Daniel Gustafsson and Erik Rijkers, per report from nick@cleaton

Discussion: https://postgr.es/m/159275646273.679.16940709892308114570@wrigleys.postgresql.org

If restart_lsn of logical replication slot gets behind more than
max_slot_wal_keep_size from the current LSN, the logical replication slot
would be invalidated and its restart_lsn is reset to an invalid LSN.
If this logical replication slot with an invalid restart_lsn was specified as
the source slot in pg_copy_logical_replication_slot(), the function caused
the assertion failure unexpectedly.

This assertion was added because restart_lsn should not be invalid before.
But in v13, it can be invalid thanks to max_slot_wal_keep_size. So since this
assertion is no longer useful, this commit removes it.

This commit also changes the errcode in the error message that
pg_copy_logical_replication_slot() emits when the slot with an invalid
restart_lsn is specified, to more appropriate one.

Back-patch to v13 where max_slot_wal_keep_size was added and
the assertion was no longer valid.

Author: Fujii Masao
Reviewed-by: Alvaro Herrera, Kyotaro Horiguchi
Discussion: https://postgr.es/m/f91de4fb-a7ab-b90e-8132-74796e049d51@oss.nttdata.com

I forgot that INT64_FORMAT can't be used with sscanf on Windows.
Use the same trick of sscanf'ing into a temp variable as we do in
some other places in zic.c.

The upstream IANA code avoids the portability problem by relying on
<inttypes.h>'s SCNdFAST64 macro.  Once we're requiring C99 in all
branches, we should do likewise and drop this set of diffs from
upstream.  For now, though, a hack seems fine, since we do not
actually care about leapseconds anyway.

Discussion: https://postgr.es/m/4e5d1a5b-143e-e70e-a99d-a3b01c1ae7c3@2ndquadrant.com

In pg_replication_slot, change output from normal/reserved/lost to
reserved/extended/unreserved/ lost, which better expresses the possible
states particularly near the time where segments are no longer safe but
checkpoint has not run yet.

Under the new definition, reserved means the slot is consuming WAL
that's still under the normal WAL size constraints; extended means it's
consuming WAL that's being protected by wal_keep_segments or the slot
itself, whose size is below max_slot_wal_keep_size; unreserved means the
WAL is no longer safe, but checkpoint has not yet removed those files.
Such as slot is in imminent danger, but can still continue for a little
while and may catch up to the reserved WAL space.

Also, there were some bugs in the calculations used to report the
status; fixed those.

Backpatch to 13.
Reported-by: Fujii Masao <masao.fujii@oss.nttdata.com>
Author: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Fujii Masao <masao.fujii@oss.nttdata.com>
Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/20200616.120236.1809496990963386593.horikyota.ntt@gmail.com

We put it aside as invalidated_at, which let us show "lost" in
pg_replication slot.  Prior to this change, the state value was reported
as NULL.

Backpatch to 13.

Author: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/20200617.101707.1735599255100002667.horikyota.ntt@gmail.com
Discussion: https://postgr.es/m/20200407.120905.1507671100168805403.horikyota.ntt@gmail.com

The current definition is dangerous.  No bugs exist in our code at
present, but backpatch to 11 nonetheless where it was introduced.

Author: Álvaro Herrera <alvherre@alvh.no-ip.org>

The description of InsertPgAttributeTuple() does not match its handling
of pg_attribute contents with NULL values for a long time, with 911e7020
making things more inconsistent.  This adjusts the description to match
the reality.

Author: Daniel Gustafsson
Discussion: https://postgr.es/m/4E4E4B33-9FDF-4D21-B77A-642D027AEAD9@yesql.se

Reported-by: Justin Pryzby
Discussion: https://postgr.es/m/20200620220402.GZ17995@telsasoft.com
Backport-through: 13

explain_get_index_name() applied quote_identifier() to the index name.
This is fine for text output, but the non-text output formats all have
their own quoting conventions and would much rather start from the
actual index name.  For example in JSON you'd get something like

       "Index Name": "\"My Index\"",

which is surely not desirable, especially when the same does not
happen for table names.  Hence, move the responsibility for applying
quoting out to the callers, where it can go into already-existing
special code paths for text format.

This changes the API spec for users of explain_get_index_name_hook:
before, they were supposed to apply quote_identifier() if necessary,
now they should not.  Research suggests that the only publicly
available user of the hook is hypopg, and it actually forgot to
apply quoting anyway, so it's fine.  (In any case, there's no
behavioral change for the output of a hook as seen in non-text
EXPLAIN formats, so this won't break any case that programs should
be relying on.)

Digging in the commit logs, it appears that quoting was included in
explain_get_index_name's duties when commit 604ffd28 invented it;
and that was fine at the time because we only had text output format.
This should have been rethought when non-text formats were invented,
but it wasn't.

This is a fairly clear bug for users of non-text EXPLAIN formats,
so back-patch to all supported branches.

Per bug #16502 from Maciek Sakrejda.  Patch by me (based on
investigation by Euler Taveira); thanks to Julien Rouhaud for review.

Discussion: https://postgr.es/m/16502-57bd1c9f913ed1d1@postgresql.org

Some fields related to pg_opclass and pg_opfamily were using incorrect
markups, listing them as structname instead of structfield.

Author: Fabien Coelho
Discussion: https://postgr.es/m/alpine.DEB.2.22.394.2006210903560.859381@pseudo

Both INDEX_CLEANUP and TRUNCATE have been available since v12, and are
enabled by default except if respectively vacuum_index_cleanup and
vacuum_truncate are disabled for a given relation.  This change adds
support for disabling these options from vacuumdb.

Author: Nathan Bossart
Reviewed-by: Michael Paquier, Masahiko Sawada
Discussion: https://postgr.es/m/6F7F17EF-B1F2-4681-8D03-BA96365717C0@amazon.com

Discussion: https://postgr.es/m/20200620232145.GB17995%40telsasoft.com
Author: Justin Pryzby
Backpatch-through: 13

Defining duplicates as "close by" to each other was unclear.  Simplify
the definition.

Backpatch: 13-, where deduplication was introduced (by commit 0d861bbb)

Reported-by: Peter Geoghegan
Discussion: https://postgr.es/m/CAH2-WzmwhYbxuoL0WjTLaiCxW3gj6qadeNpBhWAo_KZsE5-FGw%40mail.gmail.com

spg_mask() didn't take into account that pd_lower equal to SizeOfPageHeaderData
is still valid value.  This commit fixes that.  Backpatch to 11, where
spg_mask() pg_lower check was introduced.

Reported-by: Michael Paquier
Discussion: https://postgr.es/m/20200615131405.GM52676%40paquier.xyz
Backpatch-through: 11

911e7020 added opclass options and adjusted documentation for each
particular affected opclass.  However, documentation for extendability was
not adjusted.  This commit adjusts documentation for interfaces of index AMs
and opclasses.

Discussion: https://postgr.es/m/CAH2-WzmQnW6%2Bz5F9AW%2BSz%2BzEcEvXofTwh_A9J3%3D_WA-FBP0wYg%40mail.gmail.com
Author: Alexander Korotkov
Reported-by: Peter Geoghegan
Reviewed-by: Peter Geoghegan

The function has been reading global variable forceSyncCommit, mirroring
the intent of the caller that passed forceSync=forceSyncCommit.  The
other caller, RecordTransactionCommitPrepared(), passed false.  Since
COMMIT PREPARED can't share a transaction with any command, it certainly
doesn't share a transaction with a command that sets forceSyncCommit.

Reviewed by Michael Paquier.

Discussion: https://postgr.es/m/20200617032615.GC2916904@rfd.leadboat.com

The function parameter column_no is not used in CopyReadBinaryAttribute,
this can be removed.

Commit 0e319c7a removed the usage of column_no parameter in function
CopyReadBinaryAttribute but forgot to remove the parameter.

Reported-by: Vignesh C
Author: Vignesh C
Discussion: https://postgr.es/m/CALDaNm1TYSNTfqx_jfz9_mwEZ2Er=dZnu++duXpC1uQo1cG=WA@mail.gmail.com

A few places calling fwrite and gzwrite were not setting errno to ENOSPC
when reporting errors, as is customary; this led to some failures being
reported as
"could not write file: Success"
which makes us look silly.  Make a few of these places in pg_dump and
pg_basebackup use our customary pattern.

Backpatch-to: 9.5
Author: Justin Pryzby <pryzby@telsasoft.com>
Author: Tom Lane <tgl@sss.pgh.pa.us>
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/20200611153753.GU14879@telsasoft.com

The IANA time zone folk have deprecated use of a "posixrules" file in
the tz database.  While for now it's our choice whether to keep
supplying one in our own builds, installations built with
--with-system-tzdata will soon be needing to cope with that file not
being present, at least on some platforms.

This causes a problem for the horology test, which expected the
nonstandard POSIX zone spec "CST7CDT" to apply pre-2007 US daylight
savings rules.  That does happen if the posixrules file supplies such
information, but otherwise the test produces undesired results.
To fix, add an explicit transition date rule that matches 2005 practice.
(We could alternatively have switched the test to use some real time
zone, but it seems useful to have coverage of this type of zone spec.)

While at it, update a documentation example that also relied on
"CST7CDT"; use a real-world zone name instead.  Also, document why
the zone names EST5EDT, CST6CDT, MST7MDT, PST8PDT aren't subject to
similar failures when "posixrules" is missing.

Back-patch to all supported branches, since the hazard is the same
for all.

Discussion: https://postgr.es/m/1665379.1592581287@sss.pgh.pa.us

Mostly in response to Jürgen Purtz critique of previous definitions,
though I added many other changes.

Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Reviewed-by: Jürgen Purtz <juergen@purtz.de>
Reviewed-by: Justin Pryzby <pryzby@telsasoft.com>
Reviewed-by: Erik Rijkers <er@xs4all.nl>
Discussion: https://postgr.es/m/c1e06008-2132-30f4-9b38-877e8683d418@purtz.de

It was possible for deduplication's single value strategy to mistakenly
believe that a very small duplicate tuple counts as one of the six large
tuples that it aims to leave behind after the page finally splits.  This
could cause slightly suboptimal space utilization with very low
cardinality indexes, though only under fairly narrow conditions.

To fix, be particular about what kind of tuple counts as a
maxpostingsize-capped tuple.  This avoids confusion in the event of a
small tuple that gets "wedged" between two large tuples, where all
tuples on the page are duplicates of the same value.

Discussion: https://postgr.es/m/CAH2-Wz=Y+sgSFc-O3LpiZX-POx2bC+okec2KafERHuzdVa7-rQ@mail.gmail.com
Backpatch: 13-, where deduplication was introduced (by commit 0d861bbb)

This commit fixes the following issues.

1. There is the case where the slot is dropped while trying to invalidate it.
    InvalidateObsoleteReplicationSlots() did not handle this case, and
    which could cause checkpoint to fail.

2. InvalidateObsoleteReplicationSlots() could emit the same log message
    multiple times unnecessary. It should be logged only once.

3. When marking the slot as used, we always searched the target slot from
    all the replication slots even if we already found it. This could cause
    useless waste of cycles.

Back-patch to v13 where these issues were added as a part of
max_slot_wal_keep_size code.

Author: Fujii Masao
Reviewed-by: Kyotaro Horiguchi, Alvaro Herrera
Discussion: https://postgr.es/m/66c05b67-3396-042c-1b41-bfa6c3ddcf82@oss.nttdata.com

Since 1f39bce0, HashAgg nodes have had the ability to spill to disk when
memory consumption exceeds work_mem. That commit added new properties to
EXPLAIN ANALYZE to show the maximum memory usage and disk usage, however,
it didn't quite go as far as showing that information for parallel
workers.  Since workers may have experienced something very different from
the main process, we should show this information per worker, as is done
in Sort.

Reviewed-by: Justin Pryzby
Reviewed-by: Jeff Davis
Discussion: https://postgr.es/m/CAApHDvpEKbfZa18mM1TD7qV6PG+w97pwCWq5tVD0dX7e11gRJw@mail.gmail.com
Backpatch-through: 13, where the hashagg spilling code was added.

Users of spinlocks should use spin.h, not s_lock.h. And lwlock.h
hasn't utilized spinlocks for quite a while.

Discussion: https://postgr.es/m/20200618183041.upyrd25eosecyf3x@alap3.anarazel.de

This was a danger only for --disable-spinlocks in combination with
atomic operations unsupported by the current platform.

While atomics.c was careful to signal that a separate semaphore ought
to be used when spinlock emulation is active, spin.c didn't actually
implement that mechanism. That's my (Andres') fault, it seems to have
gotten lost during the development of the atomic operations support.

Fix that issue and add test for nesting atomic operations inside a
spinlock.

Author: Andres Freund
Discussion: https://postgr.es/m/20200605023302.g6v3ydozy5txifji@alap3.anarazel.de
Backpatch: 9.5-

As s_lock_test, the already existing test for spinlocks, isn't run in
an automated fashion (and doesn't test a normal backend environment),
adding tests that are run as part of a normal regression run is a good
idea. Particularly in light of several recent and upcoming spinlock
related fixes.

Currently the new tests are run as part of the pre-existing
test_atomic_ops() test. That perhaps can be quibbled about, but for
now seems ok.

The only operations that s_lock_test tests but the new tests don't are
the detection of a stuck spinlock and S_LOCK_FREE (which is otherwise
unused, not implemented on all platforms, and will be removed).

This currently contains a test for more than INT_MAX spinlocks (only
run with --disable-spinlocks), to ensure the recent commit fixing a
bug with more than INT_MAX spinlock initializations is correct. That
test is somewhat slow, so we might want to disable it after a few
days.

It might be worth retiring s_lock_test after this. The added coverage
of a stuck spinlock probably isn't worth the added complexity?

Author: Andres Freund
Discussion: https://postgr.es/m/20200606023103.avzrctgv7476xj7i@alap3.anarazel.de

We'd glossed over most of this complexity for years, but it's hard
to avoid writing it all down now, so that we can explain what happens
when there's no "posixrules" file in the IANA time zone database.
That was at best a tiny minority situation till now, but it's likely
to become quite common in the future, so we'd better explain it.

Nonetheless, we don't really encourage people to use POSIX zone specs;
picking a named zone is almost always what you really want, unless
perhaps you're stuck with an out-of-date zone database.  Therefore,
let's shove all this detail into an appendix.

Patch by me; thanks to Robert Haas for help with some awkward wording.

Discussion: https://postgr.es/m/1390.1562258309@sss.pgh.pa.us

Advancing a replication slot did not recompute the oldest xmin and LSN
values across replication slots, preventing resource removal like
segments not recycled at checkpoint time.  The original commit that
introduced the slot advancing in 9c7d06d6 never did the update of those
oldest values, and b0afdcad removed this code.

This commit adds a TAP test to check segment recycling with advancing
for physical slots, enforcing an extra segment switch before advancing
to check if the segment gets correctly recycled after a checkpoint.

Reported-by: Andres Freund
Reviewed-by: Alexey Kondratov, Kyptaro Horiguchi
Discussion: https://postgr.es/m/20200609171904.kpltxxvjzislidks@alap3.anarazel.de
Backpatch-through: 11

The previous implementation returned 1 for all negative numbers, which
is not sensible under any definition.

Discussion: https://www.postgresql.org/message-id/flat/6ce1df0e-86a3-e544-743a-f357ff663f68%402ndquadrant.com

Move from int4 to numeric test.  (They were originally int4 functions,
but were reimplemented for numeric in
04a4821a.)  Add some tests for edge
cases.

Discussion: https://www.postgresql.org/message-id/flat/6ce1df0e-86a3-e544-743a-f357ff663f68%402ndquadrant.com

testtablespace is an extra path used as tablespace location in the main
regression test suite, computed from --outputdir as defined by the
caller of pg_regress (current directory if undefined).

This special handling was introduced as of f10589e5 to be specific to
MSVC, as we let pg_regress' Makefile handle this cleanup in other
environments.  This moves the cleanup to the MSVC script running
regression tests instead where needed: check, installcheck and
upgradecheck.  I have also checked this patch on MSVC with repeated runs
of each target.

Author: Kyotaro Horiguchi, Michael Paquier
Discussion: https://postgr.es/m/20200219.142519.437573253063431435.horikyota.ntt@gmail.com

This absorbs a leap-second-related bug fix in localtime.c, and
teaches zic to handle an expiration marker in the leapseconds file.
Neither are of any interest to us (for the foreseeable future
anyway), but we need to stay more or less in sync with upstream.

Also adjust some over-eager changes in the README from commit 95733841.
I have no intention of making changes that require C99 in this code,
until such time as all the live back branches require C99.  Otherwise
back-patching will get too exciting.

For the same reason, absorb assorted whitespace and other cosmetic
changes from HEAD into the back branches; mostly this reflects use of
improved versions of pgindent.

All in all then, quite a boring update.  But I figured I'd get it
done while I was looking at this code.

Oversight in commit 0d861bbb.

Once the counter goes negative we ended up with spinlocks that errored
out on first use (due to check in tas_sema).

Author: Andres Freund
Reviewed-By: Robert Haas
Discussion: https://postgr.es/m/20200606023103.avzrctgv7476xj7i@alap3.anarazel.de
Backpatch: 9.5-

On platforms without support for 64bit atomic operations where we also
cannot rely on 64bit reads to have single copy atomicity, such atomics
are implemented using a spinlock based fallback. That means it's not
safe to even read such atomics from within a signal handler (since the
signal handler might run when the spinlock already is held).

To avoid this issue defer global barrier processing out of the signal
handler. Instead of checking local / shared barrier generation to
determine whether to set ProcSignalBarrierPending, introduce
PROCSIGNAL_BARRIER and always set ProcSignalBarrierPending when
receiving such a signal. Additionally avoid redundant work in
ProcessProcSignalBarrier if ProcSignalBarrierPending is unnecessarily.

Also do a small amount of other polishing.

Author: Andres Freund
Reviewed-By: Robert Haas
Discussion: https://postgr.es/m/20200609193723.eu5ilsjxwdpyxhgz@alap3.anarazel.de
Backpatch: 13-, where the code was introduced.

Don't use fread(), since that doesn't necessarily set errno. We could
use read() instead, but it's even better to use pg_pread(), which
allows us to avoid some extra calls to seek to the desired location in
the file.

Also, advertise a wait event while reading from a file, as we do for
most other places where we're reading data from files.

Patch by me, reviewed by Hamid Akhtar.

Discussion: http://postgr.es/m/CA+TgmobBw-3573vMosGj06r72ajHsYeKtksT_oTxH8XvTL7DxA@mail.gmail.com

Merge two calls to sendDir() that are exactly the same except for
the fifth argument. Adjust comments to match.

Also, don't bother checking whether tblspc_map_file is NULL. We
initialize it in all cases, so it can't be.

Patch by me, reviewed by Amit Kapila and Kyotaro Horiguchi.

Discussion: http://postgr.es/m/CA+TgmoYq+59SJ2zBbP891ngWPA9fymOqntqYcweSDYXS2a620A@mail.gmail.com

Commit 72d422a5 made xlog.c call
sendTablespace() with the 'sizeonly' argument set to true, which
required basebackup.c to export sendTablespace(). However, that's
kind of ugly, so instead defer the call to sendTablespace() until
basebackup.c regains control. That way, it can still be a static
function.

Patch by me, reviewed by Amit Kapila and Kyotaro Horiguchi.

Discussion: http://postgr.es/m/CA+TgmoYq+59SJ2zBbP891ngWPA9fymOqntqYcweSDYXS2a620A@mail.gmail.com

Add a separate enum for use in the locking APIs, which were the only
user.

Discussion: https://www.postgresql.org/message-id/flat/a6f91ead-0ce4-2a34-062b-7ab9813ea308%402ndquadrant.com