails due to inability to read passwords.
</para>
</listitem>
<listitem>
<!--
2017-08-16 [e4892c681] pg_dump: Support using synchronized snapshots on standby
-->
<para>
Support using synchronized snapshots when dumping from a standby
server (Petr Jelinek)
</para>
</listitem>
<listitem>
<!--
2017-03-22 [96a7128b7] Sync pg_dump and pg_dumpall output
-->
<para>
Issue <function>fsync()</> on the output files generated by
<application>pg_dump</> and
<application>pg_dumpall</> (Michael Paquier)
</para>
<para>
This provides more security that the output is safely stored on
disk before the program exits. This can be disabled with
the new <option>--no-sync</> option.
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title><xref linkend="app-pgbasebackup"></title>
<itemizedlist>
<listitem>
<!--
2016-10-23 [56c7d8d45] Allow pg_basebackup to stream transaction log in tar mod
2016-12-21 [ecbdc4c55] Forbid invalid combination of options in pg_basebackup.
-->
<para>
Allow <application>pg_basebackup</> to stream write-ahead log in
tar mode (Magnus Hagander)
</para>
<para>
The <acronym>WAL</> will be stored in a separate tar file from
the base backup.
</para>
</listitem>
<listitem>
<!--
2017-01-16 [e7b020f78] Make pg_basebackup use temporary replication slots
-->
<para>
Make <application>pg_basebackup</> use temporary replication slots
(Magnus Hagander)
</para>
<para>
Temporary replication slots will be used by default when
<application>pg_basebackup</> uses WAL streaming with default
options.
</para>
</listitem>
<listitem>
<!--
2016-09-29 [bc34223bc] pg_basebackup pg_receivexlog: Issue fsync more carefully
-->
<para>
Be more careful about fsync'ing in all required places
in <application>pg_basebackup</> and
<application>pg_receivewal</> (Michael Paquier)
</para>
</listitem>
<listitem>
<!--
2016-09-29 [6ed2d8584] pg_basebackup: Add -\-nosync option
2016-10-19 [5d58c07a4] initdb pg_basebackup: Rename -\-noxxx options to -\-no-x
-->
<para>
Add <application>pg_basebackup</> option <option>--no-sync</> to
disable fsync (Michael Paquier)
</para>
</listitem>
<listitem>
<!--
2016-09-28 [6ad8ac602] Exclude additional directories in pg_basebackup
-->
<para>
Improve <application>pg_basebackup</>'s handling of which
directories to skip (David Steele)
</para>
</listitem>
</itemizedlist>
</sect4>
<sect4>
<title><application><xref linkend="app-pg-ctl"></></title>
<itemizedlist>
<listitem>
<!--
2016-09-21 [e7010ce47] pg_ctl: Add wait option to promote action
-->
<para>
Add wait option for <application><xref linkend="app-pg-ctl"></>'s
promote operation (Peter Eisentraut)
</para>
</listitem>
<listitem>
<!--
2016-10-19 [0be22457d] pg_ctl: Add long options for -w and -W
-->
<para>
Add long options for <application>pg_ctl</> wait (<option>--wait</>)
and no-wait (<option>--no-wait</>) (Vik Fearing)
</para>
</listitem>
<listitem>
<!--
2016-10-19 [caf936b09] pg_ctl: Add long option for -o
-->
<para>
Add long option for <application>pg_ctl</> server options
(<option>--options</>) (Peter Eisentraut)
</para>
</listitem>
<listitem>
<!--
2017-06-28 [f13ea95f9] Change pg_ctl to detect server-ready by watching status
-->
<para>
Make <literal>pg_ctl start --wait</> detect server-ready by
watching <filename>postmaster.pid</>, not by attempting connections
(Tom Lane)
</para>
<para>
The postmaster has been changed to report its ready-for-connections
status in <filename>postmaster.pid</>, and <application>pg_ctl</>
now examines that file to detect whether startup is complete.
This is more efficient and reliable than the old method, and it
eliminates postmaster log entries about rejected connection
attempts during startup.
</para>
</listitem>
<listitem>
<!--
2017-06-26 [c61559ec3] Reduce pg_ctl's reaction time when waiting for postmaste
-->
<para>
Reduce <application>pg_ctl</>'s reaction time when waiting for
postmaster start/stop (Tom Lane)
</para>
<para>
<application>pg_ctl</> now probes ten times per second when waiting
for a postmaster state change, rather than once per second.
</para>
</listitem>
<listitem>
<!--
2017-07-05 [1bac5f552] pg_ctl: Make failure to complete operation a nonzero exi
-->
<para>
Ensure that <application>pg_ctl</> exits with nonzero status if an
operation being waited for does not complete within the timeout
(Peter Eisentraut)
</para>
<para>
The <literal>start</> and <literal>promote</> operations now return
exit status 1, not 0, in such cases. The <literal>stop</> operation
has always done that.
</para>
</listitem>
</itemizedlist>
</sect4>
</sect3>
<sect3>
<title>Source Code</title>
<itemizedlist>
<listitem>
<!--
2016-08-15 [ca9112a42] Stamp HEAD as 10devel.
-->
<para>
Change to two-part release version numbering (Peter Eisentraut, Tom
Lane)
</para>
<para>
Release numbers will now have two parts (e.g., <literal>10.1</>)
rather than three (e.g., <literal>9.6.3</>).
Major versions will now increase just the first number, and minor
releases will increase just the second number.
Release branches will be referred to by single numbers
(e.g., <literal>10</> rather than <literal>9.6</>).
This change is intended to reduce user confusion about what is a
major or minor release of <productname>PostgreSQL</>.
</para>
</listitem>
<listitem>
<!--
2017-06-16 [cea258b63] Teach pgindent to skip files generated by bison or flex
2017-06-21 [8ff6d4ec7] Adjust pgindent script to use pg_bsd_indent 2.0.
2017-06-21 [e3860ffa4] Initial pgindent run with pg_bsd_indent version 2.0.
2017-06-21 [c7b8998eb] Phase 2 of pgindent updates.
2017-06-21 [382ceffdf] Phase 3 of pgindent updates.
2017-06-21 [81f056c72] Remove entab and associated detritus.
-->
<para>
Improve behavior of <application>pgindent</>
(Piotr Stefaniak, Tom Lane)
</para>
<para>
We have switched to a new version of <application>pg_bsd_indent</>
based on recent improvements made by the FreeBSD project. This
fixes numerous small bugs that led to odd C code formatting
decisions. Most notably, lines within parentheses (such as in a
multi-line function call) are now uniformly indented to match the
opening paren, even if that would result in code extending past the
right margin.
</para>
</listitem>
<listitem>
<!--
2017-03-23 [eccfef81e] ICU support
-->
<para>
Allow the <link linkend="configure"><acronym>ICU</></> library to
optionally be used for collation support (Peter Eisentraut)
</para>
<para>
The <acronym>ICU</> library has versioning that allows detection
of collation changes between versions. It is enabled via configure
option <option>--with-icu</>. The default still uses the operating
system's native collation library.
</para>
</listitem>
<listitem>
<!--
2016-11-04 [c8ead2a39] Provide DLLEXPORT markers for C functions via PG_FUNCTIO
-->
<para>
Automatically mark all <link
linkend="xfunc-c"><function>PG_FUNCTION_INFO_V1</></> functions
as <literal>DLLEXPORT</>-ed on
<systemitem class="osname">Windows</> (Laurenz Albe)
</para>
<para>
If third-party code is using <literal>extern</> function
declarations, they should also add <literal>DLLEXPORT</> markers
to those declarations.
</para>
</listitem>
<listitem>
<!--
2016-11-08 [1833f1a1c] Simplify code by getting rid of SPI_push, SPI_pop, SPI_r
-->
<para>
Remove <acronym>SPI</> functions <function>SPI_push()</>,
<function>SPI_pop()</>, <function>SPI_push_conditional()</>,
<function>SPI_pop_conditional()</>,
and <function>SPI_restore_connection()</> as unnecessary (Tom Lane)
</para>
<para>
Their functionality now happens automatically. There are now no-op
macros by these names so that external modules don't need to be
updated immediately, but eventually such calls should be removed.
</para>
<para>
A side effect of this change is that <function>SPI_palloc()</> and
allied functions now require an active SPI connection; they do not
degenerate to simple <function>palloc()</> if there is none. That
previous behavior was not very useful and posed risks of unexpected
memory leaks.
</para>
</listitem>
<listitem>
<!--
2016-12-02 [13e14a78e] Management of free memory pages.
2016-12-02 [13df76a53] Introduce dynamic shared memory areas.
2016-12-19 [e13029a5c] Provide a DSA area for all parallel queries.
-->
<para>
Allow shared memory to be dynamically allocated (Thomas Munro,
Robert Haas)
</para>
</listitem>
<listitem>
<!--
2017-02-27 [58b25e981] Add "Slab" MemoryContext implementation for efficient eq
-->
<para>
Add slab-like memory allocator for efficient fixed-size allocations
(Tomas Vondra)
</para>
</listitem>
<listitem>
<!--
2016-10-09 [ecb0d20a9] Use unnamed POSIX semaphores, if available, on Linux and
-->
<para>
Use <acronym>POSIX</> semaphores rather than SysV semaphores
on <systemitem class="osname">Linux</> and <systemitem
class="osname">FreeBSD</> (Tom Lane)
</para>
<para>
This avoids platform-specific limits on SysV semaphore usage.
</para>
</listitem>
<listitem>
<!--
2017-04-07 [e8fdbd58f] Improve 64bit atomics support.
-->
<para>
Improve support for 64-bit atomics (Andres Freund)
</para>
</listitem>
<listitem>
<!--
2017-03-10 [f8f1430ae] Enable 64 bit atomics on ARM64.
-->
<para>
Enable 64-bit atomic operations on <acronym>ARM64</> (Roman
Shaposhnik)
</para>
</listitem>
<listitem>
<!--
2017-01-02 [1d63f7d2d] Use clock_gettime(), if available, in instr_time measure
-->
<para>
Switch to using <function>clock_gettime()</>, if available, for
duration measurements (Tom Lane)
</para>
<para>
<function>gettimeofday()</> is still used
if <function>clock_gettime()</> is not available.
</para>
</listitem>
<listitem>
<!--
2016-12-05 [fe0a0b599] Replace PostmasterRandom() with a stronger source, secon
2017-07-03 [bf723a274] Forbid gen_random_uuid() with -\-disable-strong-random
-->
<para>
Add more robust random number generators to be used for
cryptographically secure uses (Magnus Hagander, Michael Paquier,
Heikki Linnakangas)
</para>
<para>
If no strong random number generator can be
found, <link linkend="configure">configure</> will fail unless
the <option>--disable-strong-random</> option is used. However, with
this option, <link linkend="pgcrypto"><application>pgcrypto</></>
functions requiring a strong random number generator will be disabled.
</para>
</listitem>
<listitem>
<!--
2017-08-15 [d7ab908fb] Distinguish wait-for-connection from wait-for-write-read
-->
<para>
Allow <function>WaitLatchOrSocket()</> to wait for socket
connection on Windows (Andres Freund)
</para>
</listitem>
<listitem>
<!--
2017-04-06 [3f902354b] Clean up after insufficiently-researched optimization of
-->
<para>
<filename>tupconvert.c</> functions no longer convert tuples just to
embed a different composite-type OID in them (Ashutosh Bapat, Tom Lane)
</para>
<para>
The majority of callers don't care about the composite-type OID;
but if the result tuple is to be used as a composite Datum, steps
should be taken to make sure the correct OID is inserted in it.
</para>
</listitem>
<listitem>
<!--
2016-10-11 [2b860f52e] Remove "sco" and "unixware" ports.
-->
<para>
Remove <systemitem class="osname">SCO</> and <systemitem
class="osname">Unixware</> ports (Tom Lane)
</para>
</listitem>
<listitem>
<!--
2016-08-18 [e8306745e] doc: Speed up XSLT builds
2016-08-24 [0e4cc1fc5] doc: Fix XSLT speedup with older upstream stylesheet ver
-->
<para>
Overhaul documentation <link linkend="docguide-toolsets">build
process</> (Alexander Lakhin)
</para>
</listitem>
<listitem>
<!--
2017-04-06 [510074f9f] Remove use of Jade and DSSSL
-->
<para>
Use <acronym>XSLT</> to build the <productname>PostgreSQL</>
documentation (Peter Eisentraut)
</para>
<para>
Previously <application>Jade</>, <acronym>DSSSL</>, and
<application>JadeTex</> were used.
</para>
</listitem>
<listitem>
<!--
2016-11-15 [e36ddab11] Build HTML documentation using XSLT stylesheets by defau
-->
<para>
Build <acronym>HTML</> documentation using <acronym>XSLT</>
stylesheets by default (Peter Eisentraut)
</para>
</listitem>
</itemizedlist>
</sect3>
<sect3>
<title>Additional Modules</title>
<itemizedlist>
<listitem>
<!--
2016-09-29 [8e91e12bc] Allow contrib/file_fdw to read from a program, like COPY
-->
<para>
Allow <link linkend="file-fdw"><application>file_fdw</></> to read
from program output as well as files (Corey Huinker, Adam Gomaa)
</para>
</listitem>
<listitem>
<!--
2016-10-21 [7012b132d] postgres_fdw: Push down aggregates to remote servers.
-->
<para>
In <link linkend="postgres-fdw"><application>postgres_fdw</></>,
push aggregate functions to the remote server, when possible
(Jeevan Chalke, Ashutosh Bapat)
</para>
<para>
This reduces the amount of data that must be passed from the remote
server, and offloads aggregate computation from the requesting server.
</para>
</listitem>
<listitem>
<!--
2017-03-16 [b30fb56b0] postgres_fdw: Push down FULL JOINs with restriction clau
2017-04-24 [332bec1e6] postgres_fdw: Fix join push down with extensions
-->
<para>
In <application>postgres_fdw</>, push joins to the remote server in
more cases (David Rowley, Ashutosh Bapat, Etsuro Fujita)
</para>
</listitem>
<listitem>
<!--
2016-08-26 [ae025a159] Support OID system column in postgres_fdw.
-->
<para>
Properly support <type>OID</> columns in
<application>postgres_fdw</> tables (Etsuro Fujita)
</para>
<para>
Previously <type>OID</> columns always returned zeros.
</para>
</listitem>
<listitem>
<!--
2017-03-21 [f7946a92b] Add btree_gist support for enum types.
-->
<para>
Allow <link linkend="btree-gist"><application>btree_gist</></>
and <link linkend="btree-gin"><application>btree_gin</></> to
index enum types (Andrew Dunstan)
</para>
<para>
This allows enums to be used in exclusion constraints.
</para>
</listitem>
<listitem>
<!--
2016-11-29 [11da83a0e] Add uuid to the set of types supported by contrib/btree_
-->
<para>
Add indexing support to <application>btree_gist</> for the
<type>UUID</> data type (Paul Jungwirth)
</para>
</listitem>
<listitem>
<!--
2017-03-09 [3717dc149] Add amcheck extension to contrib.
-->
<para>
Add <link linkend="amcheck"><application>amcheck</></> which can
check the validity of B-tree indexes (Peter Geoghegan)
</para>
</listitem>
<listitem>
<!--
2017-03-27 [a6f22e835] Show ignored constants as "$N" rather than "?" in pg_sta
-->
<para>
Show ignored constants as <literal>$N</> rather than <literal>?</>
in
<link
linkend="pgstatstatements"><application>pg_stat_statements</></>
(Lukas Fittl)
</para>
</listitem>
<listitem>
<!--
2016-09-27 [f31a931fa] Improve contrib/cube's handling of zero-D cubes, infinit
-->
<para>
Improve <link linkend="cube"><application>cube</></>'s handling
of zero-dimensional cubes (Tom Lane)
</para>
<para>
This also improves handling of <literal>infinite</> and
<literal>NaN</> values.
</para>
</listitem>
<listitem>
<!--
2016-09-29 [6e654546f] Don't bother to lock bufmgr partitions in pg_buffercache
-->
<para>
Allow <link
linkend="pgbuffercache"><application>pg_buffercache</></> to run
with fewer locks (Ivan Kartyshov)
</para>
<para>
This makes it less disruptive when run on production systems.
</para>
</listitem>
<listitem>
<!--
2017-02-03 [e759854a0] pgstattuple: Add pgstathashindex.
-->
<para>
Add <link linkend="pgstattuple"><application>pgstattuple</></>
function <function>pgstathashindex()</> to view hash index
statistics (Ashutosh Sharma)
</para>
</listitem>
<listitem>
<!--
2016-09-29 [fd321a1df] Remove superuser checks in pgstattuple
-->
<para>
Use <command>GRANT</> permissions to
control <application>pgstattuple</> function usage (Stephen Frost)
</para>
<para>
This allows DBAs to allow non-superusers to run these functions.
</para>
</listitem>
<listitem>
<!--
2016-10-28 [d4b5d4cad] pgstattuple: Don't take heavyweight locks when examining
-->
<para>
Reduce locking when <application>pgstattuple</> examines hash
indexes (Amit Kapila)
</para>
</listitem>
<listitem>
<!--
2017-03-17 [fef2bcdcb] pageinspect: Add page_checksum function
-->
<para>
Add <link linkend="pageinspect"><application>pageinspect</></>
function <function>page_checksum()</> to show a page's checksum
(Tomas Vondra)
</para>
</listitem>
<listitem>
<!--
2017-04-04 [193f5f9e9] pageinspect: Add bt_page_items function with bytea argum
-->
<para>
Add <application>pageinspect</>
function <function>bt_page_items()</> to print page items from a
page image (Tomas Vondra)
</para>
</listitem>
<listitem>
<!--
2017-02-02 [08bf6e529] pageinspect: Support hash indexes.
-->
<para>
Add hash index support to <application>pageinspect</> (Jesper
Pedersen, Ashutosh Sharma)
</para>
</listitem>
</itemizedlist>
</sect3>
</sect2>
</sect1>
Reference in a new issue View git blame Copy permalink