Use extended statistics to estimate (Var op Var) clauses

From 6dcd4f2fc78be2fca8a4e934fc2e24bcd8340c4c Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Sun, 13 Jun 2021 21:59:16 +0200
Subject: [PATCH] Handling Expr op Expr clauses in extended stats

---
 src/backend/optimizer/path/clausesel.c        |  46 ++++-
 src/backend/statistics/extended_stats.c       |  81 +++++++--
 src/backend/statistics/mcv.c                  | 169 +++++++++++++-----
 .../statistics/extended_stats_internal.h      |   2 +-
 src/test/regress/expected/stats_ext.out       |  96 ++++++++++
 src/test/regress/sql/stats_ext.sql            |  26 +++
 6 files changed, 350 insertions(+), 70 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index d263ecf082..d732a9dc93 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,46 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be (Expr op Expr), which goes here thanks to the
+			 * optimization at the beginning of clauselist_selectivity.
+			 * So try applying extended stats first, then fall back to
+			 * restriction_selectivity.
+			 *
+			 * XXX Kinda does the same thing as clauselist_selectivity, but
+			 * for a single clause. Maybe we could call that, but need to
+			 * be careful not to cause infinite loop.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				Bitmapset *estimatedclauses = NULL;
+				List *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					/*
+					 * Estimate as many clauses as possible using extended statistics.
+					 *
+					 * 'estimatedclauses' is populated with the 0-based list position
+					 * index of clauses estimated here, and that should be ignored below.
+					 */
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index b05e818ba9..59a9a64b09 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1352,14 +1352,15 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		Node	   *clause_expr,
+				   *clause_expr2;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_expr, &clause_expr2, NULL, NULL))
 			return false;
 
 		/*
@@ -1399,13 +1400,44 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
+		/* clause_expr is always valid */
+		Assert(clause_expr);
+
 		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
 		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		{
+			if (!statext_is_compatible_clause_internal(root, clause_expr,
+													   relid, attnums, exprs))
+				return false;
+		}
+		else
+		{
+			/* Otherwise we have (Expr op Const) or (Const op Expr) or (Expr op Expr). */
+			*exprs = lappend(*exprs, clause_expr);
+		}
+
+		/* without second expression, it has to be Expr op Const, so we're done */
+		if (!clause_expr2)
+			return true;
+
+		/*
+		 * Do the same thing as for the first expression.
+		 *
+		 * XXX Need to be careful later about matching both expressions at
+		 * the same time.
+		 */
+		if (IsA(clause_expr2, Var))
+		{
+			if (!statext_is_compatible_clause_internal(root, clause_expr2,
+													   relid, attnums, exprs))
+				return false;
+		}
+		else
+		{
+			/* Otherwise we have (Expr op Const) or (Const op Expr). */
+			*exprs = lappend(*exprs, clause_expr2);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1421,7 +1453,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL, NULL))
 			return false;
 
 		/*
@@ -2010,19 +2042,20 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  *		Split an operator expression's arguments into Expr and Const parts.
  *
  * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Expr) or (Expr op Expr), possibly with a RelabelType on top. When the
+ * expression matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
  * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
+examine_opclause_args(List *args, Node **expr1p, Node **expr2p, Const **cstp,
 					  bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	Node	   *expr1 = NULL;
+	Node	   *expr2 = NULL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2042,22 +2075,36 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		expr1 = (Node *) leftop;
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		expr1 = (Node *) rightop;
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		expr1 = (Node *) leftop;
+		expr2 = (Node *) rightop;
+		expronleft = false;
+
+		/*
+		 * FIXME Both variables have to be for the same relation (otherwise
+		 * it's a join clause, and we don't deal with those yet. Need to
+		 * call pull_varnos on both sides or something like that.
+		 */
+		
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (expr1p)
+		*expr1p = expr1;
+
+	if (expr2p)
+		*expr2p = expr2;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index ef118952c7..59538f804e 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1645,78 +1645,153 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_expr, &clause_expr2, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
-
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
-
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
+			if (cst)	/* Expr op Const */
 			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				int idx;
 
-				Assert(idx >= 0);
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * First check whether the constant is below the lower
+					 * boundary (in that case we can skip the bucket, because
+					 * there's no overlap).
+					 *
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * First check whether the constant is below the lower
+					 * boundary (in that case we can skip the bucket, because
+					 * there's no overlap).
+					 *
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1743,7 +1818,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_expr, NULL, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 55cd9252a5..52b842bbe5 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,7 +99,7 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
+extern bool examine_opclause_args(List *args, Node **expr1p, Node **expr2p,
 								  Const **cstp, bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 8c214d8dfc..84b93299de 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1900,6 +1900,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2047,6 +2059,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      3750 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2456,6 +2480,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 ANALYZE mcv_lists;
@@ -2489,6 +2519,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       2500 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -2583,6 +2619,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
       1094 |      0
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 ANALYZE mcv_lists_bool;
@@ -2738,6 +2786,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 (1 row)
 
 DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2797,6 +2857,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2843,6 +2921,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index e033080d4f..3ace3faa09 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -946,6 +946,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -999,6 +1003,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1198,6 +1206,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1216,6 +1226,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1291,6 +1303,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1376,6 +1392,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1403,6 +1423,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
@@ -1417,6 +1440,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 DROP TABLE mcv_lists_multi;
 
-- 
2.31.1

From 788a909e09b372797c4b7b443e0e89c5d5181ec0 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas@2ndquadrant.com>
Date: Tue, 20 Jul 2021 20:15:13 +0200
Subject: [PATCH] Handling Expr op Expr clauses in extended stats

---
 src/backend/optimizer/path/clausesel.c        |  37 +++-
 src/backend/statistics/extended_stats.c       |  83 ++++++---
 src/backend/statistics/mcv.c                  | 172 +++++++++++++-----
 .../statistics/extended_stats_internal.h      |   4 +-
 src/test/regress/expected/stats_ext.out       |  96 ++++++++++
 src/test/regress/sql/stats_ext.sql            |  26 +++
 6 files changed, 341 insertions(+), 77 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index d263ecf082..6a7e9ceea5 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,37 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be a single (Expr op Expr) clause, which goes here due
+			 * to the optimization at the beginning of clauselist_selectivity.
+			 * So we try applying extended stats first, and then fall back to
+			 * restriction_selectivity.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				List	   *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					Bitmapset  *estimatedclauses = NULL;
+
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 2e55913bc8..606cf8c588 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1347,19 +1347,27 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		return true;
 	}
 
-	/* (Var/Expr op Const) or (Const op Var/Expr) */
+	/*
+	 * Three opclause variants are supported: (Expr op Const), (Const op Expr),
+	 * (Expr op Expr). That means we may need to analyze one or two expressions
+	 * to make sure the opclause is compatible with extended stats.
+	 */
 	if (is_opclause(clause))
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		ListCell   *lc;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
-		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		/*
+		 * Check if the expression has the right shape. This returns either one
+		 * or two expressions, depending on whether there is a Const.
+		 */
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
 		/*
@@ -1399,13 +1407,31 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
-		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
-		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		/*
+		 * There's always at least one expression, otherwise the clause would
+		 * not be considered compatible.
+		 */
+		Assert(list_length(clause_exprs) >= 1);
+
+		/*
+		 * Check all expressions by recursing. Var expressions are handled as
+		 * a special case (to match it to attnums etc.)
+		 */
+		foreach (lc, clause_exprs)
+		{
+			Node *clause_expr = (Node *) lfirst(lc);
+
+			if (IsA(clause_expr, Var))
+			{
+				/* if the Var is incompatible, the whole clause is incompatible */
+				if (!statext_is_compatible_clause_internal(root, clause_expr,
+														   relid, attnums, exprs))
+					return false;
+			}
+			else	/* generic expression */
+				*exprs = lappend(*exprs, clause_expr);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1415,15 +1441,21 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		ScalarArrayOpExpr *expr = (ScalarArrayOpExpr *) clause;
 		Node	   *clause_expr;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
+		/* There has to be one expression exactly. */
+		Assert(list_length(clause_exprs) == 1);
+
+		clause_expr = (Node *) linitial(clause_exprs);
+
 		/*
 		 * If it's not one of the supported operators ("=", "<", ">", etc.),
 		 * just ignore the clause, as it's not compatible with MCV lists.
@@ -2009,20 +2041,19 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * examine_opclause_args
  *		Split an operator expression's arguments into Expr and Const parts.
  *
- * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Attempts to match the arguments to either (Expr op Const) or (Const op Expr)
+ * or (Expr op Expr), possibly with a RelabelType on top. When the expression
+ * matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
- * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
+ * non-null pointers (exprsp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
-					  bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	List	   *exprs = NIL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2042,22 +2073,26 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		exprs = lappend(exprs, leftop);
+		exprs = lappend(exprs, rightop);
+		expronleft = false;
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (exprsp)
+		*exprsp = exprs;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index ef118952c7..85f650f572 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1645,78 +1645,154 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+			if (cst)	/* Expr op Const */
+			{
+				int idx;
 
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert(list_length(clause_exprs) == 1);
+				clause_expr = (Node *) linitial(clause_exprs);
 
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
-			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
 
-				Assert(idx >= 0);
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(list_length(clause_exprs) == 2);
+
+				clause_expr = (Node *) linitial(clause_exprs);
+				clause_expr2 = (Node *) lsecond(clause_exprs);
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1726,6 +1802,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
 			Oid			collid;
@@ -1743,11 +1820,14 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
 			Assert(expronleft);
+			Assert(list_length(clause_exprs) == 1);
+
+			clause_expr = (Node *) linitial(clause_exprs);
 
 			/* XXX what if (cst->constisnull == NULL)? */
 			if (!cst->constisnull)
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 55cd9252a5..1f30fa9060 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,8 +99,8 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
-								  Const **cstp, bool *expronleftp);
+extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
+								  bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 62b05c79f9..93e4edbf3e 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1904,6 +1904,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2051,6 +2063,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      3750 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2460,6 +2484,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 ANALYZE mcv_lists;
@@ -2493,6 +2523,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       2500 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -2587,6 +2623,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
       1094 |      0
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 ANALYZE mcv_lists_bool;
@@ -2742,6 +2790,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 (1 row)
 
 DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2801,6 +2861,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2847,6 +2925,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index e033080d4f..3ace3faa09 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -946,6 +946,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -999,6 +1003,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1198,6 +1206,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1216,6 +1226,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1291,6 +1303,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1376,6 +1392,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1403,6 +1423,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
@@ -1417,6 +1440,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 DROP TABLE mcv_lists_multi;
 
-- 
2.31.1

From 4abb267dfdd46e5c2157ff6d201c11d068331519 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Sun, 15 Aug 2021 13:23:13 +0200
Subject: [PATCH 1/2] patch 20210720

---
 src/backend/optimizer/path/clausesel.c        |  37 +++-
 src/backend/statistics/extended_stats.c       |  83 ++++++---
 src/backend/statistics/mcv.c                  | 172 +++++++++++++-----
 .../statistics/extended_stats_internal.h      |   4 +-
 src/test/regress/expected/stats_ext.out       |  96 ++++++++++
 src/test/regress/sql/stats_ext.sql            |  26 +++
 6 files changed, 341 insertions(+), 77 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index d263ecf082..6a7e9ceea5 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,37 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be a single (Expr op Expr) clause, which goes here due
+			 * to the optimization at the beginning of clauselist_selectivity.
+			 * So we try applying extended stats first, and then fall back to
+			 * restriction_selectivity.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				List	   *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					Bitmapset  *estimatedclauses = NULL;
+
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 2e55913bc8..606cf8c588 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1347,19 +1347,27 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		return true;
 	}
 
-	/* (Var/Expr op Const) or (Const op Var/Expr) */
+	/*
+	 * Three opclause variants are supported: (Expr op Const), (Const op Expr),
+	 * (Expr op Expr). That means we may need to analyze one or two expressions
+	 * to make sure the opclause is compatible with extended stats.
+	 */
 	if (is_opclause(clause))
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		ListCell   *lc;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
-		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		/*
+		 * Check if the expression has the right shape. This returns either one
+		 * or two expressions, depending on whether there is a Const.
+		 */
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
 		/*
@@ -1399,13 +1407,31 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
-		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
-		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		/*
+		 * There's always at least one expression, otherwise the clause would
+		 * not be considered compatible.
+		 */
+		Assert(list_length(clause_exprs) >= 1);
+
+		/*
+		 * Check all expressions by recursing. Var expressions are handled as
+		 * a special case (to match it to attnums etc.)
+		 */
+		foreach (lc, clause_exprs)
+		{
+			Node *clause_expr = (Node *) lfirst(lc);
+
+			if (IsA(clause_expr, Var))
+			{
+				/* if the Var is incompatible, the whole clause is incompatible */
+				if (!statext_is_compatible_clause_internal(root, clause_expr,
+														   relid, attnums, exprs))
+					return false;
+			}
+			else	/* generic expression */
+				*exprs = lappend(*exprs, clause_expr);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1415,15 +1441,21 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		ScalarArrayOpExpr *expr = (ScalarArrayOpExpr *) clause;
 		Node	   *clause_expr;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
+		/* There has to be one expression exactly. */
+		Assert(list_length(clause_exprs) == 1);
+
+		clause_expr = (Node *) linitial(clause_exprs);
+
 		/*
 		 * If it's not one of the supported operators ("=", "<", ">", etc.),
 		 * just ignore the clause, as it's not compatible with MCV lists.
@@ -2009,20 +2041,19 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * examine_opclause_args
  *		Split an operator expression's arguments into Expr and Const parts.
  *
- * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Attempts to match the arguments to either (Expr op Const) or (Const op Expr)
+ * or (Expr op Expr), possibly with a RelabelType on top. When the expression
+ * matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
- * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
+ * non-null pointers (exprsp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
-					  bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	List	   *exprs = NIL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2042,22 +2073,26 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		exprs = lappend(exprs, leftop);
+		exprs = lappend(exprs, rightop);
+		expronleft = false;
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (exprsp)
+		*exprsp = exprs;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index ef118952c7..85f650f572 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1645,78 +1645,154 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+			if (cst)	/* Expr op Const */
+			{
+				int idx;
 
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert(list_length(clause_exprs) == 1);
+				clause_expr = (Node *) linitial(clause_exprs);
 
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
-			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
 
-				Assert(idx >= 0);
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(list_length(clause_exprs) == 2);
+
+				clause_expr = (Node *) linitial(clause_exprs);
+				clause_expr2 = (Node *) lsecond(clause_exprs);
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1726,6 +1802,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
 			Oid			collid;
@@ -1743,11 +1820,14 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
 			Assert(expronleft);
+			Assert(list_length(clause_exprs) == 1);
+
+			clause_expr = (Node *) linitial(clause_exprs);
 
 			/* XXX what if (cst->constisnull == NULL)? */
 			if (!cst->constisnull)
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 55cd9252a5..1f30fa9060 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,8 +99,8 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
-								  Const **cstp, bool *expronleftp);
+extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
+								  bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 7fb54de53d..301227079c 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1904,6 +1904,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2051,6 +2063,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      3750 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2460,6 +2484,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 ANALYZE mcv_lists;
@@ -2493,6 +2523,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       2500 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -2587,6 +2623,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
       1094 |      0
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 ANALYZE mcv_lists_bool;
@@ -2742,6 +2790,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 (1 row)
 
 DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2801,6 +2861,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2847,6 +2925,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index d563c4654c..b0cb0f1924 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -946,6 +946,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -999,6 +1003,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1198,6 +1206,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1216,6 +1226,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1291,6 +1303,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1376,6 +1392,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1403,6 +1423,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
@@ -1417,6 +1440,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 DROP TABLE mcv_lists_multi;
 
-- 
2.31.1

From e3fa97bfc187fa86f0fe2a7fe816dc79339a2a57 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Wed, 18 Aug 2021 11:53:45 +0200
Subject: [PATCH 2/2] fix: don't treat Var op Var as simple clauses

---
 src/backend/statistics/extended_stats.c | 23 +++++++++++++++++++++--
 1 file changed, 21 insertions(+), 2 deletions(-)

diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 606cf8c588..c66845887d 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1677,6 +1677,22 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 	return true;
 }
 
+static bool
+stat_clause_is_simple(Node *clause)
+{
+	List *vars;
+
+	if (IsA(clause, RestrictInfo))
+		clause = (Node *) ((RestrictInfo *) clause)->clause;
+
+	if (!is_opclause(clause) && !IsA(clause, Var))
+		return false;
+
+	vars = pull_var_clause(clause, 0);
+
+	return (list_length(vars) == 1);
+}
+
 /*
  * statext_mcv_clauselist_selectivity
  *		Estimate clauses using the best multi-column statistics.
@@ -1797,6 +1813,8 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		listidx = -1;
 		foreach(l, clauses)
 		{
+			Node *clause = (Node *) lfirst(l);
+
 			/* Increment the index before we decide if to skip the clause. */
 			listidx++;
 
@@ -1836,12 +1854,13 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 			if ((list_attnums[listidx] == NULL &&
 				 list_length(list_exprs[listidx]) == 1) ||
 				(list_exprs[listidx] == NIL &&
-				 bms_membership(list_attnums[listidx]) == BMS_SINGLETON))
+				 bms_membership(list_attnums[listidx]) == BMS_SINGLETON &&
+				 stat_clause_is_simple(clause)))
 				simple_clauses = bms_add_member(simple_clauses,
 												list_length(stat_clauses));
 
 			/* add clause to list and mark it as estimated */
-			stat_clauses = lappend(stat_clauses, (Node *) lfirst(l));
+			stat_clauses = lappend(stat_clauses, clause);
 			*estimatedclauses = bms_add_member(*estimatedclauses, listidx);
 
 			/*
-- 
2.31.1

From 708587e495483aff4b84487103a545b6e860d0c0 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Thu, 26 Aug 2021 23:01:04 +0200
Subject: [PATCH 1/3] Improve estimates for Var op Var with the same Var

When estimating (Var op Var) conditions, we can treat the case with the
same Var on both sides as a special case, and we can provide better
selectivity estimate than for the generic case.

For example for (a = a) we know it's 1.0, because all rows are expected
to match. Similarly for (a != a) , wich has selectivity 0.0. And the
same logic can be applied to inequality comparisons, like (a < a) etc.

In principle, those clauses are a bit strange and queries are unlikely
to use them. But query generators sometimes do silly things, and these
checks are quite cheap so it's likely a win.
---
 src/backend/utils/adt/selfuncs.c | 75 +++++++++++++++++++++++++++++++-
 1 file changed, 74 insertions(+), 1 deletion(-)

diff --git a/src/backend/utils/adt/selfuncs.c b/src/backend/utils/adt/selfuncs.c
index 0c8c05f6c2..0baeaa040d 100644
--- a/src/backend/utils/adt/selfuncs.c
+++ b/src/backend/utils/adt/selfuncs.c
@@ -210,7 +210,8 @@ static bool get_actual_variable_endpoint(Relation heapRel,
 										 MemoryContext outercontext,
 										 Datum *endpointDatum);
 static RelOptInfo *find_join_input_rel(PlannerInfo *root, Relids relids);
-
+static bool matching_restriction_variables(PlannerInfo *root, List *args,
+										   int varRelid);
 
 /*
  *		eqsel			- Selectivity of "=" for any data types.
@@ -256,6 +257,14 @@ eqsel_internal(PG_FUNCTION_ARGS, bool negate)
 		}
 	}
 
+	/*
+	 * It it's (variable = variable) with the same variable on both sides, it's
+	 * a special case and we know it does not filter anything, which means the
+	 * selectivity is 1.0.
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+		return (negate) ? 0.0 : 1.0;
+
 	/*
 	 * If expression is not variable = something or something = variable, then
 	 * punt and return a default estimate.
@@ -1408,6 +1417,20 @@ scalarineqsel_wrapper(PG_FUNCTION_ARGS, bool isgt, bool iseq)
 	Oid			consttype;
 	double		selec;
 
+	/*
+	 * Handle (variable < variable) and (variable <= variable) with the same
+	 * variable on both sides as a special case. The strict inequality should
+	 * not match anything, hence selectivity is 0.0. The other case is about
+	 * the same as equality, so selectivity is 1.0.
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+	{
+		if (iseq)
+			PG_RETURN_FLOAT8(1.0);
+		else
+			PG_RETURN_FLOAT8(0.0);
+	}
+
 	/*
 	 * If expression is not variable op something or something op variable,
 	 * then punt and return a default estimate.
@@ -4871,6 +4894,56 @@ get_restriction_variable(PlannerInfo *root, List *args, int varRelid,
 	return false;
 }
 
+
+/*
+ * matching_restriction_variable
+ *		Examine the args of a restriction clause to see if it's of the
+ *		form (variable op variable) with the save variable on both sides.
+ *
+ * Inputs:
+ *	root: the planner info
+ *	args: clause argument list
+ *	varRelid: see specs for restriction selectivity functions
+ *
+ * Returns true if the same variable is on both sides, otherwise false.
+ */
+static bool
+matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
+{
+	Node	   *left,
+			   *right;
+	VariableStatData ldata;
+	VariableStatData rdata;
+	bool		res = false;
+
+	/* Fail if not a binary opclause (probably shouldn't happen) */
+	if (list_length(args) != 2)
+		return false;
+
+	left = (Node *) linitial(args);
+	right = (Node *) lsecond(args);
+
+	/*
+	 * Examine both sides.  Note that when varRelid is nonzero, Vars of other
+	 * relations will be treated as pseudoconstants.
+	 */
+	examine_variable(root, left, varRelid, &ldata);
+	examine_variable(root, right, varRelid, &rdata);
+
+	/*
+	 * If both sides are variable, and are equal, we win.
+	 */
+	if ((ldata.rel != NULL && rdata.rel != NULL) &&
+		equal(ldata.var, rdata.var))
+		res = true;
+
+	/* We don't need the stats. */
+	ReleaseVariableStats(ldata);
+	ReleaseVariableStats(rdata);
+
+	return res;
+}
+
 /*
  * get_join_variables
  *		Apply examine_variable() to each side of a join clause.
-- 
2.31.1

From caa5c2aa0e640097a43e87f7d9557a640e6e8948 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Sun, 15 Aug 2021 13:23:13 +0200
Subject: [PATCH 2/3] patch 20210720

---
 src/backend/optimizer/path/clausesel.c        |  37 +++-
 src/backend/statistics/extended_stats.c       |  83 ++++++---
 src/backend/statistics/mcv.c                  | 172 +++++++++++++-----
 .../statistics/extended_stats_internal.h      |   4 +-
 src/test/regress/expected/stats_ext.out       |  96 ++++++++++
 src/test/regress/sql/stats_ext.sql            |  26 +++
 6 files changed, 341 insertions(+), 77 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index d263ecf082..6a7e9ceea5 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,37 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be a single (Expr op Expr) clause, which goes here due
+			 * to the optimization at the beginning of clauselist_selectivity.
+			 * So we try applying extended stats first, and then fall back to
+			 * restriction_selectivity.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				List	   *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					Bitmapset  *estimatedclauses = NULL;
+
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 2e55913bc8..606cf8c588 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1347,19 +1347,27 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		return true;
 	}
 
-	/* (Var/Expr op Const) or (Const op Var/Expr) */
+	/*
+	 * Three opclause variants are supported: (Expr op Const), (Const op Expr),
+	 * (Expr op Expr). That means we may need to analyze one or two expressions
+	 * to make sure the opclause is compatible with extended stats.
+	 */
 	if (is_opclause(clause))
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		ListCell   *lc;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
-		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		/*
+		 * Check if the expression has the right shape. This returns either one
+		 * or two expressions, depending on whether there is a Const.
+		 */
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
 		/*
@@ -1399,13 +1407,31 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
-		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
-		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		/*
+		 * There's always at least one expression, otherwise the clause would
+		 * not be considered compatible.
+		 */
+		Assert(list_length(clause_exprs) >= 1);
+
+		/*
+		 * Check all expressions by recursing. Var expressions are handled as
+		 * a special case (to match it to attnums etc.)
+		 */
+		foreach (lc, clause_exprs)
+		{
+			Node *clause_expr = (Node *) lfirst(lc);
+
+			if (IsA(clause_expr, Var))
+			{
+				/* if the Var is incompatible, the whole clause is incompatible */
+				if (!statext_is_compatible_clause_internal(root, clause_expr,
+														   relid, attnums, exprs))
+					return false;
+			}
+			else	/* generic expression */
+				*exprs = lappend(*exprs, clause_expr);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1415,15 +1441,21 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		ScalarArrayOpExpr *expr = (ScalarArrayOpExpr *) clause;
 		Node	   *clause_expr;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
+		/* There has to be one expression exactly. */
+		Assert(list_length(clause_exprs) == 1);
+
+		clause_expr = (Node *) linitial(clause_exprs);
+
 		/*
 		 * If it's not one of the supported operators ("=", "<", ">", etc.),
 		 * just ignore the clause, as it's not compatible with MCV lists.
@@ -2009,20 +2041,19 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * examine_opclause_args
  *		Split an operator expression's arguments into Expr and Const parts.
  *
- * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Attempts to match the arguments to either (Expr op Const) or (Const op Expr)
+ * or (Expr op Expr), possibly with a RelabelType on top. When the expression
+ * matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
- * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
+ * non-null pointers (exprsp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
-					  bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	List	   *exprs = NIL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2042,22 +2073,26 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		exprs = lappend(exprs, leftop);
+		exprs = lappend(exprs, rightop);
+		expronleft = false;
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (exprsp)
+		*exprsp = exprs;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index ef118952c7..85f650f572 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1645,78 +1645,154 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+			if (cst)	/* Expr op Const */
+			{
+				int idx;
 
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert(list_length(clause_exprs) == 1);
+				clause_expr = (Node *) linitial(clause_exprs);
 
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
-			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
 
-				Assert(idx >= 0);
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(list_length(clause_exprs) == 2);
+
+				clause_expr = (Node *) linitial(clause_exprs);
+				clause_expr2 = (Node *) lsecond(clause_exprs);
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1726,6 +1802,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
 			Oid			collid;
@@ -1743,11 +1820,14 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
 			Assert(expronleft);
+			Assert(list_length(clause_exprs) == 1);
+
+			clause_expr = (Node *) linitial(clause_exprs);
 
 			/* XXX what if (cst->constisnull == NULL)? */
 			if (!cst->constisnull)
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 55cd9252a5..1f30fa9060 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,8 +99,8 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
-								  Const **cstp, bool *expronleftp);
+extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
+								  bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 7fb54de53d..301227079c 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1904,6 +1904,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2051,6 +2063,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      3750 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2460,6 +2484,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 ANALYZE mcv_lists;
@@ -2493,6 +2523,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       2500 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -2587,6 +2623,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
       1094 |      0
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 ANALYZE mcv_lists_bool;
@@ -2742,6 +2790,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 (1 row)
 
 DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2801,6 +2861,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2847,6 +2925,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index d563c4654c..b0cb0f1924 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -946,6 +946,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -999,6 +1003,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1198,6 +1206,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1216,6 +1226,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1291,6 +1303,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1376,6 +1392,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1403,6 +1423,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
@@ -1417,6 +1440,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 DROP TABLE mcv_lists_multi;
 
-- 
2.31.1

From 743015add200d8041737f364281139ae0b5a5766 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Wed, 18 Aug 2021 11:53:45 +0200
Subject: [PATCH 3/3] Don't treat Var op Var as simple clauses

---
 src/backend/statistics/extended_stats.c       | 60 ++++++++++++++-----
 src/backend/statistics/mcv.c                  |  4 +-
 .../statistics/extended_stats_internal.h      |  2 +-
 3 files changed, 49 insertions(+), 17 deletions(-)

diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 606cf8c588..f555ace022 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1319,7 +1319,7 @@ choose_best_statistics(List *stats, char requiredkind,
 static bool
 statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 									  Index relid, Bitmapset **attnums,
-									  List **exprs)
+									  List **exprs, bool *issimple)
 {
 	/* Look inside any binary-compatible relabeling (as in examine_variable) */
 	if (IsA(clause, RelabelType))
@@ -1343,6 +1343,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			return false;
 
 		*attnums = bms_add_member(*attnums, var->varattno);
+		*issimple = true;
 
 		return true;
 	}
@@ -1367,7 +1368,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		 * Check if the expression has the right shape. This returns either one
 		 * or two expressions, depending on whether there is a Const.
 		 */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/*
@@ -1420,16 +1421,20 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		foreach (lc, clause_exprs)
 		{
 			Node *clause_expr = (Node *) lfirst(lc);
+			bool	tmp = false;
 
 			if (IsA(clause_expr, Var))
 			{
 				/* if the Var is incompatible, the whole clause is incompatible */
 				if (!statext_is_compatible_clause_internal(root, clause_expr,
-														   relid, attnums, exprs))
+														   relid, attnums, exprs,
+														   &tmp))
 					return false;
 			}
 			else	/* generic expression */
 				*exprs = lappend(*exprs, clause_expr);
+
+			*issimple = (*issimple && tmp);
 		}
 
 		return true;
@@ -1448,7 +1453,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/* There has to be one expression exactly. */
@@ -1496,7 +1501,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IN Array clauses by recursing. */
 		if (IsA(clause_expr, Var))
 			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IN Array. */
 		*exprs = lappend(*exprs, clause_expr);
@@ -1533,8 +1539,11 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			 */
 			if (!statext_is_compatible_clause_internal(root,
 													   (Node *) lfirst(lc),
-													   relid, attnums, exprs))
+													   relid, attnums, exprs,
+													   issimple))
 				return false;
+
+			*issimple = false;
 		}
 
 		return true;
@@ -1548,7 +1557,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IS NULL clauses by recursing. */
 		if (IsA(nt->arg, Var))
 			return statext_is_compatible_clause_internal(root, (Node *) (nt->arg),
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IS NULL. */
 		*exprs = lappend(*exprs, nt->arg);
@@ -1584,13 +1594,15 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
  */
 static bool
 statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
-							 Bitmapset **attnums, List **exprs)
+							 Bitmapset **attnums, List **exprs, bool *issimple)
 {
 	RangeTblEntry *rte = root->simple_rte_array[relid];
 	RestrictInfo *rinfo = (RestrictInfo *) clause;
 	int			clause_relid;
 	Oid			userid;
 
+	*issimple = true;
+
 	/*
 	 * Special-case handling for bare BoolExpr AND clauses, because the
 	 * restrictinfo machinery doesn't build RestrictInfos on top of AND
@@ -1608,8 +1620,11 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 		foreach(lc, expr->args)
 		{
 			if (!statext_is_compatible_clause(root, (Node *) lfirst(lc),
-											  relid, attnums, exprs))
+											  relid, attnums, exprs,
+											  issimple))
 				return false;
+
+			*issimple = false;
 		}
 
 		return true;
@@ -1630,7 +1645,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 
 	/* Check the clause and determine what attributes it references. */
 	if (!statext_is_compatible_clause_internal(root, (Node *) rinfo->clause,
-											   relid, attnums, exprs))
+											   relid, attnums, exprs, issimple))
 		return false;
 
 	/*
@@ -1720,6 +1735,7 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	ListCell   *l;
 	Bitmapset **list_attnums;	/* attnums extracted from the clause */
 	List	  **list_exprs;		/* expressions matched to any statistic */
+	bool	  *list_simple;		/* marks simple expressions */
 	int			listidx;
 	Selectivity sel = (is_or) ? 0.0 : 1.0;
 
@@ -1733,6 +1749,9 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	/* expressions extracted from complex expressions */
 	list_exprs = (List **) palloc(sizeof(Node *) * list_length(clauses));
 
+	/* expressions determined to be simple (single expression) */
+	list_simple = (bool *) palloc(sizeof(bool) * list_length(clauses));
+
 	/*
 	 * Pre-process the clauses list to extract the attnums and expressions
 	 * seen in each item.  We need to determine if there are any clauses which
@@ -1750,17 +1769,21 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		Node	   *clause = (Node *) lfirst(l);
 		Bitmapset  *attnums = NULL;
 		List	   *exprs = NIL;
+		bool		issimple = false;
 
 		if (!bms_is_member(listidx, *estimatedclauses) &&
-			statext_is_compatible_clause(root, clause, rel->relid, &attnums, &exprs))
+			statext_is_compatible_clause(root, clause, rel->relid,
+										 &attnums, &exprs, &issimple))
 		{
 			list_attnums[listidx] = attnums;
 			list_exprs[listidx] = exprs;
+			list_simple[listidx] = issimple;
 		}
 		else
 		{
 			list_attnums[listidx] = NULL;
 			list_exprs[listidx] = NIL;
+			list_simple[listidx] = false;
 		}
 
 		listidx++;
@@ -1797,6 +1820,8 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		listidx = -1;
 		foreach(l, clauses)
 		{
+			Node *clause = (Node *) lfirst(l);
+
 			/* Increment the index before we decide if to skip the clause. */
 			listidx++;
 
@@ -1836,12 +1861,12 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 			if ((list_attnums[listidx] == NULL &&
 				 list_length(list_exprs[listidx]) == 1) ||
 				(list_exprs[listidx] == NIL &&
-				 bms_membership(list_attnums[listidx]) == BMS_SINGLETON))
+				 list_simple[listidx]))
 				simple_clauses = bms_add_member(simple_clauses,
 												list_length(stat_clauses));
 
 			/* add clause to list and mark it as estimated */
-			stat_clauses = lappend(stat_clauses, (Node *) lfirst(l));
+			stat_clauses = lappend(stat_clauses, clause);
 			*estimatedclauses = bms_add_member(*estimatedclauses, listidx);
 
 			/*
@@ -2050,13 +2075,15 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp,
+					  bool *issimplep)
 {
 	List	   *exprs = NIL;
 	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
+	bool		issimple;
 
 	/* enforced by statext_is_compatible_clause_internal */
 	Assert(list_length(args) == 2);
@@ -2082,12 +2109,14 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
+		issimple = true;
 	}
 	else
 	{
 		exprs = lappend(exprs, leftop);
 		exprs = lappend(exprs, rightop);
 		expronleft = false;
+		issimple = true;
 	}
 
 	/* return pointers to the extracted parts if requested */
@@ -2100,6 +2129,9 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 	if (expronleftp)
 		*expronleftp = expronleft;
 
+	if (issimplep)
+		*issimplep = issimple;
+
 	return true;
 }
 
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index 85f650f572..ad8172db90 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1654,7 +1654,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			if (cst)	/* Expr op Const */
@@ -1820,7 +1820,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 1f30fa9060..d86cc4184b 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -100,7 +100,7 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									int numattrs, AttrNumber *attnums);
 
 extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
-								  bool *expronleftp);
+								  bool *expronleftp, bool *issimplep);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
-- 
2.31.1

From 71837f82db88baa9e0250d3868078140cf11f5a7 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Thu, 26 Aug 2021 23:01:04 +0200
Subject: [PATCH 1/3] Improve estimates for Var op Var with the same Var

When estimating (Var op Var) conditions, we can treat the case with the
same Var on both sides as a special case, and we can provide better
selectivity estimate than for the generic case.

For example for (a = a) we know it's 1.0, because all rows are expected
to match. Similarly for (a != a) , wich has selectivity 0.0. And the
same logic can be applied to inequality comparisons, like (a < a) etc.

In principle, those clauses are a bit strange and queries are unlikely
to use them. But query generators sometimes do silly things, and these
checks are quite cheap so it's likely a win.
---
 src/backend/utils/adt/selfuncs.c | 77 +++++++++++++++++++++++++++++++-
 1 file changed, 76 insertions(+), 1 deletion(-)

diff --git a/src/backend/utils/adt/selfuncs.c b/src/backend/utils/adt/selfuncs.c
index 0c8c05f6c2..22608aadc1 100644
--- a/src/backend/utils/adt/selfuncs.c
+++ b/src/backend/utils/adt/selfuncs.c
@@ -210,7 +210,8 @@ static bool get_actual_variable_endpoint(Relation heapRel,
 										 MemoryContext outercontext,
 										 Datum *endpointDatum);
 static RelOptInfo *find_join_input_rel(PlannerInfo *root, Relids relids);
-
+static bool matching_restriction_variables(PlannerInfo *root, List *args,
+										   int varRelid);
 
 /*
  *		eqsel			- Selectivity of "=" for any data types.
@@ -256,6 +257,14 @@ eqsel_internal(PG_FUNCTION_ARGS, bool negate)
 		}
 	}
 
+	/*
+	 * It it's (variable = variable) with the same variable on both sides, it's
+	 * a special case and we know it's not expected to filter anything, so we
+	 * estimate the selectivity as 1.0 (or 0.0 if it's negated).
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+		return (negate) ? 0.0 : 1.0;
+
 	/*
 	 * If expression is not variable = something or something = variable, then
 	 * punt and return a default estimate.
@@ -1408,6 +1417,22 @@ scalarineqsel_wrapper(PG_FUNCTION_ARGS, bool isgt, bool iseq)
 	Oid			consttype;
 	double		selec;
 
+	/*
+	 * Handle (variable < variable) and (variable <= variable) with the same
+	 * variable on both sides as a special case. The strict inequality should
+	 * not match any rows, hence selectivity is 0.0. The other case is about
+	 * the same as equality, so selectivity is 1.0.
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+	{
+		/* The case with equality matches all rows, so estimate it as 1.0. */
+		if (iseq)
+			PG_RETURN_FLOAT8(1.0);
+
+		/* Strict inequality matches nothing, so selectivity is 0.0. */
+		PG_RETURN_FLOAT8(0.0);
+	}
+
 	/*
 	 * If expression is not variable op something or something op variable,
 	 * then punt and return a default estimate.
@@ -4871,6 +4896,56 @@ get_restriction_variable(PlannerInfo *root, List *args, int varRelid,
 	return false;
 }
 
+
+/*
+ * matching_restriction_variable
+ *		Examine the args of a restriction clause to see if it's of the
+ *		form (variable op variable) with the same variable on both sides.
+ *
+ * Inputs:
+ *	root: the planner info
+ *	args: clause argument list
+ *	varRelid: see specs for restriction selectivity functions
+ *
+ * Returns true if the same variable is on both sides, otherwise false.
+ */
+static bool
+matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
+{
+	Node	   *left,
+			   *right;
+	VariableStatData ldata;
+	VariableStatData rdata;
+	bool		res = false;
+
+	/* Fail if not a binary opclause (probably shouldn't happen) */
+	if (list_length(args) != 2)
+		return false;
+
+	left = (Node *) linitial(args);
+	right = (Node *) lsecond(args);
+
+	/*
+	 * Examine both sides.  Note that when varRelid is nonzero, Vars of other
+	 * relations will be treated as pseudoconstants.
+	 */
+	examine_variable(root, left, varRelid, &ldata);
+	examine_variable(root, right, varRelid, &rdata);
+
+	/*
+	 * If both sides are variable, and are equal, we win.
+	 */
+	if ((ldata.rel != NULL && rdata.rel != NULL) &&
+		equal(ldata.var, rdata.var))
+		res = true;
+
+	/* We don't need the stats. */
+	ReleaseVariableStats(ldata);
+	ReleaseVariableStats(rdata);
+
+	return res;
+}
+
 /*
  * get_join_variables
  *		Apply examine_variable() to each side of a join clause.
-- 
2.31.1

From ee19131d26afb9029798fa42d0fcd46ac72589de Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Sun, 15 Aug 2021 13:23:13 +0200
Subject: [PATCH 2/3] main patch

---
 src/backend/optimizer/path/clausesel.c        |  37 +++-
 src/backend/statistics/extended_stats.c       |  83 ++++++---
 src/backend/statistics/mcv.c                  | 172 +++++++++++++-----
 .../statistics/extended_stats_internal.h      |   4 +-
 src/test/regress/expected/stats_ext.out       |  96 ++++++++++
 src/test/regress/sql/stats_ext.sql            |  26 +++
 6 files changed, 341 insertions(+), 77 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index d263ecf082..6a7e9ceea5 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,37 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be a single (Expr op Expr) clause, which goes here due
+			 * to the optimization at the beginning of clauselist_selectivity.
+			 * So we try applying extended stats first, and then fall back to
+			 * restriction_selectivity.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				List	   *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					Bitmapset  *estimatedclauses = NULL;
+
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 2e55913bc8..606cf8c588 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1347,19 +1347,27 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		return true;
 	}
 
-	/* (Var/Expr op Const) or (Const op Var/Expr) */
+	/*
+	 * Three opclause variants are supported: (Expr op Const), (Const op Expr),
+	 * (Expr op Expr). That means we may need to analyze one or two expressions
+	 * to make sure the opclause is compatible with extended stats.
+	 */
 	if (is_opclause(clause))
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		ListCell   *lc;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
-		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		/*
+		 * Check if the expression has the right shape. This returns either one
+		 * or two expressions, depending on whether there is a Const.
+		 */
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
 		/*
@@ -1399,13 +1407,31 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
-		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
-		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		/*
+		 * There's always at least one expression, otherwise the clause would
+		 * not be considered compatible.
+		 */
+		Assert(list_length(clause_exprs) >= 1);
+
+		/*
+		 * Check all expressions by recursing. Var expressions are handled as
+		 * a special case (to match it to attnums etc.)
+		 */
+		foreach (lc, clause_exprs)
+		{
+			Node *clause_expr = (Node *) lfirst(lc);
+
+			if (IsA(clause_expr, Var))
+			{
+				/* if the Var is incompatible, the whole clause is incompatible */
+				if (!statext_is_compatible_clause_internal(root, clause_expr,
+														   relid, attnums, exprs))
+					return false;
+			}
+			else	/* generic expression */
+				*exprs = lappend(*exprs, clause_expr);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1415,15 +1441,21 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		ScalarArrayOpExpr *expr = (ScalarArrayOpExpr *) clause;
 		Node	   *clause_expr;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
+		/* There has to be one expression exactly. */
+		Assert(list_length(clause_exprs) == 1);
+
+		clause_expr = (Node *) linitial(clause_exprs);
+
 		/*
 		 * If it's not one of the supported operators ("=", "<", ">", etc.),
 		 * just ignore the clause, as it's not compatible with MCV lists.
@@ -2009,20 +2041,19 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * examine_opclause_args
  *		Split an operator expression's arguments into Expr and Const parts.
  *
- * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Attempts to match the arguments to either (Expr op Const) or (Const op Expr)
+ * or (Expr op Expr), possibly with a RelabelType on top. When the expression
+ * matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
- * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
+ * non-null pointers (exprsp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
-					  bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	List	   *exprs = NIL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2042,22 +2073,26 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		exprs = lappend(exprs, leftop);
+		exprs = lappend(exprs, rightop);
+		expronleft = false;
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (exprsp)
+		*exprsp = exprs;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index ef118952c7..85f650f572 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1645,78 +1645,154 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+			if (cst)	/* Expr op Const */
+			{
+				int idx;
 
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert(list_length(clause_exprs) == 1);
+				clause_expr = (Node *) linitial(clause_exprs);
 
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
-			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
 
-				Assert(idx >= 0);
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(list_length(clause_exprs) == 2);
+
+				clause_expr = (Node *) linitial(clause_exprs);
+				clause_expr2 = (Node *) lsecond(clause_exprs);
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1726,6 +1802,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
 			Oid			collid;
@@ -1743,11 +1820,14 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
 			Assert(expronleft);
+			Assert(list_length(clause_exprs) == 1);
+
+			clause_expr = (Node *) linitial(clause_exprs);
 
 			/* XXX what if (cst->constisnull == NULL)? */
 			if (!cst->constisnull)
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 55cd9252a5..1f30fa9060 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,8 +99,8 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
-								  Const **cstp, bool *expronleftp);
+extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
+								  bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 7fb54de53d..301227079c 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1904,6 +1904,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2051,6 +2063,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      3750 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2460,6 +2484,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 ANALYZE mcv_lists;
@@ -2493,6 +2523,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       2500 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -2587,6 +2623,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
       1094 |      0
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 ANALYZE mcv_lists_bool;
@@ -2742,6 +2790,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 (1 row)
 
 DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2801,6 +2861,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2847,6 +2925,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index d563c4654c..b0cb0f1924 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -946,6 +946,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -999,6 +1003,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1198,6 +1206,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1216,6 +1226,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1291,6 +1303,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1376,6 +1392,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1403,6 +1423,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
@@ -1417,6 +1440,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 DROP TABLE mcv_lists_multi;
 
-- 
2.31.1

From 7926ea227c961e5c5ba3f4cefd254d794cf5f62d Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Wed, 18 Aug 2021 11:53:45 +0200
Subject: [PATCH 3/3] Don't treat Var op Var as simple clauses

---
 src/backend/statistics/extended_stats.c       | 86 +++++++++++++++----
 src/backend/statistics/mcv.c                  |  4 +-
 .../statistics/extended_stats_internal.h      |  2 +-
 3 files changed, 74 insertions(+), 18 deletions(-)

diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 606cf8c588..5f283a0afb 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1315,11 +1315,14 @@ choose_best_statistics(List *stats, char requiredkind,
  * statext_is_compatible_clause. It needs to be split like this because
  * of recursion.  The attnums bitmap is an input/output parameter collecting
  * attribute numbers from all compatible clauses (recursively).
+ *
+ * XXX The issimple variable is expected to be initialized by the caller, we
+ * just update it while recursively analyzing the current clause.
  */
 static bool
 statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 									  Index relid, Bitmapset **attnums,
-									  List **exprs)
+									  List **exprs, bool *issimple)
 {
 	/* Look inside any binary-compatible relabeling (as in examine_variable) */
 	if (IsA(clause, RelabelType))
@@ -1367,7 +1370,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		 * Check if the expression has the right shape. This returns either one
 		 * or two expressions, depending on whether there is a Const.
 		 */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/*
@@ -1416,6 +1419,12 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/*
 		 * Check all expressions by recursing. Var expressions are handled as
 		 * a special case (to match it to attnums etc.)
+		 *
+		 * An opclause is simple if it's (Expr op Const) or (Const op Expr). We
+		 * have already checked the overall shape in examine_opclause_args, but
+		 * we haven't checked the expressions are simple (i.e. pretty much Var),
+		 * so we need to check that now. If we discover a complex expression, we
+		 * consider the whole clause complex.
 		 */
 		foreach (lc, clause_exprs)
 		{
@@ -1425,11 +1434,17 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			{
 				/* if the Var is incompatible, the whole clause is incompatible */
 				if (!statext_is_compatible_clause_internal(root, clause_expr,
-														   relid, attnums, exprs))
+														   relid, attnums, exprs,
+														   issimple))
 					return false;
 			}
 			else	/* generic expression */
+			{
 				*exprs = lappend(*exprs, clause_expr);
+
+				/* switch to false if there are any complex clauses */
+				*issimple = false;
+			}
 		}
 
 		return true;
@@ -1448,7 +1463,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/* There has to be one expression exactly. */
@@ -1496,7 +1511,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IN Array clauses by recursing. */
 		if (IsA(clause_expr, Var))
 			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IN Array. */
 		*exprs = lappend(*exprs, clause_expr);
@@ -1525,6 +1541,14 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		BoolExpr   *expr = (BoolExpr *) clause;
 		ListCell   *lc;
 
+		/*
+		 * All AND/OR clauses are considered complex, even if all arguments are
+		 * simple clauses. For NOT clauses we need to check the argument and then
+		 * we can update the flag.
+		 */
+		if (!is_notclause(clause))
+			*issimple = false;
+
 		foreach(lc, expr->args)
 		{
 			/*
@@ -1533,7 +1557,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			 */
 			if (!statext_is_compatible_clause_internal(root,
 													   (Node *) lfirst(lc),
-													   relid, attnums, exprs))
+													   relid, attnums, exprs,
+													   issimple))
 				return false;
 		}
 
@@ -1548,7 +1573,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IS NULL clauses by recursing. */
 		if (IsA(nt->arg, Var))
 			return statext_is_compatible_clause_internal(root, (Node *) (nt->arg),
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IS NULL. */
 		*exprs = lappend(*exprs, nt->arg);
@@ -1584,22 +1610,35 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
  */
 static bool
 statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
-							 Bitmapset **attnums, List **exprs)
+							 Bitmapset **attnums, List **exprs, bool *issimple)
 {
 	RangeTblEntry *rte = root->simple_rte_array[relid];
 	RestrictInfo *rinfo = (RestrictInfo *) clause;
 	int			clause_relid;
 	Oid			userid;
 
+	/*
+	 * Clauses are considered simple by default, and we mark them as complex
+	 * when we discover a complex part.
+	 */
+	*issimple = true;
+
 	/*
 	 * Special-case handling for bare BoolExpr AND clauses, because the
 	 * restrictinfo machinery doesn't build RestrictInfos on top of AND
 	 * clauses.
+	 *
+	 * AND clauses are considered complex, even if all arguments are
+	 * simple clauses.
 	 */
 	if (is_andclause(clause))
 	{
 		BoolExpr   *expr = (BoolExpr *) clause;
 		ListCell   *lc;
+		bool		tmp = false;	/* ignored result */
+
+		/* AND clauses are complex, even if the arguments are simple. */
+		*issimple = false;
 
 		/*
 		 * Check that each sub-clause is compatible.  We expect these to be
@@ -1608,7 +1647,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 		foreach(lc, expr->args)
 		{
 			if (!statext_is_compatible_clause(root, (Node *) lfirst(lc),
-											  relid, attnums, exprs))
+											  relid, attnums, exprs, &tmp))
 				return false;
 		}
 
@@ -1630,7 +1669,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 
 	/* Check the clause and determine what attributes it references. */
 	if (!statext_is_compatible_clause_internal(root, (Node *) rinfo->clause,
-											   relid, attnums, exprs))
+											   relid, attnums, exprs, issimple))
 		return false;
 
 	/*
@@ -1720,6 +1759,7 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	ListCell   *l;
 	Bitmapset **list_attnums;	/* attnums extracted from the clause */
 	List	  **list_exprs;		/* expressions matched to any statistic */
+	bool	  *list_simple;		/* marks simple expressions */
 	int			listidx;
 	Selectivity sel = (is_or) ? 0.0 : 1.0;
 
@@ -1733,6 +1773,9 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	/* expressions extracted from complex expressions */
 	list_exprs = (List **) palloc(sizeof(Node *) * list_length(clauses));
 
+	/* expressions determined to be simple (single expression) */
+	list_simple = (bool *) palloc(sizeof(bool) * list_length(clauses));
+
 	/*
 	 * Pre-process the clauses list to extract the attnums and expressions
 	 * seen in each item.  We need to determine if there are any clauses which
@@ -1750,17 +1793,21 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		Node	   *clause = (Node *) lfirst(l);
 		Bitmapset  *attnums = NULL;
 		List	   *exprs = NIL;
+		bool		issimple = false;
 
 		if (!bms_is_member(listidx, *estimatedclauses) &&
-			statext_is_compatible_clause(root, clause, rel->relid, &attnums, &exprs))
+			statext_is_compatible_clause(root, clause, rel->relid,
+										 &attnums, &exprs, &issimple))
 		{
 			list_attnums[listidx] = attnums;
 			list_exprs[listidx] = exprs;
+			list_simple[listidx] = issimple;
 		}
 		else
 		{
 			list_attnums[listidx] = NULL;
 			list_exprs[listidx] = NIL;
+			list_simple[listidx] = false;
 		}
 
 		listidx++;
@@ -1797,6 +1844,8 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		listidx = -1;
 		foreach(l, clauses)
 		{
+			Node *clause = (Node *) lfirst(l);
+
 			/* Increment the index before we decide if to skip the clause. */
 			listidx++;
 
@@ -1835,13 +1884,12 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 			/* record simple clauses (single column or expression) */
 			if ((list_attnums[listidx] == NULL &&
 				 list_length(list_exprs[listidx]) == 1) ||
-				(list_exprs[listidx] == NIL &&
-				 bms_membership(list_attnums[listidx]) == BMS_SINGLETON))
+				 list_simple[listidx])
 				simple_clauses = bms_add_member(simple_clauses,
 												list_length(stat_clauses));
 
 			/* add clause to list and mark it as estimated */
-			stat_clauses = lappend(stat_clauses, (Node *) lfirst(l));
+			stat_clauses = lappend(stat_clauses, clause);
 			*estimatedclauses = bms_add_member(*estimatedclauses, listidx);
 
 			/*
@@ -2050,13 +2098,15 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp,
+					  bool *issimplep)
 {
 	List	   *exprs = NIL;
 	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
+	bool		issimple;
 
 	/* enforced by statext_is_compatible_clause_internal */
 	Assert(list_length(args) == 2);
@@ -2076,18 +2126,21 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
+		issimple = true;
 	}
 	else if (IsA(leftop, Const))
 	{
 		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
+		issimple = true;
 	}
 	else
 	{
 		exprs = lappend(exprs, leftop);
 		exprs = lappend(exprs, rightop);
 		expronleft = false;
+		issimple = false;
 	}
 
 	/* return pointers to the extracted parts if requested */
@@ -2100,6 +2153,9 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 	if (expronleftp)
 		*expronleftp = expronleft;
 
+	if (issimplep)
+		*issimplep = (*issimplep && issimple);
+
 	return true;
 }
 
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index 85f650f572..ad8172db90 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1654,7 +1654,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			if (cst)	/* Expr op Const */
@@ -1820,7 +1820,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 1f30fa9060..d86cc4184b 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -100,7 +100,7 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									int numattrs, AttrNumber *attnums);
 
 extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
-								  bool *expronleftp);
+								  bool *expronleftp, bool *issimplep);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
-- 
2.31.1

From 91371c3e4252580a30530d5b625f72d5525c5c73 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Thu, 26 Aug 2021 23:01:04 +0200
Subject: [PATCH 1/5] Improve estimates for Var op Var with the same Var

When estimating (Var op Var) conditions, we can treat the case with the
same Var on both sides as a special case, and we can provide better
selectivity estimate than for the generic case.

For example for (a = a) we know it's 1.0, because all rows are expected
to match. Similarly for (a != a) , wich has selectivity 0.0. And the
same logic can be applied to inequality comparisons, like (a < a) etc.

In principle, those clauses are a bit strange and queries are unlikely
to use them. But query generators sometimes do silly things, and these
checks are quite cheap so it's likely a win.
---
 src/backend/utils/adt/selfuncs.c | 77 +++++++++++++++++++++++++++++++-
 1 file changed, 76 insertions(+), 1 deletion(-)

diff --git a/src/backend/utils/adt/selfuncs.c b/src/backend/utils/adt/selfuncs.c
index 10895fb287..59038895ec 100644
--- a/src/backend/utils/adt/selfuncs.c
+++ b/src/backend/utils/adt/selfuncs.c
@@ -210,7 +210,8 @@ static bool get_actual_variable_endpoint(Relation heapRel,
 										 MemoryContext outercontext,
 										 Datum *endpointDatum);
 static RelOptInfo *find_join_input_rel(PlannerInfo *root, Relids relids);
-
+static bool matching_restriction_variables(PlannerInfo *root, List *args,
+										   int varRelid);
 
 /*
  *		eqsel			- Selectivity of "=" for any data types.
@@ -256,6 +257,14 @@ eqsel_internal(PG_FUNCTION_ARGS, bool negate)
 		}
 	}
 
+	/*
+	 * It it's (variable = variable) with the same variable on both sides, it's
+	 * a special case and we know it's not expected to filter anything, so we
+	 * estimate the selectivity as 1.0 (or 0.0 if it's negated).
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+		return (negate) ? 0.0 : 1.0;
+
 	/*
 	 * If expression is not variable = something or something = variable, then
 	 * punt and return a default estimate.
@@ -1408,6 +1417,22 @@ scalarineqsel_wrapper(PG_FUNCTION_ARGS, bool isgt, bool iseq)
 	Oid			consttype;
 	double		selec;
 
+	/*
+	 * Handle (variable < variable) and (variable <= variable) with the same
+	 * variable on both sides as a special case. The strict inequality should
+	 * not match any rows, hence selectivity is 0.0. The other case is about
+	 * the same as equality, so selectivity is 1.0.
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+	{
+		/* The case with equality matches all rows, so estimate it as 1.0. */
+		if (iseq)
+			PG_RETURN_FLOAT8(1.0);
+
+		/* Strict inequality matches nothing, so selectivity is 0.0. */
+		PG_RETURN_FLOAT8(0.0);
+	}
+
 	/*
 	 * If expression is not variable op something or something op variable,
 	 * then punt and return a default estimate.
@@ -4871,6 +4896,56 @@ get_restriction_variable(PlannerInfo *root, List *args, int varRelid,
 	return false;
 }
 
+
+/*
+ * matching_restriction_variable
+ *		Examine the args of a restriction clause to see if it's of the
+ *		form (variable op variable) with the same variable on both sides.
+ *
+ * Inputs:
+ *	root: the planner info
+ *	args: clause argument list
+ *	varRelid: see specs for restriction selectivity functions
+ *
+ * Returns true if the same variable is on both sides, otherwise false.
+ */
+static bool
+matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
+{
+	Node	   *left,
+			   *right;
+	VariableStatData ldata;
+	VariableStatData rdata;
+	bool		res = false;
+
+	/* Fail if not a binary opclause (probably shouldn't happen) */
+	if (list_length(args) != 2)
+		return false;
+
+	left = (Node *) linitial(args);
+	right = (Node *) lsecond(args);
+
+	/*
+	 * Examine both sides.  Note that when varRelid is nonzero, Vars of other
+	 * relations will be treated as pseudoconstants.
+	 */
+	examine_variable(root, left, varRelid, &ldata);
+	examine_variable(root, right, varRelid, &rdata);
+
+	/*
+	 * If both sides are variable, and are equal, we win.
+	 */
+	if ((ldata.rel != NULL && rdata.rel != NULL) &&
+		equal(ldata.var, rdata.var))
+		res = true;
+
+	/* We don't need the stats. */
+	ReleaseVariableStats(ldata);
+	ReleaseVariableStats(rdata);
+
+	return res;
+}
+
 /*
  * get_join_variables
  *		Apply examine_variable() to each side of a join clause.
-- 
2.31.1

From 8e454f1aae54317d1d346012c06179ba62fbef47 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Mon, 13 Dec 2021 02:28:27 +0100
Subject: [PATCH 2/5] simplification

---
 src/backend/utils/adt/selfuncs.c | 46 +++++++++++++++++---------------
 1 file changed, 24 insertions(+), 22 deletions(-)

diff --git a/src/backend/utils/adt/selfuncs.c b/src/backend/utils/adt/selfuncs.c
index 59038895ec..1031a0fd9e 100644
--- a/src/backend/utils/adt/selfuncs.c
+++ b/src/backend/utils/adt/selfuncs.c
@@ -260,7 +260,7 @@ eqsel_internal(PG_FUNCTION_ARGS, bool negate)
 	/*
 	 * It it's (variable = variable) with the same variable on both sides, it's
 	 * a special case and we know it's not expected to filter anything, so we
-	 * estimate the selectivity as 1.0 (or 0.0 if it's negated).
+	 * estimate the selectivity as 1.0 (or 0.0 when it's negated).
 	 */
 	if (matching_restriction_variables(root, args, varRelid))
 		return (negate) ? 0.0 : 1.0;
@@ -4896,11 +4896,11 @@ get_restriction_variable(PlannerInfo *root, List *args, int varRelid,
 	return false;
 }
 
-
 /*
  * matching_restriction_variable
- *		Examine the args of a restriction clause to see if it's of the
- *		form (variable op variable) with the same variable on both sides.
+ *		Check if the two arguments of a restriction clause refer to the same
+ *		variable, i.e. if the condition is of the form (variable op variable).
+ *		We can deduce selectivity for such (in)equality clauses.
  *
  * Inputs:
  *	root: the planner info
@@ -4914,9 +4914,7 @@ matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
 {
 	Node	   *left,
 			   *right;
-	VariableStatData ldata;
-	VariableStatData rdata;
-	bool		res = false;
+	Var		   *lvar;
 
 	/* Fail if not a binary opclause (probably shouldn't happen) */
 	if (list_length(args) != 2)
@@ -4925,25 +4923,29 @@ matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
 	left = (Node *) linitial(args);
 	right = (Node *) lsecond(args);
 
-	/*
-	 * Examine both sides.  Note that when varRelid is nonzero, Vars of other
-	 * relations will be treated as pseudoconstants.
-	 */
-	examine_variable(root, left, varRelid, &ldata);
-	examine_variable(root, right, varRelid, &rdata);
+	/* Look inside any binary-compatible relabeling */
+
+	if (IsA(left, RelabelType))
+		left = (Node *) ((RelabelType *) left)->arg;
+
+	if (IsA(right, RelabelType))
+		right = (Node *) ((RelabelType *) right)->arg;
+
+	/* We only care about simple Vars from our relation for now */
+	if (!IsA(left, Var) || !IsA(right, Var))
+		return false;
+
+	lvar = (Var *) left;
 
 	/*
-	 * If both sides are variable, and are equal, we win.
+	 * We only check one variable, because we call equals() later, which checks
+	 * the other variable automatically.
 	 */
-	if ((ldata.rel != NULL && rdata.rel != NULL) &&
-		equal(ldata.var, rdata.var))
-		res = true;
-
-	/* We don't need the stats. */
-	ReleaseVariableStats(ldata);
-	ReleaseVariableStats(rdata);
+	if ((varRelid != 0) && (varRelid != lvar->varno))
+		return false;
 
-	return res;
+	/* The two variables need to match */
+	return equal(left, right);
 }
 
 /*
-- 
2.31.1

From d1b54830eea71b6afb465d4c86d9a4a92742dc98 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Mon, 13 Dec 2021 02:48:58 +0100
Subject: [PATCH 3/5] relax the restrictions

---
 src/backend/utils/adt/selfuncs.c | 25 ++++++++++++++++---------
 1 file changed, 16 insertions(+), 9 deletions(-)

diff --git a/src/backend/utils/adt/selfuncs.c b/src/backend/utils/adt/selfuncs.c
index 1031a0fd9e..6b0dde66bf 100644
--- a/src/backend/utils/adt/selfuncs.c
+++ b/src/backend/utils/adt/selfuncs.c
@@ -4914,7 +4914,7 @@ matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
 {
 	Node	   *left,
 			   *right;
-	Var		   *lvar;
+	Bitmapset  *varnos;
 
 	/* Fail if not a binary opclause (probably shouldn't happen) */
 	if (list_length(args) != 2)
@@ -4931,17 +4931,24 @@ matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
 	if (IsA(right, RelabelType))
 		right = (Node *) ((RelabelType *) right)->arg;
 
-	/* We only care about simple Vars from our relation for now */
-	if (!IsA(left, Var) || !IsA(right, Var))
+	/*
+	 * Check if it's safe to compare the two expressions. The expressions
+	 * must not contain any volatile expressions, and must belong to the
+	 * same relation.
+	 *
+	 * XXX We do a small trick - we validate just one expression, and then
+	 * check it's equal to the other side.
+	 */
+	if (contain_volatile_functions(left))
 		return false;
 
-	lvar = (Var *) left;
+	/* Check it belongs to a single relation. */
+	varnos = pull_varnos(root, left);
 
-	/*
-	 * We only check one variable, because we call equals() later, which checks
-	 * the other variable automatically.
-	 */
-	if ((varRelid != 0) && (varRelid != lvar->varno))
+	if (bms_num_members(varnos) != 1)
+		return false;
+
+	if ((varRelid != 0) && (!bms_is_member(varRelid, varnos)))
 		return false;
 
 	/* The two variables need to match */
-- 
2.31.1

From 1f6949e9eadcb961738e4314414682e179462d16 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Sun, 15 Aug 2021 13:23:13 +0200
Subject: [PATCH 4/5] main patch

---
 src/backend/optimizer/path/clausesel.c        |  37 +++-
 src/backend/statistics/extended_stats.c       |  83 ++++++---
 src/backend/statistics/mcv.c                  | 172 +++++++++++++-----
 .../statistics/extended_stats_internal.h      |   4 +-
 src/test/regress/expected/stats_ext.out       |  96 ++++++++++
 src/test/regress/sql/stats_ext.sql            |  26 +++
 6 files changed, 341 insertions(+), 77 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index d263ecf082..6a7e9ceea5 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,37 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be a single (Expr op Expr) clause, which goes here due
+			 * to the optimization at the beginning of clauselist_selectivity.
+			 * So we try applying extended stats first, and then fall back to
+			 * restriction_selectivity.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				List	   *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					Bitmapset  *estimatedclauses = NULL;
+
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 69ca52094f..e1fcfd0711 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1351,19 +1351,27 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		return true;
 	}
 
-	/* (Var/Expr op Const) or (Const op Var/Expr) */
+	/*
+	 * Three opclause variants are supported: (Expr op Const), (Const op Expr),
+	 * (Expr op Expr). That means we may need to analyze one or two expressions
+	 * to make sure the opclause is compatible with extended stats.
+	 */
 	if (is_opclause(clause))
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		ListCell   *lc;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
-		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		/*
+		 * Check if the expression has the right shape. This returns either one
+		 * or two expressions, depending on whether there is a Const.
+		 */
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
 		/*
@@ -1403,13 +1411,31 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
-		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
-		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		/*
+		 * There's always at least one expression, otherwise the clause would
+		 * not be considered compatible.
+		 */
+		Assert(list_length(clause_exprs) >= 1);
+
+		/*
+		 * Check all expressions by recursing. Var expressions are handled as
+		 * a special case (to match it to attnums etc.)
+		 */
+		foreach (lc, clause_exprs)
+		{
+			Node *clause_expr = (Node *) lfirst(lc);
+
+			if (IsA(clause_expr, Var))
+			{
+				/* if the Var is incompatible, the whole clause is incompatible */
+				if (!statext_is_compatible_clause_internal(root, clause_expr,
+														   relid, attnums, exprs))
+					return false;
+			}
+			else	/* generic expression */
+				*exprs = lappend(*exprs, clause_expr);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1419,15 +1445,21 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		ScalarArrayOpExpr *expr = (ScalarArrayOpExpr *) clause;
 		Node	   *clause_expr;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
+		/* There has to be one expression exactly. */
+		Assert(list_length(clause_exprs) == 1);
+
+		clause_expr = (Node *) linitial(clause_exprs);
+
 		/*
 		 * If it's not one of the supported operators ("=", "<", ">", etc.),
 		 * just ignore the clause, as it's not compatible with MCV lists.
@@ -2013,20 +2045,19 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * examine_opclause_args
  *		Split an operator expression's arguments into Expr and Const parts.
  *
- * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Attempts to match the arguments to either (Expr op Const) or (Const op Expr)
+ * or (Expr op Expr), possibly with a RelabelType on top. When the expression
+ * matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
- * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
+ * non-null pointers (exprsp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
-					  bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	List	   *exprs = NIL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2046,22 +2077,26 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		exprs = lappend(exprs, leftop);
+		exprs = lappend(exprs, rightop);
+		expronleft = false;
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (exprsp)
+		*exprsp = exprs;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index b350fc5f7b..ac7d7b8978 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1644,78 +1644,154 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+			if (cst)	/* Expr op Const */
+			{
+				int idx;
 
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert(list_length(clause_exprs) == 1);
+				clause_expr = (Node *) linitial(clause_exprs);
 
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
-			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
 
-				Assert(idx >= 0);
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(list_length(clause_exprs) == 2);
+
+				clause_expr = (Node *) linitial(clause_exprs);
+				clause_expr2 = (Node *) lsecond(clause_exprs);
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1725,6 +1801,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
 			Oid			collid;
@@ -1742,11 +1819,14 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
 			Assert(expronleft);
+			Assert(list_length(clause_exprs) == 1);
+
+			clause_expr = (Node *) linitial(clause_exprs);
 
 			/* XXX what if (cst->constisnull == NULL)? */
 			if (!cst->constisnull)
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 55cd9252a5..1f30fa9060 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,8 +99,8 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
-								  Const **cstp, bool *expronleftp);
+extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
+								  bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index c60ba45aba..cabc97c43b 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1917,6 +1917,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2064,6 +2076,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      3750 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2473,6 +2497,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 ANALYZE mcv_lists;
@@ -2506,6 +2536,12 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       2500 |   2500
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -2600,6 +2636,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
       1094 |      0
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 ANALYZE mcv_lists_bool;
@@ -2755,6 +2803,18 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 (1 row)
 
 DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2814,6 +2874,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2860,6 +2938,24 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index 6fb37962a7..9210f144cb 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -956,6 +956,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1009,6 +1013,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1208,6 +1216,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1226,6 +1236,8 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1301,6 +1313,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1386,6 +1402,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1413,6 +1433,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
@@ -1427,6 +1450,9 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE b = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AND b = 0 AND c = 0 AND d = 0');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
 
 DROP TABLE mcv_lists_multi;
 
-- 
2.31.1

From 87fb7e95f6694c7f9f79c28fc660211af9cfa15b Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Wed, 18 Aug 2021 11:53:45 +0200
Subject: [PATCH 5/5] Don't treat Var op Var as simple clauses

---
 src/backend/statistics/extended_stats.c       | 86 +++++++++++++++----
 src/backend/statistics/mcv.c                  |  4 +-
 .../statistics/extended_stats_internal.h      |  2 +-
 3 files changed, 74 insertions(+), 18 deletions(-)

diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index e1fcfd0711..bf4aa682ae 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1319,11 +1319,14 @@ choose_best_statistics(List *stats, char requiredkind,
  * statext_is_compatible_clause. It needs to be split like this because
  * of recursion.  The attnums bitmap is an input/output parameter collecting
  * attribute numbers from all compatible clauses (recursively).
+ *
+ * XXX The issimple variable is expected to be initialized by the caller, we
+ * just update it while recursively analyzing the current clause.
  */
 static bool
 statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 									  Index relid, Bitmapset **attnums,
-									  List **exprs)
+									  List **exprs, bool *issimple)
 {
 	/* Look inside any binary-compatible relabeling (as in examine_variable) */
 	if (IsA(clause, RelabelType))
@@ -1371,7 +1374,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		 * Check if the expression has the right shape. This returns either one
 		 * or two expressions, depending on whether there is a Const.
 		 */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/*
@@ -1420,6 +1423,12 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/*
 		 * Check all expressions by recursing. Var expressions are handled as
 		 * a special case (to match it to attnums etc.)
+		 *
+		 * An opclause is simple if it's (Expr op Const) or (Const op Expr). We
+		 * have already checked the overall shape in examine_opclause_args, but
+		 * we haven't checked the expressions are simple (i.e. pretty much Var),
+		 * so we need to check that now. If we discover a complex expression, we
+		 * consider the whole clause complex.
 		 */
 		foreach (lc, clause_exprs)
 		{
@@ -1429,11 +1438,17 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			{
 				/* if the Var is incompatible, the whole clause is incompatible */
 				if (!statext_is_compatible_clause_internal(root, clause_expr,
-														   relid, attnums, exprs))
+														   relid, attnums, exprs,
+														   issimple))
 					return false;
 			}
 			else	/* generic expression */
+			{
 				*exprs = lappend(*exprs, clause_expr);
+
+				/* switch to false if there are any complex clauses */
+				*issimple = false;
+			}
 		}
 
 		return true;
@@ -1452,7 +1467,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/* There has to be one expression exactly. */
@@ -1500,7 +1515,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IN Array clauses by recursing. */
 		if (IsA(clause_expr, Var))
 			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IN Array. */
 		*exprs = lappend(*exprs, clause_expr);
@@ -1529,6 +1545,14 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		BoolExpr   *expr = (BoolExpr *) clause;
 		ListCell   *lc;
 
+		/*
+		 * All AND/OR clauses are considered complex, even if all arguments are
+		 * simple clauses. For NOT clauses we need to check the argument and then
+		 * we can update the flag.
+		 */
+		if (!is_notclause(clause))
+			*issimple = false;
+
 		foreach(lc, expr->args)
 		{
 			/*
@@ -1537,7 +1561,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			 */
 			if (!statext_is_compatible_clause_internal(root,
 													   (Node *) lfirst(lc),
-													   relid, attnums, exprs))
+													   relid, attnums, exprs,
+													   issimple))
 				return false;
 		}
 
@@ -1552,7 +1577,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IS NULL clauses by recursing. */
 		if (IsA(nt->arg, Var))
 			return statext_is_compatible_clause_internal(root, (Node *) (nt->arg),
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IS NULL. */
 		*exprs = lappend(*exprs, nt->arg);
@@ -1588,22 +1614,35 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
  */
 static bool
 statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
-							 Bitmapset **attnums, List **exprs)
+							 Bitmapset **attnums, List **exprs, bool *issimple)
 {
 	RangeTblEntry *rte = root->simple_rte_array[relid];
 	RestrictInfo *rinfo = (RestrictInfo *) clause;
 	int			clause_relid;
 	Oid			userid;
 
+	/*
+	 * Clauses are considered simple by default, and we mark them as complex
+	 * when we discover a complex part.
+	 */
+	*issimple = true;
+
 	/*
 	 * Special-case handling for bare BoolExpr AND clauses, because the
 	 * restrictinfo machinery doesn't build RestrictInfos on top of AND
 	 * clauses.
+	 *
+	 * AND clauses are considered complex, even if all arguments are
+	 * simple clauses.
 	 */
 	if (is_andclause(clause))
 	{
 		BoolExpr   *expr = (BoolExpr *) clause;
 		ListCell   *lc;
+		bool		tmp = false;	/* ignored result */
+
+		/* AND clauses are complex, even if the arguments are simple. */
+		*issimple = false;
 
 		/*
 		 * Check that each sub-clause is compatible.  We expect these to be
@@ -1612,7 +1651,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 		foreach(lc, expr->args)
 		{
 			if (!statext_is_compatible_clause(root, (Node *) lfirst(lc),
-											  relid, attnums, exprs))
+											  relid, attnums, exprs, &tmp))
 				return false;
 		}
 
@@ -1634,7 +1673,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 
 	/* Check the clause and determine what attributes it references. */
 	if (!statext_is_compatible_clause_internal(root, (Node *) rinfo->clause,
-											   relid, attnums, exprs))
+											   relid, attnums, exprs, issimple))
 		return false;
 
 	/*
@@ -1724,6 +1763,7 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	ListCell   *l;
 	Bitmapset **list_attnums;	/* attnums extracted from the clause */
 	List	  **list_exprs;		/* expressions matched to any statistic */
+	bool	  *list_simple;		/* marks simple expressions */
 	int			listidx;
 	Selectivity sel = (is_or) ? 0.0 : 1.0;
 
@@ -1737,6 +1777,9 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	/* expressions extracted from complex expressions */
 	list_exprs = (List **) palloc(sizeof(Node *) * list_length(clauses));
 
+	/* expressions determined to be simple (single expression) */
+	list_simple = (bool *) palloc(sizeof(bool) * list_length(clauses));
+
 	/*
 	 * Pre-process the clauses list to extract the attnums and expressions
 	 * seen in each item.  We need to determine if there are any clauses which
@@ -1754,17 +1797,21 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		Node	   *clause = (Node *) lfirst(l);
 		Bitmapset  *attnums = NULL;
 		List	   *exprs = NIL;
+		bool		issimple = false;
 
 		if (!bms_is_member(listidx, *estimatedclauses) &&
-			statext_is_compatible_clause(root, clause, rel->relid, &attnums, &exprs))
+			statext_is_compatible_clause(root, clause, rel->relid,
+										 &attnums, &exprs, &issimple))
 		{
 			list_attnums[listidx] = attnums;
 			list_exprs[listidx] = exprs;
+			list_simple[listidx] = issimple;
 		}
 		else
 		{
 			list_attnums[listidx] = NULL;
 			list_exprs[listidx] = NIL;
+			list_simple[listidx] = false;
 		}
 
 		listidx++;
@@ -1801,6 +1848,8 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		listidx = -1;
 		foreach(l, clauses)
 		{
+			Node *clause = (Node *) lfirst(l);
+
 			/* Increment the index before we decide if to skip the clause. */
 			listidx++;
 
@@ -1839,13 +1888,12 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 			/* record simple clauses (single column or expression) */
 			if ((list_attnums[listidx] == NULL &&
 				 list_length(list_exprs[listidx]) == 1) ||
-				(list_exprs[listidx] == NIL &&
-				 bms_membership(list_attnums[listidx]) == BMS_SINGLETON))
+				 list_simple[listidx])
 				simple_clauses = bms_add_member(simple_clauses,
 												list_length(stat_clauses));
 
 			/* add clause to list and mark it as estimated */
-			stat_clauses = lappend(stat_clauses, (Node *) lfirst(l));
+			stat_clauses = lappend(stat_clauses, clause);
 			*estimatedclauses = bms_add_member(*estimatedclauses, listidx);
 
 			/*
@@ -2054,13 +2102,15 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp,
+					  bool *issimplep)
 {
 	List	   *exprs = NIL;
 	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
+	bool		issimple;
 
 	/* enforced by statext_is_compatible_clause_internal */
 	Assert(list_length(args) == 2);
@@ -2080,18 +2130,21 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
+		issimple = true;
 	}
 	else if (IsA(leftop, Const))
 	{
 		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
+		issimple = true;
 	}
 	else
 	{
 		exprs = lappend(exprs, leftop);
 		exprs = lappend(exprs, rightop);
 		expronleft = false;
+		issimple = false;
 	}
 
 	/* return pointers to the extracted parts if requested */
@@ -2104,6 +2157,9 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 	if (expronleftp)
 		*expronleftp = expronleft;
 
+	if (issimplep)
+		*issimplep = (*issimplep && issimple);
+
 	return true;
 }
 
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index ac7d7b8978..eb3267b164 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1653,7 +1653,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			if (cst)	/* Expr op Const */
@@ -1819,7 +1819,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 1f30fa9060..d86cc4184b 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -100,7 +100,7 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									int numattrs, AttrNumber *attnums);
 
 extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
-								  bool *expronleftp);
+								  bool *expronleftp, bool *issimplep);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
-- 
2.31.1

From 6287b69f931a44c89c20208c5b41f79117d258f2 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Wed, 20 Jul 2022 11:09:46 +0200
Subject: [PATCH 1/4] Add regression tests for (Var op Var)

A number of queries to test estimation of Var op Var clauses. The number
of new tests is likely an overkill, intended to showcase effects of
parts of this patch series.
---
 src/test/regress/expected/stats_ext.out | 2354 +++++++++++++++++++++--
 src/test/regress/sql/stats_ext.sql      |  650 +++++++
 2 files changed, 2847 insertions(+), 157 deletions(-)

diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 042316aeed8..009c7cfd848 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1988,6 +1988,78 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        343 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = a');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != a');
+ estimated | actual 
+-----------+--------
+      4975 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= a');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < a');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > a');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
+ estimated | actual 
+-----------+--------
+        25 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != c');
+ estimated | actual 
+-----------+--------
+      4975 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= c');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= c');
+ estimated | actual 
+-----------+--------
+      1667 |   1250
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2135,6 +2207,78 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
        200 |    200
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = a');
+ estimated | actual 
+-----------+--------
+      5000 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != a');
+ estimated | actual 
+-----------+--------
+      4975 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= a');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < a');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > a');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
+ estimated | actual 
+-----------+--------
+        25 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != c');
+ estimated | actual 
+-----------+--------
+      4975 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+ estimated | actual 
+-----------+--------
+      1667 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= c');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= c');
+ estimated | actual 
+-----------+--------
+      1667 |   1250
+(1 row)
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
  estimated | actual 
 -----------+--------
@@ -2544,288 +2688,1944 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
       3750 |   2500
 (1 row)
 
--- create statistics
-CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
-ANALYZE mcv_lists;
--- test pg_mcv_list_items with MCV list containing variable-length data and NULLs
-SELECT m.*
-  FROM pg_statistic_ext s, pg_statistic_ext_data d,
-       pg_mcv_list_items(d.stxdmcv) m
- WHERE s.stxname = 'mcv_lists_stats'
-   AND d.stxoid = s.oid;
- index |      values      |  nulls  | frequency | base_frequency 
--------+------------------+---------+-----------+----------------
-     0 | {NULL,x,x}       | {t,f,f} |       0.5 |           0.25
-     1 | {NULL,NULL,NULL} | {t,t,t} |       0.5 |           0.25
-(2 rows)
-
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = ''x'' OR d = ''x''');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
  estimated | actual 
 -----------+--------
-      2500 |   2500
+        25 |   2500
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''x'' OR d = ''x''');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
  estimated | actual 
 -----------+--------
-      2500 |   2500
+         8 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a < a');
  estimated | actual 
 -----------+--------
-      2500 |   2500
+         8 |      0
 (1 row)
 
--- mcv with pass-by-ref fixlen types, e.g. uuid
-CREATE TABLE mcv_lists_uuid (
-    a UUID,
-    b UUID,
-    c UUID
-)
-WITH (autovacuum_enabled = off);
-INSERT INTO mcv_lists_uuid (a, b, c)
-     SELECT
-         md5(mod(i,100)::text)::uuid,
-         md5(mod(i,50)::text)::uuid,
-         md5(mod(i,25)::text)::uuid
-     FROM generate_series(1,5000) s(i);
-ANALYZE mcv_lists_uuid;
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a = a');
  estimated | actual 
 -----------+--------
-         1 |     50
+         1 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND c = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT a = a');
  estimated | actual 
 -----------+--------
-         1 |     50
+        25 |      0
 (1 row)
 
-CREATE STATISTICS mcv_lists_uuid_stats (mcv) ON a, b, c
-  FROM mcv_lists_uuid;
-ANALYZE mcv_lists_uuid;
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
  estimated | actual 
 -----------+--------
-        50 |     50
+      1683 |   2500
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND c = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
  estimated | actual 
 -----------+--------
-        50 |     50
+      1683 |   2500
 (1 row)
 
-DROP TABLE mcv_lists_uuid;
--- mcv with arrays
-CREATE TABLE mcv_lists_arrays (
-    a TEXT[],
-    b NUMERIC[],
-    c INT[]
-)
-WITH (autovacuum_enabled = off);
-INSERT INTO mcv_lists_arrays (a, b, c)
-     SELECT
-         ARRAY[md5((i/100)::text), md5((i/100-1)::text), md5((i/100+1)::text)],
-         ARRAY[(i/100-1)::numeric/1000, (i/100)::numeric/1000, (i/100+1)::numeric/1000],
-         ARRAY[(i/100-1), i/100, (i/100+1)]
-     FROM generate_series(1,5000) s(i);
-CREATE STATISTICS mcv_lists_arrays_stats (mcv) ON a, b, c
-  FROM mcv_lists_arrays;
-ANALYZE mcv_lists_arrays;
--- mcv with bool
-CREATE TABLE mcv_lists_bool (
-    a BOOL,
-    b BOOL,
-    c BOOL
-)
-WITH (autovacuum_enabled = off);
-INSERT INTO mcv_lists_bool (a, b, c)
-     SELECT
-         (mod(i,2) = 0), (mod(i,4) = 0), (mod(i,8) = 0)
-     FROM generate_series(1,10000) s(i);
-ANALYZE mcv_lists_bool;
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE a AND b AND c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
  estimated | actual 
 -----------+--------
-       156 |   1250
+        50 |   2500
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
  estimated | actual 
 -----------+--------
-       156 |      0
+      4975 |   2500
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND NOT b AND c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
  estimated | actual 
 -----------+--------
-       469 |      0
+      4975 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
  estimated | actual 
 -----------+--------
-      1094 |      0
+      1658 |      0
 (1 row)
 
-CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
-  FROM mcv_lists_bool;
-ANALYZE mcv_lists_bool;
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE a AND b AND c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a < a');
  estimated | actual 
 -----------+--------
-      1250 |   1250
+      1658 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a = a');
  estimated | actual 
 -----------+--------
          1 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND NOT b AND c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT a = a');
  estimated | actual 
 -----------+--------
-         1 |      0
+      4950 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
  estimated | actual 
 -----------+--------
-         1 |      0
+      4983 |      0
 (1 row)
 
--- mcv covering just a small fraction of data
-CREATE TABLE mcv_lists_partial (
-    a INT,
-    b INT,
-    c INT
-);
--- 10 frequent groups, each with 100 elements
-INSERT INTO mcv_lists_partial (a, b, c)
-     SELECT
-         mod(i,10),
-         mod(i,10),
-         mod(i,10)
-     FROM generate_series(0,999) s(i);
--- 100 groups that will make it to the MCV list (includes the 10 frequent ones)
-INSERT INTO mcv_lists_partial (a, b, c)
-     SELECT
-         i,
-         i,
-         i
-     FROM generate_series(0,99) s(i);
--- 4000 groups in total, most of which won't make it (just a single item)
-INSERT INTO mcv_lists_partial (a, b, c)
-     SELECT
-         i,
-         i,
-         i
-     FROM generate_series(0,3999) s(i);
-ANALYZE mcv_lists_partial;
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
  estimated | actual 
 -----------+--------
-         1 |    102
+      4983 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
  estimated | actual 
 -----------+--------
-       300 |    102
+      4975 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 AND b = 10 AND c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
  estimated | actual 
 -----------+--------
-         1 |      2
+      5000 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 OR b = 10 OR c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
  estimated | actual 
 -----------+--------
-         6 |      2
+      1667 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
  estimated | actual 
 -----------+--------
-         1 |      0
+       556 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a < a');
  estimated | actual 
 -----------+--------
-       204 |    104
+       556 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0 AND c = 0) OR (a = 1 AND b = 1 AND c = 1) OR (a = 2 AND b = 2 AND c = 2)');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a = a');
  estimated | actual 
 -----------+--------
-         1 |    306
+         1 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0) OR (a = 0 AND c = 0) OR (b = 0 AND c = 0)');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT a = a');
  estimated | actual 
 -----------+--------
-         6 |    102
+      1658 |      0
 (1 row)
 
-CREATE STATISTICS mcv_lists_partial_stats (mcv) ON a, b, c
-  FROM mcv_lists_partial;
-ANALYZE mcv_lists_partial;
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
  estimated | actual 
 -----------+--------
-       102 |    102
+      2778 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 0');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
  estimated | actual 
 -----------+--------
-        96 |    102
+      2778 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 AND b = 10 AND c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
  estimated | actual 
 -----------+--------
-         2 |      2
+      1683 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 OR b = 10 OR c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
  estimated | actual 
 -----------+--------
-         2 |      2
+      4983 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
  estimated | actual 
 -----------+--------
-         1 |      0
+      1667 |   2500
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 10');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
  estimated | actual 
 -----------+--------
-       102 |    104
+       556 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0 AND c = 0) OR (a = 1 AND b = 1 AND c = 1) OR (a = 2 AND b = 2 AND c = 2)');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a < a');
  estimated | actual 
 -----------+--------
-       306 |    306
+       556 |      0
 (1 row)
 
-SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0) OR (a = 0 AND c = 0) OR (b = 0 AND c = 0)');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a = a');
  estimated | actual 
 -----------+--------
-       108 |    102
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a < a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a < a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
+ estimated | actual 
+-----------+--------
+         8 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c < c');
+ estimated | actual 
+-----------+--------
+         8 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
+ estimated | actual 
+-----------+--------
+        12 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+        25 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4975 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+ estimated | actual 
+-----------+--------
+      4975 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c < c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
+ estimated | actual 
+-----------+--------
+      2488 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
+ estimated | actual 
+-----------+--------
+      4975 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+-- create statistics
+CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
+ANALYZE mcv_lists;
+-- test pg_mcv_list_items with MCV list containing variable-length data and NULLs
+SELECT m.*
+  FROM pg_statistic_ext s, pg_statistic_ext_data d,
+       pg_mcv_list_items(d.stxdmcv) m
+ WHERE s.stxname = 'mcv_lists_stats'
+   AND d.stxoid = s.oid;
+ index |      values      |  nulls  | frequency | base_frequency 
+-------+------------------+---------+-----------+----------------
+     0 | {NULL,x,x}       | {t,f,f} |       0.5 |           0.25
+     1 | {NULL,NULL,NULL} | {t,t,t} |       0.5 |           0.25
+(2 rows)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = ''x'' OR d = ''x''');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''x'' OR d = ''x''');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
+ estimated | actual 
+-----------+--------
+      2500 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+ estimated | actual 
+-----------+--------
+        25 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
+ estimated | actual 
+-----------+--------
+         8 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a < a');
+ estimated | actual 
+-----------+--------
+         8 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+        25 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4975 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+ estimated | actual 
+-----------+--------
+      4975 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a < a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
+ estimated | actual 
+-----------+--------
+      4975 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a < a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a < a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a < a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a < a');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a = a');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT a = a');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
+ estimated | actual 
+-----------+--------
+         8 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c < c');
+ estimated | actual 
+-----------+--------
+         8 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
+ estimated | actual 
+-----------+--------
+        12 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+        25 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4975 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+ estimated | actual 
+-----------+--------
+      4975 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c < c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
+ estimated | actual 
+-----------+--------
+      2488 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
+ estimated | actual 
+-----------+--------
+      4975 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+ estimated | actual 
+-----------+--------
+      1667 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c < c');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
+ estimated | actual 
+-----------+--------
+       833 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
+ estimated | actual 
+-----------+--------
+      1658 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
+ estimated | actual 
+-----------+--------
+      2778 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
+ estimated | actual 
+-----------+--------
+      1683 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
+ estimated | actual 
+-----------+--------
+      4983 |   2500
+(1 row)
+
+-- mcv with pass-by-ref fixlen types, e.g. uuid
+CREATE TABLE mcv_lists_uuid (
+    a UUID,
+    b UUID,
+    c UUID
+)
+WITH (autovacuum_enabled = off);
+INSERT INTO mcv_lists_uuid (a, b, c)
+     SELECT
+         md5(mod(i,100)::text)::uuid,
+         md5(mod(i,50)::text)::uuid,
+         md5(mod(i,25)::text)::uuid
+     FROM generate_series(1,5000) s(i);
+ANALYZE mcv_lists_uuid;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+ estimated | actual 
+-----------+--------
+         1 |     50
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND c = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+ estimated | actual 
+-----------+--------
+         1 |     50
+(1 row)
+
+CREATE STATISTICS mcv_lists_uuid_stats (mcv) ON a, b, c
+  FROM mcv_lists_uuid;
+ANALYZE mcv_lists_uuid;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+ estimated | actual 
+-----------+--------
+        50 |     50
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_uuid WHERE a = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND b = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc'' AND c = ''1679091c-5a88-0faf-6fb5-e6087eb1b2dc''');
+ estimated | actual 
+-----------+--------
+        50 |     50
+(1 row)
+
+DROP TABLE mcv_lists_uuid;
+-- mcv with arrays
+CREATE TABLE mcv_lists_arrays (
+    a TEXT[],
+    b NUMERIC[],
+    c INT[]
+)
+WITH (autovacuum_enabled = off);
+INSERT INTO mcv_lists_arrays (a, b, c)
+     SELECT
+         ARRAY[md5((i/100)::text), md5((i/100-1)::text), md5((i/100+1)::text)],
+         ARRAY[(i/100-1)::numeric/1000, (i/100)::numeric/1000, (i/100+1)::numeric/1000],
+         ARRAY[(i/100-1), i/100, (i/100+1)]
+     FROM generate_series(1,5000) s(i);
+CREATE STATISTICS mcv_lists_arrays_stats (mcv) ON a, b, c
+  FROM mcv_lists_arrays;
+ANALYZE mcv_lists_arrays;
+-- mcv with bool
+CREATE TABLE mcv_lists_bool (
+    a BOOL,
+    b BOOL,
+    c BOOL
+)
+WITH (autovacuum_enabled = off);
+INSERT INTO mcv_lists_bool (a, b, c)
+     SELECT
+         (mod(i,2) = 0), (mod(i,4) = 0), (mod(i,8) = 0)
+     FROM generate_series(1,10000) s(i);
+ANALYZE mcv_lists_bool;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE a AND b AND c');
+ estimated | actual 
+-----------+--------
+       156 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND c');
+ estimated | actual 
+-----------+--------
+       156 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND NOT b AND c');
+ estimated | actual 
+-----------+--------
+       469 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
+ estimated | actual 
+-----------+--------
+      1094 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a');
+ estimated | actual 
+-----------+--------
+      9950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a');
+ estimated | actual 
+-----------+--------
+     10000 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a');
+ estimated | actual 
+-----------+--------
+      3333 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a');
+ estimated | actual 
+-----------+--------
+      3333 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a');
+ estimated | actual 
+-----------+--------
+      3333 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a');
+ estimated | actual 
+-----------+--------
+      3333 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b');
+ estimated | actual 
+-----------+--------
+        50 |   7500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b');
+ estimated | actual 
+-----------+--------
+      3333 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b');
+ estimated | actual 
+-----------+--------
+      3333 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b');
+ estimated | actual 
+-----------+--------
+      3333 |   7500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b');
+ estimated | actual 
+-----------+--------
+      3333 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
+ estimated | actual 
+-----------+--------
+      9950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a AND b != b');
+ estimated | actual 
+-----------+--------
+      9950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a AND b < b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b AND b != c');
+ estimated | actual 
+-----------+--------
+        50 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b AND b < c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b AND b <= c');
+ estimated | actual 
+-----------+--------
+      1111 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b AND b > c');
+ estimated | actual 
+-----------+--------
+      1111 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b AND b >= c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c AND b = b');
+ estimated | actual 
+-----------+--------
+      9950 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
+ estimated | actual 
+-----------+--------
+        50 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < c AND b < b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
+ estimated | actual 
+-----------+--------
+      1111 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
+  FROM mcv_lists_bool;
+ANALYZE mcv_lists_bool;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE a AND b AND c');
+ estimated | actual 
+-----------+--------
+      1250 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND NOT b AND c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a');
+ estimated | actual 
+-----------+--------
+      9950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a');
+ estimated | actual 
+-----------+--------
+     10000 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a');
+ estimated | actual 
+-----------+--------
+      3333 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a');
+ estimated | actual 
+-----------+--------
+      3333 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a');
+ estimated | actual 
+-----------+--------
+      3333 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a');
+ estimated | actual 
+-----------+--------
+      3333 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b');
+ estimated | actual 
+-----------+--------
+        50 |   7500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b');
+ estimated | actual 
+-----------+--------
+      3333 |  10000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b');
+ estimated | actual 
+-----------+--------
+      3333 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b');
+ estimated | actual 
+-----------+--------
+      3333 |   7500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b');
+ estimated | actual 
+-----------+--------
+      3333 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
+ estimated | actual 
+-----------+--------
+      9950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a AND b != b');
+ estimated | actual 
+-----------+--------
+      9950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a AND b < b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b AND b != c');
+ estimated | actual 
+-----------+--------
+        50 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b AND b < c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b AND b <= c');
+ estimated | actual 
+-----------+--------
+      1111 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b AND b > c');
+ estimated | actual 
+-----------+--------
+      1111 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b AND b >= c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c AND b = b');
+ estimated | actual 
+-----------+--------
+      9950 |   3750
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
+ estimated | actual 
+-----------+--------
+        50 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < c AND b < b');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
+ estimated | actual 
+-----------+--------
+      1111 |   1250
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
+ estimated | actual 
+-----------+--------
+      1111 |      0
+(1 row)
+
+-- mcv covering just a small fraction of data
+CREATE TABLE mcv_lists_partial (
+    a INT,
+    b INT,
+    c INT
+);
+-- 10 frequent groups, each with 100 elements
+INSERT INTO mcv_lists_partial (a, b, c)
+     SELECT
+         mod(i,10),
+         mod(i,10),
+         mod(i,10)
+     FROM generate_series(0,999) s(i);
+-- 100 groups that will make it to the MCV list (includes the 10 frequent ones)
+INSERT INTO mcv_lists_partial (a, b, c)
+     SELECT
+         i,
+         i,
+         i
+     FROM generate_series(0,99) s(i);
+-- 4000 groups in total, most of which won't make it (just a single item)
+INSERT INTO mcv_lists_partial (a, b, c)
+     SELECT
+         i,
+         i,
+         i
+     FROM generate_series(0,3999) s(i);
+ANALYZE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 0');
+ estimated | actual 
+-----------+--------
+         1 |    102
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 0');
+ estimated | actual 
+-----------+--------
+       300 |    102
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 AND b = 10 AND c = 10');
+ estimated | actual 
+-----------+--------
+         1 |      2
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 OR b = 10 OR c = 10');
+ estimated | actual 
+-----------+--------
+         6 |      2
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 10');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 10');
+ estimated | actual 
+-----------+--------
+       204 |    104
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0 AND c = 0) OR (a = 1 AND b = 1 AND c = 1) OR (a = 2 AND b = 2 AND c = 2)');
+ estimated | actual 
+-----------+--------
+         1 |    306
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0) OR (a = 0 AND c = 0) OR (b = 0 AND c = 0)');
+ estimated | actual 
+-----------+--------
+         6 |    102
+(1 row)
+
+CREATE STATISTICS mcv_lists_partial_stats (mcv) ON a, b, c
+  FROM mcv_lists_partial;
+ANALYZE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 0');
+ estimated | actual 
+-----------+--------
+       102 |    102
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 0');
+ estimated | actual 
+-----------+--------
+        96 |    102
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 AND b = 10 AND c = 10');
+ estimated | actual 
+-----------+--------
+         2 |      2
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 10 OR b = 10 OR c = 10');
+ estimated | actual 
+-----------+--------
+         2 |      2
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 AND b = 0 AND c = 10');
+ estimated | actual 
+-----------+--------
+         1 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE a = 0 OR b = 0 OR c = 10');
+ estimated | actual 
+-----------+--------
+       102 |    104
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0 AND c = 0) OR (a = 1 AND b = 1 AND c = 1) OR (a = 2 AND b = 2 AND c = 2)');
+ estimated | actual 
+-----------+--------
+       306 |    306
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0 AND b = 0) OR (a = 0 AND c = 0) OR (b = 0 AND c = 0)');
+ estimated | actual 
+-----------+--------
+       108 |    102
+(1 row)
+
+DROP TABLE mcv_lists_partial;
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+ estimated | actual 
+-----------+--------
+      9950 |   2500
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+ estimated | actual 
+-----------+--------
+        50 |   2500
 (1 row)
 
-DROP TABLE mcv_lists_partial;
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -2885,6 +4685,126 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       2649 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != d');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b AND c > d');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND c >= d');
+ estimated | actual 
+-----------+--------
+       556 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b OR c = d');
+ estimated | actual 
+-----------+--------
+        50 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != d');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b OR c > d');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR c >= d');
+ estimated | actual 
+-----------+--------
+      2778 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AND c = d');
+ estimated | actual 
+-----------+--------
+        25 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AND c > d');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
+ estimated | actual 
+-----------+--------
+       556 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
+ estimated | actual 
+-----------+--------
+        50 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
+ estimated | actual 
+-----------+--------
+      2778 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a > b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a >= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -2931,6 +4851,126 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
       1571 |   1572
 (1 row)
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+ estimated | actual 
+-----------+--------
+         1 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != d');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b AND c > d');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND c >= d');
+ estimated | actual 
+-----------+--------
+       556 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b OR c = d');
+ estimated | actual 
+-----------+--------
+        50 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != d');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b OR c > d');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR c >= d');
+ estimated | actual 
+-----------+--------
+      2778 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AND c = d');
+ estimated | actual 
+-----------+--------
+        25 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
+ estimated | actual 
+-----------+--------
+      4950 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AND c > d');
+ estimated | actual 
+-----------+--------
+       556 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
+ estimated | actual 
+-----------+--------
+       556 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
+ estimated | actual 
+-----------+--------
+        50 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
+ estimated | actual 
+-----------+--------
+      5000 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
+ estimated | actual 
+-----------+--------
+      2778 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
+ estimated | actual 
+-----------+--------
+      2778 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a > b');
+ estimated | actual 
+-----------+--------
+      1667 |      0
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a >= b');
+ estimated | actual 
+-----------+--------
+      1667 |   5000
+(1 row)
+
 DROP TABLE mcv_lists_multi;
 -- statistics on integer expressions
 CREATE TABLE expr_stats (a int, b int, c int);
diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql
index 6b954c9e500..1b4d57e0462 100644
--- a/src/test/regress/sql/stats_ext.sql
+++ b/src/test/regress/sql/stats_ext.sql
@@ -989,6 +989,30 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1042,6 +1066,30 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = ''1'' OR c = 1 OR d IS NOT NULL');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= c');
+
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52, NULL) AND b IN ( ''1'', ''2'', NULL)');
@@ -1241,6 +1289,221 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
+
+
 -- create statistics
 CREATE STATISTICS mcv_lists_stats (mcv) ON a, b, d FROM mcv_lists;
 
@@ -1259,6 +1522,219 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = 1 OR b = '
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND (b = ''x'' OR d = ''x'')');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
+
 -- mcv with pass-by-ref fixlen types, e.g. uuid
 CREATE TABLE mcv_lists_uuid (
     a UUID,
@@ -1334,6 +1810,66 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a AND b != b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a AND b < b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b AND b != c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b AND b < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b AND b <= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b AND b > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b AND b >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c AND b = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < c AND b < b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
+
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
   FROM mcv_lists_bool;
 
@@ -1347,6 +1883,66 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND b AND NOT c');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a AND b != b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a AND b < b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b AND b != c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b AND b < c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b AND b <= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b AND b > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b AND b >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c AND b = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < c AND b < b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
+
 -- mcv covering just a small fraction of data
 CREATE TABLE mcv_lists_partial (
     a INT,
@@ -1419,6 +2015,10 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_partial WHERE (a = 0
 
 DROP TABLE mcv_lists_partial;
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
+
 -- check the ability to use multiple MCV lists
 CREATE TABLE mcv_lists_multi (
 	a INTEGER,
@@ -1447,6 +2047,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AN
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b AND c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND c >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b OR c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b OR c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR c >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AND c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a > b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a >= b');
+
 -- create separate MCV statistics
 CREATE STATISTICS mcv_lists_multi_1 (mcv) ON a, b FROM mcv_lists_multi;
 CREATE STATISTICS mcv_lists_multi_2 (mcv) ON c, d FROM mcv_lists_multi;
@@ -1461,6 +2086,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 AN
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE (a = 0 AND b = 0) OR (c = 0 AND d = 0)');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR b = 0 OR c = 0 OR d = 0');
 
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b AND c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND c >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b OR c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b OR c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR c >= d');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AND c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AND c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
+
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a > b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
+SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a >= b');
+
 DROP TABLE mcv_lists_multi;
 
 
-- 
2.34.3

From 5c73433f80cace5aefbac5f913e9e69379220632 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Thu, 26 Aug 2021 23:01:04 +0200
Subject: [PATCH 2/4] Improve Var op Var estimates with the same Var

When estimating (Var op Var) clauses, we can treat the case with the
same Var on both sides as a special case, and we can provide better
selectivity estimate than for the generic case.

For example for (a = a) clause we know it's (1.0 - nullfrac), because
all non-NULL rows are expected to match. Similarly for (a != a), wich
has selectivity 0.0. And the same reasoning can be applied to inequality
comparisons, like (a < a) etc.

This of course depends on the operators having the usual meaning (e.g.
"=" equality) and not some custom operator with different behavior, but
if they use the same estimator it's reasonable to assume that (we
already do that in various other places).

In principle, such clauses are a bit strange and queries are unlikely to
use them very often. But query generators sometimes do this, and these
checks are quite cheap so it's likely a win.
---
 src/backend/utils/adt/selfuncs.c        | 103 +++++-
 src/test/regress/expected/stats_ext.out | 436 ++++++++++++------------
 2 files changed, 320 insertions(+), 219 deletions(-)

diff --git a/src/backend/utils/adt/selfuncs.c b/src/backend/utils/adt/selfuncs.c
index d35e5605de8..9289e9b8da8 100644
--- a/src/backend/utils/adt/selfuncs.c
+++ b/src/backend/utils/adt/selfuncs.c
@@ -210,7 +210,8 @@ static bool get_actual_variable_endpoint(Relation heapRel,
 										 MemoryContext outercontext,
 										 Datum *endpointDatum);
 static RelOptInfo *find_join_input_rel(PlannerInfo *root, Relids relids);
-
+static bool matching_restriction_variables(PlannerInfo *root, List *args,
+										   int varRelid);
 
 /*
  *		eqsel			- Selectivity of "=" for any data types.
@@ -256,6 +257,31 @@ eqsel_internal(PG_FUNCTION_ARGS, bool negate)
 		}
 	}
 
+	/*
+	 * If it's (variable = variable) with the same variable on both sides, it's
+	 * a special case and we know it's not expected to filter anything, so we
+	 * estimate the selectivity as (1.0 - nullfrac), or 0.0 when it's negated.
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+	{
+		VariableStatData	vardata;
+		double				null_frac = 0.0;
+
+		examine_variable(root, linitial(args), varRelid, &vardata);
+
+		if (HeapTupleIsValid(vardata.statsTuple))
+		{
+			Form_pg_statistic	stats;
+
+			stats = (Form_pg_statistic) GETSTRUCT(vardata.statsTuple);
+			null_frac = stats->stanullfrac;
+		}
+
+		ReleaseVariableStats(vardata);
+
+		return (negate) ? 0.0 : (1.0 - null_frac);
+	}
+
 	/*
 	 * If expression is not variable = something or something = variable, then
 	 * punt and return a default estimate.
@@ -1408,6 +1434,22 @@ scalarineqsel_wrapper(PG_FUNCTION_ARGS, bool isgt, bool iseq)
 	Oid			consttype;
 	double		selec;
 
+	/*
+	 * Handle (variable < variable) and (variable <= variable) with the same
+	 * variable on both sides as a special case. The strict inequality should
+	 * not match any rows, hence selectivity is 0.0. The other case is about
+	 * the same as equality, so selectivity is 1.0.
+	 */
+	if (matching_restriction_variables(root, args, varRelid))
+	{
+		/* The case with equality matches all rows, so estimate it as 1.0. */
+		if (iseq)
+			PG_RETURN_FLOAT8(1.0);
+
+		/* Strict inequality matches nothing, so selectivity is 0.0. */
+		PG_RETURN_FLOAT8(0.0);
+	}
+
 	/*
 	 * If expression is not variable op something or something op variable,
 	 * then punt and return a default estimate.
@@ -4906,6 +4948,65 @@ get_restriction_variable(PlannerInfo *root, List *args, int varRelid,
 	return false;
 }
 
+/*
+ * matching_restriction_variable
+ *		Check if the two arguments of a restriction clause refer to the same
+ *		variable, i.e. if the condition is of the form (variable op variable).
+ *		We can deduce selectivity for such (in)equality clauses.
+ *
+ * Inputs:
+ *	root: the planner info
+ *	args: clause argument list
+ *	varRelid: see specs for restriction selectivity functions
+ *
+ * Returns true if the same variable is on both sides, otherwise false.
+ */
+static bool
+matching_restriction_variables(PlannerInfo *root, List *args, int varRelid)
+{
+	Node	   *left,
+			   *right;
+	Bitmapset  *varnos;
+
+	/* Fail if not a binary opclause (probably shouldn't happen) */
+	if (list_length(args) != 2)
+		return false;
+
+	left = (Node *) linitial(args);
+	right = (Node *) lsecond(args);
+
+	/* Look inside any binary-compatible relabeling */
+
+	if (IsA(left, RelabelType))
+		left = (Node *) ((RelabelType *) left)->arg;
+
+	if (IsA(right, RelabelType))
+		right = (Node *) ((RelabelType *) right)->arg;
+
+	/*
+	 * Check if it's safe to compare the two expressions. The expressions
+	 * must not contain any volatile expressions, and must belong to the
+	 * same relation.
+	 *
+	 * XXX We do a small trick - we validate just one expression, and then
+	 * check it's equal to the other side.
+	 */
+	if (contain_volatile_functions(left))
+		return false;
+
+	/* Check it belongs to a single relation. */
+	varnos = pull_varnos(root, left);
+
+	if (bms_num_members(varnos) != 1)
+		return false;
+
+	if ((varRelid != 0) && (!bms_is_member(varRelid, varnos)))
+		return false;
+
+	/* The two variables need to match */
+	return equal(left, right);
+}
+
 /*
  * get_join_variables
  *		Apply examine_variable() to each side of a join clause.
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index 009c7cfd848..fea3d18442c 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -1997,31 +1997,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = a');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != a');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= a');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
@@ -2216,31 +2216,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = a');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != a');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= a');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
@@ -2697,13 +2697,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a < a');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a = a');
@@ -2715,31 +2715,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT a = a');
  estimated | actual 
 -----------+--------
-        25 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
  estimated | actual 
 -----------+--------
-        50 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4975 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
@@ -2751,13 +2751,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a < a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a = a');
@@ -2769,19 +2769,19 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
@@ -2793,7 +2793,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      5000 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
@@ -2805,13 +2805,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a = a');
@@ -2823,31 +2823,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
@@ -2859,13 +2859,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a = a');
@@ -2877,31 +2877,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
@@ -2913,13 +2913,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a = a');
@@ -2931,31 +2931,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
@@ -2967,13 +2967,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a = a');
@@ -2985,43 +2985,43 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c < c');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
@@ -3033,31 +3033,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
  estimated | actual 
 -----------+--------
-        25 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
  estimated | actual 
 -----------+--------
-        50 |   2500
+      2513 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4975 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
@@ -3069,13 +3069,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c < c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
@@ -3087,31 +3087,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
  estimated | actual 
 -----------+--------
-      4975 |   2500
+      4988 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      5000 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
@@ -3123,13 +3123,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
@@ -3141,31 +3141,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
@@ -3177,13 +3177,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
@@ -3195,31 +3195,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
@@ -3231,13 +3231,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
@@ -3249,31 +3249,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
@@ -3285,13 +3285,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
@@ -3303,31 +3303,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 -- create statistics
@@ -3372,13 +3372,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a < a');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a = a');
@@ -3390,31 +3390,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT a = a');
  estimated | actual 
 -----------+--------
-        25 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
  estimated | actual 
 -----------+--------
-        50 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4975 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
@@ -3426,13 +3426,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a < a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a = a');
@@ -3444,19 +3444,19 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
@@ -3468,7 +3468,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      5000 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
@@ -3480,13 +3480,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a = a');
@@ -3498,31 +3498,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
@@ -3534,13 +3534,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a = a');
@@ -3552,31 +3552,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
@@ -3588,13 +3588,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a = a');
@@ -3606,31 +3606,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
@@ -3642,13 +3642,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a < a');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a = a');
@@ -3660,43 +3660,43 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT a = a');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c < c');
  estimated | actual 
 -----------+--------
-         8 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
@@ -3708,31 +3708,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
  estimated | actual 
 -----------+--------
-        25 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
  estimated | actual 
 -----------+--------
-        50 |   2500
+      2513 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4975 |   2500
+        25 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
@@ -3744,13 +3744,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c < c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
@@ -3762,31 +3762,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
  estimated | actual 
 -----------+--------
-      4975 |   2500
+      4988 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      5000 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
@@ -3798,13 +3798,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
@@ -3816,31 +3816,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
@@ -3852,13 +3852,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
@@ -3870,31 +3870,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
@@ -3906,13 +3906,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
@@ -3924,31 +3924,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
@@ -3960,13 +3960,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c < c');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
@@ -3978,31 +3978,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c =
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
  estimated | actual 
 -----------+--------
-      1658 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
  estimated | actual 
 -----------+--------
-      2778 |   2500
+      1667 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
  estimated | actual 
 -----------+--------
-      1683 |   2500
+      3333 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4983 |   2500
+      1667 |   2500
 (1 row)
 
 -- mcv with pass-by-ref fixlen types, e.g. uuid
@@ -4102,7 +4102,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a');
  estimated | actual 
 -----------+--------
-      9950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a');
@@ -4114,25 +4114,25 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a !=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a');
  estimated | actual 
 -----------+--------
-      3333 |  10000
+     10000 |  10000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a');
  estimated | actual 
 -----------+--------
-      3333 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a');
  estimated | actual 
 -----------+--------
-      3333 |  10000
+     10000 |  10000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a');
  estimated | actual 
 -----------+--------
-      3333 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
@@ -4174,37 +4174,37 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
  estimated | actual 
 -----------+--------
-      9950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a AND b != b');
  estimated | actual 
 -----------+--------
-      9950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a AND b < b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
@@ -4252,31 +4252,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
  estimated | actual 
 -----------+--------
-        50 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < c AND b < b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= c');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
  estimated | actual 
 -----------+--------
-      1111 |   1250
+      3333 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 CREATE STATISTICS mcv_lists_bool_stats (mcv) ON a, b, c
@@ -4309,7 +4309,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a AND
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a');
  estimated | actual 
 -----------+--------
-      9950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a');
@@ -4321,25 +4321,25 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a !=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a');
  estimated | actual 
 -----------+--------
-      3333 |  10000
+     10000 |  10000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a');
  estimated | actual 
 -----------+--------
-      3333 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a');
  estimated | actual 
 -----------+--------
-      3333 |  10000
+     10000 |  10000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a');
  estimated | actual 
 -----------+--------
-      3333 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
@@ -4381,37 +4381,37 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
  estimated | actual 
 -----------+--------
-      9950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != a AND b != b');
  estimated | actual 
 -----------+--------
-      9950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < a AND b < b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
@@ -4459,31 +4459,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
  estimated | actual 
 -----------+--------
-        50 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < c AND b < b');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= a AND b <= c');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
  estimated | actual 
 -----------+--------
-      1111 |   1250
+      3333 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 -- mcv covering just a small fraction of data
@@ -4742,13 +4742,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AN
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AND c > d');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
@@ -4760,19 +4760,19 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b A
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
  estimated | actual 
 -----------+--------
-        50 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
  estimated | actual 
 -----------+--------
-      5000 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
@@ -4908,13 +4908,13 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AN
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AND c > d');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
@@ -4926,19 +4926,19 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b A
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
  estimated | actual 
 -----------+--------
-        50 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
  estimated | actual 
 -----------+--------
-      5000 |      0
+      4975 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
  estimated | actual 
 -----------+--------
-      2778 |      0
+      1667 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
-- 
2.34.3

From 0848ae05d4d1e1c2fbefaa5a43a4910efe9bfe98 Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Sun, 15 Aug 2021 13:23:13 +0200
Subject: [PATCH 3/4] Modify selectivity estimation for Var op Var

This modifies estimation of (Var op Var) clauses with distinct Vars from
the same relation. We recognize the clauses as compatible with extended
statistics, and estimate them using the extended MCV etc.
---
 src/backend/optimizer/path/clausesel.c        |  37 +++-
 src/backend/statistics/extended_stats.c       |  83 +++++--
 src/backend/statistics/mcv.c                  | 172 +++++++++++----
 .../statistics/extended_stats_internal.h      |   4 +-
 src/test/regress/expected/stats_ext.out       | 208 +++++++++---------
 5 files changed, 323 insertions(+), 181 deletions(-)

diff --git a/src/backend/optimizer/path/clausesel.c b/src/backend/optimizer/path/clausesel.c
index 06f836308d0..90ee73f89df 100644
--- a/src/backend/optimizer/path/clausesel.c
+++ b/src/backend/optimizer/path/clausesel.c
@@ -714,6 +714,7 @@ clause_selectivity_ext(PlannerInfo *root,
 	Selectivity s1 = 0.5;		/* default for any unhandled clause type */
 	RestrictInfo *rinfo = NULL;
 	bool		cacheable = false;
+	Node	   *src = clause;
 
 	if (clause == NULL)			/* can this still happen? */
 		return s1;
@@ -871,11 +872,37 @@ clause_selectivity_ext(PlannerInfo *root,
 		}
 		else
 		{
-			/* Estimate selectivity for a restriction clause. */
-			s1 = restriction_selectivity(root, opno,
-										 opclause->args,
-										 opclause->inputcollid,
-										 varRelid);
+			/*
+			 * It might be a single (Expr op Expr) clause, which goes here due
+			 * to the optimization at the beginning of clauselist_selectivity.
+			 * So we try applying extended stats first, and then fall back to
+			 * restriction_selectivity.
+			 */
+			bool	estimated = false;
+
+			if (use_extended_stats)
+			{
+				List	   *clauses = list_make1(src);
+				RelOptInfo *rel = find_single_rel_for_clauses(root, clauses);
+
+				if (rel && rel->rtekind == RTE_RELATION && rel->statlist != NIL)
+				{
+					Bitmapset  *estimatedclauses = NULL;
+
+					s1 = statext_clauselist_selectivity(root, clauses, varRelid,
+														jointype, sjinfo, rel,
+														&estimatedclauses, false);
+
+					estimated = (bms_num_members(estimatedclauses) == 1);
+				}
+			}
+
+			/* Estimate selectivity for a restriction clause (fallback). */
+			if (!estimated)
+				s1 = restriction_selectivity(root, opno,
+											 opclause->args,
+											 opclause->inputcollid,
+											 varRelid);
 		}
 
 		/*
diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index d2aa8d0ca3f..7bfde068aff 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1352,19 +1352,27 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		return true;
 	}
 
-	/* (Var/Expr op Const) or (Const op Var/Expr) */
+	/*
+	 * Three opclause variants are supported: (Expr op Const), (Const op Expr),
+	 * (Expr op Expr). That means we may need to analyze one or two expressions
+	 * to make sure the opclause is compatible with extended stats.
+	 */
 	if (is_opclause(clause))
 	{
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		OpExpr	   *expr = (OpExpr *) clause;
-		Node	   *clause_expr;
+		ListCell   *lc;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
-		/* Check if the expression has the right shape */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		/*
+		 * Check if the expression has the right shape. This returns either one
+		 * or two expressions, depending on whether there is a Const.
+		 */
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
 		/*
@@ -1404,13 +1412,31 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			!get_func_leakproof(get_opcode(expr->opno)))
 			return false;
 
-		/* Check (Var op Const) or (Const op Var) clauses by recursing. */
-		if (IsA(clause_expr, Var))
-			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+		/*
+		 * There's always at least one expression, otherwise the clause would
+		 * not be considered compatible.
+		 */
+		Assert(list_length(clause_exprs) >= 1);
+
+		/*
+		 * Check all expressions by recursing. Var expressions are handled as
+		 * a special case (to match it to attnums etc.)
+		 */
+		foreach (lc, clause_exprs)
+		{
+			Node *clause_expr = (Node *) lfirst(lc);
+
+			if (IsA(clause_expr, Var))
+			{
+				/* if the Var is incompatible, the whole clause is incompatible */
+				if (!statext_is_compatible_clause_internal(root, clause_expr,
+														   relid, attnums, exprs))
+					return false;
+			}
+			else	/* generic expression */
+				*exprs = lappend(*exprs, clause_expr);
+		}
 
-		/* Otherwise we have (Expr op Const) or (Const op Expr). */
-		*exprs = lappend(*exprs, clause_expr);
 		return true;
 	}
 
@@ -1420,15 +1446,21 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		RangeTblEntry *rte = root->simple_rte_array[relid];
 		ScalarArrayOpExpr *expr = (ScalarArrayOpExpr *) clause;
 		Node	   *clause_expr;
+		List	   *clause_exprs;
 
 		/* Only expressions with two arguments are considered compatible. */
 		if (list_length(expr->args) != 2)
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_expr, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
 			return false;
 
+		/* There has to be one expression exactly. */
+		Assert(list_length(clause_exprs) == 1);
+
+		clause_expr = (Node *) linitial(clause_exprs);
+
 		/*
 		 * If it's not one of the supported operators ("=", "<", ">", etc.),
 		 * just ignore the clause, as it's not compatible with MCV lists.
@@ -2015,20 +2047,19 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * examine_opclause_args
  *		Split an operator expression's arguments into Expr and Const parts.
  *
- * Attempts to match the arguments to either (Expr op Const) or (Const op
- * Expr), possibly with a RelabelType on top. When the expression matches this
- * form, returns true, otherwise returns false.
+ * Attempts to match the arguments to either (Expr op Const) or (Const op Expr)
+ * or (Expr op Expr), possibly with a RelabelType on top. When the expression
+ * matches this form, returns true, otherwise returns false.
  *
  * Optionally returns pointers to the extracted Expr/Const nodes, when passed
- * non-null pointers (exprp, cstp and expronleftp). The expronleftp flag
+ * non-null pointers (exprsp, cstp and expronleftp). The expronleftp flag
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, Node **exprp, Const **cstp,
-					  bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
 {
-	Node	   *expr;
-	Const	   *cst;
+	List	   *exprs = NIL;
+	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
@@ -2048,22 +2079,26 @@ examine_opclause_args(List *args, Node **exprp, Const **cstp,
 
 	if (IsA(rightop, Const))
 	{
-		expr = (Node *) leftop;
+		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
 	}
 	else if (IsA(leftop, Const))
 	{
-		expr = (Node *) rightop;
+		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
 	}
 	else
-		return false;
+	{
+		exprs = lappend(exprs, leftop);
+		exprs = lappend(exprs, rightop);
+		expronleft = false;
+	}
 
 	/* return pointers to the extracted parts if requested */
-	if (exprp)
-		*exprp = expr;
+	if (exprsp)
+		*exprsp = exprs;
 
 	if (cstp)
 		*cstp = cst;
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index f10642df4f7..5806cfbb564 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1645,78 +1645,154 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			Node	   *clause_expr2;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
-			int			idx;
 			Oid			collid;
 
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
-			/* match the attribute/expression to a dimension of the statistic */
-			idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+			if (cst)	/* Expr op Const */
+			{
+				int idx;
 
-			Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert(list_length(clause_exprs) == 1);
+				clause_expr = (Node *) linitial(clause_exprs);
 
-			/*
-			 * Walk through the MCV items and evaluate the current clause. We
-			 * can skip items that were already ruled out, and terminate if
-			 * there are no remaining MCV items that might possibly match.
-			 */
-			for (i = 0; i < mcvlist->nitems; i++)
-			{
-				bool		match = true;
-				MCVItem    *item = &mcvlist->items[i];
+				/* match the attribute/expression to a dimension of the statistic */
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
 
-				Assert(idx >= 0);
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * When the MCV item or the Const value is NULL we can treat
-				 * this as a mismatch. We must not call the operator because
-				 * of strictness.
+				 * Walk through the MCV items and evaluate the current clause. We
+				 * can skip items that were already ruled out, and terminate if
+				 * there are no remaining MCV items that might possibly match.
 				 */
-				if (item->isnull[idx] || cst->constisnull)
+				for (i = 0; i < mcvlist->nitems; i++)
 				{
-					matches[i] = RESULT_MERGE(matches[i], is_or, false);
-					continue;
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					Assert(idx >= 0);
+
+					/*
+					 * When the MCV item or the Const value is NULL we can treat
+					 * this as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || cst->constisnull)
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap. Once
+					 * the value gets false for AND-lists, or true for OR-lists,
+					 * we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics, but
+					 * we can use the collation for the attribute itself, as
+					 * stored in varcollid. We do reset the statistics after a
+					 * type change (including collation change), so this is OK.
+					 * For expressions, we use the collation extracted from the
+					 * expression itself.
+					 */
+					if (expronleft)
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   item->values[idx],
+															   cst->constvalue));
+					else
+						match = DatumGetBool(FunctionCall2Coll(&opproc,
+															   collid,
+															   cst->constvalue,
+															   item->values[idx]));
+
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
 				}
+			}
+			else	/* Expr op Expr */
+			{
+				int			idx;
+				int			idx2;
+
+				Assert(list_length(clause_exprs) == 2);
+
+				clause_expr = (Node *) linitial(clause_exprs);
+				clause_expr2 = (Node *) lsecond(clause_exprs);
+
+				Assert(clause_expr2);
+				Assert(!expronleft);
 
 				/*
-				 * Skip MCV items that can't change result in the bitmap. Once
-				 * the value gets false for AND-lists, or true for OR-lists,
-				 * we don't need to look at more clauses.
+				 * Match the expressions to a dimension of the statistic.
+				 *
+				 * XXX Can the collations differ?
 				 */
-				if (RESULT_IS_FINAL(matches[i], is_or))
-					continue;
+				idx = mcv_match_expression(clause_expr, keys, exprs, &collid);
+				idx2 = mcv_match_expression(clause_expr2, keys, exprs, &collid);
+
+				Assert((idx >= 0) && (idx < bms_num_members(keys) + list_length(exprs)));
+				Assert((idx2 >= 0) && (idx2 < bms_num_members(keys) + list_length(exprs)));
 
 				/*
-				 * First check whether the constant is below the lower
-				 * boundary (in that case we can skip the bucket, because
-				 * there's no overlap).
-				 *
-				 * We don't store collations used to build the statistics, but
-				 * we can use the collation for the attribute itself, as
-				 * stored in varcollid. We do reset the statistics after a
-				 * type change (including collation change), so this is OK.
-				 * For expressions, we use the collation extracted from the
-				 * expression itself.
+				 * Walk through the MCV items and evaluate the current clause.
+				 * We can skip items that were already ruled out, and
+				 * terminate if there are no remaining MCV items that might
+				 * possibly match.
 				 */
-				if (expronleft)
+				for (i = 0; i < mcvlist->nitems; i++)
+				{
+					bool		match = true;
+					MCVItem    *item = &mcvlist->items[i];
+
+					/*
+					 * When either of the MCV items is NULL we can treat this
+					 * as a mismatch. We must not call the operator because
+					 * of strictness.
+					 */
+					if (item->isnull[idx] || item->isnull[idx2])
+					{
+						matches[i] = RESULT_MERGE(matches[i], is_or, false);
+						continue;
+					}
+
+					/*
+					 * Skip MCV items that can't change result in the bitmap.
+					 * Once the value gets false for AND-lists, or true for
+					 * OR-lists, we don't need to look at more clauses.
+					 */
+					if (RESULT_IS_FINAL(matches[i], is_or))
+						continue;
+
+					/*
+					 * We don't store collations used to build the statistics,
+					 * but we can use the collation for the attribute itself,
+					 * as stored in varcollid. We do reset the statistics after
+					 * a type change (including collation change), so this is
+					 * OK. We may need to relax this after allowing extended
+					 * statistics on expressions.
+					 */
 					match = DatumGetBool(FunctionCall2Coll(&opproc,
 														   collid,
 														   item->values[idx],
-														   cst->constvalue));
-				else
-					match = DatumGetBool(FunctionCall2Coll(&opproc,
-														   collid,
-														   cst->constvalue,
-														   item->values[idx]));
+														   item->values[idx2]));
 
-				/* update the match bitmap with the result */
-				matches[i] = RESULT_MERGE(matches[i], is_or, match);
+					/* update the match bitmap with the result */
+					matches[i] = RESULT_MERGE(matches[i], is_or, match);
+				}
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -1726,6 +1802,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 
 			/* valid only after examine_opclause_args returns true */
 			Node	   *clause_expr;
+			List	   *clause_exprs;
 			Const	   *cst;
 			bool		expronleft;
 			Oid			collid;
@@ -1743,11 +1820,14 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_expr, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
 			Assert(expronleft);
+			Assert(list_length(clause_exprs) == 1);
+
+			clause_expr = (Node *) linitial(clause_exprs);
 
 			/* XXX what if (cst->constisnull == NULL)? */
 			if (!cst->constisnull)
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index 71f852c157b..bd4bf7fdf14 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -99,8 +99,8 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									MultiSortSupport mss,
 									int numattrs, AttrNumber *attnums);
 
-extern bool examine_opclause_args(List *args, Node **exprp,
-								  Const **cstp, bool *expronleftp);
+extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
+								  bool *expronleftp);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out
index fea3d18442c..73ba59d75f1 100644
--- a/src/test/regress/expected/stats_ext.out
+++ b/src/test/regress/expected/stats_ext.out
@@ -2246,37 +2246,37 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= a');
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a = c');
  estimated | actual 
 -----------+--------
-        25 |   1250
+      1250 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a != c');
  estimated | actual 
 -----------+--------
-      4975 |   3750
+      3750 |   3750
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a > c');
  estimated | actual 
 -----------+--------
-      1667 |   3750
+      3750 |   3750
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a >= c');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a < c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a <= c');
  estimated | actual 
 -----------+--------
-      1667 |   1250
+      1250 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IN (1, 2, 51, 52) AND b IN ( ''1'', ''2'')');
@@ -3366,7 +3366,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE a IS NULL AND
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND a > a');
@@ -3396,31 +3396,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a > a');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a < a');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR a = a');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT a = a');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND a > a');
@@ -3450,31 +3450,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a > a');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a < a');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR a = a');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND a > a');
@@ -3504,31 +3504,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a > a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a < a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR a = a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND a > a');
@@ -3558,31 +3558,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a > a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a < a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR a = a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND a > a');
@@ -3612,31 +3612,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a > a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a < a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR a = a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND a > a');
@@ -3666,25 +3666,25 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a > a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a < a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR a = a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT a = a');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c > c');
@@ -3702,7 +3702,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c <
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND c = c');
  estimated | actual 
 -----------+--------
-        12 |   2500
+      1250 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT c = c');
@@ -3714,31 +3714,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c > c');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c < c');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR c = c');
  estimated | actual 
 -----------+--------
-      2513 |   2500
+      3750 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b = d OR NOT c = c');
  estimated | actual 
 -----------+--------
-        25 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c > c');
@@ -3756,7 +3756,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c <
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND c = c');
  estimated | actual 
 -----------+--------
-      2488 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT c = c');
@@ -3768,31 +3768,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c > c');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c < c');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR c = c');
  estimated | actual 
 -----------+--------
-      4988 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b != d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c > c');
@@ -3810,7 +3810,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c <
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND c = c');
  estimated | actual 
 -----------+--------
-       833 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT c = c');
@@ -3822,31 +3822,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c > c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c < c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR c = c');
  estimated | actual 
 -----------+--------
-      3333 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b < d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c > c');
@@ -3864,7 +3864,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c <
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND c = c');
  estimated | actual 
 -----------+--------
-       833 |   2500
+      1250 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT c = c');
@@ -3876,31 +3876,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c > c');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c < c');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR c = c');
  estimated | actual 
 -----------+--------
-      3333 |   2500
+      3750 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b <= d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c > c');
@@ -3918,7 +3918,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c <
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND c = c');
  estimated | actual 
 -----------+--------
-       833 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT c = c');
@@ -3930,31 +3930,31 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c > c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c < c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR c = c');
  estimated | actual 
 -----------+--------
-      3333 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b > d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c > c');
@@ -3972,7 +3972,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c <
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND c = c');
  estimated | actual 
 -----------+--------
-       833 |   2500
+      1250 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT c = c');
@@ -3984,25 +3984,25 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d AND NOT
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c > c');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c < c');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR c = c');
  estimated | actual 
 -----------+--------
-      3333 |   2500
+      3750 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists WHERE b >= d OR NOT c = c');
  estimated | actual 
 -----------+--------
-      1667 |   2500
+      2500 |   2500
 (1 row)
 
 -- mcv with pass-by-ref fixlen types, e.g. uuid
@@ -4345,37 +4345,37 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
  estimated | actual 
 -----------+--------
-      9950 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b');
  estimated | actual 
 -----------+--------
-        50 |   7500
+      7500 |   7500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b');
  estimated | actual 
 -----------+--------
-      3333 |  10000
+     10000 |  10000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b');
  estimated | actual 
 -----------+--------
-      3333 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b');
  estimated | actual 
 -----------+--------
-      3333 |   7500
+      7500 |   7500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b');
  estimated | actual 
 -----------+--------
-      3333 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = a AND b = b');
@@ -4417,43 +4417,43 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
  estimated | actual 
 -----------+--------
-        50 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != b AND b != c');
  estimated | actual 
 -----------+--------
-        50 |   1250
+      1250 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a < b AND b < c');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <= b AND b <= c');
  estimated | actual 
 -----------+--------
-      1111 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > b AND b > c');
  estimated | actual 
 -----------+--------
-      1111 |   1250
+      1250 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= b AND b >= c');
  estimated | actual 
 -----------+--------
-      1111 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = c AND b = b');
  estimated | actual 
 -----------+--------
-      9950 |   3750
+      3750 |   3750
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a != c AND b != b');
@@ -4477,7 +4477,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a <=
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a > a AND b > c');
  estimated | actual 
 -----------+--------
-      3333 |   1250
+      1250 |   1250
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a >= a AND b >= c');
@@ -4617,13 +4617,13 @@ DROP TABLE mcv_lists_partial;
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b');
  estimated | actual 
 -----------+--------
-      9950 |   2500
+      2500 |   2500
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_bool WHERE NOT a = b AND b = c');
  estimated | actual 
 -----------+--------
-        50 |   2500
+      2500 |   2500
 (1 row)
 
 -- check the ability to use multiple MCV lists
@@ -4854,55 +4854,55 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = 0 OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b AND c = d');
  estimated | actual 
 -----------+--------
-         1 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != d');
  estimated | actual 
 -----------+--------
-      4950 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b AND c > d');
  estimated | actual 
 -----------+--------
-       556 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND c >= d');
  estimated | actual 
 -----------+--------
-       556 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = b OR c = d');
  estimated | actual 
 -----------+--------
-        50 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != d');
  estimated | actual 
 -----------+--------
-      5000 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b OR c > d');
  estimated | actual 
 -----------+--------
-      2778 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR c >= d');
  estimated | actual 
 -----------+--------
-      2778 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a AND c = d');
  estimated | actual 
 -----------+--------
-        25 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b AND c != c');
@@ -4920,7 +4920,7 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a AN
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b AND d >= c');
  estimated | actual 
 -----------+--------
-       556 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR c = d');
@@ -4932,43 +4932,43 @@ SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a = a OR
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a != b OR c != c');
  estimated | actual 
 -----------+--------
-      4975 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < a OR c > d');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b OR d >= c');
  estimated | actual 
 -----------+--------
-      2778 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a < b');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a > b');
  estimated | actual 
 -----------+--------
-      1667 |      0
+         1 |      0
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a <= b');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 SELECT * FROM check_estimated_rows('SELECT * FROM mcv_lists_multi WHERE a >= b');
  estimated | actual 
 -----------+--------
-      1667 |   5000
+      5000 |   5000
 (1 row)
 
 DROP TABLE mcv_lists_multi;
-- 
2.34.3

From 8c497b61344c499c69791571b2c8b2589c4facde Mon Sep 17 00:00:00 2001
From: Tomas Vondra <tomas.vondra@postgresql.org>
Date: Tue, 19 Jul 2022 22:05:36 +0200
Subject: [PATCH 4/4] Don't treat Var op Var as simple clauses in MCV

Until now, statext_mcv_clauselist_selectivity defined simple clauses as
those with a single attnum. This worked fine for (Var op Const) clauses,
but for (Var op Var) that's no longer correct. We need to consider them
complex even if both sides reference the same attribute.

Note: Perhaps we should just treat (Var op Var) clauses referencing the
same attribute as incompatible with extended statistics. After all it's
not providing any information about other attributes. Although, it does
restrict the relevant part of MCV, because (Var = Var) implies non-NULL.
---
 src/backend/statistics/extended_stats.c       | 95 ++++++++++++++++---
 src/backend/statistics/mcv.c                  |  4 +-
 .../statistics/extended_stats_internal.h      |  2 +-
 3 files changed, 83 insertions(+), 18 deletions(-)

diff --git a/src/backend/statistics/extended_stats.c b/src/backend/statistics/extended_stats.c
index 7bfde068aff..bbf564075c2 100644
--- a/src/backend/statistics/extended_stats.c
+++ b/src/backend/statistics/extended_stats.c
@@ -1320,11 +1320,14 @@ choose_best_statistics(List *stats, char requiredkind, bool inh,
  * statext_is_compatible_clause. It needs to be split like this because
  * of recursion.  The attnums bitmap is an input/output parameter collecting
  * attribute numbers from all compatible clauses (recursively).
+ *
+ * XXX The issimple variable is expected to be initialized by the caller, we
+ * just update it while recursively analyzing the current clause.
  */
 static bool
 statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 									  Index relid, Bitmapset **attnums,
-									  List **exprs)
+									  List **exprs, bool *issimple)
 {
 	/* Look inside any binary-compatible relabeling (as in examine_variable) */
 	if (IsA(clause, RelabelType))
@@ -1372,7 +1375,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		 * Check if the expression has the right shape. This returns either one
 		 * or two expressions, depending on whether there is a Const.
 		 */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/*
@@ -1421,20 +1424,37 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/*
 		 * Check all expressions by recursing. Var expressions are handled as
 		 * a special case (to match it to attnums etc.)
+		 *
+		 * An opclause is simple if it's (Expr op Const) or (Const op Expr). We
+		 * have already checked the overall shape in examine_opclause_args, but
+		 * we haven't checked the expressions are simple (i.e. pretty much Var),
+		 * so we need to check that now. If we discover a complex expression, we
+		 * consider the whole clause complex.
 		 */
 		foreach (lc, clause_exprs)
 		{
 			Node *clause_expr = (Node *) lfirst(lc);
 
+			/*
+			 * XXX Shouldn't this try removing RelabelType, just like we do in
+			 * statext_is_compatible_clause_internal? Otherwise we might treat
+			 * Var (with a RelabelType on top) as complex expression.
+			 */
 			if (IsA(clause_expr, Var))
 			{
 				/* if the Var is incompatible, the whole clause is incompatible */
 				if (!statext_is_compatible_clause_internal(root, clause_expr,
-														   relid, attnums, exprs))
+														   relid, attnums, exprs,
+														   issimple))
 					return false;
 			}
 			else	/* generic expression */
+			{
 				*exprs = lappend(*exprs, clause_expr);
+
+				/* switch to false if there are any complex clauses */
+				*issimple = false;
+			}
 		}
 
 		return true;
@@ -1453,7 +1473,7 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			return false;
 
 		/* Check if the expression has the right shape (one Var, one Const) */
-		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL))
+		if (!examine_opclause_args(expr->args, &clause_exprs, NULL, NULL, issimple))
 			return false;
 
 		/* There has to be one expression exactly. */
@@ -1501,7 +1521,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IN Array clauses by recursing. */
 		if (IsA(clause_expr, Var))
 			return statext_is_compatible_clause_internal(root, clause_expr,
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IN Array. */
 		*exprs = lappend(*exprs, clause_expr);
@@ -1530,6 +1551,18 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		BoolExpr   *expr = (BoolExpr *) clause;
 		ListCell   *lc;
 
+		/*
+		 * All AND/OR clauses are considered complex, even if all arguments are
+		 * simple clauses. For NOT clauses we need to check the argument and then
+		 * we can update the flag.
+		 *
+		 * XXX Maybe for AND/OR we should check if all arguments reference the
+		 * same attnum, and consider them complex only when there are multiple
+		 * attnum values (i.e. different Vars)?
+		 */
+		if (!is_notclause(clause))
+			*issimple = false;
+
 		foreach(lc, expr->args)
 		{
 			/*
@@ -1538,7 +1571,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 			 */
 			if (!statext_is_compatible_clause_internal(root,
 													   (Node *) lfirst(lc),
-													   relid, attnums, exprs))
+													   relid, attnums, exprs,
+													   issimple))
 				return false;
 		}
 
@@ -1553,7 +1587,8 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
 		/* Check Var IS NULL clauses by recursing. */
 		if (IsA(nt->arg, Var))
 			return statext_is_compatible_clause_internal(root, (Node *) (nt->arg),
-														 relid, attnums, exprs);
+														 relid, attnums, exprs,
+														 issimple);
 
 		/* Otherwise we have Expr IS NULL. */
 		*exprs = lappend(*exprs, nt->arg);
@@ -1589,22 +1624,35 @@ statext_is_compatible_clause_internal(PlannerInfo *root, Node *clause,
  */
 static bool
 statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
-							 Bitmapset **attnums, List **exprs)
+							 Bitmapset **attnums, List **exprs, bool *issimple)
 {
 	RangeTblEntry *rte = root->simple_rte_array[relid];
 	RestrictInfo *rinfo = (RestrictInfo *) clause;
 	int			clause_relid;
 	Oid			userid;
 
+	/*
+	 * Clauses are considered simple by default, and we mark them as complex
+	 * when we discover a complex part.
+	 */
+	*issimple = true;
+
 	/*
 	 * Special-case handling for bare BoolExpr AND clauses, because the
 	 * restrictinfo machinery doesn't build RestrictInfos on top of AND
 	 * clauses.
+	 *
+	 * AND clauses are considered complex, even if all arguments are
+	 * simple clauses.
 	 */
 	if (is_andclause(clause))
 	{
 		BoolExpr   *expr = (BoolExpr *) clause;
 		ListCell   *lc;
+		bool		tmp = false;	/* ignored result */
+
+		/* AND clauses are complex, even if the arguments are simple. */
+		*issimple = false;
 
 		/*
 		 * Check that each sub-clause is compatible.  We expect these to be
@@ -1613,7 +1661,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 		foreach(lc, expr->args)
 		{
 			if (!statext_is_compatible_clause(root, (Node *) lfirst(lc),
-											  relid, attnums, exprs))
+											  relid, attnums, exprs, &tmp))
 				return false;
 		}
 
@@ -1635,7 +1683,7 @@ statext_is_compatible_clause(PlannerInfo *root, Node *clause, Index relid,
 
 	/* Check the clause and determine what attributes it references. */
 	if (!statext_is_compatible_clause_internal(root, (Node *) rinfo->clause,
-											   relid, attnums, exprs))
+											   relid, attnums, exprs, issimple))
 		return false;
 
 	/*
@@ -1725,6 +1773,7 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	ListCell   *l;
 	Bitmapset **list_attnums;	/* attnums extracted from the clause */
 	List	  **list_exprs;		/* expressions matched to any statistic */
+	bool	  *list_simple;		/* marks simple expressions */
 	int			listidx;
 	Selectivity sel = (is_or) ? 0.0 : 1.0;
 	RangeTblEntry *rte = planner_rt_fetch(rel->relid, root);
@@ -1739,6 +1788,9 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 	/* expressions extracted from complex expressions */
 	list_exprs = (List **) palloc(sizeof(Node *) * list_length(clauses));
 
+	/* expressions determined to be simple (single expression) */
+	list_simple = (bool *) palloc(sizeof(bool) * list_length(clauses));
+
 	/*
 	 * Pre-process the clauses list to extract the attnums and expressions
 	 * seen in each item.  We need to determine if there are any clauses which
@@ -1756,17 +1808,21 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		Node	   *clause = (Node *) lfirst(l);
 		Bitmapset  *attnums = NULL;
 		List	   *exprs = NIL;
+		bool		issimple = false;
 
 		if (!bms_is_member(listidx, *estimatedclauses) &&
-			statext_is_compatible_clause(root, clause, rel->relid, &attnums, &exprs))
+			statext_is_compatible_clause(root, clause, rel->relid,
+										 &attnums, &exprs, &issimple))
 		{
 			list_attnums[listidx] = attnums;
 			list_exprs[listidx] = exprs;
+			list_simple[listidx] = issimple;
 		}
 		else
 		{
 			list_attnums[listidx] = NULL;
 			list_exprs[listidx] = NIL;
+			list_simple[listidx] = false;
 		}
 
 		listidx++;
@@ -1803,6 +1859,8 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 		listidx = -1;
 		foreach(l, clauses)
 		{
+			Node *clause = (Node *) lfirst(l);
+
 			/* Increment the index before we decide if to skip the clause. */
 			listidx++;
 
@@ -1841,13 +1899,12 @@ statext_mcv_clauselist_selectivity(PlannerInfo *root, List *clauses, int varReli
 			/* record simple clauses (single column or expression) */
 			if ((list_attnums[listidx] == NULL &&
 				 list_length(list_exprs[listidx]) == 1) ||
-				(list_exprs[listidx] == NIL &&
-				 bms_membership(list_attnums[listidx]) == BMS_SINGLETON))
+				 list_simple[listidx])
 				simple_clauses = bms_add_member(simple_clauses,
 												list_length(stat_clauses));
 
 			/* add clause to list and mark it as estimated */
-			stat_clauses = lappend(stat_clauses, (Node *) lfirst(l));
+			stat_clauses = lappend(stat_clauses, clause);
 			*estimatedclauses = bms_add_member(*estimatedclauses, listidx);
 
 			/*
@@ -2056,13 +2113,15 @@ statext_clauselist_selectivity(PlannerInfo *root, List *clauses, int varRelid,
  * specifies on which side of the operator we found the expression node.
  */
 bool
-examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp)
+examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp,
+					  bool *issimplep)
 {
 	List	   *exprs = NIL;
 	Const	   *cst = NULL;
 	bool		expronleft;
 	Node	   *leftop,
 			   *rightop;
+	bool		issimple;
 
 	/* enforced by statext_is_compatible_clause_internal */
 	Assert(list_length(args) == 2);
@@ -2082,18 +2141,21 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 		exprs = lappend(exprs, leftop);
 		cst = (Const *) rightop;
 		expronleft = true;
+		issimple = true;
 	}
 	else if (IsA(leftop, Const))
 	{
 		exprs = lappend(exprs, rightop);
 		cst = (Const *) leftop;
 		expronleft = false;
+		issimple = true;
 	}
 	else
 	{
 		exprs = lappend(exprs, leftop);
 		exprs = lappend(exprs, rightop);
 		expronleft = false;
+		issimple = false;
 	}
 
 	/* return pointers to the extracted parts if requested */
@@ -2106,6 +2168,9 @@ examine_opclause_args(List *args, List **exprsp, Const **cstp, bool *expronleftp
 	if (expronleftp)
 		*expronleftp = expronleft;
 
+	if (issimplep)
+		*issimplep = (*issimplep && issimple);
+
 	return true;
 }
 
diff --git a/src/backend/statistics/mcv.c b/src/backend/statistics/mcv.c
index 5806cfbb564..09a30dc1122 100644
--- a/src/backend/statistics/mcv.c
+++ b/src/backend/statistics/mcv.c
@@ -1654,7 +1654,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			if (cst)	/* Expr op Const */
@@ -1820,7 +1820,7 @@ mcv_get_match_bitmap(PlannerInfo *root, List *clauses,
 			fmgr_info(get_opcode(expr->opno), &opproc);
 
 			/* extract the var/expr and const from the expression */
-			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft))
+			if (!examine_opclause_args(expr->args, &clause_exprs, &cst, &expronleft, NULL))
 				elog(ERROR, "incompatible clause");
 
 			/* ScalarArrayOpExpr has the Var always on the left */
diff --git a/src/include/statistics/extended_stats_internal.h b/src/include/statistics/extended_stats_internal.h
index bd4bf7fdf14..bfe8fdf264a 100644
--- a/src/include/statistics/extended_stats_internal.h
+++ b/src/include/statistics/extended_stats_internal.h
@@ -100,7 +100,7 @@ extern SortItem *build_sorted_items(StatsBuildData *data, int *nitems,
 									int numattrs, AttrNumber *attnums);
 
 extern bool examine_opclause_args(List *args, List **exprs, Const **cstp,
-								  bool *expronleftp);
+								  bool *expronleftp, bool *issimplep);
 
 extern Selectivity mcv_combine_selectivities(Selectivity simple_sel,
 											 Selectivity mcv_sel,
-- 
2.34.3