Amit Kapila [Tue, 12 Sep 2023 04:53:17 +0000 (10:23 +0530)]
Fix uninitialized access to InitialRunningXacts during decoding after ERROR.
The transactions and subtransactions array that was allocated under
snapshot builder memory context and recorded during decoding was not
cleared in case of errors. This can result in an assertion failure if we
attempt to retry logical decoding within the same session. To address this
issue, we register a callback function under the snapshot builder memory
context to clear the recorded transactions and subtransactions array along
with the context.
This problem doesn't exist in PG16 and HEAD as instead of using
InitialRunningXacts, we added the list of transaction IDs and
sub-transaction IDs, that have modified catalogs and are running during
snapshot serialization, to the serialized snapshot (see commit 7f13ac8123).
Author: Hou Zhijie Reviewed-by: Amit Kapila
Backpatch-through: 11
Discussion: http://postgr.es/m/18055-ab3beed9f4b7b7d6@postgresql.org
The new test added by commit 68a59f9e9 disables the subscription and
manually drops the associated replication slot. However, since
disabling the subsubscription doesn't wait for a walsender to release
the replication slot and exit, pg_drop_replication_slot() could
fail. Avoid failure by adding a wait for the replication slot to
become inactive.
Reported-by: Hou Zhijie, as per buildfarm Reviewed-by: Hou Zhijie
Discussion: https://postgr.es/m/OS0PR01MB571682316378379AA34854F694E9A%40OS0PR01MB5716.jpnprd01.prod.outlook.com
Backpatch-through: 15
Extend the PG_TEST_EXTRA documentation to mention resource intensive
tests as well. The previous wording only mentioned special software
and security in the main paragraph, with resource usage listed on one
of the tests in the list.
Backpatch to v15 where f47ed79cc8 added wal_consistenct_checking as
a PG_TEST_EXTRA target.
Author: Nazir Bilal Yavuz <byavuz81@gmail.com>
Discussion: https://postgr.es/m/CAN55FZ0OthTuBdiNkaX2BvxuHdK4Y1MVEb8_uEuD1yHMPmT9Og@mail.gmail.com
Backpatch-through: 15
Michael Paquier [Thu, 7 Sep 2023 05:12:29 +0000 (14:12 +0900)]
pg_basebackup: Generate valid temporary slot names under PQbackendPID()
pgbouncer can cause PQbackendPID() to return negative values due to it
filling be_pid with random bytes (even these days pid_max can only be
set up to 2^22 on 64b machines on Linux, for example, so this cannot
happen with normal PID numbers). When this happens, pg_basebackup may
generate a temporary slot name that may not be accepted by the parser,
leading to spurious failures, like:
pg_basebackup: error: could not send replication command
ERROR: replication slot name "pg_basebackup_-1201966863" contains
invalid character
This commit fixes that problem by formatting the result from
PQbackendPID() as an unsigned integer when creating the temporary
replication slot name, so as the invalid character is gone and the
command can be parsed.
Thomas Munro [Wed, 6 Sep 2023 23:47:42 +0000 (11:47 +1200)]
Disable 031_recovery_conflict.pl in 15 and 16.
This test fails due to known bugs in the test and the server. Those
will be fixed in master shortly and possibly back-patched a bit later,
but in the meantime it is unhelpful for package maintainers if the tests
randomly fail, and it's not a good time to make complex changes in 16.
This had already been done for older branches prior to 15's release.
Now we're about to release 16, and Debian's test builds are regularly
failing on one architecture, so let's do the same for 15 and 16.
Reported-by: Christoph Berg <myon@debian.org> Reported-by: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Discussion: https://postgr.es/m/CALj2ACVr8au2J_9D88UfRCi0JdWhyQDDxAcSVav0B0irx9nXEg%40mail.gmail.com
Michael Paquier [Mon, 4 Sep 2023 05:55:51 +0000 (14:55 +0900)]
Fix out-of-bound read in gtsvector_picksplit()
This could lead to an imprecise choice when splitting an index page of a
GiST index on a tsvector, deciding which entries should remain on the
old page and which entries should move to a new page.
This is wrong since tsearch2 has been moved into core with commit 140d4ebcb46e, so backpatch all the way down. This error has been
spotted by valgrind.
Author: Alexander Lakhin
Discussion: https://postgr.es/m/17950-6c80a8d2b94ec695@postgresql.org
Backpatch-through: 11
Michael Paquier [Sun, 3 Sep 2023 23:04:43 +0000 (08:04 +0900)]
Fix handling of shared statistics with dropped databases
Dropping a database while a connection is attempted on it was able to
lead to the presence of valid database entries in shared statistics.
The issue is that MyDatabaseId was getting set too early than it should,
as, if the connection attempted on the dropped database fails when
renamed or dropped, the shutdown callback of the shared statistics would
finish by re-inserting a correct entry related to the database already
dropped.
As analyzed by the bug reporters, this issue could lead to phantom
entries in the database list maintained by the autovacuum launcher
(in rebuild_database_list()) if the database dropped was part of the
database list when it was still valid. After the database was dropped,
it would remain the highest on the list of databases to considered by
the autovacuum worker as things to process. This would prevent
autovacuum jobs to happen on all the other databases still present.
The commit fixes this issue by delaying setting MyDatabaseId until the
database existence has been re-checked with the second scan on
pg_database after getting a shared lock on it, and by switching
pgstat_update_dbstats() so as nothing happens if MyDatabaseId is not
valid.
Issue introduced by 5891c7a8ed8f, so backpatch down to 15.
Reported-by: Will Mortensen, Jacob Speidel Analyzed-by: Will Mortensen, Jacob Speidel
Author: Andres Freund
Discussion: https://postgr.es/m/17973-bca1f7d5c14f601e@postgresql.org
Backpatch-through: 15
Michael Paquier [Tue, 29 Aug 2023 23:03:51 +0000 (08:03 +0900)]
Avoid possible overflow with ltsGetFreeBlock() in logtape.c
nFreeBlocks, defined as a long, stores the number of free blocks in a
logical tape. ltsGetFreeBlock() has been using an int to store the
value of nFreeBlocks, which could lead to overflows on platforms where
long and int are not the same size (in short everything except Windows
where long is 4 bytes).
The problematic intermediate variable is switched to be a long instead
of an int.
Issue introduced by c02fdc9223015, so backpatch down to 13.
Author: Ranier vilela Reviewed-by: Peter Geoghegan, David Rowley
Discussion: https://postgr.es/m/CAEudQApLDWCBR_xmwNjGBrDo+f+S4E87x3s7-+hoaKqYdtC4JQ@mail.gmail.com
Backpatch-through: 13
After commit b0bea38705, syslogger prints 63 warnings about failing to
close a listen socket at postmaster startup. That's because the
syslogger process forks before the ListenSockets array is initialized,
so ClosePostmasterPorts() calls "close(0)" 64 times. The first call
succeeds, because fd 0 is stdin.
This has been like this since commit 9a86f03b4e in version 13, which
moved the SysLogger_Start() call to before initializing ListenSockets.
We just didn't notice until commit b0bea38705 added the LOG message.
Reported by Michael Paquier and Jeff Janes.
Author: Michael Paquier
Discussion: https://www.postgresql.org/message-id/ZOvvuQe0rdj2slA9%40paquier.xyz
Discussion: https://www.postgresql.org/message-id/ZO0fgDwVw2SUJiZx@paquier.xyz#482670177eb4eaf4c9f03c1eed963e5f
Backpatch-through: 13
Tom Lane [Thu, 24 Aug 2023 16:02:40 +0000 (12:02 -0400)]
Avoid unnecessary plancache revalidation of utility statements.
Revalidation of a plancache entry (after a cache invalidation event)
requires acquiring a snapshot. Normally that is harmless, but not
if the cached statement is one that needs to run without acquiring a
snapshot. We were already aware of that for TransactionStmts,
but for some reason hadn't extrapolated to the other statements that
PlannedStmtRequiresSnapshot() knows mustn't set a snapshot. This can
lead to unexpected failures of commands such as SET TRANSACTION
ISOLATION LEVEL. We can fix it in the same way, by excluding those
command types from revalidation.
However, we can do even better than that: there is no need to
revalidate for any statement type for which parse analysis, rewrite,
and plan steps do nothing interesting, which is nearly all utility
commands. To mechanize this, invent a parser function
stmt_requires_parse_analysis() that tells whether parse analysis does
anything beyond wrapping a CMD_UTILITY Query around the raw parse
tree. If that's what it does, then rewrite and plan will just
skip the Query, so that it is not possible for the same raw parse
tree to produce a different plan tree after cache invalidation.
stmt_requires_parse_analysis() is basically equivalent to the
existing function analyze_requires_snapshot(), except that for
obscure reasons that function omits ReturnStmt and CallStmt.
It is unclear whether those were oversights or intentional.
I have not been able to demonstrate a bug from not acquiring a
snapshot while analyzing these commands, but at best it seems mighty
fragile. It seems safer to acquire a snapshot for parse analysis of
these commands too, which allows making stmt_requires_parse_analysis
and analyze_requires_snapshot equivalent.
In passing this fixes a second bug, which is that ResetPlanCache
would exclude ReturnStmts and CallStmts from revalidation.
That's surely *not* safe, since they contain parsable expressions.
Per bug #18059 from Pavel Kulakov. Back-patch to all supported
branches.
Andres Freund [Wed, 23 Aug 2023 22:15:29 +0000 (15:15 -0700)]
ci: Make compute resources for CI configurable
See prior commit for an explanation for the goal of the change and why it had
to be split into two commits.
Reviewed-by: Daniel Gustafsson <daniel@yesql.se> Reviewed-by: Nazir Bilal Yavuz <byavuz81@gmail.com>
Discussion: https://postgr.es/m/20230808021541.7lbzdefvma7qmn3w@awork3.anarazel.de
Backpatch: 15-, where CI support was added
Andres Freund [Wed, 23 Aug 2023 22:15:29 +0000 (15:15 -0700)]
ci: Prepare to make compute resources for CI configurable
cirrus-ci will soon restrict the amount of free resources every user gets (as
have many other CI providers). For most users of CI that should not be an
issue. But e.g. for cfbot it will be an issue.
To allow configuring different resources on a per-repository basis, introduce
infrastructure for overriding the task execution environment. Unfortunately
this is not entirely trivial, as yaml anchors have to be defined before their
use, and cirrus-ci only allows injecting additional contents at the end of
.cirrus.yml.
To deal with that, move the definition of the CI tasks to
.cirrus.tasks.yml. The main .cirrus.yml is loaded first, then, if defined, the
file referenced by the REPO_CI_CONFIG_GIT_URL variable, will be added,
followed by the contents of .cirrus.tasks.yml. That allows
REPO_CI_CONFIG_GIT_URL to override the yaml anchors defined in .cirrus.yml.
Unfortunately git's default merge / rebase strategy does not handle copied
files, just renamed ones. To avoid painful rebasing over this change, this
commit just renames .cirrus.yml to .cirrus.tasks.yml, without adding a new
.cirrus.yml. That's done in the followup commit, which moves the relevant
portion of .cirrus.tasks.yml to .cirrus.yml. Until that is done,
REPO_CI_CONFIG_GIT_URL does not fully work.
The subsequent commit adds documentation for how to configure custom compute
resources to src/tools/ci/README
Reviewed-by: Daniel Gustafsson <daniel@yesql.se> Reviewed-by: Nazir Bilal Yavuz <byavuz81@gmail.com>
Discussion: https://postgr.es/m/20230808021541.7lbzdefvma7qmn3w@awork3.anarazel.de
Backpatch: 15-, where CI support was added
Andres Freund [Wed, 23 Aug 2023 19:30:38 +0000 (12:30 -0700)]
ci: Use VMs for SanityCheck and CompilerWarnings
The main reason for this change is to reduce different ways of executing
tasks, making it easier to use custom compute resources for cfbot. A secondary
benefit is that the tasks seem slightly faster this way, apparently the
increased startup overhead is outweighed by reduced runtime overhead.
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Discussion: https://postgr.es/m/20230808021541.7lbzdefvma7qmn3w@awork3.anarazel.de
Backpatch: 15-, where CI support was added
Andres Freund [Wed, 23 Aug 2023 19:30:23 +0000 (12:30 -0700)]
ci: Move execution method of tasks into yaml templates
This is done in preparation for making the compute resources for CI
configurable. It also looks cleaner.
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Discussion: https://postgr.es/m/20230808021541.7lbzdefvma7qmn3w@awork3.anarazel.de
Backpatch: 15-, where CI support was added
Andres Freund [Wed, 23 Aug 2023 19:30:11 +0000 (12:30 -0700)]
ci: Don't specify amount of memory
The number of CPUs is the cost-determining factor. Most instance types that
run tests have more memory/core than what we specified, there's no real
benefit in wasting that.
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Discussion: https://postgr.es/m/20230808021541.7lbzdefvma7qmn3w@awork3.anarazel.de
Backpatch: 15-, where CI support was added
doc: Replace list of drivers and PLs with wiki link
The list of external language drivers and procedural languages was
never complete or exhaustive, and rather than attempting to manage
it the content has migrated to the wiki. This replaces the tables
altogether with links to the wiki as we regularly get requests for
adding various projects, which we reject without any clear policy
for why or how the content should be managed.
The threads linked to below are the most recent discussions about
this, the archives contain many more.
Backpatch to all supported branches since the list on the wiki
applies to all branches.
Andrew Dunstan [Tue, 22 Aug 2023 15:57:08 +0000 (11:57 -0400)]
Cache by-reference missing values in a long lived context
Attribute missing values might be needed past the lifetime of the tuple
descriptors from which they are extracted. To avoid possibly using
pointers for by-reference values which might thus be left dangling, we
cache a datumCopy'd version of the datum in the TopMemoryContext. Since
we first search for the value this only needs to be done once per
session for any such value.
Original complaint from Tom Lane, idea for mitigation by Andrew Dunstan,
tweaked by Tom Lane.
Backpatch to version 11 where missing values were introduced.
Michael Paquier [Mon, 21 Aug 2023 04:33:08 +0000 (13:33 +0900)]
Fix pg_stat_reset_single_table_counters() for shared relations
This commit fixes the function of $subject for shared relations. This
feature has been added by e042678. Unfortunately, this new behavior got
removed by 5891c7a when moving statistics to shared memory.
Andres Freund [Sat, 19 Aug 2023 19:40:45 +0000 (12:40 -0700)]
ci: macos: use cached macports install
A significant chunk of the time on the macos CI task is spent installing
packages using homebrew. The downloads of the packages are cached, but the
installation needs to happen every time. We can't cache the whole homebrew
installation, because it is too large due to pre-installed packages.
Speed this up by installing packages using macports and caching the
installation as .dmg. That's a lot faster than unpacking a tarball.
In addition, don't install llvm - it wasn't enabled when building, so it's
just a waste of time/space.
This substantially speeds up the mac CI time, both in the cold cache and in
the warm cache case (the latter from ~1m20s to ~5s).
It doesn't seem great to have diverging sources of packages for CI between
branches, so backpatch to 15 (where CI was added).
Discussion: https://postgr.es/m/20230805202539.r3umyamsnctysdc7@awork3.anarazel.de
Backpatch: 15-, where CI was added
The fix itself is fine, but the test revealed other problems related
to parallel query that are not easily fixable. Remove the test for
now to fix the buildfarm.
Noah Misch [Mon, 7 Aug 2023 13:05:56 +0000 (06:05 -0700)]
Reject substituting extension schemas or owners matching ["$'\].
Substituting such values in extension scripts facilitated SQL injection
when @extowner@, @extschema@, or @extschema:...@ appeared inside a
quoting construct (dollar quoting, '', or ""). No bundled extension was
vulnerable. Vulnerable uses do appear in a documentation example and in
non-bundled extensions. Hence, the attack prerequisite was an
administrator having installed files of a vulnerable, trusted,
non-bundled extension. Subject to that prerequisite, this enabled an
attacker having database-level CREATE privilege to execute arbitrary
code as the bootstrap superuser. By blocking this attack in the core
server, there's no need to modify individual extensions. Back-patch to
v11 (all supported versions).
Reported by Micah Gate, Valerie Woolard, Tim Carey-Smith, and Christoph
Berg.
David Rowley [Mon, 7 Aug 2023 10:15:23 +0000 (22:15 +1200)]
Don't Memoize lateral joins with volatile join conditions
The use of Memoize was already disabled in normal joins when the join
conditions had volatile functions per the code in
match_opclause_to_indexcol(). Ordinarily, the parameterization for the
inner side of a nested loop will be an Index Scan or at least eventually
lead to an index scan (perhaps nested several joins deep). However, for
lateral joins, that's not the case and seq scans can be parameterized
too, so we can't rely on match_opclause_to_indexcol().
Here we explicitly check the parameterization for volatile functions and
don't consider the generation of a Memoize path when such functions
are present.
Author: Richard Guo
Discussion: https://postgr.es/m/CAMbWs49nHFnHbpepLsv_yF3qkpCS4BdB-v8HoJVv8_=Oat0u_w@mail.gmail.com
Backpatch-through: 14, where Memoize was introduced
Dean Rasheed [Mon, 7 Aug 2023 08:24:27 +0000 (09:24 +0100)]
Fix RLS policy usage in MERGE.
If MERGE executes an UPDATE action on a table with row-level security,
the code incorrectly applied the WITH CHECK clauses from the target
table's INSERT policies to new rows, instead of the clauses from the
table's UPDATE policies. In addition, it failed to check new rows
against the target table's SELECT policies, if SELECT permissions were
required (likely to always be the case).
In addition, if MERGE executes a DO NOTHING action for matched rows,
the code incorrectly applied the USING clauses from the target table's
DELETE policies to existing target tuples. These policies were applied
as checks that would throw an error, if they did not pass.
Fix this, so that a MERGE UPDATE action applies the same RLS policies
as a plain UPDATE query with a WHERE clause, and a DO NOTHING action
does not apply any RLS checks (other than adding clauses from SELECT
policies to the join).
Reported-by: Alexander Kozhemyakin
Bug: #18035
Discussion: https://postgr.es/m/18035-64af5cdcb5adf2a9@postgresql.org
Backpatch-through: 12, where SQL/JSON path was added.
Disallow replacing joins with scans in problematic cases.
Commit e7cb7ee14, which introduced the infrastructure for FDWs and
custom scan providers to replace joins with scans, failed to add support
handling of pseudoconstant quals assigned to replaced joins in
createplan.c, leading to an incorrect plan without a gating Result node
when postgres_fdw replaced a join with such a qual.
To fix, we could add the support by 1) modifying the ForeignPath and
CustomPath structs to store the list of RestrictInfo nodes to apply to
the join, as in JoinPaths, if they represent foreign and custom scans
replacing a join with a scan, and by 2) modifying create_scan_plan() in
createplan.c to use that list in that case, instead of the
baserestrictinfo list, to get pseudoconstant quals assigned to the join;
but #1 would cause an ABI break. So fix by modifying the infrastructure
to just disallow replacing joins with such quals.
Back-patch to all supported branches.
Reported by Nishant Sharma. Patch by me, reviewed by Nishant Sharma and
Richard Guo.
Tom Lane [Thu, 27 Jul 2023 16:07:48 +0000 (12:07 -0400)]
Raise fixed token-length limit in hba.c.
Historically, hba.c limited tokens in the authentication configuration
files (pg_hba.conf and pg_ident.conf) to less than 256 bytes. We have
seen a few reports of this limit causing problems; notably, for
moderately-complex LDAP configurations. Increase the limit to 10240
bytes as a low-risk stop-gap solution.
In v13 and earlier, this also requires raising MAX_LINE, the limit
on overall line length. I'm hesitant to make this code consume
too much stack space, so I only raised that to 20480 bytes.
Amit Kapila [Tue, 25 Jul 2023 03:20:37 +0000 (08:50 +0530)]
Fix the display of UNKNOWN message type in apply worker.
We include the message type while displaying an error context in the
apply worker. Now, while retrieving the message type string if the
message type is unknown we throw an error that will hide the original
error. So, instead, we need to simply return the string indicating an
unknown message type.
Some of the test_decoding test output was extremely wide, because it
deals with massive toasted values, and the aligned mode causes psql to
produce 200kB of whitespace and dashes. Change to unaligned mode
temporarily to avoid that behavior.
Tom Lane [Thu, 20 Jul 2023 18:23:46 +0000 (14:23 -0400)]
Guard against null plan pointer in CachedPlanIsSimplyValid().
If both the passed-in plan pointer and plansource->gplan are
NULL, CachedPlanIsSimplyValid would think that the plan pointer
is possibly-valid and try to dereference it. For the one extant
call site in plpgsql, this situation doesn't normally happen
which is why we've not noticed. However, it appears to be possible
if the previous use of the cached plan failed, as per report from
Justin Pryzby. Add an extra check to prevent crashing.
Back-patch to v13 where this code was added.
Tom Lane [Wed, 19 Jul 2023 15:00:34 +0000 (11:00 -0400)]
Doc: improve description of IN and row-constructor comparisons.
IN and NOT IN work fine on records and arrays, so just say that
they accept "expressions" not "scalar expressions". I think that
that phrasing was meant to say that they don't work on set-returning
expressions, but that's not the common meaning of "scalar".
Revise the description of row-constructor comparisons to make it
perhaps a bit less confusing. (This partially reverts some
dubious wording changes made by commit f56651519.)
Per gripe from Ilya Nenashev. Back-patch to supported branches.
In HEAD and v16, also drop a NOTE about pre-8.2 behavior, which
is hopefully no longer of interest to anybody.
Tom Lane [Tue, 18 Jul 2023 15:59:39 +0000 (11:59 -0400)]
Doc: fix out-of-date example of SPI usage.
The "count" argument of SPI_exec() only limits execution when
the query is actually returning rows. This was not the case
before PG 9.0, so this example was correct when written; but
we missed updating it in commit 2ddc600f8. Extend the example
to show the behavior both with and without RETURNING.
While here, improve the commentary and markup for the rest
of the example.
David G. Johnston and Tom Lane, per report from Curt Kolovson.
Back-patch to all supported branches.
Michael Paquier [Tue, 18 Jul 2023 05:04:48 +0000 (14:04 +0900)]
Fix indentation in twophase.c
This has been missed in cb0cca1, noticed before buildfarm member koel
has been able to complain while poking at a different patch. Like the
other commit, backpatch all the way down to limit the odds of merge
conflicts.
Michael Paquier [Tue, 18 Jul 2023 04:44:29 +0000 (13:44 +0900)]
Fix recovery of 2PC transaction during crash recovery
A crash in the middle of a checkpoint with some two-phase state data
already flushed to disk by this checkpoint could cause a follow-up crash
recovery to recover twice the same transaction, once from what has been
found in pg_twophase/ at the beginning of recovery and a second time
when replaying its corresponding record.
This would lead to FATAL failures in the startup process during
recovery, where the same transaction would have a state recovered twice
instead of once:
LOG: recovering prepared transaction 731 from shared memory
LOG: recovering prepared transaction 731 from shared memory
FATAL: lock ExclusiveLock on object 731/0/0 is already held
This issue is fixed by skipping the addition of any 2PC state coming
from a record whose equivalent 2PC state file has already been loaded in
TwoPhaseState at the beginning of recovery by restoreTwoPhaseData(),
which is OK as long as the system has not reached a consistent state.
The timing to get a messed up recovery processing is very racy, and
would very unlikely happen. The thread that has reported the issue has
demonstrated the bug using injection points to force a PANIC in the
middle of a checkpoint.
Issue introduced in 728bd99, so backpatch all the way down.
Michael Paquier [Fri, 14 Jul 2023 02:16:06 +0000 (11:16 +0900)]
Add indisreplident to fields refreshed by RelationReloadIndexInfo()
RelationReloadIndexInfo() is a fast-path used for index reloads in the
relation cache, and it has always forgotten about updating
indisreplident, which is something that would happen after an index is
selected for a replica identity. This can lead to incorrect cache
information provided when executing a command in a transaction context
that updates indisreplident.
None of the code paths currently on HEAD that need to check upon
pg_index.indisreplident fetch its value from the relation cache, always
relying on a fresh copy on the syscache. Unfortunately, this may not be
the case of out-of-core code, that could see out-of-date value.
Author: Shruthi Gowda Reviewed-by: Robert Haas, Dilip Kumar, Michael Paquier
Discussion: https://postgr.es/m/CAASxf_PBcxax0wW-3gErUyftZ0XrCs3Lrpuhq4-Z3Fak1DoW7Q@mail.gmail.com
Backpatch-through: 11
Michael Paquier [Fri, 14 Jul 2023 01:13:15 +0000 (10:13 +0900)]
Fix updates of indisvalid for partitioned indexes
indisvalid is switched to true for partitioned indexes when all its
partitions have valid indexes when attaching a new partition, up to the
top-most parent if all its leaves are themselves valid when dealing with
multiple layers of partitions.
The copy of the tuple from pg_index used to switch indisvalid to true
came from the relation cache, which is incorrect. Particularly, in the
case reported by Shruthi Gowda, executing a series of commands in a
single transaction would cause the validation of partitioned indexes to
use an incorrect version of a pg_index tuple, as indexes are reloaded
after an invalidation request with RelationReloadIndexInfo(), a much
faster version than a full index cache rebuild. In this case, the
limited information updated in the cache leads to an incorrect version
of the tuple used. One of the symptoms reported was the following
error, with a replica identity update, for instance:
"ERROR: attempted to update invisible tuple"
This is incorrect since 8b08f7d, so backpatch all the way down.
Andres Freund [Thu, 13 Jul 2023 20:03:31 +0000 (13:03 -0700)]
Handle DROP DATABASE getting interrupted
Until now, when DROP DATABASE got interrupted in the wrong moment, the removal
of the pg_database row would also roll back, even though some irreversible
steps have already been taken. E.g. DropDatabaseBuffers() might have thrown
out dirty buffers, or files could have been unlinked. But we continued to
allow connections to such a corrupted database.
To fix this, mark databases invalid with an in-place update, just before
starting to perform irreversible steps. As we can't add a new column in the
back branches, we use pg_database.datconnlimit = -2 for this purpose.
An invalid database cannot be connected to anymore, but can still be
dropped.
Unfortunately we can't easily add output to psql's \l to indicate that some
database is invalid, it doesn't fit in any of the existing columns.
Add tests verifying that a interrupted DROP DATABASE is handled correctly in
the backend and in various tools.
Reported-by: Evgeny Morozov <postgresql3@realityexists.net>
Author: Andres Freund <andres@anarazel.de> Reviewed-by: Daniel Gustafsson <daniel@yesql.se> Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Discussion: https://postgr.es/m/20230509004637.cgvmfwrbht7xm7p6@awork3.anarazel.de
Discussion: https://postgr.es/m/20230314174521.74jl6ffqsee5mtug@awork3.anarazel.de
Backpatch: 11-, bug present in all supported versions
Andres Freund [Thu, 13 Jul 2023 20:03:31 +0000 (13:03 -0700)]
Release lock after encountering bogs row in vac_truncate_clog()
When vac_truncate_clog() encounters bogus datfrozenxid / datminmxid values, it
returns early. Unfortunately, until now, it did not release
WrapLimitsVacuumLock. If the backend later tries to acquire
WrapLimitsVacuumLock, the session / autovacuum worker hangs in an
uncancellable way. Similarly, other sessions will hang waiting for the
lock. However, if the backend holding the lock exited or errored out for some
reason, the lock was released.
The bug was introduced as a side effect of 566372b3d643.
It is interesting that there are no production reports of this problem. That
is likely due to a mix of bugs leading to bogus values having gotten less
common, process exit releasing locks and instances of hangs being hard to
debug for "normal" users.
Tom Lane [Thu, 13 Jul 2023 17:07:51 +0000 (13:07 -0400)]
Remove unnecessary pfree() in g_intbig_compress().
GiST compress functions (like all GiST opclass functions) are
supposed to be called in short-lived memory contexts, so that
minor memory leaks in them are not of concern, and indeed
explicit pfree's are likely slightly counterproductive.
But this one in g_intbig_compress() is more than
slightly counterproductive, because it's guarded by
"if (in != DatumGetArrayTypeP(entry->key))" which means
that if this test succeeds, we've detoasted the datum twice.
(And to add insult to injury, the extra detoast result is
leaked.) Let's just drop the whole stanza, relying on the
GiST temporary context mechanism to clean up in good time.
The analogous bit in g_int_compress() is
if (r != (ArrayType *) DatumGetPointer(entry->key))
pfree(r);
which doesn't have the gratuitous-detoast problem so
I left it alone. Perhaps there is a case for removing
unnecessary pfree's more widely, but I'm not sure if it's
worth the code churn.
The potential extra decompress seems expensive enough to
justify calling this a (minor) performance bug and
back-patching.
Konstantin Knizhnik, Matthias van de Meent, Tom Lane
Tom Lane [Mon, 10 Jul 2023 16:14:34 +0000 (12:14 -0400)]
Be more rigorous about local variables in PostgresMain().
Since PostgresMain calls sigsetjmp, any local variables that are not
marked "volatile" have a risk of unspecified behavior. In practice
this means that when control returns via longjmp, such variables might
get reset to their values as of the time of sigsetjmp, depending on
whether the compiler chose to put them in registers or on the stack.
We were careful about this for "send_ready_for_query", but not the
other local variables.
In the case of the timeout_enabled flags, resetting them to
their initial "false" states is actually good, since we do
"disable_all_timeouts()" in the longjmp cleanup code path. If that
does not happen, we risk uselessly calling "disable_timeout()" later,
which is harmless but a little bit expensive. Let's explicitly reset
these flags so that the behavior is correct and platform-independent.
(This change means that we really don't need the new "volatile"
markings after all, but let's install them anyway since any change
in this logic could re-introduce a problem.)
There is no issue for "firstchar" and "input_message" because those
are explicitly reinitialized each time through the query processing
loop. To make that clearer, move them to be declared inside the loop.
That leaves us with all the function-lifespan locals except the
sigjmp_buf itself marked as volatile, which seems like a good policy
to have going forward.
Because of the possibility of extra disable_timeout() calls, this
seems worth back-patching.
Michael Paquier [Mon, 10 Jul 2023 00:40:14 +0000 (09:40 +0900)]
Fix ALTER EXTENSION SET SCHEMA with objects outside an extension's schema
As coded, the code would use as a base comparison the namespace OID from
the first object scanned in pg_depend when switching its namespace
dependency entry to the new one, and use it as a base of comparison for
any follow-up checks. It would also be used as the old namespace OID to
switch *from* for the extension's pg_depend entry. Hence, if the first
object scanned has a namespace different than the one stored in the
extension, we would finish by:
- Not checking that the extension objects map with the extension's
schema.
- Not switching the extension -> namespace dependency entry to the new
namespace provided by the user, making ALTER EXTENSION ineffective.
This issue exists since this command has been introduced in d9572c4 for
relocatable extension, so backpatch all the way down to 11. The test
case has been provided by Heikki, that I have tweaked a bit to show the
effects on pg_depend for the extension.
Reported-by: Heikki Linnakangas
Author: Michael Paquier, Heikki Linnakangas
Discussion: https://postgr.es/m/20eea594-a05b-4c31-491b-007b6fceef28@iki.fi
Backpatch-through: 11
David Rowley [Sun, 9 Jul 2023 04:15:25 +0000 (16:15 +1200)]
Doc: update old reference to "result cache"
During the PostgreSQL 14 cycle, the Memoize executor node was briefly
called "Result Cache" until it was renamed in 83f4fcc65. That commit
missed one reference.
Reported-by: Paul A Jungwirth
Packpatch-through: 14, where Memoize was added
Discussion: https://postgr.es/m/CA+renyX=40YXhsfPTzn13oNOPO3TJ12CK9GX-2P2pvnQiScefA@mail.gmail.com
Commit e213de8e78 fixed a problem with path lengths to a tempdir on
Windows, but caused problems on at least some Unix systems where the
system tempdir is on a different file system. To work around this, only
used the system temdir for the destination of pg_replslot on Windows,
and otherwise restore the old behaviour.
Andrew Dunstan [Sat, 8 Jul 2023 15:21:58 +0000 (11:21 -0400)]
Use shorter location for pg_replslot in pg_basebackup test
The symlink to a longer location tripped up some Windows limit on
buildfarm animal fairywren when running with meson, which uses slightly
longer paths.
Backpatch to release 14 to keep the script in sync. Before that the
script skipped all symlink related tests on Windows.
Andrew Dunstan [Thu, 6 Jul 2023 16:27:40 +0000 (12:27 -0400)]
Skip pg_baseback long filename test if path too long on Windows
On Windows, it's sometimes difficult to create a file with a path longer
than 255 chars, and if it can be created it might not be seen by the
archiver. This can be triggered by the test for tar backups with
filenames greater than 100 bytes. So we skip that test if the path would
exceed 255.
WAL-log the creation of the init fork of unlogged indexes.
We create a file, so we better WAL-log it. In practice, all the
built-in index AMs and all extensions that I'm aware of write a
metapage to the init fork, which is WAL-logged, and replay of the
metapage implicitly creates the fork too. But if ambuildempty() didn't
write any page, we would miss it.
This can be seen with dummy_index_am. Set up replication, create a
'dummy_index_am' index on an unlogged table, and look at the files
created in the replica: the init fork is not created on the
replica. Dummy_index_am doesn't do anything with the relation files,
however, so it doesn't lead to any user-visible errors.
Backpatch to all supported versions.
Reviewed-by: Robert Haas
Discussion: https://www.postgresql.org/message-id/6e5bbc08-cdfc-b2b3-9e23-1a914b9850a9%40iki.fi
This bug was only present on v15. MarkGUCPrefixReserved() is new in
v15, and in v16, it was rewritten to use a hash table and the new
implementation did not have this bug.
Amit Kapila [Thu, 6 Jul 2023 02:58:27 +0000 (08:28 +0530)]
Revert the commits related to allowing page lock to conflict among parallel group members.
This commit reverts the work done by commits 3ba59ccc89 and 72e78d831a.
Those commits were incorrect in asserting that we never acquire any other
heavy-weight lock after acquring page lock other than relation extension
lock. We can acquire a lock on catalogs while doing catalog look up after
acquring page lock.
This won't impact any existing feature but we need to think some other way
to achieve this before parallelizing other write operations or even
improving the parallelism in vacuum (like allowing multiple workers
for an index).
Reported-by: Jaime Casanova
Author: Amit Kapila
Backpatch-through: 13
Discussion: https://postgr.es/m/CAJKUy5jffnRKNvRHKQ0LynRb0RJC-o4P8Ku3x9vGAVLwDBWumQ@mail.gmail.com
llvm_release_context() called llvm_enter_fatal_on_oom(), but was missing
the corresponding llvm_leave_fatal_on_oom() call. As a result, if JIT was
used at all, we were almost always in the "fatal-on-oom" state.
It only makes a difference if you use an extension written in C++, and
run out of memory in a C++ 'new' call. In that case, you would get a
PostgreSQL FATAL error, instead of the default behavior of throwing a
C++ exception.
Back-patch to all supported versions.
Reviewed-by: Daniel Gustafsson
Discussion: https://www.postgresql.org/message-id/54b78cca-bc84-dad8-4a7e-5b56f764fab5@iki.fi
Commit 7b64e4b3 taught DropSubscription() to drop stats entry of
subscription that is not associated with a replication slot for apply
worker at DROP SUBSCRIPTION but missed covering the case where the
subscription is not associated with replication slots for both apply
worker and tablesync worker.
Also add a test to verify that the stats for slot-less subscription is
removed at DROP SUBSCRIPTION time.
Ensure that creation of an empty relfile is fsync'd at checkpoint.
If you create a table and don't insert any data into it, the relation file
is never fsync'd. You don't lose data, because an empty table doesn't have
any data to begin with, but if you crash and lose the file, subsequent
operations on the table will fail with "could not open file" error.
To fix, register an fsync request in mdcreate(), like we do for mdwrite().
Per discussion, we probably should also fsync the containing directory
after creating a new file. But that's a separate and much wider issue.
Backpatch to all supported versions.
Reviewed-by: Andres Freund, Thomas Munro
Discussion: https://www.postgresql.org/message-id/d47d8122-415e-425c-d0a2-e0160829702d%40iki.fi
Thomas Munro [Tue, 4 Jul 2023 03:16:34 +0000 (15:16 +1200)]
Re-bin segment when memory pages are freed.
It's OK to be lazy about re-binning memory segments when allocating,
because that can only leave segments in a bin that's too high. We'll
search higher bins if necessary while allocating next time, and
also eventually re-bin, so no memory can become unreachable that way.
However, when freeing memory, the largest contiguous range of free pages
might go up, so we should re-bin eagerly to make sure we don't leave the
segment in a bin that is too low for get_best_segment() to find.
The re-binning code is moved into a function of its own, so it can be
called whenever free pages are returned to the segment's free page map.
Back-patch to all supported releases.
Author: Dongming Liu <ldming101@gmail.com> Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier version) Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Discussion: https://postgr.es/m/CAL1p7e8LzB2LSeAXo2pXCW4%2BRya9s0sJ3G_ReKOU%3DAjSUWjHWQ%40mail.gmail.com
Thomas Munro [Mon, 3 Jul 2023 04:20:01 +0000 (16:20 +1200)]
Fix race in SSI interaction with gin fast path.
The ginfast.c code previously checked for conflicts in before locking
the relevant buffer, leaving a window where a RW conflict could be
missed. Re-order.
There was also a place where buffer ID and block number were confused
while trying to predicate-lock a page, noted by visual inspection.
Back-patch to all supported releases. Fixes one more problem discovered
with the reproducer from bug #17949, in this case when Dmitry tried
other index types.
Thomas Munro [Mon, 3 Jul 2023 04:18:20 +0000 (16:18 +1200)]
Fix race in SSI interaction with bitmap heap scan.
When performing a bitmap heap scan, we don't want to miss concurrent
writes that occurred after we observed the heap's rs_nblocks, but before
we took predicate locks on index pages. Therefore, we can't skip
fetching any heap tuples that are referenced by the index, because we
need to test them all with CheckForSerializableConflictOut(). The
old optimization that would ignore any references to blocks >=
rs_nblocks gets in the way of that requirement, because it means that
concurrent writes in that window are ignored.
Removing that optimization shouldn't affect correctness at any isolation
level, because any new tuples shouldn't be visible to an MVCC snapshot.
There also shouldn't be any error-causing references to heap blocks past
the end, because we should have held at least an AccessShareLock on the
table before the index scan. It can't get smaller while our transaction
is running. For now, though, we'll keep the optimization at lower
levels to avoid making unnecessary changes in a bug fix.
Back-patch to all supported releases. In release 11, the code is in a
different place but not fundamentally different. Fixes one aspect of
bug #17949.
Thomas Munro [Mon, 3 Jul 2023 04:16:27 +0000 (16:16 +1200)]
Fix race in SSI interaction with empty btrees.
When predicate-locking btrees, we have a special case for completely
empty btrees, since there is no page to lock. This was racy, because,
without buffer lock held, a matching key could be inserted between the
_bt_search() and the PredicateLockRelation() calls.
Fix, by rechecking _bt_search() after taking the relation-level SIREAD
lock, if using SERIALIZABLE isolation and an empty btree is discovered.
Back-patch to all supported releases. Fixes one aspect of bug #17949.
Tomas Vondra [Mon, 3 Jul 2023 16:16:58 +0000 (18:16 +0200)]
Remove expensive test of postgres_fdw batch inserts
The test inserted 70k rows into a foreign table, in order to verify
correct behavior with more than 65535 parameters, and was added in
response to a bug report.
However, this is rather expensive, especially when running the tests
under valgrind, CLOBBER_CACHE_ALWAYS etc. It doesn't seem worth it to
keep running the test, so remove it from all branches (14+).
Andrew Dunstan [Mon, 3 Jul 2023 14:06:26 +0000 (10:06 -0400)]
Improve pg_basebackup long file name test Windows robustness
Creation of a file with a very long name can create problems on Windows
due to its file path limits. Work around that by creating the file via a
symlink with a shorter name.
Michael Paquier [Mon, 3 Jul 2023 01:06:14 +0000 (10:06 +0900)]
Make PG_TEST_NOCLEAN work for temporary directories in TAP tests
When set, this environment variable was only effective for data
directories but not for all the other temporary files created by
PostgreSQL::Test::Utils. Keeping the temporary files after a successful
run can be useful for debugging purposes.
The documentation is updated to reflect the new behavior, with contents
available in doc/ since v16 and in src/test/perl/README since v15.
Author: Jacob Champion Reviewed-by: Daniel Gustafsson
Discussion: https://postgr.es/m/CAAWbhmgHtDH1SGZ+Fw05CsXtE0mzTmjbuUxLB9mY9iPKgM6cUw@mail.gmail.com
Discussion: https://postgr.es/m/YyPd9unV14SX2bLF@paquier.xyz
Backpatch-through: 11
Thomas Munro [Sun, 2 Jul 2023 22:53:44 +0000 (10:53 +1200)]
Silence "missing contrecord" error.
Commit dd38ff28ad added a new error message "missing contrecord" when
we fail to reassemble a record. Unfortunately that caused noisy
messages to be logged by pg_waldump at end of segment, and by walsender
when asked to shut down on a segment boundary.
Remove the new error message, so that this condition signals end-of-
WAL without a message. It's arguably a reportable condition that should
not be silenced while performing crash recovery, but fixing that without
introducing noise in the other cases will require more research.
Back-patch to 15.
Reported-by: Tomas Vondra <tomas.vondra@enterprisedb.com>
Discussion: https://postgr.es/m/6a1df56e-4656-b3ce-4b7a-a9cb41df8189%40enterprisedb.com
Tomas Vondra [Sun, 2 Jul 2023 18:29:01 +0000 (20:29 +0200)]
Fix oversight in handling of modifiedCols since f24523672d
Commit f24523672d fixed a memory leak by moving the modifiedCols bitmap
into the per-row memory context. In the case of AFTER UPDATE triggers,
the bitmap is however referenced from an event kept until the end of the
query, resulting in a use-after-free bug.
Fixed by copying the bitmap into the AfterTriggerEvents memory context,
which is the one where we keep the trigger events. There's only one
place that needs to do the copy, but the memory context may not exist
yet. Doing that in a separate function seems more readable.
Report by Alexander Pyhalov, fix by me. Backpatch to 13, where the
bitmap was added to the event by commit 71d60e2aa0.
Reported-by: Alexander Pyhalov
Backpatch-through: 13
Discussion: https://postgr.es/m/acddb17c89b0d6cb940eaeda18c08bbe@postgrespro.ru
Tomas Vondra [Sun, 2 Jul 2023 16:54:09 +0000 (18:54 +0200)]
Fix memory leak in Incremental Sort rescans
The Incremental Sort had a couple issues, resulting in leaking memory
during rescans, possibly triggering OOM. The code had a couple of
related flaws:
1. During rescans, the sort states were reset but then also set to NULL
(despite the comment saying otherwise). ExecIncrementalSort then
sees NULL and initializes a new sort state, leaking the memory used
by the old one.
2. Initializing the sort state also automatically rebuilt the info about
presorted keys, leaking the already initialized info. presorted_keys
was also unnecessarily reset to NULL.
Patch by James Coleman, based on patches by Laurenz Albe and Tom Lane.
Backpatch to 13, where Incremental Sort was introduced.
Author: James Coleman, Laurenz Albe, Tom Lane Reported-by: Laurenz Albe, Zu-Ming Jiang
Backpatch-through: 13
Discussion: https://postgr.es/m/b2bd02dff61af15e3526293e2771f874cf2a3be7.camel%40cybertec.at
Discussion: https://postgr.es/m/db03c582-086d-e7cd-d4a1-3bc722f81765%40inf.ethz.ch
Michael Paquier [Fri, 30 Jun 2023 04:54:55 +0000 (13:54 +0900)]
Fix marking of indisvalid for partitioned indexes at creation
The logic that introduced partitioned indexes missed a few things when
invalidating a partitioned index when these are created, still the code
is written to handle recursions:
1) If created from scratch because a mapping index could not be found,
the new index created could be itself invalid, if for example it was a
partitioned index with one of its leaves invalid.
2) A CCI was missing when indisvalid is set for a parent index, leading
to inconsistent trees when recursing across more than one level for a
partitioned index creation if an invalidation of the parent was
required.
This could lead to the creation of a partition index tree where some of
the partitioned indexes are marked as invalid, but some of the parents
are marked valid, which is not something that should happen (as
validatePartitionedIndex() defines, indisvalid is switched to true for a
partitioned index iff all its partitions are themselves valid).
This patch makes sure that indisvalid is set to false on a partitioned
index if at least one of its partition is invalid. The flag is set to
true if *all* its partitions are valid.
The regression test added in this commit abuses of a failed concurrent
index creation, marked as invalid, that maps with an index created on
its partitioned table afterwards.
Reported-by: Alexander Lakhin Reviewed-by: Alexander Lakhin
Discussion: https://postgr.es/m/14987634-43c0-0cb3-e075-94d423607e08@gmail.com
Backpatch-through: 11
Michael Paquier [Thu, 29 Jun 2023 22:49:07 +0000 (07:49 +0900)]
Fix pg_depend entry to AMs after ALTER TABLE .. SET ACCESS METHOD
ALTER TABLE .. SET ACCESS METHOD was not registering a dependency to the
new access method with the relation altered in its rewrite phase, making
possible the drop of an access method even if there are relations that
depend on it. During the rewrite, a temporary relation is created to
build the new relation files before swapping the new and old files, and,
while the temporary relation was registering a correct dependency to the
new AM, the old relation did not do that. A dependency on the access
method is added when the relation files are swapped, which is the point
where pg_class is updated.
Materialized views and tables use the same code path, hence both were
impacted.
Backpatch down to 15, where this command has been introduced.
Reported-by: Alexander Lakhin Reviewed-by: Nathan Bossart, Andres Freund
Discussion: https://postgr.es/m/18000-9145c25b1af475ca@postgresql.org
Backpatch-through: 15
Tom Lane [Thu, 29 Jun 2023 14:19:10 +0000 (10:19 -0400)]
Fix order of operations in ExecEvalFieldStoreDeForm().
If the given composite datum is toasted out-of-line,
DatumGetHeapTupleHeader will perform database accesses to detoast it.
That can invalidate the result of get_cached_rowtype, as documented
(perhaps not plainly enough) in that function's API spec; which leads
to strange errors or crashes when we try to use the TupleDesc to read
the tuple. In short then, trying to update a field of a composite
column could fail intermittently if the overall column value is wide
enough to require toasting.
We can fix the bug at no cost by just changing the order of
operations, since we don't need the TupleDesc until after detoasting.
(Other callers of get_cached_rowtype appear to get this right already,
so there's only one bug.)
Note that the added regression test case reveals this bug reliably
only with debug_discard_caches/CLOBBER_CACHE_ALWAYS.
Per bug #17994 from Alexander Lakhin. Sadly, this patch does not fix
the missing-values issue revealed in the bug discussion; we'll need
some more work to cover that.
It was walking into the ColumnDef->compression field, which is not a
node but a string. This code is currently not reachable (because the
compression field is only set in situations that don't go through
raw_expression_tree_walker()), but if it had been, this could have
behaved erratically.
Michael Paquier [Wed, 28 Jun 2023 06:57:43 +0000 (15:57 +0900)]
Ignore invalid indexes when enforcing index rules in ALTER TABLE ATTACH PARTITION
A portion of ALTER TABLE .. ATTACH PARTITION is to ensure that the
partition being attached to the partitioned table has a correct set of
indexes, so as there is a consistent index mapping between the
partitioned table and its new-to-be partition. However, as introduced
in 8b08f7d, the current logic could choose an invalid index as a match,
which is something that can exist when dealing with more than two levels
of partitioning, like attaching a partitioned table (that has
partitions, with an index created by CREATE INDEX ON ONLY) to another
partitioned table.
A partitioned index with indisvalid set to false is equivalent to an
incomplete partition tree, meaning that an invalid partitioned index
does not have indexes defined in all its partitions. Hence, choosing an
invalid partitioned index can create inconsistent partition index trees,
where the parent attaching to is valid, but its partition may be
invalid.
In the report from Alexander Lakhin, this showed up as an assertion
failure when validating an index. Without assertions enabled, the
partition index tree would be actually broken, as indisvalid should
be switched to true for a partitioned index once all its partitions are
themselves valid. With two levels of partitioning, the top partitioned
table used a valid index and was able to link to an invalid index stored
on its partition, itself a partitioned table.
I have studied a few options here (like the possibility to switch
indisvalid to false for the parent), but came down to the conclusion
that we'd better rely on a simple rule: invalid indexes had better never
be chosen, so as the partition attached uses and creates indexes that
the parent expects. Some regression tests are added to provide some
coverage. Note that the existing coverage is not impacted.
This is a problem since partitioned indexes exist, so backpatch all the
way down to v11.
Reported-by: Alexander Lakhin
Discussion: https://postgr.es/14987634-43c0-0cb3-e075-94d423607e08@gmail.com
Backpatch-through: 11
Tom Lane [Sat, 24 Jun 2023 21:18:08 +0000 (17:18 -0400)]
Check for interrupts and stack overflow in TParserGet().
TParserGet() recurses for some token types, meaning it's possible
to drive it to stack overflow. Since this is a minority behavior,
I chose to add the check_stack_depth() call to the two places that
recurse rather than doing it during every single call.
While at it, add CHECK_FOR_INTERRUPTS(), because this can run
unpleasantly long for long inputs.
Per bug #17995 from Zuming Jiang. This is old, so back-patch
to all supported branches.
Michael Paquier [Fri, 23 Jun 2023 08:50:23 +0000 (17:50 +0900)]
Fix incorrect error message in libpq_pipeline
One of the tests for the pipeline mode with portal description expects a
non-NULL PQgetResult, but used an incorrect error message on failure,
telling that PQgetResult being NULL was the expected result.
Amit Kapila [Thu, 22 Jun 2023 07:05:10 +0000 (12:35 +0530)]
Doc: Clarify the behavior of triggers/rules in a logical subscriber.
By default, triggers and rules do not fire on a logical replication
subscriber based on the "session_replication_role" GUC being set to
"replica". However, the docs in the logical replication section assumed
that the reader understood how this GUC worked. This modifies the docs to
be more explicit and links back to the GUC itself.
Author: Jonathan Katz, Peter Smith Reviewed-by: Vignesh C, Euler Taveira
Backpatch-through: 11
Discussion: https://postgr.es/m/5bb2c9a2-499f-e1a2-6e33-5ce96b35cc4a@postgresql.org
David Rowley [Thu, 22 Jun 2023 00:47:15 +0000 (12:47 +1200)]
Doc: mention that extended stats aren't used for joins
Statistics defined by the CREATE STATISTICS command are only used to
assist with the selectivity estimations of base relations, never for
joins. Here we mention this fact in the notes section of the CREATE
STATISTICS command.
Peter Geoghegan [Thu, 22 Jun 2023 00:41:56 +0000 (17:41 -0700)]
nbtree VACUUM: cope with topparent inconsistencies.
Avoid "right sibling %u of block %u is not next child" errors when
vacuuming a corrupt nbtree index. Just LOG the issue and press on.
That way VACUUM will have a decent chance of finishing off all required
processing for the index (and for the table as a whole).
This is similar to recent work from commit 5abff197, as well as work
from commit 5b861baa (later backpatched as commit 43e409ce), which
taught nbtree VACUUM to keep going when its "re-find" check fails. The
hardening added by this commit takes place directly after the "re-find"
check, right before the critical section for the first stage of page
deletion.
Tom Lane [Wed, 21 Jun 2023 15:07:11 +0000 (11:07 -0400)]
Avoid Assert failure when processing empty statement in aborted xact.
exec_parse_message() wants to create a cached plan in all cases,
including for empty input. The empty-input path does not have
a test for being in an aborted transaction, making it possible
that plancache.c will fail due to trying to do database lookups
even though there's no real work to do.
One solution would be to throw an aborted-transaction error in
this path too, but it's not entirely clear whether the lack of
such an error was intentional or whether some clients might be
relying on non-error behavior. Instead, let's hack plancache.c
so that it treats empty statements with the same logic it
already had for transaction control commands, ensuring that it
can soldier through even in an already-aborted transaction.
Per bug #17983 from Alexander Lakhin. Back-patch to all
supported branches.
Michael Paquier [Wed, 21 Jun 2023 07:16:20 +0000 (16:16 +0900)]
Disable use of archiving in 009_twophase.pl
This partially reverts 68cb5af, as using archiving to enforce the
rename of the last partial segment of the old timeline at promotion to
use .partial as suffix is impacting the tests when it does switchovers.
As showed by the logs gathered by the CI in the tests that failed, a new
standby may fail to find the WAL segment it needs to follow a promoted
instance with its timeline jump, as it got renamed to .partial.
This problem would manifest as a run timeout with 009_twophase.pl, as
the new standby repeatedly requests a segment from the promoted primary
that it would not find.
Amit Kapila [Wed, 21 Jun 2023 04:54:41 +0000 (10:24 +0530)]
Fix the errhint message and docs for drop subscription failure.
The existing errhint message and docs were missing the fact that we can't
disassociate from the slot unless the subscription is disabled.
Author: Robert Sjöblom, Peter Smith Reviewed-by: Peter Eisentraut, Amit Kapila
Backpatch-through: 11
Discussion: https://postgr.es/m/807bdf85-61ea-88e2-5712-6d9fcd4eabff@fortnox.se
Tom Lane [Tue, 20 Jun 2023 21:47:36 +0000 (17:47 -0400)]
Fix hash join when inner hashkey expressions contain Params.
If the inner-side expressions contain PARAM_EXEC Params, we must
re-hash whenever the values of those Params change. The executor
mechanism for that exists already, but we failed to invoke it because
finalize_plan() neglected to search the Hash.hashkeys field for
Params. This allowed a previous scan's hash table to be re-used
when it should not be, leading to rows missing from the join's output.
(I believe incorrectly-included join rows are impossible however,
since checking the real hashclauses would reject false matches.)
This bug is very ancient, dating probably to d24d75ff1 of 7.4.
Sadly, this simple fix depends on the plan representational changes
made by 2abd7ae9b, so it will only work back to v12. I thought
about trying to make some kind of hack for v11, but I'm leery
of putting code significantly different from what is used in the
newer branches into a nearly-EOL branch. Seeing that the bug
escaped detection for a full twenty years, problematic cases
must be rare; so I don't feel too awful about leaving v11 as-is.
Per bug #17985 from Zuming Jiang. Back-patch to v12.
Michael Paquier [Tue, 20 Jun 2023 01:25:41 +0000 (10:25 +0900)]
Enable archiving in recovery TAP test 009_twophase.pl
This is a follow-up of f663b00, that has been committed to v13 and v14,
tweaking the TAP test for two-phase transactions so as it provides
coverage for the bug that has been fixed. This change is done in its
own commit for clarity, as v15 and HEAD did not show the problematic
behavior, still missed coverage for it.
While on it, this adds a comment about the dependency of the last
partial segment rename and RecoverPreparedTransactions() at the end of
recovery, as that can be easy to miss.
Author: Michael Paquier Reviewed-by: Kyotaro Horiguchi
Discussion: https://postgr.es/m/743b9b45a2d4013bd90b6a5cba8d6faeb717ee34.camel@cybertec.at
Backpatch-through: 13
David Rowley [Mon, 19 Jun 2023 01:01:29 +0000 (13:01 +1200)]
Don't use partial unique indexes for unique proofs in the planner
Here we adjust relation_has_unique_index_for() so that it no longer makes
use of partial unique indexes as uniqueness proofs. It is incorrect to
use these as the predicates used by check_index_predicates() to set
predOK makes use of not only baserestrictinfo quals as proofs, but also
qual from join conditions. For relation_has_unique_index_for()'s case, we
need to know the relation is unique for a given set of columns before any
joins are evaluated, so if predOK was only set to true due to some join
qual, then it's unsafe to use such indexes in
relation_has_unique_index_for(). The final plan may not even make use
of that index, which could result in reading tuples that are not as
unique as the planner previously expected them to be.
Bug: #17975 Reported-by: Tor Erik Linnerud
Backpatch-through: 11, all supported versions
Discussion: https://postgr.es/m/17975-98a90c156f25c952%40postgresql.org