Properly set function column count for table value functions in Orca

The funccolcount value wasn't being set by Orca when populating the plan to execute, which caused running explain <qry> for such queries to fail with "invalid attnum 2 for relation". Interestingly, the queries I tested would execute fine--only explain would fail.
mperezfuster · Feb 16, 2024 · 456f6cf · 456f6cf
1 parent acdb2c1
commit 456f6cf
Show file tree

Hide file tree

Showing 4 changed files with 97 additions and 1 deletion.
diff --git a/src/backend/gpopt/translate/CTranslatorDXLToPlStmt.cpp b/src/backend/gpopt/translate/CTranslatorDXLToPlStmt.cpp
@@ -1683,6 +1683,7 @@ CTranslatorDXLToPlStmt::TranslateDXLTvf(
 	rtfunc->funccoltypes = NIL;
 	rtfunc->funccoltypmods = NIL;
 	rtfunc->funccolcollations = NIL;
+	rtfunc->funccolcount = gpdb::ListLength(target_list);
 	ForEach(lc_target_entry, target_list)
 	{
 		TargetEntry *target_entry = (TargetEntry *) lfirst(lc_target_entry);
@@ -1785,7 +1786,6 @@ CTranslatorDXLToPlStmt::TranslateDXLTvfToRangeTblEntry(
 		const_expr->constvalue = gpdb::DatumFromPointer(str);
 
 		rtfunc->funcexpr = (Node *) const_expr;
-		rtfunc->funccolcount = (int) num_of_cols;
 	}
 	else
 	{
@@ -1832,6 +1832,7 @@ CTranslatorDXLToPlStmt::TranslateDXLTvfToRangeTblEntry(
 		rtfunc->funcexpr = (Node *) func_expr;
 	}
 
+	rtfunc->funccolcount = (int) num_of_cols;
 	rtfunc->funcparams = funcparams;
 	// GPDB_91_MERGE_FIXME: collation
 	// set rtfunc->funccoltypemods & rtfunc->funccolcollations?

diff --git a/src/test/regress/expected/gporca.out b/src/test/regress/expected/gporca.out
@@ -14921,5 +14921,45 @@ set session authorization ruser;
 set gp_max_system_slices=10;
 ERROR:  permission denied to set parameter "gp_max_system_slices"
 reset session authorization;
+-- Test that set returning function with multiple columns works with explain
+CREATE FUNCTION srf_attnum() RETURNS TABLE(v1 int, v2 int)
+    LANGUAGE plpgsql NO SQL
+    AS $_$
+BEGIN
+    DROP TABLE IF EXISTS tbl_2_cols;
+    CREATE TEMP TABLE tbl_2_cols (col1 int, col2 int) DISTRIBUTED RANDOMLY;
+    RETURN QUERY SELECT * from tbl_2_cols;
+END;
+$_$;
+NOTICE:  specifying "NO SQL" acts as no operation.
+explain select distinct v1 from srf_attnum();
+                               QUERY PLAN                                
+-------------------------------------------------------------------------
+ HashAggregate  (cost=12.75..22.75 rows=1000 width=4)
+   Group Key: v1
+   ->  Function Scan on srf_attnum  (cost=0.25..10.25 rows=1000 width=4)
+ Optimizer: Postgres-based planner
+(4 rows)
+
+select distinct v1 from srf_attnum();
+NOTICE:  table "tbl_2_cols" does not exist, skipping
+ v1 
+----
+(0 rows)
+
+explain select distinct v2 from srf_attnum();
+                               QUERY PLAN                                
+-------------------------------------------------------------------------
+ HashAggregate  (cost=12.75..22.75 rows=1000 width=4)
+   Group Key: v2
+   ->  Function Scan on srf_attnum  (cost=0.25..10.25 rows=1000 width=4)
+ Optimizer: Postgres-based planner
+(4 rows)
+
+select distinct v2 from srf_attnum();
+ v2 
+----
+(0 rows)
+
 drop user ruser;
 drop table foo, bar;
diff --git a/src/test/regress/expected/gporca_optimizer.out b/src/test/regress/expected/gporca_optimizer.out
@@ -15016,5 +15016,45 @@ set session authorization ruser;
 set gp_max_system_slices=10;
 ERROR:  permission denied to set parameter "gp_max_system_slices"
 reset session authorization;
+-- Test that set returning function with multiple columns works with explain
+CREATE FUNCTION srf_attnum() RETURNS TABLE(v1 int, v2 int)
+    LANGUAGE plpgsql NO SQL
+    AS $_$
+BEGIN
+    DROP TABLE IF EXISTS tbl_2_cols;
+    CREATE TEMP TABLE tbl_2_cols (col1 int, col2 int) DISTRIBUTED RANDOMLY;
+    RETURN QUERY SELECT * from tbl_2_cols;
+END;
+$_$;
+NOTICE:  specifying "NO SQL" acts as no operation.
+explain select distinct v1 from srf_attnum();
+                               QUERY PLAN                               
+------------------------------------------------------------------------
+ HashAggregate  (cost=0.00..0.13 rows=1000 width=4)
+   Group Key: v1
+   ->  Function Scan on srf_attnum  (cost=0.00..0.00 rows=1000 width=4)
+ Optimizer: GPORCA
+(4 rows)
+
+select distinct v1 from srf_attnum();
+NOTICE:  table "tbl_2_cols" does not exist, skipping
+ v1 
+----
+(0 rows)
+
+explain select distinct v2 from srf_attnum();
+                               QUERY PLAN                               
+------------------------------------------------------------------------
+ HashAggregate  (cost=0.00..0.13 rows=1000 width=4)
+   Group Key: v2
+   ->  Function Scan on srf_attnum  (cost=0.00..0.00 rows=1000 width=4)
+ Optimizer: GPORCA
+(4 rows)
+
+select distinct v2 from srf_attnum();
+ v2 
+----
+(0 rows)
+
 drop user ruser;
 drop table foo, bar;
diff --git a/src/test/regress/sql/gporca.sql b/src/test/regress/sql/gporca.sql
@@ -3688,6 +3688,21 @@ set session authorization ruser;
 set gp_max_system_slices=10;
 reset session authorization;
 
+-- Test that set returning function with multiple columns works with explain
+CREATE FUNCTION srf_attnum() RETURNS TABLE(v1 int, v2 int)
+    LANGUAGE plpgsql NO SQL
+    AS $_$
+BEGIN
+    DROP TABLE IF EXISTS tbl_2_cols;
+    CREATE TEMP TABLE tbl_2_cols (col1 int, col2 int) DISTRIBUTED RANDOMLY;
+    RETURN QUERY SELECT * from tbl_2_cols;
+END;
+$_$;
+explain select distinct v1 from srf_attnum();
+select distinct v1 from srf_attnum();
+explain select distinct v2 from srf_attnum();
+select distinct v2 from srf_attnum();
+
 drop user ruser;
 drop table foo, bar;