Draft

Author: danolivo

src/backend/utils/adt/selfuncs.c

@@ -3965,6 +3965,7 @@ estimate_multivariate_ndistinct(PlannerInfo *root, RelOptInfo *rel,
 	MVNDistinct *stats;
 	StatisticExtInfo *matched_info = NULL;
 	RangeTblEntry *rte = planner_rt_fetch(rel->relid, root);
+	List	   *statlist;
 
 	/* bail out immediately if the table has no extended statistics */
 	if (!rel->statlist)
@@ -3973,7 +3974,10 @@ estimate_multivariate_ndistinct(PlannerInfo *root, RelOptInfo *rel,
 	/* look for the ndistinct statistics object matching the most vars */
 	nmatches_vars = 0;			/* we require at least two matches */
 	nmatches_exprs = 0;
-	foreach(lc, rel->statlist)
+
+attempt:
+	statlist = list_copy(rel->statlist);
+	foreach(lc, statlist)
 	{
 		ListCell   *lc2;
 		StatisticExtInfo *info = (StatisticExtInfo *) lfirst(lc);
@@ -4200,7 +4204,11 @@ estimate_multivariate_ndistinct(PlannerInfo *root, RelOptInfo *rel,
 		 * statistics includes all combinations of attributes.
 		 */
 		if (!item)
-			elog(ERROR, "corrupt MVNDistinct entry");
+		{
+			statOid = InvalidOid;
+			statlist = list_delete_ptr(statlist, matched_info);
+			goto attempt;
+		}
 
 		/* Form the output varinfo list, keeping only unmatched ones */
 		foreach(lc, *varinfos)

src/test/regress/sql/stats_ext.sql

@@ -1683,6 +1683,26 @@ SELECT statistics_name, most_common_vals FROM pg_stats_ext x
 SELECT statistics_name, most_common_vals FROM pg_stats_ext_exprs x
     WHERE tablename = 'stats_ext_tbl' ORDER BY ROW(x.*);
 
+CREATE TABLE ext_linear(x int, y int, z int);
+INSERT INTO ext_linear (x,y,z) SELECT gs%47,gs%71,gs%53
+  FROM generate_series(1,100) AS gs;
+CREATE STATISTICS x_y_linear ON x,y,z FROM ext_linear
+  WITH (method = 'linear');
+VACUUM ANALYZE ext_linear;
+
+-- Statistics built with the 'linear' method don't have x,z combination:
+-- don't complain, but not use this statistic slot.
+SELECT check_estimated_rows('SELECT count(*) FROM ext_linear GROUP BY x,z;');
+-- have statistic combination on (x,y) - get correct estimation.
+SELECT check_estimated_rows('SELECT count(*) FROM ext_linear GROUP BY x,y;');
+CREATE STATISTICS x_y_linear_1 ON x,z,y FROM ext_linear WITH (method = 'linear');
+DROP STATISTICS x_y_linear;
+VACUUM ANALYZE ext_linear;
+-- TODO: Now we have right combination in the second statistic slot and use it
+-- getting correct groups number estimation
+SELECT check_estimated_rows('SELECT count(*) FROM ext_linear GROUP BY x,z;');
+DROP TABLE ext_linear;
+
 -- Tidy up
 DROP OPERATOR <<< (int, int);
 DROP FUNCTION op_leak(int, int);