public inbox for [email protected]
help / color / mirror / Atom feedFrom: Amit Langote <[email protected]>
To: Tom Lane <[email protected]>
Cc: Tender Wang <[email protected]>
Cc: Alexander Lakhin <[email protected]>
Cc: Tomas Vondra <[email protected]>
Cc: Robert Haas <[email protected]>
Cc: Alvaro Herrera <[email protected]>
Cc: Andres Freund <[email protected]>
Cc: Daniel Gustafsson <[email protected]>
Cc: David Rowley <[email protected]>
Cc: PostgreSQL Hackers <[email protected]>
Cc: Thom Brown <[email protected]>
Subject: Re: generic plans and "initial" pruning
Date: Mon, 17 Nov 2025 21:50:01 +0900
Message-ID: <CA+HiwqFpEHBjosRackQhm6yKKnHgqm8Ewkn=qsctT1N0PqVSrg@mail.gmail.com> (raw)
In-Reply-To: <CA+HiwqEF9SgKyQ1HrYOURpv8DGRGHDNwBT9Y6yEBVCW+=kh_=w@mail.gmail.com>
References: <CA+HiwqFpZ80UJKr4tZus4Omgg7YESzFXKSwSHRW2Ap2=XSVyUA@mail.gmail.com>
<[email protected]>
<CA+HiwqH8N-SxEB6SysEBsYNgV_KJs66k9Z2SNmqVzbBP-60yWg@mail.gmail.com>
<[email protected]>
<CA+HiwqEmG9YCQvG6uux7sO=jKFSAW6hA4Ea-ymfD+JhJAe4PWQ@mail.gmail.com>
<CA+HiwqE2FfJfH=siLiR3kJ13tmXZORAGTWsZc2r52o1_5BDv+g@mail.gmail.com>
<[email protected]>
<CA+HiwqFhkpXHAA=4NY5SqYXX08uq=nYtXcSByNZF=2MAy1UA7A@mail.gmail.com>
<CA+HiwqHCcSoYfpMjFshaU1bj6NjreiDvMSDpVSeBmqk-kbWrPw@mail.gmail.com>
<CA+HiwqHOejJk0_qMuM5g38h70hY_JvHMAKwnH3k=urfTXauPQA@mail.gmail.com>
<CA+HiwqFsGKM82oaMby3VWYXf_XFpDAMeT+6SXgj-45HpTrS1dA@mail.gmail.com>
<CA+HiwqFA5hUWYktt3VMh4zQOYMxqH-MpdX8eemfM+o-9dY-zbQ@mail.gmail.com>
<CA+HiwqEn7bbUXaXO=SmUujBjJSHfS31cwQroHRBwT0sR=66bgg@mail.gmail.com>
<CA+HiwqGGLDTd1ZTK1c0zv4La7XOVSVMqBuNtscJeh6FyUQvFvA@mail.gmail.com>
<CA+HiwqE2JFiqqrXdyJVQWY-fMGwzDkLqjXQdUKbPaCpDpxd_2g@mail.gmail.com>
<CA+HiwqFp3jZGSz==QjeuV62_62F6+V6b62=Uqvy99sW_gsgWBA@mail.gmail.com>
<CAHewXNkUz9XGG8nnoxZaw35e+5bQVVP=eeJE4cW4V2e+P9ndFw@mail.gmail.com>
<CA+HiwqFKSpfYruzcVz-5CcFxg7gMa+ycXjMa2aPz_J_P4LGXTg@mail.gmail.com>
<[email protected]>
<CA+HiwqEQ1oME-hcDXwC9rGQb=u7MdUFG3Sc=Qg27uH480v10FA@mail.gmail.com>
<[email protected]>
<CA+HiwqGXMLSQyJvynWF40yNwBAx-pXtxemReP8L+C+kaUa5v5A@mail.gmail.com>
<CA+HiwqGBfMgcxokEH_mg6s=ttLFm54dj4hT6yXydU2t0g6oQ3g@mail.gmail.com>
<CA+HiwqEEkGfMc_LSJhfz96o-czVS4B59Vhw6i1_t58ZGqhP8VA@mail.gmail.com>
<CA+HiwqHAd+9nptjxP6=KrcKA1BMsS6pbB3B2oaojwdyH_wBWCA@mail.gmail.com>
<CA+HiwqE7_YpU--EsrhvNqcZ+10+92EGFaX5609AUJb9ENLntnQ@mail.gmail.com>
<CA+HiwqEF9SgKyQ1HrYOURpv8DGRGHDNwBT9Y6yEBVCW+=kh_=w@mail.gmail.com>
On Wed, Nov 12, 2025 at 11:17 PM Amit Langote <[email protected]> wrote:
> The key idea is to avoid taking unnecessary locks when reusing a
> cached plan. To achieve that, we need to perform initial partition
> pruning during cached plan reuse in plancache.c so that only surviving
> partitions are locked. This requires some plumbing to reuse the result
> of this "early" pruning during executor startup, because repeating the
> pruning logic would be both inefficient and potentially inconsistent
> -- what if you get different results the second time? (I don't have
> proof that this can happen, but some earlier emails mention the
> theoretical risk, so better to be safe.)
>
> So this patch introduces ExecutorPrep(), which allows executor
> metadata such as initial pruning results (valid subplan indexes) and
> full unpruned_relids to be computed ahead of execution and reused
> later by ExecutorStart() and during QueryDesc setup in parallel
> workers using the results shared by the leader. The parallel query bit
> was discussed previously at [1], though I didn’t have a solution I
> liked then.
>
...
> The patch set is structured as follows:
>
> * Refactor partition pruning initialization (0001): separates the
> setup of the pruning state from its execution by introducing
> ExecCreatePartitionPruneStates(). This makes the pruning logic easier
> to reuse and adds flexibility to do only the setup but skip pruning in
> some cases.
>
> * Introduce ExecutorPrep infrastructure (0002): adds ExecutorPrep()
> and ExecPrep as a formal way to perform executor setup ahead of
> execution. This enables caching or transferring pruning results and
> other metadata without triggering execution. ExecutorStart() can now
> consume precomputed prep state from the EState created during
> ExecutorPrep(). ExecPrepCleanup() handles cleanup when the plan is
> invalidated during prep and so not executed; the state is cleaned up
> in the regular ExecutorEnd() path otherwise.
In v1 patch, I had not made ExecutorStart() call ExecutorPrep() to do
the prep work (creating EState, setting up es_relations, checking
permissions) when QueryDesc did not carry the results of
ExecutorPrep() from some earlier stage. Instead, InitPlan() would
detect that prep was absent and perform the missing setup itself. On
second thought it is cleaner for ExecutorStart() to detect the absence
of prep and call ExecutorPrep() directly, matching how prep would be
created when coming from plancache et al.
v2 changes the patch to do that.
> * Enable pruning-aware locking in cached / generic plan reuse (0004):
> extends GetCachedPlan() and CheckCachedPlan() to call ExecutorPrep()
> on each PlannedStmt in the CachedPlan, locking only surviving
> partitions. Adds CachedPlanPrepData to pass this through plan cache
> APIs and down to execution via QueryDesc. Also reinstates the
> firstResultRel locking rule added in 28317de72 but later lost due to
> revert of the earlier pruning patch, to ensure correctness when all
> target partitions are pruned.
Looking at the changes to executor/function.c, I also noticed that I
had mistakenly allocated the ExecutorPrep state in
SQLFunctionCache.fcontext whereas the correct context for execution
related state is SQLFunctionCache.subcontext. In the updated patch,
I've made postquel_start() reparent the prep EState's es_query_cxt to
subcontext from fcontext. I also did not have a test case that
exercised cached plan reuse for SQL functions, so I added one. I split
the function.c's GetCachedPlan() + CachedPlanPrepData plumbing into a
new patch 0005 so it can be reviewed separately, since it is the only
non-mechanical call-site change.
> Benchmark results:
>
> echo "plan_cache_mode = force_generic_plan" >> $PGDATA/postgresql.conf
> for p in 32 64 128 256 512 1024; do pgbench -i --partitions=$p >
> /dev/null 2>&1; echo -ne "$p\t"; pgbench -n -S -T10 -Mprepared | grep
> tps; done
>
> Master
>
> 32 tps = 23841.822407 (without initial connection time)
> 64 tps = 21578.619816 (without initial connection time)
> 128 tps = 18090.500707 (without initial connection time)
> 256 tps = 14152.248201 (without initial connection time)
> 512 tps = 9432.708423 (without initial connection time)
> 1024 tps = 5873.696475 (without initial connection time)
>
> Patched
>
> 32 tps = 24724.245798 (without initial connection time)
> 64 tps = 24858.206407 (without initial connection time)
> 128 tps = 24652.655269 (without initial connection time)
> 256 tps = 23656.756615 (without initial connection time)
> 512 tps = 22299.865769 (without initial connection time)
> 1024 tps = 21911.704317 (without initial connection time)
Re-ran to include 0 partition case and more partitions than 1024:
echo "plan_cache_mode = force_generic_plan" >> $PGDATA/postgresql.conf
for p in 0 8 16 32 64 128 256 512 1024 2048 4096; do pgbench -i
--partitions=$p > /dev/null 2>&1; echo -ne "$p\t"; pgbench -n -S -T10
-Mprepared | grep tps; done
Master
0 tps = 23600.068719 (without initial connection time)
8 tps = 22548.439906 (without initial connection time)
16 tps = 22807.337363 (without initial connection time)
32 tps = 22837.789996 (without initial connection time)
64 tps = 22915.846820 (without initial connection time)
128 tps = 22958.472655 (without initial connection time)
256 tps = 22432.432730 (without initial connection time)
512 tps = 20327.618690 (without initial connection time)
1024 tps = 20554.932475 (without initial connection time)
2048 tps = 19947.061061 (without initial connection time)
4096 tps = 17294.369829 (without initial connection time)
Patched
0 tps = 23869.906654 (without initial connection time)
8 tps = 22682.498914 (without initial connection time)
16 tps = 22714.445711 (without initial connection time)
32 tps = 21653.589371 (without initial connection time)
64 tps = 20571.267545 (without initial connection time)
128 tps = 17138.088269 (without initial connection time)
256 tps = 13027.168426 (without initial connection time)
512 tps = 8689.486966 (without initial connection time)
1024 tps = 5450.525617 (without initial connection time)
2048 tps = 3034.383108 (without initial connection time)
4096 tps = 1560.110609 (without initial connection time)
Tabular format (+ve pct_change means patched better)
partitions master patched pct_change
----------------------------------------------------
0 23869.91 23600.07 -1.1%
8 22682.50 22548.44 -0.6%
16 22714.45 22807.34 +0.4%
32 21653.59 22837.79 +5.5%
64 20571.27 22915.85 +11.4%
128 17138.09 22958.47 +34.0%
256 13027.17 22432.43 +72.2%
512 8689.49 20327.62 +133.9%
1024 5450.53 20554.93 +277.1%
2048 3034.38 19947.06 +557.4%
4096 1560.11 17294.37 +1008.5%
I also did some runs for custom plans. The custom plan path should
behave about the same on master and patched since the early
ExecutorPrep() business only applies to generic plan reuse cases.
echo "plan_cache_mode = force_custom_plan" >> $PGDATA/postgresql.conf
for p in 0 8 16 32 64 128 256 512 1024 2048 4096; do pgbench -i
--partitions=$p > /dev/null 2>&1; echo -ne "$p\t"; pgbench -n -S -T10
-Mprepared | grep tps; done
Master
pgbench -n -S -T10 -Mprepared | grep tps; done
0 tps = 22346.419557 (without initial connection time)
8 tps = 20959.115560 (without initial connection time)
16 tps = 21390.573290 (without initial connection time)
32 tps = 21358.292393 (without initial connection time)
64 tps = 21288.742635 (without initial connection time)
128 tps = 21167.721447 (without initial connection time)
256 tps = 21256.618661 (without initial connection time)
512 tps = 19401.261197 (without initial connection time)
1024 tps = 19169.135145 (without initial connection time)
2048 tps = 19504.102179 (without initial connection time)
4096 tps = 18880.855783 (without initial connection time)
Patched
0 tps = 22852.634752 (without initial connection time)
8 tps = 21596.432690 (without initial connection time)
16 tps = 21428.779996 (without initial connection time)
32 tps = 20629.225272 (without initial connection time)
64 tps = 21301.644733 (without initial connection time)
128 tps = 21098.543942 (without initial connection time)
256 tps = 21394.364662 (without initial connection time)
512 tps = 19475.152170 (without initial connection time)
1024 tps = 19585.768438 (without initial connection time)
2048 tps = 19810.211969 (without initial connection time)
4096 tps = 19160.981608 (without initial connection time)
In tabular format:
partitions master patched pct_change
----------------------------------------------------
0 22346.42 22852.63 +2.3%
8 20959.12 21596.43 +3.0%
16 21390.57 21428.78 +0.2%
32 21358.29 20629.23 -3.4%
64 21288.74 21301.64 +0.1%
128 21167.72 21098.54 -0.3%
256 21256.62 21394.36 +0.6%
512 19401.26 19475.15 +0.4%
1024 19169.14 19585.77 +2.2%
2048 19504.10 19810.21 +1.6%
4096 18880.86 19160.98 +1.5%
Numbers look within noise range as expected.
--
Thanks, Amit Langote
Attachments:
[application/octet-stream] v2-0005-Make-SQL-function-executor-track-ExecutorPrep-sta.patch (6.5K, 2-v2-0005-Make-SQL-function-executor-track-ExecutorPrep-sta.patch)
download | inline diff:
From eef8d1af46ca8deefbf8eb95428d37fc900a0944 Mon Sep 17 00:00:00 2001
From: Amit Langote <[email protected]>
Date: Mon, 17 Nov 2025 17:40:26 +0900
Subject: [PATCH v2 5/5] Make SQL function executor track ExecutorPrep state
Extend the SQL function executor to use the ExecutorPrep results
returned by GetCachedPlan(). init_execution_state() now passes a
CachedPlanPrepData to GetCachedPlan() and stores the per statement
ExecPrep pointers in the execution_state nodes.
At execution time, postquel_start() reparents the prep estate's
es_query_cxt under the function's subcontext so that prep state
follows the usual per call context hierarchy.
This allows SQL language functions to participate in the same
ExecutorPrep machinery as other plan cache users, which a later
patch will use to support pruning aware locking.
Add a regression test where rule rewrite expands a single UPDATE
into multiple PlannedStmts, exercising the SQL function plan cache
and the generic plan reuse path that now invokes ExecutorPrep.
---
src/backend/executor/functions.c | 33 +++++++++++++++++++++++--
src/test/regress/expected/plancache.out | 31 +++++++++++++++++++++++
src/test/regress/sql/plancache.sql | 29 ++++++++++++++++++++++
3 files changed, 91 insertions(+), 2 deletions(-)
diff --git a/src/backend/executor/functions.c b/src/backend/executor/functions.c
index 633310c5f5b..ed7352fce61 100644
--- a/src/backend/executor/functions.c
+++ b/src/backend/executor/functions.c
@@ -72,6 +72,7 @@ typedef struct execution_state
bool setsResult; /* true if this query produces func's result */
bool lazyEval; /* true if should fetch one row at a time */
PlannedStmt *stmt; /* plan for this query */
+ ExecPrep *prep; /* ExecutorPrep() output for this plan */
QueryDesc *qd; /* null unless status == RUN */
} execution_state;
@@ -657,6 +658,8 @@ init_execution_state(SQLFunctionCachePtr fcache)
execution_state *lasttages = NULL;
int nstmts;
ListCell *lc;
+ CachedPlanPrepData cprep = {0};
+ int i;
/*
* Clean up after previous query, if there was one.
@@ -695,10 +698,20 @@ init_execution_state(SQLFunctionCachePtr fcache)
* CurrentResourceOwner will be the same when ShutdownSQLFunction runs.)
*/
fcache->cowner = CurrentResourceOwner;
+
+ /*
+ * Have ExecutorPrep() allocate under fcache->fcontext. The prep
+ * EStates it creates will initially live there; postquel_start()
+ * will later reparent their es_query_cxt into fcache->subcontext
+ * when using them for execution.
+ */
+ cprep.context = fcache->fcontext;
+ cprep.owner = fcache->cowner;
fcache->cplan = GetCachedPlan(plansource,
fcache->paramLI,
fcache->cowner,
- NULL);
+ NULL,
+ &cprep);
/*
* If necessary, make esarray[] bigger to hold the needed state.
@@ -719,9 +732,12 @@ init_execution_state(SQLFunctionCachePtr fcache)
/*
* Build execution_state list to match the number of contained plans.
*/
+ i = 0;
foreach(lc, fcache->cplan->stmt_list)
{
PlannedStmt *stmt = lfirst_node(PlannedStmt, lc);
+ ExecPrep *prep = cprep.prep_list ? list_nth(cprep.prep_list, i++) :
+ NULL;
execution_state *newes;
/*
@@ -763,6 +779,7 @@ init_execution_state(SQLFunctionCachePtr fcache)
newes->setsResult = false; /* might change below */
newes->lazyEval = false; /* might change below */
newes->stmt = stmt;
+ newes->prep = prep;
newes->qd = NULL;
if (stmt->canSetTag)
@@ -1361,8 +1378,20 @@ postquel_start(execution_state *es, SQLFunctionCachePtr fcache)
else
dest = None_Receiver;
+ if (es->prep)
+ {
+ /*
+ * Prep EStates were built under fcache->fcontext. For execution,
+ * make their es_query_cxt a child of fcache->subcontext so they
+ * follow the usual per call lifetime.
+ */
+ EState *prep_estate = es->prep->prep_estate;
+
+ MemoryContextSetParent(prep_estate->es_query_cxt, fcache->subcontext);
+ }
+
es->qd = CreateQueryDesc(es->stmt,
- NULL,
+ es->prep,
fcache->func->src,
GetActiveSnapshot(),
InvalidSnapshot,
diff --git a/src/test/regress/expected/plancache.out b/src/test/regress/expected/plancache.out
index 4e59188196c..8c68691df91 100644
--- a/src/test/regress/expected/plancache.out
+++ b/src/test/regress/expected/plancache.out
@@ -398,3 +398,34 @@ select name, generic_plans, custom_plans from pg_prepared_statements
(1 row)
drop table test_mode;
+-- exercise sql-function plan cache when rewrite expands a single statement
+-- into multiple planned statements. this forces cachedplan->stmt_list to
+-- contain more than one entry and checks that executor state for the first
+-- rewritten statement does not destroy state needed by the second one.
+set plan_cache_mode = force_generic_plan;
+create table sqlf_base(id int, val int);
+create table sqlf_log(id int, note text);
+insert into sqlf_base values (1, 10);
+create rule sqlf_base_upd_log as
+on update to sqlf_base do also
+ insert into sqlf_log(id, note)
+ values (new.id, 'logged by rule');
+create or replace function sqlf_execprep_test(a int, v int)
+returns void
+language sql
+as $$
+ update sqlf_base set val = v where id = a;
+$$;
+select sqlf_execprep_test(1, 20);
+ sqlf_execprep_test
+--------------------
+
+(1 row)
+
+select sqlf_execprep_test(1, 30);
+ sqlf_execprep_test
+--------------------
+
+(1 row)
+
+reset plan_cache_mode;
diff --git a/src/test/regress/sql/plancache.sql b/src/test/regress/sql/plancache.sql
index 4b2f11dcc64..56ebbbdecd2 100644
--- a/src/test/regress/sql/plancache.sql
+++ b/src/test/regress/sql/plancache.sql
@@ -223,3 +223,32 @@ select name, generic_plans, custom_plans from pg_prepared_statements
where name = 'test_mode_pp';
drop table test_mode;
+
+-- exercise sql-function plan cache when rewrite expands a single statement
+-- into multiple planned statements. this forces cachedplan->stmt_list to
+-- contain more than one entry and checks that executor state for the first
+-- rewritten statement does not destroy state needed by the second one.
+
+set plan_cache_mode = force_generic_plan;
+
+create table sqlf_base(id int, val int);
+create table sqlf_log(id int, note text);
+
+insert into sqlf_base values (1, 10);
+
+create rule sqlf_base_upd_log as
+on update to sqlf_base do also
+ insert into sqlf_log(id, note)
+ values (new.id, 'logged by rule');
+
+create or replace function sqlf_execprep_test(a int, v int)
+returns void
+language sql
+as $$
+ update sqlf_base set val = v where id = a;
+$$;
+
+select sqlf_execprep_test(1, 20);
+select sqlf_execprep_test(1, 30);
+
+reset plan_cache_mode;
--
2.47.3
[application/octet-stream] v2-0001-Refactor-partition-pruning-initialization-for-cla.patch (7.7K, 3-v2-0001-Refactor-partition-pruning-initialization-for-cla.patch)
download | inline diff:
From 243d407de86b0a73b9bd8c8dbc541f630eb33747 Mon Sep 17 00:00:00 2001
From: Amit Langote <[email protected]>
Date: Tue, 11 Nov 2025 21:18:24 +0900
Subject: [PATCH v2 1/5] Refactor partition pruning initialization for clarity
and modularity
Move the creation of PartitionPruneState structures out of
ExecDoInitialPruning() into a new ExecCreatePartitionPruneStates()
function. This separates the setup of pruning state from the execution
of initial pruning logic, making the code clearer and easier to
maintain.
Also simplify handling of unpruned relids by moving responsibility
for recording them in EState into CreatePartitionPruneState(),
avoiding the need to pass all_leafpart_rtis as an out parameter.
This refactoring allows callers to reuse the pruning setup logic
without always triggering pruning, a capability useful for future use
cases that may only need metadata initialization.
---
src/backend/executor/execPartition.c | 70 +++++++++++++++++-----------
src/include/executor/execPartition.h | 1 +
2 files changed, 43 insertions(+), 28 deletions(-)
diff --git a/src/backend/executor/execPartition.c b/src/backend/executor/execPartition.c
index aa12e9ad2ea..88b150c8d77 100644
--- a/src/backend/executor/execPartition.c
+++ b/src/backend/executor/execPartition.c
@@ -182,8 +182,7 @@ static char *ExecBuildSlotPartitionKeyDescription(Relation rel,
static List *adjust_partition_colnos(List *colnos, ResultRelInfo *leaf_part_rri);
static List *adjust_partition_colnos_using_map(List *colnos, AttrMap *attrMap);
static PartitionPruneState *CreatePartitionPruneState(EState *estate,
- PartitionPruneInfo *pruneinfo,
- Bitmapset **all_leafpart_rtis);
+ PartitionPruneInfo *pruneinfo);
static void InitPartitionPruneContext(PartitionPruneContext *context,
List *pruning_steps,
PartitionDesc partdesc,
@@ -1772,6 +1771,9 @@ adjust_partition_colnos_using_map(List *colnos, AttrMap *attrMap)
*
* Functions:
*
+ * ExecCreatePartitionPruneStates
+ * Create PartitionPruneState for all PartitionPruneInfos in the EState
+ *
* ExecDoInitialPruning:
* Perform runtime "initial" pruning, if necessary, to determine the set
* of child subnodes that need to be initialized during ExecInitNode() for
@@ -1796,6 +1798,29 @@ adjust_partition_colnos_using_map(List *colnos, AttrMap *attrMap)
*-------------------------------------------------------------------------
*/
+/*
+ * ExecCreatePartitionPruneStates
+ *
+ * Create a PartitionPruneState for each PartitionPruneInfo in the estate,
+ * and save them in estate->es_part_prune_states. This setup is required
+ * before any initial or runtime pruning can occur.
+ */
+void
+ExecCreatePartitionPruneStates(EState *estate)
+{
+ ListCell *lc;
+
+ foreach(lc, estate->es_part_prune_infos)
+ {
+ PartitionPruneInfo *pruneinfo = lfirst_node(PartitionPruneInfo, lc);
+ PartitionPruneState *prunestate;
+
+ /* Create and save the PartitionPruneState. */
+ prunestate = CreatePartitionPruneState(estate, pruneinfo);
+ estate->es_part_prune_states = lappend(estate->es_part_prune_states,
+ prunestate);
+ }
+}
/*
* ExecDoInitialPruning
@@ -1803,11 +1828,11 @@ adjust_partition_colnos_using_map(List *colnos, AttrMap *attrMap)
* of child subnodes that need to be initialized during ExecInitNode() for
* plan nodes that support partition pruning.
*
- * This function iterates over each PartitionPruneInfo entry in
- * estate->es_part_prune_infos. For each entry, it creates a PartitionPruneState
- * and adds it to es_part_prune_states. ExecInitPartitionExecPruning() accesses
+ * This function iterates over each PartitionPruneState in
+ * estate->es_part_prune_states, which must have been populated earlier by
+ * ExecCreatePartitionPruneStates(). ExecInitPartitionExecPruning() accesses
* these states through their corresponding indexes in es_part_prune_states and
- * assign each state to the parent node's PlanState, from where it will be used
+ * assigns each state to the parent node's PlanState, from where it will be used
* for "exec" pruning.
*
* If initial pruning steps exist for a PartitionPruneInfo entry, this function
@@ -1825,20 +1850,13 @@ ExecDoInitialPruning(EState *estate)
{
ListCell *lc;
- foreach(lc, estate->es_part_prune_infos)
+ Assert(estate->es_part_prune_results == NULL);
+ foreach(lc, estate->es_part_prune_states)
{
- PartitionPruneInfo *pruneinfo = lfirst_node(PartitionPruneInfo, lc);
- PartitionPruneState *prunestate;
+ PartitionPruneState *prunestate = (PartitionPruneState *) lfirst(lc);
Bitmapset *validsubplans = NULL;
- Bitmapset *all_leafpart_rtis = NULL;
Bitmapset *validsubplan_rtis = NULL;
- /* Create and save the PartitionPruneState. */
- prunestate = CreatePartitionPruneState(estate, pruneinfo,
- &all_leafpart_rtis);
- estate->es_part_prune_states = lappend(estate->es_part_prune_states,
- prunestate);
-
/*
* Perform initial pruning steps, if any, and save the result
* bitmapset or NULL as described in the header comment.
@@ -1846,8 +1864,6 @@ ExecDoInitialPruning(EState *estate)
if (prunestate->do_initial_prune)
validsubplans = ExecFindMatchingSubPlans(prunestate, true,
&validsubplan_rtis);
- else
- validsubplan_rtis = all_leafpart_rtis;
estate->es_unpruned_relids = bms_add_members(estate->es_unpruned_relids,
validsubplan_rtis);
@@ -1965,14 +1981,12 @@ ExecInitPartitionExecPruning(PlanState *planstate,
* parent plan node's PlanState.
*
* If initial pruning steps are to be skipped (e.g., during EXPLAIN
- * (GENERIC_PLAN)), *all_leafpart_rtis will be populated with the RT indexes of
- * all leaf partitions whose scanning subnode is included in the parent plan
- * node's list of child plans. The caller must add these RT indexes to
- * estate->es_unpruned_relids.
+ * (GENERIC_PLAN)), the RT indexes of all leaf partitions whose scanning
+ * subnode is included in the parent plan node's list of child plans are
+ * added to estate->es_unpruned_relids.
*/
static PartitionPruneState *
-CreatePartitionPruneState(EState *estate, PartitionPruneInfo *pruneinfo,
- Bitmapset **all_leafpart_rtis)
+CreatePartitionPruneState(EState *estate, PartitionPruneInfo *pruneinfo)
{
PartitionPruneState *prunestate;
int n_part_hierarchies;
@@ -2206,8 +2220,8 @@ CreatePartitionPruneState(EState *estate, PartitionPruneInfo *pruneinfo,
pinfo->execparamids);
/*
- * Return all leaf partition indexes if we're skipping pruning in
- * the EXPLAIN (GENERIC_PLAN) case.
+ * Add all leaf partition indexes to es_unpruned_relids if we're
+ * skipping pruning in the EXPLAIN (GENERIC_PLAN) case.
*/
if (pinfo->initial_pruning_steps && !prunestate->do_initial_prune)
{
@@ -2219,8 +2233,8 @@ CreatePartitionPruneState(EState *estate, PartitionPruneInfo *pruneinfo,
Index rtindex = pprune->leafpart_rti_map[part_index];
if (rtindex)
- *all_leafpart_rtis = bms_add_member(*all_leafpart_rtis,
- rtindex);
+ estate->es_unpruned_relids =
+ bms_add_member(estate->es_unpruned_relids, rtindex);
}
}
diff --git a/src/include/executor/execPartition.h b/src/include/executor/execPartition.h
index 3b3f46aced0..ba8cc594fc9 100644
--- a/src/include/executor/execPartition.h
+++ b/src/include/executor/execPartition.h
@@ -130,6 +130,7 @@ typedef struct PartitionPruneState
PartitionPruningData *partprunedata[FLEXIBLE_ARRAY_MEMBER];
} PartitionPruneState;
+extern void ExecCreatePartitionPruneStates(EState *estate);
extern void ExecDoInitialPruning(EState *estate);
extern PartitionPruneState *ExecInitPartitionExecPruning(PlanState *planstate,
int n_total_subplans,
--
2.47.3
[application/octet-stream] v2-0004-Use-pruning-aware-locking-in-cached-plans.patch (24.0K, 4-v2-0004-Use-pruning-aware-locking-in-cached-plans.patch)
download | inline diff:
From 74dc075dc8f844e036fc38e005fc512b6dd54bc9 Mon Sep 17 00:00:00 2001
From: Amit Langote <[email protected]>
Date: Tue, 11 Nov 2025 22:30:52 +0900
Subject: [PATCH v2 4/5] Use pruning-aware locking in cached plans
Extend GetCachedPlan() to perform ExecutorPrep() on each planned
statement, capturing unpruned relids and initial pruning results.
Use this data to acquire execution locks only on surviving partitions,
avoiding unnecessary locking of pruned tables even when using cached
plans.
Introduce CachedPlanPrepData to carry ExecutorPrep results
through the plan caching layer. Adjust call sites in SPI,
functions, portals, and EXPLAIN to propagate this data.
This ensures pruning decisions made during initial pruning are
consistently reused without redoing pruning logic in executor paths
like parallel workers. It also lays the groundwork for
pruning-dependent lock behavior during plan reuse.
To maintain correctness when all target partitions are pruned, also
reinstate the firstResultRel locking behavior lost in commit
28317de72. That commit required the first ModifyTable target to
remain initialized for executor assumptions to hold. We now
explicitly track these relids in PlannerGlobal and PlannedStmt so they
are locked even if pruned, preserving that rule across cached plan
reuse.
---
src/backend/commands/prepare.c | 19 +-
src/backend/executor/nodeModifyTable.c | 4 +-
src/backend/executor/spi.c | 26 ++-
src/backend/optimizer/plan/planner.c | 1 +
src/backend/optimizer/plan/setrefs.c | 3 +
src/backend/tcop/postgres.c | 9 +-
src/backend/utils/cache/plancache.c | 234 ++++++++++++++++++++++++-
src/include/nodes/pathnodes.h | 3 +
src/include/nodes/plannodes.h | 10 ++
src/include/utils/plancache.h | 24 ++-
10 files changed, 312 insertions(+), 21 deletions(-)
diff --git a/src/backend/commands/prepare.c b/src/backend/commands/prepare.c
index afd449c73ba..23332d19b37 100644
--- a/src/backend/commands/prepare.c
+++ b/src/backend/commands/prepare.c
@@ -154,6 +154,7 @@ ExecuteQuery(ParseState *pstate,
{
PreparedStatement *entry;
CachedPlan *cplan;
+ CachedPlanPrepData cprep = {0};
List *plan_list;
ParamListInfo paramLI = NULL;
EState *estate = NULL;
@@ -193,7 +194,10 @@ ExecuteQuery(ParseState *pstate,
entry->plansource->query_string);
/* Replan if needed, and increment plan refcount for portal */
- cplan = GetCachedPlan(entry->plansource, paramLI, NULL, NULL);
+ /* Keep ExecutorPrep state with the portal and its resowner. */
+ cprep.context = portal->portalContext;
+ cprep.owner = portal->resowner;
+ cplan = GetCachedPlan(entry->plansource, paramLI, NULL, NULL, &cprep);
plan_list = cplan->stmt_list;
/*
@@ -205,7 +209,7 @@ ExecuteQuery(ParseState *pstate,
query_string,
entry->plansource->commandTag,
plan_list,
- NIL,
+ cprep.prep_list,
cplan);
/*
@@ -575,6 +579,7 @@ ExplainExecuteQuery(ExecuteStmt *execstmt, IntoClause *into, ExplainState *es,
PreparedStatement *entry;
const char *query_string;
CachedPlan *cplan;
+ CachedPlanPrepData cprep = {0};
List *plan_list;
List *prep_list;
ListCell *p;
@@ -633,8 +638,14 @@ ExplainExecuteQuery(ExecuteStmt *execstmt, IntoClause *into, ExplainState *es,
}
/* Replan if needed, and acquire a transient refcount */
+ /* ExecutorPrep state is local to this EXPLAIN EXECUTE call. */
+ cprep.context = CurrentMemoryContext;
+ cprep.owner = CurrentResourceOwner;
+ if (es->generic)
+ cprep.eflags = EXEC_FLAG_EXPLAIN_GENERIC;
cplan = GetCachedPlan(entry->plansource, paramLI,
- CurrentResourceOwner, pstate->p_queryEnv);
+ CurrentResourceOwner, pstate->p_queryEnv,
+ &cprep);
INSTR_TIME_SET_CURRENT(planduration);
INSTR_TIME_SUBTRACT(planduration, planstart);
@@ -653,7 +664,7 @@ ExplainExecuteQuery(ExecuteStmt *execstmt, IntoClause *into, ExplainState *es,
}
plan_list = cplan->stmt_list;
- prep_list = NIL;
+ prep_list = cprep.prep_list;
/* Explain each query */
i = 0;
diff --git a/src/backend/executor/nodeModifyTable.c b/src/backend/executor/nodeModifyTable.c
index 4c5647ac38a..c5812612f8d 100644
--- a/src/backend/executor/nodeModifyTable.c
+++ b/src/backend/executor/nodeModifyTable.c
@@ -4648,8 +4648,8 @@ ExecInitModifyTable(ModifyTable *node, EState *estate, int eflags)
* as a reference for building the ResultRelInfo of the target partition.
* In either case, it doesn't matter which result relation is kept, so we
* just keep the first one, if all others have been pruned. See also,
- * ExecDoInitialPruning(), which ensures that this first result relation
- * has been locked.
+ * AcquireExecutorLocksUnpruned(), which ensures that this first result
+ * relation has been locked.
*/
i = 0;
foreach(l, node->resultRelations)
diff --git a/src/backend/executor/spi.c b/src/backend/executor/spi.c
index 7a3cb944d6f..d580f1e0425 100644
--- a/src/backend/executor/spi.c
+++ b/src/backend/executor/spi.c
@@ -1579,6 +1579,7 @@ SPI_cursor_open_internal(const char *name, SPIPlanPtr plan,
{
CachedPlanSource *plansource;
CachedPlan *cplan;
+ CachedPlanPrepData cprep = {0};
List *stmt_list;
char *query_string;
Snapshot snapshot;
@@ -1659,7 +1660,11 @@ SPI_cursor_open_internal(const char *name, SPIPlanPtr plan,
*/
/* Replan if needed, and increment plan refcount for portal */
- cplan = GetCachedPlan(plansource, paramLI, NULL, _SPI_current->queryEnv);
+ /* ExecutorPrep state lives in this portal's context. */
+ cprep.context = portal->portalContext;
+ cprep.owner = portal->resowner;
+ cplan = GetCachedPlan(plansource, paramLI, NULL, _SPI_current->queryEnv,
+ &cprep);
stmt_list = cplan->stmt_list;
if (!plan->saved)
@@ -1685,7 +1690,7 @@ SPI_cursor_open_internal(const char *name, SPIPlanPtr plan,
query_string,
plansource->commandTag,
stmt_list,
- NIL,
+ cprep.prep_list, /* lives in portalContext */
cplan);
/*
@@ -2078,6 +2083,7 @@ SPI_plan_get_cached_plan(SPIPlanPtr plan)
{
CachedPlanSource *plansource;
CachedPlan *cplan;
+ CachedPlanPrepData cprep = {0};
SPICallbackArg spicallbackarg;
ErrorContextCallback spierrcontext;
@@ -2101,9 +2107,13 @@ SPI_plan_get_cached_plan(SPIPlanPtr plan)
error_context_stack = &spierrcontext;
/* Get the generic plan for the query */
+ /* ExecutorPrep() state lives in caller's active context. */
+ cprep.context = CurrentMemoryContext;
+ cprep.owner = CurrentResourceOwner;
cplan = GetCachedPlan(plansource, NULL,
plan->saved ? CurrentResourceOwner : NULL,
- _SPI_current->queryEnv);
+ _SPI_current->queryEnv,
+ &cprep);
Assert(cplan == plansource->gplan);
/* Pop the error context stack */
@@ -2501,6 +2511,7 @@ _SPI_execute_plan(SPIPlanPtr plan, const SPIExecuteOptions *options,
CachedPlanSource *plansource = (CachedPlanSource *) lfirst(lc1);
List *stmt_list;
ListCell *lc2;
+ CachedPlanPrepData cprep = {0};
List *prep_list;
int i;
@@ -2577,11 +2588,16 @@ _SPI_execute_plan(SPIPlanPtr plan, const SPIExecuteOptions *options,
* Replan if needed, and increment plan refcount. If it's a saved
* plan, the refcount must be backed by the plan_owner.
*/
+
+ /* ExecutorPrep state is per _SPI_execute_plan call. */
+ cprep.context = CurrentMemoryContext;
+ cprep.owner = CurrentResourceOwner;
cplan = GetCachedPlan(plansource, options->params,
- plan_owner, _SPI_current->queryEnv);
+ plan_owner, _SPI_current->queryEnv,
+ &cprep);
stmt_list = cplan->stmt_list;
- prep_list = NIL;
+ prep_list = cprep.prep_list;
/*
* If we weren't given a specific snapshot to use, and the statement
diff --git a/src/backend/optimizer/plan/planner.c b/src/backend/optimizer/plan/planner.c
index c4fd646b999..4c76e78c1da 100644
--- a/src/backend/optimizer/plan/planner.c
+++ b/src/backend/optimizer/plan/planner.c
@@ -608,6 +608,7 @@ standard_planner(Query *parse, const char *query_string, int cursorOptions,
glob->prunableRelids);
result->permInfos = glob->finalrteperminfos;
result->resultRelations = glob->resultRelations;
+ result->firstResultRels = glob->firstResultRels;
result->appendRelations = glob->appendRelations;
result->subplans = glob->subplans;
result->rewindPlanIDs = glob->rewindPlanIDs;
diff --git a/src/backend/optimizer/plan/setrefs.c b/src/backend/optimizer/plan/setrefs.c
index ccdc9bc264a..229b39060ae 100644
--- a/src/backend/optimizer/plan/setrefs.c
+++ b/src/backend/optimizer/plan/setrefs.c
@@ -1274,6 +1274,9 @@ set_plan_refs(PlannerInfo *root, Plan *plan, int rtoffset)
lappend_int(root->glob->resultRelations,
splan->rootRelation);
}
+ root->glob->firstResultRels =
+ lappend_int(root->glob->firstResultRels,
+ linitial_int(splan->resultRelations));
}
break;
case T_Append:
diff --git a/src/backend/tcop/postgres.c b/src/backend/tcop/postgres.c
index d3964a12a14..249829f59a0 100644
--- a/src/backend/tcop/postgres.c
+++ b/src/backend/tcop/postgres.c
@@ -1639,6 +1639,7 @@ exec_bind_message(StringInfo input_message)
int16 *rformats = NULL;
CachedPlanSource *psrc;
CachedPlan *cplan;
+ CachedPlanPrepData cprep = {0};
Portal portal;
char *query_string;
char *saved_stmt_name;
@@ -2021,7 +2022,11 @@ exec_bind_message(StringInfo input_message)
* will be generated in MessageContext. The plan refcount will be
* assigned to the Portal, so it will be released at portal destruction.
*/
- cplan = GetCachedPlan(psrc, params, NULL, NULL);
+
+ /* ExecutorPrep() state lives in portal context. */
+ cprep.context = portal->portalContext;
+ cprep.owner = portal->resowner;
+ cplan = GetCachedPlan(psrc, params, NULL, NULL, &cprep);
/*
* Now we can define the portal.
@@ -2034,7 +2039,7 @@ exec_bind_message(StringInfo input_message)
query_string,
psrc->commandTag,
cplan->stmt_list,
- NIL,
+ cprep.prep_list,
cplan);
/* Portal is defined, set the plan ID based on its contents. */
diff --git a/src/backend/utils/cache/plancache.c b/src/backend/utils/cache/plancache.c
index 6661d2c6b73..c1cfd47422c 100644
--- a/src/backend/utils/cache/plancache.c
+++ b/src/backend/utils/cache/plancache.c
@@ -93,7 +93,7 @@ static bool StmtPlanRequiresRevalidation(CachedPlanSource *plansource);
static bool BuildingPlanRequiresSnapshot(CachedPlanSource *plansource);
static List *RevalidateCachedQuery(CachedPlanSource *plansource,
QueryEnvironment *queryEnv);
-static bool CheckCachedPlan(CachedPlanSource *plansource);
+static bool PrepAndCheckCachedPlan(CachedPlanSource *plansource, CachedPlanPrepData *cprep);
static CachedPlan *BuildCachedPlan(CachedPlanSource *plansource, List *qlist,
ParamListInfo boundParams, QueryEnvironment *queryEnv);
static bool choose_custom_plan(CachedPlanSource *plansource,
@@ -101,6 +101,8 @@ static bool choose_custom_plan(CachedPlanSource *plansource,
static double cached_plan_cost(CachedPlan *plan, bool include_planner);
static Query *QueryListGetPrimaryStmt(List *stmts);
static void AcquireExecutorLocks(List *stmt_list, bool acquire);
+static void AcquireExecutorLocksUnpruned(List *stmt_list, bool acquire,
+ CachedPlanPrepData *cprep);
static void AcquirePlannerLocks(List *stmt_list, bool acquire);
static void ScanQueryForLocks(Query *parsetree, bool acquire);
static bool ScanQueryWalker(Node *node, bool *acquire);
@@ -137,6 +139,26 @@ ResourceOwnerForgetPlanCacheRef(ResourceOwner owner, CachedPlan *plan)
/* GUC parameter */
int plan_cache_mode = PLAN_CACHE_MODE_AUTO;
+/*
+ * Lock acquisition policy for execution locks.
+ *
+ * LOCK_ALL acquires locks on all relations mentioned in the plan,
+ * reproducing the behavior of AcquireExecutorLocks().
+ *
+ * LOCK_UNPRUNED restricts locking to only the unpruned relations. That
+ * includes those mentioned in PlannedStmt.unprunableRelids and the leaf
+ * partitions remaining after performing initial pruning.
+ */
+typedef enum LockPolicy
+{
+ LOCK_ALL,
+ LOCK_UNPRUNED,
+} LockPolicy;
+
+static void AcquireExecutorLocksWithPolicy(List *stmt_list,
+ LockPolicy policy, bool acquire,
+ CachedPlanPrepData *cprep);
+
/*
* InitPlanCache: initialize module during InitPostgres.
*
@@ -938,7 +960,12 @@ RevalidateCachedQuery(CachedPlanSource *plansource,
}
/*
- * CheckCachedPlan: see if the CachedPlanSource's generic plan is valid.
+ * PrepAndCheckCachedPlan: see if the CachedPlanSource's generic plan is valid.
+ *
+ * If 'cprep' is not NULL, ExecutorPrep() is applied to each PlannedStmt to
+ * compute the set of partitions that survive initial runtime pruning in order
+ * to only lock them. The resulting ExecPrep structures are saved in cprep for
+ * later reuse by ExecutorStart().
*
* Caller must have already called RevalidateCachedQuery to verify that the
* querytree is up to date.
@@ -947,7 +974,7 @@ RevalidateCachedQuery(CachedPlanSource *plansource,
* (We must do this for the "true" result to be race-condition-free.)
*/
static bool
-CheckCachedPlan(CachedPlanSource *plansource)
+PrepAndCheckCachedPlan(CachedPlanSource *plansource, CachedPlanPrepData *cprep)
{
CachedPlan *plan = plansource->gplan;
@@ -975,13 +1002,15 @@ CheckCachedPlan(CachedPlanSource *plansource)
*/
if (plan->is_valid)
{
+ LockPolicy policy = !cprep ? LOCK_ALL : LOCK_UNPRUNED;
+
/*
* Plan must have positive refcount because it is referenced by
* plansource; so no need to fear it disappears under us here.
*/
Assert(plan->refcount > 0);
- AcquireExecutorLocks(plan->stmt_list, true);
+ AcquireExecutorLocksWithPolicy(plan->stmt_list, policy, true, cprep);
/*
* If plan was transient, check to see if TransactionXmin has
@@ -1003,7 +1032,7 @@ CheckCachedPlan(CachedPlanSource *plansource)
}
/* Oops, the race case happened. Release useless locks. */
- AcquireExecutorLocks(plan->stmt_list, false);
+ AcquireExecutorLocksWithPolicy(plan->stmt_list, policy, false, cprep);
}
/*
@@ -1283,6 +1312,10 @@ cached_plan_cost(CachedPlan *plan, bool include_planner)
* On return, the plan is valid and we have sufficient locks to begin
* execution.
*
+ * If 'cprep' is not NULL and a generic plan is reused, the function prepares
+ * each PlannedStmt via ExecutorPrep() and stores the results in
+ * cprep->prep_list. These are intended to be passed later to ExecutorStart().
+ *
* On return, the refcount of the plan has been incremented; a later
* ReleaseCachedPlan() call is expected. If "owner" is not NULL then
* the refcount has been reported to that ResourceOwner (note that this
@@ -1293,7 +1326,8 @@ cached_plan_cost(CachedPlan *plan, bool include_planner)
*/
CachedPlan *
GetCachedPlan(CachedPlanSource *plansource, ParamListInfo boundParams,
- ResourceOwner owner, QueryEnvironment *queryEnv)
+ ResourceOwner owner, QueryEnvironment *queryEnv,
+ CachedPlanPrepData *cprep)
{
CachedPlan *plan = NULL;
List *qlist;
@@ -1315,7 +1349,9 @@ GetCachedPlan(CachedPlanSource *plansource, ParamListInfo boundParams,
if (!customplan)
{
- if (CheckCachedPlan(plansource))
+ if (cprep)
+ cprep->params = boundParams;
+ if (PrepAndCheckCachedPlan(plansource, cprep))
{
/* We want a generic plan, and we already have a valid one */
plan = plansource->gplan;
@@ -1902,6 +1938,38 @@ QueryListGetPrimaryStmt(List *stmts)
return NULL;
}
+/*
+ * AcquireExecutorLocksWithPolicy
+ * Acquire or release execution locks for a cached plan according to
+ * the specified policy.
+ *
+ * LOCK_ALL reproduces AcquireExecutorLocks(), locking every relation in
+ * each PlannedStmt's rtable. LOCK_UNPRUNED restricts locking to the
+ * unprunable rels and partitions that survive initial runtime pruning.
+ *
+ * When LOCK_UNPRUNED is used on acquire, ExecutorPrep() is invoked for
+ * each PlannedStmt and the resulting ExecPrep pointers are appended to
+ * cprep->prep_list in cprep->context. On release, the same ExecPrep
+ * list is consulted to determine which relations to unlock and is then
+ * cleaned up with ExecPrepCleanup().
+ */
+static void
+AcquireExecutorLocksWithPolicy(List *stmt_list, LockPolicy policy, bool acquire,
+ CachedPlanPrepData *cprep)
+{
+ switch (policy)
+ {
+ case LOCK_ALL:
+ AcquireExecutorLocks(stmt_list, acquire);
+ break;
+ case LOCK_UNPRUNED:
+ AcquireExecutorLocksUnpruned(stmt_list, acquire, cprep);
+ break;
+ default:
+ elog(ERROR, "invalid LockPolicy");
+ }
+}
+
/*
* AcquireExecutorLocks: acquire locks needed for execution of a cached plan;
* or release them if acquire is false.
@@ -1954,6 +2022,158 @@ AcquireExecutorLocks(List *stmt_list, bool acquire)
}
}
+/*
+ * LockRelids
+ * Acquire or release locks on the specified relids, which reference
+ * entries in the provided range table.
+ *
+ * Helper for AcquireExecutorLocksUnpruned().
+ */
+static void
+LockRelids(List *rtable, Bitmapset *relids, bool acquire)
+{
+ int rtindex = -1;
+
+ while ((rtindex = bms_next_member(relids, rtindex)) >= 0)
+ {
+ RangeTblEntry *rte = list_nth_node(RangeTblEntry, rtable, rtindex - 1);
+
+ Assert(rte->rtekind == RTE_RELATION ||
+ (rte->rtekind == RTE_SUBQUERY && OidIsValid(rte->relid)));
+
+ /*
+ * Acquire the appropriate type of lock on each relation OID. Note
+ * that we don't actually try to open the rel, and hence will not
+ * fail if it's been dropped entirely --- we'll just transiently
+ * acquire a non-conflicting lock.
+ */
+ if (acquire)
+ LockRelationOid(rte->relid, rte->rellockmode);
+ else
+ UnlockRelationOid(rte->relid, rte->rellockmode);
+ }
+}
+
+/*
+ * AcquireExecutorLocksUnpruned
+ * Acquire or release execution locks for only unpruned relations
+ * referenced by the given PlannedStmts.
+ *
+ * On acquire, this:
+ * - locks unprunable rels listed in PlannedStmt.unprunableRelids
+ * - runs ExecutorPrep() to perform initial runtime pruning
+ * - locks the surviving partitions reported in the prep estate
+ * - appends the ExecPrep pointer for each PlannedStmt to cprep->prep_list
+ *
+ * On release, it:
+ * - looks up the ExecPrep object for each PlannedStmt from cprep->prep_list
+ * (which must already be populated)
+ * - unlocks the same relations identified during acquire
+ * - calls ExecPrepCleanup() on each ExecPrep
+ *
+ * prep_list is extended during acquire and must match stmt_list one-to-one
+ * when releasing locks. Memory allocation for ExecPrep happens in
+ * cprep->context. Locks are acquired using cprep->owner.
+ */
+
+static void
+AcquireExecutorLocksUnpruned(List *stmt_list, bool acquire,
+ CachedPlanPrepData *cprep)
+{
+ MemoryContext oldcontext = MemoryContextSwitchTo(cprep->context);
+ ListCell *lc1;
+ List *prep_list;
+ int i;
+
+ Assert(cprep);
+
+ /*
+ * When releasing locks, use the ExecPrep list (if any) created during
+ * acquisition to determine which relids to unlock. The list must match
+ * the PlannedStmt list one-to-one.
+ */
+ prep_list = cprep->prep_list;
+ Assert(acquire || list_length(prep_list) == list_length(stmt_list));
+
+ i = 0;
+ foreach(lc1, stmt_list)
+ {
+ PlannedStmt *plannedstmt = lfirst_node(PlannedStmt, lc1);
+ ExecPrep *prep;
+
+ if (plannedstmt->commandType == CMD_UTILITY)
+ {
+ /* Same as AcquireExecutorLocks(). */
+ Query *query = UtilityContainsQuery(plannedstmt->utilityStmt);
+
+ if (query)
+ ScanQueryForLocks(query, acquire);
+
+ /* Keep the list one-to-one with stmt_list. */
+ if (acquire)
+ cprep->prep_list = lappend(cprep->prep_list, NULL);
+ continue;
+ }
+
+ /*
+ * Lock tables mentioned in the original query and other unprunable
+ * relations that were added to the plan via inheritance expansion.
+ */
+ LockRelids(plannedstmt->rtable, plannedstmt->unprunableRelids, acquire);
+
+ /* Lock partitions surviving runtime initial pruning. */
+ if (acquire)
+ {
+ prep = ExecutorPrep(plannedstmt, cprep->params, cprep->owner, true,
+ cprep->eflags);
+ Assert(prep || plannedstmt->partPruneInfos == NULL);
+ cprep->prep_list = lappend(cprep->prep_list, prep);
+ }
+ else
+ prep = list_nth(prep_list, i++);
+
+ Assert(prep == NULL || prep->prep_estate);
+ if (prep)
+ {
+ EState *prep_estate = prep->prep_estate;
+
+ /*
+ * es_unpruned_relids includes plannedstmt->unprunableRelids,
+ * which we've already locked. Filter them out to avoid double-locking.
+ */
+ Bitmapset *lock_relids = bms_difference(prep_estate->es_unpruned_relids,
+ plannedstmt->unprunableRelids);
+
+ /*
+ * firstResultRels may contain pruned partitions that must still be
+ * locked to satisfy executor assumptions (see comments in
+ * ExecInitModifyTable(). Ensure they’re included here.
+ */
+ if (plannedstmt->resultRelations)
+ {
+ ListCell *lc2;
+
+ foreach(lc2, plannedstmt->firstResultRels)
+ {
+ Index firstResultRel = lfirst_int(lc2);
+
+ if (!bms_is_member(firstResultRel, lock_relids))
+ lock_relids = bms_add_member(lock_relids, firstResultRel);
+ }
+ }
+
+ LockRelids(plannedstmt->rtable, lock_relids, acquire);
+ bms_free(lock_relids);
+ }
+
+ /* Clean up prep if releasing locks. */
+ if (!acquire)
+ ExecPrepCleanup(prep);
+ }
+
+ MemoryContextSwitchTo(oldcontext);
+}
+
/*
* AcquirePlannerLocks: acquire locks needed for planning of a querytree list;
* or release them if acquire is false.
diff --git a/src/include/nodes/pathnodes.h b/src/include/nodes/pathnodes.h
index 30d889b54c5..6fb86dc05f6 100644
--- a/src/include/nodes/pathnodes.h
+++ b/src/include/nodes/pathnodes.h
@@ -141,6 +141,9 @@ typedef struct PlannerGlobal
/* "flat" list of integer RT indexes */
List *resultRelations;
+ /* "flat" list of integer RT indexes (one per ModifyTable node) */
+ List *firstResultRels;
+
/* "flat" list of AppendRelInfos */
List *appendRelations;
diff --git a/src/include/nodes/plannodes.h b/src/include/nodes/plannodes.h
index c4393a94321..eb211f1ba56 100644
--- a/src/include/nodes/plannodes.h
+++ b/src/include/nodes/plannodes.h
@@ -123,6 +123,16 @@ typedef struct PlannedStmt
/* integer list of RT indexes, or NIL */
List *resultRelations;
+ /*
+ * rtable indexes of first target relation in each ModifyTable node in the
+ * plan for INSERT/UPDATE/DELETE/MERGE. NIL if resultRelations is NIL.
+ *
+ * These are used by AcquireExecutorLocksUnpruned() to ensure that the
+ * first result rel for each ModifyTable remains locked even if pruned;
+ * see ExecInitModifyTable() for the executor side assumptions.
+ */
+ List *firstResultRels;
+
/* list of AppendRelInfo nodes */
List *appendRelations;
diff --git a/src/include/utils/plancache.h b/src/include/utils/plancache.h
index a82b66d4bc2..c7b8ec4be39 100644
--- a/src/include/utils/plancache.h
+++ b/src/include/utils/plancache.h
@@ -197,6 +197,27 @@ typedef struct CachedExpression
} CachedExpression;
+/*
+ * CachedPlanPrepData
+ * Carries ExecutorPrep results for each PlannedStmt in a CachedPlan,
+ * along with context and owner information needed to allocate them.
+ *
+ * prep_list is indexed one-to-one with CachedPlan->stmt_list, and is
+ * populated when GetCachedPlan() prepares a reused generic plan. The
+ * same list is later used to determine which relations to unlock when
+ * releasing execution locks.
+ *
+ * ExecutorPrep state is allocated in 'context' and owned by 'owner'.
+ */
+typedef struct CachedPlanPrepData
+{
+ List *prep_list; /* one ExecPrep per PlannedStmt, or NULL */
+ ParamListInfo params; /* params visible to ExecutorPrep */
+ MemoryContext context; /* where to allocate ExecPrep objects */
+ ResourceOwner owner; /* ResourceOwner for ExecutorPrep state */
+ int eflags; /* executor flags to pass to ExecutorPrep */
+} CachedPlanPrepData;
+
extern void InitPlanCache(void);
extern void ResetPlanCache(void);
@@ -240,7 +261,8 @@ extern List *CachedPlanGetTargetList(CachedPlanSource *plansource,
extern CachedPlan *GetCachedPlan(CachedPlanSource *plansource,
ParamListInfo boundParams,
ResourceOwner owner,
- QueryEnvironment *queryEnv);
+ QueryEnvironment *queryEnv,
+ CachedPlanPrepData *cprep);
extern void ReleaseCachedPlan(CachedPlan *plan, ResourceOwner owner);
extern bool CachedPlanAllowsSimpleValidityCheck(CachedPlanSource *plansource,
--
2.47.3
[application/octet-stream] v2-0003-Reuse-partition-pruning-results-in-parallel-worke.patch (9.1K, 5-v2-0003-Reuse-partition-pruning-results-in-parallel-worke.patch)
download | inline diff:
From d9d95e09961dcb8236e5fe7b2da4a37fda8e5944 Mon Sep 17 00:00:00 2001
From: Amit Langote <[email protected]>
Date: Tue, 11 Nov 2025 22:17:47 +0900
Subject: [PATCH v2 3/5] Reuse partition pruning results in parallel workers
Pass the leader's initial partition pruning results and unpruned
relids to parallel workers and reuse them via ExecutorPrep(). This
avoids repeating pruning logic in workers, which is not only
redundant but also risks divergence due to nondeterminism in pruning
steps or parameter evaluation timing.
Introduce ExecCheckInitialPruningResults() to verify that the results
match what the worker would compute. This check helps catch
inconsistencies across leader and worker pruning logic.
While valuable on its own, this change also lays the foundation for
future optimizations where the leader may take locks only on
surviving partitions. Ensuring that workers follow identical pruning
decisions makes such selective locking safe.
---
src/backend/executor/execParallel.c | 67 +++++++++++++++++++++++++++-
src/backend/executor/execPartition.c | 35 +++++++++++++++
src/include/executor/execPartition.h | 1 +
3 files changed, 102 insertions(+), 1 deletion(-)
diff --git a/src/backend/executor/execParallel.c b/src/backend/executor/execParallel.c
index aedbd9566d6..751590adcc9 100644
--- a/src/backend/executor/execParallel.c
+++ b/src/backend/executor/execParallel.c
@@ -24,6 +24,7 @@
#include "postgres.h"
#include "executor/execParallel.h"
+#include "executor/execPartition.h"
#include "executor/executor.h"
#include "executor/nodeAgg.h"
#include "executor/nodeAppend.h"
@@ -65,6 +66,8 @@
#define PARALLEL_KEY_QUERY_TEXT UINT64CONST(0xE000000000000008)
#define PARALLEL_KEY_JIT_INSTRUMENTATION UINT64CONST(0xE000000000000009)
#define PARALLEL_KEY_WAL_USAGE UINT64CONST(0xE00000000000000A)
+#define PARALLEL_KEY_PARTITION_PRUNE_RESULTS UINT64CONST(0xE00000000000000B)
+#define PARALLEL_KEY_UNPRUNED_RELIDS UINT64CONST(0xE00000000000000C)
#define PARALLEL_TUPLE_QUEUE_SIZE 65536
@@ -608,12 +611,18 @@ ExecInitParallelPlan(PlanState *planstate, EState *estate,
FixedParallelExecutorState *fpes;
char *pstmt_data;
char *pstmt_space;
+ char *part_prune_results_data;
+ char *part_prune_results_space;
+ char *unpruned_relids_data;
+ char *unpruned_relids_space;
char *paramlistinfo_space;
BufferUsage *bufusage_space;
WalUsage *walusage_space;
SharedExecutorInstrumentation *instrumentation = NULL;
SharedJitInstrumentation *jit_instrumentation = NULL;
int pstmt_len;
+ int part_prune_results_len;
+ int unpruned_relids_len;
int paramlistinfo_len;
int instrumentation_len = 0;
int jit_instrumentation_len = 0;
@@ -642,6 +651,8 @@ ExecInitParallelPlan(PlanState *planstate, EState *estate,
/* Fix up and serialize plan to be sent to workers. */
pstmt_data = ExecSerializePlan(planstate->plan, estate);
+ part_prune_results_data = nodeToString(estate->es_part_prune_results);
+ unpruned_relids_data = nodeToString(estate->es_unpruned_relids);
/* Create a parallel context. */
pcxt = CreateParallelContext("postgres", "ParallelQueryMain", nworkers);
@@ -668,6 +679,16 @@ ExecInitParallelPlan(PlanState *planstate, EState *estate,
shm_toc_estimate_chunk(&pcxt->estimator, pstmt_len);
shm_toc_estimate_keys(&pcxt->estimator, 1);
+ /* Estimate space for serialized part_prune_results. */
+ part_prune_results_len = strlen(part_prune_results_data) + 1;
+ shm_toc_estimate_chunk(&pcxt->estimator, part_prune_results_len);
+ shm_toc_estimate_keys(&pcxt->estimator, 1);
+
+ /* Estimate space for serialized unpruned_relids. */
+ unpruned_relids_len = strlen(unpruned_relids_data) + 1;
+ shm_toc_estimate_chunk(&pcxt->estimator, unpruned_relids_len);
+ shm_toc_estimate_keys(&pcxt->estimator, 1);
+
/* Estimate space for serialized ParamListInfo. */
paramlistinfo_len = EstimateParamListSpace(estate->es_param_list_info);
shm_toc_estimate_chunk(&pcxt->estimator, paramlistinfo_len);
@@ -769,6 +790,16 @@ ExecInitParallelPlan(PlanState *planstate, EState *estate,
memcpy(pstmt_space, pstmt_data, pstmt_len);
shm_toc_insert(pcxt->toc, PARALLEL_KEY_PLANNEDSTMT, pstmt_space);
+ /* Store serialized part_prune_results */
+ part_prune_results_space = shm_toc_allocate(pcxt->toc, part_prune_results_len);
+ memcpy(part_prune_results_space, part_prune_results_data, part_prune_results_len);
+ shm_toc_insert(pcxt->toc, PARALLEL_KEY_PARTITION_PRUNE_RESULTS, part_prune_results_space);
+
+ /* Store serialized unpruned_relids */
+ unpruned_relids_space = shm_toc_allocate(pcxt->toc, unpruned_relids_len);
+ memcpy(unpruned_relids_space, unpruned_relids_data, unpruned_relids_len);
+ shm_toc_insert(pcxt->toc, PARALLEL_KEY_UNPRUNED_RELIDS, unpruned_relids_space);
+
/* Store serialized ParamListInfo. */
paramlistinfo_space = shm_toc_allocate(pcxt->toc, paramlistinfo_len);
shm_toc_insert(pcxt->toc, PARALLEL_KEY_PARAMLISTINFO, paramlistinfo_space);
@@ -1263,10 +1294,15 @@ ExecParallelGetQueryDesc(shm_toc *toc, DestReceiver *receiver,
int instrument_options)
{
char *pstmtspace;
+ char *part_prune_results_space;
+ char *unpruned_relids_space;
char *paramspace;
PlannedStmt *pstmt;
+ List *part_prune_results;
+ Bitmapset *unpruned_relids;
ParamListInfo paramLI;
char *queryString;
+ ExecPrep *prep = NULL;
/* Get the query string from shared memory */
queryString = shm_toc_lookup(toc, PARALLEL_KEY_QUERY_TEXT, false);
@@ -1279,9 +1315,38 @@ ExecParallelGetQueryDesc(shm_toc *toc, DestReceiver *receiver,
paramspace = shm_toc_lookup(toc, PARALLEL_KEY_PARAMLISTINFO, false);
paramLI = RestoreParamList(¶mspace);
+ /* Reconstruct leader-supplied part_prune_results and unpruned_relids. */
+ part_prune_results_space =
+ shm_toc_lookup(toc, PARALLEL_KEY_PARTITION_PRUNE_RESULTS, false);
+ part_prune_results = (List *) stringToNode(part_prune_results_space);
+ unpruned_relids_space =
+ shm_toc_lookup(toc, PARALLEL_KEY_UNPRUNED_RELIDS, false);
+ unpruned_relids = (Bitmapset *) stringToNode(unpruned_relids_space);
+
+ /*
+ * If pruning was done in the leader, build a prep estate in the worker
+ * and inject the leader's pruning results into it for reuse.
+ */
+ if (pstmt->partPruneInfos)
+ {
+ prep = ExecutorPrep(pstmt, paramLI, CurrentResourceOwner, false, 0);
+ Assert(prep->prep_estate);
+
+ prep->prep_estate->es_part_prune_results = part_prune_results;
+ prep->prep_estate->es_unpruned_relids =
+ bms_add_members(prep->prep_estate->es_unpruned_relids,
+ unpruned_relids);
+
+ /*
+ * Verify that the pruning results passed from the leader match
+ * what the worker would independently compute.
+ */
+ ExecCheckInitialPruningResults(prep->prep_estate);
+ }
+
/* Create a QueryDesc for the query. */
return CreateQueryDesc(pstmt,
- NULL,
+ prep,
queryString,
GetActiveSnapshot(), InvalidSnapshot,
receiver, paramLI, NULL, instrument_options);
diff --git a/src/backend/executor/execPartition.c b/src/backend/executor/execPartition.c
index 187a480e508..3b450e3373f 100644
--- a/src/backend/executor/execPartition.c
+++ b/src/backend/executor/execPartition.c
@@ -1872,6 +1872,41 @@ ExecDoInitialPruning(EState *estate)
}
}
+/*
+ * ExecCheckInitialPruningResults
+ * Verify partition pruning results passed from the leader process.
+ *
+ * This is intended to be called during parallel worker query setup.
+ * It recomputes initial pruning results locally and compares them with
+ * those received from the leader. Any mismatch may indicate a divergence
+ * between leader and worker logic or environment.
+ */
+void
+ExecCheckInitialPruningResults(EState *estate)
+{
+ ListCell *lc;
+ int i;
+
+ Assert(estate->es_part_prune_results != NULL);
+ i = 0;
+ foreach(lc, estate->es_part_prune_states)
+ {
+ PartitionPruneState *prunestate = (PartitionPruneState *) lfirst(lc);
+ Bitmapset *reuse_validsubplans =
+ list_nth_node(Bitmapset, estate->es_part_prune_results, i);
+ Bitmapset *validsubplans = NULL;
+ Bitmapset *validsubplan_rtis = NULL;
+
+ if (prunestate->do_initial_prune)
+ validsubplans = ExecFindMatchingSubPlans(prunestate, true,
+ &validsubplan_rtis);
+ if (bms_nonempty_difference(validsubplans, reuse_validsubplans))
+ elog(ERROR, "different validsubplns in parallel worker");
+ if (bms_nonempty_difference(validsubplan_rtis, estate->es_unpruned_relids))
+ elog(ERROR, "different unprunable_relids in parallel worker");
+ }
+}
+
/*
* ExecInitPartitionExecPruning
* Initialize the data structures needed for runtime "exec" partition
diff --git a/src/include/executor/execPartition.h b/src/include/executor/execPartition.h
index ba8cc594fc9..126efd008e5 100644
--- a/src/include/executor/execPartition.h
+++ b/src/include/executor/execPartition.h
@@ -132,6 +132,7 @@ typedef struct PartitionPruneState
extern void ExecCreatePartitionPruneStates(EState *estate);
extern void ExecDoInitialPruning(EState *estate);
+extern void ExecCheckInitialPruningResults(EState *estate);
extern PartitionPruneState *ExecInitPartitionExecPruning(PlanState *planstate,
int n_total_subplans,
int part_prune_index,
--
2.47.3
[application/octet-stream] v2-0002-Introduce-ExecutorPrep-and-refactor-executor-star.patch (28.7K, 6-v2-0002-Introduce-ExecutorPrep-and-refactor-executor-star.patch)
download | inline diff:
From 11e0262e31e35539f50e96531559db6cd7e32160 Mon Sep 17 00:00:00 2001
From: Amit Langote <[email protected]>
Date: Tue, 11 Nov 2025 21:47:46 +0900
Subject: [PATCH v2 2/5] Introduce ExecutorPrep and refactor executor startup
Factor permission checks, range table initialization, and initial
partition pruning out of InitPlan() into a new ExecutorPrep()
helper. ExecutorPrep builds an EState containing the executor
metadata needed before plan execution, including partition
pruning state where partPruneInfos are present.
ExecutorStart() now expects QueryDesc->prep to point at such an
ExecPrep object. If no prep was supplied by the caller, it
invokes ExecutorPrep() itself and adopts the resulting EState
for the duration of the query. This keeps the executor startup
behaviour unchanged while making the setup work callable
separately when needed.
CreateQueryDesc() grows a prep argument and stores it in the
QueryDesc. Portals, SPI, SQL functions, and EXPLAIN are wired
to carry an optional ExecPrep pointer alongside the PlannedStmt
list, but most callers still pass NULL and let ExecutorStart()
perform the setup lazily.
Add the ExecPrep struct and ExecPrepCleanup() to encapsulate
ownership of the prepared EState and any caller specific
cleanup hook. Update executor/README and related comments to
document the new control flow and the separation between
preparation and execution.
---
src/backend/commands/copyto.c | 2 +-
src/backend/commands/createas.c | 2 +-
src/backend/commands/explain.c | 7 +-
src/backend/commands/extension.c | 1 +
src/backend/commands/matview.c | 2 +-
src/backend/commands/portalcmds.c | 1 +
src/backend/commands/prepare.c | 11 +-
src/backend/executor/README | 8 +-
src/backend/executor/execMain.c | 179 +++++++++++++++++++++++----
src/backend/executor/execParallel.c | 1 +
src/backend/executor/execPartition.c | 3 +
src/backend/executor/functions.c | 1 +
src/backend/executor/spi.c | 10 ++
src/backend/tcop/postgres.c | 2 +
src/backend/tcop/pquery.c | 27 +++-
src/backend/utils/mmgr/portalmem.c | 2 +
src/include/commands/explain.h | 3 +-
src/include/executor/execdesc.h | 3 +-
src/include/executor/executor.h | 11 ++
src/include/nodes/execnodes.h | 48 +++++++
src/include/utils/portal.h | 2 +
21 files changed, 286 insertions(+), 40 deletions(-)
diff --git a/src/backend/commands/copyto.c b/src/backend/commands/copyto.c
index cef452584e5..5efbb0949c2 100644
--- a/src/backend/commands/copyto.c
+++ b/src/backend/commands/copyto.c
@@ -870,7 +870,7 @@ BeginCopyTo(ParseState *pstate,
((DR_copy *) dest)->cstate = cstate;
/* Create a QueryDesc requesting no output */
- cstate->queryDesc = CreateQueryDesc(plan, pstate->p_sourcetext,
+ cstate->queryDesc = CreateQueryDesc(plan, NULL, pstate->p_sourcetext,
GetActiveSnapshot(),
InvalidSnapshot,
dest, NULL, NULL, 0);
diff --git a/src/backend/commands/createas.c b/src/backend/commands/createas.c
index 1ccc2e55c64..9eabe4920cd 100644
--- a/src/backend/commands/createas.c
+++ b/src/backend/commands/createas.c
@@ -334,7 +334,7 @@ ExecCreateTableAs(ParseState *pstate, CreateTableAsStmt *stmt,
UpdateActiveSnapshotCommandId();
/* Create a QueryDesc, redirecting output to our tuple receiver */
- queryDesc = CreateQueryDesc(plan, pstate->p_sourcetext,
+ queryDesc = CreateQueryDesc(plan, NULL, pstate->p_sourcetext,
GetActiveSnapshot(), InvalidSnapshot,
dest, params, queryEnv, 0);
diff --git a/src/backend/commands/explain.c b/src/backend/commands/explain.c
index 7e699f8595e..d6ab3697dd9 100644
--- a/src/backend/commands/explain.c
+++ b/src/backend/commands/explain.c
@@ -370,7 +370,7 @@ standard_ExplainOneQuery(Query *query, int cursorOptions,
}
/* run it (if needed) and produce output */
- ExplainOnePlan(plan, into, es, queryString, params, queryEnv,
+ ExplainOnePlan(plan, NULL, into, es, queryString, params, queryEnv,
&planduration, (es->buffers ? &bufusage : NULL),
es->memory ? &mem_counters : NULL);
}
@@ -492,7 +492,8 @@ ExplainOneUtility(Node *utilityStmt, IntoClause *into, ExplainState *es,
* to call it.
*/
void
-ExplainOnePlan(PlannedStmt *plannedstmt, IntoClause *into, ExplainState *es,
+ExplainOnePlan(PlannedStmt *plannedstmt, ExecPrep *prep,
+ IntoClause *into, ExplainState *es,
const char *queryString, ParamListInfo params,
QueryEnvironment *queryEnv, const instr_time *planduration,
const BufferUsage *bufusage,
@@ -548,7 +549,7 @@ ExplainOnePlan(PlannedStmt *plannedstmt, IntoClause *into, ExplainState *es,
dest = None_Receiver;
/* Create a QueryDesc for the query */
- queryDesc = CreateQueryDesc(plannedstmt, queryString,
+ queryDesc = CreateQueryDesc(plannedstmt, prep, queryString,
GetActiveSnapshot(), InvalidSnapshot,
dest, params, queryEnv, instrument_option);
diff --git a/src/backend/commands/extension.c b/src/backend/commands/extension.c
index 93ef1ad106f..3cca6d45ec1 100644
--- a/src/backend/commands/extension.c
+++ b/src/backend/commands/extension.c
@@ -993,6 +993,7 @@ execute_sql_string(const char *sql, const char *filename)
QueryDesc *qdesc;
qdesc = CreateQueryDesc(stmt,
+ NULL,
sql,
GetActiveSnapshot(), NULL,
dest, NULL, NULL, 0);
diff --git a/src/backend/commands/matview.c b/src/backend/commands/matview.c
index ef7c0d624f1..30cbf9f264f 100644
--- a/src/backend/commands/matview.c
+++ b/src/backend/commands/matview.c
@@ -437,7 +437,7 @@ refresh_matview_datafill(DestReceiver *dest, Query *query,
UpdateActiveSnapshotCommandId();
/* Create a QueryDesc, redirecting output to our tuple receiver */
- queryDesc = CreateQueryDesc(plan, queryString,
+ queryDesc = CreateQueryDesc(plan, NULL, queryString,
GetActiveSnapshot(), InvalidSnapshot,
dest, NULL, NULL, 0);
diff --git a/src/backend/commands/portalcmds.c b/src/backend/commands/portalcmds.c
index ec96c2efcd3..ac1ddd25aba 100644
--- a/src/backend/commands/portalcmds.c
+++ b/src/backend/commands/portalcmds.c
@@ -118,6 +118,7 @@ PerformCursorOpen(ParseState *pstate, DeclareCursorStmt *cstmt, ParamListInfo pa
queryString,
CMDTAG_SELECT, /* cursor's query is always a SELECT */
list_make1(plan),
+ list_make1(NULL),
NULL);
/*----------
diff --git a/src/backend/commands/prepare.c b/src/backend/commands/prepare.c
index 34b6410d6a2..afd449c73ba 100644
--- a/src/backend/commands/prepare.c
+++ b/src/backend/commands/prepare.c
@@ -205,6 +205,7 @@ ExecuteQuery(ParseState *pstate,
query_string,
entry->plansource->commandTag,
plan_list,
+ NIL,
cplan);
/*
@@ -575,6 +576,7 @@ ExplainExecuteQuery(ExecuteStmt *execstmt, IntoClause *into, ExplainState *es,
const char *query_string;
CachedPlan *cplan;
List *plan_list;
+ List *prep_list;
ListCell *p;
ParamListInfo paramLI = NULL;
EState *estate = NULL;
@@ -585,6 +587,7 @@ ExplainExecuteQuery(ExecuteStmt *execstmt, IntoClause *into, ExplainState *es,
MemoryContextCounters mem_counters;
MemoryContext planner_ctx = NULL;
MemoryContext saved_ctx = NULL;
+ int i;
if (es->memory)
{
@@ -650,14 +653,20 @@ ExplainExecuteQuery(ExecuteStmt *execstmt, IntoClause *into, ExplainState *es,
}
plan_list = cplan->stmt_list;
+ prep_list = NIL;
/* Explain each query */
+ i = 0;
foreach(p, plan_list)
{
PlannedStmt *pstmt = lfirst_node(PlannedStmt, p);
+ ExecPrep *prep = prep_list ?
+ (ExecPrep *) list_nth(prep_list, i) : NULL;
+ i++;
if (pstmt->commandType != CMD_UTILITY)
- ExplainOnePlan(pstmt, into, es, query_string, paramLI, pstate->p_queryEnv,
+ ExplainOnePlan(pstmt, prep,
+ into, es, query_string, paramLI, pstate->p_queryEnv,
&planduration, (es->buffers ? &bufusage : NULL),
es->memory ? &mem_counters : NULL);
else
diff --git a/src/backend/executor/README b/src/backend/executor/README
index 54f4782f31b..95b5ec58c55 100644
--- a/src/backend/executor/README
+++ b/src/backend/executor/README
@@ -291,10 +291,16 @@ Query Processing Control Flow
This is a sketch of control flow for full query processing:
+ ExecutorPrep
+ May be run before ExecutorStart (e.g., for plan validation), or
+ implicitly from ExecutorStart if not done earlier. Performs range
+ table initialization, permission checks, and initial partition pruning.
+ Returns an ExecPrep wrapper with EState that ExecutorStart may reuse.
+
CreateQueryDesc
ExecutorStart
- CreateExecutorState
+ CreateExecutorState (or reuse one from ExecPrep if present)
creates per-query context
switch to per-query context to run ExecInitNode
AfterTriggerBeginQuery
diff --git a/src/backend/executor/execMain.c b/src/backend/executor/execMain.c
index 27c9eec697b..39de0b93a1c 100644
--- a/src/backend/executor/execMain.c
+++ b/src/backend/executor/execMain.c
@@ -171,8 +171,26 @@ standard_ExecutorStart(QueryDesc *queryDesc, int eflags)
/*
* Build EState, switch into per-query memory context for startup.
+ *
+ * If ExecutorPrep() ran earlier (e.g., to do initial pruning during plan
+ * validity checking), reuse its EState to avoid redoing range table setup
+ * and pruning. Otherwise, create a fresh EState as usual.
*/
- estate = CreateExecutorState();
+ if (queryDesc->prep == NULL)
+ queryDesc->prep = ExecutorPrep(queryDesc->plannedstmt,
+ queryDesc->params,
+ CurrentResourceOwner,
+ true,
+ eflags);
+ Assert(queryDesc->prep);
+ estate = queryDesc->prep->prep_estate;
+
+ /*
+ * Executor is adopting the prep's EState. Mark it so ExecPrepCleanup()
+ * doesn't try to free it redundantly.
+ */
+ queryDesc->prep->owns_estate = false;
+
queryDesc->estate = estate;
oldcontext = MemoryContextSwitchTo(estate->es_query_cxt);
@@ -263,6 +281,136 @@ standard_ExecutorStart(QueryDesc *queryDesc, int eflags)
MemoryContextSwitchTo(oldcontext);
}
+/*
+ * ExecutorPrep: prepare executor state for a PlannedStmt outside ExecutorStart.
+ *
+ * Performs range table initialization, permission checks, and initial
+ * partition pruning if partPruneInfos are present and do_initial_pruning is
+ * true.
+ *
+ * This is intended for callers that need executor metadata ahead of actual
+ * execution. Typical use cases include:
+ * - determining which relations must be locked during plan cache validation;
+ * - initializing unpruned relids and valid subplans in parallel workers
+ * using state copied from the leader.
+ *
+ * The executor can reuse the resulting state to avoid redundant setup during
+ * ExecutorStart().
+ *
+ * Returns an ExecPrep wrapper that owns the EState and can be reused
+ * or cleaned up later.
+ */
+ExecPrep *
+ExecutorPrep(PlannedStmt *pstmt, ParamListInfo params, ResourceOwner owner,
+ bool do_initial_pruning, int eflags)
+{
+ ResourceOwner oldowner;
+ EState *estate;
+ bool snapshot_set;
+
+ if (pstmt->commandType == CMD_UTILITY)
+ return NULL;
+
+ /* Pruning may use expressions that require an active snapshot. */
+ snapshot_set = false;
+ if (!ActiveSnapshotSet())
+ {
+ PushActiveSnapshot(GetTransactionSnapshot());
+ snapshot_set = true;
+ }
+ Assert(ActiveSnapshotSet());
+
+ estate = CreateExecutorState();
+ estate->es_plannedstmt = pstmt;
+ estate->es_part_prune_infos = pstmt->partPruneInfos;
+ estate->es_param_list_info = params;
+ estate->es_top_eflags = eflags;
+
+ /*
+ * Do permissions checks.
+ */
+ ExecCheckPermissions(pstmt->rtable, pstmt->permInfos, true);
+
+ /*
+ * Initialize range table.
+ */
+ ExecInitRangeTable(estate, pstmt->rtable, pstmt->permInfos,
+ bms_copy(pstmt->unprunableRelids));
+
+ /*
+ * Ensure locks taken during initial pruning are tracked under the given
+ * ResourceOwner (e.g., one associated with CachedPlan validation).
+ */
+ oldowner = CurrentResourceOwner;
+ CurrentResourceOwner = owner;
+
+ /*
+ * Set up PartitionPruneState structures needed for both initial and
+ * runtime partition pruning. These structures are built from the
+ * PartitionPruneInfo entries in the plan tree.
+ *
+ * If do_initial_pruning is true, also perform initial pruning to compute
+ * the subset of child subplans that will be executed. The results,
+ * which are bitmapsets of selected child indexes, are saved in
+ * es_part_prune_results. This list is parallel to es_part_prune_infos.
+ *
+ * In parallel workers, do_initial_pruning should be false -- they receive
+ * es_part_prune_results from the leader process and should only initialize
+ * the PartitionPruneStates.
+ */
+ ExecCreatePartitionPruneStates(estate);
+ if (do_initial_pruning)
+ ExecDoInitialPruning(estate);
+
+ CurrentResourceOwner = oldowner;
+
+ /* Release snapshot if we got one */
+ if (snapshot_set)
+ PopActiveSnapshot();
+
+ return CreateExecPrep(estate, CurrentMemoryContext, NULL, NULL);
+}
+
+/*
+ * CreateExecPrep: initialize ExecPrep wrapper with optional cleanup metadata.
+ */
+ExecPrep *
+CreateExecPrep(EState *estate, MemoryContext context,
+ execprep_cleanup_fn cleanup, void *cleanup_arg)
+{
+ ExecPrep *prep = palloc0(sizeof(ExecPrep));
+
+ prep->prep_estate = estate;
+ prep->context = context;
+ prep->cleanup = cleanup;
+ prep->cleanup_arg = cleanup_arg;
+ prep->owns_estate = true;
+
+ return prep;
+}
+
+/*
+ * ExecPrepCleanup: free ExecPrep resources not adopted by the executor.
+ *
+ * Only frees the EState if it wasn't taken over by ExecutorStart().
+ * Always runs the optional user-defined cleanup callback.
+ */
+void
+ExecPrepCleanup(ExecPrep *prep)
+{
+ if (prep == NULL)
+ return;
+
+ if (prep->prep_estate && prep->owns_estate)
+ {
+ ExecCloseRangeTableRelations(prep->prep_estate);
+ FreeExecutorState(prep->prep_estate);
+ }
+
+ if (prep->cleanup)
+ prep->cleanup(prep->cleanup_arg);
+}
+
/* ----------------------------------------------------------------
* ExecutorRun
*
@@ -824,7 +972,6 @@ ExecCheckXactReadOnly(PlannedStmt *plannedstmt)
PreventCommandIfParallelMode(CreateCommandName((Node *) plannedstmt));
}
-
/* ----------------------------------------------------------------
* InitPlan
*
@@ -838,37 +985,15 @@ InitPlan(QueryDesc *queryDesc, int eflags)
CmdType operation = queryDesc->operation;
PlannedStmt *plannedstmt = queryDesc->plannedstmt;
Plan *plan = plannedstmt->planTree;
- List *rangeTable = plannedstmt->rtable;
EState *estate = queryDesc->estate;
PlanState *planstate;
TupleDesc tupType;
ListCell *l;
int i;
- /*
- * Do permissions checks
- */
- ExecCheckPermissions(rangeTable, plannedstmt->permInfos, true);
-
- /*
- * initialize the node's execution state
- */
- ExecInitRangeTable(estate, rangeTable, plannedstmt->permInfos,
- bms_copy(plannedstmt->unprunableRelids));
-
- estate->es_plannedstmt = plannedstmt;
- estate->es_part_prune_infos = plannedstmt->partPruneInfos;
-
- /*
- * Perform runtime "initial" pruning to identify which child subplans,
- * corresponding to the children of plan nodes that contain
- * PartitionPruneInfo such as Append, will not be executed. The results,
- * which are bitmapsets of indexes of the child subplans that will be
- * executed, are saved in es_part_prune_results. These results correspond
- * to each PartitionPruneInfo entry, and the es_part_prune_results list is
- * parallel to es_part_prune_infos.
- */
- ExecDoInitialPruning(estate);
+ /* ExecutorPrep() must have been done. */
+ Assert(queryDesc->prep);
+ Assert(estate == queryDesc->prep->prep_estate);
/*
* Next, build the ExecRowMark array from the PlanRowMark(s), if any.
diff --git a/src/backend/executor/execParallel.c b/src/backend/executor/execParallel.c
index f098a5557cf..aedbd9566d6 100644
--- a/src/backend/executor/execParallel.c
+++ b/src/backend/executor/execParallel.c
@@ -1281,6 +1281,7 @@ ExecParallelGetQueryDesc(shm_toc *toc, DestReceiver *receiver,
/* Create a QueryDesc for the query. */
return CreateQueryDesc(pstmt,
+ NULL,
queryString,
GetActiveSnapshot(), InvalidSnapshot,
receiver, paramLI, NULL, instrument_options);
diff --git a/src/backend/executor/execPartition.c b/src/backend/executor/execPartition.c
index 88b150c8d77..187a480e508 100644
--- a/src/backend/executor/execPartition.c
+++ b/src/backend/executor/execPartition.c
@@ -2368,6 +2368,9 @@ InitExecPartitionPruneContexts(PartitionPruneState *prunestate,
Assert(parent_plan != NULL);
estate = parent_plan->state;
+ /* Wouldn't be available at ExecutorPrep() time. */
+ prunestate->econtext->ecxt_param_exec_vals = estate->es_param_exec_vals;
+
/*
* No need to fix subplans maps if initial pruning didn't eliminate any
* subplans.
diff --git a/src/backend/executor/functions.c b/src/backend/executor/functions.c
index 630d708d2a3..633310c5f5b 100644
--- a/src/backend/executor/functions.c
+++ b/src/backend/executor/functions.c
@@ -1362,6 +1362,7 @@ postquel_start(execution_state *es, SQLFunctionCachePtr fcache)
dest = None_Receiver;
es->qd = CreateQueryDesc(es->stmt,
+ NULL,
fcache->func->src,
GetActiveSnapshot(),
InvalidSnapshot,
diff --git a/src/backend/executor/spi.c b/src/backend/executor/spi.c
index 653500b38dc..7a3cb944d6f 100644
--- a/src/backend/executor/spi.c
+++ b/src/backend/executor/spi.c
@@ -1685,6 +1685,7 @@ SPI_cursor_open_internal(const char *name, SPIPlanPtr plan,
query_string,
plansource->commandTag,
stmt_list,
+ NIL,
cplan);
/*
@@ -2500,6 +2501,8 @@ _SPI_execute_plan(SPIPlanPtr plan, const SPIExecuteOptions *options,
CachedPlanSource *plansource = (CachedPlanSource *) lfirst(lc1);
List *stmt_list;
ListCell *lc2;
+ List *prep_list;
+ int i;
spicallbackarg.query = plansource->query_string;
@@ -2578,6 +2581,7 @@ _SPI_execute_plan(SPIPlanPtr plan, const SPIExecuteOptions *options,
plan_owner, _SPI_current->queryEnv);
stmt_list = cplan->stmt_list;
+ prep_list = NIL;
/*
* If we weren't given a specific snapshot to use, and the statement
@@ -2615,12 +2619,17 @@ _SPI_execute_plan(SPIPlanPtr plan, const SPIExecuteOptions *options,
}
}
+ i = 0;
foreach(lc2, stmt_list)
{
PlannedStmt *stmt = lfirst_node(PlannedStmt, lc2);
+ ExecPrep *prep = prep_list ?
+ list_nth(prep_list, i) : NULL;
bool canSetTag = stmt->canSetTag;
DestReceiver *dest;
+ i++;
+
/*
* Reset output state. (Note that if a non-SPI receiver is used,
* _SPI_current->processed will stay zero, and that's what we'll
@@ -2690,6 +2699,7 @@ _SPI_execute_plan(SPIPlanPtr plan, const SPIExecuteOptions *options,
snap = InvalidSnapshot;
qdesc = CreateQueryDesc(stmt,
+ prep,
plansource->query_string,
snap, crosscheck_snapshot,
dest,
diff --git a/src/backend/tcop/postgres.c b/src/backend/tcop/postgres.c
index 2bd89102686..d3964a12a14 100644
--- a/src/backend/tcop/postgres.c
+++ b/src/backend/tcop/postgres.c
@@ -1232,6 +1232,7 @@ exec_simple_query(const char *query_string)
query_string,
commandTag,
plantree_list,
+ NIL,
NULL);
/*
@@ -2033,6 +2034,7 @@ exec_bind_message(StringInfo input_message)
query_string,
psrc->commandTag,
cplan->stmt_list,
+ NIL,
cplan);
/* Portal is defined, set the plan ID based on its contents. */
diff --git a/src/backend/tcop/pquery.c b/src/backend/tcop/pquery.c
index fde78c55160..82c295502b0 100644
--- a/src/backend/tcop/pquery.c
+++ b/src/backend/tcop/pquery.c
@@ -37,6 +37,7 @@ Portal ActivePortal = NULL;
static void ProcessQuery(PlannedStmt *plan,
+ ExecPrep *prep,
const char *sourceText,
ParamListInfo params,
QueryEnvironment *queryEnv,
@@ -66,6 +67,7 @@ static void DoPortalRewind(Portal portal);
*/
QueryDesc *
CreateQueryDesc(PlannedStmt *plannedstmt,
+ ExecPrep *prep,
const char *sourceText,
Snapshot snapshot,
Snapshot crosscheck_snapshot,
@@ -78,6 +80,7 @@ CreateQueryDesc(PlannedStmt *plannedstmt,
qd->operation = plannedstmt->commandType; /* operation */
qd->plannedstmt = plannedstmt; /* plan */
+ qd->prep = prep; /* executor prep output */
qd->sourceText = sourceText; /* query text */
qd->snapshot = RegisterSnapshot(snapshot); /* snapshot */
/* RI check snapshot */
@@ -112,6 +115,13 @@ FreeQueryDesc(QueryDesc *qdesc)
UnregisterSnapshot(qdesc->snapshot);
UnregisterSnapshot(qdesc->crosscheck_snapshot);
+ /* ExecPrep cleanup if necessary */
+ if (qdesc->prep)
+ {
+ ExecPrepCleanup(qdesc->prep);
+ qdesc->prep = NULL;
+ }
+
/* Only the QueryDesc itself need be freed */
pfree(qdesc);
}
@@ -123,6 +133,7 @@ FreeQueryDesc(QueryDesc *qdesc)
* PORTAL_ONE_RETURNING, or PORTAL_ONE_MOD_WITH portal
*
* plan: the plan tree for the query
+ * prep: ExecPrep for the plan (output of ExecutorPrep())
* sourceText: the source text of the query
* params: any parameters needed
* dest: where to send results
@@ -135,6 +146,7 @@ FreeQueryDesc(QueryDesc *qdesc)
*/
static void
ProcessQuery(PlannedStmt *plan,
+ ExecPrep *prep,
const char *sourceText,
ParamListInfo params,
QueryEnvironment *queryEnv,
@@ -146,7 +158,7 @@ ProcessQuery(PlannedStmt *plan,
/*
* Create the QueryDesc object
*/
- queryDesc = CreateQueryDesc(plan, sourceText,
+ queryDesc = CreateQueryDesc(plan, prep, sourceText,
GetActiveSnapshot(), InvalidSnapshot,
dest, params, queryEnv, 0);
@@ -489,6 +501,9 @@ PortalStart(Portal portal, ParamListInfo params,
* the destination to DestNone.
*/
queryDesc = CreateQueryDesc(linitial_node(PlannedStmt, portal->stmts),
+ portal->preps ?
+ (ExecPrep *) linitial(portal->preps) :
+ NULL,
portal->sourceText,
GetActiveSnapshot(),
InvalidSnapshot,
@@ -1185,6 +1200,7 @@ PortalRunMulti(Portal portal,
{
bool active_snapshot_set = false;
ListCell *stmtlist_item;
+ int i;
/*
* If the destination is DestRemoteExecute, change to DestNone. The
@@ -1205,9 +1221,14 @@ PortalRunMulti(Portal portal,
* Loop to handle the individual queries generated from a single parsetree
* by analysis and rewrite.
*/
+ i = 0;
foreach(stmtlist_item, portal->stmts)
{
PlannedStmt *pstmt = lfirst_node(PlannedStmt, stmtlist_item);
+ ExecPrep *prep = portal->preps ?
+ list_nth(portal->preps, i) : NULL;
+
+ i++;
/*
* If we got a cancel signal in prior command, quit
@@ -1265,7 +1286,7 @@ PortalRunMulti(Portal portal,
if (pstmt->canSetTag)
{
/* statement can set tag string */
- ProcessQuery(pstmt,
+ ProcessQuery(pstmt, prep,
portal->sourceText,
portal->portalParams,
portal->queryEnv,
@@ -1274,7 +1295,7 @@ PortalRunMulti(Portal portal,
else
{
/* stmt added by rewrite cannot set tag */
- ProcessQuery(pstmt,
+ ProcessQuery(pstmt, prep,
portal->sourceText,
portal->portalParams,
portal->queryEnv,
diff --git a/src/backend/utils/mmgr/portalmem.c b/src/backend/utils/mmgr/portalmem.c
index 943da087c9f..313f8ef2fdc 100644
--- a/src/backend/utils/mmgr/portalmem.c
+++ b/src/backend/utils/mmgr/portalmem.c
@@ -284,6 +284,7 @@ PortalDefineQuery(Portal portal,
const char *sourceText,
CommandTag commandTag,
List *stmts,
+ List *preps,
CachedPlan *cplan)
{
Assert(PortalIsValid(portal));
@@ -298,6 +299,7 @@ PortalDefineQuery(Portal portal,
portal->qc.nprocessed = 0;
portal->commandTag = commandTag;
portal->stmts = stmts;
+ portal->preps = preps;
portal->cplan = cplan;
portal->status = PORTAL_DEFINED;
}
diff --git a/src/include/commands/explain.h b/src/include/commands/explain.h
index 6e51d50efc7..6aa8b275aa2 100644
--- a/src/include/commands/explain.h
+++ b/src/include/commands/explain.h
@@ -63,7 +63,8 @@ extern void ExplainOneUtility(Node *utilityStmt, IntoClause *into,
ExplainState *es, ParseState *pstate,
ParamListInfo params);
-extern void ExplainOnePlan(PlannedStmt *plannedstmt, IntoClause *into,
+extern void ExplainOnePlan(PlannedStmt *plannedstmt, ExecPrep *prep,
+ IntoClause *into,
ExplainState *es, const char *queryString,
ParamListInfo params, QueryEnvironment *queryEnv,
const instr_time *planduration,
diff --git a/src/include/executor/execdesc.h b/src/include/executor/execdesc.h
index 86db3dc8d0d..c18530f5d11 100644
--- a/src/include/executor/execdesc.h
+++ b/src/include/executor/execdesc.h
@@ -18,7 +18,6 @@
#include "nodes/execnodes.h"
#include "tcop/dest.h"
-
/* ----------------
* query descriptor:
*
@@ -35,6 +34,7 @@ typedef struct QueryDesc
/* These fields are provided by CreateQueryDesc */
CmdType operation; /* CMD_SELECT, CMD_UPDATE, etc. */
PlannedStmt *plannedstmt; /* planner's output (could be utility, too) */
+ ExecPrep *prep; /* output of ExecutorPrep() or NULL */
const char *sourceText; /* source text of the query */
Snapshot snapshot; /* snapshot to use for query */
Snapshot crosscheck_snapshot; /* crosscheck for RI update/delete */
@@ -57,6 +57,7 @@ typedef struct QueryDesc
/* in pquery.c */
extern QueryDesc *CreateQueryDesc(PlannedStmt *plannedstmt,
+ ExecPrep *prep,
const char *sourceText,
Snapshot snapshot,
Snapshot crosscheck_snapshot,
diff --git a/src/include/executor/executor.h b/src/include/executor/executor.h
index fa2b657fb2f..3579926d4e8 100644
--- a/src/include/executor/executor.h
+++ b/src/include/executor/executor.h
@@ -20,6 +20,7 @@
#include "nodes/lockoptions.h"
#include "nodes/parsenodes.h"
#include "utils/memutils.h"
+#include "utils/resowner.h"
/*
@@ -234,6 +235,16 @@ ExecGetJunkAttribute(TupleTableSlot *slot, AttrNumber attno, bool *isNull)
*/
extern void ExecutorStart(QueryDesc *queryDesc, int eflags);
extern void standard_ExecutorStart(QueryDesc *queryDesc, int eflags);
+
+extern ExecPrep *ExecutorPrep(PlannedStmt *pstmt,
+ ParamListInfo params,
+ ResourceOwner owner,
+ bool do_initial_pruning,
+ int eflags);
+extern ExecPrep *CreateExecPrep(EState *estate, MemoryContext context,
+ execprep_cleanup_fn cleanup, void *cleanup_arg);
+extern void ExecPrepCleanup(ExecPrep *prep);
+
extern void ExecutorRun(QueryDesc *queryDesc,
ScanDirection direction, uint64 count);
extern void standard_ExecutorRun(QueryDesc *queryDesc,
diff --git a/src/include/nodes/execnodes.h b/src/include/nodes/execnodes.h
index 18ae8f0d4bb..8bdecd631bf 100644
--- a/src/include/nodes/execnodes.h
+++ b/src/include/nodes/execnodes.h
@@ -772,6 +772,54 @@ typedef struct EState
List *es_insert_pending_modifytables;
} EState;
+/*
+ * ExecPrep: encapsulates executor preparation results for a PlannedStmt.
+ *
+ * ExecutorPrep() factors out executor setup steps such as initializing the
+ * range table, checking permissions, and executing initial partition pruning.
+ * ExecutorStart() can reuse the prepared EState instead of repeating that
+ * work, and other callers (such as plan cache validation) can use it without
+ * running the full plan.
+ */
+
+/*
+ * Optional callback to clean up user-specific resources associated with
+ * ExecPrep.
+ */
+typedef void (*execprep_cleanup_fn)(void *prep);
+
+typedef struct ExecPrep
+{
+ /*
+ * Context in which this struct and all subsidiary allocations were made.
+ * This context must remain alive until ExecPrepCleanup is called.
+ */
+ MemoryContext context;
+
+ /*
+ * Partially-initialized executor state used for permission checks and
+ * pruning. May be adopted directly by ExecutorStart(), in which case
+ * ExecPrepCleanup will skip freeing it.
+ */
+ EState *prep_estate;
+
+ /*
+ * True if ExecPrepCleanup() must free the EState. If the executor adopts
+ * prep_estate, this is set to false to avoid double-free.
+ */
+ bool owns_estate;
+
+ /*
+ * Optional caller-supplied cleanup hook to run during ExecPrepCleanup.
+ * Useful for releasing external resources associated with the prep.
+ */
+ execprep_cleanup_fn cleanup;
+
+ /*
+ * Opaque pointer to pass to the cleanup hook.
+ */
+ void *cleanup_arg;
+} ExecPrep;
/*
* ExecRowMark -
diff --git a/src/include/utils/portal.h b/src/include/utils/portal.h
index 5ffa6fd5cc8..013bcc3bd8e 100644
--- a/src/include/utils/portal.h
+++ b/src/include/utils/portal.h
@@ -137,6 +137,7 @@ typedef struct PortalData
CommandTag commandTag; /* command tag for original query */
QueryCompletion qc; /* command completion data for executed query */
List *stmts; /* list of PlannedStmts */
+ List *preps; /* list of ExecPreps where needed */
CachedPlan *cplan; /* CachedPlan, if stmts are from one */
ParamListInfo portalParams; /* params to pass to query */
@@ -240,6 +241,7 @@ extern void PortalDefineQuery(Portal portal,
const char *sourceText,
CommandTag commandTag,
List *stmts,
+ List *preps,
CachedPlan *cplan);
extern PlannedStmt *PortalGetPrimaryStmt(Portal portal);
extern void PortalCreateHoldStore(Portal portal);
--
2.47.3
view thread (114+ messages) latest in thread
reply
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: [email protected]
Cc: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
Subject: Re: generic plans and "initial" pruning
In-Reply-To: <CA+HiwqFpEHBjosRackQhm6yKKnHgqm8Ewkn=qsctT1N0PqVSrg@mail.gmail.com>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox