In SQL, the ability to use parallel query was previous contingent on
fcache->readonly_func, which is only set for non-volatile functions;
but the volatility of a function has no bearing on whether queries
inside it can use parallelism.  Remove that condition.

SPI_execute and SPI_execute_with_args always run the plan just once,
though not necessarily to completion.  Given the changes in commit
691b8d59, it's sensible to pass
CURSOR_OPT_PARALLEL_OK here, so do that.  This improves access to
parallelism for any caller that uses these functions to execute
queries.  Such callers include plperl, plpython, pltcl, and plpgsql,
though it's not the case that they all use these functions
exclusively.

In plpgsql, allow parallel query for plain SELECT queries (as
opposed to PERFORM, which already worked) and for plain expressions
(which probably won't go through the executor at all, because they
will likely be simple expressions, but if they do then this helps).

Rafia Sabih and Robert Haas, reviewed by Dilip Kumar and Amit Kapila

Discussion: http://postgr.es/m/CAOGQiiMfJ+4SQwgG=6CVHWoisiU0+7jtXSuiyXBM3y=A=eJzmg@mail.gmail.com

Detected by buildfarm member prion

Buildfarm issues and other reported issues

Previously manual VACUUM did not report the number of skipped frozen
pages even when VERBOSE option is specified. But this information is
helpful to monitor the VACUUM activity, and also autovacuum reports that
number in the log file when the condition of log_autovacuum_min_duration
is met.

This commit changes VACUUM VERBOSE so that it reports the number
of frozen pages that it skips.

Author: Masahiko Sawada
Reviewed-by: Yugo Nagata and Jim Nasby

Discussion: http://postgr.es/m/CAD21AoDZQKCxo0L39Mrq08cONNkXQKXuh=2DP1Q8ebmt35SoaA@mail.gmail.com

Add support for explicitly declared statistic objects (CREATE
STATISTICS), allowing collection of statistics on more complex
combinations that individual table columns.  Companion commands DROP
STATISTICS and ALTER STATISTICS ... OWNER TO / SET SCHEMA / RENAME are
added too.  All this DDL has been designed so that more statistic types
can be added later on, such as multivariate most-common-values and
multivariate histograms between columns of a single table, leaving room
for permitting columns on multiple tables, too, as well as expressions.

This commit only adds support for collection of n-distinct coefficient
on user-specified sets of columns in a single table.  This is useful to
estimate number of distinct groups in GROUP BY and DISTINCT clauses;
estimation errors there can cause over-allocation of memory in hashed
aggregates, for instance, so it's a worthwhile problem to solve.  A new
special pseudo-type pg_ndistinct is used.

(num-distinct estimation was deemed sufficiently useful by itself that
this is worthwhile even if no further statistic types are added
immediately; so much so that another version of essentially the same
functionality was submitted by Kyotaro Horiguchi:
https://postgr.es/m/20150828.173334.114731693.horiguchi.kyotaro@lab.ntt.co.jp
though this commit does not use that code.)

Author: Tomas Vondra.  Some code rework by Álvaro.
Reviewed-by: Dean Rasheed, David Rowley, Kyotaro Horiguchi, Jeff Janes,
    Ideriha Takeshi
Discussion: https://postgr.es/m/543AFA15.4080608@fuzzy.cz
    https://postgr.es/m/20170320190220.ixlaueanxegqd5gr@alvherre.pgsql

Commit 7aea8e4f allowed a parallel
plan to be generated when for a RETURN QUERY or RETURN QUERY EXECUTE
statement in a PL/pgsql block, but that's a bad idea because plplgsql
asks the executor for 50 rows at a time.  That means that we'll always
be running serially a plan that was intended for parallel execution,
which is not a good idea.  Fix by not requesting a parallel plan from
the outset.

Per discussion, back-patch to 9.6.  There is a slight risk that, due
to optimizer error, somebody could have a case where the parallel plan
executed serially is actually faster than the supposedly-best serial
plan, but the consensus seems to be that that's not sufficient
justification for leaving 9.6 unpatched.

Discussion: http://postgr.es/m/CA+TgmoZ_ZuH+auEeeWnmtorPsgc_SmP+XWbDsJ+cWvWBSjNwDQ@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmobXEhvHbJtWDuPZM9bVSLiTj-kShxQJ2uM5GPDze9fRYA@mail.gmail.com

If your connection to the database server is lost while a COMMIT is
in progress, it may be difficult to figure out whether the COMMIT was
successful or not.  This function will tell you, provided that you
don't wait too long to ask.  It may be useful in other situations,
too.

Craig Ringer, reviewed by Simon Riggs and by me

Discussion: http://postgr.es/m/CAMsr+YHQiWNEi0daCTboS40T+V5s_+dst3PYv_8v2wNVH+Xx4g@mail.gmail.com

For normal commits and aborts we already reset PgXact->xmin
Avoiding touching highly contented shmem improves concurrent
performance.

Simon Riggs

Discussion: CANP8+jJdXE9b+b9F8CQT-LuxxO0PBCB-SZFfMVAdp+akqo4zfg@mail.gmail.com

Always return tupleslot and tupledesc from libpqrcv_exec.  This avoids
requiring callers to handle that separately.

Author: Petr Jelinek <petr.jelinek@2ndquadrant.com>
Reported-by: Michael Banck <michael.banck@credativ.de>

If a user has a SCRAM verifier in pg_authid.rolpassword, there's no reason
we cannot attempt to perform SCRAM authentication instead of MD5. The worst
that can happen is that the client doesn't support SCRAM, and the
authentication will fail. But previously, it would fail for sure, because
we would not even try. SCRAM is strictly more secure than MD5, so there's
no harm in trying it. This allows for a more graceful transition from MD5
passwords to SCRAM, as user passwords can be changed to SCRAM verifiers
incrementally, without changing pg_hba.conf.

Refactor the code in auth.c to support that better. Notably, we now have to
look up the user's pg_authid entry before sending the password challenge,
also when performing MD5 authentication. Also simplify the concept of a
"doomed" authentication. Previously, if a user had a password, but it had
expired, we still performed SCRAM authentication (but always returned error
at the end) using the salt and iteration count from the expired password.
Now we construct a fake salt, like we do when the user doesn't have a
password or doesn't exist at all. That simplifies get_role_password(), and
we can don't need to distinguish the  "user has expired password", and
"user does not exist" cases in auth.c.

On second thoughts, also rename uaSASL to uaSCRAM. It refers to the
mechanism specified in pg_hba.conf, and while we use SASL for SCRAM
authentication at the protocol level, the mechanism should be called SCRAM,
not SASL. As a comparison, we have uaLDAP, even though it looks like the
plain 'password' authentication at the protocol level.

Discussion: https://www.postgresql.org/message-id/6425.1489506016@sss.pgh.pa.us
Reviewed-by: Michael Paquier

Assert-enabled build crashes but without asserts it works by wrong way:
it may not reset forcing full page write and preventing from starting
exclusive backup with the same name as cancelled.
Patch replaces pair of booleans
nonexclusive_backup_running/exclusive_backup_running to single enum to
correctly describe backup state.

Backpatch to 9.6 where bug was introduced

Reported-by: David Steele
Authors: Michael Paquier, David Steele
Reviewed-by: Anastasia Lubennikova

https://commitfest.postgresql.org/13/1068/

Buildfarm member anole sees this union as empty, and doesn't like it.

Reported-by: Thomas Munro <thomas.munro@enterprisedb.com>

This only happened with single-byte encodings.

Commit 249cf070 assigned to one of
the labels in the middle the value that should have been assigned
to the first member of the enum.  Rushabh's patch didn't have that
defect as submitted, but I managed to mess it up while editing.
Repair.

Add a column collprovider to pg_collation that determines which library
provides the collation data.  The existing choices are default and libc,
and this adds an icu choice, which uses the ICU4C library.

The pg_locale_t type is changed to a union that contains the
provider-specific locale handles.  Users of locale information are
changed to look into that struct for the appropriate handle to use.

Also add a collversion column that records the version of the collation
when it is created, and check at run time whether it is still the same.
This detects potentially incompatible library upgrades that can corrupt
indexes and other structures.  This is currently only supported by
ICU-provided collations.

initdb initializes the default collation set as before from the `locale
-a` output but also adds all available ICU locales with a "-x-icu"
appended.

Currently, ICU-provided collations can only be explicitly named
collations.  The global database locales are still always libc-provided.

ICU support is enabled by configure --with-icu.
Reviewed-by: Thomas Munro <thomas.munro@enterprisedb.com>
Reviewed-by: Andreas Karlsson <andreas@proxel.se>

This provides infrastructure for looking up arbitrary, user-supplied
XIDs without a risk of scary-looking failures from within the clog
module.  Normally, the oldest XID that can be safely looked up in CLOG
is the same as the oldest XID that can reused without causing
wraparound, and the latter is already tracked.  However, while
truncation is in progress, the values are different, so we must
keep track of them separately.

Craig Ringer, reviewed by Simon Riggs and by me.

Discussion: http://postgr.es/m/CAMsr+YHQiWNEi0daCTboS40T+V5s_+dst3PYv_8v2wNVH+Xx4g@mail.gmail.com

They have been deprecated since PostgreSQL 9.1.
Reviewed-by: Magnus Hagander <magnus@hagander.net>
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>

Previously, it was unsafe to execute a plan in parallel if
ExecutorRun() might be called with a non-zero row count.  However,
it's quite easy to fix things up so that we can support that case,
provided that it is known that we will never call ExecutorRun() a
second time for the same QueryDesc.  Add infrastructure to signal
this, and cross-checks to make sure that a caller who claims this is
true doesn't later reneg.

While that pattern never happens with queries received directly from a
client -- there's no way to know whether multiple Execute messages
will be sent unless the first one requests all the rows -- it's pretty
common for queries originating from procedural languages, which often
limit the result to a single tuple or to a user-specified number of
tuples.

This commit doesn't actually enable parallelism in any additional
cases, because currently none of the places that would be able to
benefit from this infrastructure pass CURSOR_OPT_PARALLEL_OK in the
first place, but it makes it much more palatable to pass
CURSOR_OPT_PARALLEL_OK in places where we currently don't, because it
eliminates some cases where we'd end up having to run the parallel
plan serially.

Patch by me, based on some ideas from Rafia Sabih and corrected by
Rafia Sabih based on feedback from Dilip Kumar and myself.

Discussion: http://postgr.es/m/CA+TgmobXEhvHbJtWDuPZM9bVSLiTj-kShxQJ2uM5GPDze9fRYA@mail.gmail.com

GIN vacuum during cleaning posting tree can lock this whole tree for a long
time with by holding LockBufferForCleanup() on root. Patch changes it with
two ways: first, cleanup lock will be taken only if there is an empty page
(which should be deleted) and, second, it tries to lock only subtree, not the
whole posting tree.

Author: Andrey Borodin with minor editorization by me
Reviewed-by: Jeff Davis, me

https://commitfest.postgresql.org/13/896/

Author: Petr Jelinek <petr.jelinek@2ndquadrant.com>

related to 7c4f5240, per build farm

Author: Petr Jelinek <petr.jelinek@2ndquadrant.com>

Jon Nelson

Adds write_lag, flush_lag and replay_lag cols to pg_stat_replication.

Implements a lag tracker module that reports the lag times based upon
measurements of the time taken for recent WAL to be written, flushed and
replayed and for the sender to hear about it. These times
represent the commit lag that was (or would have been) introduced by each
synchronous commit level, if the remote server was configured as a
synchronous standby.  For an asynchronous standby, the replay_lag column
approximates the delay before recent transactions became visible to queries.
If the standby server has entirely caught up with the sending server and
there is no more WAL activity, the most recently measured lag times will
continue to be displayed for a short time and then show NULL.

Physical replication lag tracking is automatic. Logical replication tracking
is possible but is the responsibility of the logical decoding plugin.
Tracking is a private module operating within each walsender individually,
with values reported to shared memory. Module not used outside of walsender.

Design and code is good enough now to commit - kudos to the author.
In many ways a difficult topic, with important and subtle behaviour so this
shoudl be expected to generate discussion and multiple open items: Test now!

Author: Thomas Munro, following designs by Fujii Masao and Simon Riggs
Review: Simon Riggs, Ian Barwick and Craig Ringer

Add functionality for a new subscription to copy the initial data in the
tables and then sync with the ongoing apply process.

For the copying, add a new internal COPY option to have the COPY source
data provided by a callback function.  The initial data copy works on
the subscriber by receiving COPY data from the publisher and then
providing it locally into a COPY that writes to the destination table.

A WAL receiver can now execute full SQL commands.  This is used here to
obtain information about tables and publications.

Several new options were added to CREATE and ALTER SUBSCRIPTION to
control whether and when initial table syncing happens.

Change pg_dump option --no-create-subscription-slots to
--no-subscription-connect and use the new CREATE SUBSCRIPTION
... NOCONNECT option for that.

Author: Petr Jelinek <petr.jelinek@2ndquadrant.com>
Tested-by: Erik Rijkers <er@xs4all.nl>

Author: Emil Iggland

Internally, we have supported the option to either wait for all of the
WAL associated with a backup to be archived, or to return immediately.
This option is useful to users of pg_stop_backup() as well, when they
are reading the stop backup record position and checking that the WAL
they need has been archived independently.

This patch adds an additional, optional, argument to pg_stop_backup()
which allows the user to indicate if they wish to wait for the WAL to be
archived or not.  The default matches current behavior, which is to
wait.

Author: David Steele, with some minor changes, doc updates by me.
Reviewed by: Takayuki Tsunakawa, Fujii Masao
Discussion: https://postgr.es/m/758e3fd1-45b4-5e28-75cd-e9e7f93a4c02@pgmasters.net

The pg_hba.conf rules were changed in
8df9bd0b.

There's no mechanism for such a sort to become a top-N sort, so we
should pass -1 rather than limit_tuples to cost_sort().

Rushabh Lathia, per a report from Mithun Cy

Discussion: http://postgr.es/m/CAGPqQf1akRcSgC9=6iwx=sEPap9UvPpHJLzg8_N+OuHdb6fL+g@mail.gmail.com

This changes all the RADIUS related parameters (radiusserver,
radiussecret, radiusport, radiusidentifier) to be plural and to accept a
comma separated list of servers, which will be tried in order.

Reviewed by Adam Brightwell

Craig Ringer

Replace ignoreVacuum parameter with more flexible flags.

Author: Eiji Seki
Review: Haribabu Kommi

Previously AELs were registered against the top-level xid, which could
cause locks to be held much longer than necessary in some cases during
Hot Standby replay. We now record locks directly against their appropriate
xids. Requires few code changes because original code allowed for this
situation but didn’t fully implement it.

Discussion: CAKJS1f9vJ841HY=wonnLVbfkTWGYWdPN72VMxnArcGCjF3SywA@mail.gmail.com

Author: Simon Riggs and David Rowley

Logical replication no longer uses the "replication" keyword.  It just
matches database entries in the normal way.  The "replication" keyword
now only applies to physical replication.
Reviewed-by: Petr Jelinek <petr.jelinek@2ndquadrant.com>

Previously, the new owner had to be a superuser.  The new rules are more
refined similar to other objects.
Reviewed-by: Petr Jelinek <petr.jelinek@2ndquadrant.com>

Before exiting any files are fsync'ed. A --no-sync option is also
provided for a faster exit if desired.

Michael Paquier.

Reviewed by Albe Laurenz

Discussion: https://postgr.es/m/CAB7nPqS1uZ=Ov+UruW6jr3vB-S_DLVMPc0dQpV-fTDjmm0ZQMg@mail.gmail.com

A hot standby replica keeps a list of Access Exclusive locks for a top
level transaction. These locks are released when the top level transaction
ends. Searching of this list is O(N^2), and each transaction had to pay the
price of searching this list for locks, even if it didn't take any AE
locks itself.

This patch optimizes this case by having the master server track which
transactions took AE locks, and passes that along to the standby server in
the commit/abort record. This allows the standby to only try to release
locks for transactions which actually took any, avoiding the majority of
the performance issue.

Refactor MyXactAccessedTempRel into MyXactFlags to allow minimal additional
cruft with this.

Analysis and initial patch by David Rowley
Author: David Rowley and Simon Riggs