[prev in list] [next in list] [prev in thread] [next in thread]
List: postgresql-announce
Subject: PostgreSQL Weekly News - January 31, 2021
From: PWN via PostgreSQL Announce <announce-noreply () postgresql ! org>
Date: 2021-02-01 9:07:55
Message-ID: 161217047588.14653.2369631999036024003 () wrigleys ! postgresql ! org
[Download RAW message or body]
[Attachment #2 (multipart/alternative)]
# PostgreSQL Weekly News - January 31, 2021
Crunchy PostgreSQL Operator 4.6.0, a system for deploying and managing open
source PostgreSQL clusters on Kubernetes, released.
[https://access.crunchydata.com/documentation/postgres-operator/latest/releases/4.6.0/ \
](https://access.crunchydata.com/documentation/postgres-operator/latest/releases/4.6.0/)
pgAdmin4 4.30, a web- and native GUI control center for PostgreSQL, released.
[https://www.pgadmin.org/docs/pgadmin4/dev/release_notes_4_30.html](https://www.pgadmin.org/docs/pgadmin4/dev/release_notes_4_30.html)
# PostgreSQL Product News
# PostgreSQL Jobs for January
[https://archives.postgresql.org/pgsql-jobs/2021-01/](https://archives.postgresql.org/pgsql-jobs/2021-01/)
# PostgreSQL in the News
Planet PostgreSQL: [https://planet.postgresql.org/](https://planet.postgresql.org/)
PostgreSQL Weekly News is brought to you this week by David Fetter
Submit news and announcements by Sunday at 3:00pm PST8PDT to david@fetter.org.
# Applied Patches
Tom Lane pushed:
- Update time zone data files to tzdata release 2021a. DST law changes in Russia
(Volgograd zone) and South Sudan. Historical corrections for Australia,
Bahamas, Belize, Bermuda, Ghana, Israel, Kenya, Nigeria, Palestine,
Seychelles, and Vanuatu. Notably, the Australia/Currie zone has been corrected
to the point where it is identical to Australia/Hobart.
[https://git.postgresql.org/pg/commitdiff/c7edf4ac246b67073563354c2808c78868cbac36]( \
https://git.postgresql.org/pg/commitdiff/c7edf4ac246b67073563354c2808c78868cbac36)
- Make storage/standby.h compile standalone again. This file has failed
headerscheck/cpluspluscheck verification since commit 0650ff230, as a result
of referencing typedef TimestampTz without including the appropriate header.
[https://git.postgresql.org/pg/commitdiff/a4b03de589c1df0845e9732da203f505f2eedb6d]( \
https://git.postgresql.org/pg/commitdiff/a4b03de589c1df0845e9732da203f505f2eedb6d)
- Doc: improve documentation of pg_proc.protrftypes. Add a "references" link
pointing to pg_type, as we have for other arrays of type OIDs. Wordsmith the
explanation a bit. Joel Jacobson, additional editing by me Discussion:
[https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com](https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com)
[https://git.postgresql.org/pg/commitdiff/951862eda57e5dc8f78c97b3c30fe2032a5562b8]( \
https://git.postgresql.org/pg/commitdiff/951862eda57e5dc8f78c97b3c30fe2032a5562b8)
- Fix broken ruleutils support for function TRANSFORM clauses. I chanced to
notice that this dumped core due to a faulty Assert. To add insult to injury,
the output has been misformatted since v11. Obviously we need some regression
testing here. Discussion:
[https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com](https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com)
[https://git.postgresql.org/pg/commitdiff/07d46fceb4254b00e79f3d06419cbae13b0ecb5a]( \
https://git.postgresql.org/pg/commitdiff/07d46fceb4254b00e79f3d06419cbae13b0ecb5a)
- Don't clobber the calling user's credentials cache in Kerberos test.
Embarrassing oversight in this test script, which fortunately is not run by
default. Report and patch by Jacob Champion. Discussion:
[https://postgr.es/m/1fcb175bafef6560f47a8c31229fa7c938486b8d.camel@vmware.com](https://postgr.es/m/1fcb175bafef6560f47a8c31229fa7c938486b8d.camel@vmware.com)
[https://git.postgresql.org/pg/commitdiff/881933f194221abcce07fb134ebe8685e5bb58dd]( \
https://git.postgresql.org/pg/commitdiff/881933f194221abcce07fb134ebe8685e5bb58dd)
- Improve performance of repeated CALLs within plpgsql procedures. This patch
essentially is cleaning up technical debt left behind by the original
implementation of plpgsql procedures, particularly commit d92bc83c4. That
patch (or more precisely, follow-on patches fixing its worst bugs) forced us
to re-plan CALL and DO statements each time through, if we're in a non-atomic
context. That wasn't for any fundamental reason, but just because use of a
saved plan requires having a ResourceOwner to hold a reference count for the
plan, and we had no suitable resowner at hand, nor would the available APIs
support using one if we did. While it's not that expensive to create a "plan"
for CALL/DO, the cycles do add up in repeated executions. This patch
therefore makes the following API changes: `*` GetCachedPlan/ReleaseCachedPlan
are modified to let the caller specify which resowner to use to pin the plan,
rather than forcing use of CurrentResourceOwner. `*` spi.c gains a
"SPI_execute_plan_extended" entry point that lets callers say which resowner
to use to pin the plan. This borrows the idea of an options struct from the
recently added SPI_prepare_extended, hopefully allowing future options to be
added without more API breaks. This supersedes SPI_execute_plan_with_paramlist
(which I've marked deprecated) as well as SPI_execute_plan_with_receiver
(which is new in v14, so I just took it out altogether). `*` I also took the
opportunity to remove the crude hack of letting plpgsql reach into SPI private
data structures to mark SPI plans as "no_snapshot". It's better to treat that
as an option of SPI_prepare_extended. Now, when running a non-atomic
procedure or DO block that contains any CALL or DO commands, plpgsql creates a
ResourceOwner that will be used to pin the plans of the CALL/DO commands. (In
an atomic context, we just use CurrentResourceOwner, as before.) Having done
this, we can just save CALL/DO plans normally, whether or not they are used
across transaction boundaries. This seems to be good for something like 2X
speedup of a CALL of a trivial procedure with a few simple argument
expressions. By restricting the creation of an extra ResourceOwner like this,
there's essentially zero penalty in cases that can't benefit. Pavel Stehule,
with some further hacking by me Discussion:
[https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail. \
com](https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/ee895a655ce4341546facd6f23e3e8f2931b96bf]( \
https://git.postgresql.org/pg/commitdiff/ee895a655ce4341546facd6f23e3e8f2931b96bf)
- Code review for psql's helpSQL() function. The loops to identify word
boundaries could access past the end of the input string. Likely that would
never result in an actual crash, but it makes valgrind unhappy. The logic to
try different numbers of words didn't work when the input has two words but we
only have a match to the first, eg "\h with select". (We must "continue" the
pass loop, not "break".) The logic to compute nl_count was bizarrely managed,
and in at least two code paths could end up calling PageOutput with nl_count =
0, resulting in failing to paginate output that should have been fed to the
pager. Also, in v12 and up, the nl_count calculation hadn't been updated to
account for the addition of a URL. The PQExpBuffer holding the command syntax
details wasn't freed, resulting in a session-lifespan memory leak. While
here, improve some comments, choose a more descriptive name for a variable,
fix inconsistent datatype choice for another variable. Per bug #16837 from
Alexander Lakhin. This code is very old, so back-patch to all supported
branches. Kyotaro Horiguchi and Tom Lane Discussion:
[https://postgr.es/m/16837-479bcd56040c71b3@postgresql.org](https://postgr.es/m/16837-479bcd56040c71b3@postgresql.org)
[https://git.postgresql.org/pg/commitdiff/f76a85000bba2f1b1c926cbbe525e47b246215f1]( \
https://git.postgresql.org/pg/commitdiff/f76a85000bba2f1b1c926cbbe525e47b246215f1)
- Suppress compiler warnings from commit ee895a655. For obscure reasons, some
buildfarm members are now generating complaints about plpgsql_call_handler's
"retval" variable possibly being used uninitialized. It seems no less safe
than it was before that commit, but these complaints are (mostly?) new. I
trust that initializing the variable where it's declared will be enough to
shut that up. I also notice that some compilers are warning about setjmp
clobber of the same variable, which is maybe a bit more defensible. Mark it
volatile to silence that. Also, rearrange the logic to give
procedure_resowner a single point of initialization, in hopes of silencing
some setjmp-clobber warnings about that. (Marking it volatile would serve
too, but its sibling variables are depending on single assignment, so let's
stick with that method.) Discussion:
[https://postgr.es/m/E1l4F1z-0000cN-Lx@gemulon.postgresql.org](https://postgr.es/m/E1l4F1z-0000cN-Lx@gemulon.postgresql.org)
[https://git.postgresql.org/pg/commitdiff/7292fd8f1c781278021407276474d9188845113d]( \
https://git.postgresql.org/pg/commitdiff/7292fd8f1c781278021407276474d9188845113d)
- Rethink recently-added SPI interfaces. SPI_execute_with_receiver and
SPI_cursor_parse_open_with_paramlist are new in v14 (cf. commit 2f48ede08).
Before they can get out the door, let's change their APIs to follow the
practice recently established by SPI_prepare_extended etc: shove all optional
arguments into a struct that callers are supposed to pre-zero. The hope is to
allow future addition of more options without either API breakage or a
continuing proliferation of new SPI entry points. With that in mind, choose
slightly more generic names for them: SPI_execute_extended and
SPI_cursor_parse_open respectively. Discussion:
[https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail. \
com](https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/d5a83d79c9f9b660a6a5a77afafe146d3c8c6f46]( \
https://git.postgresql.org/pg/commitdiff/d5a83d79c9f9b660a6a5a77afafe146d3c8c6f46)
- Doc: improve documentation for UNNEST(). Per a user question, spell out that
UNNEST() returns array elements in storage order; also provide an example to
clarify the behavior for multi-dimensional arrays. While here, also clarify
the SELECT reference page's description of WITH ORDINALITY. These details
were already given in 7.2.1.4, but a reference page should not omit details.
Back-patch to v13; there's not room in the table in older versions.
Discussion:
[https://postgr.es/m/FF1FB31F-0507-4F18-9559-2DE6E07E3B43@gmail.com](https://postgr.es/m/FF1FB31F-0507-4F18-9559-2DE6E07E3B43@gmail.com)
[https://git.postgresql.org/pg/commitdiff/662affcfe9e816584e3d8602b3b4005236931bbb]( \
https://git.postgresql.org/pg/commitdiff/662affcfe9e816584e3d8602b3b4005236931bbb)
- Make ecpg's rjulmdy() and rmdyjul() agree with their declarations. We had
`"short *mdy"` in the extern declarations, but "short mdy[3]" in the actual
function definitions. Per C99 these are equivalent, but recent versions of
gcc have started to issue warnings about the inconsistency. Clean it up
before the warnings get any more widespread. Back-patch, in case anyone wants
to build older PG versions with bleeding-edge compilers. Discussion:
[https://postgr.es/m/2401575.1611764534@sss.pgh.pa.us](https://postgr.es/m/2401575.1611764534@sss.pgh.pa.us)
[https://git.postgresql.org/pg/commitdiff/1b242f42ba1e732b202f35265ab1a1614ce7d859]( \
https://git.postgresql.org/pg/commitdiff/1b242f42ba1e732b202f35265ab1a1614ce7d859)
- Fix hash partition pruning with asymmetric partition sets.
perform_pruning_combine_step() was not taught about the number of partition
indexes used in hash partitioning; more embarrassingly,
get_matching_hash_bounds() also had it wrong. These errors are masked in the
common case where all the partitions have the same modulus and no partition is
missing. However, with missing or unequal-size partitions, we could
erroneously prune some partitions that need to be scanned, leading to silently
wrong query answers. While a minimal-footprint fix for this could be to
export get_partition_bound_num_indexes and make the incorrect functions use
it, I'm of the opinion that that function should never have existed in the
first place. It's not reasonable data structure design that
PartitionBoundInfoData lacks any explicit record of the length of its
indexes[] array. Perhaps that was all right when it could always be assumed
equal to ndatums, but something should have been done about it as soon as that
stopped being true. Putting in an explicit "nindexes" field makes both
partition_bounds_equal() and partition_bounds_copy() simpler, safer, and
faster than before, and removes explicit knowledge of the
number-of-partition-indexes rules from some other places too. This change
also makes get_hash_partition_greatest_modulus obsolete. I left that in place
in case any external code uses it, but no core code does anymore. Per bug
#16840 from Michał Albrycht. Back-patch to v11 where the hash partitioning
code came in. (In the back branches, add the new field at the end of
PartitionBoundInfoData to minimize ABI risks.) Discussion:
[https://postgr.es/m/16840-571a22976f829ad4@postgresql.org](https://postgr.es/m/16840-571a22976f829ad4@postgresql.org)
[https://git.postgresql.org/pg/commitdiff/1d9351a87c9a9e82c7091aab03d9299982670ce0]( \
https://git.postgresql.org/pg/commitdiff/1d9351a87c9a9e82c7091aab03d9299982670ce0)
- Silence another gcc 11 warning. Per buildfarm and local experimentation,
bleeding-edge gcc isn't convinced that the MemSet in
reorder_function_arguments() is safe. Shut it up by adding an explicit check
that pronargs isn't negative, and by changing MemSet to memset. (It appears
that either change is enough to quiet the warning at -O2, but let's do both to
be sure.)
[https://git.postgresql.org/pg/commitdiff/1046dbedde2fc3fe55f007ff3255ab65ab98f858]( \
https://git.postgresql.org/pg/commitdiff/1046dbedde2fc3fe55f007ff3255ab65ab98f858)
- Doc: improve cross-references for SET/SHOW. The corresponding functions
set_config and current_setting were mostly not hyperlinked. Clarify their
descriptions a tad, too. Discussion:
[https://postgr.es/m/161183356250.4077.687338658090583892@wrigleys.postgresql.org](https://postgr.es/m/161183356250.4077.687338658090583892@wrigleys.postgresql.org)
[https://git.postgresql.org/pg/commitdiff/f743a2bbd4e511ad58b6ce01f81841e5e1611474]( \
https://git.postgresql.org/pg/commitdiff/f743a2bbd4e511ad58b6ce01f81841e5e1611474)
Amit Kapila pushed:
- Fix ALTER PUBLICATION...DROP TABLE behavior. Commit 69bd60672 fixed the
initialization of streamed transactions for RelationSyncEntry. It forgot to
initialize the publication actions while invalidating the RelationSyncEntry
due to which even though the relation is dropped from a particular publication
we still publish its changes. Fix it by initializing pubactions when entry got
invalidated. Author: Japin Li and Bharath Rupireddy Reviewed-by: Amit Kapila
Discussion:
[https://postgr.es/m/CALj2ACV+0UFpcZs5czYgBpujM9p0Hg1qdOZai_43OU7bqHU_xw@mail.gmail. \
com](https://postgr.es/m/CALj2ACV+0UFpcZs5czYgBpujM9p0Hg1qdOZai_43OU7bqHU_xw@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/40ab64c1ec1cb9bd73695f519cf66ddbb97d8144]( \
https://git.postgresql.org/pg/commitdiff/40ab64c1ec1cb9bd73695f519cf66ddbb97d8144)
David Rowley pushed:
- Fix hypothetical bug in heap backward scans. Both heapgettup() and
heapgettup_pagemode() incorrectly set the first page to scan in a backward
scan in which the number of pages to scan was specified by
heap_setscanlimits(). The code incorrectly started the scan at the end of the
relation when startBlk was 0, or otherwise at startBlk - 1, neither of which
is correct when only scanning a subset of pages. The fix here checks if
heap_setscanlimits() has changed the number of pages to scan and if so we set
the first page to scan as the final page in the specified range during
backward scans. Proper adjustment of this code was forgotten when
heap_setscanlimits() was added in 7516f5259 back in 9.5. However, practice,
nowhere in core code performs backward scans after having used
heap_setscanlimits(), yet, it is possible an extension uses the heap functions
in this way, hence backpatch. An upcoming patch does use heap_setscanlimits()
with backward scans, so this must be fixed before that can go in. Author:
David Rowley Discussion:
[https://postgr.es/m/CAApHDvpGc9h0_oVD2CtgBcxCS1N-qDYZSeBRnUh+0CWJA9cMaA@mail.gmail. \
com](https://postgr.es/m/CAApHDvpGc9h0_oVD2CtgBcxCS1N-qDYZSeBRnUh+0CWJA9cMaA@mail.gmail.com)
Backpatch-through: 9.5, all supported versions
[https://git.postgresql.org/pg/commitdiff/16dfe253e31f75b60e93acc0c2b5bbf19936c074]( \
https://git.postgresql.org/pg/commitdiff/16dfe253e31f75b60e93acc0c2b5bbf19936c074)
Peter Eisentraut pushed:
- Remove duplicate include. Reported-by: Ashutosh Sharma <ashu.coek88@gmail.com>
Discussion:
[https://www.postgresql.org/message-id/flat/CAE9k0PkORqHHGKY54-sFyDpP90yAf%2B05Auc4f \
s9EAn4J%2BuBeUQ%40mail.gmail.com](https://www.postgresql.org/message-id/flat/CAE9k0PkORqHHGKY54-sFyDpP90yAf%2B05Auc4fs9EAn4J%2BuBeUQ%40mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/ecc4b1318c92594abdf9bd2933fb6feaa9d56229]( \
https://git.postgresql.org/pg/commitdiff/ecc4b1318c92594abdf9bd2933fb6feaa9d56229)
- Remove gratuitous uses of deprecated SELECT INTO. CREATE TABLE AS has been
preferred over SELECT INTO (outside of ecpg and PL/pgSQL) for a long time.
There were still a few uses of SELECT INTO in tests and documentation, some
old, some more recent. This changes them to CREATE TABLE AS. Some
occurrences in the tests remain where they are specifically testing SELECT
INTO parsing or similar. Discussion:
[https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40e \
nterprisedb.com](https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40enterprisedb.com)
[https://git.postgresql.org/pg/commitdiff/b034ef9b376dbe712caa076541d6a750f37d85ce]( \
https://git.postgresql.org/pg/commitdiff/b034ef9b376dbe712caa076541d6a750f37d85ce)
- Fix typo.
[https://git.postgresql.org/pg/commitdiff/2592be8be5aed6448a1e2b386fffd75ad69eb6c5]( \
https://git.postgresql.org/pg/commitdiff/2592be8be5aed6448a1e2b386fffd75ad69eb6c5)
- Allow GRANTED BY clause in normal GRANT and REVOKE statements. The SQL
standard allows a GRANTED BY clause on GRANT and REVOKE (privilege) statements
that can specify CURRENT_USER or CURRENT_ROLE. In PostgreSQL, both of these
are the default behavior. Since we already have all the parsing support for
this for the GRANT (role) statement, we might as well add basic support for
this for the privilege variant as well. This allows us to check off SQL
feature T332. In the future, perhaps more interesting things could be done
with this, too. Reviewed-by: Simon Riggs <simon@2ndquadrant.com> Discussion:
[https://www.postgresql.org/message-id/flat/f2feac44-b4c5-f38f-3699-2851d6a76dc9@2nd \
quadrant.com](https://www.postgresql.org/message-id/flat/f2feac44-b4c5-f38f-3699-2851d6a76dc9@2ndquadrant.com)
[https://git.postgresql.org/pg/commitdiff/6aaaa76bb47db11cd6f567eafa3d1ee81ca59556]( \
https://git.postgresql.org/pg/commitdiff/6aaaa76bb47db11cd6f567eafa3d1ee81ca59556)
- doc: Clarify status of SELECT INTO on reference page. The documentation as
well as source code comments weren't entirely clear whether SELECT INTO was
truly deprecated (thus in theory destined to be removed eventually), or just a
less recommended variant. After discussion, it appears that other
implementations also use SELECT INTO in direct SQL in a way similar to
PostgreSQL, so it seems worth keeping it for compatibility. Update the
language in the documentation to that effect. Discussion:
[https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40e \
nterprisedb.com](https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40enterprisedb.com)
[https://git.postgresql.org/pg/commitdiff/65330622441d7ee08f768c4326825ae903f2595a]( \
https://git.postgresql.org/pg/commitdiff/65330622441d7ee08f768c4326825ae903f2595a)
- Add primary keys and unique constraints to system catalogs. For those system
catalogs that have a unique indexes, make a primary key and unique constraint,
using ALTER TABLE ... PRIMARY KEY/UNIQUE USING INDEX. This can be helpful for
GUI tools that look for a primary key, and it might in the future allow
declaring foreign keys, for making schema diagrams. The constraint creation
statements are automatically created by genbki.pl from DECLARE_UNIQUE_INDEX
directives. To specify which one of the available unique indexes is the
primary key, use the new directive DECLARE_UNIQUE_INDEX_PKEY instead. By
convention, we usually make a catalog's OID column its primary key, if it has
one. Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Discussion:
[https://www.postgresql.org/message-id/flat/dc5f44d9-5ec1-a596-0251-dadadcdede98@2nd \
quadrant.com](https://www.postgresql.org/message-id/flat/dc5f44d9-5ec1-a596-0251-dadadcdede98@2ndquadrant.com)
[https://git.postgresql.org/pg/commitdiff/dfb75e478cacb33d277432e0df5e2f9a2a2b52d9]( \
https://git.postgresql.org/pg/commitdiff/dfb75e478cacb33d277432e0df5e2f9a2a2b52d9)
Robert Haas pushed:
- Remove CheckpointLock. Up until now, we've held this lock when performing a
checkpoint or restartpoint, but commit
076a055acf3c55314de267c62b03191586d79cf6 back in 2004 and commit
7e48b77b1cebb9a43f9fdd6b17128a0ba36132f9 from 2009, taken together, have
removed all need for this. In the present code, there's only ever one process
entitled to attempt a checkpoint: either the checkpointer, during normal
operation, or the postmaster, during single-user operation. So, we don't need
the lock. One possible concern in making this change is that it means that a
substantial amount of code where HOLD_INTERRUPTS() was previously in effect
due to the preceding LWLockAcquire() will now be running without that. This
could mean that ProcessInterrupts() gets called in places from which it didn't
before. However, this seems unlikely to do very much, because the checkpointer
doesn't have any signal mapped to die(), so it's not clear how, for example,
ProcDiePending = true could happen in the first place. Similarly with
ClientConnectionLost and recovery conflicts. Also, if there are any such
problems, we might want to fix them rather than reverting this, since running
lots of code with interrupt handling suspended is generally bad. Patch by me,
per an inquiry by Amul Sul. Review by Tom Lane and Michael Paquier.
Discussion:
[http://postgr.es/m/CAAJ_b97XnBBfYeSREDJorFsyoD1sHgqnNuCi=02mNQBUMnA=FA@mail.gmail.c \
om](http://postgr.es/m/CAAJ_b97XnBBfYeSREDJorFsyoD1sHgqnNuCi=02mNQBUMnA=FA@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/d18e75664a2fda2e4d5cc433d68e37fc0e9499f2]( \
https://git.postgresql.org/pg/commitdiff/d18e75664a2fda2e4d5cc433d68e37fc0e9499f2)
- Move StartupCLOG() calls to just after we initialize ShmemVariableCache.
Previously, the hot_standby=off code path did this at end of recovery, while
the hot_standby=on code path did it at the beginning of recovery. It's better
to do this in only one place because (a) it's simpler, (b) StartupCLOG() is
trivial so trying to postpone the work isn't useful, and (c) this will make it
possible to simplify some other logic. Patch by me, reviewed by Heikki
Linnakangas. Discussion:
[http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.c \
om](http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/1f113abdf87cd085dee3927960bb4f70442b7250]( \
https://git.postgresql.org/pg/commitdiff/1f113abdf87cd085dee3927960bb4f70442b7250)
- In clog_redo(), don't set XactCtl->shared->latest_page_number. The comment is
no longer accurate, and hasn't been entirely accurate since Hot Standby was
introduced. The original idea here was that StartupCLOG() wouldn't be called
until the end of recovery and therefore this value would be uninitialized when
this code is reached, but Hot Standby made that true only when
hot_standby=off, and commit 1f113abdf87cd085dee3927960bb4f70442b7250 means
that this value is now always initialized before replay even starts. The
original purpose of this code was to bypass the sanity check in
SimpleLruTruncate(), which will no longer occur: now, if something is wrong,
that sanity check might trip during recovery. That's probably a good thing,
because in the current code base latest_page_number should always be
initialized and therefore we expect that the sanity check should pass. If it
doesn't, something has gone wrong, and complaining about it is appropriate.
Patch by me, reviewed by Heikki Linnakangas. Discussion:
[http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.c \
om](http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/0fcc2decd485a61321a3220d8f76cb108b082009]( \
https://git.postgresql.org/pg/commitdiff/0fcc2decd485a61321a3220d8f76cb108b082009)
- In TrimCLOG(), don't reset XactCtl->shared->latest_page_number. Since the CLOG
page number is not recorded directly in the checkpoint record, we have to use
ShmemVariableCache->nextXid to figure out the latest CLOG page number at the
start of recovery. However, as recovery progresses, replay of CLOG/EXTEND
records will update our notion of the latest page number, and we should rely
on that being accurate rather than recomputing the value based on an updated
notion of nextXid. ShmemVariableCache->nextXid is only an approximation during
recovery anyway, whereas CLOG/EXTEND records are an authoritative
representation of how the SLRU has been updated. Commit
0fcc2decd485a61321a3220d8f76cb108b082009 makes this simplification possible,
as before that change clog_redo() might have injected a bogus value here, and
we'd want to get rid of that before entering normal running. Patch by me,
reviewed by Heikki Linnakangas. Discussion:
[http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.c \
om](http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/69059d3b2f0754c8e661ba479f7121e6631cdf4a]( \
https://git.postgresql.org/pg/commitdiff/69059d3b2f0754c8e661ba479f7121e6631cdf4a)
Andres Freund pushed:
- Fix two typos in snapbuild.c. Reported-by: Heikki Linnakangas
<hlinnaka@iki.fi> Discussion:
[https://postgr.es/m/c94be044-818f-15e3-1ad3-7a7ae2dfed0a@iki.fi](https://postgr.es/m/c94be044-818f-15e3-1ad3-7a7ae2dfed0a@iki.fi)
[https://git.postgresql.org/pg/commitdiff/55ef8555f0c1207bac25050e7297bbb969c84233]( \
https://git.postgresql.org/pg/commitdiff/55ef8555f0c1207bac25050e7297bbb969c84233)
Fujii Masao pushed:
- postgres_fdw: Add functions to discard cached connections. This commit
introduces two new functions postgres_fdw_disconnect() and
postgres_fdw_disconnect_all(). The former function discards the cached
connections to the specified foreign server. The latter discards all the
cached connections. If the connection is used in the current transaction, it's
not closed and a warning message is emitted. For example, these functions are
useful when users want to explicitly close the foreign server connections that
are no longer necessary and then to prevent them from eating up the foreign
servers connections capacity. Author: Bharath Rupireddy, tweaked a bit by
Fujii Masao Reviewed-by: Alexey Kondratov, Zhijie Hou, Zhihong Yu, Fujii Masao
Discussion:
[https://postgr.es/m/CALj2ACVvrp5=AVp2PupEm+nAC8S4buqR3fJMmaCoc7ftT0aD2A@mail.gmail. \
com](https://postgr.es/m/CALj2ACVvrp5=AVp2PupEm+nAC8S4buqR3fJMmaCoc7ftT0aD2A@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/411ae64997dc3a42d19eda6721c581841ce2cb82]( \
https://git.postgresql.org/pg/commitdiff/411ae64997dc3a42d19eda6721c581841ce2cb82)
- postgres_fdw: Stabilize regression test for postgres_fdw_disconnect_all(). The
regression test added in commit 411ae64997 caused buildfarm failures. The
cause of them was that the order of warning messages output in the test was
not stable. To fix this, this commit sets client_min_messages to ERROR
temporarily when performing the test generating those warnings. Per buildfarm
failures. Discussion:
[https://postgr.es/m/2147113.1611644754@sss.pgh.pa.us](https://postgr.es/m/2147113.1611644754@sss.pgh.pa.us)
[https://git.postgresql.org/pg/commitdiff/6adc5376dca4ef8b7d591c0ee7338cb9ff660216]( \
https://git.postgresql.org/pg/commitdiff/6adc5376dca4ef8b7d591c0ee7338cb9ff660216)
- postgres_fdw: Fix test failure with
-DENFORCE_REGRESSION_TEST_NAME_RESTRICTIONS. The roles created by regression
test should have names starting with "regress_", and the test introduced in
commit 411ae64997 did not do that. Per buildfarm member longfin. Discussion:
[https://postgr.es/m/73fc5ae4-3c54-1262-4533-f8c547de2e60@oss.nttdata.com](https://postgr.es/m/73fc5ae4-3c54-1262-4533-f8c547de2e60@oss.nttdata.com)
[https://git.postgresql.org/pg/commitdiff/0c3fc09fe359a6dc46f1870ceccf60ec60396bc9]( \
https://git.postgresql.org/pg/commitdiff/0c3fc09fe359a6dc46f1870ceccf60ec60396bc9)
- postgres_fdw: Fix tests for CLOBBER_CACHE_ALWAYS. The regression tests added
in commits 708d165ddb and 411ae64997 caused buildfarm failures when
CLOBBER_CACHE_ALWAYS was enabled. This commit stabilizes those tests. The
foreign server connections established by postgres_fdw behaves differently
depending on whether CLOBBER_CACHE_ALWAYS is enabled or not. If it's not
enabled, those connections are cached. On the other hand, if it's enabled,
when the connections are established outside transaction block, they are not
cached (i.e., they are immediately closed at the end of query that established
them). So the subsequent postgres_fdw_get_connections() cannot list those
connections and postgres_fdw_disconnect() cannot close them (because they are
already closed). When the connections are established inside transaction
block, they are cached whether CLOBBER_CACHE_ALWAYS was enabled or not. But if
it's enabled, they are immediately marked as invalid, otherwise not. This
causes the subsequent postgres_fdw_get_connections() to return different
result in "valid" column depending on whether CLOBBER_CACHE_ALWAYS was enabled
or not. This commit prevents the above differences of behavior from affecting
the regression tests. Per buildfarm failure on trilobite. Original patch by
Bharath Rupireddy. I (Fujii Masao) extracted the regression test fix from that
and revised it a bit. Reported-by: Tom Lane Author: Bharath Rupireddy
Reviewed-by: Fujii Masao Discussion:
[https://postgr.es/m/2688508.1611865371@sss.pgh.pa.us](https://postgr.es/m/2688508.1611865371@sss.pgh.pa.us)
[https://git.postgresql.org/pg/commitdiff/f77717b2985aa529a185e6988de26b885ca10ddb]( \
https://git.postgresql.org/pg/commitdiff/f77717b2985aa529a185e6988de26b885ca10ddb)
Michaël Paquier pushed:
- Fix memory leak when deallocating prepared statement in postgres_fdw. The leak
is minor, so no backpatch is done. Oversight in 21734d2. Reported-by: Tom
Lane
[https://git.postgresql.org/pg/commitdiff/7b4c660466dd8a1d25ca316ac02099a7cf0e5896]( \
https://git.postgresql.org/pg/commitdiff/7b4c660466dd8a1d25ca316ac02099a7cf0e5896)
- Refactor code in tablecmds.c to check and process tablespace moves. Two code
paths of tablecmds.c (for relations with storage and without storage) use the
same logic to check if the move of a relation to a new tablespace is allowed
or not and to update pg_class.reltablespace and pg_class.relfilenode. A
potential TABLESPACE clause for REINDEX, CLUSTER and VACUUM FULL needs similar
checks to make sure that nothing is moved around in illegal ways (no mapped
relations, shared relations only in pg_global, no move of temp tables owned by
other backends). This reorganizes the existing code of ALTER TABLE so as all
this logic is controlled by two new routines that can be reused for the other
commands able to move relations across tablespaces, limiting the number of
code paths in need of the same protections. This also removes some code that
was duplicated for tables with and without storage for ALTER TABLE. Author:
Alexey Kondratov, Michael Paquier Discussion:
[https://postgr.es/m/YA+9mAMWYLXJMVPL@paquier.xyz](https://postgr.es/m/YA+9mAMWYLXJMVPL@paquier.xyz)
[https://git.postgresql.org/pg/commitdiff/4c9c359d38ff1e2de388eedd860785be6a49201c]( \
https://git.postgresql.org/pg/commitdiff/4c9c359d38ff1e2de388eedd860785be6a49201c)
- doc: Remove reference to views for TRUNCATE privilege. The page about
privilege rights mentioned that TRUNCATE could be applied to views or even
other relation types. This is confusing as this command can be used only on
tables and on partitioned tables. Oversight in afc4a78. Reported-by: Harisai
Hari Reviewed-by: Laurenz Albe Discussion:
[https://postgr.es/m/161157636877.14625.15340884663716426087@wrigleys.postgresql.org \
](https://postgr.es/m/161157636877.14625.15340884663716426087@wrigleys.postgresql.org)
Backpatch-through: 12
[https://git.postgresql.org/pg/commitdiff/32bef758296142ce0fff944c3478d63fdefdaf33]( \
https://git.postgresql.org/pg/commitdiff/32bef758296142ce0fff944c3478d63fdefdaf33)
- Refactor SQL functions of SHA-2 in cryptohashfuncs.c. The same code pattern
was repeated four times when compiling a SHA-2 hash. This refactoring has the
advantage to issue a compilation warning if a new value is added to
pg_cryptohash_type, so as anybody doing an addition in this area would need to
consider if support for a new SQL function is needed or not. Author: Sehrope
Sarkuni, Michael Paquier Discussion:
[https://postgr.es/m/YA7DvLRn2xnTgsMc@paquier.xyz](https://postgr.es/m/YA7DvLRn2xnTgsMc@paquier.xyz)
[https://git.postgresql.org/pg/commitdiff/f854c69a5b36ba7aa85bee9e9590c3e517970156]( \
https://git.postgresql.org/pg/commitdiff/f854c69a5b36ba7aa85bee9e9590c3e517970156)
- Fix crash of `pg_stat_statements_info()` without library loaded. Other code
paths are already protected against this case, and _PG_init() warns about that
in pg_stat_statements.c. While on it, I have checked the other extensions of
the tree but did not notice any holes. Oversight in 9fbc3f3. Author: Jaime
Casanova Reviewed-by: Julien Rouhaud Discussion:
[`https://postgr.es/m/CAJKUy5gF4=_=qhJ1VX_tSGFfjKHb9BvzhRYWSApJD=Bfwp2SBw@mail.gmail \
.com`](https://postgr.es/m/CAJKUy5gF4=_=qhJ1VX_tSGFfjKHb9BvzhRYWSApJD=Bfwp2SBw@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/bca96ddab57c7c64ce442b0f9f35558741c7f8bd]( \
https://git.postgresql.org/pg/commitdiff/bca96ddab57c7c64ce442b0f9f35558741c7f8bd)
- Adjust comments of CheckRelationTableSpaceMove() and SetRelationTableSpace().
4c9c359, that introduced those two functions, has been overoptimistic on the
point that only ShareUpdateExclusiveLock would be required when moving a
relation to a new tablespace. AccessExclusiveLock is a requirement, but
ShareUpdateExclusiveLock may be used under specific conditions like REINDEX
CONCURRENTLY where waits on past transactions make the operation safe even
with a lower-level lock. The current code does only the former, so update the
existing comments to reflect that. Once a REINDEX (TABLESPACE) is introduced,
those comments would require an extra refresh to mention their new use case.
While on it, fix an incorrect variable name. Per discussion with Álvaro
Herrera. Discussion:
[https://postgr.es/m/20210127140741.GA14174@alvherre.pgsql](https://postgr.es/m/20210127140741.GA14174@alvherre.pgsql)
[https://git.postgresql.org/pg/commitdiff/24843297a96d7be16cc3f4b090aacfc6e5e6839e]( \
https://git.postgresql.org/pg/commitdiff/24843297a96d7be16cc3f4b090aacfc6e5e6839e)
- doc: Improve wording of section for repslot statistics. This documentation has
been added in 9868167, so no backpatch is needed. Author: Justin Pryzby,
Michael Paquier Discussion:
[https://postgr.es/m/20201222041153.GK30237@telsasoft.com](https://postgr.es/m/20201222041153.GK30237@telsasoft.com)
[https://git.postgresql.org/pg/commitdiff/2a5862f013647dd185de3e1b496fd4f614d12f0e]( \
https://git.postgresql.org/pg/commitdiff/2a5862f013647dd185de3e1b496fd4f614d12f0e)
Peter Geoghegan pushed:
- Fix GiST index deletion assert issue. Avoid calling heap_index_delete_tuples()
with an empty deltids array to avoid an assertion failure. This issue was
arguably an oversight in commit b5f58cf2, though the failing assert itself was
added by my recent commit d168b666. No backpatch, though, since the oversight
is harmless in the back branches. Author: Peter Geoghegan <pg@bowt.ie>
Reported-By: Jaime Casanova <jcasanov@systemguards.com.ec> Discussion:
[https://postgr.es/m/CAJKUy5jscES84n3puE=sYngyF+zpb4wv8UMtuLnLPv5z=6yyNw@mail.gmail. \
com](https://postgr.es/m/CAJKUy5jscES84n3puE=sYngyF+zpb4wv8UMtuLnLPv5z=6yyNw@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/e42b3c3bd6a9c6233ac4c8a2e9b040367ba2f97c]( \
https://git.postgresql.org/pg/commitdiff/e42b3c3bd6a9c6233ac4c8a2e9b040367ba2f97c)
- Reduce the default value of vacuum_cost_page_miss. When commit f425b605
introduced cost based vacuum delays back in 2004, the defaults reflected
then-current trends in hardware, as well as certain historical limitations in
PostgreSQL. There have been enormous improvements in both areas since that
time. The cost limit GUC defaults finally became much more representative of
current trends following commit cbccac37, which decreased
autovacuum_vacuum_cost_delay's default by 10x for PostgreSQL 12 (it went from
20ms to only 2ms). The relative costs have shifted too. This should also be
accounted for by the defaults. More specifically, the relative importance of
avoiding dirtying pages within VACUUM has greatly increased, primarily due to
main memory capacity scaling and trends in flash storage. Within Postgres
itself, improvements like sequential access during index vacuuming (at least
in nbtree and GiST indexes) have also been contributing factors. To reflect
all this, decrease the default of vacuum_cost_page_miss to 2. Since the
default of vacuum_cost_page_dirty remains 20, dirtying a page is now
considered 10x "costlier" than a page miss by default. Author: Peter
Geoghegan <pg@bowt.ie> Discussion:
[https://postgr.es/m/CAH2-WzmLPFnkWT8xMjmcsm7YS3+_Qi3iRWAb2+_Bc8UhVyHfuA@mail.gmail. \
com](https://postgr.es/m/CAH2-WzmLPFnkWT8xMjmcsm7YS3+_Qi3iRWAb2+_Bc8UhVyHfuA@mail.gmail.com)
[https://git.postgresql.org/pg/commitdiff/e19594c5c059d2e071b67d87ae84f569a52d2e32]( \
https://git.postgresql.org/pg/commitdiff/e19594c5c059d2e071b67d87ae84f569a52d2e32)
- Remove unused _bt_delitems_delete() argument. The latestRemovedXid values used
by nbtree deletion operations are determined by _bt_delitems_delete()'s
caller, so there is no reason to pass a separate heapRel argument. Oversight
in commit d168b666823.
[https://git.postgresql.org/pg/commitdiff/dc43492e46c7145a476cb8ca6200fc8eefe673ef]( \
https://git.postgresql.org/pg/commitdiff/dc43492e46c7145a476cb8ca6200fc8eefe673ef)
Andrew Gierth pushed:
- Don't add bailout adjustment for non-strict deserialize calls. When building
aggregate expression steps, strict checks need a bailout jump for when a null
value is encountered, so there is a list of steps that require later
adjustment. Adding entries to that list for steps that aren't actually strict
would be harmless, except that there is an Assert which catches them. This
leads to spurious errors on asserts builds, for data sets that trigger
parallel aggregation of an aggregate with a non-strict deserialization
function (no such aggregates exist in the core system). Repair by not adding
the adjustment entry when it's not needed. Backpatch back to 11 where the
code was introduced. Per a report from Darafei (Komzpa) of the PostGIS
project; analysis and patch by me. Discussion:
[https://postgr.es/m/87mty7peb3.fsf@news-spur.riddles.org.uk](https://postgr.es/m/87mty7peb3.fsf@news-spur.riddles.org.uk)
[https://git.postgresql.org/pg/commitdiff/a3367aa3c4552170004c92369681730d85e384c6]( \
https://git.postgresql.org/pg/commitdiff/a3367aa3c4552170004c92369681730d85e384c6)
Heikki Linnakangas pushed:
- Add mbverifystr() functions specific to each encoding. This makes
pg_verify_mbstr() function faster, by allowing more efficient
encoding-specific implementations. All the implementations included in this
commit are pretty naive, they just call the same encoding-specific verifychar
functions that were used previously, but that already gives a performance
boost because the tight character-at-a-time loop is simpler. Reviewed-by:
John Naylor Discussion:
[https://www.postgresql.org/message-id/e7861509-3960-538a-9025-b75a61188e01@iki.fi]( \
https://www.postgresql.org/message-id/e7861509-3960-538a-9025-b75a61188e01@iki.fi) \
[https://git.postgresql.org/pg/commitdiff/b80e10638e36b9d2f0b39170c613837af2ca2aac](https://git.postgresql.org/pg/commitdiff/b80e10638e36b9d2f0b39170c613837af2ca2aac)
- Add direct conversion routines between EUC_TW and Big5. Conversions between
EUC_TW and Big5 were previously implemented by converting the whole input to
MIC first, and then from MIC to the target encoding. Implement functions to
convert directly between the two. The reason to do this now is that I'm
working on a patch that will change the conversion function signature so that
if the input is invalid, we convert as much as we can and return the number of
bytes successfully converted. That's not possible if we use an intermediary
format, because if an error happens in the intermediary -> final conversion,
we lose track of the location of the invalid character in the original input.
Avoiding the intermediate step makes the conversions faster, too.
Reviewed-by: John Naylor Discussion:
[https://www.postgresql.org/message-id/b9e3167f-f84b-7aa4-5738-be578a4db924%40iki.fi \
](https://www.postgresql.org/message-id/b9e3167f-f84b-7aa4-5738-be578a4db924%40iki.fi)
[https://git.postgresql.org/pg/commitdiff/6c5576075b0f93f2235ac8a82290fe3b6e82300d]( \
https://git.postgresql.org/pg/commitdiff/6c5576075b0f93f2235ac8a82290fe3b6e82300d)
Álvaro Herrera pushed:
- pgbench: Remove dead code. doConnect() never returns connections in state
CONNECTION_BAD, so checking for that is pointless. Remove the code that does.
This code has been dead since ba708ea3dc84, 20 years ago. Discussion:
[https://postgr.es/m/20210126195224.GA20361@alvherre.pgsql](https://postgr.es/m/20210126195224.GA20361@alvherre.pgsql)
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
[https://git.postgresql.org/pg/commitdiff/6819b9042fe69154ff3fd2337c5028038d3cfab4]( \
https://git.postgresql.org/pg/commitdiff/6819b9042fe69154ff3fd2337c5028038d3cfab4)
- Remove bogus restriction from BEFORE UPDATE triggers. In trying to protect the
user from inconsistent behavior, commit 487e9861d0cf "Enable BEFORE row-level
triggers for partitioned tables" tried to prevent BEFORE UPDATE FOR EACH ROW
triggers from moving the row from one partition to another. However, it turns
out that the restriction is wrong in two ways: first, it fails spuriously,
preventing valid situations from working, as in bug #16794; and second, they
don't protect from any misbehavior, because tuple routing would cope anyway.
Fix by removing that restriction. We keep the same restriction on BEFORE
INSERT FOR EACH ROW triggers, though. It is valid and useful there. In the
future we could remove it by having tuple reroute work for inserts as it does
for updates. Backpatch to 13. Author: Álvaro Herrera
<alvherre@alvh.no-ip.org> Reported-by: Phillip Menke <pg@pmenke.de>
Discussion:
[https://postgr.es/m/16794-350a655580fbb9ae@postgresql.org](https://postgr.es/m/16794-350a655580fbb9ae@postgresql.org)
[https://git.postgresql.org/pg/commitdiff/6f5c8a8ec23f8ab00da4d2b77bfc8af2a578c4d3]( \
https://git.postgresql.org/pg/commitdiff/6f5c8a8ec23f8ab00da4d2b77bfc8af2a578c4d3)
Thomas Munro pushed:
- Retire pg_standby. pg_standby was useful more than a decade ago, but now it is
obsolete. It has been proposed that we retire it many times. Now seems like a
good time to finally do it, because "waiting restore commands" are
incompatible with a proposed recovery prefetching feature. Discussion:
[https://postgr.es/m/20201029024412.GP5380%40telsasoft.com](https://postgr.es/m/20201029024412.GP5380%40telsasoft.com)
Author: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Heikki Linnakangas
<hlinnaka@iki.fi> Reviewed-by: Peter Eisentraut
<peter.eisentraut@enterprisedb.com> Reviewed-by: Michael Paquier
<michael@paquier.xyz> Reviewed-by: Fujii Masao <masao.fujii@oss.nttdata.com>
[https://git.postgresql.org/pg/commitdiff/514b411a2b5226167add9ab139d3a96dbe98035d]( \
https://git.postgresql.org/pg/commitdiff/514b411a2b5226167add9ab139d3a96dbe98035d)
- Remove documentation of waiting restore_command. Following the removal of
pg_standby, also remove the documentation section that describes how to write
your own "waiting restore_command" along the same lines. Discussion:
[https://postgr.es/m/20201029024412.GP5380%40telsasoft.com](https://postgr.es/m/20201029024412.GP5380%40telsasoft.com)
Reviewed-by: Michael Paquier <michael@paquier.xyz>
[https://git.postgresql.org/pg/commitdiff/5c6d184213bb9feac0800eee1a16769bec9e8d58]( \
https://git.postgresql.org/pg/commitdiff/5c6d184213bb9feac0800eee1a16769bec9e8d58)
Alexander Korotkov pushed:
- Document behavior of the `.**` jsonpath accessor in the lax mode. When the `.**`
jsonpath accessor handles the array, it selects both array and each of its
elements. When using lax mode, subsequent accessors automatically unwrap
arrays. So, the content of each array element may be selected twice. Even
though this behavior is counterintuitive, it's correct because everything
works as designed. This commit documents it. Backpatch to 12 where the
jsonpath language was introduced. Reported-by: Thomas Kellerer Bug: #16828
Discussion:
[https://postgr.es/m/16828-2b0229babfad2d8c%40postgresql.org](https://postgr.es/m/16828-2b0229babfad2d8c%40postgresql.org)
Discussion:
[https://postgr.es/m/CAPpHfdtS-nNidT%3DEqZbAYOPcnNOWh_sd6skVdu2CAQUGdvpT8Q%40mail.gm \
ail.com](https://postgr.es/m/CAPpHfdtS-nNidT%3DEqZbAYOPcnNOWh_sd6skVdu2CAQUGdvpT8Q%40mail.gmail.com)
Author: Alexandex Korotkov, revised by Tom Lane Reviewed-by: Alvaro Herrera,
Thomas Kellerer, Tom Lane Backpatch-through: 12
[https://git.postgresql.org/pg/commitdiff/b41645460af563cfd4e4f57f354058cf69ef3b14]( \
https://git.postgresql.org/pg/commitdiff/b41645460af563cfd4e4f57f354058cf69ef3b14)
- Fix parsing of complex morphs to tsquery. When to_tsquery() or
websearch_to_tsquery() meet a complex morph containing multiple words residing
adjacent position, these words are connected with OP_AND operator. That leads
to surprising results. For instace, both websearch_to_tsquery('"pg_class
pg"') and to_tsquery('pg_class <-> pg') produce '( pg & class ) <-> pg'
tsquery. This tsquery requires 'pg' and 'class' words to reside on the same
position and doesn't match to to_tsvector('pg_class pg'). It appears to be
ridiculous behavior, which needs to be fixed. This commit makes to_tsquery()
or websearch_to_tsquery() connect words residing adjacent position with
OP_PHRASE. Therefore, now those words are normally chained with other
OP_PHRASE operator. The examples of above now produces 'pg <-> class <-> pg'
tsquery, which matches to to_tsvector('pg_class pg'). Another effect of this
commit is that complex morph word positions now need to match the tsvector
even if there is no surrounding OP_PHRASE. This behavior change generally
looks like an improvement but making this commit not backpatchable.
Reported-by: Barry Pederson Bug: #16592 Discussion:
[https://postgr.es/m/16592-70b110ff9731c07d@postgresql.org](https://postgr.es/m/16592-70b110ff9731c07d@postgresql.org)
Discussion:
[https://postgr.es/m/CAPpHfdv0EzVhf6CWfB1_TTZqXV_2Sn-jSY3zSd7ePH%3D-%2B1V2DQ%40mail. \
gmail.com](https://postgr.es/m/CAPpHfdv0EzVhf6CWfB1_TTZqXV_2Sn-jSY3zSd7ePH%3D-%2B1V2DQ%40mail.gmail.com)
Author: Alexander Korotkov Reviewed-by: Tom Lane, Neil Chen
[https://git.postgresql.org/pg/commitdiff/0c4f355c6a5fd437f71349f2f3d5d491382572b7]( \
https://git.postgresql.org/pg/commitdiff/0c4f355c6a5fd437f71349f2f3d5d491382572b7)
Noah Misch pushed:
- Fix CREATE INDEX CONCURRENTLY for simultaneous prepared transactions. In a
cluster having used CREATE INDEX CONCURRENTLY while having enabled prepared
transactions, queries that use the resulting index can silently fail to find
rows. Fix this for future CREATE INDEX CONCURRENTLY by making it wait for
prepared transactions like it waits for ordinary transactions. This expands
the VirtualTransactionId structure domain to admit prepared transactions. It
may be necessary to reindex to recover from past occurrences. Back-patch to
9.5 (all supported versions). Andrey Borodin, reviewed (in earlier versions)
by Tom Lane and Michael Paquier. Discussion:
[https://postgr.es/m/2E712143-97F7-4890-B470-4A35142ABC82@yandex-team.ru](https://postgr.es/m/2E712143-97F7-4890-B470-4A35142ABC82@yandex-team.ru)
[https://git.postgresql.org/pg/commitdiff/8a54e12a38d1545d249f1402f66c8cde2837d97c]( \
https://git.postgresql.org/pg/commitdiff/8a54e12a38d1545d249f1402f66c8cde2837d97c)
- Fix error with CREATE PUBLICATION, wal_level=minimal, and new tables. CREATE
PUBLICATION has failed spuriously when applied to a permanent relation created
or rewritten in the current transaction. Make the same change to another site
having the same semantic intent; the second instance has no user-visible
consequences. Back-patch to v13, where commit
c6b92041d38512a4176ed76ad06f713d2e6c01a8 broke this. Kyotaro Horiguchi
Discussion:
[https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com](https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com)
[https://git.postgresql.org/pg/commitdiff/360bd2321b1ad9c47466bc485ee5eb2f4515372e]( \
https://git.postgresql.org/pg/commitdiff/360bd2321b1ad9c47466bc485ee5eb2f4515372e)
- Revive "snapshot too old" with wal_level=minimal and SET TABLESPACE. Given a
permanent relation rewritten in the current transaction, the
old_snapshot_threshold mechanism assumed the relation had never been subject
to early pruning. Hence, a query could fail to report "snapshot too old" when
the rewrite followed an early truncation. ALTER TABLE SET TABLESPACE is
probably the only rewrite mechanism capable of exposing this bug. REINDEX
sets indcheckxmin, avoiding the problem. CLUSTER has zeroed page LSNs since
before old_snapshot_threshold existed, so old_snapshot_threshold has never
cooperated with it. ALTER TABLE ... SET DATA TYPE makes the table look empty
to every past snapshot, which is strictly worse. Back-patch to v13, where
commit c6b92041d38512a4176ed76ad06f713d2e6c01a8 broke this. Kyotaro Horiguchi
and Noah Misch Discussion:
[https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com](https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com)
[https://git.postgresql.org/pg/commitdiff/7da83415e5bc01bba0093f34f4f612b70c70b678]( \
https://git.postgresql.org/pg/commitdiff/7da83415e5bc01bba0093f34f4f612b70c70b678)
# Pending Patches
Amit Langote sent in another revision of a patch to set
ForeignScanState.resultRelInfo lazily, set ResultRelInfo.ri_PartitionRoot in all
result relations, and initialize result relation information lazily.
Takamichi Osumi sent in another revision of a patch to ensure that archive
recovery not miss data.
Amit Langote sent in another revision of a patch to create foreign key triggers
in partitioned tables, and enforce foreign keys correctly during cross-partition
updates.
Pavel Borisov sent in another revision of a patch to generate HASH and LIST
partitions automatically with a new CONFIGURATION directive.
Bucoo sent in another revision of a patch to support parallel
union/distinct/aggregate using batch hashagg and sort.
Vigneshwaran C sent in another revision of a patch to identify missing
publications from publisher during CREATE/ALTER SUBSCRIPTION operations.
Bertrand Drouvot sent in a patch to add an --extension flag to pg_dump to dump
specific extensions.
David Cramer sent in two more revisions of a patch to throw an error and
rollback on a failed transaction instead of silently rolling back.
Konstantin Knizhnik sent in two more revisions of a patch to have auto_explain
create statistics.
Michaël Paquier and Sehrope Sarkuni traded patches to surface sha1 to SQL.
Takayuki Tsunakawa, Iwata Aya, and Kirk Jamison traded patches to add tracing to
libpq.
Masahiro Ikeda sent in three more revisions of a patch to add statistics related
to write/sync wal records, and make the wal receiver report WAL statistics.
Masahiko Sawada sent in two more revisions of a patch to choose vacuum strategy
before heap and index vacuums, skip btree bulkdelete if the index doesn't grow,
and disable index cleanup for anti-wraparound and aggressive vacuums.
Peter Smith and Amit Kapila traded patches to enable tablesync workers to use
multiple transactions apiece.
Bruce Momjian sent in four more revisions of a patch to implement key
management.
Dilip Kumar sent in three more revisions of a patch to provide a new interface
to get the recovery pause status.
David Rowley sent in another revision of a patch to add TID Range Scans to
support efficient scanning ranges of TIDs.
Bertrand Drouvot sent in another revision of a patch to implement minimal
logical decoding on standbys.
Surafel Temesgen sent in another revision of a patch to implement
system-versioned temporal tables.
Jim Finnerty sent in another revision of a patch to implement 64-bit XIDs.
Amit Langote sent in another revision of a patch to overhaul how updates compute
new tuples, and revise how inherited updates and deletes are handled.
Li Japin sent in a patch to fix the documentation of ALTER SUBSCRIPTION ... SET
PUBLICATION.
Daniel Gustafsson sent in another revision of a patch to support checksum
enable/disable in a running cluster.
Takashi Menjo sent in another revision of a patch to make it possible to use
NVRAM for WAL buffers.
Masahiko Sawada sent in another revision of a patch to implement transactions
involving multiple postgres foreign servers.
Dean Rasheed sent in another revision of a patch to make it possible to create
and use extended statistics on expressions.
Amit Langote sent in another revision of a patch to export
get_partition_for_tuple() and avoid using SPI for some RI checks.
Denis Laxalde sent in a patch to Set default transactions to read-only at
server start in pg_upgrade.
Amit Khandekar sent in another revision of a patch to speed up xor'ing of two
gist index signatures for tsvectors by using popcount64() on chunks instead of
xor'ing char values, and avoid function pointer dereferencing for
pg_popcount32/64() calls.
Konstantin Knizhnik sent in another revision of a patch to extending the
auto_explain extension to generate extended statistics in case of bad
selectivity estimation, and take into account extended statistics when computing
join selectivity.
Michail Nikolaev sent in another revision of a patch to add full support for
index LP_DEAD hint bits on standby.
Nathan Bossart sent in another revision of a patch to Avoid marking segments as
ready-for-archival too early.
Anastasia Lubennikova sent in another revision of a patch intended to fix a
bug that manifested as pg_upgrade fails with non-standard ACL.
Paul Guo sent in a patch to speed up pg_rewind by fsync()ing only the affected
files/directories, and using copy_file_range() for file copying.
Kyotaro HORIGUCHI and Heikki Linnakangas traded patches to add catcache
expiration.
Heikki Linnakangas and Daniel Gustafsson traded patches to support
enabling/disabling page checksums in a running instance.
Mark Rofail sent in two more revisions of a patch to implement foreign key
arrays.
Vigneshwaran C sent in three more revisions of a patch to make it possible to
print a backtrace of any postgres process using a new function
pg_print_backtrace() that is part of the instance in which the function is
called.
Michaël Paquier and Daniel Gustafsson traded patches to make it possible to use
NSS as a TLS provider for libpq.
Bharath Rupireddy sent in two more revisions of a patch to avoid catalogue
accesses in slot_store_error_callback and conversion_error_callback.
Takamichi Osumi sent in a patch to log when/where wal_level is changed to
'minimal' from upper level to invalidate the old backups or make alerts to
users.
Amul Sul sent in another revision of a patch to implement wal prohibit state
using global barrier.
Greg Nancarrow sent in another revision of a patch to make it possible to use
parallel selects in INSERT ... SELECT, and parallel INSERT or SELECT where it's
safe to do so.
Justin Pryzby sent in another revision of a patch to make it possible to run
CREATE INDEX CONCURRENTLY on a partitioned table.
Andrew Dunstan sent in two revisions of a patch to implement PostgresNodePath, a
subclass of PostgresNode which runs commands in the context of a given
PostgreSQL install path.
David Rowley sent in another revision of a patch to add a Result Cache executor
node.
Nathan Bossart sent in two more revisions of a patch to add a PROCESS_TOAST
option to VACUUM.
Paul Martinez sent in a patch to clarify the messages that occur when a
replication connection is not allowed.
Jacob Champion sent in a patch to save a role's original authenticated identity
for logging.
Mark Dilger sent in three more revisions of a patch to implement a new
pg_amcheck contrib extension.
Hou Zhijie sent in another revision of a patch to add a new GUC and
corresponding table option enable_parallel_dml.
Marco Montagna sent in a patch to reduce likelihood of fdw prepared statement
collisions by prepending a random number to the names generated for same.
Peter Eisentraut sent in another revision of a patch to pg_dump to fix dumping
of inherited generated columns.
Antonin Houska sent in another revision of a patch to clean up orphaned files
using undo logs.
Dilip Kumar and Justin Pryzby traded patches to implement custom table
compression methods.
Álvaro Herrera sent in another revision of a patch to add
pg_atomic_monotonic_advance_u64, use atomics in xlog.c, and add barriers.
Fabien COELHO sent in another revision of a patch to add barriers to pgbench.
Thomas Munro sent in two revisions of a patch to use a global barrier to fix
DROP TABLESPACE on Windows, and use condition variables for ProcSignalBarriers.
Andrew Dunstan sent in another revision of a patch to allow matching whole DN
from a client certificate.
Thomas Munro sent in a patch to hold interrupts while running dsm_detach()
callbacks.
Joel Jacobson sent in a patch to improve the documentation of oid columns that
can be zero in system views.
Vigneshwaran C sent in another revision of a patch to add schema support for
CREATE PUBLICATION.
Tom Lane sent in a patch to record foreign key relationships for the system
catalogs.
Euler Taveira de Oliveira sent in another revision of a patch to implement row
filtering for logical replication.
[Attachment #5 (text/html)]
<!doctype html>
<html>
<head>
<meta name="viewport" content="width=device-width">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>PostgreSQL Weekly News - January 31, 2021</title>
<style>
@media only screen and (max-width: 620px) {
table[class=body] h1 {
font-size: 28px !important;
margin-bottom: 10px !important;
}
table[class=body] p,
table[class=body] ul,
table[class=body] ol,
table[class=body] td,
table[class=body] span,
table[class=body] a {
font-size: 16px !important;
}
table[class=body] .wrapper,
table[class=body] .article {
padding: 10px !important;
}
table[class=body] .content {
padding: 0 !important;
}
table[class=body] .container {
padding: 0 !important;
width: 100% !important;
}
table[class=body] .main {
border-left-width: 0 !important;
border-radius: 0 !important;
border-right-width: 0 !important;
}
table[class=body] .btn table {
width: 100% !important;
}
table[class=body] .btn a {
width: 100% !important;
}
table[class=body] .img-responsive {
height: auto !important;
max-width: 100% !important;
width: auto !important;
}
}
@media all {
.ExternalClass {
width: 100%;
}
.ExternalClass,
.ExternalClass p,
.ExternalClass span,
.ExternalClass font,
.ExternalClass td,
.ExternalClass div {
line-height: 100%;
}
.apple-link a {
color: inherit !important;
font-family: inherit !important;
font-size: inherit !important;
font-weight: inherit !important;
line-height: inherit !important;
text-decoration: none !important;
}
#MessageViewBody a {
color: inherit;
text-decoration: none;
font-size: inherit;
font-family: inherit;
font-weight: inherit;
line-height: inherit;
}
.btn-primary table td:hover {
background-color: #34495e !important;
}
.btn-primary a:hover {
background-color: #34495e !important;
border-color: #34495e !important;
}
}
</style>
</head>
<body class="" style="background-color: #f6f6f6; font-family: sans-serif; \
-webkit-font-smoothing: antialiased; font-size: 14px; line-height: 1.4; margin: 0; \
padding: 0; -ms-text-size-adjust: 100%; -webkit-text-size-adjust: 100%;"> <table \
border="0" cellpadding="0" cellspacing="0" class="body" style="border-collapse: \
separate; mso-table-lspace: 0pt; mso-table-rspace: 0pt; width: 100%; \
background-color: #f6f6f6;"> <tr>
<td style="font-family: sans-serif; font-size: 14px; vertical-align: \
top;"> </td> <td class="container" style="font-family: sans-serif; font-size: \
14px; vertical-align: top; display: block; Margin: 0 auto; max-width: 580px; padding: \
10px; width: 580px;">
<div class="content" style="box-sizing: border-box; display: block; Margin: \
0 auto; max-width: 580px; padding: 10px;">
<span class="preheader" style="color: transparent; display: none; height: \
0; max-height: 0; max-width: 0; opacity: 0; overflow: hidden; mso-hide: all; \
visibility: hidden; width: 0;"></span> <table class="main" style="border-collapse: \
separate; mso-table-lspace: 0pt; mso-table-rspace: 0pt; width: 100%; background: \
#ffffff; border-radius: 3px;">
<tr>
<td class="wrapper" style="font-family: sans-serif; font-size: 14px; \
vertical-align: top; box-sizing: border-box; padding: 20px;">
<table border="0" cellpadding="0" cellspacing="0" \
style="border-collapse: separate; mso-table-lspace: 0pt; mso-table-rspace: 0pt; \
width: 100%;"> <tr>
<td style="font-family: sans-serif; font-size: 14px; \
vertical-align: top;">
<div>
<h1 style="color: #000; font-family: sans-serif; line-height: 1.4; margin: 0; \
margin-bottom: 30px; font-size: 25px; font-weight: 300; text-align: \
center">PostgreSQL Weekly News - January 31, 2021</h1> </div>
<h1 style="color: #000; font-family: sans-serif; line-height: 1.4; margin: 0; \
margin-bottom: 30px; font-size: 25px; font-weight: 300; text-align: \
center">PostgreSQL Weekly News - January 31, 2021</h1> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Crunchy PostgreSQL Operator 4.6.0, a system for deploying and managing open \
source PostgreSQL clusters on Kubernetes, released. <a \
href="https://access.crunchydata.com/documentation/postgres-operator/latest/releases/4.6.0/" \
style="color: #3498db; text-decoration: \
underline">https://access.crunchydata.com/documentation/postgres-operator/latest/releases/4.6.0/</a></p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">pgAdmin4 4.30, a web- and native GUI control center for \
PostgreSQL, released. <a \
href="https://www.pgadmin.org/docs/pgadmin4/dev/release_notes_4_30.html" \
style="color: #3498db; text-decoration: \
underline">https://www.pgadmin.org/docs/pgadmin4/dev/release_notes_4_30.html</a></p> \
<h1 style="color: #000; font-family: sans-serif; line-height: 1.4; margin: 0; \
margin-bottom: 30px; font-size: 25px; font-weight: 300; text-align: \
center">PostgreSQL Product News</h1> <h1 style="color: #000; font-family: sans-serif; \
line-height: 1.4; margin: 0; margin-bottom: 30px; font-size: 25px; font-weight: 300; \
text-align: center">PostgreSQL Jobs for January</h1> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"><a \
href="https://archives.postgresql.org/pgsql-jobs/2021-01/" style="color: #3498db; \
text-decoration: underline">https://archives.postgresql.org/pgsql-jobs/2021-01/</a></p>
<h1 style="color: #000; font-family: sans-serif; line-height: 1.4; margin: 0; \
margin-bottom: 30px; font-size: 25px; font-weight: 300; text-align: \
center">PostgreSQL in the News</h1> <p style="font-family: sans-serif; font-size: \
14px; font-weight: normal; margin: 0; margin-bottom: 15px">Planet PostgreSQL: <a \
href="https://planet.postgresql.org/" style="color: #3498db; text-decoration: \
underline">https://planet.postgresql.org/</a></p> <p style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px">PostgreSQL \
Weekly News is brought to you this week by David Fetter</p> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Submit news and announcements by Sunday at 3:00pm PST8PDT to \
david@fetter.org.</p> <h1 style="color: #000; font-family: sans-serif; line-height: \
1.4; margin: 0; margin-bottom: 30px; font-size: 25px; font-weight: 300; text-align: \
center">Applied Patches</h1> <p style="font-family: sans-serif; font-size: 14px; \
font-weight: normal; margin: 0; margin-bottom: 15px">Tom Lane pushed:</p> <ul \
style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px"> <li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Update time zone data files to tzdata release 2021a. DST law \
changes in Russia (Volgograd zone) and South Sudan. Historical corrections for \
Australia, Bahamas, Belize, Bermuda, Ghana, Israel, Kenya, Nigeria, Palestine,
Seychelles, and Vanuatu. Notably, the Australia/Currie zone has been corrected
to the point where it is identical to Australia/Hobart.
<a href="https://git.postgresql.org/pg/commitdiff/c7edf4ac246b67073563354c2808c78868cbac36" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/c7edf4ac246b67073563354c2808c78868cbac36</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Make storage/standby.h compile standalone again. This file has \
failed headerscheck/cpluspluscheck verification since commit 0650ff230, as a result
of referencing typedef TimestampTz without including the appropriate header.
<a href="https://git.postgresql.org/pg/commitdiff/a4b03de589c1df0845e9732da203f505f2eedb6d" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/a4b03de589c1df0845e9732da203f505f2eedb6d</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Doc: improve documentation of pg_proc.protrftypes. Add a \
"references" link pointing to pg_type, as we have for other arrays of type OIDs. \
Wordsmith the explanation a bit. Joel Jacobson, additional editing by me \
Discussion: <a href="https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/951862eda57e5dc8f78c97b3c30fe2032a5562b8" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/951862eda57e5dc8f78c97b3c30fe2032a5562b8</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fix broken ruleutils support for function TRANSFORM clauses. I \
chanced to notice that this dumped core due to a faulty Assert. To add insult to \
injury, the output has been misformatted since v11. Obviously we need some \
regression testing here. Discussion:
<a href="https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/d1cc628c-3953-4209-957b-29427acc38c8@www.fastmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/07d46fceb4254b00e79f3d06419cbae13b0ecb5a" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/07d46fceb4254b00e79f3d06419cbae13b0ecb5a</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Don't clobber the calling user's credentials cache in Kerberos \
test. Embarrassing oversight in this test script, which fortunately is not run by
default. Report and patch by Jacob Champion. Discussion:
<a href="https://postgr.es/m/1fcb175bafef6560f47a8c31229fa7c938486b8d.camel@vmware.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/1fcb175bafef6560f47a8c31229fa7c938486b8d.camel@vmware.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/881933f194221abcce07fb134ebe8685e5bb58dd" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/881933f194221abcce07fb134ebe8685e5bb58dd</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Improve performance of repeated CALLs within plpgsql procedures. \
This patch essentially is cleaning up technical debt left behind by the original
implementation of plpgsql procedures, particularly commit d92bc83c4. That
patch (or more precisely, follow-on patches fixing its worst bugs) forced us
to re-plan CALL and DO statements each time through, if we're in a non-atomic
context. That wasn't for any fundamental reason, but just because use of a
saved plan requires having a ResourceOwner to hold a reference count for the
plan, and we had no suitable resowner at hand, nor would the available APIs
support using one if we did. While it's not that expensive to create a "plan"
for CALL/DO, the cycles do add up in repeated executions. This patch
therefore makes the following API changes: <code>*</code> \
GetCachedPlan/ReleaseCachedPlan are modified to let the caller specify which \
resowner to use to pin the plan, rather than forcing use of CurrentResourceOwner. \
<code>*</code> spi.c gains a "SPI_execute_plan_extended" entry point that lets \
callers say which resowner to use to pin the plan. This borrows the idea of an \
options struct from the recently added SPI_prepare_extended, hopefully allowing \
future options to be added without more API breaks. This supersedes \
SPI_execute_plan_with_paramlist (which I've marked deprecated) as well as \
SPI_execute_plan_with_receiver (which is new in v14, so I just took it out \
altogether). <code>*</code> I also took the opportunity to remove the crude hack of \
letting plpgsql reach into SPI private data structures to mark SPI plans as \
"no_snapshot". It's better to treat that as an option of SPI_prepare_extended. \
Now, when running a non-atomic procedure or DO block that contains any CALL or DO \
commands, plpgsql creates a ResourceOwner that will be used to pin the plans of the \
CALL/DO commands. (In an atomic context, we just use CurrentResourceOwner, as \
before.) Having done this, we can just save CALL/DO plans normally, whether or not \
they are used across transaction boundaries. This seems to be good for something \
like 2X speedup of a CALL of a trivial procedure with a few simple argument
expressions. By restricting the creation of an extra ResourceOwner like this,
there's essentially zero penalty in cases that can't benefit. Pavel Stehule,
with some further hacking by me Discussion:
<a href="https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/ee895a655ce4341546facd6f23e3e8f2931b96bf" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/ee895a655ce4341546facd6f23e3e8f2931b96bf</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Code review for psql's helpSQL() function. The loops to identify \
word boundaries could access past the end of the input string. Likely that would
never result in an actual crash, but it makes valgrind unhappy. The logic to
try different numbers of words didn't work when the input has two words but we
only have a match to the first, eg "\h with select". (We must "continue" the
pass loop, not "break".) The logic to compute nl_count was bizarrely managed,
and in at least two code paths could end up calling PageOutput with nl_count =
0, resulting in failing to paginate output that should have been fed to the
pager. Also, in v12 and up, the nl_count calculation hadn't been updated to
account for the addition of a URL. The PQExpBuffer holding the command syntax
details wasn't freed, resulting in a session-lifespan memory leak. While
here, improve some comments, choose a more descriptive name for a variable,
fix inconsistent datatype choice for another variable. Per bug #16837 from
Alexander Lakhin. This code is very old, so back-patch to all supported
branches. Kyotaro Horiguchi and Tom Lane Discussion:
<a href="https://postgr.es/m/16837-479bcd56040c71b3@postgresql.org" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/16837-479bcd56040c71b3@postgresql.org</a> <a \
href="https://git.postgresql.org/pg/commitdiff/f76a85000bba2f1b1c926cbbe525e47b246215f1" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/f76a85000bba2f1b1c926cbbe525e47b246215f1</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Suppress compiler warnings from commit ee895a655. For obscure \
reasons, some buildfarm members are now generating complaints about \
plpgsql_call_handler's "retval" variable possibly being used uninitialized. It \
seems no less safe than it was before that commit, but these complaints are \
(mostly?) new. I trust that initializing the variable where it's declared will be \
enough to shut that up. I also notice that some compilers are warning about setjmp
clobber of the same variable, which is maybe a bit more defensible. Mark it
volatile to silence that. Also, rearrange the logic to give
procedure_resowner a single point of initialization, in hopes of silencing
some setjmp-clobber warnings about that. (Marking it volatile would serve
too, but its sibling variables are depending on single assignment, so let's
stick with that method.) Discussion:
<a href="https://postgr.es/m/E1l4F1z-0000cN-Lx@gemulon.postgresql.org" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/E1l4F1z-0000cN-Lx@gemulon.postgresql.org</a> <a \
href="https://git.postgresql.org/pg/commitdiff/7292fd8f1c781278021407276474d9188845113d" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/7292fd8f1c781278021407276474d9188845113d</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Rethink recently-added SPI interfaces. SPI_execute_with_receiver \
and SPI_cursor_parse_open_with_paramlist are new in v14 (cf. commit 2f48ede08).
Before they can get out the door, let's change their APIs to follow the
practice recently established by SPI_prepare_extended etc: shove all optional
arguments into a struct that callers are supposed to pre-zero. The hope is to
allow future addition of more options without either API breakage or a
continuing proliferation of new SPI entry points. With that in mind, choose
slightly more generic names for them: SPI_execute_extended and
SPI_cursor_parse_open respectively. Discussion:
<a href="https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAFj8pRCLPdDAETvR7Po7gC5y_ibkn_-bOzbeJb39WHms01194Q@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/d5a83d79c9f9b660a6a5a77afafe146d3c8c6f46" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/d5a83d79c9f9b660a6a5a77afafe146d3c8c6f46</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Doc: improve documentation for UNNEST(). Per a user question, \
spell out that UNNEST() returns array elements in storage order; also provide an \
example to clarify the behavior for multi-dimensional arrays. While here, also \
clarify the SELECT reference page's description of WITH ORDINALITY. These details
were already given in 7.2.1.4, but a reference page should not omit details.
Back-patch to v13; there's not room in the table in older versions.
Discussion:
<a href="https://postgr.es/m/FF1FB31F-0507-4F18-9559-2DE6E07E3B43@gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/FF1FB31F-0507-4F18-9559-2DE6E07E3B43@gmail.com</a> <a \
href="https://git.postgresql.org/pg/commitdiff/662affcfe9e816584e3d8602b3b4005236931bbb" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/662affcfe9e816584e3d8602b3b4005236931bbb</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Make ecpg's rjulmdy() and rmdyjul() agree with their \
declarations. We had <code>"short *mdy"</code> in the extern declarations, but \
"short mdy[3]" in the actual function definitions. Per C99 these are equivalent, \
but recent versions of gcc have started to issue warnings about the inconsistency. \
Clean it up before the warnings get any more widespread. Back-patch, in case anyone \
wants to build older PG versions with bleeding-edge compilers. Discussion:
<a href="https://postgr.es/m/2401575.1611764534@sss.pgh.pa.us" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/2401575.1611764534@sss.pgh.pa.us</a> <a \
href="https://git.postgresql.org/pg/commitdiff/1b242f42ba1e732b202f35265ab1a1614ce7d859" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/1b242f42ba1e732b202f35265ab1a1614ce7d859</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fix hash partition pruning with asymmetric partition sets. \
perform_pruning_combine_step() was not taught about the number of partition indexes \
used in hash partitioning; more embarrassingly, get_matching_hash_bounds() also had \
it wrong. These errors are masked in the common case where all the partitions have \
the same modulus and no partition is missing. However, with missing or unequal-size \
partitions, we could erroneously prune some partitions that need to be scanned, \
leading to silently wrong query answers. While a minimal-footprint fix for this \
could be to export get_partition_bound_num_indexes and make the incorrect functions \
use it, I'm of the opinion that that function should never have existed in the
first place. It's not reasonable data structure design that
PartitionBoundInfoData lacks any explicit record of the length of its
indexes[] array. Perhaps that was all right when it could always be assumed
equal to ndatums, but something should have been done about it as soon as that
stopped being true. Putting in an explicit "nindexes" field makes both
partition_bounds_equal() and partition_bounds_copy() simpler, safer, and
faster than before, and removes explicit knowledge of the
number-of-partition-indexes rules from some other places too. This change
also makes get_hash_partition_greatest_modulus obsolete. I left that in place
in case any external code uses it, but no core code does anymore. Per bug
#16840 from Michał Albrycht. Back-patch to v11 where the hash partitioning
code came in. (In the back branches, add the new field at the end of
PartitionBoundInfoData to minimize ABI risks.) Discussion:
<a href="https://postgr.es/m/16840-571a22976f829ad4@postgresql.org" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/16840-571a22976f829ad4@postgresql.org</a> <a \
href="https://git.postgresql.org/pg/commitdiff/1d9351a87c9a9e82c7091aab03d9299982670ce0" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/1d9351a87c9a9e82c7091aab03d9299982670ce0</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Silence another gcc 11 warning. Per buildfarm and local \
experimentation, bleeding-edge gcc isn't convinced that the MemSet in
reorder_function_arguments() is safe. Shut it up by adding an explicit check
that pronargs isn't negative, and by changing MemSet to memset. (It appears
that either change is enough to quiet the warning at -O2, but let's do both to
be sure.)
<a href="https://git.postgresql.org/pg/commitdiff/1046dbedde2fc3fe55f007ff3255ab65ab98f858" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/1046dbedde2fc3fe55f007ff3255ab65ab98f858</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Doc: improve cross-references for SET/SHOW. The corresponding \
functions set_config and current_setting were mostly not hyperlinked. Clarify their
descriptions a tad, too. Discussion:
<a href="https://postgr.es/m/161183356250.4077.687338658090583892@wrigleys.postgresql.org" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/161183356250.4077.687338658090583892@wrigleys.postgresql.org</a>
<a href="https://git.postgresql.org/pg/commitdiff/f743a2bbd4e511ad58b6ce01f81841e5e1611474" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/f743a2bbd4e511ad58b6ce01f81841e5e1611474</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Amit Kapila pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px">Fix ALTER PUBLICATION...DROP \
TABLE behavior. Commit 69bd60672 fixed the initialization of streamed transactions \
for RelationSyncEntry. It forgot to initialize the publication actions while \
invalidating the RelationSyncEntry due to which even though the relation is dropped \
from a particular publication we still publish its changes. Fix it by initializing \
pubactions when entry got invalidated. Author: Japin Li and Bharath Rupireddy \
Reviewed-by: Amit Kapila Discussion:
<a href="https://postgr.es/m/CALj2ACV+0UFpcZs5czYgBpujM9p0Hg1qdOZai_43OU7bqHU_xw@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CALj2ACV+0UFpcZs5czYgBpujM9p0Hg1qdOZai_43OU7bqHU_xw@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/40ab64c1ec1cb9bd73695f519cf66ddbb97d8144" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/40ab64c1ec1cb9bd73695f519cf66ddbb97d8144</a></li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">David Rowley pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px">Fix hypothetical bug in heap \
backward scans. Both heapgettup() and heapgettup_pagemode() incorrectly set the \
first page to scan in a backward scan in which the number of pages to scan was \
specified by heap_setscanlimits(). The code incorrectly started the scan at the end \
of the relation when startBlk was 0, or otherwise at startBlk - 1, neither of which
is correct when only scanning a subset of pages. The fix here checks if
heap_setscanlimits() has changed the number of pages to scan and if so we set
the first page to scan as the final page in the specified range during
backward scans. Proper adjustment of this code was forgotten when
heap_setscanlimits() was added in 7516f5259 back in 9.5. However, practice,
nowhere in core code performs backward scans after having used
heap_setscanlimits(), yet, it is possible an extension uses the heap functions
in this way, hence backpatch. An upcoming patch does use heap_setscanlimits()
with backward scans, so this must be fixed before that can go in. Author:
David Rowley Discussion:
<a href="https://postgr.es/m/CAApHDvpGc9h0_oVD2CtgBcxCS1N-qDYZSeBRnUh+0CWJA9cMaA@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAApHDvpGc9h0_oVD2CtgBcxCS1N-qDYZSeBRnUh+0CWJA9cMaA@mail.gmail.com</a>
Backpatch-through: 9.5, all supported versions
<a href="https://git.postgresql.org/pg/commitdiff/16dfe253e31f75b60e93acc0c2b5bbf19936c074" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/16dfe253e31f75b60e93acc0c2b5bbf19936c074</a></li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Peter Eisentraut pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Remove duplicate include. Reported-by: Ashutosh Sharma <a style="color: \
#3498db; text-decoration: \
underline">ashu.coek88@gmail.com</a>
Discussion:
<a href="https://www.postgresql.org/message-id/flat/CAE9k0PkORqHHGKY54-sFyDpP90yAf%2B05Auc4fs9EAn4J%2BuBeUQ%40mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/flat/CAE9k0PkORqHHGKY54-sFyDpP90yAf%2B05Auc4fs9EAn4J%2BuBeUQ%40mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/ecc4b1318c92594abdf9bd2933fb6feaa9d56229" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/ecc4b1318c92594abdf9bd2933fb6feaa9d56229</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Remove gratuitous uses of deprecated SELECT INTO. CREATE TABLE \
AS has been preferred over SELECT INTO (outside of ecpg and PL/pgSQL) for a long \
time. There were still a few uses of SELECT INTO in tests and documentation, some
old, some more recent. This changes them to CREATE TABLE AS. Some
occurrences in the tests remain where they are specifically testing SELECT
INTO parsing or similar. Discussion:
<a href="https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40enterprisedb.com" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40enterprisedb.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/b034ef9b376dbe712caa076541d6a750f37d85ce" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/b034ef9b376dbe712caa076541d6a750f37d85ce</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fix typo. <a \
href="https://git.postgresql.org/pg/commitdiff/2592be8be5aed6448a1e2b386fffd75ad69eb6c5" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/2592be8be5aed6448a1e2b386fffd75ad69eb6c5</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Allow GRANTED BY clause in normal GRANT and REVOKE statements. \
The SQL standard allows a GRANTED BY clause on GRANT and REVOKE (privilege) \
statements that can specify CURRENT_USER or CURRENT_ROLE. In PostgreSQL, both of \
these are the default behavior. Since we already have all the parsing support for
this for the GRANT (role) statement, we might as well add basic support for
this for the privilege variant as well. This allows us to check off SQL
feature T332. In the future, perhaps more interesting things could be done
with this, too. Reviewed-by: Simon Riggs <a style="color: #3498db; \
text-decoration: underline">simon@2ndquadrant.com</a> \
Discussion: <a href="https://www.postgresql.org/message-id/flat/f2feac44-b4c5-f38f-3699-2851d6a76dc9@2ndquadrant.com" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/flat/f2feac44-b4c5-f38f-3699-2851d6a76dc9@2ndquadrant.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/6aaaa76bb47db11cd6f567eafa3d1ee81ca59556" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/6aaaa76bb47db11cd6f567eafa3d1ee81ca59556</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">doc: Clarify status of SELECT INTO on reference page. The \
documentation as well as source code comments weren't entirely clear whether SELECT \
INTO was truly deprecated (thus in theory destined to be removed eventually), or \
just a less recommended variant. After discussion, it appears that other
implementations also use SELECT INTO in direct SQL in a way similar to
PostgreSQL, so it seems worth keeping it for compatibility. Update the
language in the documentation to that effect. Discussion:
<a href="https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40enterprisedb.com" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/flat/96dc0df3-e13a-a85d-d045-d6e2c85218da%40enterprisedb.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/65330622441d7ee08f768c4326825ae903f2595a" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/65330622441d7ee08f768c4326825ae903f2595a</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Add primary keys and unique constraints to system catalogs. For \
those system catalogs that have a unique indexes, make a primary key and unique \
constraint, using ALTER TABLE ... PRIMARY KEY/UNIQUE USING INDEX. This can be \
helpful for GUI tools that look for a primary key, and it might in the future allow
declaring foreign keys, for making schema diagrams. The constraint creation
statements are automatically created by genbki.pl from DECLARE_UNIQUE_INDEX
directives. To specify which one of the available unique indexes is the
primary key, use the new directive DECLARE_UNIQUE_INDEX_PKEY instead. By
convention, we usually make a catalog's OID column its primary key, if it has
one. Reviewed-by: Tom Lane <a style="color: #3498db; text-decoration: \
underline">tgl@sss.pgh.pa.us</a> \
Discussion: <a href="https://www.postgresql.org/message-id/flat/dc5f44d9-5ec1-a596-0251-dadadcdede98@2ndquadrant.com" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/flat/dc5f44d9-5ec1-a596-0251-dadadcdede98@2ndquadrant.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/dfb75e478cacb33d277432e0df5e2f9a2a2b52d9" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/dfb75e478cacb33d277432e0df5e2f9a2a2b52d9</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Robert Haas pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Remove CheckpointLock. Up until now, we've held this lock when performing a \
checkpoint or restartpoint, but commit 076a055acf3c55314de267c62b03191586d79cf6 back \
in 2004 and commit 7e48b77b1cebb9a43f9fdd6b17128a0ba36132f9 from 2009, taken \
together, have removed all need for this. In the present code, there's only ever one \
process entitled to attempt a checkpoint: either the checkpointer, during normal
operation, or the postmaster, during single-user operation. So, we don't need
the lock. One possible concern in making this change is that it means that a
substantial amount of code where HOLD_INTERRUPTS() was previously in effect
due to the preceding LWLockAcquire() will now be running without that. This
could mean that ProcessInterrupts() gets called in places from which it didn't
before. However, this seems unlikely to do very much, because the checkpointer
doesn't have any signal mapped to die(), so it's not clear how, for example,
ProcDiePending = true could happen in the first place. Similarly with
ClientConnectionLost and recovery conflicts. Also, if there are any such
problems, we might want to fix them rather than reverting this, since running
lots of code with interrupt handling suspended is generally bad. Patch by me,
per an inquiry by Amul Sul. Review by Tom Lane and Michael Paquier.
Discussion:
<a href="http://postgr.es/m/CAAJ_b97XnBBfYeSREDJorFsyoD1sHgqnNuCi=02mNQBUMnA=FA@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">http://postgr.es/m/CAAJ_b97XnBBfYeSREDJorFsyoD1sHgqnNuCi=02mNQBUMnA=FA@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/d18e75664a2fda2e4d5cc433d68e37fc0e9499f2" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/d18e75664a2fda2e4d5cc433d68e37fc0e9499f2</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Move StartupCLOG() calls to just after we initialize \
ShmemVariableCache. Previously, the hot_standby=off code path did this at end of \
recovery, while the hot_standby=on code path did it at the beginning of recovery. \
It's better to do this in only one place because (a) it's simpler, (b) StartupCLOG() \
is trivial so trying to postpone the work isn't useful, and (c) this will make it
possible to simplify some other logic. Patch by me, reviewed by Heikki
Linnakangas. Discussion:
<a href="http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/1f113abdf87cd085dee3927960bb4f70442b7250" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/1f113abdf87cd085dee3927960bb4f70442b7250</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">In clog_redo(), don't set \
XactCtl->shared->latest_page_number. The comment is no longer accurate, and \
hasn't been entirely accurate since Hot Standby was introduced. The original idea \
here was that StartupCLOG() wouldn't be called until the end of recovery and \
therefore this value would be uninitialized when this code is reached, but Hot \
Standby made that true only when hot_standby=off, and commit \
1f113abdf87cd085dee3927960bb4f70442b7250 means that this value is now always \
initialized before replay even starts. The original purpose of this code was to \
bypass the sanity check in SimpleLruTruncate(), which will no longer occur: now, if \
something is wrong, that sanity check might trip during recovery. That's probably a \
good thing, because in the current code base latest_page_number should always be
initialized and therefore we expect that the sanity check should pass. If it
doesn't, something has gone wrong, and complaining about it is appropriate.
Patch by me, reviewed by Heikki Linnakangas. Discussion:
<a href="http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/0fcc2decd485a61321a3220d8f76cb108b082009" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/0fcc2decd485a61321a3220d8f76cb108b082009</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">In TrimCLOG(), don't reset \
XactCtl->shared->latest_page_number. Since the CLOG page number is not \
recorded directly in the checkpoint record, we have to use \
ShmemVariableCache->nextXid to figure out the latest CLOG page number at the \
start of recovery. However, as recovery progresses, replay of CLOG/EXTEND records \
will update our notion of the latest page number, and we should rely on that being \
accurate rather than recomputing the value based on an updated notion of nextXid. \
ShmemVariableCache->nextXid is only an approximation during recovery anyway, \
whereas CLOG/EXTEND records are an authoritative representation of how the SLRU has \
been updated. Commit 0fcc2decd485a61321a3220d8f76cb108b082009 makes this \
simplification possible, as before that change clog_redo() might have injected a \
bogus value here, and we'd want to get rid of that before entering normal running. \
Patch by me, reviewed by Heikki Linnakangas. Discussion:
<a href="http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">http://postgr.es/m/CA+TgmoZYig9+AQodhF5sRXuKkJ=RgFDugLr3XX_dz_F-p=TwTg@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/69059d3b2f0754c8e661ba479f7121e6631cdf4a" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/69059d3b2f0754c8e661ba479f7121e6631cdf4a</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Andres Freund pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px">Fix two typos in snapbuild.c. \
Reported-by: Heikki Linnakangas <a style="color: #3498db; text-decoration: \
underline">hlinnaka@iki.fi</a> \
Discussion: <a href="https://postgr.es/m/c94be044-818f-15e3-1ad3-7a7ae2dfed0a@iki.fi" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/c94be044-818f-15e3-1ad3-7a7ae2dfed0a@iki.fi</a> <a \
href="https://git.postgresql.org/pg/commitdiff/55ef8555f0c1207bac25050e7297bbb969c84233" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/55ef8555f0c1207bac25050e7297bbb969c84233</a></li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fujii Masao pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">postgres_fdw: Add functions to discard cached connections. This commit \
introduces two new functions postgres_fdw_disconnect() and \
postgres_fdw_disconnect_all(). The former function discards the cached connections \
to the specified foreign server. The latter discards all the cached connections. If \
the connection is used in the current transaction, it's not closed and a warning \
message is emitted. For example, these functions are useful when users want to \
explicitly close the foreign server connections that are no longer necessary and \
then to prevent them from eating up the foreign servers connections capacity. \
Author: Bharath Rupireddy, tweaked a bit by Fujii Masao Reviewed-by: Alexey \
Kondratov, Zhijie Hou, Zhihong Yu, Fujii Masao Discussion:
<a href="https://postgr.es/m/CALj2ACVvrp5=AVp2PupEm+nAC8S4buqR3fJMmaCoc7ftT0aD2A@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CALj2ACVvrp5=AVp2PupEm+nAC8S4buqR3fJMmaCoc7ftT0aD2A@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/411ae64997dc3a42d19eda6721c581841ce2cb82" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/411ae64997dc3a42d19eda6721c581841ce2cb82</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">postgres_fdw: Stabilize regression test for \
postgres_fdw_disconnect_all(). The regression test added in commit 411ae64997 caused \
buildfarm failures. The cause of them was that the order of warning messages output \
in the test was not stable. To fix this, this commit sets client_min_messages to \
ERROR temporarily when performing the test generating those warnings. Per buildfarm
failures. Discussion:
<a href="https://postgr.es/m/2147113.1611644754@sss.pgh.pa.us" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/2147113.1611644754@sss.pgh.pa.us</a> <a \
href="https://git.postgresql.org/pg/commitdiff/6adc5376dca4ef8b7d591c0ee7338cb9ff660216" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/6adc5376dca4ef8b7d591c0ee7338cb9ff660216</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">postgres_fdw: Fix test failure with
-DENFORCE_REGRESSION_TEST_NAME_RESTRICTIONS. The roles created by regression
test should have names starting with "regress_", and the test introduced in
commit 411ae64997 did not do that. Per buildfarm member longfin. Discussion:
<a href="https://postgr.es/m/73fc5ae4-3c54-1262-4533-f8c547de2e60@oss.nttdata.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/73fc5ae4-3c54-1262-4533-f8c547de2e60@oss.nttdata.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/0c3fc09fe359a6dc46f1870ceccf60ec60396bc9" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/0c3fc09fe359a6dc46f1870ceccf60ec60396bc9</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">postgres_fdw: Fix tests for CLOBBER_CACHE_ALWAYS. The \
regression tests added in commits 708d165ddb and 411ae64997 caused buildfarm \
failures when CLOBBER_CACHE_ALWAYS was enabled. This commit stabilizes those tests. \
The foreign server connections established by postgres_fdw behaves differently
depending on whether CLOBBER_CACHE_ALWAYS is enabled or not. If it's not
enabled, those connections are cached. On the other hand, if it's enabled,
when the connections are established outside transaction block, they are not
cached (i.e., they are immediately closed at the end of query that established
them). So the subsequent postgres_fdw_get_connections() cannot list those
connections and postgres_fdw_disconnect() cannot close them (because they are
already closed). When the connections are established inside transaction
block, they are cached whether CLOBBER_CACHE_ALWAYS was enabled or not. But if
it's enabled, they are immediately marked as invalid, otherwise not. This
causes the subsequent postgres_fdw_get_connections() to return different
result in "valid" column depending on whether CLOBBER_CACHE_ALWAYS was enabled
or not. This commit prevents the above differences of behavior from affecting
the regression tests. Per buildfarm failure on trilobite. Original patch by
Bharath Rupireddy. I (Fujii Masao) extracted the regression test fix from that
and revised it a bit. Reported-by: Tom Lane Author: Bharath Rupireddy
Reviewed-by: Fujii Masao Discussion:
<a href="https://postgr.es/m/2688508.1611865371@sss.pgh.pa.us" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/2688508.1611865371@sss.pgh.pa.us</a> <a \
href="https://git.postgresql.org/pg/commitdiff/f77717b2985aa529a185e6988de26b885ca10ddb" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/f77717b2985aa529a185e6988de26b885ca10ddb</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Michaël Paquier pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px">Fix \
memory leak when deallocating prepared statement in postgres_fdw. The leak is minor, \
so no backpatch is done. Oversight in 21734d2. Reported-by: Tom Lane
<a href="https://git.postgresql.org/pg/commitdiff/7b4c660466dd8a1d25ca316ac02099a7cf0e5896" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/7b4c660466dd8a1d25ca316ac02099a7cf0e5896</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Refactor code in tablecmds.c to check and process tablespace \
moves. Two code paths of tablecmds.c (for relations with storage and without \
storage) use the same logic to check if the move of a relation to a new tablespace \
is allowed or not and to update pg_class.reltablespace and pg_class.relfilenode. A
potential TABLESPACE clause for REINDEX, CLUSTER and VACUUM FULL needs similar
checks to make sure that nothing is moved around in illegal ways (no mapped
relations, shared relations only in pg_global, no move of temp tables owned by
other backends). This reorganizes the existing code of ALTER TABLE so as all
this logic is controlled by two new routines that can be reused for the other
commands able to move relations across tablespaces, limiting the number of
code paths in need of the same protections. This also removes some code that
was duplicated for tables with and without storage for ALTER TABLE. Author:
Alexey Kondratov, Michael Paquier Discussion:
<a href="https://postgr.es/m/YA+9mAMWYLXJMVPL@paquier.xyz" style="color: #3498db; \
text-decoration: underline">https://postgr.es/m/YA+9mAMWYLXJMVPL@paquier.xyz</a> <a \
href="https://git.postgresql.org/pg/commitdiff/4c9c359d38ff1e2de388eedd860785be6a49201c" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/4c9c359d38ff1e2de388eedd860785be6a49201c</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">doc: Remove reference to views for TRUNCATE privilege. The page \
about privilege rights mentioned that TRUNCATE could be applied to views or even
other relation types. This is confusing as this command can be used only on
tables and on partitioned tables. Oversight in afc4a78. Reported-by: Harisai
Hari Reviewed-by: Laurenz Albe Discussion:
<a href="https://postgr.es/m/161157636877.14625.15340884663716426087@wrigleys.postgresql.org" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/161157636877.14625.15340884663716426087@wrigleys.postgresql.org</a>
Backpatch-through: 12
<a href="https://git.postgresql.org/pg/commitdiff/32bef758296142ce0fff944c3478d63fdefdaf33" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/32bef758296142ce0fff944c3478d63fdefdaf33</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Refactor SQL functions of SHA-2 in cryptohashfuncs.c. The same \
code pattern was repeated four times when compiling a SHA-2 hash. This refactoring \
has the advantage to issue a compilation warning if a new value is added to
pg_cryptohash_type, so as anybody doing an addition in this area would need to
consider if support for a new SQL function is needed or not. Author: Sehrope
Sarkuni, Michael Paquier Discussion:
<a href="https://postgr.es/m/YA7DvLRn2xnTgsMc@paquier.xyz" style="color: #3498db; \
text-decoration: underline">https://postgr.es/m/YA7DvLRn2xnTgsMc@paquier.xyz</a> <a \
href="https://git.postgresql.org/pg/commitdiff/f854c69a5b36ba7aa85bee9e9590c3e517970156" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/f854c69a5b36ba7aa85bee9e9590c3e517970156</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fix crash of <code>pg_stat_statements_info()</code> without \
library loaded. Other code paths are already protected against this case, and \
_PG_init() warns about that in pg_stat_statements.c. While on it, I have checked \
the other extensions of the tree but did not notice any holes. Oversight in \
9fbc3f3. Author: Jaime Casanova Reviewed-by: Julien Rouhaud Discussion:
<a href="https://postgr.es/m/CAJKUy5gF4=_=qhJ1VX_tSGFfjKHb9BvzhRYWSApJD=Bfwp2SBw@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline"><code>https://postgr.es/m/CAJKUy5gF4=_=qhJ1VX_tSGFfjKHb9BvzhRYWSApJD=Bfwp2SBw@mail.gmail.com</code></a>
<a href="https://git.postgresql.org/pg/commitdiff/bca96ddab57c7c64ce442b0f9f35558741c7f8bd" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/bca96ddab57c7c64ce442b0f9f35558741c7f8bd</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Adjust comments of CheckRelationTableSpaceMove() and \
SetRelationTableSpace(). 4c9c359, that introduced those two functions, has been \
overoptimistic on the point that only ShareUpdateExclusiveLock would be required \
when moving a relation to a new tablespace. AccessExclusiveLock is a requirement, \
but ShareUpdateExclusiveLock may be used under specific conditions like REINDEX
CONCURRENTLY where waits on past transactions make the operation safe even
with a lower-level lock. The current code does only the former, so update the
existing comments to reflect that. Once a REINDEX (TABLESPACE) is introduced,
those comments would require an extra refresh to mention their new use case.
While on it, fix an incorrect variable name. Per discussion with Álvaro
Herrera. Discussion:
<a href="https://postgr.es/m/20210127140741.GA14174@alvherre.pgsql" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/20210127140741.GA14174@alvherre.pgsql</a> <a \
href="https://git.postgresql.org/pg/commitdiff/24843297a96d7be16cc3f4b090aacfc6e5e6839e" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/24843297a96d7be16cc3f4b090aacfc6e5e6839e</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">doc: Improve wording of section for repslot statistics. This \
documentation has been added in 9868167, so no backpatch is needed. Author: Justin \
Pryzby, Michael Paquier Discussion:
<a href="https://postgr.es/m/20201222041153.GK30237@telsasoft.com" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/20201222041153.GK30237@telsasoft.com</a> <a \
href="https://git.postgresql.org/pg/commitdiff/2a5862f013647dd185de3e1b496fd4f614d12f0e" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/2a5862f013647dd185de3e1b496fd4f614d12f0e</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Peter Geoghegan pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px">Fix \
GiST index deletion assert issue. Avoid calling heap_index_delete_tuples() with an \
empty deltids array to avoid an assertion failure. This issue was arguably an \
oversight in commit b5f58cf2, though the failing assert itself was added by my \
recent commit d168b666. No backpatch, though, since the oversight is harmless in \
the back branches. Author: Peter Geoghegan <a style="color: #3498db; \
text-decoration: underline">pg@bowt.ie</a>
Reported-By: Jaime Casanova <a style="color: #3498db; text-decoration: \
underline">jcasanov@syst& \
#101;mguards.com.ec</a> \
Discussion: <a href="https://postgr.es/m/CAJKUy5jscES84n3puE=sYngyF+zpb4wv8UMtuLnLPv5z=6yyNw@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAJKUy5jscES84n3puE=sYngyF+zpb4wv8UMtuLnLPv5z=6yyNw@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/e42b3c3bd6a9c6233ac4c8a2e9b040367ba2f97c" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/e42b3c3bd6a9c6233ac4c8a2e9b040367ba2f97c</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Reduce the default value of vacuum_cost_page_miss. When commit \
f425b605 introduced cost based vacuum delays back in 2004, the defaults reflected
then-current trends in hardware, as well as certain historical limitations in
PostgreSQL. There have been enormous improvements in both areas since that
time. The cost limit GUC defaults finally became much more representative of
current trends following commit cbccac37, which decreased
autovacuum_vacuum_cost_delay's default by 10x for PostgreSQL 12 (it went from
20ms to only 2ms). The relative costs have shifted too. This should also be
accounted for by the defaults. More specifically, the relative importance of
avoiding dirtying pages within VACUUM has greatly increased, primarily due to
main memory capacity scaling and trends in flash storage. Within Postgres
itself, improvements like sequential access during index vacuuming (at least
in nbtree and GiST indexes) have also been contributing factors. To reflect
all this, decrease the default of vacuum_cost_page_miss to 2. Since the
default of vacuum_cost_page_dirty remains 20, dirtying a page is now
considered 10x "costlier" than a page miss by default. Author: Peter
Geoghegan <a style="color: #3498db; text-decoration: \
underline">pg@bowt.ie</a> Discussion: \
<a href="https://postgr.es/m/CAH2-WzmLPFnkWT8xMjmcsm7YS3+_Qi3iRWAb2+_Bc8UhVyHfuA@mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAH2-WzmLPFnkWT8xMjmcsm7YS3+_Qi3iRWAb2+_Bc8UhVyHfuA@mail.gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/e19594c5c059d2e071b67d87ae84f569a52d2e32" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/e19594c5c059d2e071b67d87ae84f569a52d2e32</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Remove unused _bt_delitems_delete() argument. The \
latestRemovedXid values used by nbtree deletion operations are determined by \
_bt_delitems_delete()'s caller, so there is no reason to pass a separate heapRel \
argument. Oversight in commit d168b666823.
<a href="https://git.postgresql.org/pg/commitdiff/dc43492e46c7145a476cb8ca6200fc8eefe673ef" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/dc43492e46c7145a476cb8ca6200fc8eefe673ef</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Andrew Gierth pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px">Don't add bailout adjustment \
for non-strict deserialize calls. When building aggregate expression steps, strict \
checks need a bailout jump for when a null value is encountered, so there is a list \
of steps that require later adjustment. Adding entries to that list for steps that \
aren't actually strict would be harmless, except that there is an Assert which \
catches them. This leads to spurious errors on asserts builds, for data sets that \
trigger parallel aggregation of an aggregate with a non-strict deserialization
function (no such aggregates exist in the core system). Repair by not adding
the adjustment entry when it's not needed. Backpatch back to 11 where the
code was introduced. Per a report from Darafei (Komzpa) of the PostGIS
project; analysis and patch by me. Discussion:
<a href="https://postgr.es/m/87mty7peb3.fsf@news-spur.riddles.org.uk" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/87mty7peb3.fsf@news-spur.riddles.org.uk</a> <a \
href="https://git.postgresql.org/pg/commitdiff/a3367aa3c4552170004c92369681730d85e384c6" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/a3367aa3c4552170004c92369681730d85e384c6</a></li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Heikki Linnakangas pushed:</p> <ul style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> \
<li style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px">Add \
mbverifystr() functions specific to each encoding. This makes pg_verify_mbstr() \
function faster, by allowing more efficient encoding-specific implementations. All \
the implementations included in this commit are pretty naive, they just call the \
same encoding-specific verifychar functions that were used previously, but that \
already gives a performance boost because the tight character-at-a-time loop is \
simpler. Reviewed-by: John Naylor Discussion:
<a href="https://www.postgresql.org/message-id/e7861509-3960-538a-9025-b75a61188e01@iki.fi" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/e7861509-3960-538a-9025-b75a61188e01@iki.fi</a>
<a href="https://git.postgresql.org/pg/commitdiff/b80e10638e36b9d2f0b39170c613837af2ca2aac" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/b80e10638e36b9d2f0b39170c613837af2ca2aac</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Add direct conversion routines between EUC_TW and Big5. \
Conversions between EUC_TW and Big5 were previously implemented by converting the \
whole input to MIC first, and then from MIC to the target encoding. Implement \
functions to convert directly between the two. The reason to do this now is that \
I'm working on a patch that will change the conversion function signature so that
if the input is invalid, we convert as much as we can and return the number of
bytes successfully converted. That's not possible if we use an intermediary
format, because if an error happens in the intermediary -> final conversion,
we lose track of the location of the invalid character in the original input.
Avoiding the intermediate step makes the conversions faster, too.
Reviewed-by: John Naylor Discussion:
<a href="https://www.postgresql.org/message-id/b9e3167f-f84b-7aa4-5738-be578a4db924%40iki.fi" \
style="color: #3498db; text-decoration: \
underline">https://www.postgresql.org/message-id/b9e3167f-f84b-7aa4-5738-be578a4db924%40iki.fi</a>
<a href="https://git.postgresql.org/pg/commitdiff/6c5576075b0f93f2235ac8a82290fe3b6e82300d" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/6c5576075b0f93f2235ac8a82290fe3b6e82300d</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Álvaro Herrera pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">pgbench: Remove dead code. doConnect() never returns connections in state \
CONNECTION_BAD, so checking for that is pointless. Remove the code that does. This \
code has been dead since ba708ea3dc84, 20 years ago. Discussion: <a \
href="https://postgr.es/m/20210126195224.GA20361@alvherre.pgsql" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/20210126195224.GA20361@alvherre.pgsql</a> \
Reviewed-by: Tom Lane <a style="color: #3498db; text-decoration: \
underline">tgl@sss.pgh.pa.us</a>
<a href="https://git.postgresql.org/pg/commitdiff/6819b9042fe69154ff3fd2337c5028038d3cfab4" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/6819b9042fe69154ff3fd2337c5028038d3cfab4</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Remove bogus restriction from BEFORE UPDATE triggers. In trying \
to protect the user from inconsistent behavior, commit 487e9861d0cf "Enable BEFORE \
row-level triggers for partitioned tables" tried to prevent BEFORE UPDATE FOR EACH \
ROW triggers from moving the row from one partition to another. However, it turns
out that the restriction is wrong in two ways: first, it fails spuriously,
preventing valid situations from working, as in bug #16794; and second, they
don't protect from any misbehavior, because tuple routing would cope anyway.
Fix by removing that restriction. We keep the same restriction on BEFORE
INSERT FOR EACH ROW triggers, though. It is valid and useful there. In the
future we could remove it by having tuple reroute work for inserts as it does
for updates. Backpatch to 13. Author: Álvaro Herrera
<a style="color: #3498db; text-decoration: \
underline">alvherre@alvh.no-ip.org</a> \
Reported-by: Phillip Menke <a style="color: #3498db; text-decoration: \
underline">pg@pmenke.de</a> \
Discussion: <a href="https://postgr.es/m/16794-350a655580fbb9ae@postgresql.org" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/16794-350a655580fbb9ae@postgresql.org</a> <a \
href="https://git.postgresql.org/pg/commitdiff/6f5c8a8ec23f8ab00da4d2b77bfc8af2a578c4d3" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/6f5c8a8ec23f8ab00da4d2b77bfc8af2a578c4d3</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Thomas Munro pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Retire pg_standby. pg_standby was useful more than a decade ago, but now it is \
obsolete. It has been proposed that we retire it many times. Now seems like a good \
time to finally do it, because "waiting restore commands" are incompatible with a \
proposed recovery prefetching feature. Discussion: <a \
href="https://postgr.es/m/20201029024412.GP5380%40telsasoft.com" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/20201029024412.GP5380%40telsasoft.com</a> Author: \
Justin Pryzby <a style="color: #3498db; text-decoration: \
underline">pryzby@telsasoft.com</a> \
Reviewed-by: Heikki Linnakangas <a style="color: #3498db; text-decoration: \
underline">hlinnaka@iki.fi</a> \
Reviewed-by: Peter Eisentraut <a style="color: #3498db; text-decoration: \
underline">peter.eisent \
4;aut@enterprisedb.com</a> \
Reviewed-by: Michael Paquier <a style="color: #3498db; text-decoration: \
underline">michael@paquier.xyz</a> \
Reviewed-by: Fujii Masao <a style="color: #3498db; text-decoration: \
underline">masao.fujii@o& \
#115;s.nttdata.com</a> <a \
href="https://git.postgresql.org/pg/commitdiff/514b411a2b5226167add9ab139d3a96dbe98035d" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/514b411a2b5226167add9ab139d3a96dbe98035d</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Remove documentation of waiting restore_command. Following the \
removal of pg_standby, also remove the documentation section that describes how to \
write your own "waiting restore_command" along the same lines. Discussion:
<a href="https://postgr.es/m/20201029024412.GP5380%40telsasoft.com" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/20201029024412.GP5380%40telsasoft.com</a> \
Reviewed-by: Michael Paquier <a style="color: #3498db; text-decoration: \
underline">michael@paquier.xyz</a>
<a href="https://git.postgresql.org/pg/commitdiff/5c6d184213bb9feac0800eee1a16769bec9e8d58" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/5c6d184213bb9feac0800eee1a16769bec9e8d58</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Alexander Korotkov pushed:</p> <ul style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> \
<li style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Document behavior of the <code>.**</code> jsonpath accessor in the lax mode. \
When the <code>.**</code> jsonpath accessor handles the array, it selects both array \
and each of its elements. When using lax mode, subsequent accessors automatically \
unwrap arrays. So, the content of each array element may be selected twice. Even
though this behavior is counterintuitive, it's correct because everything
works as designed. This commit documents it. Backpatch to 12 where the
jsonpath language was introduced. Reported-by: Thomas Kellerer Bug: #16828
Discussion:
<a href="https://postgr.es/m/16828-2b0229babfad2d8c%40postgresql.org" style="color: \
#3498db; text-decoration: \
underline">https://postgr.es/m/16828-2b0229babfad2d8c%40postgresql.org</a> \
Discussion: <a href="https://postgr.es/m/CAPpHfdtS-nNidT%3DEqZbAYOPcnNOWh_sd6skVdu2CAQUGdvpT8Q%40mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAPpHfdtS-nNidT%3DEqZbAYOPcnNOWh_sd6skVdu2CAQUGdvpT8Q%40mail.gmail.com</a>
Author: Alexandex Korotkov, revised by Tom Lane Reviewed-by: Alvaro Herrera,
Thomas Kellerer, Tom Lane Backpatch-through: 12
<a href="https://git.postgresql.org/pg/commitdiff/b41645460af563cfd4e4f57f354058cf69ef3b14" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/b41645460af563cfd4e4f57f354058cf69ef3b14</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fix parsing of complex morphs to tsquery. When to_tsquery() or \
websearch_to_tsquery() meet a complex morph containing multiple words residing \
adjacent position, these words are connected with OP_AND operator. That leads to \
surprising results. For instace, both websearch_to_tsquery('"pg_class pg"') and \
to_tsquery('pg_class <-> pg') produce '( pg & class ) <-> pg' \
tsquery. This tsquery requires 'pg' and 'class' words to reside on the same \
position and doesn't match to to_tsvector('pg_class pg'). It appears to be \
ridiculous behavior, which needs to be fixed. This commit makes to_tsquery() or \
websearch_to_tsquery() connect words residing adjacent position with OP_PHRASE. \
Therefore, now those words are normally chained with other OP_PHRASE operator. The \
examples of above now produces 'pg <-> class <-> pg' tsquery, which \
matches to to_tsvector('pg_class pg'). Another effect of this commit is that \
complex morph word positions now need to match the tsvector even if there is no \
surrounding OP_PHRASE. This behavior change generally looks like an improvement but \
making this commit not backpatchable. Reported-by: Barry Pederson Bug: #16592 \
Discussion: <a href="https://postgr.es/m/16592-70b110ff9731c07d@postgresql.org" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/16592-70b110ff9731c07d@postgresql.org</a> Discussion:
<a href="https://postgr.es/m/CAPpHfdv0EzVhf6CWfB1_TTZqXV_2Sn-jSY3zSd7ePH%3D-%2B1V2DQ%40mail.gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/CAPpHfdv0EzVhf6CWfB1_TTZqXV_2Sn-jSY3zSd7ePH%3D-%2B1V2DQ%40mail.gmail.com</a>
Author: Alexander Korotkov Reviewed-by: Tom Lane, Neil Chen
<a href="https://git.postgresql.org/pg/commitdiff/0c4f355c6a5fd437f71349f2f3d5d491382572b7" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/0c4f355c6a5fd437f71349f2f3d5d491382572b7</a></p>
</li>
</ul>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Noah Misch pushed:</p> <ul style="font-family: sans-serif; \
font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px"> <li \
style="list-style-position: inside; margin-left: 5px"> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: 15px">Fix \
CREATE INDEX CONCURRENTLY for simultaneous prepared transactions. In a cluster \
having used CREATE INDEX CONCURRENTLY while having enabled prepared transactions, \
queries that use the resulting index can silently fail to find rows. Fix this for \
future CREATE INDEX CONCURRENTLY by making it wait for prepared transactions like it \
waits for ordinary transactions. This expands the VirtualTransactionId structure \
domain to admit prepared transactions. It may be necessary to reindex to recover \
from past occurrences. Back-patch to 9.5 (all supported versions). Andrey Borodin, \
reviewed (in earlier versions) by Tom Lane and Michael Paquier. Discussion:
<a href="https://postgr.es/m/2E712143-97F7-4890-B470-4A35142ABC82@yandex-team.ru" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/2E712143-97F7-4890-B470-4A35142ABC82@yandex-team.ru</a>
<a href="https://git.postgresql.org/pg/commitdiff/8a54e12a38d1545d249f1402f66c8cde2837d97c" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/8a54e12a38d1545d249f1402f66c8cde2837d97c</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fix error with CREATE PUBLICATION, wal_level=minimal, and new \
tables. CREATE PUBLICATION has failed spuriously when applied to a permanent \
relation created or rewritten in the current transaction. Make the same change to \
another site having the same semantic intent; the second instance has no \
user-visible consequences. Back-patch to v13, where commit
c6b92041d38512a4176ed76ad06f713d2e6c01a8 broke this. Kyotaro Horiguchi
Discussion:
<a href="https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/360bd2321b1ad9c47466bc485ee5eb2f4515372e" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/360bd2321b1ad9c47466bc485ee5eb2f4515372e</a></p>
</li>
<li style="list-style-position: inside; margin-left: 5px">
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Revive "snapshot too old" with wal_level=minimal and SET \
TABLESPACE. Given a permanent relation rewritten in the current transaction, the
old_snapshot_threshold mechanism assumed the relation had never been subject
to early pruning. Hence, a query could fail to report "snapshot too old" when
the rewrite followed an early truncation. ALTER TABLE SET TABLESPACE is
probably the only rewrite mechanism capable of exposing this bug. REINDEX
sets indcheckxmin, avoiding the problem. CLUSTER has zeroed page LSNs since
before old_snapshot_threshold existed, so old_snapshot_threshold has never
cooperated with it. ALTER TABLE ... SET DATA TYPE makes the table look empty
to every past snapshot, which is strictly worse. Back-patch to v13, where
commit c6b92041d38512a4176ed76ad06f713d2e6c01a8 broke this. Kyotaro Horiguchi
and Noah Misch Discussion:
<a href="https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com" \
style="color: #3498db; text-decoration: \
underline">https://postgr.es/m/20210113.160705.2225256954956139776.horikyota.ntt@gmail.com</a>
<a href="https://git.postgresql.org/pg/commitdiff/7da83415e5bc01bba0093f34f4f612b70c70b678" \
style="color: #3498db; text-decoration: \
underline">https://git.postgresql.org/pg/commitdiff/7da83415e5bc01bba0093f34f4f612b70c70b678</a></p>
</li>
</ul>
<h1 style="color: #000; font-family: sans-serif; line-height: 1.4; margin: 0; \
margin-bottom: 30px; font-size: 25px; font-weight: 300; text-align: center">Pending \
Patches</h1> <p style="font-family: sans-serif; font-size: 14px; font-weight: normal; \
margin: 0; margin-bottom: 15px">Amit Langote sent in another revision of a patch to \
set ForeignScanState.resultRelInfo lazily, set ResultRelInfo.ri_PartitionRoot in all
result relations, and initialize result relation information lazily.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Takamichi Osumi sent in another revision of a patch to ensure \
that archive recovery not miss data.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Amit Langote sent in another revision of a patch to create \
foreign key triggers in partitioned tables, and enforce foreign keys correctly during \
cross-partition updates.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Pavel Borisov sent in another revision of a patch to generate \
HASH and LIST partitions automatically with a new CONFIGURATION directive.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Bucoo sent in another revision of a patch to support parallel \
union/distinct/aggregate using batch hashagg and sort.</p> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Vigneshwaran C sent in another revision of a patch to identify missing \
publications from publisher during CREATE/ALTER SUBSCRIPTION operations.</p> <p \
style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Bertrand Drouvot sent in a patch to add an --extension flag to \
pg_dump to dump specific extensions.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">David Cramer sent in two more revisions of a patch to throw an \
error and rollback on a failed transaction instead of silently rolling back.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Konstantin Knizhnik sent in two more revisions of a patch to \
have auto_explain create statistics.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Michaël Paquier and Sehrope Sarkuni traded patches to surface \
sha1 to SQL.</p> <p style="font-family: sans-serif; font-size: 14px; font-weight: \
normal; margin: 0; margin-bottom: 15px">Takayuki Tsunakawa, Iwata Aya, and Kirk \
Jamison traded patches to add tracing to libpq.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Masahiro Ikeda sent in three more revisions of a patch to add \
statistics related to write/sync wal records, and make the wal receiver report WAL \
statistics.</p> <p style="font-family: sans-serif; font-size: 14px; font-weight: \
normal; margin: 0; margin-bottom: 15px">Masahiko Sawada sent in two more revisions of \
a patch to choose vacuum strategy before heap and index vacuums, skip btree \
bulkdelete if the index doesn't grow, and disable index cleanup for anti-wraparound \
and aggressive vacuums.</p> <p style="font-family: sans-serif; font-size: 14px; \
font-weight: normal; margin: 0; margin-bottom: 15px">Peter Smith and Amit Kapila \
traded patches to enable tablesync workers to use multiple transactions apiece.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Bruce Momjian sent in four more revisions of a patch to \
implement key management.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Dilip Kumar sent in three more revisions of a patch to provide a \
new interface to get the recovery pause status.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">David Rowley sent in another revision of a patch to add TID \
Range Scans to support efficient scanning ranges of TIDs.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Bertrand Drouvot sent in another revision of a patch to \
implement minimal logical decoding on standbys.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Surafel Temesgen sent in another revision of a patch to \
implement system-versioned temporal tables.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Jim Finnerty sent in another revision of a patch to implement \
64-bit XIDs.</p> <p style="font-family: sans-serif; font-size: 14px; font-weight: \
normal; margin: 0; margin-bottom: 15px">Amit Langote sent in another revision of a \
patch to overhaul how updates compute new tuples, and revise how inherited updates \
and deletes are handled.</p> <p style="font-family: sans-serif; font-size: 14px; \
font-weight: normal; margin: 0; margin-bottom: 15px">Li Japin sent in a patch to fix \
the documentation of ALTER SUBSCRIPTION ... SET PUBLICATION.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Daniel Gustafsson sent in another revision of a patch to support \
checksum enable/disable in a running cluster.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Takashi Menjo sent in another revision of a patch to make it \
possible to use NVRAM for WAL buffers.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Masahiko Sawada sent in another revision of a patch to implement \
transactions involving multiple postgres foreign servers.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Dean Rasheed sent in another revision of a patch to make it \
possible to create and use extended statistics on expressions.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Amit Langote sent in another revision of a patch to export \
get_partition_for_tuple() and avoid using SPI for some RI checks.</p> <p \
style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Denis Laxalde sent in a patch to Set default transactions to \
read-only at server start in pg_upgrade.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Amit Khandekar sent in another revision of a patch to speed up \
xor'ing of two gist index signatures for tsvectors by using popcount64() on chunks \
instead of xor'ing char values, and avoid function pointer dereferencing for
pg_popcount32/64() calls.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Konstantin Knizhnik sent in another revision of a patch to \
extending the auto_explain extension to generate extended statistics in case of bad
selectivity estimation, and take into account extended statistics when computing
join selectivity.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Michail Nikolaev sent in another revision of a patch to add full \
support for index LP_DEAD hint bits on standby.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Nathan Bossart sent in another revision of a patch to Avoid \
marking segments as ready-for-archival too early.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Anastasia Lubennikova sent in another revision of a patch \
intended to fix a bug that manifested as pg_upgrade fails with non-standard ACL.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Paul Guo sent in a patch to speed up pg_rewind by fsync()ing \
only the affected files/directories, and using copy_file_range() for file \
copying.</p> <p style="font-family: sans-serif; font-size: 14px; font-weight: normal; \
margin: 0; margin-bottom: 15px">Kyotaro HORIGUCHI and Heikki Linnakangas traded \
patches to add catcache expiration.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Heikki Linnakangas and Daniel Gustafsson traded patches to \
support enabling/disabling page checksums in a running instance.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Mark Rofail sent in two more revisions of a patch to implement \
foreign key arrays.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Vigneshwaran C sent in three more revisions of a patch to make \
it possible to print a backtrace of any postgres process using a new function
pg_print_backtrace() that is part of the instance in which the function is
called.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Michaël Paquier and Daniel Gustafsson traded patches to make it \
possible to use NSS as a TLS provider for libpq.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Bharath Rupireddy sent in two more revisions of a patch to avoid \
catalogue accesses in slot_store_error_callback and conversion_error_callback.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Takamichi Osumi sent in a patch to log when/where wal_level is \
changed to 'minimal' from upper level to invalidate the old backups or make alerts to
users.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Amul Sul sent in another revision of a patch to implement wal \
prohibit state using global barrier.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Greg Nancarrow sent in another revision of a patch to make it \
possible to use parallel selects in INSERT ... SELECT, and parallel INSERT or SELECT \
where it's safe to do so.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Justin Pryzby sent in another revision of a patch to make it \
possible to run CREATE INDEX CONCURRENTLY on a partitioned table.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Andrew Dunstan sent in two revisions of a patch to implement \
PostgresNodePath, a subclass of PostgresNode which runs commands in the context of a \
given PostgreSQL install path.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">David Rowley sent in another revision of a patch to add a Result \
Cache executor node.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Nathan Bossart sent in two more revisions of a patch to add a \
PROCESS_TOAST option to VACUUM.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Paul Martinez sent in a patch to clarify the messages that occur \
when a replication connection is not allowed.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Jacob Champion sent in a patch to save a role's original \
authenticated identity for logging.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Mark Dilger sent in three more revisions of a patch to implement \
a new pg_amcheck contrib extension.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Hou Zhijie sent in another revision of a patch to add a new GUC \
and corresponding table option enable_parallel_dml.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Marco Montagna sent in a patch to reduce likelihood of fdw \
prepared statement collisions by prepending a random number to the names generated \
for same.</p> <p style="font-family: sans-serif; font-size: 14px; font-weight: \
normal; margin: 0; margin-bottom: 15px">Peter Eisentraut sent in another revision of \
a patch to pg_dump to fix dumping of inherited generated columns.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Antonin Houska sent in another revision of a patch to clean up \
orphaned files using undo logs.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Dilip Kumar and Justin Pryzby traded patches to implement custom \
table compression methods.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Álvaro Herrera sent in another revision of a patch to add \
pg_atomic_monotonic_advance_u64, use atomics in xlog.c, and add barriers.</p> <p \
style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Fabien COELHO sent in another revision of a patch to add \
barriers to pgbench.</p> <p style="font-family: sans-serif; font-size: 14px; \
font-weight: normal; margin: 0; margin-bottom: 15px">Thomas Munro sent in two \
revisions of a patch to use a global barrier to fix DROP TABLESPACE on Windows, and \
use condition variables for ProcSignalBarriers.</p> <p style="font-family: \
sans-serif; font-size: 14px; font-weight: normal; margin: 0; margin-bottom: \
15px">Andrew Dunstan sent in another revision of a patch to allow matching whole DN \
from a client certificate.</p> <p style="font-family: sans-serif; font-size: 14px; \
font-weight: normal; margin: 0; margin-bottom: 15px">Thomas Munro sent in a patch to \
hold interrupts while running dsm_detach() callbacks.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Joel Jacobson sent in a patch to improve the documentation of \
oid columns that can be zero in system views.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Vigneshwaran C sent in another revision of a patch to add schema \
support for CREATE PUBLICATION.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Tom Lane sent in a patch to record foreign key relationships for \
the system catalogs.</p>
<p style="font-family: sans-serif; font-size: 14px; font-weight: normal; margin: 0; \
margin-bottom: 15px">Euler Taveira de Oliveira sent in another revision of a patch to \
implement row filtering for logical replication.</p>
</td>
</tr>
</table>
</td>
</tr>
</table>
<div class="footer" style="clear: both; Margin-top: 10px; text-align: \
center; width: 100%;">
<table border="0" cellpadding="0" cellspacing="0" \
style="border-collapse: separate; mso-table-lspace: 0pt; mso-table-rspace: 0pt; \
width: 100%;"> <tr>
<td class="content-block" style="font-family: sans-serif; \
vertical-align: top; padding-bottom: 10px; padding-top: 10px; font-size: 12px; color: \
#999999; text-align: center;">
<span class="apple-link" style="color: #999999; font-size: 12px; \
text-align: center;"> This email was sent to you from PWN. It was delivered on their \
behalf by the PostgreSQL project. Any questions about the content of the message \
should be sent to PWN.
</span>
<br><br>
You were sent this email as a subscriber of the <em>pgsql-announce</em> mailinglist, \
for the content tag PWN.
To unsubscribe from
further emails, or change which emails you want to receive, please click the personal \
unsubscribe link that you can find in the headers of this email, or visit
<a href="https://lists.postgresql.org/unsubscribe/" style="color: #3498db; \
text-decoration: underline">https://lists.postgresql.org/unsubscribe/</a>.
</td>
</tr>
</table>
</div>
</div>
</td>
<td style="font-family: sans-serif; font-size: 14px; vertical-align: \
top;"> </td> </tr>
</table>
</body>
</html>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic