1cbbee033 added BUFFER_USAGE_LIMIT to the VACUUM and ANALYZE commands, so
here we permit that option to be specified in vacuumdb.
In passing, adjust the documents for vacuum_buffer_usage_limit and the
BUFFER_USAGE_LIMIT VACUUM option to mention "kB" rather than "KB". Do the
same for the ERROR message in ExecVacuum() and
check_vacuum_buffer_usage_limit(). Without that we might tell a user that
the valid minimum value is 128 KB only to reject that because we accept
only "kB" and not "KB".
Also, add a small reminder comment in vacuum.h to try to trigger the
memory of anyone adding new fields to VacuumParams that they might want to
consider if vacuumdb needs to grow a new option too.
Author: Melanie Plageman
Reviewed-by: Justin Pryzby
Reviewed-by: David Rowley
Discussion: https://postgr.es/m/ZAzTg3iEnubscvbf@telsasoft.com
While we already had some form of bulk extension for relations, it was fairly
limited. It only amortized the cost of acquiring the extension lock, the
relation itself was still extended one-by-one. Bulk extension was also solely
triggered by contention, not by the amount of data inserted.
To address this, use ExtendBufferedRelBy(), introduced in 31966b151e, to
extend the relation. We try to extend the relation by multiple blocks in two
situations:
1) The caller tells RelationGetBufferForTuple() that it will need multiple
pages. For now that's only used by heap_multi_insert(), see commit FIXME.
2) If there is contention on the extension lock, use the number of waiters for
the lock as a multiplier for the number of blocks to extend by. This is
similar to what we already did. Previously we additionally multiplied the
numbers of waiters by 20, but with the new relation extension
infrastructure I could not see a benefit in doing so.
Using the freespacemap to provide empty pages can cause significant
contention, and adds measurable overhead, even if there is no contention. To
reduce that, remember the blocks the relation was extended by in the
BulkInsertState, in the extending backend. In case 1) from above, the blocks
the extending backend needs are not entered into the FSM, as we know that we
will need those blocks.
One complication with using the FSM to record empty pages, is that we need to
insert blocks into the FSM, when we already hold a buffer content lock. To
avoid doing IO while holding a content lock, release the content lock before
recording free space. Currently that opens a small window in which another
backend could fill the block, if a concurrent VACUUM records the free
space. If that happens, we retry, similar to the already existing case when
otherBuffer is provided. In the future it might be worth closing the race by
preventing VACUUM from recording the space in newly extended pages.
This change provides very significant wins (3x at 16 clients, on my
workstation) for concurrent COPY into a single relation. Even single threaded
COPY is measurably faster, primarily due to not dirtying pages while
extending, if supported by the operating system (see commit 4d330a61bb). Even
single-row INSERTs benefit, although to a much smaller degree, as the relation
extension lock rarely is the primary bottleneck.
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
Add new options to the VACUUM and ANALYZE commands called
BUFFER_USAGE_LIMIT to allow users more control over how large to make the
buffer access strategy that is used to limit the usage of buffers in
shared buffers. Larger rings can allow VACUUM to run more quickly but
have the drawback of VACUUM possibly evicting more buffers from shared
buffers that might be useful for other queries running on the database.
Here we also add a new GUC named vacuum_buffer_usage_limit which controls
how large to make the access strategy when it's not specified in the
VACUUM/ANALYZE command. This defaults to 256KB, which is the same size as
the access strategy was prior to this change. This setting also
controls how large to make the buffer access strategy for autovacuum.
Per idea by Andres Freund.
Author: Melanie Plageman
Reviewed-by: David Rowley
Reviewed-by: Andres Freund
Reviewed-by: Justin Pryzby
Reviewed-by: Bharath Rupireddy
Discussion: https://postgr.es/m/20230111182720.ejifsclfwymw2reb@awork3.anarazel.de
A future commit will use this information to determine how aggressively to
extend the relation by. In heap_multi_insert() we know accurately how many
pages we need once we need to extend the relation, providing an accurate lower
bound for how much to extend.
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
Allow autovacuum to reload the config file more often so that cost-based
delay parameters can take effect while VACUUMing a relation. Previously,
autovacuum workers only reloaded the config file once per relation
vacuumed, so config changes could not take effect until beginning to
vacuum the next table.
Now, check if a reload is pending roughly once per block, when checking
if we need to delay.
In order for autovacuum workers to safely update their own cost delay
and cost limit parameters without impacting performance, we had to
rethink when and how these values were accessed.
Previously, an autovacuum worker's wi_cost_limit was set only at the
beginning of vacuuming a table, after reloading the config file.
Therefore, at the time that autovac_balance_cost() was called, workers
vacuuming tables with no cost-related storage parameters could still
have different values for their wi_cost_limit_base and wi_cost_delay.
Now that the cost parameters can be updated while vacuuming a table,
workers will (within some margin of error) have no reason to have
different values for cost limit and cost delay (in the absence of
cost-related storage parameters). This removes the rationale for keeping
cost limit and cost delay in shared memory. Balancing the cost limit
requires only the number of active autovacuum workers vacuuming a table
with no cost-based storage parameters.
Author: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com>
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/CAAKRu_ZngzqnEODc7LmS1NH04Kt6Y9huSjz5pp7%2BDXhrjDA0gw%40mail.gmail.com
Vacuum code run both by autovacuum workers and a backend doing
VACUUM/ANALYZE previously inspected VacuumCostLimit and VacuumCostDelay,
which are the global variables backing the GUCs vacuum_cost_limit and
vacuum_cost_delay.
Autovacuum workers needed to override these variables with their
own values, derived from autovacuum_vacuum_cost_limit and
autovacuum_vacuum_cost_delay and worker cost limit balancing logic.
This led to confusing code which, in some cases, both derived and
set a new value of VacuumCostLimit from VacuumCostLimit.
In preparation for refreshing these GUC values more often, introduce
new, independent global variables and add a function to update them
using the GUCs and existing logic.
Per suggestion by Kyotaro Horiguchi
Author: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com>
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/CAAKRu_ZngzqnEODc7LmS1NH04Kt6Y9huSjz5pp7%2BDXhrjDA0gw%40mail.gmail.com
While vacuuming a table in failsafe mode, VacuumCostActive should
not be re-enabled. This currently isn't a problem because vacuum
cost parameters are only refreshed in between vacuuming tables and
failsafe status is reset for every table.
In preparation for allowing vacuum cost parameters to be updated
more frequently, elevate LVRelState->failsafe_active to a global,
VacuumFailsafeActive, which will be checked when determining whether
or not to re-enable vacuum cost-related delays.
Author: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com>
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/CAAKRu_ZngzqnEODc7LmS1NH04Kt6Y9huSjz5pp7%2BDXhrjDA0gw%40mail.gmail.com
tsquery's GETQUERY() macro is only safe to apply to a tsquery
that is known non-empty; otherwise it gives a pointer to garbage.
Before commit 5a617d75d, ts_headline() avoided this pitfall, but
only in a very indirect, nonobvious way. (hlCover could not reach
its TS_execute call, because if the query contains no lexemes
then hlFirstIndex would surely return -1.) After that commit,
it fell into the trap, resulting in weird errors such as
"unrecognized operator" and/or valgrind complaints. In HEAD,
fix this by not calling TS_execute_locations() at all for an
empty query. In the back branches, add a defensive check to
hlCover() --- that's not fixing any live bug, but I judge the
code a bit too fragile as-is.
Also, both mark_hl_fragments() and mark_hl_words() were careless
about the possibility of empty search text: in the cases where
no match has been found, they'd end up telling mark_fragment() to
mark from word indexes 0 to 0 inclusive, even when there is no
word 0. This is harmless since we over-allocated the prs->words
array, but it does annoy valgrind. Fix so that the end index is -1
and thus mark_fragment() will do nothing in such cases.
Bottom line is that this fixes a live bug in HEAD, but in the
back branches it's only getting rid of a valgrind nitpick.
Back-patch anyway.
Per report from Alexander Lakhin.
Discussion: https://postgr.es/m/c27f642d-020b-01ff-ae61-086af287c4fd@gmail.com
Starting with commit 7db0cd2145, RelationGetBufferForTuple() did a
visibilitymap_pin() while holding an exclusive buffer content lock on a newly
extended page, when using COPY FREEZE. We elsewhere try hard to avoid to doing
IO while holding a content lock. And until 14f98e0af9, that happened while
holding the relation extension lock.
Practically, this isn't a huge issue, because COPY FREEZE is restricted to
relations created or truncated in the current session, so it's unlikely
there's a lot of contention.
We can't avoid doing IO while holding the content lock by pinning the VM
earlier, because we don't know which page it will be on.
While we could just ignore the issue in this case, a future commit will add
bulk relation extension, which needs to enter pages into the FSM while also
trying to hold onto a buffer lock.
To address this issue, use visibilitymap_pin_ok() to see if the relevant
buffer is already pinned. If not, release the buffer, pin the VM buffer, and
acquire the lock again. This opens up a small window for other backends to
insert data onto the page - as the page is not entered into the freespacemap,
other backends won't see it normally, but a concurrent vacuum could enter the
page, if it started just after the relation is extended. In case the page is
used by another backend, retry. This is very similar to how locking
"otherBuffer" is already dealt with.
Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com>
Discussion: http://postgr.es/m/20230325025740.wzvchp2kromw4zqz@awork3.anarazel.de
GetVisibilityMapPins() insisted on the buffer1/buffer2 being in a specific
order. This required checks at the callsite. As a subsequent patch will add
another callsite, move related logic into GetVisibilityMapPins().
Discussion: https://postgr.es/m/20230403190030.fk2frxv6faklrseb@awork3.anarazel.de
zstd compression supports a special mode for finding matched in distant
past, which may result in better compression ratio, at the expense of
using more memory (the window size is 128MB).
To enable this optional mode, use the "long" keyword when specifying the
compression method (--compress=zstd:long).
Author: Justin Pryzby
Reviewed-by: Tomas Vondra, Jacob Champion
Discussion: https://postgr.es/m/20230224191840.GD1653@telsasoft.com
Discussion: https://postgr.es/m/20220327205020.GM28503@telsasoft.com
vacuum() is used for both the VACUUM command and for autovacuum. There
were many prechecks being done inside vacuum() that were just not relevant
to autovacuum. Let's move the bulk of these into ExecVacuum() so that
they're only executed when running the VACUUM command. This removes a
small amount of overhead when autovacuum vacuums a table.
While we are at it, allocate VACUUM's BufferAccessStrategy in ExecVacuum()
and pass it into vacuum() instead of expecting vacuum() to make it if it's
not already made by the calling function. To make this work, we need to
create the vacuum memory context slightly earlier, so we now need to pass
that down to vacuum() so that it's available for use in other memory
allocations.
Author: Melanie Plageman
Reviewed-by: David Rowley
Discussion: https://postgr.es/m/20230405211534.4skgskbilnxqrmxg@awork3.anarazel.de
A few places are not converted. Some because they are tackled in later
commits (e.g. hio.c, xlogutils.c), some because they are more
complicated (e.g. brin_pageops.c). Having a few users of ReadBuffer(P_NEW) is
good anyway, to ensure the backward compat path stays working.
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
32fbe0239 changed things so we didn't bother allocating the
BufferAccessStrategy during VACUUM (ONLY_DATABASE_STATS); and VACUUM
(FULL), however, it forgot to consider that VACUUM (FULL, ANALYZE) is a
possible combination. That change would have resulted in such a command
allowing ANALYZE to make full use of shared buffers, which wasn't
intended, so fix that.
Reported-by: Melanie Plageman
Discussion: https://postgr.es/m/CAAKRu_bJRKe+v_=OqwC+5sA3j5qv8rqdAwy3+yHaO3wmtfrCRg@mail.gmail.com
pg_stat_statements relies on EState->es_processed to count the number of
rows processed by ExecutorRun(). This proves to be a problem under the
extended query protocol when the result of a query is fetched through
more than one call of ExecutorRun(), as es_processed is reset each time
ExecutorRun() is called. This causes pg_stat_statements to report the
number of rows calculated in the last execute fetch, rather than the
global sum of all the rows processed.
As pquery.c tells, this is a problem when a portal does not use
holdStore. For example, DMLs with RETURNING would report a correct
tuple count as these do one execution cycle when the query is first
executed to fill in the portal's store with one ExecutorRun(), feeding
on the portal's store for each follow-up execute fetch depending on the
fetch size requested by the client.
The fix proposed for this issue is simple with the addition of an extra
counter in EState that's preserved across multiple ExecutorRun() calls,
incremented with the value calculated in es_processed. This approach is
not back-patchable, unfortunately.
Note that libpq does not currently give any way to control the fetch
size when using the extended v3 protocol, meaning that in-core testing
is not possible yet. This issue can be easily verified with the JDBC
driver, though, with *autocommit disabled*. Hence, having in-core tests
requires more features, left for future discussion:
- At least two new libpq routines splitting PQsendQueryGuts(), one for
the bind/describe and a second for a series of execute fetches with a
custom fetch size, likely in a fashion similar to what JDBC does.
- A psql meta-command for the execute phase. This part is not strictly
mandatory, still it could be handy.
Reported-by: Andrew Dunstan (original discovery by Simon Siggs)
Author: Sami Imseih
Reviewed-by: Tom Lane, Michael Paquier
Discussion: https://postgr.es/m/EBE6C507-9EB6-4142-9E4D-38B1673363A7@amazon.com
Discussion: https://postgr.es/m/c90890e7-9c89-c34f-d3c5-d5c763a34bd8@dunslane.net
The primary bottlenecks for relation extension are:
1) The extension lock is held while acquiring a victim buffer for the new
page. Acquiring a victim buffer can require writing out the old page
contents including possibly needing to flush WAL.
2) When extending via ReadBuffer() et al, we write a zero page during the
extension, and then later write out the actual page contents. This can
nearly double the write rate.
3) The existing bulk relation extension infrastructure in hio.c just amortized
the cost of acquiring the relation extension lock, but none of the other
costs.
Unfortunately 1) cannot currently be addressed in a central manner as the
callers to ReadBuffer() need to acquire the extension lock. To address that,
this this commit moves the responsibility for acquiring the extension lock
into bufmgr.c functions. That allows to acquire the relation extension lock
for just the required time. This will also allow us to improve relation
extension further, without changing callers.
The reason we write all-zeroes pages during relation extension is that we hope
to get ENOSPC errors earlier that way (largely works, except for CoW
filesystems). It is easier to handle out-of-space errors gracefully if the
page doesn't yet contain actual tuples. This commit addresses 2), by using the
recently introduced smgrzeroextend(), which extends the relation, without
dirtying the kernel page cache for all the extended pages.
To address 3), this commit introduces a function to extend a relation by
multiple blocks at a time.
There are three new exposed functions: ExtendBufferedRel() for extending the
relation by a single block, ExtendBufferedRelBy() to extend a relation by
multiple blocks at once, and ExtendBufferedRelTo() for extending a relation up
to a certain size.
To avoid duplicating code between ReadBuffer(P_NEW) and the new functions,
ReadBuffer(P_NEW) now implements relation extension with
ExtendBufferedRel(), using a flag to tell ExtendBufferedRel() that the
relation lock is already held.
Note that this commit does not yet lead to a meaningful performance or
scalability improvement - for that uses of ReadBuffer(P_NEW) will need to be
converted to ExtendBuffered*(), which will be done in subsequent commits.
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
A future patch will add support for extending relations by multiple blocks at
once. To be concurrency safe, the buffers for those blocks need to be marked
as BM_IO_IN_PROGRESS. Until now we only had infrastructure for recovering from
an IO error for a single buffer. This commit extends that infrastructure to
multiple buffers by using the resource owner infrastructure.
This commit increases the size of the ResourceOwnerData struct, which appears
to have a just about measurable overhead in very extreme workloads. Medium
term we are planning to substantially shrink the size of
ResourceOwnerData. Short term the increase is small enough to not worry about
it for now.
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
Discussion: https://postgr.es/m/20221029200025.w7bvlgvamjfo6z44@awork3.anarazel.de
Merge and hash joins can support antijoin with the non-nullable input
on the right, using very simple combinations of their existing logic
for right join and anti join. This gives the planner more freedom
about how to order the join. It's particularly useful for hash join,
since we may now have the option to hash the smaller table instead
of the larger.
Richard Guo, reviewed by Ronan Dunklau and myself
Discussion: https://postgr.es/m/CAMbWs48xh9hMzXzSy3VaPzGAz+fkxXXTUbCLohX1_L8THFRm2Q@mail.gmail.com
Previously we held buffer locks for two buffer mapping partitions at the same
time to change the identity of buffers. Particularly for extending relations
needing to hold the extension lock while acquiring a victim buffer is
painful.But it also creates a bottleneck for workloads that just involve
reads.
Now we instead first acquire a victim buffer and write it out, if
necessary. Then we remove that buffer from the old partition with just the old
partition's partition lock held and insert it into the new partition with just
that partition's lock held.
By separating out the victim buffer acquisition, future commits will be able
to change relation extensions to scale better.
On my workstation, a micro-benchmark exercising buffered reads strenuously and
under a lot of concurrency, sees a >2x improvement.
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
Commit 47bb9db75 taught AcquireExecutorLocks to re-acquire locks
on views using data from their RTE_SUBQUERY replacements, but
it now seems like we should make AcquirePlannerLocks do the same.
In this way, if a view has been redefined, we will notice that
a bit earlier while checking validity of a cached plan and thereby
avoid some wasted work.
Report and patch by Amit Langote.
Discussion: https://postgr.es/m/CA+HiwqH0xZOQ+GQAdKeckY1R4NOeHdzhtfxkAMJLSchpapNk5w@mail.gmail.com
This adds a few more assertions against a buffer being local in places we
don't expect, and extracts the check for a buffer being pinned exactly once
from LockBufferForCleanup() into its own function. Later commits will use this
function.
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: http://postgr.es/m/419312fd-9255-078c-c3e3-f0525f911d7f@iki.fi
smgrzeroextend() uses FileFallocate() to efficiently extend files by multiple
blocks. When extending by a small number of blocks, use FileZero() instead, as
using posix_fallocate() for small numbers of blocks is inefficient for some
file systems / operating systems. FileZero() is also used as the fallback for
FileFallocate() on platforms / filesystems that don't support fallocate.
A big advantage of using posix_fallocate() is that it typically won't cause
dirty buffers in the kernel pagecache. So far the most common pattern in our
code is that we smgrextend() a page full of zeroes and put the corresponding
page into shared buffers, from where we later write out the actual contents of
the page. If the kernel, e.g. due to memory pressure or elapsed time, already
wrote back the all-zeroes page, this can lead to doubling the amount of writes
reaching storage.
There are no users of smgrzeroextend() as of this commit. That will follow in
future commits.
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
Reviewed-by: John Naylor <john.naylor@enterprisedb.com>
Discussion: https://postgr.es/m/20221029025420.eplyow6k7tgu6he3@awork3.anarazel.de
In v13 and v14, the ENABLE/DISABLE TRIGGER USER variant malfunctioned
on cloned triggers, failing to find the clones because it thought they
were system triggers. Other variants of ENABLE/DISABLE TRIGGER would
improperly apply a superuserness check. Fix by adjusting the is-it-
a-system-trigger check to match reality in those branches. (As far
as I can find, this is the only place that got it wrong.)
There's no such bug in v15/HEAD, because we revised the catalog
representation of system triggers to be what this code was expecting.
However, add the test case to these branches anyway, because this area
is visibly pretty fragile. Also remove an obsoleted comment.
The recent v15/HEAD commit 6949b921d fixed a nearby bug. I now see
that my commit message for that was inaccurate: the behavior of
recursing to clone triggers is older than v15, but it didn't apply
to the case in v13/v14 because in those branches parent partitioned
tables have no pg_trigger entries for foreign-key triggers. But add
the test case from that commit to v13/v14, just to show what is
happening there.
Per bug #17886 from DzmitryH.
Discussion: https://postgr.es/m/17886-5406d5d828aa4aa3@postgresql.org
The read path needs to be able to initialize pages anyway, as relation
extensions are not durable. By avoiding initializing pages, we can, in a
future patch, extend the relation by multiple blocks at once.
Using smgrextend() for {vm,fsm}_extend() is not a good idea in general, as at
least one page of the VM/FSM will be read immediately after, always causing a
cache miss, requiring us to read content we just wrote.
Discussion: https://postgr.es/m/20230301223515.pucbj7nb54n4i4nv@awork3.anarazel.de
This is mainly a light renumbering to match the sections in the
standard.
The comments for SQL_IMPLEMENTATION_INFO and SQL_SIZING are no longer
applicable because the required information has been moved into part
9.
fill_seq_fork_with_data(), used to initialize a new sequence relation, only
locked the buffer after calling PageInit(), even though PageInit() modifies
page contents.
This is unlikely to cause real-world issues, as the relation is exclusively
locked at that point, and the buffer not yet marked dirty, so other processes
should not access the buffer.
This issue looks to have been present since the introduction of sequences in
e8647c45d6.
Given the low risk, it does not seem worth backpatching the fix.
Discussion: https://postgr.es/m/20230404185501.wdkmo3k7bedlx6qk@awork3.anarazel.de
Convert to BCP47 language tags before storing in the catalog, except
during binary upgrade or when the locale comes from an existing
collation or template database.
The resulting language tags can vary slightly between ICU
versions. For instance, "@colBackwards=yes" is converted to
"und-u-kb-true" in older versions of ICU, and to the simpler (but
equivalent) "und-u-kb" in newer versions.
The process of canonicalizing to a language tag also understands more
input locale string formats than ucol_open(). For instance,
"fr_CA.UTF-8" is misinterpreted by ucol_open() and the region is
ignored; effectively treating it the same as the locale "fr" and
opening the wrong collator. Canonicalization properly interprets the
language and region, resulting in the language tag "fr-CA", which can
then be understood by ucol_open().
This commit fixes a problem in prior versions due to ucol_open()
misinterpreting locale strings as described above. For instance,
creating an ICU collation with locale "fr_CA.UTF-8" would store that
string directly in the catalog, which would later be passed to (and
misinterpreted by) ucol_open(). After this commit, the locale string
will be canonicalized to language tag "fr-CA" in the catalog, which
will be properly understood by ucol_open(). Because this fix affects
the resulting collator, we cannot change the locale string stored in
the catalog for existing databases or collations; otherwise we'd risk
corrupting indexes. Therefore, only canonicalize locales for
newly-created (not upgraded) collations/databases. For similar
reasons, do not backport.
Discussion: https://postgr.es/m/8c7af6820aed94dc7bc259d2aa7f9663518e6137.camel@j-davis.com
Reviewed-by: Peter Eisentraut
This option is normally false, but can be set to true to obtain
the legacy behavior where the subscription runs with the permissions
of the subscription owner rather than the permissions of the
table owner. The advantages of this mode are (1) it doesn't require
that the subscription owner have permission to SET ROLE to each
table owner and (2) since no role switching occurs, the
SECURITY_RESTRICTED_OPERATION restrictions do not apply.
On the downside, it allows any table owner to easily usurp
the privileges of the subscription owner - basically, to take
over their account. Because that's generally quite undesirable,
we don't make this mode the default, but we do make it available,
just in case the new behavior causes too many problems for someone.
Discussion: http://postgr.es/m/CA+TgmoZ-WEeG6Z14AfH7KhmpX2eFh+tZ0z+vf0=eMDdbda269g@mail.gmail.com
Up until now, logical replication actions have been performed as the
subscription owner, who will generally be a superuser. Commit
cec57b1a0f documented hazards
associated with that situation, namely, that any user who owns a
table on the subscriber side could assume the privileges of the
subscription owner by attaching a trigger, expression index, or
some other kind of executable code to it. As a remedy, it suggested
not creating configurations where users who are not fully trusted
own tables on the subscriber.
Although that will work, it basically precludes using logical
replication in the way that people typically want to use it,
namely, to replicate a database from one node to another
without necessarily having any restrictions on which database
users can own tables. So, instead, change logical replication to
execute INSERT, UPDATE, DELETE, and TRUNCATE operations as the
table owner when they are replicated.
Since this involves switching the active user frequently within
a session that is authenticated as the subscription user, also
impose SECURITY_RESTRICTED_OPERATION restrictions on logical
replication code. As an exception, if the table owner can SET
ROLE to the subscription owner, these restrictions have no
security value, so don't impose them in that case.
Subscription owners are now required to have the ability to
SET ROLE to every role that owns a table that the subscription
is replicating. If they don't, replication will fail. Superusers,
who normally own subscriptions, satisfy this property by default.
Non-superusers users who own subscriptions will need to be
granted the roles that own relevant tables.
Patch by me, reviewed (but not necessarily in its entirety) by
Jelte Fennema, Jeff Davis, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoaSCkg9ww9oppPqqs+9RVqCexYCE6Aq=UsYPfnOoDeFkw@mail.gmail.com
- At the last minute and for no particularly good reason, I changed the
WITHOUT token to be marked especially for lookahead, from the one in
WITHOUT TIME to the one in WITHOUT UNIQUE. Study of upcoming patches
(where a new WITHOUT ARRAY WRAPPER clause is added) showed me that the
former was better, so put it back the way the original patch had it.
- update exprTypmod() for JsonConstructorExpr to return the typmod of
the RETURNING clause, as a comment there suggested. Perhaps it's
possible for this to make a difference with datetime types, but I
didn't try to build a test case.
- The nodeFuncs.c support code for new nodes was calling walker()
directly instead of the WALK() macro as introduced by commit 1c27d16e6e.
Modernize that. Also add exprLocation() support for a couple of nodes
that missed it. Lastly, reorder the code more sensibly.
The WITHOUT_LA -> WITHOUT change means that stored rules containing
either WITHOUT TIME ZONE or WITHOUT UNIQUE KEYS would change
representation. Therefore, bump catversion.
Discussion: https://postgr.es/m/20230329181708.e64g2tpy7jyufqkr@alvherre.pgsql
The trace point was using the relfileno / fork / block for the to-be-read-in
buffer. Some upcoming work would make that more expensive to provide. We still
have buffer-flush-start/done, which can serve most tracing needs that
buffer-write-dirty could serve.
Discussion: https://postgr.es/m/f5164e7a-eef6-8972-75a3-8ac622ed0c6e@iki.fi
Commit 61b313e4 made VACUUM pass down a heaprel to index AM bulkdelete
and vacuumcleanup routines. Although this was primarily intended as
preparation for logical decoding on standbys, it also made it easy to
correct an old deficiency in how we determine how to cleanup SP-GiST
redirect and placeholder tuples.
Pass the heaprel to GlobalVisTestFor() during cleanup of redirect and
placeholder tuples, rather than pessimistically passing NULL.
Author: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
Discussion: https://postgr.es/m/02392033-f030-a3c8-c7d0-5c27eb529fec@gmail.com
Commit 61b313e4 made nbtree consistently pass down a heaprel to low
level routines like _bt_getbuf(). Although this was primarily intended
as preparation for logical decoding on standbys, it also made it easy to
correct an old deficiency in how nbtree VACUUM determines whether or not
it's now safe to recycle deleted pages.
Pass the heaprel to GlobalVisTestFor() in nbtree routines that deal with
recycle safety. nbtree now makes less pessimistic assumptions about
recycle safety within non-catalog relations. This enhancement
complements the recycling enhancement added by commit 9dd963ae25.
nbtree remains just as pessimistic as ever when it comes to recycle
safety within indexes on catalog relations. There is no fundamental
reason why we need to treat catalog relations differently, though. The
behavioral inconsistency is a consequence of the way that nbtree uses
nextXID values to implement what Lanin and Shasha call "the drain
technique". Note in particular that it has nothing to do with whether
or not index tuples might still be required for an older MVCC snapshot.
Author: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
Discussion: https://postgr.es/m/CAH2-WzkaiDxCje0yPuH=3Uh2p1V_2pFGY==xfbZoZu7Ax_NB8g@mail.gmail.com
In USE_VALGRIND builds, add code to print the text of the current query
to the valgrind log whenever the valgrind error count has increased
during the query. Valgrind will already have printed its report,
if the error is distinct from ones already seen, so that this works
out fairly well as a way of annotating the log.
Onur Tirtir and Tom Lane
Discussion: https://postgr.es/m/AM9PR83MB0498531E804DC8DF8CFF0E8FE9D09@AM9PR83MB0498.EURPRD83.prod.outlook.com
The new name better reflects what the field is - the size of the buffers[]
array. ring_size sounded more like it is in units of bytes.
An upcoming commit allows a BufferAccessStrategy of custom sizes, so it
seems relevant to improve this beforehand.
Author: Melanie Plageman
Reviewed-by: David Rowley
Discussion: https://postgr.es/m/CAAKRu_YefVbhg4rAxU2frYFdTap61UftH-kUNQZBvAs%2BYK81Zg%40mail.gmail.com
Traditionally, vacuum always makes use of a buffer access strategy 32
buffers in size. This means that running vacuums tend not to cause too
many shared buffers to become dirty, however, this can cause vacuums to
run much more slowly than they otherwise could as WAL flushes will occur
more frequently due to having to flush WAL out to the LSN of the dirty
page before that page can be written to disk.
When we are performing failsafe VACUUMs (as added in 1e55e7d17), we really
want to make the vacuum work go as quickly as possible, so here we disable
the buffer access strategy when entering failsafe mode while vacuuming a
relation.
Per idea and analyis from Andres Freund.
In passing, also include some changes I had intended for 32fbe0239.
Author: Melanie Plageman
Reviewed-by: Justin Pryzby, David Rowley
Discussion: https://postgr.es/m/20230111182720.ejifsclfwymw2reb%40awork3.anarazel.de
VACUUM FULL and VACUUM ONLY_DATABASE_STATS will not use the vacuum
strategy ring created in vacuum(), so don't waste effort making it in
those cases.
There are other conceivable cases where the buffer strategy also won't be
used, but those are probably less common and not worth troubling over too
much. For example VACUUM (PROCESS_MAIN false, PROCESS_TOAST false).
There are other cases too, but many of these are only discovered once
inside vacuum_rel().
Author: Melanie Plageman
Reviewed-by: David Rowley
Discussion: https://postgr.es/m/CAAKRu_ZLRuzkM3zKogiZAz2hUony37yLY4aaLb8fPf8fgqs5Og@mail.gmail.com
Using global variables because we don't want to pass these values around
as parameters does not really seem like a great idea, so let's remove
these two global variables and adjust a few functions to accept these
values as parameters instead.
This is part of a wider patch which intends to allow the size of the
buffer access strategy that vacuum uses to be adjusted.
Author: Melanie Plageman
Reviewed-by: Bharath Rupireddy
Discussion: https://postgr.es/m/CAAKRu_b1q_07uquUtAvLqTM%3DW9nzee7QbtzHwA4XdUo7KX_Cnw%40mail.gmail.com
This commit only implements one prerequisite part for allowing logical
decoding. The commit message contains an explanation of the overall design,
which later commits will refer back to.
Overall design:
1. We want to enable logical decoding on standbys, but replay of WAL
from the primary might remove data that is needed by logical decoding,
causing error(s) on the standby. To prevent those errors, a new replication
conflict scenario needs to be addressed (as much as hot standby does).
2. Our chosen strategy for dealing with this type of replication slot
is to invalidate logical slots for which needed data has been removed.
3. To do this we need the latestRemovedXid for each change, just as we
do for physical replication conflicts, but we also need to know
whether any particular change was to data that logical replication
might access. That way, during WAL replay, we know when there is a risk of
conflict and, if so, if there is a conflict.
4. We can't rely on the standby's relcache entries for this purpose in
any way, because the startup process can't access catalog contents.
5. Therefore every WAL record that potentially removes data from the
index or heap must carry a flag indicating whether or not it is one
that might be accessed during logical decoding.
Why do we need this for logical decoding on standby?
First, let's forget about logical decoding on standby and recall that
on a primary database, any catalog rows that may be needed by a logical
decoding replication slot are not removed.
This is done thanks to the catalog_xmin associated with the logical
replication slot.
But, with logical decoding on standby, in the following cases:
- hot_standby_feedback is off
- hot_standby_feedback is on but there is no a physical slot between
the primary and the standby. Then, hot_standby_feedback will work,
but only while the connection is alive (for example a node restart
would break it)
Then, the primary may delete system catalog rows that could be needed
by the logical decoding on the standby (as it does not know about the
catalog_xmin on the standby).
So, it’s mandatory to identify those rows and invalidate the slots
that may need them if any. Identifying those rows is the purpose of
this commit.
Implementation:
When a WAL replay on standby indicates that a catalog table tuple is
to be deleted by an xid that is greater than a logical slot's
catalog_xmin, then that means the slot's catalog_xmin conflicts with
the xid, and we need to handle the conflict. While subsequent commits
will do the actual conflict handling, this commit adds a new field
isCatalogRel in such WAL records (and a new bit set in the
xl_heap_visible flags field), that is true for catalog tables, so as to
arrange for conflict handling.
The affected WAL records are the ones that already contain the
snapshotConflictHorizon field, namely:
- gistxlogDelete
- gistxlogPageReuse
- xl_hash_vacuum_one_page
- xl_heap_prune
- xl_heap_freeze_page
- xl_heap_visible
- xl_btree_reuse_page
- xl_btree_delete
- spgxlogVacuumRedirect
Due to this new field being added, xl_hash_vacuum_one_page and
gistxlogDelete do now contain the offsets to be deleted as a
FLEXIBLE_ARRAY_MEMBER. This is needed to ensure correct alignment.
It's not needed on the others struct where isCatalogRel has
been added.
This commit just introduces the WAL format changes mentioned above. Handling
the actual conflicts will follow in future commits.
Bumps XLOG_PAGE_MAGIC as the several WAL records are changed.
Author: "Drouvot, Bertrand" <bertranddrouvot.pg@gmail.com>
Author: Andres Freund <andres@anarazel.de> (in an older version)
Author: Amit Khandekar <amitdkhan.pg@gmail.com> (in an older version)
Reviewed-by: "Drouvot, Bertrand" <bertranddrouvot.pg@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Reviewed-by: Fabrízio de Royes Mello <fabriziomello@gmail.com>
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
This is done in preparation for logical decoding on standby, which needs to
include whether visibility affecting WAL records are about a (user) catalog
table. Which is only known for the table, not the indexes.
It's also nice to be able to pass the heap relation to GlobalVisTestFor() in
vacuumRedirectAndPlaceholder().
Author: "Drouvot, Bertrand" <bertranddrouvot.pg@gmail.com>
Discussion: https://postgr.es/m/21b700c3-eecf-2e05-a699-f8c78dd31ec7@gmail.com
If visibilitymap_set() is called with flags containing a higher bit than
VISIBILITYMAP_ALL_FROZEN, the state of neighboring pages is affected. While
there was an assertion that *some* valid bits were set, it did not check
that *only* valid bits were. Change that.
Discussion: https://postgr.es/m/20230331043300.gux3s5wzrapqi4oe@awork3.anarazel.de
PageInit() while holding the extension lock is unnecessary after 0d1fe9f74e
started to use RBM_ZERO_AND_LOCK - nobody can look at the new page before we
release the page lock. PageInit() zeroes the page, which isn't that cheap, so
deferring it until after the extension lock is released seems like a good idea.
Doing visibilitymap_pin() while holding the extension lock, introduced in
7db0cd2145, looks like an accident. Due to the restrictions on
HEAP_INSERT_FROZEN it's unlikely to be a performance issue, but it still seems
better to move it out. We also are doing the visibilitymap_pin() while
holding the buffer lock, which will be fixed in a separate commit.
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Discussion: http://postgr.es/m/419312fd-9255-078c-c3e3-f0525f911d7f@iki.fi
This patch introduces the SQL standard IS JSON predicate. It operates
on text and bytea values representing JSON, as well as on the json and
jsonb types. Each test has IS and IS NOT variants and supports a WITH
UNIQUE KEYS flag. The tests are:
IS JSON [VALUE]
IS JSON ARRAY
IS JSON OBJECT
IS JSON SCALAR
These should be self-explanatory.
The WITH UNIQUE KEYS flag makes these return false when duplicate keys
exist in any object within the value, not necessarily directly contained
in the outermost object.
Author: Nikita Glukhov <n.gluhov@postgrespro.ru>
Author: Teodor Sigaev <teodor@sigaev.ru>
Author: Oleg Bartunov <obartunov@gmail.com>
Author: Alexander Korotkov <aekorotkov@gmail.com>
Author: Amit Langote <amitlangote09@gmail.com>
Author: Andrew Dunstan <andrew@dunslane.net>
Reviewers have included (in no particular order) Andres Freund, Alexander
Korotkov, Pavel Stehule, Andrew Alsup, Erik Rijkers, Zihong Yu,
Himanshu Upadhyaya, Daniel Gustafsson, Justin Pryzby.
Discussion: https://postgr.es/m/CAF4Au4w2x-5LTnN_bxky-mq4=WOqsGsxSpENCzHRAzSnEd8+WQ@mail.gmail.com
Discussion: https://postgr.es/m/cd0bb935-0158-78a7-08b5-904886deac4b@postgrespro.ru
Discussion: https://postgr.es/m/20220616233130.rparivafipt6doj3@alap3.anarazel.de
Discussion: https://postgr.es/m/abd9b83b-aa66-f230-3d6d-734817f0995d%40postgresql.org
I realized that the third overflow case I posited in commit b0e9e4d76
actually should be handled in a different way: rather than tolerating
the idea that the quotient could round to 1, we should clamp so that
the output cannot be more than "count" when we know that the operand is
less than bound2. That being the case, we don't need an overflow-aware
increment in that code path, which leads me to revert the movement of
the pg_add_s32_overflow() call. (The diff in width_bucket_float8
might be easier to read by comparing against b0e9e4d76^.)
What's more, width_bucket_numeric also has this problem of the quotient
potentially rounding to 1, so add a clamp there too.
As before, I'm not quite convinced that a back-patch is warranted.
Discussion: https://postgr.es/m/391415.1680268470@sss.pgh.pa.us
Up through v11 it was sensible to use the "oid" system column as
a foreign key column, but since that was removed there's no visible
usefulness in making any of the remaining system columns a foreign
key. Moreover, since the TupleTableSlot rewrites in v12, such cases
actively fail because of implicit assumptions that only user columns
appear in foreign keys. The lack of complaints about that seems
like good evidence that no one is trying to do it. Hence, rather
than trying to repair those assumptions (of which there are at least
two, maybe more), let's just forbid the case up front.
Per this patch, a system column in either the referenced or
referencing side of a foreign key will draw this error; however,
putting one in the referenced side would have failed later anyway,
since we don't allow unique indexes to be made on system columns.
Per bug #17877 from Alexander Lakhin. Back-patch to v12; the
case still appears to work in v11, so we shouldn't break it there.
Discussion: https://postgr.es/m/17877-4bcc658e33df6de1@postgresql.org
The totalrows/totaldeadrows outputs were left uninitialized in cases
where we find no analyzable child tables of a partitioned table. This
could lead to setting the partitioned table's pg_class.reltuples value
to garbage. It's not clear that that would have any very bad effects
in practice, but fix it anyway because it's making valgrind unhappy.
Reported and diagnosed by Alexander Lakhin (bug #17880).
Discussion: https://postgr.es/m/17880-9282037c923d856e@postgresql.org
Among other things, this should make it easier to calculate a useful cache hit
ratio by excluding buffer reads via buffer access strategies. As buffer access
strategies reuse buffers (and thus evict the prior buffer contents), it is
normal to see reads on repeated scans of the same data.
Author: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-by: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/CAAKRu_beMa9Hzih40%3DXPYqhDVz6tsgUGTrhZXRo%3Dunp%2Bszb%3DUA%40mail.gmail.com
When calling generateSerialExtraStmts(), we would pass in the
constraint->options. In some cases, generateSerialExtraStmts() would
modify the referenced List to remove elements from it, but doing so is
invalid without assigning the list back to all variables that point to it.
In the particular reported problem case, the List became empty, in which
cases it became NIL, but the passed in constraint->options didn't get to
find out about that and was left pointing to free'd memory.
To fix this, just perform a list_copy() inside generateSerialExtraStmts().
We could just do a list_copy() just before we perform the delete from the
list, however, that seems less robust. Let's make sure the generated
CreateSeqStmt gets a completely different copy of the list to be safe.
Bug: #17879
Reported-by: Fei Changhong
Diagnosed-by: Fei Changhong
Discussion: https://postgr.es/m/17879-b7dfb5debee58ff5@postgresql.org
Backpatch-through: 11, all supported versions
Full and right outer joins were not supported in the initial
implementation of Parallel Hash Join because of deadlock hazards (see
discussion). Therefore FULL JOIN inhibited parallelism, as the other
join strategies can't do that in parallel either.
Add a new PHJ phase PHJ_BATCH_SCAN that scans for unmatched tuples on
the inner side of one batch's hash table. For now, sidestep the
deadlock problem by terminating parallelism there. The last process to
arrive at that phase emits the unmatched tuples, while others detach and
are free to go and work on other batches, if there are any, but
otherwise they finish the join early.
That unfairness is considered acceptable for now, because it's better
than no parallelism at all. The build and probe phases are run in
parallel, and the new scan-for-unmatched phase, while serial, is usually
applied to the smaller of the two relations and is either limited by
some multiple of work_mem, or it's too big and is partitioned into
batches and then the situation is improved by batch-level parallelism.
Author: Melanie Plageman <melanieplageman@gmail.com>
Author: Thomas Munro <thomas.munro@gmail.com>
Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Discussion: https://postgr.es/m/CA%2BhUKG%2BA6ftXPz4oe92%2Bx8Er%2BxpGZqto70-Q_ERwRaSyA%3DafNg%40mail.gmail.com
In instr_time.h it is stated that:
* When summing multiple measurements, it's recommended to leave the
* running sum in instr_time form (ie, use INSTR_TIME_ADD or
* INSTR_TIME_ACCUM_DIFF) and convert to a result format only at the end.
The reason for that is that converting to microseconds is not cheap, and can
loose precision. Therefore this commit changes 'PendingWalStats' to use
'instr_time' instead of 'PgStat_Counter' while accumulating 'wal_write_time'
and 'wal_sync_time'.
Author: Nazir Bilal Yavuz <byavuz81@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/1feedb83-7aa9-cb4b-5086-598349d3f555@gmail.com
There's no need for callers to pass aggregate names so that the function
can resolve them to OIDs, when callers can just pass aggregate OIDs
directly to begin with.
We were transferring partPruneInfos from PlannerInfo into PlannerGlobal
wrong, essentially relying on all of them being transferred, and
adjusting their list indexes based on that. But apparently it's
possible that some of them are skipped, so that strategy leads to a
corrupted execution tree. Instead, adjust each Append/MergeAppend's
partpruneinfo index as we copy from one list to the other, which seems
safer anyway. This requires adjusting the RT offset of the RTE
referenced in each partPruneInfo ahead of actually adjusting the RTE
itself, which seems a bit too ad-hoc.
This problem was introduced by commit ec38694894. However, it may be
that we no longer require the change introduced there, so perhaps we
should revert both the present commit and that one.
Problem noticed by sqlsmith.
Author: Amit Langote <amitlangote09@gmail.com
Discussion: https://postgr.es/m/CA+HiwqG6tbc2oadsbyyy24b2AL295XHQgyLRWghmA7u_SL1K8A@mail.gmail.com
Quoting Melanie:
> Since if buffer is INT_MAX, then the -(buffer + 1) version invokes
> undefined behavior while the -buffer - 1 version doesn't.
All other places were already using the correct version. I (Andres), copied
the code into more places in a patch. Melanie caught it in review, but to
prevent more people from copying the bad code, fix it. Even if it is a
theoretical issue.
We really ought to wrap these accesses in a helper function...
As this is a theoretical issue, don't backpatch.
Reported-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/CAAKRu_aW2SX_LWtwHgfnqYpBrunMLfE9PD6-ioPpkh92XH0qpg@mail.gmail.com
This role can be granted to non-superusers to allow them to issue
CREATE SUBSCRIPTION. The non-superuser must additionally have CREATE
permissions on the database in which the subscription is to be
created.
Most forms of ALTER SUBSCRIPTION, including ALTER SUBSCRIPTION .. SKIP,
now require only that the role performing the operation own the
subscription, or inherit the privileges of the owner. However, to
use ALTER SUBSCRIPTION ... RENAME or ALTER SUBSCRIPTION ... OWNER TO,
you also need CREATE permission on the database. This is similar to
what we do for schemas. To change the owner of a schema, you must also
have permission to SET ROLE to the new owner, similar to what we do
for other object types.
Non-superusers are required to specify a password for authentication
and the remote side must use the password, similar to what is required
for postgres_fdw and dblink. A superuser who wants a non-superuser to
own a subscription that does not rely on password authentication may
set the new password_required=false property on that subscription. A
non-superuser may not set password_required=false and may not modify a
subscription that already has password_required=false.
This new password_required subscription property works much like the
eponymous postgres_fdw property. In both cases, the actual semantics
are that a password is not required if either (1) the property is set
to false or (2) the relevant user is the superuser.
Patch by me, reviewed by Andres Freund, Jeff Davis, Mark Dilger,
and Stephen Frost (but some of those people did not fully endorse
all of the decisions that the patch makes).
Discussion: http://postgr.es/m/CA+TgmoaDH=0Xj7OBiQnsHTKcF2c4L+=gzPBUKSJLh8zed2_+Dg@mail.gmail.com
The original coding of this function paid little attention to the
possibility of overflow. There were actually three different hazards:
1. The range from bound1 to bound2 could exceed DBL_MAX, which on
IEEE-compliant machines produces +Infinity in the subtraction.
At best we'd lose all precision in the result, and at worst
produce NaN due to dividing Inf/Inf. The range can't exceed
twice DBL_MAX though, so we can fix this case by scaling all the
inputs by 0.5.
2. We computed count * (operand - bound1), which is also at risk of
float overflow, before dividing. Safer is to do the division first,
producing a quotient that should be in [0,1), and even after allowing
for roundoff error can't be outside [0,1]; then multiplying by count
can't produce a result overflowing an int. (width_bucket_numeric does
the multiplication first on the grounds that that improves accuracy of
its result, but I don't think that a similar argument can be made in
float arithmetic.)
3. If the division result does round to 1, and count is INT_MAX,
the final addition of 1 would overflow an int. We took care
of that in the operand >= bound2 case but did not consider that
it could be possible in the main path. Fix that by moving the
overflow-aware addition of 1 so it is done that way in all cases.
The fix for point 2 creates a possibility that values very close to
a bucket boundary will be rounded differently than they were before.
I'm not troubled by that for HEAD, but it is an argument against
putting this into the stable branches. Given that the cases being
fixed here are fairly extreme and unlikely to be hit in normal use,
it seems best not to back-patch.
Mats Kindahl and Tom Lane
Discussion: https://postgr.es/m/17876-61f280d1601f978d@postgresql.org
The tts_flag is named TTS_FLAG_SHOULDFREE, so use that instead of
TTS_SHOULDFREE, which is the name of the macro that checks for that flag.
Additionally, 4da597edf got rid of the TupleTableSlot.tts_tuple field but
forgot to update a comment which referenced that field. Fix that.
Reported-by: Zhen Mingyang <zhenmingyang@yeah.net>
Reported-by: Richard Guo <guofenglinux@gmail.com>
Discussion: https://postgr.es/m/1a96696c.9d3.187193989c3.Coremail.zhenmingyang@yeah.net
gistBuildCallback tried to fetch the size of an index tuple that
might have already been freed by gistProcessEmptyingQueue.
While this seems to usually be harmless in production builds,
in principle it could result in a SIGSEGV, or more likely a bogus
value for indtuplesSize leading to poor page-split decisions later
in the build.
The memory management here is confusing and could stand to be
refactored, but for the moment it seems to be enough to fetch
the tuple size sooner. AFAICT the indtuples[Size] totals aren't
used in between these places; even if they were, the updated
values shouldn't be any worse to use. So just move the
incrementing of the totals up.
It's not very clear why our valgrind-using buildfarm animals
haven't noticed this problem, because the relevant code path
does seem to be exercised according to the code coverage report.
I think the reason that we didn't fix this bug after the first
report is that I'd wanted to try to understand that better.
However, now that it's been re-discovered let's just be pragmatic
and fix it already.
Original report by Alexander Lakhin (bug #16329),
later rediscovered by Egor Chindyaskin (bug #17874).
Patch by Alexander Lakhin (commentary by Pavel Borisov and me).
Back-patch to all supported branches.
Discussion: https://postgr.es/m/16329-7a6aa9b6fa1118a1@postgresql.org
Discussion: https://postgr.es/m/17874-63ca6c7ce42d2103@postgresql.org
This function has been a no-op for over a decade. Even if bufmgr
regains a need to be called during commit, it seems unlikely that
the most appropriate call points would be precisely here, so it's not
doing us much good as a placeholder either. Now, removing it probably
doesn't save any noticeable number of cycles --- but the main call is
inside the commit critical section, and the less work done there the
better.
Matthias van de Meent
Discussion: https://postgr.es/m/CAEze2Wi1=tLKbxZnXzcD+8fYKyKqBtivVakLQC_mYBsP4Y8qVA@mail.gmail.com
This commit introduces the SQL/JSON standard-conforming constructors for
JSON types:
JSON_ARRAY()
JSON_ARRAYAGG()
JSON_OBJECT()
JSON_OBJECTAGG()
Most of the functionality was already present in PostgreSQL-specific
functions, but these include some new functionality such as the ability
to skip or include NULL values, and to allow duplicate keys or throw
error when they are found, as well as the standard specified syntax to
specify output type and format.
Author: Nikita Glukhov <n.gluhov@postgrespro.ru>
Author: Teodor Sigaev <teodor@sigaev.ru>
Author: Oleg Bartunov <obartunov@gmail.com>
Author: Alexander Korotkov <aekorotkov@gmail.com>
Author: Amit Langote <amitlangote09@gmail.com>
Reviewers have included (in no particular order) Andres Freund, Alexander
Korotkov, Pavel Stehule, Andrew Alsup, Erik Rijkers, Zihong Yu,
Himanshu Upadhyaya, Daniel Gustafsson, Justin Pryzby.
Discussion: https://postgr.es/m/CAF4Au4w2x-5LTnN_bxky-mq4=WOqsGsxSpENCzHRAzSnEd8+WQ@mail.gmail.com
Discussion: https://postgr.es/m/cd0bb935-0158-78a7-08b5-904886deac4b@postgrespro.ru
Discussion: https://postgr.es/m/20220616233130.rparivafipt6doj3@alap3.anarazel.de
Discussion: https://postgr.es/m/abd9b83b-aa66-f230-3d6d-734817f0995d%40postgresql.org
In ancient times, these belonged to arguments or fields that were
actually of type long, but now they are not anymore, so this "L"
decoration is just confusing. (Some other 0L and other "L" constants
remain, where they are actually associated with a long type.)
When there are multiple publications for a subscription and one of those
publishes via the parent table by using publish_via_partition_root and the
other one directly publishes the child table, we end up copying the same
data twice during initial synchronization. The reason for this was that we
get both the parent and child tables from the publisher and try to copy
the data for both of them.
This patch extends the function pg_get_publication_tables() to take a
publication list as its input parameter. This allows us to exclude a
partition table whose ancestor is published by the same publication list.
This problem does exist in back-branches but we decide to fix it there in
a separate commit if required. The fix for back-branches requires quite
complicated changes to fetch the required table information from the
publisher as we can't update the function pg_get_publication_tables() in
back-branches. We are not sure whether we want to deviate and complicate
the code in back-branches for this problem as there are no field reports
yet.
Author: Wang wei
Reviewed-by: Peter Smith, Jacob Champion, Kuroda Hayato, Vignesh C, Osumi Takamichi, Amit Kapila
Discussion: https://postgr.es/m/OS0PR01MB57167F45D481F78CDC5986F794B99@OS0PR01MB5716.jpnprd01.prod.outlook.com
For ICU collations, ensure that the locale's language exists in ICU,
and that the locale can be opened.
Basic validation helps avoid minor mistakes and misspellings, which
often fall back to the root locale instead of the intended
locale. It's even more important to avoid such mistakes in ICU
versions 54 and earlier, where the same (misspelled) locale string
could fall back to different locales depending on the environment.
Discussion: https://postgr.es/m/11b1eeb7e7667fdd4178497aeb796c48d26e69b9.camel@j-davis.com
Discussion: https://postgr.es/m/df2efad0cae7c65180df8e5ebb709e5eb4f2a82b.camel@j-davis.com
Reviewed-by: Peter Eisentraut
MERGE planning could fail with "variable not found in subplan target
list" if the target table is partitioned and all its partitions are
excluded at plan time, or in the case where it has no partitions but
used to have some. This happened because distribute_row_identity_vars
thought it didn't need to make the target table's reltarget list
fully valid; but if we generate a join plan then that is required
because the dummy Result node's tlist will be made from the reltarget.
The same logic appears in distribute_row_identity_vars in v14,
but AFAICS the problem is unreachable in that branch for lack of
MERGE. In other updating statements, the target table is always
inner-joined to any other tables, so if the target is known dummy
then the whole plan reduces to dummy, so no join nodes are created.
So I'll refrain from back-patching this code change to v14 for now.
Per report from Alvaro Herrera.
Discussion: https://postgr.es/m/20230328112248.6as34mlx5sr4kltg@alvherre.pgsql
To support older ICU versions, we rely on
icu_set_collation_attributes() to do error checking that is handled
directly by ucol_open() in newer ICU versions. Commit 3b50275b12
introduced a slight inconsistency, where the error report includes the
fixed-up locale string, rather than the locale string passed to
pg_ucol_open().
Refactor slightly so that pg_ucol_open() handles the errors from both
ucol_open() and icu_set_collation_attributes(), making it easier to
see any differences between the error reports. It also makes
pg_ucol_open() responsible for closing the UCollator on error, which
seems like the right place.
Discussion: https://postgr.es/m/04182066-7655-344a-b8b7-040b1b2490fb%40enterprisedb.com
Reviewed-by: Peter Eisentraut
Change the columns attndims, attstattarget, and attinhcount from int32
to int16, and reorder a bit. This saves some space (currently 4
bytes) in pg_attribute and tuple descriptors, which translates into
small performance benefits and/or room for new columns in pg_attribute
needed by future features.
attndims and attinhcount are never realistically used with values
larger than int16. Just to be sure, add some overflow checks.
attstattarget is currently limited explicitly to 10000.
For consistency, pg_constraint.coninhcount is also changed like
attinhcount.
Discussion: https://www.postgresql.org/message-id/flat/d07ffc2b-e0e8-77f7-38fb-be921dff71af%40enterprisedb.com
Two new macros are added with their respective functions switched to
use them. These are for functions with millisecond stats, with and
without "xact" in their names (for the stats that can be tracked within
a transaction).
While on it, prefix the macro for float8 on database entries with "_MS",
as it does a us->ms conversion, based on a suggestion from Andres
Freund.
Author: Bertrand Drouvot
Discussion: https://postgr.es/m/6e2efb4f-6fd0-807e-f6bf-94207db8183a@gmail.com
find_composite_type_dependencies() ignored indexes, which is a poor
decision because an expression index could have a stored column of
a composite (or other container) type even when the underlying table
does not. Teach it to detect such cases and error out. We have to
work a bit harder than for other relations because the pg_depend entry
won't identify the specific index column of concern, but it's not much
new code.
This does not address bug #17872's original complaint that dropping
a column in such a type might lead to violations of the uniqueness
property that a unique index is supposed to ensure. That seems of
much less concern to me because it won't lead to crashes.
Per bug #17872 from Alexander Lakhin. Back-patch to all supported
branches.
Discussion: https://postgr.es/m/17872-d0fbb799dc3fd85d@postgresql.org
Replace the hardcoded value with a GUC such that the iteration
count can be raised in order to increase protection against
brute-force attacks. The hardcoded value for SCRAM iteration
count was defined to be 4096, which is taken from RFC 7677, so
set the default for the GUC to 4096 to match. In RFC 7677 the
recommendation is at least 15000 iterations but 4096 is listed
as a SHOULD requirement given that it's estimated to yield a
0.5s processing time on a mobile handset of the time of RFC
writing (late 2015).
Raising the iteration count of SCRAM will make stored passwords
more resilient to brute-force attacks at a higher computational
cost during connection establishment. Lowering the count will
reduce computational overhead during connections at the tradeoff
of reducing strength against brute-force attacks.
There are however platforms where even a modest iteration count
yields a too high computational overhead, with weaker password
encryption schemes chosen as a result. In these situations,
SCRAM with a very low iteration count still gives benefits over
weaker schemes like md5, so we allow the iteration count to be
set to one at the low end.
The new GUC is intentionally generically named such that it can
be made to support future SCRAM standards should they emerge.
At that point the value can be made into key:value pairs with
an undefined key as a default which will be backwards compatible
with this.
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Reviewed-by: Jonathan S. Katz <jkatz@postgresql.org>
Discussion: https://postgr.es/m/F72E7BC7-189F-4B17-BF47-9735EB72C364@yesql.se
The nested-arrays code path in ExecEvalArrayExpr() used palloc to
allocate the result array, whereas every other array-creating function
has used palloc0 since 18c0b4ecc. This mostly works, but unused bits
past the end of the nulls bitmap may end up undefined. That causes
valgrind complaints with -DWRITE_READ_PARSE_PLAN_TREES, and could
cause planner misbehavior as cited in 18c0b4ecc. There seems no very
good reason why we should strive to avoid palloc0 in just this one case,
so fix it the easy way with s/palloc/palloc0/.
While looking at that I noted that we also failed to check for overflow
of "nbytes" and "nitems" while summing the sizes of the sub-arrays,
potentially allowing a crash due to undersized output allocation.
For "nbytes", follow the policy used by other array-munging code of
checking for overflow after each addition. (As elsewhere, the last
addition of the array's overhead space doesn't need an extra check,
since palloc itself will catch a value between 1Gb and 2Gb.)
For "nitems", there's no very good reason to sum the inputs at all,
since we can perfectly well use ArrayGetNItems' result instead of
ignoring it.
Per discussion of this bug, also remove redundant zeroing of the
nulls bitmap in array_set_element and array_set_slice.
Patch by Alexander Lakhin and myself, per bug #17858 from Alexander
Lakhin; thanks also to Richard Guo. These bugs are a dozen years old,
so back-patch to all supported branches.
Discussion: https://postgr.es/m/17858-8fd287fd3663d051@postgresql.org
When extracting an attr from a cached tuple in the syscache with
SysCacheGetAttr the isnull parameter must be checked in case the
attr cannot be NULL. For cases when this is known beforehand, a
wrapper is introduced which perform the errorhandling internally
on behalf of the caller, invoking an elog in case of a NULL attr.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/AD76405E-DB45-46B6-941F-17B1EB3A9076@yesql.se
The "partitions_total" and "partitions_done" fields were updated
as though the current level of partitioning was the only one.
In multi-level cases, not only could partitions_total change
over the course of the command, but partitions_done could go
backwards or exceed the currently-reported partitions_total.
Fix by setting partitions_total to the total number of direct
and indirect children once at command start, and then just
incrementing partitions_done at appropriate points. Invent
a new progress monitoring function "pgstat_progress_incr_param"
to simplify doing the latter. We can avoid adding cost for the
former when doing CREATE INDEX, because ProcessUtility already
enumerates the children and it's pretty easy to pass the count
down to DefineIndex. In principle the same could be done in
ALTER TABLE, but that's structurally difficult; for now, just
eat the cost of an extra find_all_inheritors scan in that case.
Ilya Gladyshev and Justin Pryzby
Discussion: https://postgr.es/m/a15f904a70924ffa4ca25c3c744cff31e0e6e143.camel@gmail.com
This provides a very simple way to see the generic plan for a
parameterized query. Without this, it's necessary to define
a prepared statement and temporarily change plan_cache_mode,
which is a bit tedious.
One thing that's a bit of a hack perhaps is that we disable
execution-time partition pruning when the GENERIC_PLAN option
is given. That's because the pruning code may attempt to
fetch the value of one of the parameters, which would fail.
Laurenz Albe, reviewed by Julien Rouhaud, Christoph Berg,
Michel Pelletier, Jim Jones, and myself
Discussion: https://postgr.es/m/0a29b954b10b57f0d135fe12aa0909bd41883eb0.camel@cybertec.at
ICU versions 53 and earlier rely on icu_set_collation_attributes() to
process the attributes in the locale string. Avoid leaking the
already-opened UCollator object if an error is encountered.
Discussion: https://postgr.es/m/04182066-7655-344a-b8b7-040b1b2490fb%40enterprisedb.com
Reviewed-by: Peter Eisentraut
The fields of NLSVERSIONINFOEX are of type DWORD, which is unsigned
long, so the results of the computations being printed are also of
type unsigned long.
This commit renames the members of a few pgstat structures related to
functions and relations, by respectively removing their prefix "f_" and
"t_". The statistics for functions and relations and handled in their
own file, and pgstatfuncs.c associates each field in a structure
variable named based on the object type handled, so no information is
lost with this rename.
This will help with some of the refactoring aimed for pgstatfuncs.c, as
this makes more consistent the field names with the SQL functions
retrieving them.
Author: Bertrand Drouvot
Reviewed-by: Michael Paquier, Melanie Plageman
Discussion: https://postgr.es/m/9142f62a-a422-145c-bde0-b5bc498a4ada@gmail.com
Add pgstat counter to track row updates that result in the successor
version going to a new heap page, leaving behind an original version
whose t_ctid points to the new version. The current count is shown by
the n_tup_newpage_upd column of each of the pg_stat_*_tables views.
The new n_tup_newpage_upd column complements the existing n_tup_hot_upd
and n_tup_upd columns. Tables that have high n_tup_newpage_upd values
(relative to n_tup_upd) are good candidates for tuning heap fillfactor.
Corey Huinker, with small tweaks by me.
Author: Corey Huinker <corey.huinker@gmail.com>
Reviewed-By: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/CADkLM=ded21M9iZ36hHm-vj2rE2d=zcKpUQMds__Xm2pxLfHKA@mail.gmail.com
The "und" locale is an alternative spelling of the root locale, but it
was not recognized until ICU 55. To maintain common behavior across
all supported ICU versions, check for "und" and replace with "root"
before opening.
Previously, the lack of support for "und" was dangerous, because
versions 54 and older fall back to the environment when a locale is
not found. If the user specified "und" for the language (which is
expected and documented), it could not only resolve to the wrong
collator, but it could unexpectedly change (which could lead to
corrupt indexes).
This effectively reverts commit d72900bded, which worked around the
problem for the built-in "unicode" collation, and is no longer
necessary.
Discussion: https://postgr.es/m/60da0cecfb512a78b8666b31631a636215d8ce73.camel@j-davis.com
Discussion: https://postgr.es/m/0c6fa66f2753217d2a40480a96bd2ccf023536a1.camel@j-davis.com
Reviewed-by: Peter Eisentraut
Hide details of supporting older ICU versions in a wrapper
function. The current code only needs to handle
icu_set_collation_attributes(), but a subsequent commit will add
additional version-specific code.
Discussion: https://postgr.es/m/7ee414ad-deb5-1144-8a0e-b34ae3b71cd5@enterprisedb.com
Reviewed-by: Peter Eisentraut
We fail to apply updates and deletes when the REPLICA IDENTITY FULL is
used for the table having generated columns. We didn't use to ignore
generated columns while doing tuple comparison among the tuples from
the publisher and subscriber during apply of updates and deletes.
Author: Onder Kalaci
Reviewed-by: Shi yu, Amit Kapila
Backpatch-through: 12
Discussion: https://postgr.es/m/CACawEhVQC9WoofunvXg12aXtbqKnEgWxoRx3+v8q32AWYsdpGg@mail.gmail.com
This patch allows copying tables in the binary format during table
synchronization when the binary option for a subscription is enabled.
Previously, tables are copied in text format even if the subscription is
created with the binary option enabled. Copying tables in binary format
may reduce the time spent depending on column types.
A binary copy for initial table synchronization is supported only when
both publisher and subscriber are v16 or later.
Author: Melih Mutlu
Reviewed-by: Peter Smith, Shi yu, Euler Taveira, Vignesh C, Kuroda Hayato, Osumi Takamichi, Bharath Rupireddy, Hou Zhijie
Discussion: https://postgr.es/m/CAGPVpCQvAziCLknEnygY0v1-KBtg%2BOm-9JHJYZOnNPKFJPompw%40mail.gmail.com
* Commit 3048898e dropped -ING from PHJ wait event names. Update the
corresponding barrier phases names to match.
* Rename the "DONE" phases to "FREE". That's symmetrical with
"ALLOCATE", and names the activity that actually happens in that phase
(as we do for the other phases) rather than a state. The bug fixed by
commit 8d578b9b might have been more obvious with this name.
* Rename the batch/bucket growth barriers' "ALLOCATE" phases to
"REALLOCATE", a better description of what they do.
* Update the high level comments about phases to highlight phases
are executed by a single process with an asterisk (mostly memory
management phases).
No behavior change, as this is just improving internal identifiers. The
only user-visible sign of this is that a couple of wait events' display
names change from "...Allocate" to "...Reallocate" in pg_stat_activity,
to stay in sync with the internal names.
Reviewed-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/CA%2BhUKG%2BMDpwF2Eo2LAvzd%3DpOh81wUTsrwU1uAwR-v6OGBB6%2B7g%40mail.gmail.com
Currently, in read committed transaction isolation mode (default), we have the
following sequence of actions when tuple_update()/tuple_delete() finds
the tuple updated by concurrent transaction.
1. Attempt to update/delete tuple with tuple_update()/tuple_delete(), which
returns TM_Updated.
2. Lock tuple with tuple_lock().
3. Re-evaluate plan qual (recheck if we still need to update/delete and
calculate the new tuple for update).
4. Second attempt to update/delete tuple with tuple_update()/tuple_delete().
This attempt should be successful, since the tuple was previously locked.
This patch eliminates step 2 by taking the lock during first
tuple_update()/tuple_delete() call. Heap table access method saves some
efforts by checking the updated tuple once instead of twice. Future
undo-based table access methods, which will start from the latest row version,
can immediately place a lock there.
The code in nodeModifyTable.c is simplified by removing the nested switch/case.
Discussion: https://postgr.es/m/CAPpHfdua-YFw3XTprfutzGp28xXLigFtzNbuFY8yPhqeq6X5kg%40mail.gmail.com
Reviewed-by: Aleksander Alekseev, Pavel Borisov, Vignesh C, Mason Sharp
Reviewed-by: Andres Freund, Chris Travers
When we lock tuple using table_tuple_lock() then we at the same time fetch
the locked tuple to the slot. In this case we can skip extra
table_tuple_fetch_row_version() thank to we've already fetched the 'old' tuple
and nobody can change it concurrently since it's locked.
Discussion: https://postgr.es/m/CAPpHfdua-YFw3XTprfutzGp28xXLigFtzNbuFY8yPhqeq6X5kg%40mail.gmail.com
Reviewed-by: Aleksander Alekseev, Pavel Borisov, Vignesh C, Mason Sharp
Reviewed-by: Andres Freund, Chris Travers
RelationCopyStorageUsingBuffer() did not free the strategies used to access
the source / target relation. They memory was released at the end of the
transaction, but when using a template database with a lot of relations, the
temporary leak can become big prohibitively big.
RelationCopyStorageUsingBuffer() acquired the buffer for the target relation
with RBM_NORMAL, therefore requiring a read of a block guaranteed to be
zero. Use RBM_ZERO_AND_LOCK instead.
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Discussion: https://postgr.es/m/20230321070113.o2vqqxogjykwgfrr@awork3.anarazel.de
Backpatch: 15-, where STRATEGY WAL_LOG was introduced
The files generated by gen_node_support.pl for query jumbling
(queryjumblefuncs.funcs.c and queryjumblefuncs.switch.c) were not being
removed on make maintainer-clean (they need to remain around after a
simple "clean"). This commit makes the operation consistent with the
copy, equal, out and read files.
While on it, update a comment in the nodes'README where a reference to
queryjumblefuncs.funcs.c was missing.
Reported-by: Nathan Bossart
Reviewed-by: Richard Guo, Daniel Gustafsson
Discussion: https://postgr.es/m/ZBgAfTHcL6W7zGdW@paquier.xyz
As demonstrated by David Johnston, the Memoize cache hit ratio calculation
wasn't quite correct.
This change only affects the estimated hit ratio when the estimated number
of entries to cache is estimated not to fit inside the cache. For
example, if we expect 2000 distinct cache key values and only expect to be
able to cache 1000 of those at once due to memory constraints, with an
estimate of 10000 calls, if we could store all entries then the hit ratio
should be 80% to account for the first 2000 of the 10000 calls to be a
cache miss due to the value not being cached yet. If we can only store
1000 entries for each of the 2000 distinct possible values at once then
the 80% should be reduced by half to make the final estimate of 40%.
Previously, the calculation would have produced an estimated hit ratio of
30%, which wasn't correct.
Apply to master only so as not to destabilize plans in the back branches.
Reported-by: David G. Johnston
Discussion: https://postgr.es/m/CAKFQuwZEmcNk3YQo2Xj4EDUOdY6qakad31rOD1Vc4q1_s68-Ew@mail.gmail.com
Discussion: https://postgr.es/m/CAApHDvrV44LwiF4W_qf_RpbGYWSgp1kF=cZr+kTRRaALUfmXqw@mail.gmail.com
We fail to apply updates and deletes when the REPLICA IDENTITY FULL is
used for the table having dropped columns. We didn't use to ignore dropped
columns while doing tuple comparison among the tuples from the publisher
and subscriber during apply of updates and deletes.
Author: Onder Kalaci, Shi yu
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/CACawEhVQC9WoofunvXg12aXtbqKnEgWxoRx3+v8q32AWYsdpGg@mail.gmail.com
With unlucky timing and parallel_leader_participation=off (not the
default), PHJ could attempt to access per-batch shared state just as it
was being freed. There was code intended to prevent that by checking
for a cleared pointer, but it was racy. Fix, by introducing an extra
barrier phase. The new phase PHJ_BUILD_RUNNING means that it's safe to
access the per-batch state to find a batch to help with, and
PHJ_BUILD_DONE means that it is too late. The last to detach will free
the array of per-batch state as before, but now it will also atomically
advance the phase, so that late attachers can avoid the hazard. This
mirrors the way per-batch hash tables are freed (see phases
PHJ_BATCH_PROBING and PHJ_BATCH_DONE).
An earlier attempt to fix this (commit 3b8981b6, later reverted) missed
one special case. When the inner side is empty (the "empty inner
optimization), the build barrier would only make it to
PHJ_BUILD_HASHING_INNER phase before workers attempted to detach from
the hashtable. In that case, fast-forward the build barrier to
PHJ_BUILD_RUNNING before proceeding, so that our later assertions hold
and we can still negotiate who is cleaning up.
Revealed by build farm failures, where BarrierAttach() failed a sanity
check assertion, because the memory had been clobbered by dsa_free().
In non-assert builds, the result could be a segmentation fault.
Back-patch to all supported releases.
Author: Thomas Munro <thomas.munro@gmail.com>
Author: Melanie Plageman <melanieplageman@gmail.com>
Reported-by: Michael Paquier <michael@paquier.xyz>
Reported-by: David Geier <geidav.pg@gmail.com>
Tested-by: David Geier <geidav.pg@gmail.com>
Discussion: https://postgr.es/m/20200929061142.GA29096%40paquier.xyz
@extschema:name@ extends the existing @extschema@ feature so that
we can also insert the schema name of some required extension,
thus making cross-extension references robust even if they are in
different schemas.
However, this has the same hazard as @extschema@: if the schema
name is embedded literally in an installed object, rather than being
looked up once during extension script execution, then it's no longer
safe to relocate the other extension to another schema. To deal with
that without restricting things unnecessarily, add a "no_relocate"
option to extension control files. This allows an extension to
specify that it cannot handle relocation of some of its required
extensions, even if in themselves those extensions are relocatable.
We detect "no_relocate" requests of dependent extensions during
ALTER EXTENSION SET SCHEMA.
Regina Obe, reviewed by Sandro Santilli and myself
Discussion: https://postgr.es/m/003001d8f4ae$402282c0$c0678840$@pcorp.us
When determining whether an index update may be skipped by using HOT, we
can ignore attributes indexed by block summarizing indexes without
references to individual tuples that need to be cleaned up.
A new type TU_UpdateIndexes provides a signal to the executor to
determine which indexes to update - no indexes, all indexes, or only the
summarizing indexes.
This also removes rd_indexattr list, and replaces it with rd_attrsvalid
flag. The list was not used anywhere, and a simple flag is sufficient.
This was originally committed as 5753d4ee32, but then got reverted by
e3fcca0d0d because of correctness issues.
Original patch by Josef Simanek, various fixes and improvements by Tomas
Vondra and me.
Authors: Matthias van de Meent, Josef Simanek, Tomas Vondra
Reviewed-by: Tomas Vondra, Alvaro Herrera
Discussion: https://postgr.es/m/05ebcb44-f383-86e3-4f31-0a97a55634cf@enterprisedb.com
Discussion: https://postgr.es/m/CAFp7QwpMRGcDAQumN7onN9HjrJ3u4X3ZRXdGFT0K5G2JWvnbWg%40mail.gmail.com
When calculating distance in brin_minmax_multi_distance_inet(), the
netmask was applied incorrectly. This results in (seemingly) incorrect
ordering of values, triggering an assert.
For builds without asserts this is mostly harmless - we may merge other
ranges, possibly resulting in slightly less efficient index. But it's
still correct and the greedy algorithm doesn't guarantee optimality
anyway.
Backpatch to 14, where minmax-multi indexes were introduced.
Reported by Dmitry Dolgov, investigation and fix by me.
Reported-by: Dmitry Dolgov
Backpatch-through: 14
Discussion: https://postgr.es/m/17774-c6f3e36dd4471e67@postgresql.org
The Memoize executor node stores the cache key values along with the
tuple(s) which were found in the outer node which match each key value,
however, when the planner tried to estimate how many entries could be
stored in the cache, it didn't take into account that the cache key must
also be stored. In many cases, this won't make a large difference as the
key is likely small in comparison to the tuple(s) being stored, however,
it's not impossible to craft cases where the key could take more memory
than the tuple(s) stored for it.
Here we adjust the planner so it takes into account the estimated amount
of memory to store the cache key. Effectively, this change will reduce
the estimated cache hit ratio when it's thought that not all items will
fit in the cache, thus Memoize will become more expensive in such cases.
The executor already takes into account the memory consumed by the cache
key, so here we only need to adjust the planner.
Discussion: https://postgr.es/m/CAApHDvqGErGuyBfQvBQrTCHDbzLTqoiW=_G9sOzeFxWEc_7auA@mail.gmail.com
When probing the Memoize cache to check if the current cache key values
exist in the cache, we perform an evaluation of the expressions making up
the cache key before probing the hash table for those values. This
operation could leak memory as it is possible that the cache key is an
expression which requires allocation of memory, as was the case in bug
17844.
Here we fix this by correctly switching to the per tuple context before
evaluating the cache expressions so that the memory is freed next time the
per tuple context is reset.
Bug: 17844
Reported-by: Alexey Ermakov
Discussion: https://postgr.es/m/17844-d2f6f9e75a622bed@postgresql.org
Backpatch-through: 14, where Memoize was introduced
nodeRead() will have created a Node struct that's only allocated big
enough for the specific node type, so copying sizeof(union ValUnion)
can be copying too much. This provokes valgrind complaints, and with
very bad luck could perhaps result in SIGSEGV.
While at it, tidy up _equalA_Const to avoid duplicate checks of isnull.
Per report from Alexander Lakhin. This code is new as of a6bc33019,
so no need to back-patch.
Discussion: https://postgr.es/m/4995256b-cc65-170e-0b22-60ad2cd535f1@gmail.com
Add versions of timestamptz + interval, timestamptz - interval, and
generate_series(timestamptz, ...) in which a timezone can be specified
explicitly instead of defaulting to the TimeZone GUC setting.
The new functions for the first two are named date_add and
date_subtract. This might seem too generic, but we could use
overloading to add additional variants if that seems useful.
Along the way, improve the docs' pretty inadequate explanation
of how timestamptz +- interval works.
Przemysław Sztoch and Gurjeet Singh; cosmetic changes and most of
the docs work by me
Discussion: https://postgr.es/m/01a84551-48dd-1359-bf7e-f6b0203a6bd0@sztoch.pl
We already had five copies of essentially the same logic, and an
upcoming patch introduces yet another use-case. That's past my
threshold of pain, so introduce a common subroutine. There's not
that much net code savings, but the chance of typos should go down.
Inspired by a patch from Przemysław Sztoch, but different in detail.
Discussion: https://postgr.es/m/01a84551-48dd-1359-bf7e-f6b0203a6bd0@sztoch.pl
Check whether the datctype is C to determine whether t_isspace() and
related functions use isspace() or iswspace().
Previously, t_isspace() checked whether the database default collation
was C; which is incorrect when the default collation uses the ICU
provider.
Discussion: https://postgr.es/m/79e4354d9eccfdb00483146a6b9f6295202e7890.camel@j-davis.com
Reviewed-by: Peter Eisentraut
Backpatch-through: 15
Mainly move some detail from errmsg to errdetail, remove explicit
mention of superuser where appropriate, since that is implied in most
permission checks, and make messages more uniform.
Author: Nathan Bossart <nathandbossart@gmail.com>
Discussion: https://www.postgresql.org/message-id/20230316234701.GA903298@nathanxps13
Currently, there are quite a few places in reorderbuffer.c that tries to
access top-transaction for a subtransaction. This makes the code to access
top-transaction consistent and easier to follow.
Author: Peter Smith
Reviewed-by: Vignesh C, Sawada Masahiko
Discussion: https://postgr.es/m/CAHut+PuCznOyTqBQwjRUu-ibG-=KHyCv-0FTcWQtZUdR88umfg@mail.gmail.com
The logic added in 9d9c02ccd to determine when a qual can be used as a
WindowClause run condition failed to correctly check for subqueries in the
qual. This was being done correctly for normal subquery qual pushdowns,
it's just that 9d9c02ccd failed to follow the lead on that.
This also fixes various other cases where transforming the qual into a
WindowClause run condition in the subquery should have been disallowed.
Bug: #17826
Reported-by: Anban Company
Discussion: https://postgr.es/m/17826-7d8750952f19a5f5@postgresql.org
Backpatch-through: 15, where 9d9c02ccd was introduced.
Further to commit 6a9229da, checking for NULL is now redundant. An "out
of memory" error would have been thrown already by palloc() and treated
as FATAL, so we can delete a few more lines.
Back-patch to all releases, like those other commits.
Reported-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/4040668.1679013388%40sss.pgh.pa.us
create_append_path() would only apply get_baserel_parampathinfo
when the path is for a partitioned table, but it's also potentially
useful for paths for UNION ALL appendrels. Specifically, that
supports building a Memoize path atop this one.
While we're in the vicinity, delete some dead code in
create_merge_append_plan(): there's no need for it to support
parameterized MergeAppend paths, and it doesn't look like that
is going to change anytime soon. It'll be easy enough to undo
this when/if it becomes useful.
Richard Guo
Discussion: https://postgr.es/m/CAMbWs4_ABSu4PWG2rE1q10tJugEXHWgru3U8dAgkoFvgrb6aEA@mail.gmail.com
gcc 12+ has complaints like the following:
../../../../../pgsql/src/backend/utils/adt/network.c: In function 'inetnot':
../../../../../pgsql/src/backend/utils/adt/network.c:1893:34: warning: writing 1 byte into a region of size 0 [-Wstringop-overflow=]
1893 | pdst[nb] = ~pip[nb];
| ~~~~~~~~~^~~~~~~~~~
../../../../../pgsql/src/include/utils/inet.h:27:23: note: at offset -1 into destination object 'ipaddr' of size 16
27 | unsigned char ipaddr[16]; /* up to 128 bits of address */
| ^~~~~~
../../../../../pgsql/src/include/utils/inet.h:27:23: note: at offset -1 into destination object 'ipaddr' of size 16
This is due to a compiler bug:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104986
It has been a year since the bug has been reported without getting fixed. As
the warnings are verbose and use of gcc 12 is becoming more common, it seems
worth working around the bug. Particularly because a simple reformulation of
the loop condition fixes the issue and isn't any less readable.
Author: Tom Lane <tgl@sss.pgh.pa.us>
Author: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/144536.1648326206@sss.pgh.pa.us
Backpatch: 11-
A comment was left behind claiming that we needed to use malloc() rather
than palloc() because the corresponding free would run in another
thread, but that's not true anymore. Remove that comment. And, with
the reason being gone, we might as well actually use palloc().
Back-patch to supported releases, like d41a178b.
Discussion: https://postgr.es/m/CA%2BhUKG%2BpdM9v3Jv4tc2BFx2jh_daY3uzUyAGBhtDkotEQDNPYw%40mail.gmail.com
DecodeDateTime and DecodeTimeOnly had support for date input in the
style "Y2023M03D16", which the comments claimed to be an "ISO" format.
However, so far as I can find there is no such format in ISO 8601;
they write units before numbers in intervals, but not in datetimes.
Furthermore, the lesser-known ISO 8601-2 spec actually defines an
incompatible format "2023Y03M16D". None of our documentation mentions
such a format either. So let's just drop it.
That leaves us with only two cases for a prefix unit specifier in
datetimes: Julian dates written as Jnnnn, and the "T" separator
defined by ISO 8601. Add checks to catch misuse of these specifiers,
that is consecutive specifiers or a dangling specifier at the end of
the string. We do not however disallow a specifier that is separated
from the field that it disambiguates (by noise words or unrelated
fields). That being the case, remove some overly-aggressive error
checks from the ISOTIME cases.
Joseph Koshakow, editorialized a bit by me; thanks also to
Peter Eisentraut for some standards-reading.
Discussion: https://postgr.es/m/CAAvxfHf2Q1gKLiHGnuPOiyf0ASvKUM4BnMfsXuwgtYEb_Gx0Zw@mail.gmail.com
The user receiving the message might not understand where the
server's "current directory" is. "Data directory" seems clearer.
(This would not be good for frontend code, but both of these
messages are only issued in the backend.)
Kyotaro Horiguchi
Discussion: https://postgr.es/m/20230316.111646.1564684434328830712.horikyota.ntt@gmail.com
This structure included only PgStat_FunctionCounts, and removing it
facilitates some upcoming refactoring for pgstatfuncs.c to use more
macros rather that mostly-duplicated functions.
Author: Bertrand Drouvot
Reviewed-by: Nathan Bossart
Discussion: https://postgr.es/m/11d531fe-52fc-c6ea-7e8e-62f1b6ec626e@gmail.com
This adds the ability to pretty-print XML documents ... according to
libxml's somewhat idiosyncratic notions of what's pretty, anyway.
One notable divergence from a strict reading of the spec is that
libxml is willing to collapse empty nodes "<node></node>" to just
"<node/>", whereas SQL and the underlying XML spec say that this
option should only result in whitespace tweaks. Nonetheless,
it seems close enough to justify using the SQL-standard syntax.
Jim Jones, reviewed by Peter Smith and myself
Discussion: https://postgr.es/m/2f5df461-dad8-6d7d-4568-08e10608a69b@uni-muenster.de
The hook can be installed by a shared_preload library.
A similar mechanism could be used for radius paswords, for example, and
the type name auth_password_hook_typ has been shosen with that in mind.
John Naylor and Andrew Dunstan
Discussion: https://postgr.es/m/469b06ed-69de-ba59-c13a-91d2372e52a9@dunslane.net
preprocess_targetlist thought PHVs couldn't appear here.
It was mistaken, as per report from Önder Kalacı.
Surveying other pull_var_clause calls, I noted no similar errors,
but I did notice that qual_is_pushdown_safe's assertion about
!contain_window_function was pointless, because the following
pull_var_clause call would complain about them anyway. In HEAD
only, remove the redundant Assert and improve the commentary.
Discussion: https://postgr.es/m/CACawEhUuum-gC_2S3sXLTcsk7bUSPSHOD+g1ZpfKaDK-KKPPWA@mail.gmail.com
Using REPLICA IDENTITY FULL on the publisher can lead to a full table scan
per tuple change on the subscription when REPLICA IDENTITY or PK index is
not available. This makes REPLICA IDENTITY FULL impractical to use apart
from some small number of use cases.
This patch allows using indexes other than PRIMARY KEY or REPLICA
IDENTITY on the subscriber during apply of update/delete. The index that
can be used must be a btree index, not a partial index, and it must have
at least one column reference (i.e. cannot consist of only expressions).
We can uplift these restrictions in the future. There is no smart
mechanism to pick the index. If there is more than one index that
satisfies these requirements, we just pick the first one. We discussed
using some of the optimizer's low-level APIs for this but ruled it out
as that can be a maintenance burden in the long run.
This patch improves the performance in the vast majority of cases and the
improvement is proportional to the amount of data in the table. However,
there could be some regression in a small number of cases where the indexes
have a lot of duplicate and dead rows. It was discussed that those are
mostly impractical cases but we can provide a table or subscription level
option to disable this feature if required.
Author: Onder Kalaci, Amit Kapila
Reviewed-by: Peter Smith, Shi yu, Hou Zhijie, Vignesh C, Kuroda Hayato, Amit Kapila
Discussion: https://postgr.es/m/CACawEhVLqmAAyPXdHEPv1ssU2c=dqOniiGz7G73HfyS7+nGV4w@mail.gmail.com
Commit 4753ef37 changed vacuum_delay_point() to use the WaitLatch() API,
to fix the problem that vacuum could keep running for a very long time
after the postmaster died.
Unfortunately, that broke commit caf626b2's support for fractional
vacuum_cost_delay, which shipped in PostgreSQL 12. WaitLatch() works in
whole milliseconds.
For now, revert the change from commit 4753ef37, but add an explicit
check for postmaster death. That's an extra system call on systems
other than Linux and FreeBSD, but that overhead doesn't matter much
considering that we willingly went to sleep and woke up again. (In
later work, we might add higher resolution timeouts to the latch API so
that we could do this with our standard programming pattern, but that
wouldn't be back-patched.)
Back-patch to 14, where commit 4753ef37 arrived.
Reported-by: Melanie Plageman <melanieplageman@gmail.com>
Discussion: https://postgr.es/m/CAAKRu_b-q0hXCBUCAATh0Z4Zi6UkiC0k2DFgoD3nC-r3SkR3tg%40mail.gmail.com
Our waitpid() emulation didn't prevent a PID from being recycled by the
OS before the call to waitpid(). The postmaster could finish up
tracking more than one child process with the same PID, and confuse
them.
Fix, by moving the guts of pgwin32_deadchild_callback() into waitpid(),
so that resources are released synchronously. The process and PID
continue to exist until we close the process handle, which only happens
once we're ready to adjust our book-keeping of running children.
This seems to explain a couple of failures on CI. It had never been
reported before, despite the code being as old as the Windows port.
Perhaps Windows started recycling PIDs more rapidly, or perhaps timing
changes due to commit 7389aad6 made it more likely to break.
Thanks to Alexander Lakhin for analysis and Andres Freund for tracking
down the root cause.
Back-patch to all supported branches.
Reported-by: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/20230208012852.bvkn2am4h4iqjogq%40awork3.anarazel.de
The band-aid applied in commit f0bedf3e4 turns out to still need
some work: it made sure we didn't set Np->last_relevant too small
(to the left of the decimal point), but it didn't prevent setting
it too large (off the end of the partially-converted string).
This could result in fetching data beyond the end of the allocated
space, which with very bad luck could cause a SIGSEGV, though
I don't see any hazard of interesting memory disclosure.
Per bug #17839 from Thiago Nunes. The bug's pretty ancient,
so back-patch to all supported versions.
Discussion: https://postgr.es/m/17839-aada50db24d7b0da@postgresql.org
Scanning the expression for compatible Vars isn't really necessary,
because the subsequent match against StatisticExtInfo entries will
eliminate expressions containing other Vars just fine. Moreover,
this code hadn't stopped to think about what to do with
PlaceHolderVars or Aggrefs in the clause; and at least for the PHV
case, that demonstrably leads to failures. Rather than work out
whether it's reasonable to ignore those, let's just remove the
whole stanza.
Per report from Richard Guo. Back-patch to v14 where this code
was added.
Discussion: https://postgr.es/m/CAMbWs48Mmvm-acGevXuwpB=g5JMqVSL6i9z5UaJyLGJqa-XPAA@mail.gmail.com
Expose the standard error functions as SQL-callable functions. These
are expected to be useful to people working with normal distributions,
and we use them here to test the distribution from random_normal().
Since these functions are defined in the POSIX and C99 standards, they
should in theory be available on all supported platforms. If that
turns out not to be the case, more work will be needed.
On all platforms tested so far, using extra_float_digits = -1 in the
regression tests is sufficient to allow for variations between
implementations. However, past experience has shown that there are
almost certainly going to be additional unexpected portability issues,
so these tests may well need further adjustments, based on the
buildfarm results.
Dean Rasheed, reviewed by Nathan Bossart and Thomas Munro.
Discussion: https://postgr.es/m/CAEZATCXv5fi7+Vu-POiyai+ucF95+YMcCMafxV+eZuN1B-=MkQ@mail.gmail.com
The majority of error exit cases in json_lex_string() failed to
set lex->token_terminator, causing problems for the error context
reporting code: it would see token_terminator less than token_start
and do something more or less nuts. In v14 and up the end result
could be as bad as a crash in report_json_context(). Older
versions accidentally avoided that fate; but all versions produce
error context lines that are far less useful than intended,
because they'd stop at the end of the prior token instead of
continuing to where the actually-bad input is.
To fix, invent some macros that make it less notationally painful
to do the right thing. Also add documentation about what the
function is actually required to do; and in >= v14, add an assertion
in report_json_context about token_terminator being sufficiently
far advanced.
Per report from Nikolay Shaplov. Back-patch to all supported
versions.
Discussion: https://postgr.es/m/7332649.x5DLKWyVIX@thinkpad-pgpro
check_agg_arguments_walker() supposed that it needn't descend into
the arguments of a lower-level aggregate function, but this is
just wrong in the presence of multiple levels of sub-select. The
oversight would lead to executor failures on queries that should
be rejected. (Prior to v11, they actually were rejected, thanks
to a "redundant" execution-time check.)
Per bug #17835 from Anban Company. Back-patch to all supported
branches.
Discussion: https://postgr.es/m/17835-4f29f3098b2d0ba4@postgresql.org
This allows for a string which if an input field matches causes the
column's default value to be inserted. The advantage of this is that
the default can be inserted in some rows and not others, for which
non-default data is available.
The file_fdw extension is also modified to take allow use of this
option.
Israel Barth Rubio
Discussion: https://postgr.es/m/CAO_rXXAcqesk6DsvioOZ5zmeEmpUN5ktZf-9=9yu+DTr0Xr8Uw@mail.gmail.com
This ensures that the row count in the command tag for a MERGE is
correctly computed in the case where UPDATEs or DELETEs are skipped
due to a BEFORE ROW trigger returning NULL (the INSERT case was
already handled correctly by ExecMergeNotMatched() calling
ExecInsert()).
Back-patch to v15, where MERGE was introduced.
Discussion: https://postgr.es/m/CAEZATCU8XEmR0JWKDtyb7iZ%3DqCffxS9uyJt0iOZ4TV4RT%2Bow1w%40mail.gmail.com
If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW
triggers, or a cross-partition UPDATE (with or without triggers), and
a concurrent UPDATE or DELETE happens, the merge code would fail.
In some cases this would lead to a crash, while in others it would
cause the wrong merge action to be executed, or no action at all. The
immediate cause of the crash was the trigger code calling
ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails
because during a merge ri_projectNew is NULL, since merge has its own
per-action projection information, which ExecGetUpdateNewTuple() knows
nothing about.
Fix by arranging for the trigger code to exit early, returning the
TM_Result and TM_FailureData information, if a concurrent modification
is detected, allowing the merge code to do the necessary EPQ handling
in its own way. Similarly, prevent the cross-partition update code
from doing any EPQ processing for a merge, allowing the merge code to
work out what it needs to do.
This leads to a number of simplifications in nodeModifyTable.c. Most
notably, the ModifyTableContext->GetUpdateNewTuple() callback is no
longer needed, and mergeGetUpdateNewTuple() can be deleted, since
there is no longer any requirement for get-update-new-tuple during a
merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer
needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of
ExecCrossPartitionUpdate() can be restored to how they were in v14,
before the merge code was added, and ExecMergeMatched() no longer
needs any special-case handling for cross-partition updates.
While at it, tidy up ExecUpdateEpilogue() a bit, making it handle
recheckIndexes locally, rather than passing it in as a parameter,
ensuring that it is freed properly. This dates back to when it was
split off from ExecUpdate() to support merge.
Per bug #17809 from Alexander Lakhin, and follow-up investigation of
bug #17792, also from Alexander Lakhin.
Back-patch to v15, where MERGE was introduced, taking care to preserve
backwards-compatibility of the trigger API in v15 for any extensions
that might use it.
Discussion:
https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.orghttps://postgr.es/m/17792-0f89452029662c36%40postgresql.org
adjust_appendrel_attrs can't transfer nullingrel labeling to a non-Var
translation expression (mainly because it's too late to wrap such an
expression in a PlaceHolderVar). I'd supposed in commit 2489d76c4
that that restriction was unreachable because we'd not attempt to push
problematic clauses down to an appendrel child relation. I forgot that
set_append_rel_size blindly converts all the parent rel's joininfo
clauses to child clauses, and that list could well contain clauses
from above a nulling outer join.
We might eventually have to devise a direct fix for this implementation
restriction, but for now it seems enough to filter out troublesome
clauses while constructing the child's joininfo list. Such clauses
are certainly not useful while constructing paths for the child rel;
they'll have to be applied later when we join the completed appendrel
to something else. So we don't need them here, and omitting them from
the list should save a few cycles while processing the child rel.
Per bug #17832 from Marko Tiikkaja.
Discussion: https://postgr.es/m/17832-d0a8106cdf1b722e@postgresql.org
The COPY documentation is quite clear that "COPY relation TO" copies
rows from only the named table, not any inheritance children it may
have. However, if you enabled row-level security on the table then
this stopped being true, because the code forgot to apply the ONLY
modifier in the "SELECT ... FROM relation" query that it constructs
in order to allow RLS predicates to be attached. Fix that.
Report and patch by Antonin Houska (comment adjustments and test case
by me). Back-patch to all supported branches.
Discussion: https://postgr.es/m/3472.1675251957@antos
Datetime input formerly accepted combinations such as
'1995-08-06 infinity', but this seems like a clear error.
Reject any combination of regular y/m/d/h/m/s fields with
these special tokens.
Joseph Koshakow, reviewed by Keisuke Kuroda and myself
Discussion: https://postgr.es/m/CAAvxfHdm8wwXwG_FFRaJ1nTHiMWb7YXS2YKCzCt8Q0a2ZoMcHg@mail.gmail.com
Commit bdaabb9b started skipping doomed transactions when building the
list of possible conflicts for SERIALIZABLE READ ONLY. That makes
sense, because doomed transactions won't commit, but a couple of subtle
things broke:
1. If all uncommitted r/w transactions are doomed, a READ ONLY
transaction would arbitrarily not benefit from the safe snapshot
optimization. It would not be taken immediately, and yet no other
transaction would set SXACT_FLAG_RO_SAFE later.
2. In the same circumstances but with DEFERRABLE, GetSafeSnapshot()
would correctly exit its wait loop without sleeping and then take the
optimization in non-assert builds, but assert builds would fail a sanity
check that SXACT_FLAG_RO_SAFE had been set by another transaction.
This is similar to the case for PredXact->WritableSxactCount == 0. We
should opt out immediately if our possibleUnsafeConflicts list is empty
after filtering.
The code to maintain the serializable global xmin is moved down below
the new opt out site, because otherwise we'd have to reverse its effects
before returning.
Back-patch to all supported releases. Bug #17368.
Reported-by: Alexander Lakhin <exclusion@gmail.com>
Discussion: https://postgr.es/m/17116-d6ca217acc180e30%40postgresql.org
Discussion: https://postgr.es/m/20110707212159.GF76634%40csail.mit.edu
This exposes the ICU facility to add custom collation rules to a
standard collation.
New options are added to CREATE COLLATION, CREATE DATABASE, createdb,
and initdb to set the rules.
Reviewed-by: Laurenz Albe <laurenz.albe@cybertec.at>
Reviewed-by: Daniel Verite <daniel@manitou-mail.org>
Discussion: https://www.postgresql.org/message-id/flat/821c71a4-6ef0-d366-9acf-bb8e367f739f@enterprisedb.com
Reformat some of the comments in MergeAttributes(). A lot of code has
been added here over time, and the comments could use a bit of editing
to make the code flow read better.
When vacuum_defer_cleanup_age is bigger than the current xid, including the
epoch, the subtraction of vacuum_defer_cleanup_age would lead to a wrapped
around xid. While that normally is not a problem, the subsequent conversion to
a 64bit xid results in a 64bit-xid very far into the future. As that xid is
used as a horizon to detect whether rows versions are old enough to be
removed, that allows removal of rows that are still visible (i.e. corruption).
If vacuum_defer_cleanup_age was never changed from the default, there is no
chance of this bug occurring.
This bug was introduced in dc7420c2c9. A lesser version of it exists in
12-13, introduced by fb5344c969, affecting only GiST.
The 12-13 version of the issue can, in rare cases, lead to pages in a gist
index getting recycled too early, potentially causing index entries to be
found multiple times.
The fix is fairly simple - don't allow vacuum_defer_cleanup_age to retreat
further than FirstNormalTransactionId.
Patches to make similar bugs easier to find, by adding asserts to the 64bit
xid infrastructure, have been proposed, but are not suitable for backpatching.
Currently there are no tests for vacuum_defer_cleanup_age. A patch introducing
infrastructure to make writing a test easier has been posted to the list.
Reported-by: Michail Nikolaev <michail.nikolaev@gmail.com>
Reviewed-by: Matthias van de Meent <boekewurm+postgres@gmail.com>
Author: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/20230108002923.cyoser3ttmt63bfn@awork3.anarazel.de
Backpatch: 12-, but impact/fix is smaller for 12-13
4211fbd has been handling PROCESS_MAIN in vacuum_rel() with an "if/else
if" structure to avoid an extra level of indentation, but this has been
found as being rather parse to read. This commit updates the code so as
we check for PROCESS_MAIN in a single place and then handle its
subpaths, FULL or non-FULL vacuums. Some comments are added to make
that clearer for the reader.
Reported-by: Melanie Plageman
Author: Nathan Bossart
Reviewed-by: Michael Paquier, Melanie Plageman
Discussion: https://postgr.es/m/20230306194009.5cn6sp3wjotd36nu@liskov
If a view is defined atop another view, and then CREATE OR REPLACE
VIEW is used to add columns to the lower view, then when the upper
view's referencing RTE is expanded by ApplyRetrieveRule we will have
a subquery RTE with fewer eref->colnames than output columns. This
confuses various code that assumes those lists are always in sync,
as they are in plain parser output.
We have seen such problems before (cf commit d5b760ecb), and now
I think the time has come to do what was speculated about in that
commit: let's make ApplyRetrieveRule synthesize some column names to
preserve the invariant that holds in parser output. Otherwise we'll
be chasing this class of bugs indefinitely. Moreover, it appears from
testing that this actually gives us better results in the test case
d5b760ecb added, and likely in other corner cases that we lack
coverage for.
In HEAD, I replaced d5b760ecb's hack to make expandRTE exit early with
an elog(ERROR) call, since the case is now presumably unreachable.
But it seems like changing that in back branches would bring more risk
than benefit, so there I just updated the comment.
Per bug #17811 from Alexander Lakhin. Back-patch to all supported
branches.
Discussion: https://postgr.es/m/17811-d31686b78f0dffc9@postgresql.org
This routine is able to retrieve the OID of the schema used with an
extension (pg_extension.extnamespace), or InvalidOid if this information
is not available. plpgsql_check embeds a copy of this code when
performing checks on functions, as one out-of-core example.
Author: Pavel Stehule
Reviewed-by: Julien Rouhaud
Discussion: https://postgr.es/m/CAFj8pRD+9x55hjDoi285jCcjPc8uuY_D+FLn5RpXggdz+4O2sQ@mail.gmail.com
If UPDATE is forced to retry after an EvalPlanQual check, it neglected
to repeat GENERATED-column computations, even though those might well
have changed since we're dealing with a different tuple than before.
Fixing this is mostly a matter of looping back a bit further when
we retry. In v15 and HEAD that's most easily done by altering the API
of ExecUpdateAct so that it includes computing GENERATED expressions.
Also, if an UPDATE in a partitioned table turns into a cross-partition
INSERT operation, we failed to recompute GENERATED columns. That's a
bug since 8bf6ec3ba allowed partitions to have different generation
expressions; although it seems to have no ill effects before that.
Fixing this is messier because we can now have situations where the same
query needs both the UPDATE-aligned set of GENERATED columns and the
INSERT-aligned set, and it's unclear which set will be generated first
(else we could hack things by forcing the INSERT-aligned set to be
generated, which is indeed how fe9e658f4 made it work for MERGE).
The best fix seems to be to build and store separate sets of expressions
for the INSERT and UPDATE cases. That would create ABI issues in the
back branches, but so far it seems we can leave this alone in the back
branches.
Per bug #17823 from Hisahiro Kauchi. The first part of this affects all
branches back to v12 where GENERATED columns were added.
Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
While testing a fix for bug #17823, I discovered that EvalPlanQualStart
failed to copy es_rteperminfos from the parent EState, resulting in
failure if anything in EPQ execution wanted to consult that information.
This led me to conclude that commit a61b1f748 had been too haphazard
about where to fill es_rteperminfos, and that we need to be sure that
that happens exactly where es_range_table gets filled. So I changed the
signature of ExecInitRangeTable to help ensure that this new requirement
doesn't get missed. (Indeed, pgoutput.c was also failing to fill it.
Maybe we don't ever need it there, but I wouldn't bet on that.)
No test case yet; one will arrive with the fix for #17823.
But that needs to be back-patched, while this fix is HEAD-only.
Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
The functions that follow are concerned with various things, of
which the tar format is only one, so this comment doesn't really
seem helpful. The file isn't really divided into sections in the
way that this comment seems to contemplate -- or at least, not
any more.
Patch by me, reviewed by Michael Paquier.
Discussion: http://postgr.es/m/CA+TgmoZ_fFAoU6mrHt9QBs+dcYhN6yXenGTTMRebZNhtwPwHyg@mail.gmail.com
We read blocks of data from files that we're backing up in chunks,
some multiple of BLCKSZ for each read. If checksum verification fails,
we then try rereading just the one block for which validation failed.
If that block happened to be the first block of the chunk, and if
the file was concurrently truncated to remove that block, then we'd
reach a call to bbsink_archive_contents() with a buffer length of 0.
That causes an assertion failure.
As far as I can see, there are no particularly bad consequences if
this happens in a non-assert build, and it's pretty unlikely to happen
in the first place because it requires a series of somewhat unlikely
things to happen in very quick succession. However, assertion failures
are bad, so rearrange the code to avoid that possibility.
Patch by me, reviewed by Michael Paquier.
Discussion: http://postgr.es/m/CA+TgmoZ_fFAoU6mrHt9QBs+dcYhN6yXenGTTMRebZNhtwPwHyg@mail.gmail.com
Disabling this option is useful to run VACUUM (with or without FULL) on
only the toast table of a relation, bypassing the main relation. This
option is enabled by default.
Running directly VACUUM on a toast table was already possible without
this feature, by using the non-deterministic name of a toast relation
(as of pg_toast.pg_toast_N, where N would be the OID of the parent
relation) in the VACUUM command, and it required a scan of pg_class to
know the name of the toast table. So this feature is basically a
shortcut to be able to run VACUUM or VACUUM FULL on a toast relation,
using only the name of the parent relation.
A new switch called --no-process-main is added to vacuumdb, to work as
an equivalent of PROCESS_MAIN.
Regression tests are added to cover VACUUM and VACUUM FULL, looking at
pg_stat_all_tables.vacuum_count to see how many vacuums have run on
each table, main or toast.
Author: Nathan Bossart
Reviewed-by: Masahiko Sawada
Discussion: https://postgr.es/m/20221230000028.GA435655@nathanxps13
The following changes are made to pg_write_zeros(), the API able to
write series of zeros using vectored I/O:
- Add of an "offset" parameter, to write the size from this position
(the 'p' of "pwrite" seems to mean position, though POSIX does not
outline ythat directly), hence the name of the routine is incorrect if
it is not able to handle offsets.
- Avoid memset() of "zbuffer" on every call.
- Avoid initialization of the whole IOV array if not needed.
- Group the trailing write() call with the main write() call,
simplifying the function logic.
Author: Andres Freund
Reviewed-by: Michael Paquier, Bharath Rupireddy
Discussion: https://postgr.es/m/20230215005525.mrrlmqrxzjzhaipl@awork3.anarazel.de
1. Make sure that we don't decrement SxactGlobalXminCount twice when
the SXACT_FLAG_RO_SAFE optimization is reached in a parallel query.
This could trigger a sanity check failure in assert builds. Non-assert
builds recompute the count in SetNewSxactGlobalXmin(), so the problem
was hidden, explaining the lack of field reports. Add a new isolation
test to exercise that case.
2. Remove an assertion that the DOOMED flag can't be set on a partially
released SERIALIZABLEXACT. Instead, ignore the flag (our transaction
was already determined to be read-only safe, and DOOMED is in fact set
during partial release, and there was already an assertion that it
wasn't set sooner). Improve an existing isolation test so that it
reaches that case (previously it wasn't quite testing what it was
supposed to be testing; see discussion).
Back-patch to 12. Bug #17116. Defects in commit 47a338cf.
Reported-by: Alexander Lakhin <exclusion@gmail.com>
Discussion: https://postgr.es/m/17116-d6ca217acc180e30%40postgresql.org
Add support for non-decimal integer literals and underscores in
numeric literals to SQL JSON path language. This follows the rules of
ECMAScript, as referred to by the SQL standard.
Internally, all the numeric literal parsing of jsonpath goes through
numeric_in, which already supports all this, so this patch is just a
bit of lexer work and some tests and documentation.
Reviewed-by: Dean Rasheed <dean.a.rasheed@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/b11b25bb-6ec1-d42f-cedd-311eae59e1fb@enterprisedb.com
Beginning in v15, if you apply ALTER TABLE ENABLE/DISABLE TRIGGER to
a partitioned table, it also affects the partitions' cloned versions
of the affected trigger(s). The initial implementation of this
located the clones by name, but that fails on foreign-key triggers
which have names incorporating their own OIDs. We can fix that, and
also make the behavior more bulletproof in the face of user-initiated
trigger renames, by identifying the cloned triggers by tgparentid.
Following the lead of earlier commits in this area, I took care not
to break ABI in the v15 branch, even though I rather doubt there
are any external callers of EnableDisableTrigger.
While here, update the documentation, which was not touched when
the semantics were changed.
Per bug #17817 from Alan Hodgson. Back-patch to v15; older versions
do not have this behavior.
Discussion: https://postgr.es/m/17817-31dfb7c2100d9f3d@postgresql.org
Previously, meson installed modules under src/test/modules/ as part of
a normal installation, even though these files are only meant for use
by tests. This is because there is no way to set up up the build
system to install extra things only when told.
This patch fixes that with a workaround: We don't install these
modules as part of meson install, but we create a new "test" that runs
before the real tests whose action it is to install these files. The
installation is done by manual copies using a small helper script.
Author: Nazir Bilal Yavuz <byavuz81@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/2a039e8e-f31f-31e8-afe7-bab3130ad2de%40enterprisedb.com
Open long-lived data and WAL file descriptors with O_CLOEXEC. This flag
was introduced by SUSv4 (POSIX.1-2008), and by now all of our target
Unix systems have it. Our open() implementation for Windows already had
that behavior, so provide a dummy O_CLOEXEC flag on that platform.
For now, callers of open() and the "thin" wrappers in fd.c that deal in
raw descriptors need to pass in O_CLOEXEC explicitly if desired. This
commit does that for WAL files, and automatically for everything
accessed via VFDs including SMgrRelation and BufFile. (With more
discussion we might decide to turn it on automatically for the thin
open()-wrappers too to avoid risk of missing places that need it, but
these are typically used for short-lived descriptors where we don't
expect to fork/exec, and it's remotely possible that extensions could be
using these APIs and passing descriptors to subprograms deliberately, so
that hasn't been done here.)
Do the same for sockets and the postmaster pipe with FD_CLOEXEC. (Later
commits might use modern interfaces to remove these extra fcntl() calls
and more where possible, but we'll need them as a fallback for a couple
of systems, so do it that way in this initial commit.)
With this change, subprograms executed for archiving, copying etc will
no longer have access to the server's descriptors, other than the ones
that we decide to pass down.
Reviewed-by: Andres Freund <andres@anarazel.de> (earlier version)
Discussion: https://postgr.es/m/CA%2BhUKGKb6FsAdQWcRL35KJsftv%2B9zXqQbzwkfRf1i0J2e57%2BhQ%40mail.gmail.com
When I designed the Bitmapset module, I set things up so that an empty
Bitmapset could be represented either by a NULL pointer, or by an
allocated object all of whose bits are zero. I've recently come to
the conclusion that that was a bad idea and we should instead have a
convention like the longstanding invariant for Lists, whereby an empty
list is represented by NIL and nothing else.
To do this, we need to fix bms_intersect, bms_difference, and a couple
of other functions to check for having produced an empty result; but
then we can replace bms_is_empty(a) by a simple "a == NULL" test.
This is very likely a (marginal) win performance-wise, because we
call bms_is_empty many more times than those other functions put
together. However, the real reason to do it is that we have various
places that have hand-implemented a rule about "this Bitmapset
variable must be exactly NULL if empty", so that they can use
checks-for-null in place of bms_is_empty calls in particularly hot
code paths. That is a really fragile, mistake-prone way to do things,
and I'm surprised that we've seldom been bitten by it. It's not well
documented at all which variables have this property, so you can't
readily tell which code might be violating those conventions. By
making the convention universal, we can eliminate a subtle source of
bugs.
Patch by me; thanks to Nathan Bossart and Richard Guo for review.
Discussion: https://postgr.es/m/1159933.1677621588@sss.pgh.pa.us
nodeAppend.c used non-nullness of appendstate->as_valid_subplans as
a state flag to indicate whether it'd done ExecFindMatchingSubPlans
(or some sufficient approximation to that). This was pretty
questionable even in the beginning, since it wouldn't really work
right if there are no valid subplans. It got more questionable
after commit 27e1f1456 added logic that could reduce as_valid_subplans
to an empty set: at that point we were depending on unspecified
behavior of bms_del_members, namely that it'd not return an empty
set as NULL. It's about to start doing that, which breaks this
logic entirely. Hence, add a separate boolean flag to signal
whether as_valid_subplans has been computed.
Also fix a previously-cosmetic bug in nodeAgg.c, wherein it ignored
the return value of bms_del_member instead of updating its pointer.
Patch by me; thanks to Nathan Bossart and Richard Guo for review.
Discussion: https://postgr.es/m/1159933.1677621588@sss.pgh.pa.us
This function has been semi-deprecated ever since we invented
bms_next_member(). Its habit of scribbling on the input bitmapset
isn't great, plus for sufficiently large bitmapsets it would take
O(N^2) time to complete a loop. Now we have the additional problem
that reducing the input to empty while leaving it still accessible
would violate a planned invariant. So let's just get rid of it,
after updating the few extant callers to use bms_next_member().
Patch by me; thanks to Nathan Bossart and Richard Guo for review.
Discussion: https://postgr.es/m/1159933.1677621588@sss.pgh.pa.us
Some deprecated options were not marked as such in usage output. This
does so across the installed binaries in an attempt to provide consistent
markup for this.
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Discussion: https://postgr.es/m/062C6A8A-A4E8-4F52-9E31-45F0C9E9915E@yesql.se
Commit 0a20ff54f split out the GUC variables from guc.c into a new file
guc_tables.c. This updates comments referencing guc.c regarding variables
which are now in guc_tables.c.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/6B50C70C-8C1F-4F9A-A7C0-EEAFCC032406@yesql.se
When you have some invalid WAL, you often get a message like "wanted
24, got 0". This is a bit incorrect, since it really wanted *at
least* 24, not exactly 24. This updates the messages to that effect,
and also adds that detail to one message where it was available but
not printed.
Reviewed-by: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Reviewed-by: Jeevan Ladhe <jeevanladhe.os@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/726d782b-5e45-0c3e-d775-6686afe9aa83%40enterprisedb.com
This is usually harmless, but if you were very unlucky it could
provoke a segfault due to the "to" string being right up against
the end of memory. Found via valgrind testing (so we might've
found it earlier, except that our regression tests lacked any
exercise of translate()'s deletion feature).
Fix by switching the order of the test-for-end-of-string and
advance-pointer steps. While here, compute "to_ptr + tolen"
just once. (Smarter compilers might figure that out for
themselves, but let's just make sure.)
Report and fix by Daniil Anisimov, in bug #17816.
Discussion: https://postgr.es/m/17816-70f3d2764e88a108@postgresql.org
pg_input_error_info() is now a SQL function able to return a row with
more than just the error message generated for incorrect data type
inputs when these are able to handle soft failures, returning more
contents of ErrorData, as of:
- The error message (same as before).
- The error detail, if set.
- The error hint, if set.
- SQL error code.
All the regression tests that relied on pg_input_error_message() are
updated to reflect the effects of the rename.
Per discussion with Tom Lane and Andrew Dunstan.
Author: Nathan Bossart
Discussion: https://postgr.es/m/139a68e1-bd1f-a9a7-b5fe-0be9845c6311@dunslane.net
Some clang versions whine about comparing an enum variable to
a value outside the range of the enum, on the grounds that the
result must be constant. In the cases we fix here, the loops
will terminate only if the enum variable can in fact hold a
value one beyond its declared range. While that's very likely
to always be true for these enum types, it still seems like a
poor coding practice to assume it; so use "int" loop variables
instead to silence the warnings. (This matches what we've done
in other places, for example loops over the range of ForkNumber.)
While at it, let's drop the XXX_FIRST macros for these enums and just
write zeroes for the loop start values. The apparent flexibility
seems rather illusory given that iterating up to one-less-than-
the-number-of-values is only correct for a zero-based range.
Melanie Plageman
Discussion: https://postgr.es/m/20520.1677435600@sss.pgh.pa.us
We already tried to fix this in commits 3f7323cbb et al (and follow-on
fixes), but now it emerges that there are still unfixed cases;
moreover, these cases affect all branches not only pre-v14. I thought
we had eliminated all cases of making multiple clones of an UPDATE's
target list when we nuked inheritance_planner. But it turns out we
still do that in some partitioned-UPDATE cases, notably including
INSERT ... ON CONFLICT UPDATE, because ExecInitPartitionInfo thinks
it's okay to clone and modify the parent's targetlist.
This fix is based on a suggestion from Andres Freund: let's stop
abusing the ParamExecData.execPlan mechanism, which was only ever
meant to handle initplans, and instead solve the execution timing
problem by having the expression compiler move MULTIEXPR_SUBLINK steps
to the front of their expression step lists. This is feasible because
(a) all branches still in support compile the entire targetlist of
an UPDATE into a single ExprState, and (b) we know that all
MULTIEXPR_SUBLINKs do need to be evaluated --- none could be buried
inside a CASE, for example. There is a minor semantics change
concerning the order of execution of the MULTIEXPR's subquery versus
other parts of the parent targetlist, but that seems like something
we can get away with. By doing that, we no longer need to worry
about whether different clones of a MULTIEXPR_SUBLINK share output
Params; their usage of that data structure won't overlap.
Per bug #17800 from Alexander Lakhin. Back-patch to all supported
branches. In v13 and earlier, we can revert 3f7323cbb and follow-on
fixes; however, I chose to keep the SubPlan.subLinkId field added
in ccbb54c72. We don't need that anymore in the core code, but it's
cheap enough to fill, and removing a plan node field in a minor
release seems like it'd be asking for trouble.
Andres Freund and Tom Lane
Discussion: https://postgr.es/m/17800-ff90866b3906c964@postgresql.org
If a rule action contains a subquery that refers to columns from OLD
or NEW, then those are really lateral references, and the planner will
complain if it sees such things in a subquery that isn't marked as
lateral. However, at rule-definition time, the user isn't required to
mark the subquery with LATERAL, and so it can fail when the rule is
used.
Fix this by marking such subqueries as lateral in the rewriter, at the
point where they're used.
Dean Rasheed and Tom Lane, per report from Alexander Lakhin.
Back-patch to all supported branches.
Discussion: https://postgr.es/m/5e09da43-aaba-7ea7-0a51-a2eb981b058b%40gmail.com
A unique index which is created with non-distinct NULLS cannot be
used for backing a primary key constraint. Make sure to disallow
such table alterations and teach pg_dump to drop the non-distinct
NULLS clause on indexes where this has been set.
Bug: 17720
Reported-by: Reiner Peterke <zedaardv@drizzle.com>
Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/17720-dab8ee0fa85d316d@postgresql.org
Multiple cycles of starting up and shutting down the plugin within a
single session would eventually lead to "out of relcache_callback_list
slots", because pgoutput_startup blindly re-registered its cache
callbacks each time. Fix it to register them only once, as all other
users of cache callbacks already take care to do.
This has been broken all along, so back-patch to all supported branches.
Shi Yu
Discussion: https://postgr.es/m/OSZPR01MB631004A78D743D68921FFAD3FDA79@OSZPR01MB6310.jpnprd01.prod.outlook.com
Offers a generally better separation of responsibilities for collation
code. Also, a step towards multi-lib ICU, which should be based on a
clean separation of the routines required for collation providers.
Callers with NUL-terminated strings should call pg_strcoll() or
pg_strxfrm(); callers with strings and their length should call the
variants pg_strncoll() or pg_strnxfrm().
Reviewed-by: Peter Eisentraut, Peter Geoghegan
Discussion: https://postgr.es/m/a581136455c940d7bd0ff482d3a2bd51af25a94f.camel%40j-davis.com
It's possible, in admittedly-rather-contrived cases, for an eclass
to generate a derived "join" qual that constrains the post-outer-join
value(s) of some RHS variable(s) without mentioning the LHS at all.
While the mechanisms were set up to work for this, we fell foul of
the "get_common_eclass_indexes" filter installed by commit 3373c7155:
it could decide that such an eclass wasn't relevant to the join, so
that the required qual clause wouldn't get emitted there or anywhere
else.
To fix, apply get_common_eclass_indexes only at inner joins, where
its rule is still valid. At an outer join, fall back to examining all
eclasses that mention either input (or the OJ relid, though it should
be impossible for an eclass to mention that without mentioning either
input). Perhaps we can improve on that later, but the cost/benefit of
adding more complexity to skip some irrelevant eclasses is dubious.
To allow cheaply distinguishing outer from inner joins, pass the
ojrelid to generate_join_implied_equalities as a separate argument.
This also allows cleaning up some sloppiness that had crept into
the definition of its join_relids argument, and it allows accurate
calculation of nominal_join_relids for a child outer join. (The
latter oversight seems not to have been a live bug, but it certainly
could have caused problems in future.)
Also fix what might be a live bug in check_index_predicates: it was
being sloppy about what it passed to generate_join_implied_equalities.
Per report from Richard Guo.
Discussion: https://postgr.es/m/CAMbWs4-DsTBfOvXuw64GdFss2=M5cwtEhY=0DCS7t2gT7P6hSA@mail.gmail.com
Given an updatable view with a DO ALSO INSERT ... SELECT rule, a
multi-row INSERT ... VALUES query on the view fails if the VALUES list
contains any DEFAULTs that are not replaced by view defaults. This
manifests as an "unrecognized node type" error, or an Assert failure,
in an assert-enabled build.
The reason is that when RewriteQuery() attempts to replace the
remaining DEFAULT items with NULLs in any product queries, using
rewriteValuesRTEToNulls(), it assumes that the VALUES RTE is located
at the same rangetable index in each product query. However, if the
product query is an INSERT ... SELECT, then the VALUES RTE is actually
in the SELECT part of that query (at the same index), rather than the
top-level product query itself.
Fix, by descending to the SELECT in such cases. Note that we can't
simply use getInsertSelectQuery() for this, since that expects to be
given a raw rule action with OLD and NEW placeholder entries, so we
duplicate its logic instead.
While at it, beef up the checks in getInsertSelectQuery() by checking
that the jointree->fromlist node is indeed a RangeTblRef, and that the
RTE it points to has rtekind == RTE_SUBQUERY.
Per bug #17803, from Alexander Lakhin. Back-patch to all supported
branches.
Dean Rasheed, reviewed by Tom Lane.
Discussion: https://postgr.es/m/17803-53c63ed4ecb4eac6%40postgresql.org
initsplan.c figured that it could push Var-free qual clauses to
the top of the current JoinDomain, which is okay in the abstract.
But if the current domain is inside some outer join, and we later
commute an inside-the-domain outer join with one outside it,
we end up placing the pushed-up qual clause incorrectly.
In distribute_qual_to_rels, avoid this by using the syntactic scope
of the qual clause; with the exception that if we're in the top-level
join domain we can still use the full query relid set, ensuring the
resulting gating Result node goes to the top of the plan. (This is
approximately as smart as the pre-v16 code was. Perhaps we can do
better later, but it's not clear that such cases are worth a lot of
sweat.)
In process_implied_equality, we don't have a clear notion of syntactic
scope, but we do have the results of SpecialJoinInfo construction.
Thumb through those and remove any lower outer joins that might get
commuted to above the join domain. Again, we can make an exception
for the top-level join domain. It'd be possible to work harder here
(for example, by keeping outer joins that aren't shown as potentially
commutable), but I'm going to stop here for the moment. This issue
has convinced me that the current representation of join domains
probably needs further refinement, so I'm disinclined to write
inessential dependent logic just yet.
In passing, tighten the qualscope passed to process_implied_equality
by generate_base_implied_equalities_no_const; there's no need for
it to be larger than the rel we are currently considering.
Tom Lane and Richard Guo, per report from Tender Wang.
Discussion: https://postgr.es/m/CAHewXNk9eJ35ru5xATWioTV4+xZPHptjy9etdcNPjUfY9RQ+uQ@mail.gmail.com
Whe decoding a transactional logical message, logicalmsg_decode called
SnapBuildGetOrBuildSnapshot. But we may not have a consistent snapshot
yet at that point. We don't actually need the snapshot in this case
(during replay we'll have the snapshot from the transaction), so in
practice this is harmless. But in assert-enabled build this crashes.
Fixed by requesting the snapshot only in non-transactional case, where
we are guaranteed to have SNAPBUILD_CONSISTENT.
Backpatch to 11. The issue exists since 9.6.
Backpatch-through: 11
Reviewed-by: Andres Freund
Discussion: https://postgr.es/m/84d60912-6eab-9b84-5de3-41765a5449e8@enterprisedb.com
In ExecInitPartitionInfo(), the Assert when building the WITH CHECK
OPTION list for the new partition assumed that the command would be an
INSERT or UPDATE, but it can also be a MERGE. This can be triggered by
a MERGE into a partitioned table with RLS checks to enforce.
Fix, and back-patch to v15, where MERGE was introduced.
Discussion: https://postgr.es/m/CAEZATCWWFtQmW67F3XTyMU5Am10Oxa_b8oe0x%2BNu5Mo%2BCdRErg%40mail.gmail.com
This ensures that the row count in the command tag for a MERGE is
correctly computed. Previously, if MERGE updated a partitioned table,
the row count would be incorrect if any row was moved to a different
partition, since such updates were counted twice.
Back-patch to v15, where MERGE was introduced.
Discussion: https://postgr.es/m/CAEZATCWRMG7XX2QEsVL1LswmNo2d_YG8tKTLkpD3=Lp644S7rg@mail.gmail.com
SQL:2023 defines an ANY_VALUE aggregate whose purpose is to emit an
implementation-dependent (i.e. non-deterministic) value from the
aggregated rows.
Author: Vik Fearing <vik@postgresfriends.org>
Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/5cff866c-10a8-d2df-32cb-e9072e6b04a2@postgresfriends.org
WAL_LOG does a scan of the template's pg_class to determine the set of
relations that need to be copied from a template database to the new
one. However, as coded in 9c08aea, this copy strategy would load the
pages of pg_class without considering it as a permanent relation,
causing the loaded pages to never be flushed when they should. Any
modification of the template's pg_class, mostly through DDLs, would then
be missed, causing corruptions.
STRATEGY = WAL_LOG is the default over FILE_COPY since it has been
introduced, so any changes done to pg_class on a database template would
be gone. Updates of database templates should be a rare thing, so the
impact of this bug should be hopefully limited. The pre-14 default
strategy FILE_COPY is safe, and can be used as a workaround.
Ryo Matsumura has found and analyzed the issue, and Nathan has written a
test able to reproduce the failure (with few tweaks from me).
Backpatch down to 15, where STRATEGY = WAL_LOG has been introduced.
Author: Nathan Bossart, Ryo Matsumura
Reviewed-by: Dilip Kumar, Michael Paquier
Discussion: https://postgr.es/m/TYCPR01MB6868677E499C9AD5123084B5E8A39@TYCPR01MB6868.jpnprd01.prod.outlook.com
Backpatch-through: 15
If asked to decrease the size of a large (>8K) palloc chunk,
AllocSetRealloc could improperly change the Valgrind state of memory
beyond the new end of the chunk: it would mark data UNDEFINED as far
as the old end of the chunk after having done the realloc(3) call,
thus tromping on the state of memory that no longer belongs to it.
One would normally expect that memory to now be marked NOACCESS,
so that this mislabeling might prevent detection of later errors.
If realloc() had chosen to move the chunk someplace else (unlikely,
but well within its rights) we could also mismark perfectly-valid
DEFINED data as UNDEFINED, causing false-positive valgrind reports
later. Also, any malloc bookkeeping data placed within this area
might now be wrongly marked, causing additional problems.
Fix by replacing relevant uses of "oldsize" with "Min(size, oldsize)".
It's sufficient to mark as far as "size" when that's smaller, because
whatever remains in the new chunk size will be marked NOACCESS below,
and we expect realloc() to have taken care of marking the memory
beyond the new official end of the chunk.
While we're here, also rename the function's "oldsize" variable
to "oldchksize" to more clearly explain what it actually holds,
namely the distance to the end of the chunk (that is, requested size
plus trailing padding). This is more consistent with the use of
"size" and "chksize" to hold the new requested size and chunk size.
Add a new variable "oldsize" in the one stanza where we're actually
talking about the old requested size.
Oversight in commit c477f3e44. Back-patch to all supported branches,
as that was, just in case anybody wants to do valgrind testing on back
branches.
Karina Litskevich
Discussion: https://postgr.es/m/CACiT8iaAET-fmzjjZLjaJC4zwSJmrFyL7LAdHwaYyjjQOQ4hcg@mail.gmail.com
Commits 04cad8f7 and 0c088568 supported old macOS systems that didn't
define O_CLOEXEC or O_DSYNC yet, but those arrived in macOS releases
10.7 and 10.6 (respectively), which themselves reached EOL around a
decade ago. We've already made use of other POSIX features that early
macOS vintages can't compile (for example commits 623cc673, d2e15083).
A later commit will use O_CLOEXEC on POSIX systems so it would be
strange to pretend here that it's optional, and we might as well give
O_DSYNC the same treatment since the reference is also guarded by a test
for a macOS-specific macro, and we know that current Macs have it.
Discussion: https://postgr.es/m/CA%2BhUKGKb6FsAdQWcRL35KJsftv%2B9zXqQbzwkfRf1i0J2e57%2BhQ%40mail.gmail.com
Historically we've accepted interval input like 'P.1e10D'. This
is probably an accident of having used strtod() to do the parsing,
rather than something anyone intended, but it's been that way for
a long time. Commit e39f99046 broke this by trying to parse the
integer and fractional parts separately, without accounting for
the possibility of an exponent. In principle that coding allowed
for precise conversions of field values wider than 15 decimal
digits, but that does not seem like a goal worth sweating bullets
for. So, rather than trying to manage an exponent on top of the
existing complexity, let's just revert to the previous coding that
used strtod() by itself. We can still improve on the old code to
the extent of allowing the value to range up to 1.0e15 rather than
only INT_MAX. (Allowing more than that risks creating problems
due to precision loss: the converted fractional part might have
absolute value more than 1. Perhaps that could be dealt with in
some way, but it really does not seem worth additional effort.)
Per bug #17795 from Alexander Lakhin. Back-patch to v15 where
the faulty code came in.
Discussion: https://postgr.es/m/17795-748d6db3ed95d313@postgresql.org
This was not something that required consideration before MERGE
was invented; but MERGE builds a join tree that left-joins to the
result relation, meaning that remove_useless_joins will consider
removing it. That should generally be stopped by the query's use
of output variables from the result relation. However, if the
result relation is inherited (e.g. a partitioned table) then
we don't add any row identity variables to the query until
expand_inherited_rtentry, which happens after join removal.
This was exposed as of commit 3c569049b, which made it possible
to deduce that a partitioned table could contain at most one row
matching a join key, enabling removal of the not-yet-expanded
result relation. Ooops.
To fix, let's just teach join_is_removable that the query result
rel is never removable. It's a cheap enough test in any case,
and it'll save some cycles that we'd otherwise expend in proving
that it's not removable, even in the cases we got right.
Back-patch to v15 where MERGE was added. Although I think the
case cannot be reached in v15, this seems like cheap insurance.
Per investigation of a report from Alexander Lakhin.
Discussion: https://postgr.es/m/36bee393-b351-16ac-93b2-d46d83637e45@gmail.com
For no clearly good reason, make_modifytable assumed that it
could not reach its get-the-FDW-info-the-hard-way path in MERGE.
It's currently possible to demonstrate that assertion failing,
which seems to be due to an upstream planner bug; but there's no
good reason to do it like this at all. Let's apply the principle
of separation of concerns and make the MERGE check separately,
after getting or not getting the fdwroutine pointer.
Per report from Alexander Lakhin. No test case, since I think
the potential test condition will go away soon.
Discussion: https://postgr.es/m/36bee393-b351-16ac-93b2-d46d83637e45@gmail.com
The RelOptInfo->userid field (the user ID to check permissions as) of an
"otherrel" relation was being copied from its parent relation, which is
correct in most cases but wrong when the parent is a subquery. In that
case, using the value from the RTEPermissionInfo of the child itself is
the appropriate thing to do.
Coming up with a test case where user-visible behavior changes proves
hard enough, so we don't add one here.
Bug introduced by a61b1f7482, discovered by Amit while reviewing
nearby code.
Author: Amit Langote <amitlangote09@gmail.com>
Discussion: https://postgr.es/m/CA+HiwqE0WY_AhLnGtTsY7eYebG212XWbM-D8gr2A_ToOHyCywQ@mail.gmail.com
In generate_orderedappend_paths(), when match_partition_order_desc was
true, we would lcons() items to various lists in a loop over each live
partition. When the number of live partitions was large, the lcons()
could show up in profiles due to it having to perform memmove() to make
way for the new list item.
Here we adjust things so that we just perform the loop over the live
partitions backwards when match_partition_order_desc is true. This allows
us to simplify the logic in the loop. Now, as far as the guts of the loop
knows, there's no difference between match_partition_order and
match_partition_order_desc. We can just set match_partition_order to true
so that we build the correct list of paths for the asc and desc case. Per
idea from Andres Freund.
Discussion: https://postgr.es/m/20230217002351.nyt4y5tdzg6hugdt@awork3.anarazel.de
To allow testing for general support for fast bitscan intrinsics,
add symbols HAVE_BITSCAN_REVERSE and HAVE_BITSCAN_FORWARD.
Also do related cleanup in AllocSetFreeIndex(): Previously, we
tested for HAVE__BUILTIN_CLZ and copied the relevant internals of
pg_leftmost_one_pos32(), with a special fallback that does less
work than the general fallback for that function. Now that we have
a more general test, we just call pg_leftmost_one_pos32() directly
for platforms with intrinsic support. On gcc at least, there is no
difference in the binary for non-assert builds.
Discussion: https://www.postgresql.org/message-id/CAFBsxsEPc%2BFnX_0vmmQ5DHv60sk4rL_RZJ%2BMD6ei%3D76L0kFMvA%40mail.gmail.com
The setting of the process title could be seen on profiles of very
fast-to-execute queries. In many locations where we call
set_ps_display() we pass along a string constant, the length of which is
known during compilation. Here we effectively rename set_ps_display() to
set_ps_display_with_len() and then add a static inline function named
set_ps_display() which calls strlen() on the given string. This allows
the compiler to optimize away the strlen() call when dealing with
call sites passing a string constant. We can then also use memcpy()
instead of strlcpy() to copy the string into the destination buffer.
That's significantly faster than strlcpy's byte-at-a-time way of
copying.
Here we also take measures to improve some code which was adjusting the
process title to add a " waiting" suffix to it. Call sites which require
this can now just call set_ps_display_suffix() to add or adjust the suffix
and call set_ps_display_remove_suffix() to remove it again.
Reviewed-by: Andres Freund
Discussion: https://postgr.es/m/CAApHDvocBvvk-0gWNA2Gohe+sv9fMcv+fK_G+siBKJrgDG4O7g@mail.gmail.com
When evaluating clauses on multiple scan keys of a multi-column BRIN
index, we can stop processing as soon as we find a scan key eliminating
the range, and the range should not be added to tbe bitmap.
That's how it worked before 14, but since a681e3c107 the code treated
the range as matching if it matched at least the last scan key.
Backpatch to 14, where this code was introduced.
Backpatch-through: 14
Discussion: https://postgr.es/m/ebc18613-125e-60df-7520-fcbe0f9274fc%40enterprisedb.com
ruleutils.c blindly printed the user-given alias (or nothing if there
hadn't been one) for the target table of INSERT/UPDATE/DELETE queries.
That works a large percentage of the time, but not always: for queries
appearing in WITH, it's possible that we chose a different alias to
avoid conflict with outer-scope names. Since the chosen alias would
be used in any Var references to the target table, this'd lead to an
inconsistent printout with consequences such as dump/restore failures.
The correct logic for printing (or not) a relation alias was embedded
in get_from_clause_item. Factor it out to a separate function so that
we don't need a jointree node to use it. (Only a limited part of that
function can be reached from these new call sites, but this seems like
the cleanest non-duplicative factorization.)
In passing, I got rid of a redundant "\d+ rules_src" step in rules.sql.
Initial report from Jonathan Katz; thanks to Vignesh C for analysis.
This has been broken for a long time, so back-patch to all supported
branches.
Discussion: https://postgr.es/m/e947fa21-24b2-f922-375a-d4f763ef3e4b@postgresql.org
Discussion: https://postgr.es/m/CALDaNm1MMntjmT_NJGp-Z=xbF02qHGAyuSHfYHias3TqQbPF2w@mail.gmail.com
A new callback named startup_cb, called shortly after a module is
loaded, is added. This makes possible the initialization of any
additional state data required by a module. This initial state data can
be saved in a ArchiveModuleState, that is now passed down to all the
callbacks that can be defined in a module. With this design, it is
possible to have a per-module state, aimed at opening the door to the
support of more than one archive module.
The initialization of the callbacks is changed so as
_PG_archive_module_init() does not anymore give in input a
ArchiveModuleCallbacks that a module has to fill in with callback
definitions. Instead, a module now needs to return a const
ArchiveModuleCallbacks.
All the structure and callback definitions of archive modules are moved
into their own header, named archive_module.h, from pgarch.h.
Command-based archiving follows the same line, with a new set of files
named shell_archive.{c,h}.
There are a few more items that are under discussion to improve the
design of archive modules, like the fact that basic_archive calls
sigsetjmp() by itself to define its own error handling flow. These will
be adjusted later, the changes done here cover already a good portion
of what has been discussed.
Any modules created for v15 will need to be adjusted to this new
design.
Author: Nathan Bossart
Reviewed-by: Andres Freund
Discussion: https://postgr.es/m/20230130194810.6fztfgbn32e7qarj@awork3.anarazel.de
In commit 8bf6ec3ba, I mistakenly supposed that MergeAttributes'
loop over saved_schema was reprocessing column definitions that
had already been checked earlier: there is a variant syntax for
creating a child partition in which that's not true. So we need
to duplicate the full check appearing further up.
(Actually, I believe that the "if (restdef->identity)" part is
not reachable, because we reject identity on partitions earlier.
But it seems wise to keep the check, in case that's ever relaxed,
and to keep this code in sync with the other instance.)
Per report from Alexander Lakhin.
Discussion: https://postgr.es/m/4a8200ca-8378-653e-38ed-b2e1f1611aa6@gmail.com
d9d7fe68d3 made use of an existing wait event when sending data from the
apply worker, but we should have invented a new wait event since this is a
new place to wait.
This patch corrects the mistake by using a new wait event
"LogicalApplySendData".
Author: Hou Zhijie
Reviewed-by: Peter Smith
Discussion: https://postgr.es/m/CA+TgmobWzbr9H3yN3dLVckviEZKemPwd+XyCFKEgyZQZhgP66Q@mail.gmail.com
Support for regexps in database and role entries for pg_hba.conf has
been added in 8fea8683, and efb6f4a has extended support of pg-user in
pg_ident.conf, still both of them have missed a short description about
the new patterns supported in their respective sample files.
This commit closes the gap, by providing a short description of all the
new features supported for each entry type.
Reported-by: Pavel Luzanov
Reviewed-by: Jelte Fennema, Pavel Luzanov
Discussion: https://postgr.es/m/e495112d-8741-e651-64a2-ecb5728f1a56@postgrespro.ru
On MERGE / WHEN MATCHED DELETE it's not possible to get cross-partition
updates, so we don't initialize cpUpdateRetrySlot; however, the code was
not careful to ignore the value in that case. Make it do so.
Backpatch to 15.
Reported-by: Alexander Lakhin <exclusion@gmail.com>
Reviewed-by: Dean Rasheed <dean.a.rasheed@gmail.com>
Discussion: https://postgr.es/m/17792-0f89452029662c36@postgresql.org
force_parallel_mode is meant to be used to allow us to exercise the
parallel query infrastructure to ensure that it's working as we expect.
It seems some users think this GUC is for forcing the query planner into
picking a parallel plan regardless of the costs. A quick look at the
documentation would have made them realize that they were wrong, but the
GUC is likely too conveniently named which, evidently, seems to often
result in users expecting that it forces the planner into usefully
parallelizing queries.
Here we rename the GUC to something which casual users are less likely to
mistakenly think is what they need to make their query run more quickly.
For now, the old name can still be used. We'll revisit if the old name
mapping can be removed once the buildfarm configs are all updated.
Reviewed-by: John Naylor
Discussion: https://postgr.es/m/CAApHDvrsOi92_uA7PEaHZMH-S4Xv+MGhQWA+GrP8b1kjpS1HjQ@mail.gmail.com
OpenSSL 1.1.1 and newer versions have added support for RSA-PSS
certificates, which requires the use of a specific routine in OpenSSL to
determine which hash function to use when compiling it when using
channel binding in SCRAM-SHA-256. X509_get_signature_nid(), that is the
original routine the channel binding code has relied on, is not able to
determine which hash algorithm to use for such certificates. However,
X509_get_signature_info(), new to OpenSSL 1.1.1, is able to do it. This
commit switches the channel binding logic to rely on
X509_get_signature_info() over X509_get_signature_nid(), which would be
the choice when building with 1.1.1 or newer.
The error could have been triggered on the client or the server, hence
libpq and the backend need to have their related code paths patched.
Note that attempting to load an RSA-PSS certificate with OpenSSL 1.1.0
or older leads to a failure due to an unsupported algorithm.
The discovery of relying on X509_get_signature_info() comes from Jacob,
the tests have been written by Heikki (with few tweaks from me), while I
have bundled the whole together while adding the bits needed for MSVC
and meson.
This issue exists since channel binding exists, so backpatch all the way
down. Some tests are added in 15~, triggered if compiling with OpenSSL
1.1.1 or newer, where the certificate and key files can easily be
generated for RSA-PSS.
Reported-by: Gunnar "Nick" Bluth
Author: Jacob Champion, Heikki Linnakangas
Discussion: https://postgr.es/m/17760-b6c61e752ec07060@postgresql.org
Backpatch-through: 11
In commit b78f6264e I opined that it was "too risky" to delete a
relation's RelOptInfo from the planner's data structures when we have
realized that we don't need to join to it; so instead we just marked
it as a dead relation. In hindsight that judgment seems flawed: any
subsequent access to such a dead relation is arguably a bug in
itself, so leaving the RelOptInfo present just helps to mask bugs.
Let's delete it instead, allowing removal of the whole notion of a
"dead relation". So far as the regression tests can find, this
requires no other code changes, except for one Assert in equivclass.c
that was very dubiously not complaining about access to a dead rel.
Discussion: https://postgr.es/m/229905.1676062220@sss.pgh.pa.us
Late in the development of commit 2489d76c4, I (tgl) incorrectly
concluded that the new function have_unsafe_outer_join_ref couldn't
ever reach its inner loop. That should be the case if the inner
rel's parameterization is based on just one Var, but it could be
based on Vars from several relations, and then not only is the
inner loop reachable but it's wrongly coded.
Despite those errors, it still appears that the whole thing is
redundant given previous join_is_legal checks, so let's arrange
to only run it in assert-enabled builds.
Diagnosis and patch by Richard Guo, per fuzz testing by Justin Pryzby.
Discussion: https://postgr.es/m/20230212235823.GW1653@telsasoft.com
In commit ad89a5d115, we added an unhelpful 'ON' that doesn't match
the input syntax. This was discovered while adding code to support for
DDL in logical replication.
No backpatch because of the change of behavior, however improbable it
may be that somebody is depending on this.
Author: Zheng Li <zhengli10@gmail.com>
Discussion: https://postgr.es/m/CAAD30UKg8rXeGM8Oy_MAmxKBL_K5DiHXdeNF=hUefcu1C_6VfQ@mail.gmail.com
The logic in this area was recently changed in 7da51590e, however, in that
commit, I neglected to consider that the conditions in which we should
pfree the old Datum needed to be updated after that change. This could
result in trying to pfree a NULL value, as was demonstrated by Alexander
Lakhin.
Reported-by: Alexander Lakhin
Discussion: https://postgr.es/m/4103db46-d888-6d1d-e88d-87c21ed99472@gmail.com
Here we fix a faulty "if" condition which failed to correctly handle two
or more consecutive NULL transition values when checking if the new value
is DISTINCT from the old value for presorted aggregates. Given a suitably
non-strict aggregate transition function, a byref aggregate could cause a
crash due to calling the type's equality function and passing along a
(Datum) 0 value to test for equality, the equality function would then try
to dereference that 0 Datum and segfault. For byval types, there'd have
been no crash and the equality function would have seen that the two 0
Datums matched, which (only by chance) meant the calling code would have
worked correctly.
Here we ensure that we only call the equality function when neither of
the input values are NULL.
This code is all new as of 1349d2790, so no backpatch needed.
Reported-by: Fujii Masao
Discussion: https://postgr.es/m/860c6d6f-a3c5-3ae9-9da2-827177bede06@oss.nttdata.com
When an aggregate function is used as a WindowFunc and a tuple transitions
out of the window frame, we ordinarily try to make use of the aggregate
function's inverse transition function to "unaggregate" the exiting tuple.
This optimization is disabled for various cases, including when the
aggregate contains a volatile function. In such a case we'd be unable to
ensure that the transition value was calculated to the same value during
transitions and inverse transitions. Unfortunately, we did this check by
calling contain_volatile_functions() which does not recursively search
SubPlans for volatile functions. If the aggregate function's arguments or
its FILTER clause contained a subplan with volatile functions then we'd
fail to notice this.
Here we fix this by just disabling the optimization when the WindowFunc
contains any subplans. Volatile functions are not the only reason that a
subplan may have nonrepeatable results.
Bug: #17777
Reported-by: Anban Company
Discussion: https://postgr.es/m/17777-860b739b6efde977%40postgresql.org
Reviewed-by: Tom Lane
Backpatch-through: 11
This commit removes most of the Plan and Path nodes, which should never
be included in the query jumbling because we ignore these in Query
nodes. This is facilitated by making no_query_jumble an inherited
attribute, like no_copy, no_equal and no_read when the supertype of a
node is found as marked with that.
RawStmt is not used in parsed queries, so it can be removed from the
query jumbling. A couple of nodes defined in pathnodes.h, plannodes.h
and primnodes.h with NodeTag as supertype need to be marked
individually.
Forcing the execution of the query jumbling code with compute_query_id =
auto while pg_stat_statements is loaded brings the code coverage of
queryjumblefuncs.funcs.c to 95.6%.
The core code does not yet include a way to enforce the execution in
query jumbling except in pg_stat_statements, so the numbers I am
mentioning above will not reflect on the default coverage report with
just what is done in this commit.
Reported-by: Tom Lane
Reviewed-by: Tom Lane
Discussion: https://postgr.es/m/3344827.1675809127@sss.pgh.pa.us
Commit e39f99046 moved some code up closer to the start of
DecodeInterval(), without noticing that it had been implicitly
relying on previous checks to reject the case of empty input.
Given empty input, we'd now dereference a pointer that hadn't been
set, possibly leading to a core dump. (But if we fail to provoke
a SIGSEGV, nothing bad happens, and the expected syntax error is
thrown a bit later.)
Per bug #17788 from Alexander Lakhin. Back-patch to v15 where
the fault was introduced.
Discussion: https://postgr.es/m/17788-dabac9f98f7eafd5@postgresql.org
Builds on 28e626bde0 and f30d62c2fc. See the former for motivation.
Rows of the view show IO operations for a particular backend type, IO target
object, IO context combination (e.g. a client backend's operations on
permanent relations in shared buffers) and each column in the view is the
total number of IO Operations done (e.g. writes). So a cell in the view would
be, for example, the number of blocks of relation data written from shared
buffers by client backends since the last stats reset.
In anticipation of tracking WAL IO and non-block-oriented IO (such as
temporary file IO), the "op_bytes" column specifies the unit of the "reads",
"writes", and "extends" columns for a given row.
Rows for combinations of IO operation, backend type, target object and context
that never occur, are ommitted entirely. For example, checkpointer will never
operate on temporary relations.
Similarly, if an IO operation never occurs for such a combination, the IO
operation's cell will be null, to distinguish from 0 observed IO
operations. For example, bgwriter should not perform reads.
Note that some of the cells in the view are redundant with fields in
pg_stat_bgwriter (e.g. buffers_backend). For now, these have been kept for
backwards compatibility.
Bumps catversion.
Author: Melanie Plageman <melanieplageman@gmail.com>
Author: Samay Sharma <smilingsamay@gmail.com>
Reviewed-by: Maciek Sakrejda <m.sakrejda@gmail.com>
Reviewed-by: Lukas Fittl <lukas@fittl.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Justin Pryzby <pryzby@telsasoft.com>
Discussion: https://postgr.es/m/20200124195226.lth52iydq2n2uilq@alap3.anarazel.de
analyzejoins.c took care to clean out removed relids from the
clause_relids and required_relids of RestrictInfos associated with
the doomed rel ... but it paid no attention to the fact that if such a
RestrictInfo contains an OR clause, there will be sub-RestrictInfos
containing similar fields.
I'm more than a bit surprised that this oversight hasn't caused
visible problems before. In any case, it's certainly broken now,
so add logic to clean out the sub-RestrictInfos recursively.
We might need to back-patch this someday.
Per bug #17786 from Robins Tharakan.
Discussion: https://postgr.es/m/17786-f1ea7fbdab97daec@postgresql.org
One of the add_nulling_relids calls in deconstruct_distribute_oj_quals
added an OJ relid to too few Vars, while the other added it to too
many. We should consider the syntactic structure not
min_left/righthand while deciding which Vars to decorate, and when
considering pushing up a lower outer join pursuant to transforming the
second form of OJ identity 3 to the first form, we only want to
decorate Vars coming from its LHS.
In a related bug, I realized that make_outerjoininfo was failing to
check a very basic property that's needed to apply OJ identity 3:
the syntactically-upper outer join clause can't refer to the lower
join's LHS. This didn't break the join order restriction logic,
but it led to setting bogus commute_xxx bits, possibly resulting
in bogus nullingrel markings in modified quals.
Richard Guo and Tom Lane
Discussion: https://postgr.es/m/CAMbWs497CmBruMx1SOjepWEz+T5NWa4scqbdE9v7ZzSXqH_gQw@mail.gmail.com
Discussion: https://postgr.es/m/CAEP4nAx9C5gXNBfEA0JBfz7B+5f1Bawt-RWQWyhev-wdps8BZA@mail.gmail.com
Commit 28e626bde0 introduced the infrastructure for tracking more detailed IO
statistics. This commit adds the actual collection of the new IO statistics
for relations and temporary relations. See aforementioned commit for goals and
high-level design.
The changes in this commit are fairly straight-forward. The bulk of the change
is to passing sufficient information to the callsites of pgstat_count_io_op().
A somewhat unsightly detail is that it currently is hard to find a better
place to count fsyncs than in md.c, whereas the other pgstat_count_io_op()
calls are in bufmgr.c/localbuf.c. As the number of fsyncs is tied to md.c
implementation details, it's not obvious there is a better answer.
Author: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/20200124195226.lth52iydq2n2uilq@alap3.anarazel.de