Commit 9dc718bd added a "logically unchanged by UPDATE" hinting
mechanism, which is currently used within nbtree indexes only (see
commit d168b666).  This mechanism determined whether or not the incoming
item is a logically unchanged duplicate (a duplicate needed only for
MVCC versioning purposes) once per row updated per non-HOT update.  This
approach led to memory leaks which were noticeable with an UPDATE
statement that updated sufficiently many rows, at least on tables that
happen to have an expression index.

On HEAD, fix the issue by adding a cache to the executor's per-index
IndexInfo struct.

Take a different approach on Postgres 14 to avoid an ABI break: simply
pass down the hint to all indexes unconditionally with non-HOT UPDATEs.
This is deemed acceptable because the hint is currently interpreted
within btinsert() as "perform a bottom-up index deletion pass if and
when the only alternative is splitting the leaf page -- prefer to delete
any LP_DEAD-set items first".  nbtree must always treat the hint as a
noisy signal about what might work, as a strategy of last resort, with
costs imposed on non-HOT updaters.  (The same thing might not be true
within another index AM that applies the hint, which is why the original
behavior is preserved on HEAD.)

Author: Peter Geoghegan <pg@bowt.ie>
Reported-By: Klaudie Willis <Klaudie.Willis@protonmail.com>
Diagnosed-By: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/261065.1639497535@sss.pgh.pa.us
Backpatch: 14-, where the hinting mechanism was added.

One of the comments introduced in b69aba7 was worded a bit weirdly, so
improve it.

Reported-by: Sergey Shinderuk
Discussion: https://postgr.es/m/71b9a5d2-a3bf-83bc-a243-93dcf0bcfb3b@postgrespro.ru
Backpatch-through: 14

Experimenting with FIPS mode enabled, I saw

regression=# \password joe
Enter new password for user "joe":
Enter it again:
could not encrypt password: disabled for FIPS
out of memory

because PQencryptPasswordConn was still of the opinion that "out of
memory" is always appropriate to print.

Minor oversight in b69aba745.  Like that one, back-patch to v14.

The existing cryptohash facility was causing problems in some code paths
related to MD5 (frontend and backend) that relied on the fact that the
only type of error that could happen would be an OOM, as the MD5
implementation used in PostgreSQL ~13 (the in-core implementation is
used when compiling with or without OpenSSL in those older versions),
could fail only under this circumstance.

The new cryptohash facilities can fail for reasons other than OOMs, like
attempting MD5 when FIPS is enabled (upstream OpenSSL allows that up to
1.0.2, Fedora and Photon patch OpenSSL 1.1.1 to allow that), so this
would cause incorrect reports to show up.

This commit extends the cryptohash APIs so as callers of those routines
can fetch more context when an error happens, by using a new routine
called pg_cryptohash_error().  The error states are stored within each
implementation's internal context data, so as it is possible to extend
the logic depending on what's suited for an implementation.  The default
implementation requires few error states, but OpenSSL could report
various issues depending on its internal state so more is needed in
cryptohash_openssl.c, and the code is shaped so as we are always able to
grab the necessary information.

The core code is changed to adapt to the new error routine, painting
more "const" across the call stack where the static errors are stored,
particularly in authentication code paths on variables that provide
log details.  This way, any future changes would warn if attempting to
free these strings.  The MD5 authentication code was also a bit blurry
about the handling of "logdetail" (LOG sent to the postmaster), so
improve the comments related that, while on it.

The origin of the problem is 87ae9691, that introduced the centralized
cryptohash facility.  Extra changes are done for pgcrypto in v14 for the
non-OpenSSL code path to cope with the improvements done by this
commit.

Reported-by: Michael Mühlbeyer
Author: Michael Paquier
Reviewed-by: Tom Lane
Discussion: https://postgr.es/m/89B7F072-5BBE-4C92-903E-D83E865D9367@trivadis.com
Backpatch-through: 14

I had a brain fade in commit d3289915, and used 2:30AM as the
example timestamp for both spring-forward and fall-back cases.
But it's not actually ambiguous at all in the fall-back case,
because that transition is from 2AM to 1AM under USA rules.
Fix the example to use 1:30AM, which *is* ambiguous.

Noted while answering a question from Aleksander Alekseev.
Back-patch to all supported branches.

Discussion: https://postgr.es/m/2191355.1641828552@sss.pgh.pa.us

Juan José Santamaría Flecha

Backpatch to all live branches

If contrib/btree_gist is used to make a GIST index on a char(N)
(bpchar) column, and that column is retrieved via an index-only
scan, what came out had all trailing spaces removed.  Since
that doesn't happen in any other kind of table scan, this is
clearly a bug.  The cause is that gbt_bpchar_compress() strips
trailing spaces (using rtrim1) before a new index entry is made.
That was probably a good idea when this code was first written,
but since we invented index-only scans, it's not so good.

One answer could be to mark this opclass as incapable of index-only
scans.  But to do so, we'd need an extension module version bump,
followed by manual action by DBAs to install the updated version
of btree_gist.  And it's not really a desirable place to end up,
anyway.

Instead, let's fix the code by removing the unwanted space-stripping
action and adjusting the opclass's comparison logic to ignore
trailing spaces as bpchar normally does.  This will not hinder
cases that work today, since index searches with this logic will
act the same whether trailing spaces are stored or not.  It will
not by itself fix the problem of getting space-stripped results
from index-only scans, of course.  Users who care about that can
REINDEX affected indexes after installing this update, to immediately
replace all improperly-truncated index entries.  Otherwise, it can
be expected that the index's behavior will change incrementally as
old entries are replaced by new ones.

Per report from Alexander Lakhin.  Back-patch to all supported branches.

Discussion: https://postgr.es/m/696c995b-b37f-5526-f45d-04abe713179f@gmail.com

This addresses some problems in the describe queries used for extended
statistics:
- Two schema qualifications for the text type were missing for \dX.
- The list of extended statistics listed for a table through \d was
ordered based on the object OIDs, but it is more consistent with the
other commands to order by namespace and then by object name.
- A couple of aliases were not used in \d.  These are removed.

This is similar to commits 1f092a3 and 07f8a9e.

Author: Justin Pryzby
Discussion: https://postgr.es/m/20220107022235.GA14051@telsasoft.com
Backpatch-through: 14

Backpatch-through: 10

Instead of using a hardcoded or default path to the perl file the .bat
file is a wrapper for, we use a path that means the file is found in
the same directory as the .bat file.

Patch by Anton Voloshin, slightly tweaked by me.

Backpatch to all live branches

Discussion: https://postgr.es/m/2b7a674b-5fb0-d264-75ef-ecc7a31e54f8@postgrespro.ru

We disallow altering a column datatype within a regular table,
if the table's rowtype is used as a column type elsewhere,
because we lack code to go around and rewrite the other tables.
This restriction should apply to partitioned tables as well, but it
was not checked because ATRewriteTables and ATPrepAlterColumnType
were not on the same page about who should do it for which relkinds.

Per bug #17351 from Alexander Lakhin.  Back-patch to all supported
branches.

Discussion: https://postgr.es/m/17351-6db1870f3f4f612a@postgresql.org

The link used in the documentation is dead, and the only options to have
an access to this part of the SQL specification are not free.  Like any
other books referred, just remove the link to keep some neutrality but
keep its reference.

Reported-by: Erik Rijkers
Discussion: https://postgr.es/m/989abd7d-af30-ab52-1201-bf0b4f33b872@xs4all.nl
Backpatch-through: 12

get_rel_sync_entry(), which is called each time a change needs to be
logically replicated, is a rather hot code path in the WAL sender
sending logical changes.  This code path was doing a relcache access on
relkind and relpartition for each logical change, but we only need to
know this information when building or re-building the cached
information for a relation.

Some measurements prove that this is noticeable in perf profiles,
particularly when attempting to replicate changes from relations that
are not published as these cause less overhead in the WAL sender,
delaying further the replication of changes for relations that are
published.

Issue introduced in 83fd4532.

Author: Hou Zhijie
Reviewed-by: Kyotaro Horiguchi, Euler Taveira
Discussion: https://postgr.es/m/OS0PR01MB5716E863AA9E591C1F010F7A947D9@OS0PR01MB5716.jpnprd01.prod.outlook.com
Backpatch-through: 13

Under concurrency, it is possible for two sessions to be merrily locking
and releasing a tuple and marking it again as HEAP_XMAX_INVALID all the
while a third session attempts to lock it, miserably fails at it, and
then contemplates life, the universe and everything only to eventually
fail an assertion that said bit is not set.  Before SKIP LOCKED that was
indeed a reasonable expectation, but alas! commit df630b0d falsified
it.

This bug is as old as time itself, and even older, if you think time
begins with the oldest supported branch.  Therefore, backpatch to all
supported branches.

Author: Simon Riggs <simon.riggs@enterprisedb.com>
Discussion: https://postgr.es/m/CANbhV-FeEwMnN8yuMyss7if1ZKjOKfjcgqB26n8pqu1e=q0ebg@mail.gmail.com

Commit 4ace45677 failed to fix the problem fully, because the
same issue of attempting to fetch a non-returnable index column
can occur when rechecking the indexqual after using a lossy index
operator.  Moreover, it broke EXPLAIN for such indexquals (which
indicates a gap in our test cases :-().

Revert the code changes of 4ace45677 in favor of adding a new field
to struct IndexOnlyScan, containing a version of the indexqual that
can be executed against the index-returned tuple without using any
non-returnable columns.  (The restrictions imposed by check_index_only
guarantee this is possible, although we may have to recompute indexed
expressions.)  Support construction of that during setrefs.c
processing by marking IndexOnlyScan.indextlist entries as resjunk
if they can't be returned, rather than removing them entirely.
(We could alternatively require setrefs.c to look up the IndexOptInfo
again, but abusing resjunk this way seems like a reasonably safe way
to avoid needing to do that.)

This solution isn't great from an API-stability standpoint: if there
are any extensions out there that build IndexOnlyScan structs directly,
they'll be broken in the next minor releases.  However, only a very
invasive extension would be likely to do such a thing.  There's no
change in the Path representation, so typical planner extensions
shouldn't have a problem.

As before, back-patch to all supported branches.

Discussion: https://postgr.es/m/3179992.1641150853@sss.pgh.pa.us
Discussion: https://postgr.es/m/17350-b5bdcf476e5badbb@postgresql.org

Since 4f0b0966, pgss_store() does nothing if compute_query_id is disabled
or if no other module computed a query identifier, but the top comment
of this function did not reflect that.  This behavior is already
documented in its own code path, and this just removes the inconsistent
comment.

Author: Kyotaro Horiguchi
Reviewed-by: Julien Rouhaud
Discussion: https://postgr.es/m/20211122.153823.1325120762360533122.horikyota.ntt@gmail.com
Backpatch-through: 14

Reported-By: Eric Mutta
Backpatch-through: 10
Discussion: https://postgr.es/m/164052477973.21665.7888120874624887609@wrigleys.postgresql.org

If an index has both returnable and non-returnable columns, and one of
the non-returnable columns is an expression using a Var that is in a
returnable column, then a query returning that expression could result
in an index-only scan plan that attempts to read the non-returnable
column, instead of recomputing the expression from the returnable
column as intended.

To fix, redefine the "indextlist" list of an IndexOnlyScan plan node
as containing null Consts in place of any non-returnable columns.
This solves the problem by preventing setrefs.c from falsely matching
to such entries.  The executor is happy since it only cares about the
exposed types of the entries, and ruleutils.c doesn't care because a
correct plan won't reference those entries.  I considered some other
ways to prevent setrefs.c from doing the wrong thing, but this way
seems good since (a) it allows a very localized fix, (b) it makes
the indextlist structure more compact in many cases, and (c) the
indextlist is now a more faithful representation of what the index AM
will actually produce, viz. nulls for any non-returnable columns.

This is easier to hit since we introduced included columns, but it's
possible to construct failing examples without that, as per the
added regression test.  Hence, back-patch to all supported branches.

Per bug #17350 from Louis Jachiet.

Discussion: https://postgr.es/m/17350-b5bdcf476e5badbb@postgresql.org

The reverted commit attempted to fix SQL specification compliance for
the cases which 6aaaa76b left.  This however broke existing behavior
which takes precedence over spec compliance so revert. The introduced
tests are left after the revert since the codepath isn't well covered.
Per bug report 17346. Backpatch down to 14 where it was introduced.
Reported-by: Andrew Bille <andrewbille@gmail.com>
Discussion: https://postgr.es/m/17346-f72b28bd1a341060@postgresql.org

Avoid the name "test".  In the 10 branch, this could clash with
alter_table.sql, as seen in the build farm.  That other instance was
already renamed in later branches by commit 2cf8c7aa, but it's good to
future-proof the name here too.

Back-patch to 10.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/CA%2BhUKGJf4RAXUyAYVUcQawcptX%3DnhEco3SYpuPK5cCbA-F1eLA%40mail.gmail.com

Only non-HOT updates evaluate the index expression.

Reported-by: Chris Lowder

Discussion: https://postgr.es/m/163967385701.26064.15365003480975321072@wrigleys.postgresql.org

Backpatch-through: 10

catalog/pg_class.h was stating that REPLICA_IDENTITY_INDEX with a
dropped index is equivalent to REPLICA_IDENTITY_DEFAULT.  The code tells
a different story, as it is equivalent to REPLICA_IDENTITY_NOTHING.

The behavior exists since the introduction of replica identities, and
fe7fd4e9 even added tests for this case but I somewhat forgot to fix this
comment.

While on it, this commit reorganizes the documentation about replica
identities on the ALTER TABLE page, and a note is added about the case
of dropped indexes with REPLICA_IDENTITY_INDEX.

Author: Michael Paquier, Wei Wang
Reviewed-by: Euler Taveira
Discussion: https://postgr.es/m/OS3PR01MB6275464AD0A681A0793F56879E759@OS3PR01MB6275.jpnprd01.prod.outlook.com
Backpatch-through: 10

One code path related to this flavor of ALTER TABLE was checking that
the relation to detach has to be a normal table or a partitioned table,
which would fail if using the command with a different relation kind.

Views, sequences and materialized views cannot be part of a partition
tree, so these would cause the command to fail anyway, but the assertion
was triggered.  Foreign tables can be part of a partition tree, and
again the assertion would have failed.  The simplest solution is just to
remove this assertion, so as we get the same failure as the
non-concurrent code path.

While on it, add a regression test in postgres_fdw for the concurrent
partition detach of a foreign table, as per a suggestion from Alexander
Lakhin.

Issue introduced in 71f4c8c6.

Reported-by: Alexander Lakhin
Author: Michael Paquier, Alexander Lakhin
Reviewed-by: Peter Eisentraut, Kyotaro Horiguchi
Discussion: https://postgr.es/m/17339-a9e09aaf38a3457a@postgresql.org
Backpatch-through: 14

Fix the code changed by commit 5c056b0c2 so that we always generate
RelabelType, not something else, for a cast to unspecified typmod.
Otherwise planner optimizations might not happen.

It appears we missed this point because the previous experiments were
done on type numeric: the parser undesirably generates a call on the
numeric() length-coercion function, but then numeric_support()
optimizes that down to a RelabelType, so that everything seems fine.
It misbehaves for types that have a non-optimized length coercion
function, such as bpchar.

Per report from John Naylor.  Back-patch to all supported branches,
as the previous patch eventually was.  Unfortunately, that no longer
includes 9.6 ... we really shouldn't put this type of change into a
nearly-EOL branch.

Discussion: https://postgr.es/m/CAFBsxsEfbFHEkouc+FSj+3K1sHipLPbEC67L0SAe-9-da8QtYg@mail.gmail.com

edc2332 has introduced in vcregress.pl some control on the environment
variables LZ4, TAR and GZIP_PROGRAM to allow any TAP tests to be able
use those commands.  This makes the settings more consistent with
src/Makefile.global.in, as the same default gets used for Make and MSVC
builds.

Each parameter can be changed in buildenv.pl, but as a default gets
assigned after loading buldenv.pl, it is not possible to unset any of
these, and using an empty value would not work with "||=" either.  As
some environments may not have a compatible command in their PATH (tar
coming from MinGW is an issue, for one), this could break tests without
an exit path to bypass any failing test.  This commit changes things so
as the default values for LZ4, TAR and GZIP_PROGRAM are assigned before
loading buildenv.pl, not after.  This way, we keep the same amount of
compatibility as a GNU build with the same defaults, and it becomes
possible to unset any of those values.

While on it, this adds some documentation about those three variables in
the section dedicated to the TAP tests for MSVC.

Per discussion with Andrew Dunstan.

Discussion: https://postgr.es/m/YbGYe483803il3X7@paquier.xyz
Backpatch-through: 10

This could only matter if (a) long is wider than int, and (b) the heap
of free blocks exceeds UINT_MAX entries, which seems pretty unlikely.
Still, it's a theoretical bug, so backpatch to v13 where the typo came
in (in commit c02fdc92).

In passing, also make swap_nodes() use consistent datatypes.

Ma Liangzhu

Discussion: https://postgr.es/m/17336-fc4e522d26a750fd@postgresql.org

When using replication origins, pg_replication_origin_xact_setup() is an
optional choice to be able to set a LSN and a timestamp to mark the
origin, which would be additionally added to WAL for transaction commits
or aborts (including 2PC transactions).  An assertion in the code path
of PREPARE TRANSACTION assumed that this data should always be set, so
it would trigger when using replication origins without setting up an
origin LSN.  Some tests are added to cover more this kind of scenario.

Oversight in commit 1eb6d652.

Per discussion with Amit Kapila and Masahiko Sawada.

Discussion: https://postgr.es/m/YbbBfNSvMm5nIINV@paquier.xyz
Backpatch-through: 11

In commit 791090bd, I made an effort to fill in documentation
for all geometric operators listed in pg_operator.  However,
it now appears that at least some of the omissions may have been
intentional, because some of those operator entries point at
unimplemented stub functions.  Remove those from the docs again.

(In HEAD, poly_distance stays, because c5c192d7b just added an
implementation for it.)

Per complaint from Anton Voloshin.

Discussion: https://postgr.es/m/3426566.1638832718@sss.pgh.pa.us

When writing / debugging an isolation test it sometimes is useful to see which
session holds what lock etc. To make it easier, both as part of spec files and
interactively, append the session name to application_name. Since b1907d68
application_name already contains the test name, this appends the session's
name to that.

insert-conflict-specconflict did something like this manually, which can now
be removed.

As we have done lately with other test infrastructure improvements, backpatch
this change, to make it easier to backpatch tests.

Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Michael Paquier <michael@paquier.xyz>
Reviewed-By: Andrew Dunstan <andrew@dunslane.net>
Discussion: https://postgr.es/m/20211211012052.2blmzcmxnxqawd2z@alap3.anarazel.de
Backpatch: 10-, to make backpatching of tests easier.

The multirange_get_range() function fails when two boundaries of the same
range have different alignments.  Fix that by adding proper pointer alignment.

Reported-by: Alexander Lakhin
Discussion: https://postgr.es/m/17300-dced2d01ddeb1f2f%40postgresql.org
Backpatch-through: 14

One of the changes impacts the documentation, so backpatch.

Author: Peter Smith
Discussion: https://postgr.es/m/CAHut+Pu6+c+r3mY24VT7u+H+E_s6vMr5OdRiZ8NT3EOa-E5Lmw@mail.gmail.com
Backpatch-through: 14

We publish the child table's data twice for a publication that has both
child and parent tables and is published with publish_via_partition_root
as true. This happens because subscribers will initiate synchronization
using both parent and child tables, since it gets both as separate tables
in the initial table list.

Ensure that pg_publication_tables returns only parent tables in such
cases.

Author: Hou Zhijie
Reviewed-by: Greg Nancarrow, Amit Langote, Vignesh C, Amit Kapila
Backpatch-through: 13
Discussion: https://postgr.es/m/OS0PR01MB57167F45D481F78CDC5986F794B99@OS0PR01MB5716.jpnprd01.prod.outlook.com

List types numeric and timestamptz, which don't seem to have ever been
included here.  Restore bigint, which was no-doubt-accidentally deleted
in v12.  Fix some errors, or at least obsolete usages (nobody declares
float arguments as "float8*" anymore, even though they might be that
under the hood).  Re-alphabetize.  Remove the seeming claim that this
is a complete list of built-in types.

Per question from Oskar Stenberg.

Discussion: https://postgr.es/m/HE1PR03MB2971DE2527ECE1E99D6C19A8F96E9@HE1PR03MB2971.eurprd03.prod.outlook.com

This reverts commit f920f7e799c587228227ec94356c760e3f3d5f2b.

The patch in effect fixed a problem we didn't have and caused another
instead.

Backpatch to release 14 like original

Discussion: https://postgr.es/m/3655283.1638977975@sss.pgh.pa.us

Issue exposed by commit edc2332550 and the buildfarm.

Backpatch to release 14 where this usage started.

This happens because we were passing incorrect arguments to
ReorderBufferFinishPrepared().

Author: Masahiko Sawada
Reviewed-by: Vignesh C
Backpatch-through: 14
Discussion: https://postgr.es/m/CAD21AoBqhUqgDZUhUVnnwKRubPDNJ6m6fJDPgok3E5cWJLL+pA@mail.gmail.com

REINDEX CONCURRENTLY run on a toast index or a toast relation could
corrupt the target indexes rebuilt, as a backend running in parallel
that manipulates toast values would directly release the lock on the
toast relation when its local operation is done, rather than releasing
the lock once the transaction that manipulated the toast values
committed.

The fix done here is simple: we now hold a ROW EXCLUSIVE lock on the
toast relation when saving or deleting a toast value until the
transaction working on them is committed, so as a concurrent reindex
happening in parallel would be able to wait for any activity and see any
new rows inserted (or deleted).

An isolation test is added to check after the case fixed here, which is
a bit fancy by design as it relies on allow_system_table_mods to rename
the toast table and its index to fixed names.  This way, it is possible
to reindex them directly without any dependency on the OID of the
underlying relation.  Note that this could not use a DO block either, as
REINDEX CONCURRENTLY cannot be run in a transaction block.  The test is
backpatched down to 13, where it is possible, thanks to c4a7a392, to use
allow_system_table_mods in a test suite.

Reported-by: Alexey Ermakov
Analyzed-by: Andres Freund, Noah Misch
Author: Michael Paquier
Reviewed-by: Nathan Bossart
Discussion: https://postgr.es/m/17268-d2fb426e0895abd4@postgresql.org
Backpatch-through: 12

Certain settings from configuration or the Makefile infrastructure are
used by the TAP tests, but were not being set up by vcregress.pl. This
remedies those omissions. This should increase test coverage, especially
on the buildfarm.

Reviewed by Noah Misch

Discussion: https://postgr.es/m/17093da5-e40d-8335-d53a-2bd803fc38b0@dunslane.net

Backpatch to all live branches.

Further experimentation shows that commit 6051857fc is not sufficient
when using (some versions of?) OpenSSL.  The reason is obscure, but
calling shutdown(socket, SD_SEND) improves matters.

Per testing by Andrew Dunstan and Alexander Lakhin.
Back-patch as before.

Discussion: https://postgr.es/m/af5e0bf3-6a61-bb97-6cba-061ddf22ff6b@dunslane.net