Specifically, when pg_basebackup is invoked with -Tx=y, don't error
out if x could plausibly be an absolute path either on Windows or on
non-Windows systems. We don't know whether the remote system is
running the same OS as the local system, so it's not appropriate to
assume that our local rule about absolute pathnames is the same as
the rule on the remote system.
Patch by me, reviewed by Tom Lane, Andrew Dunstan, and
Davinder Singh.
Discussion: http://postgr.es/m/CA+TgmoY+jC3YiskomvYKDPK3FbrmsDU7_8+wMHt02HOdJeRb0g@mail.gmail.com
Per discussion, the existing routine name able to initialize a SRF
function with materialize mode is unpopular, so rename it. Equally, the
flags of this function are renamed, as of:
- SRF_SINGLE_USE_EXPECTED -> MAT_SRF_USE_EXPECTED_DESC
- SRF_SINGLE_BLESS -> MAT_SRF_BLESS
The previous function and flags introduced in 9e98583 are kept around
for compatibility purposes, so as any extension code already compiled
with v15 continues to work as-is. The declarations introduced here for
compatibility will be removed from HEAD in a follow-up commit.
The new names have been suggested by Andres Freund and Melanie
Plageman.
Discussion: https://postgr.es/m/20221013194820.ciktb2sbbpw7cljm@awork3.anarazel.de
Backpatch-through: 15
In the latest version of Apple's macOS SDK, <sys/socket.h>
fails to compile if "REF" is #define'd as something.
Apple may or may not agree that this is a bug, and even if
they do accept the bug report I filed, they probably won't
fix it very quickly. In the meantime, our back branches will all
fail to compile gram.y. v15 and HEAD currently escape the problem
thanks to the refactoring done in 98e93a1fc, but that's purely
accidental. Moreover, since that patch removed a widely-visible
inclusion of <netdb.h>, back-patching it seems too likely to break
third-party code.
Instead, change the token's code name to REF_P, following our usual
convention for naming parser tokens that are likely to have symbol
conflicts. The effects of that should be localized to the grammar
and immediately surrounding files, so it seems like a safer answer.
Per project policy that we want to keep recently-out-of-support
branches buildable on modern systems, back-patch all the way to 9.2.
Discussion: https://postgr.es/m/1803927.1665938411@sss.pgh.pa.us
Previously PgStat_StatReplSlotEntry contained the slotname, which was mainly
used when writing out the stats during shutdown, to identify the slot in the
serialized data (at runtime the index in ReplicationSlotCtl->replication_slots
is used, but that can change during a restart). Unfortunately the slotname was
overwritten when the slot's stats were reset.
That turned out to only cause "real" problems if the slot was active during
the reset, triggering an assertion failure at the next
pgstat_report_replslot(). In other paths the stats were re-initialized during
pgstat_acquire_replslot().
Fix this by removing slotname from PgStat_StatReplSlotEntry. Instead we can
get the slot's name from the slot itself. Besides fixing a bug, this also is
architecturally cleaner (a name is not really statistics). This is safe
because stats, for a slot removed while shut down, will not be restored at
startup.
In 15 the slotname is not removed, but renamed, to avoid changing the stats
format. In master, bump PGSTAT_FILE_FORMAT_ID.
This commit does not contain a test for the fix. I think this can only be
tested by a tap test starting pg_recvlogical in the background and checking
pg_recvlogical's output. That type of test is notoriously hard to be reliable,
so committing it shortly before the release is wrapped seems like a bad idea.
Reported-by: Jaime Casanova <jcasanov@systemguards.com.ec>
Author: Andres Freund <andres@anarazel.de>
Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Discussion: https://postgr.es/m/YxfagaTXUNa9ggLb@ahch-to
Backpatch: 15-, where the bug was introduced in 5891c7a8ed8f
This reverts commit db0d67db2401eb6238ccc04c6407a4fd4f985832 and
several follow-on fixes. The idea of making a cost-based choice
of the order of the sorting columns is not fundamentally unsound,
but it requires cost information and data statistics that we don't
really have. For example, relying on procost to distinguish the
relative costs of different sort comparators is pretty pointless
so long as most such comparator functions are labeled with cost 1.0.
Moreover, estimating the number of comparisons done by Quicksort
requires more than just an estimate of the number of distinct values
in the input: you also need some idea of the sizes of the larger
groups, if you want an estimate that's good to better than a factor of
three or so. That's data that's often unknown or not very reliable.
Worse, to arrive at estimates of the number of calls made to the
lower-order-column comparison functions, the code needs to make
estimates of the numbers of distinct values of multiple columns,
which are necessarily even less trustworthy than per-column stats.
Even if all the inputs are perfectly reliable, the cost algorithm
as-implemented cannot offer useful information about how to order
sorting columns beyond the point at which the average group size
is estimated to drop to 1.
Close inspection of the code added by db0d67db2 shows that there
are also multiple small bugs. These could have been fixed, but
there's not much point if we don't trust the estimates to be
accurate in-principle.
Finally, the changes in cost_sort's behavior made for very large
changes (often a factor of 2 or so) in the cost estimates for all
sorting operations, not only those for multi-column GROUP BY.
That naturally changes plan choices in many situations, and there's
precious little evidence to show that the changes are for the better.
Given the above doubts about whether the new estimates are really
trustworthy, it's hard to summon much confidence that these changes
are better on the average.
Since we're hard up against the release deadline for v15, let's
revert these changes for now. We can always try again later.
Note: in v15, I left T_PathKeyInfo in place in nodes.h even though
it's unreferenced. Removing it would be an ABI break, and it seems
a bit late in the release cycle for that.
Discussion: https://postgr.es/m/TYAPR01MB586665EB5FB2C3807E893941F5579@TYAPR01MB5866.jpnprd01.prod.outlook.com
This adds additional variants of palloc, pg_malloc, etc. that
encapsulate common usage patterns and provide more type safety.
Specifically, this adds palloc_object(), palloc_array(), and
repalloc_array(), which take the type name of the object to be
allocated as its first argument and cast the return as a pointer to
that type. There are also palloc0_object() and palloc0_array()
variants for initializing with zero, and pg_malloc_*() variants of all
of the above.
Inspired by the talloc library.
This is backpatched from master so that future backpatchable code can
make use of these APIs. This patch by itself does not contain any
users of these APIs.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://www.postgresql.org/message-id/flat/bb755632-2a43-d523-36f8-a1e7a389a907@enterprisedb.com
PG_COMPRESSION_OPTION_LEVEL is removed from the compression
specification logic, and instead the compression level is always
assigned with each library's default if nothing is directly given. This
centralizes the checks on the compression methods supported by a given
build, and always assigns a default compression level when parsing a
compression specification. This results in complaining at an earlier
stage than previously if a build supports a compression method or not,
aka when parsing a specification in the backend or the frontend, and not
when processing it. zstd, lz4 and zlib are able to handle in their
respective routines setting up the compression level the case of a
default value, hence the backend or frontend code (pg_receivewal or
pg_basebackup) has now no need to know what the default compression
level should be if nothing is specified: the logic is now done so as the
specification parsing assigns it. It can also be enforced by passing
down a "level" set to the default value, that the backend will accept
(the replication protocol is for example able to handle a command like
BASE_BACKUP (COMPRESSION_DETAIL 'gzip:level=-1')).
This code simplification fixes an issue with pg_basebackup --gzip
introduced by ffd5365, where the tarball of the streamed WAL segments
would be created as of pg_wal.tar.gz with uncompressed contents, while
the intention is to compress the segments with gzip at a default level.
The origin of the confusion comes from the handling of the default
compression level of gzip (-1 or Z_DEFAULT_COMPRESSION) and the value of
0 was getting assigned, which is what walmethods.c would consider
as equivalent to no compression when streaming WAL segments with its tar
methods. Assigning always the compression level removes the confusion
of some code paths considering a value of 0 set in a specification as
either no compression or a default compression level.
Note that 010_pg_basebackup.pl has to be adjusted to skip a few tests
where the shape of the compression detail string for client and
server-side compression was checked using gzip. This is a result of the
code simplification, as gzip specifications cannot be used if a build
does not support it.
Reported-by: Tom Lane
Reviewed-by: Tom Lane
Discussion: https://postgr.es/m/1400032.1662217889@sss.pgh.pa.us
Backpatch-through: 15
We should process completed IOs *before* trying to start more, so that
it is always possible to decode one more record when the decoded record
queue is empty, even if maintenance_io_concurrency is set so low that a
single earlier WAL record might have saturated the IO queue.
That bug was hidden because the effect of maintenance_io_concurrency was
arbitrarily clamped to be at least 2. Fix the ordering, and also remove
that clamp. We need a special case for 0, which is now treated the same
as recovery_prefetch=off, but otherwise the number is used directly.
This allows for testing with 1, which would have made the problem
obvious in simple test scenarios.
Also add an explicit error message for missing contrecords. It was a
bit strange that we didn't report an error already, and became a latent
bug with prefetching, since the internal state that tracks aborted
contrecords would not survive retrying, as revealed by
026_overwrite_contrecord.pl with this adjustment. Reporting an error
prevents that.
Back-patch to 15.
Reported-by: Justin Pryzby <pryzby@telsasoft.com>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Discussion: https://postgr.es/m/20220831140128.GS31833%40telsasoft.com
The addition of published column names forgot to filter on attisdropped,
leading to cases where you could see "........pg.dropped.1........"
or the like as a reportedly-published column.
While we're here, rewrite the new subquery to get a more efficient plan
for it.
Hou Zhijie, per report from Jaime Casanova. Back-patch to v15 where
the bug was introduced. (Sadly, this means we need a post-beta4
catversion bump before beta4 has even hit the streets. I see no
good alternative though.)
Discussion: https://postgr.es/m/Yxa1SU4nH2HfN3/i@ahch-to
XLogPageRead() can retry internally after a pread() system call has
succeeded, in the case of short reads, and page validation failures
while in standby mode (see commit 0668719801). Due to an oversight in
commit 3f1ce973, these cases could leave stale data in the internal
cache of xlogreader.c without marking it invalid. The main defense
against stale cached data on failure to read a page was in the error
handling path of the calling function ReadPageInternal(), but that
wasn't quite enough for errors handled internally by XLogPageRead()'s
retry loop if we then exited with XLREAD_WOULDBLOCK.
1. ReadPageInternal() now marks the cache invalid before calling the
page_read callback, by setting state->readLen to 0. It'll be set to
a non-zero value only after a successful read. It'll stay valid as
long as the caller requests data in the cached range.
2. XLogPageRead() no long performs internal retries while reading
ahead. While such retries should work, the general philosophy is
that we should give up prefetching if anything unusual happens so we
can handle it when recovery catches up, to reduce the complexity of
the system. Let's do that here too.
3. While here, a new function XLogReaderResetError() improves the
separation between xlogrecovery.c and xlogreader.c, where the former
previously clobbered the latter's internal error buffer directly.
The new function makes this more explicit, and also clears a related
flag, without which a standby would needlessly retry in the outer
function.
Thanks to Noah Misch for tracking down the conditions required for a
rare build farm failure in src/bin/pg_ctl/t/003_promote.pl, and
providing a reproducer.
Back-patch to 15.
Reported-by: Noah Misch <noah@leadboat.com>
Discussion: https://postgr.es/m/20220807003627.GA4168930%40rfd.leadboat.com
The reverts the following and makes some associated cleanups:
commit f79b803dc: Common SQL/JSON clauses
commit f4fb45d15: SQL/JSON constructors
commit 5f0adec25: Make STRING an unreserved_keyword.
commit 33a377608: IS JSON predicate
commit 1a36bc9db: SQL/JSON query functions
commit 606948b05: SQL JSON functions
commit 49082c2cc: RETURNING clause for JSON() and JSON_SCALAR()
commit 4e34747c8: JSON_TABLE
commit fadb48b00: PLAN clauses for JSON_TABLE
commit 2ef6f11b0: Reduce running time of jsonb_sqljson test
commit 14d3f24fa: Further improve jsonb_sqljson parallel test
commit a6baa4bad: Documentation for SQL/JSON features
commit b46bcf7a4: Improve readability of SQL/JSON documentation.
commit 112fdb352: Fix finalization for json_objectagg and friends
commit fcdb35c32: Fix transformJsonBehavior
commit 4cd8717af: Improve a couple of sql/json error messages
commit f7a605f63: Small cleanups in SQL/JSON code
commit 9c3d25e17: Fix JSON_OBJECTAGG uniquefying bug
commit a79153b7a: Claim SQL standard compliance for SQL/JSON features
commit a1e7616d6: Rework SQL/JSON documentation
commit 8d9f9634e: Fix errors in copyfuncs/equalfuncs support for JSON node types.
commit 3c633f32b: Only allow returning string types or bytea from json_serialize
commit 67b26703b: expression eval: Fix EEOP_JSON_CONSTRUCTOR and EEOP_JSONEXPR size.
The release notes are also adjusted.
Backpatch to release 15.
Discussion: https://postgr.es/m/40d2c882-bcac-19a9-754d-4299e1d87ac7@postgresql.org
This event can happen when using SET ACCESS METHOD, as the data files of
the materialized need a full refresh but this command tag was not
updated to reflect that. The documentation is updated to track this
behavior.
Author: Onder Kalaci
Discussion: https://postgr.es/m/CACawEhXwHN3X34FiwoYG8vXR-oyUdrp7qcfRWSzS+NPahS5gSw@mail.gmail.com
Backpatch-through: 15
Previously, we relied on HEAP2_NEW_CID records and XACT_INVALIDATION
records to know if the transaction has modified the catalog, and that
information is not serialized to snapshot. Therefore, after the restart,
if the logical decoding decodes only the commit record of the transaction
that has actually modified a catalog, we will miss adding its XID to the
snapshot. Thus, we will end up looking at catalogs with the wrong
snapshot.
To fix this problem, this changes the snapshot builder so that it
remembers the last-running-xacts list of the decoded RUNNING_XACTS record
after restoring the previously serialized snapshot. Then, we mark the
transaction as containing catalog changes if it's in the list of initial
running transactions and its commit record has XACT_XINFO_HAS_INVALS. To
avoid ABI breakage, we store the array of the initial running transactions
in the static variables InitialRunningXacts and NInitialRunningXacts,
instead of storing those in SnapBuild or ReorderBuffer.
This approach has a false positive; we could end up adding the transaction
that didn't change catalog to the snapshot since we cannot distinguish
whether the transaction has catalog changes only by checking the COMMIT
record. It doesn't have the information on which (sub) transaction has
catalog changes, and XACT_XINFO_HAS_INVALS doesn't necessarily indicate
that the transaction has catalog change. But that won't be a problem since
we use snapshot built during decoding only to read system catalogs.
On the master branch, we took a more future-proof approach by writing
catalog modifying transactions to the serialized snapshot which avoids the
above false positive. But we cannot backpatch it because of a change in
the SnapBuild.
Reported-by: Mike Oh
Author: Masahiko Sawada
Reviewed-by: Amit Kapila, Shi yu, Takamichi Osumi, Kyotaro Horiguchi, Bertrand Drouvot, Ahsan Hadi
Backpatch-through: 10
Discussion: https://postgr.es/m/81D0D8B0-E7C4-4999-B616-1E5004DBDCD2%40amazon.com
Previously, if an extension script did CREATE OR REPLACE and there was
an existing object not belonging to the extension, it would overwrite
the object and adopt it into the extension. This is problematic, first
because the overwrite is probably unintentional, and second because we
didn't change the object's ownership. Thus a hostile user could create
an object in advance of an expected CREATE EXTENSION command, and would
then have ownership rights on an extension object, which could be
modified for trojan-horse-type attacks.
Hence, forbid CREATE OR REPLACE of an existing object unless it already
belongs to the extension. (Note that we've always forbidden replacing
an object that belongs to some other extension; only the behavior for
previously-free-standing objects changes here.)
For the same reason, also fail CREATE IF NOT EXISTS when there is
an existing object that doesn't belong to the extension.
Our thanks to Sven Klemm for reporting this problem.
Security: CVE-2022-2625
Using ATSimpleRecursion() in ATPrepCmd() to do so as bbb927b4db9b did is
not correct, because ATPrepCmd() can't distinguish between triggers that
may be cloned and those that may not, so would wrongly try to recurse
for the latter category of triggers.
So this commit restores the code in EnableDisableTrigger() that
86f575948c77 had added to do the recursion, which would do it only for
triggers that may be cloned, that is, row-level triggers. This also
changes tablecmds.c such that ATExecCmd() is able to pass the value of
ONLY flag down to EnableDisableTrigger() using its new 'recurse'
parameter.
This also fixes what seems like an oversight of 86f575948c77 that the
recursion to partition triggers would only occur if EnableDisableTrigger()
had actually changed the trigger. It is more apt to recurse to inspect
partition triggers even if the parent's trigger didn't need to be
changed: only then can we be certain that all descendants share the same
state afterwards.
Backpatch all the way back to 11, like bbb927b4db9b. Care is taken not
to break ABI compatibility (and that no catversion bump is needed.)
Co-authored-by: Amit Langote <amitlangote09@gmail.com>
Reviewed-by: Dmitry Koval <d.koval@postgrespro.ru>
Discussion: https://postgr.es/m/CA+HiwqG-cZT3XzGAnEgZQLoQbyfJApVwOTQaCaas1mhpf+4V5A@mail.gmail.com
Commit fac1b470a thought we could check for set-returning functions
by testing only the top-level node in an expression tree. This is
wrong in itself, and to make matters worse it encouraged others
to make the same mistake, by exporting tlist.c's special-purpose
IS_SRF_CALL() as a widely-visible macro. I can't find any evidence
that anyone's taken the bait, but it was only a matter of time.
Use expression_returns_set() instead, and stuff the IS_SRF_CALL()
genie back in its bottle, this time with a warning label. I also
added a couple of cross-reference comments.
After a fair amount of fooling around, I've despaired of making
a robust test case that exposes the bug reliably, so no test case
here. (Note that the test case added by fac1b470a is itself
broken, in that it doesn't notice if you remove the code change.
The repro given by the bug submitter currently doesn't fail either
in v15 or HEAD, though I suspect that may indicate an unrelated bug.)
Per bug #17564 from Martijn van Oosterhout. Back-patch to v13,
as the faulty patch was.
Discussion: https://postgr.es/m/17564-c7472c2f90ef2da3@postgresql.org
We have a few commands that "can't run in a transaction block",
meaning that if they complete their processing but then we fail
to COMMIT, we'll be left with inconsistent on-disk state.
However, the existing defenses for this are only watertight for
simple query protocol. In extended protocol, we didn't commit
until receiving a Sync message. Since the client is allowed to
issue another command instead of Sync, we're in trouble if that
command fails or is an explicit ROLLBACK. In any case, sitting
in an inconsistent state while waiting for a client message
that might not come seems pretty risky.
This case wasn't reachable via libpq before we introduced pipeline
mode, but it's always been an intended aspect of extended query
protocol, and likely there are other clients that could reach it
before.
To fix, set a flag in PreventInTransactionBlock that tells
exec_execute_message to force an immediate commit. This seems
to be the approach that does least damage to existing working
cases while still preventing the undesirable outcomes.
While here, add some documentation to protocol.sgml that explicitly
says how to use pipelining. That's latent in the existing docs if
you know what to look for, but it's better to spell it out; and it
provides a place to document this new behavior.
Per bug #17434 from Yugo Nagata. It's been wrong for ages,
so back-patch to all supported branches.
Discussion: https://postgr.es/m/17434-d9f7a064ce2a88a3@postgresql.org
Previously we did this after InitPostgres, at a somewhat randomly chosen
place within PostgresMain. However, since commit a0ffa885e doing this
outside a transaction can cause a crash, if we need to check permissions
while replacing a placeholder GUC. (Besides which, a preloaded library
could itself want to do database access within _PG_init.)
To avoid needing an additional transaction start/end in every session,
move the process_session_preload_libraries call to within InitPostgres's
transaction. That requires teaching the code not to call it when
InitPostgres is called from somewhere other than PostgresMain, since
we don't want session_preload_libraries to affect background workers.
The most future-proof solution here seems to be to add an additional
flag parameter to InitPostgres; fortunately, we're not yet very worried
about API stability for v15.
Doing this also exposed the fact that we're currently honoring
session_preload_libraries in walsenders, even those not connected to
any database. This seems, at minimum, a POLA violation: walsenders
are not interactive sessions. Let's stop doing that.
(All these comments also apply to local_preload_libraries, of course.)
Per report from Gurjeet Singh (thanks also to Nathan Bossart and Kyotaro
Horiguchi for review). Backpatch to v15 where a0ffa885e came in.
Discussion: https://postgr.es/m/CABwTF4VEpwTHhRQ+q5MiC5ucngN-whN-PdcKeufX7eLSoAfbZA@mail.gmail.com
We allow users to set the values of not-yet-loaded extension GUCs,
remembering those values in "placeholder" GUC entries. When/if
the extension is loaded later in the session, we need to verify that
the user had permissions to set the GUC. That was done correctly
before commit a0ffa885e, but as of that commit, we'd check the
permissions of the active role when the LOAD happens, not the role
that had set the value. (This'd be a security bug if it had made it
into a released version.)
In principle this is simple enough to fix: we just need to remember
the exact role OID that set each GUC value, and use that not
GetUserID() when verifying permissions. Maintaining that data in
the guc.c data structures is slightly tedious, but fortunately it's
all basically just copy-n-paste of the logic for tracking the
GucSource of each setting, as we were already doing.
Another oversight is that validate_option_array_item() hadn't
been taught to check for granted GUC privileges. This appears
to manifest only in that ALTER ROLE/DATABASE RESET ALL will
fail to reset settings that the user should be allowed to reset.
Patch by myself and Nathan Bossart, per report from Nathan Bossart.
Back-patch to v15 where the faulty code came in.
Discussion: https://postgr.es/m/20220706224727.GA2158260@nathanxps13
Because they are not available we've used _fileno(stdin) in some places, but
that doesn't reliably work, because stdin might be closed. This is the
prerequisite of the subsequent commit, fixing a failure introduced in
76e38b37a5.
Author: Andres Freund <andres@anarazel.de>
Reported-By: Sandeep Thakkar <sandeep.thakkar@enterprisedb.com>
Message-Id: 20220520164558.ozb7lm6unakqzezi@alap3.anarazel.de (on pgsql-packagers)
Backpatch: 15-, where 76e38b37a5 came in
Noticed while working in this area. This code was introduced in PG15,
which is still in beta, so backpatch to there for consistency.
Backpatch-through: 15
There are a few things that we could do a little better within
get_cheapest_group_keys_order():
1. We should be using list_free() rather than pfree() on a List.
2. We should use for_each_from() instead of manually coding a for loop to
skip the first n elements of a List
3. list_truncate(list_copy(...), n) is not a great way to copy the first n
elements of a list. Let's invent list_copy_head() for that. That way we
don't need to copy the entire list just to truncate it directly
afterwards.
4. We can simplify finding the cheapest cost by setting the cheapest cost
variable to DBL_MAX. That allows us to skip special-casing the initial
iteration of the loop.
Author: David Rowley
Discussion: https://postgr.es/m/CAApHDvrGyL3ft8waEkncG9y5HDMu5TFFJB1paoTC8zi9YK97Nw@mail.gmail.com
Backpatch-through: 15, where get_cheapest_group_keys_order was added.
Justin Pryzby reported that some scenarios could cause gathering
of extended statistics to spend many seconds in an un-cancelable
qsort() operation. To fix, invent qsort_interruptible(), which is
just like qsort_arg() except that it will also do CHECK_FOR_INTERRUPTS
every so often. This bloats the backend by a couple of kB, which
seems like a good investment. (We considered just enabling
CHECK_FOR_INTERRUPTS in the existing qsort and qsort_arg functions,
but there are some callers for which that'd demonstrably be unsafe.
Opt-in seems like a better way.)
For now, just apply qsort_interruptible() in statistics collection.
There's probably more places where it could be useful, but we can
always change other call sites as we find problems.
Back-patch to v14. Before that we didn't have extended stats on
expressions, so that the problem was less severe. Also, this patch
depends on the sort_template infrastructure introduced in v14.
Tom Lane and Justin Pryzby
Discussion: https://postgr.es/m/20220509000108.GQ28830@telsasoft.com
dshash.c previously maintained flags to be able to assert that you
didn't hold any partition lock. These flags could get out of sync with
reality in error scenarios.
Get rid of all that, and make assertions about the locks themselves
instead. Since LWLockHeldByMe() loops internally, we don't want to put
that inside another loop over all partition locks. Introduce a new
debugging-only interface LWLockAnyHeldByMe() to avoid that.
This problem was noted by Tom and Andres while reviewing changes to
support the new shared memory stats system, and later showed up in
reality while working on commit 389869af.
Back-patch to 11, where dshash.c arrived.
Reported-by: Tom Lane <tgl@sss.pgh.pa.us>
Reported-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>
Reviewed-by: Zhihong Yu <zyu@yugabyte.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/20220311012712.botrpsikaufzteyt@alap3.anarazel.de
Discussion: https://postgr.es/m/CA%2BhUKGJ31Wce6HJ7xnVTKWjFUWQZPBngxfJVx4q0E98pDr3kAw%40mail.gmail.com
40af10b57 changed things so we make use of a generation memory context for
storing tuples to be sorted by tuplesort.c. That change does not play
nicely with the changes made in 9f03ca915 (back in 2014). That commit
changed things so that index_form_tuple() is called while switched into
the tuplestore's tuplecontext. In order to fetch the tuple from the index,
index_form_tuple() must do various memory allocations which are unrelated
to the storage of the final returned tuple. Although all of these
allocations are pfree'd, the fact that we now use a generation context
means that the memory for these pfree'd allocations won't be used again by
any other allocation due to generation.c's lack of freelists. This could
result in sorts used for building indexes exceeding maintenance_work_mem
by a very large amount.
Here we fix it so we no longer allocate anything apart from the tuple
itself into the generation context by adding a new version of
index_form_tuple named index_form_tuple_context, which can be called to
specify the MemoryContext to allocate the tuple into.
Discussion: https://postgr.es/m/CAApHDvrHQkiFRHiGiAS-LMOvJN-eK-s762=tVzBz8ZqUea-a_A@mail.gmail.com
Backpatch-through: 15, where 40af10b57 was added.
50e17ad28 increased the size of ExprEvalStep from 64 bytes up to 88 bytes.
Lots of effort was spent during the development of the current expression
evaluation code to make an instance of this struct as small as possible.
Making this struct larger than needed reduces CPU cache efficiency during
expression evaluation which causes noticeable slowdowns during query
execution.
In order to reduce the size of the struct, here we remove the fn_addr
field. The values from this field can be obtained via fcinfo, just with
some extra pointer dereferencing. The extra indirection does not seem to
cause any noticeable slowdowns.
Various other fields have been moved into the ScalarArrayOpExprHashTable
struct. These fields are only used when the ScalarArrayOpExprHashTable
pointer has already been dereferenced, so no additional pointer
dereferences occur for these. Here we also make hash_fcinfo_data the last
field in ScalarArrayOpExprHashTable so that we can avoid a further pointer
dereference to get the FunctionCallInfoBaseData. This also saves a call to
palloc().
50e17ad28 was added in 14, but it's too late to adjust the size of the
ExprEvalStep in that version, so here we just backpatch to 15, which is
currently in beta.
Author: Andres Freund, David Rowley
Discussion: https://postgr.es/m/20220616233130.rparivafipt6doj3@alap3.anarazel.de
Backpatch-through: 15
The new expression step types increased the size of ExprEvalStep by ~4 for all
types of expression steps, slowing down expression evaluation noticeably. Move
them out of line.
There's other issues with these expression steps, but addressing them is
largely independent of this aspect.
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Andrew Dunstan <andrew@dunslane.net>
Discussion: https://postgr.es/m/20220616233130.rparivafipt6doj3@alap3.anarazel.de
Backpatch: 15-
This reverts most of 91c0570a791, f28bf667f60, fe0972ee5e6, afdeff10526. The
only thing left is the retry loop in 019_replslot_limit.pl that avoids
spurious failures by retrying a couple times.
We haven't seen any hard evidence that this is caused by anything but slow
process shutdown. We did not find any cases where walsenders did not vanish
after waiting for longer. Therefore there's no reason for this debugging code
to remain.
Discussion: https://postgr.es/m/20220530190155.47wr3x2prdwyciah@alap3.anarazel.de
Backpatch: 15-
TransactionIdIsInProgress had a fast path to return 'false' if the
single-item CLOG cache said that the transaction was known to be
committed. However, that was wrong, because a transaction is first
marked as committed in the CLOG but doesn't become visible to others
until it has removed its XID from the proc array. That could lead to an
error:
ERROR: t_xmin is uncommitted in tuple to be updated
or for an UPDATE to go ahead without blocking, before the previous
UPDATE on the same row was made visible.
The window is usually very short, but synchronous replication makes it
much wider, because the wait for synchronous replica happens in that
window.
Another thing that makes it hard to hit is that it's hard to get such
a commit-in-progress transaction into the single item CLOG cache.
Normally, if you call TransactionIdIsInProgress on such a transaction,
it determines that the XID is in progress without checking the CLOG
and without populating the cache. One way to prime the cache is to
explicitly call pg_xact_status() on the XID. Another way is to use a
lot of subtransactions, so that the subxid cache in the proc array is
overflown, making TransactionIdIsInProgress rely on pg_subtrans and
CLOG checks.
This has been broken ever since it was introduced in 2008, but the race
condition is very hard to hit, especially without synchronous
replication. There were a couple of reports of the error starting from
summer 2021, but no one was able to find the root cause then.
TransactionIdIsKnownCompleted() is now unused. In 'master', remove it,
but I left it in place in backbranches in case it's used by extensions.
Also change pg_xact_status() to check TransactionIdIsInProgress().
Previously, it only checked the CLOG, and returned "committed" before
the transaction was actually made visible to other queries. Note that
this also means that you cannot use pg_xact_status() to reproduce the
bug anymore, even if the code wasn't fixed.
Report and analysis by Konstantin Knizhnik. Patch by Simon Riggs, with
the pg_xact_status() change added by me.
Author: Simon Riggs
Reviewed-by: Andres Freund
Discussion: https://www.postgresql.org/message-id/flat/4da7913d-398c-e2ad-d777-f752cf7f0bbb%40garret.ru
Previously, we encoded both NULL and the first byte at the base address
as 0. That confusion led to the assertion in commit e07d4ddc, which
failed when min_dynamic_shared_memory was used. Give them distinct
encodings, by switching to 1-based offsets for non-NULL pointers. Also
improve macro hygiene in passing (missing/misplaced parentheses), and
remove open-coded access to the raw offset value from freepage.c/h.
Although e07d4ddc was back-patched to 10, the only code that actually
makes use of relptr at the base address arrived in 84b1c63a, so no need
to back-patch further than 14 for now.
Reported-by: Justin Pryzby <pryzby@telsasoft.com>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Discussion: https://postgr.es/m/20220519193839.GT19626%40telsasoft.com
This reverts commits 5753d4ee32 and fe60b67250 that modified HOT to
ignore BRIN indexes. The commit message for 5753d4ee32 claims that:
When determining whether an index update may be skipped by using
HOT, we can ignore attributes indexed only by BRIN indexes. There
are no index pointers to individual tuples in BRIN, and the page
range summary will be updated anyway as it relies on visibility
info.
This is partially incorrect - it's true BRIN indexes don't point to
individual tuples, so HOT chains are not an issue, but the visibitlity
info is not sufficient to keep the index up to date. This can easily
result in corrupted indexes, as demonstrated in the hackers thread.
This does not mean relaxing the HOT restrictions for BRIN is a lost
cause, but it needs to handle the two aspects (allowing HOT chains and
updating the page range summaries) as separate. But that requires a
major changes, and it's too late for that in the current dev cycle.
Reported-by: Tomas Vondra
Discussion: https://postgr.es/m/05ebcb44-f383-86e3-4f31-0a97a55634cf@enterprisedb.com
We were not updating the partition map cache in the subscriber even when
the corresponding remote rel is changed. Due to this data was getting
incorrectly replicated for partition tables after the publisher has
changed the table schema.
Fix it by resetting the required entries in the partition map cache after
receiving a new relation mapping from the publisher.
Reported-by: Shi Yu
Author: Shi Yu, Hou Zhijie
Reviewed-by: Amit Langote, Amit Kapila
Backpatch-through: 13, where it was introduced
Discussion: https://postgr.es/m/OSZPR01MB6310F46CD425A967E4AEF736FDA49@OSZPR01MB6310.jpnprd01.prod.outlook.com
The original advice for hard-wired SetConfigOption calls was to use
PGC_S_OVERRIDE, particularly for PGC_INTERNAL GUCs. However,
that's really overkill for PGC_INTERNAL GUCs, since there is no
possibility that we need to override a user-provided setting.
Instead use PGC_S_DYNAMIC_DEFAULT in most places, so that the
value will appear with source = 'default' in pg_settings and thereby
not be shown by psql's new \dconfig command. The one exception is
that when changing in_hot_standby in a hot-standby session, we still
use PGC_S_OVERRIDE, because people felt that seeing that in \dconfig
would be a good thing.
Similarly use PGC_S_DYNAMIC_DEFAULT for the auto-tune value of
wal_buffers (if possible, that is if wal_buffers wasn't explicitly
set to -1), and for the typical 2MB value of max_stack_depth.
In combination these changes remove four not-very-interesting
entries from the typical output of \dconfig, all of which people
fingered as "why is that showing up?" in the discussion thread.
Discussion: https://postgr.es/m/3118455.1649267333@sss.pgh.pa.us
With the old logic, when the reciever had not yet attached, we would
never call shm_mq_inc_bytes_written(), even if force_flush = true
was specified. That could result in a situation where data that the
sender believes it has sent is never received.
Along the way, remove a useless function prototype for a nonexistent
function from shm_mq.h.
Commit 46846433a03dff4f2e08c8a161e54a842da360d6 introduced these
problems.
Pavan Deolasee, with a few changes by me.
Discussion: https://postgr.es/m/CABOikdPkwtLLCTnzzmpSMXo3QZa2yXq0J7Q61ssdLFAJYrOVvQ@mail.gmail.com
Commit 1a36bc9db (SQL/JSON query functions) introduced STRING as a
type_func_name_keyword, thereby breaking applications that use
"string" as a table name, column name, function parameter name, etc.
That seems like a pretty bad thing, not least because the SQL spec
says that STRING is an unreserved keyword.
This is easy enough to fix so far as the core grammar is concerned.
However, doing so causes some ECPG test cases to fail, specifically
those that use "string" as a typedef name. It turns out this is
because portions of the ECPG grammar allow type_func_name_keywords
but not unreserved_keywords as typedef names. That's pretty horrid,
and it's mildly astonishing that we've not heard complaints about it
before. We can fix two of those uses trivially, but the ones in the
var_type production are less easy. As a stopgap, hard-code STRING as
an allowed alternative in var_type.
Per report from Alastair McKinley.
Discussion: https://postgr.es/m/3661437.1653855582@sss.pgh.pa.us
Several places in the planner tried to clamp a double value to fit
in a "long" by doing
(long) Min(x, (double) LONG_MAX);
This is subtly incorrect, because it casts LONG_MAX to double and
potentially back again. If long is 64 bits then the double value
is inexact, and the platform might round it up to LONG_MAX+1
resulting in an overflow and an undesirably negative output.
While it's not hard to rewrite the expression into a safe form,
let's put it into a common function to reduce the risk of someone
doing it wrong in future.
In principle this is a bug fix, but since the problem could only
manifest with group count estimates exceeding 2^63, it seems unlikely
that anyone has actually hit this or will do so anytime soon. We're
fixing it mainly to satisfy fuzzer-type tools. That being the case,
a HEAD-only fix seems sufficient.
Andrey Lepikhov
Discussion: https://postgr.es/m/ebbc2efb-7ef9-bf2f-1ada-d6ec48f70e58@postgrespro.ru
This is a slight, convenient semantics change from what commit
0f0cfb494004 ("Fix parallel operations that prevent oldest xmin from
advancing") introduced that lets us simplify the coding in the one place
where it is used.
Backpatch to 13. This is related to commit 6fea65508a1a ("Tighten
ComputeXidHorizons' handling of walsenders") rewriting the code site
where this is used, which has not yet been backpatched, but it may well
be in the future.
Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/202204191637.eldwa2exvguw@alvherre.pgsql
Commit 923def9a53 and 52e4f0cd47 allowed to specify column lists and row
filters for publication tables. This commit extends the
pg_publication_tables view and pg_get_publication_tables function to
display that information.
This information will be useful to users and we also need this for the
later commit that prohibits combining multiple publications with different
column lists for the same table.
Author: Hou Zhijie
Reviewed By: Amit Kapila, Alvaro Herrera, Shi Yu, Takamichi Osumi
Discussion: https://postgr.es/m/202204251548.mudq7jbqnh7r@alvherre.pgsql
I started out with the intention to rename value_type to item_type to
avoid a collision with a typedef name that appears on some platforms.
Along the way, I noticed that the adjacent field "format" was not being
correctly handled by the backend/nodes/ infrastructure functions:
copyfuncs.c erroneously treated it as a scalar, while equalfuncs,
outfuncs, and readfuncs omitted handling it at all. This looks like
it might be cosmetic at the moment because the field is always NULL
after parse analysis; but that's likely a bug in itself, and the code's
certainly not very future-proof. Let's fix it while we can still do so
without forcing an initdb on beta testers.
Further study found a few other inconsistencies in the backend/nodes/
infrastructure for the recently-added JSON node types, so fix those too.
catversion bumped because of potential change in stored rules.
Discussion: https://postgr.es/m/526703.1652385613@sss.pgh.pa.us
Currently, preloaded libraries are expected to request additional
shared memory and LWLocks in _PG_init(). However, it is not unusal
for such requests to depend on MaxBackends, which won't be
initialized at that time. Such requests could also depend on GUCs
that other modules might change. This introduces a new hook where
modules can safely use MaxBackends and GUCs to request additional
shared memory and LWLocks.
Furthermore, this change restricts requests for shared memory and
LWLocks to this hook. Previously, libraries could make requests
until the size of the main shared memory segment was calculated.
Unlike before, we no longer silently ignore requests received at
invalid times. Instead, we FATAL if someone tries to request
additional shared memory or LWLocks outside of the hook.
Nathan Bossart and Julien Rouhaud
Discussion: https://postgr.es/m/20220412210112.GA2065815%40nathanxps13
Discussion: https://postgr.es/m/Yn2jE/lmDhKtkUdr@paquier.xyz
