The problem is that we don't send keep-alive messages for a long time
while processing large transactions during logical replication where we
don't send any data of such transactions. This can happen when the table
modified in the transaction is not published or because all the changes
got filtered. We do try to send the keep_alive if necessary at the end of
the transaction (via WalSndWriteData()) but by that time the
subscriber-side can timeout and exit.
To fix this we try to send the keepalive message if required after
processing certain threshold of changes.
Reported-by: Fabrice Chapuis
Author: Wang wei and Amit Kapila
Reviewed By: Masahiko Sawada, Euler Taveira, Hou Zhijie, Hayato Kuroda
Backpatch-through: 10
Discussion: https://postgr.es/m/CAA5-nLARN7-3SLU_QUxfy510pmrYK6JJb=bk3hcgemAM_pAv+w@mail.gmail.com
Presently, the server may emit a variety of log messages when inspecting
a runtime-computed GUC, mostly in the shape of one LOG message with the
default configuration, related to the startup sequence launched as such
GUCs require a load of the control file and of external shared
libraries.
For example, the server will always emit a "database system is shut
down" LOG (unless the user has set log_min_messages higher than LOG),
which is an annoying behavior as "postgres -C" is expected to only emit
in its output the parameter value we are looking for. The parameter
value is sent to stdout, while the logs are sent to stderr so we could
recommend to use a redirection, but there was not much love for this
workaround either.
To avoid such extra log messages, per discussion, this change sets
log_min_messages to FATAL internally when -C is used on a
runtime-computed GUC (even if set to PANIC in postgresql.conf). At
FATAL, the user will still receive messages explaining why a GUC value
cannot be inspected, and will know if the command is attempted on a
server already running, something not supported yet for a
runtime-computed GUC.
Reported-by: Magnus Hagander, Bruce Momjian
Author: Nathan Bossart, Michael Paquier
Discussion: https://postgr.es/m/Yni6ZHkGotUU+RSf@paquier.xyz
The current setup assumes that commands for lz4, zstd and gzip always
exist by default if not enforced by a user's environment. However,
vcpkg, as one example, installs libraries but no binaries, so this
default setup to assume that a command should always be present would
cause failures. This commit improves the detection of such external
commands as follows:
* If a ENV value is available, trust the environment/user and use it.
* If a ENV value is not available, check its execution by looking in the
current PATH, by launching a simple "$command --version" (that should be
portable enough).
** On execution failure, ignore ENV{command}.
** On execution success, set ENV{command} = "$command".
Note that this new rule applies to gzip, lz4 and zstd but not tar that
we assume will always exist. Those commands are set up in the
environment only when using bincheck and taptest. The CI includes all
those commands and I have checked that their setup is correct there. I
have also tested this change in a MSVC environment where we have none of
those commands.
While on it, remove the references to lz4 from the documentation and
vcregress.pl in ~v13. --with-lz4 has been added in v14~ so there is no
point to have this information in these older branches.
Reported-by: Andrew Dunstan
Reviewed-by: Andrew Dunstan
Discussion: https://postgr.es/m/14402151-376b-a57a-6d0c-10ad12608e12@dunslane.net
Backpatch-through: 10
697492434 added 3 new quicksort specialization functions for common
datatypes.
That commit was not very consistent in how it would determine if we're
compiling for 32-bit or 64-bit machines. It would sometimes use
USE_FLOAT8_BYVAL and at other times check if SIZEOF_DATUM == 8. This
could cause theoretical problems due to the way USE_FLOAT8_BYVAL is now
defined based on SIZEOF_VOID_P >= 8. If pointers for some reason were
ever larger than 8-bytes then we'd end up doing 32-bit comparisons
mistakenly. Let's just always check SIZEOF_DATUM >= 8.
It also seems that ssup_datum_signed_cmp is just never used on 32-bit
builds, so let's just ifdef that out to make sure we never accidentally
use that comparison function on such machines. This also allows us to
ifdef out 1 of the 3 new specialization quicksort functions in 32-bit
builds which seems to shrink down the binary by over 4KB on my machine.
In passing, also add the missing DatumGetInt32() / DatumGetInt64() macros
in the comparison functions.
Discussion: https://postgr.es/m/CAApHDvqcQExRhtRa9hJrJB_5egs3SUfOcutP3m+3HO8A+fZTPA@mail.gmail.com
Reviewed-by: John Naylor
In OpenLDAP 2.5 and later, libldap itself is always thread-safe and
there's never a libldap_r. Our existing coding dealt with that
by assuming it wouldn't find libldap_r if libldap is thread-safe.
But that rule fails to cope if there are multiple OpenLDAP versions
visible, as is likely to be the case on macOS in particular. We'd
end up using shiny new libldap in the backend and a hoary libldap_r
in libpq.
Instead, once we've found libldap, check if it's >= 2.5 (by
probing for a function introduced then) and don't bother looking
for libldap_r if so. While one can imagine library setups that
this'd still give the wrong answer for, they seem unlikely to
occur in practice.
Per report from Peter Eisentraut. Back-patch to all supported branches.
Discussion: https://postgr.es/m/fedacd7c-2a38-25c9-e7ff-dea549d0e979@enterprisedb.com
Several keywords were missing or misclassified in the original
SQL:2016 standard. This has been corrected in later technical
corrigenda. This change updates the PostgreSQL documentation
accordingly.
This also fixes a mistake in 606948b058d: The keywords JSON_SCALAR and
JSON_SERIALIZE added there are not from SQL:2016 but from future
SQL:202x, so they don't belong in that list yet.
(606948b058 also added JSON to the reserved list, which is what the
corrigendum also does, but failed to remove it from the nonreserved
list.)
This commit addresses the following issues in the TAP tests of
pg_upgrade, introduced in 322becb:
- Remove --port and --host for commands that already rely on a node's
environment PGHOST and PGPORT.
- Switch from run_log() to command_ok(), as all the commands executed in
the tests should succeed.
- Change EXTRA_REGRESS_OPTS to make it count as a shell fragment (fixing
s/OPT/OPTS on a way), to be compatible with the various Makefiles using
it as well as 027_stream_regress.pl in the recovery tests. The command
built for the execution the pg_regress command is reformatted, while on
it, to map with the recovery test doing the same thing (we should
refactor and consolidate that in the future, perhaps).
- Re-add the test for database names stressing the behavior of
backslashes with double quotes, mostly here for Windows.
Tests doable with the upgrade across different major versions still work
the same way.
Reported-by: Noah Misch
Discussion: https://postgr.es/m/20220502042718.GB1565149@rfd.leadboat.com
The parser code that transformed VALUES from row-oriented to
column-oriented lists failed if there were zero columns.
You can't write that straightforwardly (though probably you
should be able to), but the case can be reached by expanding
a "tab.*" reference to a zero-column table.
Per bug #17477 from Wang Ke. Back-patch to all supported branches.
Discussion: https://postgr.es/m/17477-0af3c6ac6b0a6ae0@postgresql.org
This reverts commit eafdf9de06
and its back-branch counterparts. Corey Huinker pointed out that
we'd discussed this exact change back in 2016 and rejected it,
on the grounds that there's at least one usage pattern with LIMIT
where an infinite endpoint can usefully be used. Perhaps that
argument needs to be re-litigated, but there's no time left before
our back-branch releases. To keep our options open, restore the
status quo ante; if we do end up deciding to change things, waiting
one more quarter won't hurt anything.
Rather than just doing a straight revert, I added a new test case
demonstrating the usage with LIMIT. That'll at least remind us of
the issue if we forget again.
Discussion: https://postgr.es/m/3603504.1652068977@sss.pgh.pa.us
Discussion: https://postgr.es/m/CADkLM=dzw0Pvdqp5yWKxMd+VmNkAMhG=4ku7GnCZxebWnzmz3Q@mail.gmail.com
It intended to, but did not, achieve this. Adopt the new standard of
setting user ID just after locking the relation. Back-patch to v10 (all
supported versions).
Reviewed by Simon Riggs. Reported by Alvaro Herrera.
Security: CVE-2022-1552
When a feature enumerates relations and runs functions associated with
all found relations, the feature's user shall not need to trust every
user having permission to create objects. BRIN-specific functionality
in autovacuum neglected to account for this, as did pg_amcheck and
CLUSTER. An attacker having permission to create non-temp objects in at
least one schema could execute arbitrary SQL functions under the
identity of the bootstrap superuser. CREATE INDEX (not a
relation-enumerating operation) and REINDEX protected themselves too
late. This change extends to the non-enumerating amcheck interface.
Back-patch to v10 (all supported versions).
Sergey Shinderuk, reviewed (in earlier versions) by Alexander Lakhin.
Reported by Alexander Lakhin.
Security: CVE-2022-1552
If a cluster is promoted (aka the control file shows a state different
than DB_IN_ARCHIVE_RECOVERY) while CreateRestartPoint() is still
processing, this function could miss an update of the control file for
"checkPoint" and "checkPointCopy" but still do the recycling and/or
removal of the past WAL segments, assuming that the to-be-updated LSN
values should be used as reference points for the cleanup. This causes
a follow-up restart attempting crash recovery to fail with a PANIC on a
missing checkpoint record if the end-of-recovery checkpoint triggered by
the promotion did not complete while the cluster abruptly stopped or
crashed before the completion of this checkpoint. The PANIC would be
caused by the redo LSN referred in the control file as located in a
segment already gone, recycled by the previous restartpoint with
"checkPoint" out-of-sync in the control file.
This commit fixes the update of the control file during restartpoints so
as "checkPoint" and "checkPointCopy" are updated even if the cluster has
been promoted while a restartpoint is running, to be on par with the set
of WAL segments actually recycled in the end of CreateRestartPoint().
This problem exists in all the stable branches. However, commit
7ff23c6, by removing the last call of CreateCheckPoint() from the
startup process, has made this bug much easier to reason about as
concurrent checkpoints are not possible anymore. No backpatch is done
yet, mostly out of caution from me as a point release is close by, but
we need to think harder about the case of concurrent checkpoints at
promotion if the bgwriter is not considered as running by the startup
process in ~v14, so this change is done only on HEAD for the moment.
Reported-by: Fujii Masao, Rui Zhao
Author: Kyotaro Horiguchi
Reviewed-by: Nathan Bossart, Michael Paquier
Discussion: https://postgr.es/m/20220316.102444.2193181487576617583.horikyota.ntt@gmail.com
Commits 4eb21763 and b74e94dc introduced a way to force every backend to
close all relation files, to fix an ancient Windows-only bug.
This commit extends that behavior to all operating systems and adds
a couple of extra barrier points, to fix a totally different class of
bug: the reuse of relfilenodes in scenarios that have no other kind of
cache invalidation to prevent file descriptor mix-ups.
In all releases, data corruption could occur when you moved a database
to another tablespace and then back again. Despite that, no back-patch
for now as the infrastructure required is too new and invasive. In
master only, since commit aa010514, it could also happen when using
CREATE DATABASE with a user-supplied OID or via pg_upgrade.
Author: Andres Freund <andres@anarazel.de>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Discussion: https://postgr.es/m/20220209220004.kb3dgtn2x2k2gtdm%40alap3.anarazel.de
With sufficiently bad luck, it was possible for IssuePendingWritebacks()
to reopen a file after we'd processed PROCSIGNAL_BARRIER_SMGRRELEASE and
before the file was unlinked by some other backend. That left a small
hole in commit 4eb21763's plan to fix all spurious errors from DROP
TABLESPACE and similar on Windows.
Fix by closing md.c's segments, instead of just closing fd.c's
descriptors, and then teaching smgrwriteback() not to open files that
aren't already open.
Reported-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Discussion: https://postgr.es/m/20220209220004.kb3dgtn2x2k2gtdm%40alap3.anarazel.de
Since 6bc8ef0b7f, the maximum number
of backends can't change as background workers are registered, but
these comments still reflect the way things worked prior to that.
Also, per recent discussion, some modules call SetConfigOption()
from _PG_init(). It's not entirely clear to me whether we want to
regard that as a fully supported operation, but since we know it's
a thing that happens, it at least deserves a mention in the comments,
so add that.
Nathan Bossart, reviewed by Anton A. Melnikov
Discussion: http://postgr.es/m/20220419154658.GA2487941@nathanxps13
The PXE_CIPHER_INIT error is used to report initialization errors, so
appending a questionmark to the error isn't entirely accurate (using a
space before the questionmark doubly so).
Discussion: https://postgr.es/m/C89D932C-501E-4473-9750-638CFCD9095E@yesql.se
Report OpenSSL errors during initialization as PXE_CIPHER_INIT since
that's just what they were, and not generic unknown errors. This also
removes the last users of the generic error, and thus it can be removed.
Discussion: http://postgr.es/m/C89D932C-501E-4473-9750-638CFCD9095E@yesql.se
Setting up an EVP context for ciphers banned under FIPS generate
two OpenSSL errors in the queue, and as we only consume one from
the queue the other is at the head for the next invocation:
postgres=# select md5('foo');
ERROR: could not compute MD5 hash: unsupported
postgres=# select md5('foo');
ERROR: could not compute MD5 hash: initialization error
Clearing the error queue when creating the context ensures that
we don't pull in an error from an earlier operation.
Discussion: https://postgr.es/m/C89D932C-501E-4473-9750-638CFCD9095E@yesql.se
SubqueryScan was always getting labeled with a rowcount estimate
appropriate for non-parallel cases. However, nodes that are
underneath a Gather should be treated as processing only one
worker's share of the rows, whether the particular node is explicitly
parallel-aware or not. Most non-scan-level node types get this
right automatically because they base their rowcount estimate on
that of their input sub-Path(s). But SubqueryScan didn't do that,
instead using the whole-relation rowcount estimate as if it were
a non-parallel-aware scan node. If there is a parallel-aware node
below the SubqueryScan, this is wrong, and it results in inflating
the cost estimates for nodes above the SubqueryScan, which can cause
us to not choose a parallel plan, or choose a silly one --- as indeed
is visible in the one regression test whose results change with this
patch. (Although that plan tree appears to contain no SubqueryScans,
there were some in it before setrefs.c deleted them.)
To fix, use path->subpath->rows not baserel->tuples as the number
of input tuples we'll process. This requires estimating the quals'
selectivity afresh, which is slightly annoying; but it shouldn't
really add much cost thanks to the caching done in RestrictInfo.
This is pretty clearly a bug fix, but I'll refrain from back-patching
as people might not appreciate plan choices changing in stable branches.
The fact that it took us this long to identify the bug suggests that
it's not a major problem.
Per report from bucoo, though this is not his proposed patch.
Discussion: https://postgr.es/m/202204121457159307248@sohu.com
The test for lz4.h used AC_CHECK_HEADERS, but nothing was using the
resulting symbol HAVE_LZ4_H. Change this to use AC_CHECK_HEADER
instead. This was probably an oversight, seeing that the nearby
similar tests do this correctly.
The testclient and uri-regress programs in the libpq test suite had
quite generic names which didn't convey much meaning. Since they are
installed as part of the MSVC test runs, ensure that their purpose
is a little bit clearer by renaming with a libpq_ prefix. While at
it rename uri-regress to uri_regress to avoid mixing dash and under-
score in the same filename.
Reported-by: Noah Misch <noah@leadboat.com>
Discussion: https://postgr.es/m/20220501080706.GA1542365@rfd.leadboat.com
The tests added in 9f8a050f68 failed nearly reliably on FreeBSD in CI, and
occasionally on the buildfarm. That turns out to be caused not by a bug in the
test, but by a longstanding bug in recovery conflict handling.
The standby timeout handler, used by ResolveRecoveryConflictWithBufferPin(),
executed SendRecoveryConflictWithBufferPin() inside a signal handler. A bad
idea, because the deadlock timeout handler (or a spurious latch set) could
have interrupted ProcWaitForSignal(). If unlucky that could cause a
self-deadlock on ProcArrayLock, if the deadlock check is in
SendRecoveryConflictWithBufferPin()->CancelDBBackends().
To fix, set a flag in StandbyTimeoutHandler(), and check the flag in
ResolveRecoveryConflictWithBufferPin().
Subsequently the recovery conflict tests will be backpatched.
Discussion: https://postgr.es/m/20220413002626.udl7lll7f3o7nre7@alap3.anarazel.de
Backpatch: 10-
The recovery conflict tests added in 9f8a050f68 surfaced a bug in the
interaction between buffer pin and deadlock recovery conflicts. To make sure
that the bugfix won't break deadlock conflict detection, add a test for that
scenario.
031_recovery_conflict.pl will later be backpatched, with this included.
Discussion: https://postgr.es/m/20220413002626.udl7lll7f3o7nre7@alap3.anarazel.de
Instability in the test for pg_walinspect revealed that
pg_get_wal_records_info_till_end_of_wal(x) would try to decode all the
records with a start LSN earlier than the flush LSN, even though that
might include a partial record at the end of the range. In that case,
read_local_xlog_page_no_wait() would return NULL when it tried to read
past the flush LSN, which would be interpreted as an error by the
caller. That caused a test failure only on a BF animal that had been
restarted recently, but could be expected to happen in the wild quite
easily depending on the alignment of various parameters.
Fix by using private data in read_local_xlog_page_no_wait() to signal
end-of-wal to the caller, so that it can be properly distinguished
from a real error.
Discussion: https://postgr.es/m/Ymd/e5eeZMNAkrXo%40paquier.xyz
Discussion: https://postgr.es/m/111657.1650910309@sss.pgh.pa.us
Authors: Thomas Munro, Bharath Rupireddy.
Amit Kapila