Author: Amul Sul <sulamul@gmail.com>
Reviewed-by: Jeevan Ladhe <jeevan.ladhe@enterprisedb.com>
Reviewed-by: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Reviewed-by: Justin Pryzby <pryzby@telsasoft.com>
Discussion: https://postgr.es/m/CAAJ_b94HaNcrPVREUuB9-qUn2uB+gfcoX3FG_Vx0S6aFse+yhw@mail.gmail.com

Backpatch all the way back to 9.6.

Author: Simon Riggs <simon.riggs@enterprisedb.com>
Discussion: https://postgr.es/m/CANbhV-EwxvdhHuOLdfG2ciYrHOHXV=mm6=fD5aMhqcH09Li3Tg@mail.gmail.com

This was forgotten in 71f4c8c6.
Reported-by: Pavel Luzanov <p.luzanov@postgrespro.ru>
Author: Amit Langote <amitlangote09@gmail.com>
Discussion: https://postgr.es/m/0688e7c3-8bc8-a3e4-9d8e-3bcbbf3e1f4d@postgrespro.ru

The all_visible_according_to_vm variable's value is inherently prone to
becoming invalidated concurrently, since it is set before we even
acquire a lock on a related heap page buffer.

Oversight in commit 7136bf34, which added the assertion in passing.

Author: Masahiko Sawada <sawada.mshk@gmail.com>
Reported-By: Tang <tanghy.fnst@fujitsu.com>
Diagnosed-By: : Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/CAD21AoDzgc8_MYrA5m1fyydomw_eVKtQiYh7sfDK4KEhdMsf_g@mail.gmail.com

In d6b8d294 I (Álvaro) was sloppy about recording whether a
partition descripor does or does not include detached partitions, when
the snapshot checking does not see the pg_inherits row marked detached.
In that case no partition was omitted, yet in the relcache entry we were
saving the partdesc as omitting partitions.  Flip that (so we save it as
a partdesc not omitting partitions, which indeed it doesn't), which
hopefully makes the code easier to reason about.

Author: Amit Langote <amitlangote09@gmail.com>
Discussion: https://postgr.es/m/CA+HiwqE7GxGU4VdzwZzfiz+Ont5SsopoFkgtrZGEdPqWRL+biA@mail.gmail.com

Improve self-referential foreign key example, per suggestion
from David Johnston.

Discussion: https://postgr.es/m/CAKFQuwZTke7+HUn4YUGqu2+gAPi4Cy18TXMrg_Z5nADkxfPNMw@mail.gmail.com

The grammar changes in commit bbe0a81d
allow SET COMPRESSION to be used with ALTER MATERIALIZED VIEW as
well as with ALTER TABLE, so update those docs to say that it works.

Also, update the documentation for the pg_column_compression()
to explain that it will return NULL when there's no relevant value.

Patch by me, per concerns from Michael Paquier.

Discussion: http://postgr.es/m/CA+Tgmob9h5u4iNL9KM0drZgkY-JL4oCVW0dWrMqtLPQ1zHkquA@mail.gmail.com

Justin Pryzby, per a complaint from Michael Paquier. Reviewed by
Dilip Kumar and by me.

Discussion: http://postgr.es/m/20210429040132.GF27406@telsasoft.com

Currently, replication slot statistics are updated at prepare, commit, and
rollback. Now, if the transaction is interrupted the stats might not get
updated. Fixed this by updating replication statistics after every
stream/spill.

In passing update the docs to change the description of some of the slot
stats.

Author: Vignesh C, Sawada Masahiko
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de

Reported-By: Erik Rijkers <er@xs4all.nl>
Discussion: https://postgr.es/m/833107370.1313189.1619647621213@webmailclassic.xs4all.nl
Backpatch: 13, where b059d2f4 introduced the issue.

Update checklist to reflect current practice:

* The platform-specific FAQ files are long gone.

* We've never routinely updated the libbind code we borrowed, either,
and there seems no reason to start now.

* Explain current practice of running pgindent twice per cycle.

Discussion: https://postgr.es/m/4038398.1620238684@sss.pgh.pa.us

During decoding of an in-progress or prepared transaction, we detect
concurrent abort with an error code ERRCODE_TRANSACTION_ROLLBACK. That is
not sufficient because a callback can decide to throw that error code
at other times as well.

Reported-by: Tom Lane
Author: Amit Kapila
Reviewed-by: Dilip Kumar
Discussion: https://postgr.es/m/CAA4eK1KCjPRS4aZHB48QMM4J8XOC1+TD8jo-4Yu84E+MjwqVhA@mail.gmail.com

Commit 0325d7a5 made this unused but forgot to remove it. Do so now.

Author: Amit Langote <amitlangote09@gmail.com>
Discussion: https://postgr.es/m/209c99fe-b9a2-94f4-cd68-a8304186a09e@lab.ntt.co.jp

When ALTER TABLE .. ALTER CONSTRAINT changes deferrability properties
changed in a partitioned table, we failed to propagate those changes
correctly to partitions and to triggers.  Repair by adding a recursion
mechanism to affect all derived constraints and all derived triggers.
(In particular, recurse to partitions even if their respective parents
are already in the desired state: it is possible for the partitions to
have been altered individually.)  Because foreign keys involve tables in
two sides, we cannot use the standard ALTER TABLE recursion mechanism,
so we invent our own by following pg_constraint.conparentid down.

When ALTER TABLE .. ALTER CONSTRAINT is invoked on the derived
pg_constraint object that's automaticaly created in a partition as a
result of a constraint added to its parent, raise an error instead of
pretending to work and then failing to modify all the affected triggers.
Before this commit such a command would be allowed but failed to affect
all triggers, so it would silently misbehave.  (Restoring dumps of
existing databases is not affected, because pg_dump does not produce
anything for such a derived constraint anyway.)

Add some tests for the case.

Backpatch to 11, where foreign key support was added to partitioned
tables by commit 3de241db.  (A related change is commit f56f8f8d
in pg12 which added support for FKs *referencing* partitioned tables;
this is what forces us to use an ad-hoc recursion mechanism for this.)

Diagnosed by Tom Lane from bug report from Ron L Johnson.  As of this
writing, no reviews were offered.

Discussion: https://postgr.es/m/75fe0761-a291-86a9-c8d8-4906da077469@gmail.com
Discussion: https://postgr.es/m/3144850.1607369633@sss.pgh.pa.us

Previously, a lot of information about type regclass existed only
in the discussion of the sequence functions.  Maybe that made sense
in the beginning, because I think originally those were the only
functions taking regclass.  But it doesn't make sense anymore.
Move that material to the "Object Identifier Types" section in
datatype.sgml, generalize it to talk about the other reg* types
as well, and add more examples.

Per bug #16991 from Federico Caselli.

Discussion: https://postgr.es/m/16991-bcaeaafa17e0a723@postgresql.org

Commit 93f41461 improved a pre-existing test case so that it would
show whether or not termination of the "remote" worker process happened.
This soon exposed that, when debug_invalidate_system_caches_always
(nee CLOBBER_CACHE_ALWAYS) is enabled, no such termination occurs.
That's because cache invalidation forces postgres_fdw connections
to be dropped at end of transaction, so that there's no worker to
terminate.  There's a race condition as to whether the worker will
manage to get out of the BackendStatusArray before we look, but at
least on buildfarm member hyrax, it's failed twice in two attempts.

Rather than re-lobotomizing the test, let's fix this by transiently
disabling debug_invalidate_system_caches_always.  (Hooray for that
being just a GUC nowadays, rather than a compile-time option.)
If this proves not to be enough to make the test stable, we can
do the other thing instead.

Discussion: https://postgr.es/m/3854538.1620081771@sss.pgh.pa.us

The OID of the constraint is used instead of the OID of the trigger --
an easy mistake to make.  Apparently the object-alter hooks are not very
well tested :-(

Backpatch to 12, where this typo was introduced by 578b2297

Discussion: https://postgr.es/m/20210503231633.GA6994@alvherre.pgsql

The previous fix for dumping of inherited generated columns
(0bf83648) must not be applied to
partitions, since, unlike normal inherited tables, they are always
dumped separately and reattached.
Reported-by: Santosh Udupi <email@hitha.net>
Discussion: https://www.postgresql.org/message-id/flat/CACLRvHZ4a-%2BSM_159%2BtcrHdEqxFrG%3DW4gwTRnwf7Oj0UNj5R2A%40mail.gmail.com

When running ALTER TABLE t2 INHERIT t1, we must check that columns in
t2 that correspond to a generated column in t1 are also generated and
have the same generation expression.  Otherwise, this would allow
creating setups that a normal CREATE TABLE sequence would not allow.

Discussion: https://www.postgresql.org/message-id/22de27f6-7096-8d96-4619-7b882932ca25@2ndquadrant.com

We don't usually mention the version number in similar situations.  So, neither
mention it here.

Reported-by: Bruce Momjian
Discussion: https://postgr.es/m/20210503234914.GO6180%40momjian.us

Properly fix:

- the "ONLY" in FROM [ONLY] isn't hashed
- the agglevelsup field in GROUPING isn't hashed
- WITH TIES not being hashed (new in PG 13)
- "DISTINCT" in "GROUP BY [DISTINCT]" isn't hashed (new in PG 14)

Reported-by: Julien Rouhaud

Discussion: https://postgr.es/m/20210425081119.ulyzxqz23ueh3wuj@nol

Before now, looking up "multirange" in the index only led to the
multirange() function.  To make this more useful, also add an entry
pointing to the range types section.

Don't phrase reports in terms of the number of tuples thus-far
returned by the index scan, but rather in terms of the chunk_seq
values found inside the tuples.

Patch by me, reviewed by Mark Dilger.

Discussion: http://postgr.es/m/CA+TgmoZUONCkdcdR778EKuE+f1r5Obieu63db2OgMPHaEvEPTQ@mail.gmail.com

Commit 824bf719 introduced a new search of the NFAs generated by
regex compilation.  I failed to think hard about the performance
characteristics of that search, with the predictable outcome
that it's bad: weird regexes can trigger exponential search time.
Worse, there's no check-for-interrupt in that code, so you can't
even cancel the query if this happens.

Fix by introducing memo-ization of the search results, so that any one
NFA state need be examined in detail just once.  This potentially uses
a lot of memory, but we can bound the memory usage by putting a limit
on the number of states for which we'll try to prove match-all-ness.
That is sane because we already have a limit (DUPINF) on the maximum
finite string length that a matchall regex can match; and patterns
that involve much more than DUPINF states would probably exceed that
limit anyway.

Also, rearrange the logic so that we check the basic is-the-graph-
all-RAINBOW-arcs property before we start the recursive search to
determine path lengths.  This will ensure that we fall out quickly
whenever the NFA couldn't possibly be matchall.

Also stick in a check-for-interrupt, just in case these measures
don't completely eliminate the risk of slowness.

Discussion: https://postgr.es/m/3483895.1619898362@sss.pgh.pa.us

If dtrace is compiled in but disabled, the lwlock dtrace probes still
evaluate their arguments.  Since PostgreSQL 13, T_NAME(lock) does
nontrivial work, so it should be avoided if not needed.  To fix, make
these calls conditional on the *_ENABLED() macro corresponding to each
probe.
Reviewed-by: Craig Ringer <craig.ringer@enterprisedb.com>
Discussion: https://www.postgresql.org/message-id/CAGRY4nwxKUS_RvXFW-ugrZBYxPFFM5kjwKT5O+0+Stuga5b4+Q@mail.gmail.com

became unused by 04942bff

One more that ought to have been part of
82c3cd97.

Previously, we were using the size of all the changes present in
ReorderBuffer to compute total_bytes after decoding a transaction and that
can lead to counting some of the transactions' changes more than once. Fix
it by using the size of the changes decoded for a transaction to compute
'total_bytes'.

Author: Sawada Masahiko
Reviewed-by: Vignesh C, Amit Kapila
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de

websearch_to_tsquery() splits text in quotes into tokens and connects them with
phrase operator on its own.  However, that leads to surprising results when the
token contains no words.

For instance, websearch_to_tsquery('"aaa: bbb"') is 'aaa <2> bbb', because
it is equivalent of to_tsquery(E'aaa <-> \':\' <-> bbb').  But
websearch_to_tsquery('"aaa: bbb"') has to be 'aaa <-> bbb' in order to match
to_tsvector('aaa: bbb').

Since 0c4f355c, we anyway connect lexemes of complex tokens with phrase
operators.  Thus, let's just websearch_to_tsquery() parse text in quotes as
a single token.  Therefore, websearch_to_tsquery() should process the quoted
text in the same way phraseto_tsquery() does.  This solution is what we exactly
need and also simplifies the code.

This commit is an incompatible change, so we don't backpatch it.

Reported-by: Valentin Gatien-Baron
Discussion: https://postgr.es/m/CA%2B0DEqiZs7gdOd4ikmg%3D0UWG%2BSwWOLxPsk_JW-sx9WNOyrb0KQ%40mail.gmail.com
Author: Alexander Korotkov
Reviewed-by: Tom Lane, Zhihong Yu

Turns out you can specify negative values using plurals:

	https://english.stackexchange.com/questions/9735/is-1-followed-by-a-singular-or-plural-noun

so the previous code was correct enough, and consistent with other usage
in our code.  Also add comment in the two places where this could be
confused.

Reported-by: Noah Misch

Diagnosed-by: 20210425115726.GA2353095@rfd.leadboat.com

While we've always allowed such cases, the documentation didn't
say you could do it.

Discussion: https://postgr.es/m/161969805833.690.13680986983883602407@wrigleys.postgresql.org

Mention specifically that you can't call aggregates, window functions,
or procedures this way (the inability to call SRFs was already
mentioned).

Also, the claim that PQfn doesn't support NULL arguments or results
has been a lie since we invented protocol 3.0.  Not sure why this
text was never updated for that, but do it now.

Discussion: https://postgr.es/m/2039442.1615317309@sss.pgh.pa.us

Reject aggregates, window functions, and procedures.  Aggregates
failed anyway, though with a somewhat obscure error message.
Window functions would hit an Assert or null-pointer dereference.
Procedures seemed to work as long as you didn't try to do
transaction control, but (a) transaction control is sort of the
point of a procedure, and (b) it's not entirely clear that no
bugs lurk in that path.  Given the lack of testing of this area,
it seems safest to be conservative in what we support.

Also reject proretset functions, as the fastpath protocol can't
support returning a set.

Also remove an easily-triggered assertion that the given OID
isn't 0; the subsequent lookups can handle that case themselves.

Per report from Theodor-Arsenij Larionov-Trichkin.
Back-patch to all supported branches.  (The procedure angle
only applies in v11+, of course.)

Discussion: https://postgr.es/m/2039442.1615317309@sss.pgh.pa.us

There were two problems:
a. We were always selecting the next available txn instead of selecting it
when it is larger than the previous transaction.
b. We were selecting the transactions which haven't made any changes to
the database (base snapshot is not set). Later it was hitting an Assert
because we don't decode such transactions and the changes in txn remain as
it is. It is better not to choose such transactions for streaming in the
first place.

Reported-by: Haiying Tang
Author: Dilip Kumar
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/OS0PR01MB61133B94E63177040F7ECDA1FB429@OS0PR01MB6113.jpnprd01.prod.outlook.com

Here we adjust the EXPLAIN ANALYZE output for Result Cache so that we
don't show any Result Cache stats for parallel workers who don't
contribute anything to Result Cache plan nodes.

I originally had ideas that workers who don't help could still have their
Result Cache stats displayed.  The idea with that was so that I could
write some parallel Result Cache regression tests that show the EXPLAIN
ANALYZE output.  However, I realized a little too late that such tests
would just not be possible to have run in a stable way on the buildfarm.

With that knowledge, before 9eacee2e went in, I had removed all of the
tests that were showing the EXPLAIN ANALYZE output of a parallel Result
Cache plan, however, I forgot to put back the code that adjusts the
EXPLAIN output to hide the Result Cache stats for parallel workers who
were not fast enough to help out before query execution was over. All
other nodes behave this way and so should Result Cache.

Additionally, with this change, it now seems safe enough to remove the SET
force_parallel_mode = off that I had added to the regression tests.

Also, perform some cleanup in the partition_prune tests. I had adjusted
the explain_parallel_append() function to sanitize the Result Cache
EXPLAIN ANALYZE output.  However, since I didn't actually include any
parallel Result Cache tests that show their EXPLAIN ANALYZE output, that
code does nothing and can be removed.

In passing, move the setting of memPeakKb into the scope where it's used.

Reported-by: Amit Khandekar
Author: David Rowley, Amit Khandekar
Discussion: https://postgr.es/m/CAJ3gD9d8SkfY95GpM1zmsOtX2-Ogx5q-WLsf8f0ykEb0hCRK3w@mail.gmail.com

As per analysis, it appears that the 'drop slot' message from the previous
test and 'create slot' message of the new test are either missed or not
yet delivered to the stats collector due to which we will still see the
stats from the old slot. This can happen rarely which could be the reason
that we are seeing some failures in the buildfarm randomly. To avoid that
we are using a different slot name for the tests in
test_decoding/sql/stats.sql.

Reported-by: Tom Lane based on buildfarm reports
Author: Sawada Masahiko
Reviewed-by: Amit Kapila, Vignesh C
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de

Don't advocate dropping a whole table when dropping a column would
serve.  While at it, try to make the layout of these messages a
bit cleaner and more consistent.

Per suggestion from Daniel Gustafsson.  No back-patch, as we generally
don't like to churn translatable messages in released branches.

Discussion: https://postgr.es/m/2798740.1619622555@sss.pgh.pa.us