Allow subquery pullup to wrap a PlaceHolderVar in another one.

tglsfdc · tglsfdc · commit 80bece312c4b · 2024-01-11T15:28:13.000-05:00
The code for wrapping subquery output expressions in PlaceHolderVars believed that if the expression already was a PlaceHolderVar, it was never necessary to wrap that in another one. That's wrong if the expression is underneath an outer join and involves a lateral reference to outside that scope: failing to add an additional PHV risks evaluating the expression at the wrong place and hence not forcing it to null when the outer join should do so. This is an oversight in commit 9e7e29c, which added logic to forcibly wrap lateral-reference Vars in PlaceHolderVars, but didn't see that the adjacent case for PlaceHolderVars needed the same treatment. The test case we have for this doesn't fail before 4be058f, but now that I see the problem I wonder if it is possible to demonstrate related errors before that. That's moot though, since all such branches are out of support. Per bug #18284 from Holger Reise. Back-patch to all supported branches. Discussion: https://postgr.es/m/18284-47505a20c23647f8@postgresql.org
diff --git a/src/backend/optimizer/prep/prepjointree.c b/src/backend/optimizer/prep/prepjointree.c
@@ -2435,8 +2435,13 @@ pullup_replace_vars_callback(Var *var,
 			else if (newnode && IsA(newnode, PlaceHolderVar) &&
 					 ((PlaceHolderVar *) newnode)->phlevelsup == 0)
 			{
-				/* No need to wrap a PlaceHolderVar with another one, either */
-				wrap = false;
+				/* The same rules apply for a PlaceHolderVar */
+				if (rcon->target_rte->lateral &&
+					!bms_is_subset(((PlaceHolderVar *) newnode)->phrels,
+								   rcon->relids))
+					wrap = true;
+				else
+					wrap = false;
 			}
 			else
 			{
diff --git a/src/test/regress/expected/join.out b/src/test/regress/expected/join.out
@@ -7013,6 +7013,33 @@ select * from
          Output: (COALESCE((COALESCE(b.q2, '42'::bigint)), d.q2))
 (24 rows)
 
+-- another case requiring nested PlaceHolderVars
+explain (verbose, costs off)
+select * from
+  (select 0 as val0) as ss0
+  left join (select 1 as val) as ss1 on true
+  left join lateral (select ss1.val as val_filtered where false) as ss2 on true;
+           QUERY PLAN           
+--------------------------------
+ Nested Loop Left Join
+   Output: 0, (1), ((1))
+   Join Filter: false
+   ->  Result
+         Output: 1
+   ->  Result
+         Output: (1)
+         One-Time Filter: false
+(8 rows)
+
+select * from
+  (select 0 as val0) as ss0
+  left join (select 1 as val) as ss1 on true
+  left join lateral (select ss1.val as val_filtered where false) as ss2 on true;
+ val0 | val | val_filtered 
+------+-----+--------------
+    0 |   1 |             
+(1 row)
+
 -- case that breaks the old ph_may_need optimization
 explain (verbose, costs off)
 select c.*,a.*,ss1.q1,ss2.q1,ss3.* from
diff --git a/src/test/regress/sql/join.sql b/src/test/regress/sql/join.sql
@@ -2522,6 +2522,18 @@ select * from
   ) on c.q2 = ss2.q1,
   lateral (select ss2.y offset 0) ss3;
 
+-- another case requiring nested PlaceHolderVars
+explain (verbose, costs off)
+select * from
+  (select 0 as val0) as ss0
+  left join (select 1 as val) as ss1 on true
+  left join lateral (select ss1.val as val_filtered where false) as ss2 on true;
+
+select * from
+  (select 0 as val0) as ss0
+  left join (select 1 as val) as ss1 on true
+  left join lateral (select ss1.val as val_filtered where false) as ss2 on true;
+
 -- case that breaks the old ph_may_need optimization
 explain (verbose, costs off)
 select c.*,a.*,ss1.q1,ss2.q1,ss3.* from

Original file line number	Diff line number	Diff line change
`@@ -2435,8 +2435,13 @@ pullup_replace_vars_callback(Var *var,`
`2435`	`2435`	`else if (newnode && IsA(newnode, PlaceHolderVar) &&`
`2436`	`2436`	`((PlaceHolderVar *) newnode)->phlevelsup == 0)`
`2437`	`2437`	`{`
`2438`		`- /* No need to wrap a PlaceHolderVar with another one, either */`
`2439`		`- wrap = false;`
	`2438`	`+ /* The same rules apply for a PlaceHolderVar */`
	`2439`	`+ if (rcon->target_rte->lateral &&`
	`2440`	`+ !bms_is_subset(((PlaceHolderVar *) newnode)->phrels,`
	`2441`	`+ rcon->relids))`
	`2442`	`+ wrap = true;`
	`2443`	`+ else`
	`2444`	`+ wrap = false;`
`2440`	`2445`	`}`
`2441`	`2446`	`else`
`2442`	`2447`	`{`