Before executing a cached generic plan, AcquireExecutorLocks() in
plancache.c locks all relations in a plan's range table to ensure the
plan is safe for execution. However, this locks runtime-prunable
relations that will later be pruned during "initial" runtime pruning,
introducing unnecessary overhead.
This commit defers locking for such relations to executor startup and
ensures that if the CachedPlan is invalidated due to concurrent DDL
during this window, replanning is triggered. Deferring these locks
avoids unnecessary locking overhead for pruned partitions, resulting
in significant speedup, particularly when many partitions are pruned
during initial runtime pruning.
* Changes to locking when executing generic plans:
AcquireExecutorLocks() now locks only unprunable relations, that is,
those found in PlannedStmt.unprunableRelids (introduced in commit
cbc127917e), to avoid locking runtime-prunable partitions
unnecessarily. The remaining locks are taken by
ExecDoInitialPruning(), which acquires them only for partitions that
survive pruning.
This deferral does not affect the locks required for permission
checking in InitPlan(), which takes place before initial pruning.
ExecCheckPermissions() now includes an Assert to verify that all
relations undergoing permission checks, none of which can be in the
set of runtime-prunable relations, are properly locked.
* Plan invalidation handling:
Deferring locks introduces a window where prunable relations may be
altered by concurrent DDL, invalidating the plan. A new function,
ExecutorStartCachedPlan(), wraps ExecutorStart() to detect and handle
invalidation caused by deferred locking. If invalidation occurs,
ExecutorStartCachedPlan() updates CachedPlan using the new
UpdateCachedPlan() function and retries execution with the updated
plan. To ensure all code paths that may be affected by this handle
invalidation properly, all callers of ExecutorStart that may execute a
PlannedStmt from a CachedPlan have been updated to use
ExecutorStartCachedPlan() instead.
UpdateCachedPlan() replaces stale plans in CachedPlan.stmt_list. A new
CachedPlan.stmt_context, created as a child of CachedPlan.context,
allows freeing old PlannedStmts while preserving the CachedPlan
structure and its statement list. This ensures that loops over
statements in upstream callers of ExecutorStartCachedPlan() remain
intact.
ExecutorStart() and ExecutorStart_hook implementations now return a
boolean value indicating whether plan initialization succeeded with a
valid PlanState tree in QueryDesc.planstate, or false otherwise, in
which case QueryDesc.planstate is NULL. Hook implementations are
required to call standard_ExecutorStart() at the beginning, and if it
returns false, they should do the same without proceeding.
* Testing:
To verify these changes, the delay_execution module tests scenarios
where cached plans become invalid due to changes in prunable relations
after deferred locks.
* Note to extension authors:
ExecutorStart_hook implementations must verify plan validity after
calling standard_ExecutorStart(), as explained earlier. For example:
if (prev_ExecutorStart)
plan_valid = prev_ExecutorStart(queryDesc, eflags);
else
plan_valid = standard_ExecutorStart(queryDesc, eflags);
if (!plan_valid)
return false;
<extension-code>
return true;
Extensions accessing child relations, especially prunable partitions,
via ExecGetRangeTableRelation() must now ensure their RT indexes are
present in es_unpruned_relids (introduced in commit cbc127917e), or
they will encounter an error. This is a strict requirement after this
change, as only relations in that set are locked.
The idea of deferring some locks to executor startup, allowing locks
for prunable partitions to be skipped, was first proposed by Tom Lane.
Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions)
Reviewed-by: David Rowley <dgrowleyml@gmail.com> (earlier versions)
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> (earlier versions)
Reviewed-by: Tomas Vondra <tomas@vondra.me>
Reviewed-by: Junwang Zhao <zhjwpku@gmail.com>
Discussion: https://postgr.es/m/CA+HiwqFGkMSge6TgC9KQzde0ohpAycLQuV7ooitEEpbKB0O_mg@mail.gmail.com
Replace the use of Constraint with a new ATAlterConstraint struct, which
allows us to pass additional information. No functionality is added by
this commit. This is necessary for future work that allows altering
constraints in other ways.
I (Álvaro) took the liberty of restructuring the code for ALTER
CONSTRAINT beyond what Amul did. The original coding before Amul's
patch was unnecessarily baroque, and this change makes things simpler
by removing one level of subroutine. Also, partly remove the assumption
that only partitioned tables are relevant (by passing sensible 'recurse'
arguments) and no longer ignore whether ONLY was specified. I say
'partly' because the current coding only walks down via the 'conparentid'
relationship, which is only used for partitioned tables; but future
patches could handle ONLY or not for other types of constraint changes
for legacy inheritance trees too.
Author: Amul Sul <sulamul@gmail.com>
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/CAAJ_b94bfgPV-8Mw_HwSBeheVwaK9=5s+7+KbBj_NpwXQFgDGg@mail.gmail.com
The Self-Join Elimination (SJE) feature removes an inner join of a plain
table to itself in the query tree if it is proven that the join can be
replaced with a scan without impacting the query result. Self-join and
inner relation get replaced with the outer in query, equivalence classes,
and planner info structures. Also, the inner restrictlist moves to the
outer one with the removal of duplicated clauses. Thus, this optimization
reduces the length of the range table list (this especially makes sense for
partitioned relations), reduces the number of restriction clauses and,
in turn, selectivity estimations, and potentially improves total planner
prediction for the query.
This feature is dedicated to avoiding redundancy, which can appear after
pull-up transformations or the creation of an EquivalenceClass-derived clause
like the below.
SELECT * FROM t1 WHERE x IN (SELECT t3.x FROM t1 t3);
SELECT * FROM t1 WHERE EXISTS (SELECT t3.x FROM t1 t3 WHERE t3.x = t1.x);
SELECT * FROM t1,t2, t1 t3 WHERE t1.x = t2.x AND t2.x = t3.x;
In the future, we could also reduce redundancy caused by subquery pull-up
after unnecessary outer join removal in cases like the one below.
SELECT * FROM t1 WHERE x IN
(SELECT t3.x FROM t1 t3 LEFT JOIN t2 ON t2.x = t1.x);
Also, it can drastically help to join partitioned tables, removing entries
even before their expansion.
The SJE proof is based on innerrel_is_unique() machinery.
We can remove a self-join when for each outer row:
1. At most, one inner row matches the join clause;
2. Each matched inner row must be (physically) the same as the outer one;
3. Inner and outer rows have the same row mark.
In this patch, we use the next approach to identify a self-join:
1. Collect all merge-joinable join quals which look like a.x = b.x;
2. Add to the list above the baseretrictinfo of the inner table;
3. Check innerrel_is_unique() for the qual list. If it returns false, skip
this pair of joining tables;
4. Check uniqueness, proved by the baserestrictinfo clauses. To prove the
possibility of self-join elimination, the inner and outer clauses must
match exactly.
The relation replacement procedure is not trivial and is partly combined
with the one used to remove useless left joins. Tests covering this feature
were added to join.sql. Some of the existing regression tests changed due
to self-join removal logic.
Discussion: https://postgr.es/m/flat/64486b0b-0404-e39e-322d-0801154901f3%40postgrespro.ru
Author: Andrey Lepikhov <a.lepikhov@postgrespro.ru>
Author: Alexander Kuzmenkov <a.kuzmenkov@postgrespro.ru>
Co-authored-by: Alexander Korotkov <aekorotkov@gmail.com>
Co-authored-by: Alena Rybakina <lena.ribackina@yandex.ru>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Simon Riggs <simon@2ndquadrant.com>
Reviewed-by: Jonathan S. Katz <jkatz@postgresql.org>
Reviewed-by: David Rowley <david.rowley@2ndquadrant.com>
Reviewed-by: Thomas Munro <thomas.munro@enterprisedb.com>
Reviewed-by: Konstantin Knizhnik <k.knizhnik@postgrespro.ru>
Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-by: Hywel Carver <hywel@skillerwhale.com>
Reviewed-by: Laurenz Albe <laurenz.albe@cybertec.at>
Reviewed-by: Ronan Dunklau <ronan.dunklau@aiven.io>
Reviewed-by: vignesh C <vignesh21@gmail.com>
Reviewed-by: Zhihong Yu <zyu@yugabyte.com>
Reviewed-by: Greg Stark <stark@mit.edu>
Reviewed-by: Jaime Casanova <jcasanov@systemguards.com.ec>
Reviewed-by: Michał Kłeczek <michal@kleczek.org>
Reviewed-by: Alena Rybakina <lena.ribackina@yandex.ru>
Reviewed-by: Alexander Korotkov <aekorotkov@gmail.com>
ExecInitModifyTable() forgot to trim MERGE-related lists to exclude
entries for result relations pruned during initial pruning, so fix
that.
While at it, make the function's use of the pruned resultRelations
list, rather than ModifyTable.resultRelations, more consistent.
Reported-by: Alexander Lakhin <exclusion@gmail.com> (via sqlsmith)
Reviewed-by: Junwang Zhao <zhjwpku@gmail.com>
Discussion: https://postgr.es/m/e72c94d9-e5f9-4753-9bc1-69d72bd54b8a@gmail.com
Commit 1dc5ebc90 allowed PL/pgSQL to perform in-place updates
of expanded-object variables that are being updated with
assignments like "x := f(x, ...)". However this was allowed
only for a hard-wired list of functions f(), since we need to
be sure that f() will not modify the variable if it fails.
It was always envisioned that we should make that extensible,
but at the time we didn't have a good way to do so. Since
then we've invented the idea of "support functions" to allow
attaching specialized optimization knowledge to functions,
and that is a perfect mechanism for doing this.
Hence, adjust PL/pgSQL to use a support function request instead
of hard-wired logic to decide if in-place update is safe.
Preserve the previous optimizations by creating support functions
for the three functions that were previously hard-wired.
Author: Tom Lane <tgl@sss.pgh.pa.us>
Reviewed-by: Andrey Borodin <x4mmm@yandex-team.ru>
Reviewed-by: Pavel Borisov <pashkin.elfe@gmail.com>
Discussion: https://postgr.es/m/CACxu=vJaKFNsYxooSnW1wEgsAO5u_v1XYBacfVJ14wgJV_PYeg@mail.gmail.com
This is similar to d575051b9a but this time for the comments in
plannodes.h to avoid long lines, which is useful if adding per-field
annotations with pg_node_attr() to these planner structures.
Some patches are under discussion to add such properties to planner
fields, which is something that may or may not happen, and this change
makes future proposals easier to work on and review, which being more
consistent in style with the parse nodes.
Author: Sami Imseih
Discussion: https://postgr.es/m/Z5xTb5iBHVGns35R@paquier.xyz
This adds a new variant of generated columns that are computed on read
(like a view, unlike the existing stored generated columns, which are
computed on write, like a materialized view).
The syntax for the column definition is
... GENERATED ALWAYS AS (...) VIRTUAL
and VIRTUAL is also optional. VIRTUAL is the default rather than
STORED to match various other SQL products. (The SQL standard makes
no specification about this, but it also doesn't know about VIRTUAL or
STORED.) (Also, virtual views are the default, rather than
materialized views.)
Virtual generated columns are stored in tuples as null values. (A
very early version of this patch had the ambition to not store them at
all. But so much stuff breaks or gets confused if you have tuples
where a column in the middle is completely missing. This is a
compromise, and it still saves space over being forced to use stored
generated columns. If we ever find a way to improve this, a bit of
pg_upgrade cleverness could allow for upgrades to a newer scheme.)
The capabilities and restrictions of virtual generated columns are
mostly the same as for stored generated columns. In some cases, this
patch keeps virtual generated columns more restricted than they might
technically need to be, to keep the two kinds consistent. Some of
that could maybe be relaxed later after separate careful
considerations.
Some functionality that is currently not supported, but could possibly
be added as incremental features, some easier than others:
- index on or using a virtual column
- hence also no unique constraints on virtual columns
- extended statistics on virtual columns
- foreign-key constraints on virtual columns
- not-null constraints on virtual columns (check constraints are supported)
- ALTER TABLE / DROP EXPRESSION
- virtual column cannot have domain type
- virtual columns are not supported in logical replication
The tests in generated_virtual.sql have been copied over from
generated_stored.sql with the keyword replaced. This way we can make
sure the behavior is mostly aligned, and the differences can be
visible. Some tests for currently not supported features are
currently commented out.
Reviewed-by: Jian He <jian.universality@gmail.com>
Reviewed-by: Dean Rasheed <dean.a.rasheed@gmail.com>
Tested-by: Shlok Kyal <shlok.kyal.oss@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/a368248e-69e4-40be-9c07-6c3b5880b0a6@eisentraut.org
This commit introduces changes to track unpruned relations explicitly,
making it possible for top-level plan nodes, such as ModifyTable and
LockRows, to avoid processing partitions pruned during initial
pruning. Scan-level nodes, such as Append and MergeAppend, already
avoid the unnecessary processing by accessing partition pruning
results directly via part_prune_index. In contrast, top-level nodes
cannot access pruning results directly and need to determine which
partitions remain unpruned.
To address this, this commit introduces a new bitmapset field,
es_unpruned_relids, which the executor uses to track the set of
unpruned relations. This field is referenced during plan
initialization to skip initializing certain nodes for pruned
partitions. It is initialized with PlannedStmt.unprunableRelids,
a new field that the planner populates with RT indexes of relations
that cannot be pruned during runtime pruning. These include relations
not subject to partition pruning and those required for execution
regardless of pruning.
PlannedStmt.unprunableRelids is computed during set_plan_refs() by
removing the RT indexes of runtime-prunable relations, identified
from PartitionPruneInfos, from the full set of relation RT indexes.
ExecDoInitialPruning() then updates es_unpruned_relids by adding
partitions that survive initial pruning.
To support this, PartitionedRelPruneInfo and PartitionedRelPruningData
now include a leafpart_rti_map[] array that maps partition indexes to
their corresponding RT indexes. The former is used in set_plan_refs()
when constructing unprunableRelids, while the latter is used in
ExecDoInitialPruning() to convert partition indexes returned by
get_matching_partitions() into RT indexes, which are then added to
es_unpruned_relids.
These changes make it possible for ModifyTable and LockRows nodes to
process only relations that remain unpruned after initial pruning.
ExecInitModifyTable() trims lists, such as resultRelations,
withCheckOptionLists, returningLists, and updateColnosLists, to
consider only unpruned partitions. It also creates ResultRelInfo
structs only for these partitions. Similarly, child RowMarks for
pruned relations are skipped.
By avoiding unnecessary initialization of structures for pruned
partitions, these changes improve the performance of updates and
deletes on partitioned tables during initial runtime pruning.
Due to ExecInitModifyTable() changes as described above, EXPLAIN on a
plan for UPDATE and DELETE that uses runtime initial pruning no longer
lists partitions pruned during initial pruning.
Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions)
Reviewed-by: Tomas Vondra <tomas@vondra.me>
Discussion: https://postgr.es/m/CA+HiwqFGkMSge6TgC9KQzde0ohpAycLQuV7ooitEEpbKB0O_mg@mail.gmail.com
Consistently use "Size" (or size_t, or in some places int64 or double)
as the type for variables holding memory allocation sizes. In most
places variables' data types were fine already, but we had an ancient
habit of computing bytes from kilobytes-units GUCs with code like
"work_mem * 1024L". That risks overflow on Win64 where they did not
make "long" as wide as "size_t". We worked around that by restricting
such GUCs' ranges, so you couldn't set work_mem et al higher than 2GB
on Win64. This patch removes that restriction, after replacing such
calculations with "work_mem * (Size) 1024" or variants of that.
It should be noted that this patch was constructed by searching
outwards from the GUCs that have MAX_KILOBYTES as upper limit.
So I can't positively guarantee there are no other places doing
memory-size arithmetic in int or long variables. I do however feel
pretty confident that increasing MAX_KILOBYTES on Win64 is safe now.
Also, nothing in our code should be dealing in multiple-gigabyte
allocations without authorization from a relevant GUC, so it seems
pretty likely that this search caught everything that could be at
risk of overflow.
Author: Vladlen Popolitov <v.popolitov@postgrespro.ru>
Co-authored-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/1a01f0-66ec2d80-3b-68487680@27595217
This commit builds on the prior change that moved PartitionPruneInfos
out of individual plan nodes into a list in PlannedStmt, making it
possible to initialize PartitionPruneStates without traversing the
plan tree and perform runtime initial pruning before ExecInitNode()
initializes the plan trees. These tasks are now handled in a new
routine, ExecDoInitialPruning(), which is called by InitPlan()
before calling ExecInitNode() on various plan trees.
ExecDoInitialPruning() performs the initial pruning and saves the
result -- a Bitmapset of indexes for surviving child subnodes -- in
es_part_prune_results, a list in EState.
PartitionPruneStates created for initial pruning are stored in
es_part_prune_states, another list in EState, for later use during
exec pruning. Both lists are parallel to es_part_prune_infos, which
holds the PartitionPruneInfos from PlannedStmt, enabling shared
indexing.
PartitionPruneStates initialized in ExecDoInitialPruning() now
include only the PartitionPruneContexts for initial pruning steps.
Exec pruning contexts are initialized later in
ExecInitPartitionExecPruning() when the parent plan node is
initialized, as the exec pruning step expressions depend on the parent
node's PlanState.
The existing function PartitionPruneFixSubPlanMap() has been
repurposed for this initialization to avoid duplicating a similar
loop structure for finding PartitionedRelPruningData to initialize
exec pruning contexts for. It has been renamed to
InitExecPruningContexts() to reflect its new primary responsibility.
The original logic to "fix subplan maps" remains intact but is now
encapsulated within the renamed function.
This commit removes two obsolete Asserts in partkey_datum_from_expr().
The ExprContext used for pruning expression evaluation is now
independent of the parent PlanState, making these Asserts unnecessary.
By centralizing pruning logic and decoupling it from the plan
initialization step (ExecInitNode()), this change sets the stage for
future patches that will use the result of initial pruning to
save the overhead of redundant processing for pruned partitions.
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Reviewed-by: Tomas Vondra <tomas@vondra.me>
Discussion: https://postgr.es/m/CA+HiwqFGkMSge6TgC9KQzde0ohpAycLQuV7ooitEEpbKB0O_mg@mail.gmail.com
This moves PartitionPruneInfo from plan nodes to PlannedStmt,
simplifying traversal by centralizing all PartitionPruneInfo
structures in a single list in it, which holds all instances for the
main query and its subqueries. Instead of plan nodes (Append or
MergeAppend) storing PartitionPruneInfo pointers, they now reference
an index in this list.
A bitmapset field is added to PartitionPruneInfo to store the RT
indexes corresponding to the apprelids field in Append or MergeAppend.
This allows execution pruning logic to verify that it operates on the
correct plan node, mainly to facilitate debugging.
Duplicated code in set_append_references() and
set_mergeappend_references() is refactored into a new function,
register_pruneinfo(). This updates RT indexes by applying rtoffet
and adds PartitionPruneInfo to the global list in PlannerGlobal.
By allowing pruning to be performed without traversing the plan tree,
this change lays the groundwork for runtime initial pruning to occur
independently of plan tree initialization.
Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org> (earlier version)
Reviewed-by: Robert Haas <robertmhaas@gmail.com>
Reviewed-by: Tomas Vondra <tomas@vondra.me>
Discussion: https://postgr.es/m/CA+HiwqFGkMSge6TgC9KQzde0ohpAycLQuV7ooitEEpbKB0O_mg@mail.gmail.com
This allows the RETURNING list of INSERT/UPDATE/DELETE/MERGE queries
to explicitly return old and new values by using the special aliases
"old" and "new", which are automatically added to the query (if not
already defined) while parsing its RETURNING list, allowing things
like:
RETURNING old.colname, new.colname, ...
RETURNING old.*, new.*
Additionally, a new syntax is supported, allowing the names "old" and
"new" to be changed to user-supplied alias names, e.g.:
RETURNING WITH (OLD AS o, NEW AS n) o.colname, n.colname, ...
This is useful when the names "old" and "new" are already defined,
such as inside trigger functions, allowing backwards compatibility to
be maintained -- the interpretation of any existing queries that
happen to already refer to relations called "old" or "new", or use
those as aliases for other relations, is not changed.
For an INSERT, old values will generally be NULL, and for a DELETE,
new values will generally be NULL, but that may change for an INSERT
with an ON CONFLICT ... DO UPDATE clause, or if a query rewrite rule
changes the command type. Therefore, we put no restrictions on the use
of old and new in any DML queries.
Dean Rasheed, reviewed by Jian He and Jeff Davis.
Discussion: https://postgr.es/m/CAEZATCWx0J0-v=Qjc6gXzR=KtsdvAE7Ow=D=mu50AgOe+pvisQ@mail.gmail.com
This changes commit 7406ab623f in that the gist strategy number
mapping support function is changed to use the CompareType enum as
input, instead of the "well-known" RT*StrategyNumber strategy numbers.
This is a bit cleaner, since you are not dealing with two sets of
strategy numbers. Also, this will enable us to subsume this system
into a more general system of using CompareType to define operator
semantics across index methods.
Discussion: https://www.postgresql.org/message-id/flat/E72EAA49-354D-4C2E-8EB9-255197F55330@enterprisedb.com
RowCompareType served as a way to describe the fundamental meaning of
an operator, notionally independent of an operator class (although so
far this was only really supported for btrees). Its original purpose
was for use inside RowCompareExpr, and it has also found some small
use outside, such as for get_op_btree_interpretation().
We want to expand this now, as a more general way to describe operator
semantics for other index access methods, including gist (to improve
GistTranslateStratnum()) and others not written yet. To avoid future
confusion, we rename the type to CompareType and the symbols from
ROWCOMPARE_XXX to COMPARE_XXX to reflect their more general purpose.
Reviewed-by: Mark Dilger <mark.dilger@enterprisedb.com>
Discussion: https://www.postgresql.org/message-id/flat/E72EAA49-354D-4C2E-8EB9-255197F55330@enterprisedb.com
This adds support for the NOT ENFORCED/ENFORCED flag for constraints,
with support for check constraints.
The plan is to eventually support this for foreign key constraints,
where it is typically more useful.
Note that CHECK constraints do not currently support ALTER operations,
so changing the enforceability of an existing constraint isn't
possible without dropping and recreating it. This could be added
later.
Author: Amul Sul <amul.sul@enterprisedb.com>
Reviewed-by: Peter Eisentraut <peter@eisentraut.org>
Reviewed-by: jian he <jian.universality@gmail.com>
Tested-by: Triveni N <triveni.n@enterprisedb.com>
Discussion: https://www.postgresql.org/message-id/flat/CAAJ_b962c5AcYW9KUt_R_ER5qs3fUGbe4az-SP-vuwPS-w-AGA@mail.gmail.com
Previously, the caller needed to allocate the memory and the
TupleHashTable would store a pointer to it. That wastes space for the
palloc overhead as well as the size of the pointer itself.
Now, the TupleHashTable relies on the caller to correctly specify the
additionalsize, and allocates that amount of space. The caller can
then request a pointer into that space.
Discussion: https://postgr.es/m/b9cbf0219a9859dc8d240311643ff4362fd9602c.camel@j-davis.com
Reviewed-by: Heikki Linnakangas
The original design for set operations involved appending the two
input relations into one and adding a flag column that allows
distinguishing which side each row came from. Then the SetOp node
pries them apart again based on the flag. This is bizarre. The
only apparent reason to do it is that when sorting, we'd only need
one Sort node not two. But since sorting is at least O(N log N),
sorting all the data is actually worse than sorting each side
separately --- plus, we have no chance of taking advantage of
presorted input. On top of that, adding the flag column frequently
requires an additional projection step that adds cycles, and then
the Append node isn't free either. Let's get rid of all of that
and make the SetOp node have two separate children, using the
existing outerPlan/innerPlan infrastructure.
This initial patch re-implements nodeSetop.c and does a bare minimum
of work on the planner side to generate correctly-shaped plans.
In particular, I've tried not to change the cost estimates here,
so that the visible changes in the regression test results will only
involve removal of useless projection steps and not any changes in
whether to use sorted vs hashed mode.
For SORTED mode, we combine successive identical tuples from each
input into groups, and then merge-join the groups. The tuple
comparisons now use SortSupport instead of simple equality, but
the group-formation part should involve roughly the same number of
tuple comparisons as before. The cross-comparisons between left and
right groups probably add to that, but I'm not sure to quantify how
many more comparisons we might need.
For HASHED mode, nodeSetop's logic is almost the same as before,
just refactored into two separate loops instead of one loop that
has an assumption that it will see all the left-hand inputs first.
In both modes, I added early-exit logic to not bother reading the
right-hand relation if the left-hand input is empty, since neither
INTERSECT nor EXCEPT modes can produce any output if the left input
is empty. This could have been done before in the hashed mode, but
not in sorted mode. Sorted mode can also stop as soon as it exhausts
the left input; any remaining right-hand tuples cannot have matches.
Also, this patch adds some infrastructure for detecting whether
child plan nodes all output the same type of tuple table slot.
If they do, the hash table logic can use slightly more efficient
code based on assuming that that's the input slot type it will see.
We'll make use of that infrastructure in other plan node types later.
Patch by me; thanks to Richard Guo and David Rowley for review.
Discussion: https://postgr.es/m/1850138.1731549611@sss.pgh.pa.us
With the repurposing of TBMIterator as an interface for both parallel
and serial iteration through TIDBitmaps in commit 7f9d4187e7,
bitmap table scans may now use it.
Modify bitmap table scan code to use the TBMIterator. This requires
moving around a bit of code, so a few variables are initialized
elsewhere.
Author: Melanie Plageman
Reviewed-by: Tomas Vondra
Discussion: https://postgr.es/m/c736f6aa-8b35-4e20-9621-62c7c82e2168%40vondra.me
Add and use TBMPrivateIterator, which replaces the current TBMIterator
for serial use cases, and repurpose TBMIterator to be a unified
interface for both the serial ("private") and parallel ("shared") TID
Bitmap iterator interfaces. This encapsulation simplifies call sites for
callers supporting both parallel and serial TID Bitmap access.
TBMIterator is not yet used in this commit.
Author: Melanie Plageman
Reviewed-by: Tomas Vondra, Heikki Linnakangas
Discussion: https://postgr.es/m/063e4eb4-32d9-439e-a0b1-75565a9835a8%40iki.fi
d4c3a156c added support that when the GROUP BY contained all of the
columns belonging to a relation's PRIMARY KEY, all other columns
belonging to that relation would be removed from the GROUP BY clause.
That's possible because all other columns are functionally dependent on
the PRIMARY KEY and those columns alone ensure the groups are distinct.
Here we expand on that optimization and allow it to work for any unique
indexes on the table rather than just the PRIMARY KEY index. This
normally requires that all columns in the index are defined with NOT NULL,
however, we can relax that requirement when the index is defined with
NULLS NOT DISTINCT.
When there are multiple suitable indexes to allow columns to be removed,
we prefer the index with the least number of columns as this allows us
to remove the highest number of GROUP BY columns. One day, we may want to
revisit that decision as it may make more sense to use the narrower set of
columns in terms of the width of the data types and stored/queried data.
This also adjusts the code to make use of RelOptInfo.indexlist rather
than looking up the catalog tables.
In passing, add another short-circuit path to allow bailing out earlier
in cases where it's certainly not possible to remove redundant GROUP BY
columns. This early exit is now cheaper to do than when this code was
originally written as 00b41463c made it cheaper to check for empty
Bitmapsets.
Patch originally by Zhang Mingli and later worked on by jian he, but after
I (David) worked on it, there was very little of the original left.
Author: Zhang Mingli, jian he, David Rowley
Reviewed-by: jian he, Andrei Lepikhov
Discussion: https://postgr.es/m/327990c8-b9b2-4b0c-bffb-462249f82de0%40Spark
This speeds up obtaining hash values for GROUP BY and hashed SubPlans by
using the ExprState support for hashing, thus allowing JIT compilation for
obtaining hash values for these operations.
This, even without JIT compilation, has been shown to improve Hash
Aggregate performance in some cases by around 15% and hashed NOT IN
queries in one case by over 30%, however, real-world cases are likely to
see smaller gains as the test cases used were purposefully designed to
have high hashing overheads by keeping the hash table small to prevent
additional memory overheads that would be a factor when working with large
hash tables.
In passing, fix a hypothetical bug in ExecBuildHash32Expr() so that the
initial value is stored directly in the ExprState's result field if
there are no expressions to hash. None of the current users of this
function use an initial value, so the bug is only hypothetical.
Reviewed-by: Andrei Lepikhov <lepihov@gmail.com>
Discussion: https://postgr.es/m/CAApHDvpYSO3kc9UryMevWqthTBrxgfd9djiAjKHMPUSQeX9vdQ@mail.gmail.com
We now create contype='n' pg_constraint rows for not-null constraints on
user tables. Only one such constraint is allowed for a column.
We propagate these constraints to other tables during operations such as
adding inheritance relationships, creating and attaching partitions and
creating tables LIKE other tables. These related constraints mostly
follow the well-known rules of conislocal and coninhcount that we have
for CHECK constraints, with some adaptations: for example, as opposed to
CHECK constraints, we don't match not-null ones by name when descending
a hierarchy to alter or remove it, instead matching by the name of the
column that they apply to. This means we don't require the constraint
names to be identical across a hierarchy.
The inheritance status of these constraints can be controlled: now we
can be sure that if a parent table has one, then all children will have
it as well. They can optionally be marked NO INHERIT, and then children
are free not to have one. (There's currently no support for altering a
NO INHERIT constraint into inheriting down the hierarchy, but that's a
desirable future feature.)
This also opens the door for having these constraints be marked NOT
VALID, as well as allowing UNIQUE+NOT NULL to be used for functional
dependency determination, as envisioned by commit e49ae8d3bc. It's
likely possible to allow DEFERRABLE constraints as followup work, as
well.
psql shows these constraints in \d+, though we may want to reconsider if
this turns out to be too noisy. Earlier versions of this patch hid
constraints that were on the same columns of the primary key, but I'm
not sure that that's very useful. If clutter is a problem, we might be
better off inventing a new \d++ command and not showing the constraints
in \d+.
For now, we omit these constraints on system catalog columns, because
they're unlikely to achieve anything.
The main difference to the previous attempt at this (b0e96f3119) is
that we now require that such a constraint always exists when a primary
key is in the column; we didn't require this previously which had a
number of unpalatable consequences. With this requirement, the code is
easier to reason about. For example:
- We no longer have "throwaway constraints" during pg_dump. We needed
those for the case where a table had a PK without a not-null
underneath, to prevent a slow scan of the data during restore of the
PK creation, which was particularly problematic for pg_upgrade.
- We no longer have to cope with attnotnull being set spuriously in
case a primary key is dropped indirectly (e.g., via DROP COLUMN).
Some bits of code in this patch were authored by Jian He.
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Author: Bernd Helmle <mailings@oopsware.de>
Reviewed-by: 何建 (jian he) <jian.universality@gmail.com>
Reviewed-by: 王刚 (Tender Wang) <tndrwang@gmail.com>
Reviewed-by: Justin Pryzby <pryzby@telsasoft.com>
Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
Reviewed-by: Dean Rasheed <dean.a.rasheed@gmail.com>
Discussion: https://postgr.es/m/202408310358.sdhumtyuy2ht@alvherre.pgsql
This allows an error cursor to be supplied for a bunch of
bad-function-definition errors that previously lacked one,
or that cheated a bit by pointing at the contained type name
when the error isn't really about that.
Bump catversion from an abundance of caution --- I don't think
this node type can actually appear in stored views/rules, but
better safe than sorry.
Jian He and Tom Lane (extracted from a larger patch by Jian,
with some additional work by me)
Discussion: https://postgr.es/m/CACJufxEmONE3P2En=jopZy1m=cCCUs65M4+1o52MW5og9oaUPA@mail.gmail.com
Move all responsibility for indicating a block is exhuasted into
table_scan_bitmap_next_tuple() and advance the main iterator in
heap-specific code. This flow control makes more sense and is a step
toward using the read stream API for bitmap heap scans.
Previously, table_scan_bitmap_next_block() returned false to indicate
table_scan_bitmap_next_tuple() should not be called for the tuples on
the page. This happened both when 1) there were no visible tuples on the
page and 2) when the block returned by the iterator was past the end of
the table. BitmapHeapNext() (generic bitmap table scan code) handled the
case when the bitmap was exhausted.
It makes more sense for table_scan_bitmap_next_tuple() to return false
when there are no visible tuples on the page and
table_scan_bitmap_next_block() to return false when the bitmap is
exhausted or there are no more blocks in the table.
As part of this new design, TBMIterateResults are no longer used as a
flow control mechanism in BitmapHeapNext(), so we removed
table_scan_bitmap_next_tuple's TBMIterateResult parameter.
Note that the prefetch iterator is still saved in the
BitmapHeapScanState node and advanced in generic bitmap table scan code.
This is because 1) it was not necessary to change the prefetch iterator
location to change the flow control in BitmapHeapNext() 2) modifying
prefetch iterator management requires several more steps better split
over multiple commits and 3) the prefetch iterator will be removed once
the read stream API is used.
Author: Melanie Plageman
Reviewed-by: Tomas Vondra, Andres Freund, Heikki Linnakangas, Mark Dilger
Discussion: https://postgr.es/m/063e4eb4-32d9-439e-a0b1-75565a9835a8%40iki.fi
Previously, a Query generated through the transform phase would have
unset stmt_location, tracking the starting point of a query string.
Extensions relying on the statement location to extract its relevant
parts in the source text string would fallback to use the whole
statement instead, leading to confusing results like in
pg_stat_statements for queries relying on nested queries, like:
- EXPLAIN, with top-level and nested query using the same query string,
and a query ID coming from the nested query when the non-top-level
entry.
- Multi-statements, with only partial portions of queries being
normalized.
- COPY TO with a query, SELECT or DMLs.
This patch improves things by keeping track of the statement locations
and propagate it to Query during transform, allowing PGSS to only show
the relevant part of the query for nested query. This leads to less
bloat in entries for non-top-level entries, as queries can now be
grouped within the same (toplevel, queryid) duos in pg_stat_statements.
The result gives a stricter one-one mapping between query IDs and its
query strings.
The regression tests introduced in 45e0ba30fc produce differences
reflecting the new logic.
Author: Anthonin Bonnefoy
Reviewed-by: Michael Paquier, Jian He
Discussion: https://postgr.es/m/CAO6_XqqM6S9bQ2qd=75W+yKATwoazxSNhv5sjW06fjGAtHbTUA@mail.gmail.com
The approach of declaring a function pointer with an empty argument
list and hoping that the compiler will not complain about casting it
to another type no longer works with C23, because foo() is now
equivalent to foo(void).
We don't need to do this here. With a few struct forward declarations
we can supply a correct argument list without having to pull in
another header file.
(This is the only new warning with C23. Together with the previous
fix a67a49648d, this makes the whole code compile cleanly under C23.)
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://www.postgresql.org/message-id/flat/95c6a9bf-d306-43d8-b880-664ef08f2944%40eisentraut.org
The decision in b6e1157e7 to ignore raw_expr when evaluating a
JsonValueExpr was incorrect. While its value is not ultimately
used (since formatted_expr's value is), failing to initialize it
can lead to problems, for instance, when the expression tree in
raw_expr contains Aggref nodes, which must be initialized to
ensure the parent Agg node works correctly.
Also, optimize eval_const_expressions_mutator()'s handling of
JsonValueExpr a bit. Currently, when formatted_expr cannot be folded
into a constant, we end up processing it twice -- once directly in
eval_const_expressions_mutator() and again recursively via
ece_generic_processing(). This recursive processing is required to
handle raw_expr. To avoid the redundant processing of formatted_expr,
we now process raw_expr directly in eval_const_expressions_mutator().
Finally, update the comment of JsonValueExpr to describe the roles of
raw_expr and formatted_expr more clearly.
Bug: #18657
Reported-by: Alexander Lakhin <exclusion@gmail.com>
Diagnosed-by: Fabio R. Sluzala <fabio3rs@gmail.com>
Diagnosed-by: Tender Wang <tndrwang@gmail.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/18657-1b90ccce2b16bdb8@postgresql.org
Backpatch-through: 16
Instead of using Node *, we can use an incomplete struct. That way,
everything has the correct type and fewer casts are required. This
technique is already used elsewhere in node type definitions.
Reviewed-by: Nathan Bossart <nathandbossart@gmail.com>
Reviewed-by: Tender Wang <tndrwang@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/637eeea8-5663-460b-a114-39572c0f6c6e%40eisentraut.org
The MergeJoin struct was tracking "mergeStrategies", which were an
array of btree strategy numbers, purely for the purpose of comparing
it later against btree strategies to determine if the scan direction
was forward or reverse. Change that. Instead, track
"mergeReversals", an array of bool, to indicate the same without an
unfortunate assumption that a strategy number refers specifically to a
btree strategy.
Author: Mark Dilger <mark.dilger@enterprisedb.com>
Discussion: https://www.postgresql.org/message-id/flat/E72EAA49-354D-4C2E-8EB9-255197F55330@enterprisedb.com
Functions make_pathkey_from_sortop() and transformWindowDefinitions(),
which receive a SortGroupClause, were determining the sort order
(ascending vs. descending) by comparing that structure's operator
strategy to BTLessStrategyNumber, but could just as easily have gotten
it from the SortGroupClause object, if it had such a field, so add
one. This reduces the number of places that hardcode the assumption
that the strategy refers specifically to a btree strategy, rather than
some other index AM's operators.
Author: Mark Dilger <mark.dilger@enterprisedb.com>
Discussion: https://www.postgresql.org/message-id/flat/E72EAA49-354D-4C2E-8EB9-255197F55330@enterprisedb.com
Commit 90189eefc1 narrowed pg_attribute.attinhcount and
pg_constraint.coninhcount from 32 to 16 bits, but kept other related
structs with 32-bit wide fields: ColumnDef and CookedConstraint contain
an int 'inhcount' field which is itself checked for overflow on
increments, but there's no check that the values aren't above INT16_MAX
before assigning to the catalog columns. This means that a creative
user can get a inconsistent table definition and override some
protections.
Fix it by changing those other structs to also use int16.
Also, modernize style by using pg_add_s16_overflow for overflow testing
instead of checking for negative values.
We also have Constraint.inhcount, which is here removed completely.
This was added by commit b0e96f3119 and not removed by its revert at
6f8bb7c1e9. It is not needed by the upcoming not-null constraints
patch.
This is mostly academic, so we agreed not to backpatch to avoid ABI
problems.
Bump catversion because of the changes to parse nodes.
Co-authored-by: Álvaro Herrera <alvherre@alvh.no-ip.org>
Co-authored-by: 何建 (jian he) <jian.universality@gmail.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/202410081611.up4iyofb5ie7@alvherre.pgsql
These fields can be set by executor nodes to record how many parallel
workers were planned to be launched and how many of them have been
actually launched within the number initially planned. This data is
able to give an approximation of the parallel worker draught a system
is facing, making easier the tuning of related configuration parameters.
These fields will be used by some follow-up patches to populate other
parts of the system with their data.
Author: Guillaume Lelarge, Benoit Lobréau
Discussion: https://postgr.es/m/783bc7f7-659a-42fa-99dd-ee0565644e25@dalibo.com
Discussion: https://postgr.es/m/CAECtzeWtTGOK0UgKXdDGpfTVSa5bd_VbUt6K6xn8P7X+_dZqKw@mail.gmail.com
This is a continuation of work like 11c34b342b, done to reduce the
bloat of pg_stat_statements by applying more normalization to query
entries. This commit is able to detect and normalize values in
VariableSetStmt, resulting in:
SET conf_param = $1
Compared to other parse nodes, VariableSetStmt is embedded in much more
places in the parser, impacting many query patterns in
pg_stat_statements. A custom jumble function is used, with an extra
field in the node to decide if arguments should be included in the
jumbling or not, a location field being not enough for this purpose.
This approach allows for a finer tuning.
Clauses relying on one or more keywords are not normalized, for example:
* DEFAULT
* FROM CURRENT
* List of keywords. SET SESSION CHARACTERISTICS AS TRANSACTION,
where it is critical to differentiate different sets of options, is a
good example of why normalization should not happen.
Some queries use VariableSetStmt for some subclauses with SET, that also
have their values normalized:
- ALTER DATABASE
- ALTER ROLE
- ALTER SYSTEM
- CREATE/ALTER FUNCTION
ba90eac7a9 has added test coverage for most of the existing SET
patterns. The expected output of these tests shows the difference this
commit creates. Normalization could be perhaps applied to more portions
of the grammar but what is done here is conservative, and good enough as
a starting point.
Author: Greg Sabino Mullane, Michael Paquier
Discussion: https://postgr.es/m/36e5bffe-e989-194f-85c8-06e7bc88e6f7@amazon.com
Discussion: https://postgr.es/m/B44FA29D-EBD0-4DD9-ABC2-16F1CB087074@amazon.com
Discussion: https://postgr.es/m/CAKAnmmJtJY2jzQN91=2QAD2eAJAA-Per61eyO48-TyxEg-q0Rg@mail.gmail.com
The previous commit fixed some ways of losing an inplace update. It
remained possible to lose one when a backend working toward a
heap_update() copied a tuple into memory just before inplace update of
that tuple. In catalogs eligible for inplace update, use LOCKTAG_TUPLE
to govern admission to the steps of copying an old tuple, modifying it,
and issuing heap_update(). This includes MERGE commands. To avoid
changing most of the pg_class DDL, don't require LOCKTAG_TUPLE when
holding a relation lock sufficient to exclude inplace updaters.
Back-patch to v12 (all supported versions). In v13 and v12, "UPDATE
pg_class" or "UPDATE pg_database" can still lose an inplace update. The
v14+ UPDATE fix needs commit 86dc90056d,
and it wasn't worth reimplementing that fix without such infrastructure.
Reviewed by Nitin Motiani and (in earlier versions) Heikki Linnakangas.
Discussion: https://postgr.es/m/20231027214946.79.nmisch@google.com
Add PERIOD clause to foreign key constraint definitions. This is
supported for range and multirange types. Temporal foreign keys check
for range containment instead of equality.
This feature matches the behavior of the SQL standard temporal foreign
keys, but it works on PostgreSQL's native ranges instead of SQL's
"periods", which don't exist in PostgreSQL (yet).
Reference actions ON {UPDATE,DELETE} {CASCADE,SET NULL,SET DEFAULT}
are not supported yet.
(previously committed as 34768ee361, reverted by 8aee330af55; this is
essentially unchanged from those)
Author: Paul A. Jungwirth <pj@illuminatedcomputing.com>
Reviewed-by: Peter Eisentraut <peter@eisentraut.org>
Reviewed-by: jian he <jian.universality@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/CA+renyUApHgSZF9-nd-a0+OPGharLQLO=mDHcY4_qQ0+noCUVg@mail.gmail.com
Add WITHOUT OVERLAPS clause to PRIMARY KEY and UNIQUE constraints.
These are backed by GiST indexes instead of B-tree indexes, since they
are essentially exclusion constraints with = for the scalar parts of
the key and && for the temporal part.
(previously committed as 46a0cd4cef, reverted by 46a0cd4cefb; the new
part is this:)
Because 'empty' && 'empty' is false, the temporal PK/UQ constraint
allowed duplicates, which is confusing to users and breaks internal
expectations. For instance, when GROUP BY checks functional
dependencies on the PK, it allows selecting other columns from the
table, but in the presence of duplicate keys you could get the value
from any of their rows. So we need to forbid empties.
This all means that at the moment we can only support ranges and
multiranges for temporal PK/UQs, unlike the original patch (above).
Documentation and tests for this are added. But this could
conceivably be extended by introducing some more general support for
the notion of "empty" for other types.
Author: Paul A. Jungwirth <pj@illuminatedcomputing.com>
Reviewed-by: Peter Eisentraut <peter@eisentraut.org>
Reviewed-by: jian he <jian.universality@gmail.com>
Discussion: https://www.postgresql.org/message-id/flat/CA+renyUApHgSZF9-nd-a0+OPGharLQLO=mDHcY4_qQ0+noCUVg@mail.gmail.com
If there are subqueries in the grouping expressions, each of these
subqueries in the targetlist and HAVING clause is expanded into
distinct SubPlan nodes. As a result, only one of these SubPlan nodes
would be converted to reference to the grouping key column output by
the Agg node; others would have to get evaluated afresh. This is not
efficient, and with grouping sets this can cause wrong results issues
in cases where they should go to NULL because they are from the wrong
grouping set. Furthermore, during re-evaluation, these SubPlan nodes
might use nulled column values from grouping sets, which is not
correct.
This issue is not limited to subqueries. For other types of
expressions that are part of grouping items, if they are transformed
into another form during preprocessing, they may fail to match lower
target items. This can also lead to wrong results with grouping sets.
To fix this issue, we introduce a new kind of RTE representing the
output of the grouping step, with columns that are the Vars or
expressions being grouped on. In the parser, we replace the grouping
expressions in the targetlist and HAVING clause with Vars referencing
this new RTE, so that the output of the parser directly expresses the
semantic requirement that the grouping expressions be gotten from the
grouping output rather than computed some other way. In the planner,
we first preprocess all the columns of this new RTE and then replace
any Vars in the targetlist and HAVING clause that reference this new
RTE with the underlying grouping expressions, so that we will have
only one instance of a SubPlan node for each subquery contained in the
grouping expressions.
Bump catversion because this changes the querytree produced by the
parser.
Thanks to Tom Lane for the idea to invent a new kind of RTE.
Per reports from Geoff Winkless, Tobias Wendorff, Richard Guo from
various threads.
Author: Richard Guo
Reviewed-by: Ashutosh Bapat, Sutou Kouhei
Discussion: https://postgr.es/m/CAMbWs4_dp7e7oTwaiZeBX8+P1rXw4ThkZxh1QG81rhu9Z47VsQ@mail.gmail.com
When WindowAgg finished one partition of a PARTITION BY, it previously
would call tuplestore_end() to purge all the stored tuples before again
calling tuplestore_begin_heap() and carefully setting up all of the
tuplestore read pointers exactly as required for the given frameOptions.
Since the frameOptions don't change between partitions, this part does
not make much sense. For queries that had very few rows per partition,
the overhead of this was very large.
It seems much better to create the tuplestore and the read pointers once
and simply call tuplestore_clear() at the end of each partition.
tuplestore_clear() moves all of the read pointers back to the start
position and deletes all the previously stored tuples.
A simple test query with 1 million partitions and 1 tuple per partition
has been shown to run around 40% faster than without this change. The
additional effort seems to have mostly been spent in malloc/free.
Making this work required adding a new bool field to WindowAggState
which had the unfortunate effect of being the 9th bool field in a group
resulting in the struct being enlarged. Here we shuffle the fields
around a little so that the two bool fields for runcondition relating
stuff fit into existing padding. Also, move the "runcondition" field to
be near those. This frees up enough space with the other bool fields so
that the newly added one fits into the padding bytes. This was done to
address a very small but apparent performance regression with queries
containing a large number of rows per partition.
Reviewed-by: Ashutosh Bapat <ashutosh.bapat.oss@gmail.com>
Reviewed-by: Tatsuo Ishii <ishii@postgresql.org>
Discussion: https://postgr.es/m/CAHoyFK9n-QCXKTUWT_xxtXninSMEv%2BgbJN66-y6prM3f4WkEHw%40mail.gmail.com
