Fix pull_varnos' miscomputation of relids set for a PlaceHolderVar.

Previously, pull_varnos() took the relids of a PlaceHolderVar as being
equal to the relids in its contents, but that fails to account for the
possibility that we have to postpone evaluation of the PHV due to outer
joins.  This could result in a malformed plan.  The known cases end up
triggering the "failed to assign all NestLoopParams to plan nodes"
sanity check in createplan.c, but other symptoms may be possible.

The right value to use is the join level we actually intend to evaluate
the PHV at.  We can get that from the ph_eval_at field of the associated
PlaceHolderInfo.  However, there are some places that call pull_varnos()
before the PlaceHolderInfos have been created; in that case, fall back
to the conservative assumption that the PHV will be evaluated at its
syntactic level.  (In principle this might result in missing some legal
optimization, but I'm not aware of any cases where it's an issue in
practice.)  Things are also a bit ticklish for calls occurring during
deconstruct_jointree(), but AFAICS the ph_eval_at fields should have
reached their final values by the time we need them.

The main problem in making this work is that pull_varnos() has no
way to get at the PlaceHolderInfos.  We can fix that easily, if a
bit tediously, in HEAD by passing it the planner "root" pointer.
In the back branches that'd cause an unacceptable API/ABI break for
extensions, so leave the existing entry points alone and add new ones
with the additional parameter.  (If an old entry point is called and
encounters a PHV, it'll fall back to using the syntactic level,
again possibly missing some valid optimization.)

Back-patch to v12.  The computation is surely also wrong before that,
but it appears that we cannot reach a bad plan thanks to join order
restrictions imposed on the subquery that the PlaceHolderVar came from.
The error only became reachable when commit 4be058fe9 allowed trivial
subqueries to be collapsed out completely, eliminating their join order
restrictions.

Per report from Stephan Springl.

Discussion: https://postgr.es/m/171041.1610849523@sss.pgh.pa.us

1 files changed, 9 insertions, 9 deletions

diff --git a/src/backend/optimizer/path/tidpath.c b/src/backend/optimizer/path/tidpath.c
index 8ef04060570..0845b460e2c 100644
--- a/src/backend/optimizer/path/tidpath.c
+++ b/src/backend/optimizer/path/tidpath.c
@@ -123,7 +123,7 @@ IsTidEqualClause(RestrictInfo *rinfo, RelOptInfo *rel)
  * other side of the clause does.
  */
 static bool
-IsTidEqualAnyClause(RestrictInfo *rinfo, RelOptInfo *rel)
+IsTidEqualAnyClause(PlannerInfo *root, RestrictInfo *rinfo, RelOptInfo *rel)
 {
 	ScalarArrayOpExpr *node;
 	Node	   *arg1,
@@ -148,7 +148,7 @@ IsTidEqualAnyClause(RestrictInfo *rinfo, RelOptInfo *rel)
 		IsCTIDVar((Var *) arg1, rel))
 	{
 		/* The other argument must be a pseudoconstant */
-		if (bms_is_member(rel->relid, pull_varnos(arg2)) ||
+		if (bms_is_member(rel->relid, pull_varnos(root, arg2)) ||
 			contain_volatile_functions(arg2))
 			return false;
 
@@ -190,7 +190,7 @@ IsCurrentOfClause(RestrictInfo *rinfo, RelOptInfo *rel)
  * (Using a List may seem a bit weird, but it simplifies the caller.)
  */
 static List *
-TidQualFromRestrictInfo(RestrictInfo *rinfo, RelOptInfo *rel)
+TidQualFromRestrictInfo(PlannerInfo *root, RestrictInfo *rinfo, RelOptInfo *rel)
 {
 	/*
 	 * We may ignore pseudoconstant clauses (they can't contain Vars, so could
@@ -210,7 +210,7 @@ TidQualFromRestrictInfo(RestrictInfo *rinfo, RelOptInfo *rel)
 	 * Check all base cases.  If we get a match, return the clause.
 	 */
 	if (IsTidEqualClause(rinfo, rel) ||
-		IsTidEqualAnyClause(rinfo, rel) ||
+		IsTidEqualAnyClause(root, rinfo, rel) ||
 		IsCurrentOfClause(rinfo, rel))
 		return list_make1(rinfo);
 
@@ -227,7 +227,7 @@ TidQualFromRestrictInfo(RestrictInfo *rinfo, RelOptInfo *rel)
  * This function is just concerned with handling AND/OR recursion.
  */
 static List *
-TidQualFromRestrictInfoList(List *rlist, RelOptInfo *rel)
+TidQualFromRestrictInfoList(PlannerInfo *root, List *rlist, RelOptInfo *rel)
 {
 	List	   *rlst = NIL;
 	ListCell   *l;
@@ -255,14 +255,14 @@ TidQualFromRestrictInfoList(List *rlist, RelOptInfo *rel)
 					List	   *andargs = ((BoolExpr *) orarg)->args;
 
 					/* Recurse in case there are sub-ORs */
-					sublist = TidQualFromRestrictInfoList(andargs, rel);
+					sublist = TidQualFromRestrictInfoList(root, andargs, rel);
 				}
 				else
 				{
 					RestrictInfo *rinfo = castNode(RestrictInfo, orarg);
 
 					Assert(!restriction_is_or_clause(rinfo));
-					sublist = TidQualFromRestrictInfo(rinfo, rel);
+					sublist = TidQualFromRestrictInfo(root, rinfo, rel);
 				}
 
 				/*
@@ -284,7 +284,7 @@ TidQualFromRestrictInfoList(List *rlist, RelOptInfo *rel)
 		else
 		{
 			/* Not an OR clause, so handle base cases */
-			rlst = TidQualFromRestrictInfo(rinfo, rel);
+			rlst = TidQualFromRestrictInfo(root, rinfo, rel);
 		}
 
 		/*
@@ -390,7 +390,7 @@ create_tidscan_paths(PlannerInfo *root, RelOptInfo *rel)
 	 * If any suitable quals exist in the rel's baserestrict list, generate a
 	 * plain (unparameterized) TidPath with them.
 	 */
-	tidquals = TidQualFromRestrictInfoList(rel->baserestrictinfo, rel);
+	tidquals = TidQualFromRestrictInfoList(root, rel->baserestrictinfo, rel);
 
 	if (tidquals)
 	{