v34-0009-Don-t-copy-PartitionBoundInfo-in-set_relation_pa.patch

From 9c9d2f973152277ca334f945cb27b7a065191260 Mon Sep 17 00:00:00 2001
From: amit <amitlangote09@gmail.com>
Date: Mon, 4 Mar 2019 16:03:49 +0900
Subject: [PATCH v34 9/9] Don't copy PartitionBoundInfo in
 set_relation_partition_info

As long as we hold a lock on the table, it shouldn't change, so
seems pointless to copy, which can get really expensive as the
number of partitions grows beyond thousands.
---
 src/backend/optimizer/util/plancat.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/src/backend/optimizer/util/plancat.c b/src/backend/optimizer/util/plancat.c
index 2b22dff690..b8f01269be 100644
--- a/src/backend/optimizer/util/plancat.c
+++ b/src/backend/optimizer/util/plancat.c
@@ -2082,13 +2082,11 @@ set_relation_partition_info(PlannerInfo *root, RelOptInfo *rel,
 							Relation relation)
 {
 	PartitionDesc partdesc;
-	PartitionKey partkey;
 
 	Assert(relation->rd_rel->relkind == RELKIND_PARTITIONED_TABLE);
 
 	partdesc = PartitionDirectoryLookup(root->glob->partition_directory,
 										relation);
-	partkey = RelationGetPartitionKey(relation);
 	rel->part_scheme = find_partition_scheme(root, relation);
 	Assert(partdesc != NULL && rel->part_scheme != NULL);
 	/*
@@ -2098,7 +2096,13 @@ set_relation_partition_info(PlannerInfo *root, RelOptInfo *rel,
 	 * only lazily updated as far as dropped partitions are concerned.
 	 */
 	if (partdesc->nparts > 0)
-		rel->boundinfo = partition_bounds_copy(partdesc->boundinfo, partkey);
+	{
+		/*
+		 * Holding onto the relcache pointer should be OK, as it won't change
+		 * under us as long as we're holding the lock on the relation.
+		 */
+		rel->boundinfo = partdesc->boundinfo;
+	}
 	rel->nparts = partdesc->nparts;
 	set_baserel_partition_key_exprs(relation, rel);
 	rel->partition_qual = RelationGetPartitionQual(relation);
-- 
2.11.0