Rethink representation of index clauses' mapping to index columns.

In commit e2c2c2e8 I made use of nested
list structures to show which clauses went with which index columns, but
on reflection that's a data structure that only an old-line Lisp hacker
could love.  Worse, it adds unnecessary complication to the many places
that don't much care which clauses go with which index columns.  Revert
to the previous arrangement of flat lists of clauses, and instead add a
parallel integer list of column numbers.  The places that care about the
pairing can chase both lists with forboth(), while the places that don't
care just examine one list the same as before.

The only real downside to this is that there are now two more lists that
need to be passed to amcostestimate functions in case they care about
column matching (which btcostestimate does, so not passing the info is not
an option).  Rather than deal with 11-argument amcostestimate functions,
pass just the IndexPath and expect the functions to extract fields from it.
That gets us down to 7 arguments which is better than 11, and it seems
more future-proof against likely additions to the information we keep
about an index path.

doc/src/sgml/indexam.sgml

@@ -288,9 +288,7 @@ amcanreturn (Relation indexRelation);
 <programlisting>
 void
 amcostestimate (PlannerInfo *root,
-                IndexOptInfo *index,
-                List *indexQuals,
-                List *indexOrderBys,
+                IndexPath *path,
                 RelOptInfo *outer_rel,
                 Cost *indexStartupCost,
                 Cost *indexTotalCost,
@@ -929,9 +927,7 @@ amrestrpos (IndexScanDesc scan);
 <programlisting>
 void
 amcostestimate (PlannerInfo *root,
-                IndexOptInfo *index,
-                List *indexQuals,
-                List *indexOrderBys,
+                IndexPath *path,
                 RelOptInfo *outer_rel,
                 Cost *indexStartupCost,
                 Cost *indexTotalCost,
@@ -939,7 +935,7 @@ amcostestimate (PlannerInfo *root,
                 double *indexCorrelation);
 </programlisting>
 
-   The first five parameters are inputs:
+   The first three parameters are inputs:
 
    <variablelist>
     <varlistentry>
@@ -952,32 +948,11 @@ amcostestimate (PlannerInfo *root,
     </varlistentry>
 
     <varlistentry>
-     <term><parameter>index</></term>
+     <term><parameter>path</></term>
      <listitem>
       <para>
-       The index being considered.
-      </para>
-     </listitem>
-    </varlistentry>
-
-    <varlistentry>
-     <term><parameter>indexQuals</></term>
-     <listitem>
-      <para>
-       List of index qual clauses (implicitly ANDed);
-       a <symbol>NIL</> list indicates no qualifiers are available.
-       Note that the list contains expression trees with RestrictInfo nodes
-       at the top, not ScanKeys.
-      </para>
-     </listitem>
-    </varlistentry>
-
-    <varlistentry>
-     <term><parameter>indexOrderBys</></term>
-     <listitem>
-      <para>
-       List of indexable ORDER BY operators, or <symbol>NIL</> if none.
-       Note that the list contains expression trees, not ScanKeys.
+       The index access path being considered.  All fields except cost and
+       selectivity values are valid.
       </para>
      </listitem>
     </varlistentry>
@@ -988,7 +963,8 @@ amcostestimate (PlannerInfo *root,
       <para>
        If the index is being considered for use in a join inner indexscan,
        the planner's information about the outer side of the join.  Otherwise
-       <symbol>NULL</>.  When non-<symbol>NULL</>, some of the qual clauses will be join clauses
+       <symbol>NULL</>.  When non-<symbol>NULL</>, some of the qual clauses
+       will be join clauses for joins
        with this rel rather than being simple restriction clauses.  Also,
        the cost estimator should expect that the index scan will be repeated
        for each row of the outer rel.
@@ -1055,7 +1031,7 @@ amcostestimate (PlannerInfo *root,
    row should usually be taken as <varname>cpu_index_tuple_cost</>.  In
    addition, an appropriate multiple of <varname>cpu_operator_cost</> should
    be charged for any comparison operators invoked during index processing
-   (especially evaluation of the <literal>indexQuals</> themselves).
+   (especially evaluation of the indexquals themselves).
   </para>
 
   <para>
@@ -1103,8 +1079,8 @@ amcostestimate (PlannerInfo *root,
      knowledge, use the standard optimizer function <function>clauselist_selectivity()</function>:
 
 <programlisting>
-*indexSelectivity = clauselist_selectivity(root, indexQuals,
-                                           index-&gt;rel-&gt;relid,
+*indexSelectivity = clauselist_selectivity(root, path-&gt;indexquals,
+                                           path-&gt;indexinfo-&gt;rel-&gt;relid,
                                            JOIN_INNER, NULL);
 </programlisting>
     </para>
@@ -1115,7 +1091,8 @@ amcostestimate (PlannerInfo *root,
      Estimate the number of index rows that will be visited during the
      scan.  For many index types this is the same as <parameter>indexSelectivity</> times
      the number of rows in the index, but it might be more.  (Note that the
-     index's size in pages and rows is available from the <structname>IndexOptInfo</> struct.)
+     index's size in pages and rows is available from the
+     <literal>path-&gt;indexinfo</> struct.)
     </para>
    </step>
 
@@ -1137,7 +1114,7 @@ amcostestimate (PlannerInfo *root,
  * Also, we charge for evaluation of the indexquals at each index row.
  * All the costs are assumed to be paid incrementally during the scan.
  */
-cost_qual_eval(&amp;index_qual_cost, indexQuals, root);
+cost_qual_eval(&amp;index_qual_cost, path-&gt;indexquals, root);
 *indexStartupCost = index_qual_cost.startup;
 *indexTotalCost = seq_page_cost * numIndexPages +
     (cpu_index_tuple_cost + index_qual_cost.per_tuple) * numIndexTuples;

src/backend/access/nbtree/nbtutils.c

@@ -616,7 +616,7 @@ _bt_advance_array_keys(IndexScanDesc scan, ScanDirection dir)
  *
  * The output keys must be sorted by index attribute.  Presently we expect
  * (but verify) that the input keys are already so sorted --- this is done
- * by group_clauses_by_indexkey() in indxpath.c.  Some reordering of the keys
+ * by match_clauses_to_index() in indxpath.c.  Some reordering of the keys
  * within each attribute may be done as a byproduct of the processing here,
  * but no other code depends on that.
  *

src/backend/nodes/outfuncs.c

@@ -1512,7 +1512,9 @@ _outIndexPath(StringInfo str, const IndexPath *node)
 	WRITE_NODE_FIELD(indexinfo);
 	WRITE_NODE_FIELD(indexclauses);
 	WRITE_NODE_FIELD(indexquals);
+	WRITE_NODE_FIELD(indexqualcols);
 	WRITE_NODE_FIELD(indexorderbys);
+	WRITE_NODE_FIELD(indexorderbycols);
 	WRITE_BOOL_FIELD(isjoininner);
 	WRITE_ENUM_FIELD(indexscandir, ScanDirection);
 	WRITE_FLOAT_FIELD(indextotalcost, "%.2f");

src/backend/optimizer/path/costsize.c

@@ -52,7 +52,9 @@
  * results into.  All the input data they need is passed as separate
  * parameters, even though much of it could be extracted from the Path.
  * An exception is made for the cost_XXXjoin() routines, which expect all
- * the non-cost fields of the passed XXXPath to be filled in.
+ * the non-cost fields of the passed XXXPath to be filled in, and similarly
+ * cost_index() assumes the passed IndexPath is valid except for its output
+ * values.
  *
  *
  * Portions Copyright (c) 1996-2011, PostgreSQL Global Development Group
@@ -208,38 +210,28 @@ cost_seqscan(Path *path, PlannerInfo *root,
  * cost_index
  *	  Determines and returns the cost of scanning a relation using an index.
  *
- * 'index' is the index to be used
- * 'indexQuals' is a list of lists of applicable qual clauses (implicit AND
- *		semantics, one sub-list per index column)
- * 'indexOrderBys' is a list of lists of lists of ORDER BY expressions for
- *		amcanorderbyop indexes (lists per pathkey and index column)
- * 'indexonly' is true if it's an index-only scan
+ * 'path' describes the indexscan under consideration, and is complete
+ *		except for the fields to be set by this routine
  * 'outer_rel' is the outer relation when we are considering using the index
- *		scan as the inside of a nestloop join (hence, some of the indexQuals
+ *		scan as the inside of a nestloop join (hence, some of the indexquals
  *		are join clauses, and we should expect repeated scans of the index);
  *		NULL for a plain index scan
  *
- * cost_index() takes an IndexPath not just a Path, because it sets a few
- * additional fields of the IndexPath besides startup_cost and total_cost.
- * These fields are needed if the IndexPath is used in a BitmapIndexScan.
+ * In addition to startup_cost and total_cost, cost_index() sets the path's
+ * indextotalcost and indexselectivity fields.  These values are needed if
+ * the IndexPath is used in a BitmapIndexScan.
  *
- * indexQuals is a list of lists of RestrictInfo nodes, but indexOrderBys
- * is a list of lists of lists of bare expressions.
- *
- * NOTE: 'indexQuals' must contain only clauses usable as index restrictions.
- * Any additional quals evaluated as qpquals may reduce the number of returned
- * tuples, but they won't reduce the number of tuples we have to fetch from
- * the table, so they don't reduce the scan cost.
+ * NOTE: path->indexquals must contain only clauses usable as index
+ * restrictions.  Any additional quals evaluated as qpquals may reduce the
+ * number of returned tuples, but they won't reduce the number of tuples
+ * we have to fetch from the table, so they don't reduce the scan cost.
  */
 void
-cost_index(IndexPath *path, PlannerInfo *root,
-		   IndexOptInfo *index,
-		   List *indexQuals,
-		   List *indexOrderBys,
-		   bool indexonly,
-		   RelOptInfo *outer_rel)
+cost_index(IndexPath *path, PlannerInfo *root, RelOptInfo *outer_rel)
 {
+	IndexOptInfo *index = path->indexinfo;
 	RelOptInfo *baserel = index->rel;
+	bool		indexonly = (path->path.pathtype == T_IndexOnlyScan);
 	Cost		startup_cost = 0;
 	Cost		run_cost = 0;
 	Cost		indexStartupCost;
@@ -271,11 +263,9 @@ cost_index(IndexPath *path, PlannerInfo *root,
 	 * the fraction of main-table tuples we will have to retrieve) and its
 	 * correlation to the main-table tuple order.
 	 */
-	OidFunctionCall9(index->amcostestimate,
+	OidFunctionCall7(index->amcostestimate,
 					 PointerGetDatum(root),
-					 PointerGetDatum(index),
-					 PointerGetDatum(indexQuals),
-					 PointerGetDatum(indexOrderBys),
+					 PointerGetDatum(path),
 					 PointerGetDatum(outer_rel),
 					 PointerGetDatum(&indexStartupCost),
 					 PointerGetDatum(&indexTotalCost),
@@ -431,7 +421,7 @@ cost_index(IndexPath *path, PlannerInfo *root,
 	{
 		QualCost	index_qual_cost;
 
-		cost_qual_eval(&index_qual_cost, indexQuals, root);
+		cost_qual_eval(&index_qual_cost, path->indexquals, root);
 		/* any startup cost still has to be paid ... */
 		cpu_per_tuple -= index_qual_cost.per_tuple;
 	}
@@ -589,7 +579,7 @@ get_indexpath_pages(Path *bitmapqual)
  * 'baserel' is the relation to be scanned
  * 'bitmapqual' is a tree of IndexPaths, BitmapAndPaths, and BitmapOrPaths
  * 'outer_rel' is the outer relation when we are considering using the bitmap
- *		scan as the inside of a nestloop join (hence, some of the indexQuals
+ *		scan as the inside of a nestloop join (hence, some of the indexquals
  *		are join clauses, and we should expect repeated scans of the table);
  *		NULL for a plain bitmap scan
  *

src/backend/optimizer/plan/createplan.c

@@ -83,10 +83,8 @@ static HashJoin *create_hashjoin_plan(PlannerInfo *root, HashPath *best_path,
 					 Plan *outer_plan, Plan *inner_plan);
 static Node *replace_nestloop_params(PlannerInfo *root, Node *expr);
 static Node *replace_nestloop_params_mutator(Node *node, PlannerInfo *root);
-static List *fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
-						 List *indexquals);
-static List *fix_indexorderby_references(PlannerInfo *root, IndexPath *index_path,
-							List *indexorderbys);
+static List *fix_indexqual_references(PlannerInfo *root, IndexPath *index_path);
+static List *fix_indexorderby_references(PlannerInfo *root, IndexPath *index_path);
 static Node *fix_indexqual_operand(Node *node, IndexOptInfo *index, int indexcol);
 static List *get_switched_clauses(List *clauses, Relids outerrelids);
 static List *order_qual_clauses(PlannerInfo *root, List *clauses);
@@ -1082,11 +1080,11 @@ create_indexscan_plan(PlannerInfo *root,
 					  bool indexonly)
 {
 	Scan	   *scan_plan;
+	List	   *indexquals = best_path->indexquals;
 	List	   *indexorderbys = best_path->indexorderbys;
 	Index		baserelid = best_path->path.parent->relid;
 	Oid			indexoid = best_path->indexinfo->indexoid;
 	List	   *qpqual;
-	List	   *indexquals;
 	List	   *stripped_indexquals;
 	List	   *fixed_indexquals;
 	List	   *fixed_indexorderbys;
@@ -1096,13 +1094,6 @@ create_indexscan_plan(PlannerInfo *root,
 	Assert(baserelid > 0);
 	Assert(best_path->path.parent->rtekind == RTE_RELATION);
 
-	/*
-	 * We need to flatten the indexquals list-of-sublists, since most of the
-	 * processing below doesn't care which index column each qual is
-	 * associated with.
-	 */
-	indexquals = flatten_clausegroups_list(best_path->indexquals);
-
 	/*
 	 * Build "stripped" indexquals structure (no RestrictInfos) to pass to
 	 * executor as indexqualorig
@@ -1111,23 +1102,14 @@ create_indexscan_plan(PlannerInfo *root,
 
 	/*
 	 * The executor needs a copy with the indexkey on the left of each clause
-	 * and with index Vars substituted for table ones.  Here we use the
-	 * unflattened list so we can conveniently tell which index column each
-	 * clause is for.
+	 * and with index Vars substituted for table ones.
 	 */
-	fixed_indexquals = fix_indexqual_references(root, best_path,
-												best_path->indexquals);
+	fixed_indexquals = fix_indexqual_references(root, best_path);
 
 	/*
 	 * Likewise fix up index attr references in the ORDER BY expressions.
 	 */
-	fixed_indexorderbys = fix_indexorderby_references(root, best_path,
-													  indexorderbys);
-
-	/*
-	 * Also produce a flat list to become the indexorderbyorig.
-	 */
-	indexorderbys = flatten_indexorderbys_list(indexorderbys);
+	fixed_indexorderbys = fix_indexorderby_references(root, best_path);
 
 	/*
 	 * If this is an innerjoin scan, the indexclauses will contain join
@@ -1506,7 +1488,7 @@ create_bitmap_subplan(PlannerInfo *root, Path *bitmapqual,
 			clamp_row_est(ipath->indexselectivity * ipath->path.parent->tuples);
 		plan->plan_width = 0;	/* meaningless */
 		*qual = get_actual_clauses(ipath->indexclauses);
-		*indexqual = get_actual_clauses(flatten_clausegroups_list(ipath->indexquals));
+		*indexqual = get_actual_clauses(ipath->indexquals);
 		foreach(l, ipath->indexinfo->indpred)
 		{
 			Expr	   *pred = (Expr *) lfirst(l);
@@ -2484,8 +2466,7 @@ replace_nestloop_params_mutator(Node *node, PlannerInfo *root)
  *	  Adjust indexqual clauses to the form the executor's indexqual
  *	  machinery needs.
  *
- * We have five tasks here:
- *	* Flatten the list-of-sublists structure of indexquals into a simple list.
+ * We have four tasks here:
  *	* Remove RestrictInfo nodes from the input clauses.
  *	* Replace any outer-relation Var or PHV nodes with nestloop Params.
  *	  (XXX eventually, that responsibility should go elsewhere?)
@@ -2494,34 +2475,25 @@ replace_nestloop_params_mutator(Node *node, PlannerInfo *root)
  *	* If the index key is on the right, commute the clause to put it on the
  *	  left.
  *
- * The result is a modified copy of the indexquals list --- the
+ * The result is a modified copy of the path's indexquals list --- the
  * original is not changed.  Note also that the copy shares no substructure
  * with the original; this is needed in case there is a subplan in it (we need
  * two separate copies of the subplan tree, or things will go awry).
  */
 static List *
-fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
-						 List *indexquals)
+fix_indexqual_references(PlannerInfo *root, IndexPath *index_path)
 {
 	IndexOptInfo *index = index_path->indexinfo;
 	List	   *fixed_indexquals;
-	ListCell   *lc1;
-	int			indexcol;
+	ListCell   *lcc,
+			   *lci;
 
 	fixed_indexquals = NIL;
 
-	/* clausegroups must correspond to index columns */
-	Assert(list_length(indexquals) <= index->ncolumns);
-
-	indexcol = 0;
-	foreach(lc1, indexquals)
-	{
-		List	   *clausegroup = (List *) lfirst(lc1);
-		ListCell   *lc2;
-
-		foreach(lc2, clausegroup)
+	forboth(lcc, index_path->indexquals, lci, index_path->indexqualcols)
 	{
-			RestrictInfo *rinfo = (RestrictInfo *) lfirst(lc2);
+		RestrictInfo *rinfo = (RestrictInfo *) lfirst(lcc);
+		int			indexcol = lfirst_int(lci);
 		Node	   *clause;
 
 		Assert(IsA(rinfo, RestrictInfo));
@@ -2542,9 +2514,9 @@ fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
 				elog(ERROR, "indexqual clause is not binary opclause");
 
 			/*
-				 * Check to see if the indexkey is on the right; if so,
-				 * commute the clause. The indexkey should be the side that
-				 * refers to (only) the base relation.
+			 * Check to see if the indexkey is on the right; if so, commute
+			 * the clause.  The indexkey should be the side that refers to
+			 * (only) the base relation.
 			 */
 			if (!bms_equal(rinfo->left_relids, index->rel->relids))
 				CommuteOpExpr(op);
@@ -2563,7 +2535,7 @@ fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
 			List	   *indexcolnos;
 			bool		var_on_left;
 			ListCell   *lca,
-						   *lci;
+					   *lcai;
 
 			/*
 			 * Re-discover which index columns are used in the rowcompare.
@@ -2576,15 +2548,15 @@ fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
 
 			/*
 			 * Trouble if adjust_rowcompare_for_index thought the
-				 * RowCompareExpr didn't match the index as-is; the clause
-				 * should have gone through that routine already.
+			 * RowCompareExpr didn't match the index as-is; the clause should
+			 * have gone through that routine already.
 			 */
 			if (newrc != (Expr *) rc)
 				elog(ERROR, "inconsistent results from adjust_rowcompare_for_index");
 
 			/*
-				 * Check to see if the indexkey is on the right; if so,
-				 * commute the clause.
+			 * Check to see if the indexkey is on the right; if so, commute
+			 * the clause.
 			 */
 			if (!var_on_left)
 				CommuteRowCompareExpr(rc);
@@ -2593,11 +2565,11 @@ fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
 			 * Now replace the indexkey expressions with index Vars.
 			 */
 			Assert(list_length(rc->largs) == list_length(indexcolnos));
-				forboth(lca, rc->largs, lci, indexcolnos)
+			forboth(lca, rc->largs, lcai, indexcolnos)
 			{
 				lfirst(lca) = fix_indexqual_operand(lfirst(lca),
 													index,
-														lfirst_int(lci));
+													lfirst_int(lcai));
 			}
 		}
 		else if (IsA(clause, ScalarArrayOpExpr))
@@ -2627,9 +2599,6 @@ fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
 		fixed_indexquals = lappend(fixed_indexquals, clause);
 	}
 
-		indexcol++;
-	}
-
 	return fixed_indexquals;
 }
 
@@ -2645,41 +2614,25 @@ fix_indexqual_references(PlannerInfo *root, IndexPath *index_path,
  * is allowed for ordering operators.
  */
 static List *
-fix_indexorderby_references(PlannerInfo *root, IndexPath *index_path,
-							List *indexorderbys)
+fix_indexorderby_references(PlannerInfo *root, IndexPath *index_path)
 {
 	IndexOptInfo *index = index_path->indexinfo;
 	List	   *fixed_indexorderbys;
-	ListCell   *lc1;
+	ListCell   *lcc,
+			   *lci;
 
 	fixed_indexorderbys = NIL;
 
-	foreach(lc1, indexorderbys)
-	{
-		List	   *percollists = (List *) lfirst(lc1);
-		ListCell   *lc2;
-		int			indexcol;
-
-		/* percollists must correspond to index columns */
-		Assert(list_length(percollists) <= index->ncolumns);
-
-		indexcol = 0;
-		foreach(lc2, percollists)
-		{
-			List	   *percollist = (List *) lfirst(lc2);
-
-			if (percollist != NIL)
+	forboth(lcc, index_path->indexorderbys, lci, index_path->indexorderbycols)
 	{
-				Node	   *clause = (Node *) linitial(percollist);
-
-				/* Should have only one clause per index column */
-				Assert(list_length(percollist) == 1);
+		Node	   *clause = (Node *) lfirst(lcc);
+		int			indexcol = lfirst_int(lci);
 
 		/*
 		 * Replace any outer-relation variables with nestloop params.
 		 *
-				 * This also makes a copy of the clause, so it's safe to
-				 * modify it in-place below.
+		 * This also makes a copy of the clause, so it's safe to modify it
+		 * in-place below.
 		 */
 		clause = replace_nestloop_params(root, clause);
 
@@ -2704,10 +2657,6 @@ fix_indexorderby_references(PlannerInfo *root, IndexPath *index_path,
 		fixed_indexorderbys = lappend(fixed_indexorderbys, clause);
 	}
 
-			indexcol++;
-		}
-	}
-
 	return fixed_indexorderbys;
 }
 

src/backend/optimizer/plan/planner.c

@@ -3296,7 +3296,7 @@ plan_cluster_use_sort(Oid tableOid, Oid indexOid)
 
 	/* Estimate the cost of index scan */
 	indexScanPath = create_index_path(root, indexInfo,
-									  NIL, NIL, NIL,
+									  NIL, NIL, NIL, NIL, NIL,
 									  ForwardScanDirection, false, NULL);
 
 	return (seqScanAndSortPath.total_cost < indexScanPath->path.total_cost);

src/backend/optimizer/util/pathnode.c

@@ -410,10 +410,14 @@ create_seqscan_path(PlannerInfo *root, RelOptInfo *rel)
  *	  Creates a path node for an index scan.
  *
  * 'index' is a usable index.
- * 'clause_groups' is a list of lists of RestrictInfo nodes
+ * 'indexclauses' is a list of RestrictInfo nodes representing clauses
  *			to be used as index qual conditions in the scan.
- * 'indexorderbys' is a list of lists of lists of bare expressions (not
- *			RestrictInfos) to be used as index ordering operators.
+ * 'indexclausecols' is an integer list of index column numbers (zero based)
+ *			the indexclauses can be used with.
+ * 'indexorderbys' is a list of bare expressions (no RestrictInfos)
+ *			to be used as index ordering operators in the scan.
+ * 'indexorderbycols' is an integer list of index column numbers (zero based)
+ *			the ordering operators can be used with.
  * 'pathkeys' describes the ordering of the path.
  * 'indexscandir' is ForwardScanDirection or BackwardScanDirection
  *			for an ordered index, or NoMovementScanDirection for
@@ -427,8 +431,10 @@ create_seqscan_path(PlannerInfo *root, RelOptInfo *rel)
 IndexPath *
 create_index_path(PlannerInfo *root,
 				  IndexOptInfo *index,
-				  List *clause_groups,
+				  List *indexclauses,
+				  List *indexclausecols,
 				  List *indexorderbys,
+				  List *indexorderbycols,
 				  List *pathkeys,
 				  ScanDirection indexscandir,
 				  bool indexonly,
@@ -437,7 +443,7 @@ create_index_path(PlannerInfo *root,
 	IndexPath  *pathnode = makeNode(IndexPath);
 	RelOptInfo *rel = index->rel;
 	List	   *indexquals,
-			   *allclauses;
+			   *indexqualcols;
 
 	/*
 	 * For a join inner scan, there's no point in marking the path with any
@@ -457,16 +463,16 @@ create_index_path(PlannerInfo *root,
 	pathnode->path.pathkeys = pathkeys;
 
 	/* Convert clauses to indexquals the executor can handle */
-	indexquals = expand_indexqual_conditions(index, clause_groups);
-
-	/* Flatten the clause-groups list to produce indexclauses list */
-	allclauses = flatten_clausegroups_list(clause_groups);
+	expand_indexqual_conditions(index, indexclauses, indexclausecols,
+								&indexquals, &indexqualcols);
 
 	/* Fill in the pathnode */
 	pathnode->indexinfo = index;
-	pathnode->indexclauses = allclauses;
+	pathnode->indexclauses = indexclauses;
 	pathnode->indexquals = indexquals;
+	pathnode->indexqualcols = indexqualcols;
 	pathnode->indexorderbys = indexorderbys;
+	pathnode->indexorderbycols = indexorderbycols;
 
 	pathnode->isjoininner = (outer_rel != NULL);
 	pathnode->indexscandir = indexscandir;
@@ -476,7 +482,7 @@ create_index_path(PlannerInfo *root,
 		/*
 		 * We must compute the estimated number of output rows for the
 		 * indexscan.  This is less than rel->rows because of the additional
-		 * selectivity of the join clauses.  Since clause_groups may contain
+		 * selectivity of the join clauses.  Since indexclauses may contain
 		 * both restriction and join clauses, we have to do a set union to get
 		 * the full set of clauses that must be considered to compute the
 		 * correct selectivity.  (Without the union operation, we might have
@@ -489,7 +495,9 @@ create_index_path(PlannerInfo *root,
 		 * Note that we force the clauses to be treated as non-join clauses
 		 * during selectivity estimation.
 		 */
-		allclauses = list_union_ptr(rel->baserestrictinfo, allclauses);
+		List	   *allclauses;
+
+		allclauses = list_union_ptr(rel->baserestrictinfo, indexclauses);
 		pathnode->rows = rel->tuples *
 			clauselist_selectivity(root,
 								   allclauses,
@@ -508,8 +516,7 @@ create_index_path(PlannerInfo *root,
 		pathnode->rows = rel->rows;
 	}
 
-	cost_index(pathnode, root, index, indexquals, indexorderbys,
-			   indexonly, outer_rel);
+	cost_index(pathnode, root, outer_rel);
 
 	return pathnode;
 }

src/include/catalog/catversion.h

@@ -53,6 +53,6 @@
  */
 
 /*							yyyymmddN */
-#define CATALOG_VERSION_NO	201112231
+#define CATALOG_VERSION_NO	201112241
 
 #endif

src/include/catalog/pg_proc.h

@@ -548,7 +548,7 @@ DATA(insert OID = 972 (  btvacuumcleanup   PGNSP PGUID 12 1 0 0 0 f f f t f v 2
 DESCR("btree(internal)");
 DATA(insert OID = 276 (  btcanreturn	   PGNSP PGUID 12 1 0 0 0 f f f t f s 1 0 16 "2281" _null_ _null_ _null_ _null_ btcanreturn _null_ _null_ _null_ ));
 DESCR("btree(internal)");
-DATA(insert OID = 1268 (  btcostestimate   PGNSP PGUID 12 1 0 0 0 f f f t f v 9 0 2278 "2281 2281 2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ btcostestimate _null_ _null_ _null_ ));
+DATA(insert OID = 1268 (  btcostestimate   PGNSP PGUID 12 1 0 0 0 f f f t f v 7 0 2278 "2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ btcostestimate _null_ _null_ _null_ ));
 DESCR("btree(internal)");
 DATA(insert OID = 2785 (  btoptions		   PGNSP PGUID 12 1 0 0 0 f f f t f s 2 0 17 "1009 16" _null_ _null_ _null_ _null_  btoptions _null_ _null_ _null_ ));
 DESCR("btree(internal)");
@@ -662,7 +662,7 @@ DATA(insert OID = 442 (  hashbulkdelete    PGNSP PGUID 12 1 0 0 0 f f f t f v 4
 DESCR("hash(internal)");
 DATA(insert OID = 425 (  hashvacuumcleanup PGNSP PGUID 12 1 0 0 0 f f f t f v 2 0 2281 "2281 2281" _null_ _null_ _null_ _null_ hashvacuumcleanup _null_ _null_ _null_ ));
 DESCR("hash(internal)");
-DATA(insert OID = 438 (  hashcostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 9 0 2278 "2281 2281 2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ hashcostestimate _null_ _null_ _null_ ));
+DATA(insert OID = 438 (  hashcostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 7 0 2278 "2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ hashcostestimate _null_ _null_ _null_ ));
 DESCR("hash(internal)");
 DATA(insert OID = 2786 (  hashoptions	   PGNSP PGUID 12 1 0 0 0 f f f t f s 2 0 17 "1009 16" _null_ _null_ _null_ _null_  hashoptions _null_ _null_ _null_ ));
 DESCR("hash(internal)");
@@ -917,7 +917,7 @@ DATA(insert OID = 776 (  gistbulkdelete    PGNSP PGUID 12 1 0 0 0 f f f t f v 4
 DESCR("gist(internal)");
 DATA(insert OID = 2561 (  gistvacuumcleanup   PGNSP PGUID 12 1 0 0 0 f f f t f v 2 0 2281 "2281 2281" _null_ _null_ _null_ _null_ gistvacuumcleanup _null_ _null_ _null_ ));
 DESCR("gist(internal)");
-DATA(insert OID = 772 (  gistcostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 9 0 2278 "2281 2281 2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ gistcostestimate _null_ _null_ _null_ ));
+DATA(insert OID = 772 (  gistcostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 7 0 2278 "2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ gistcostestimate _null_ _null_ _null_ ));
 DESCR("gist(internal)");
 DATA(insert OID = 2787 (  gistoptions	   PGNSP PGUID 12 1 0 0 0 f f f t f s 2 0 17 "1009 16" _null_ _null_ _null_ _null_  gistoptions _null_ _null_ _null_ ));
 DESCR("gist(internal)");
@@ -3870,7 +3870,7 @@ DATA(insert OID = 2739 (  ginbulkdelete    PGNSP PGUID 12 1 0 0 0 f f f t f v 4
 DESCR("gin(internal)");
 DATA(insert OID = 2740 (  ginvacuumcleanup PGNSP PGUID 12 1 0 0 0 f f f t f v 2 0 2281 "2281 2281" _null_ _null_ _null_ _null_ ginvacuumcleanup _null_ _null_ _null_ ));
 DESCR("gin(internal)");
-DATA(insert OID = 2741 (  gincostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 9 0 2278 "2281 2281 2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ gincostestimate _null_ _null_ _null_ ));
+DATA(insert OID = 2741 (  gincostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 7 0 2278 "2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ gincostestimate _null_ _null_ _null_ ));
 DESCR("gin(internal)");
 DATA(insert OID = 2788 (  ginoptions	   PGNSP PGUID 12 1 0 0 0 f f f t f s 2 0 17 "1009 16" _null_ _null_ _null_ _null_  ginoptions _null_ _null_ _null_ ));
 DESCR("gin(internal)");
@@ -4530,7 +4530,7 @@ DATA(insert OID = 4012 (  spgvacuumcleanup   PGNSP PGUID 12 1 0 0 0 f f f t f v
 DESCR("spgist(internal)");
 DATA(insert OID = 4032 (  spgcanreturn	   PGNSP PGUID 12 1 0 0 0 f f f t f s 1 0 16 "2281" _null_ _null_ _null_ _null_ spgcanreturn _null_ _null_ _null_ ));
 DESCR("spgist(internal)");
-DATA(insert OID = 4013 (  spgcostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 9 0 2278 "2281 2281 2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ spgcostestimate _null_ _null_ _null_ ));
+DATA(insert OID = 4013 (  spgcostestimate  PGNSP PGUID 12 1 0 0 0 f f f t f v 7 0 2278 "2281 2281 2281 2281 2281 2281 2281" _null_ _null_ _null_ _null_ spgcostestimate _null_ _null_ _null_ ));
 DESCR("spgist(internal)");
 DATA(insert OID = 4014 (  spgoptions	   PGNSP PGUID 12 1 0 0 0 f f f t f s 2 0 17 "1009 16" _null_ _null_ _null_ _null_  spgoptions _null_ _null_ _null_ ));
 DESCR("spgist(internal)");

src/include/nodes/relation.h

@@ -657,27 +657,33 @@ typedef struct Path
  *
  * 'indexclauses' is a list of index qualification clauses, with implicit
  * AND semantics across the list.  Each clause is a RestrictInfo node from
- * the query's WHERE or JOIN conditions.
- *
- * 'indexquals' is a list of sub-lists of the actual index qual conditions
- * that can be used with the index.  There is one possibly-empty sub-list
- * for each index column (but empty sub-lists for trailing columns can be
- * omitted).  The qual conditions are RestrictInfos, and in simple cases
- * are the same RestrictInfos that appear in the flat indexclauses list.
- * But when special indexable operators appear in 'indexclauses', they are
- * replaced by their derived indexscannable conditions in 'indexquals'.
- * Note that an entirely empty indexquals list denotes a full-index scan.
- *
- * 'indexorderbys', if not NIL, is a list of lists of lists of ORDER BY
- * expressions that have been found to be usable as ordering operators for an
- * amcanorderbyop index.  These are not RestrictInfos, just bare expressions,
+ * the query's WHERE or JOIN conditions.  An empty list implies a full
+ * index scan.
+ *
+ * 'indexquals' has the same structure as 'indexclauses', but it contains
+ * the actual index qual conditions that can be used with the index.
+ * In simple cases this is identical to 'indexclauses', but when special
+ * indexable operators appear in 'indexclauses', they are replaced by the
+ * derived indexscannable conditions in 'indexquals'.
+ *
+ * 'indexqualcols' is an integer list of index column numbers (zero-based)
+ * of the same length as 'indexquals', showing which index column each qual
+ * is meant to be used with.  'indexquals' is required to be ordered by
+ * index column, so 'indexqualcols' must form a nondecreasing sequence.
+ * (The order of multiple quals for the same index column is unspecified.)
+ *
+ * 'indexorderbys', if not NIL, is a list of ORDER BY expressions that have
+ * been found to be usable as ordering operators for an amcanorderbyop index.
+ * The list must match the path's pathkeys, ie, one expression per pathkey
+ * in the same order.  These are not RestrictInfos, just bare expressions,
  * since they generally won't yield booleans.  Also, unlike the case for
  * quals, it's guaranteed that each expression has the index key on the left
- * side of the operator.  The top list has one entry per pathkey in the
- * path's pathkeys, and the sub-lists have one sub-sublist per index column.
- * This representation is a bit of overkill, since there will be only one
- * actual expression per pathkey, but it's convenient because each sub-list
- * has the same structure as the indexquals list.
+ * side of the operator.
+ *
+ * 'indexorderbycols' is an integer list of index column numbers (zero-based)
+ * of the same length as 'indexorderbys', showing which index column each
+ * ORDER BY expression is meant to be used with.  (There is no restriction
+ * on which index column each ORDER BY can be used with.)
  *
  * 'isjoininner' is TRUE if the path is a nestloop inner scan (that is,
  * some of the index conditions are join rather than restriction clauses).
@@ -711,7 +717,9 @@ typedef struct IndexPath
 	IndexOptInfo *indexinfo;
 	List	   *indexclauses;
 	List	   *indexquals;
+	List	   *indexqualcols;
 	List	   *indexorderbys;
+	List	   *indexorderbycols;
 	bool		isjoininner;
 	ScanDirection indexscandir;
 	Cost		indextotalcost;

src/include/optimizer/cost.h

@@ -67,9 +67,8 @@ extern double clamp_row_est(double nrows);
 extern double index_pages_fetched(double tuples_fetched, BlockNumber pages,
 					double index_pages, PlannerInfo *root);
 extern void cost_seqscan(Path *path, PlannerInfo *root, RelOptInfo *baserel);
-extern void cost_index(IndexPath *path, PlannerInfo *root, IndexOptInfo *index,
-		   List *indexQuals, List *indexOrderBys,
-		   bool indexonly, RelOptInfo *outer_rel);
+extern void cost_index(IndexPath *path, PlannerInfo *root,
+					   RelOptInfo *outer_rel);
 extern void cost_bitmap_heap_scan(Path *path, PlannerInfo *root, RelOptInfo *baserel,
 					  Path *bitmapqual, RelOptInfo *outer_rel);
 extern void cost_bitmap_and_node(BitmapAndPath *path, PlannerInfo *root);

src/include/optimizer/pathnode.h

@@ -30,8 +30,10 @@ extern void add_path(RelOptInfo *parent_rel, Path *new_path);
 extern Path *create_seqscan_path(PlannerInfo *root, RelOptInfo *rel);
 extern IndexPath *create_index_path(PlannerInfo *root,
 				  IndexOptInfo *index,
-				  List *clause_groups,
+				  List *indexclauses,
+				  List *indexclausecols,
 				  List *indexorderbys,
+				  List *indexorderbycols,
 				  List *pathkeys,
 				  ScanDirection indexscandir,
 				  bool indexonly,

src/include/optimizer/paths.h

@@ -57,11 +57,10 @@ extern bool eclass_matches_any_index(EquivalenceClass *ec,
 						 RelOptInfo *rel);
 extern bool match_index_to_operand(Node *operand, int indexcol,
 					   IndexOptInfo *index);
-extern List *expand_indexqual_conditions(IndexOptInfo *index,
-							List *clausegroups);
+extern void expand_indexqual_conditions(IndexOptInfo *index,
+							List *indexclauses, List *indexclausecols,
+							List **indexquals_p, List **indexqualcols_p);
 extern void check_partial_indexes(PlannerInfo *root, RelOptInfo *rel);
-extern List *flatten_clausegroups_list(List *clausegroups);
-extern List *flatten_indexorderbys_list(List *indexorderbys);
 extern Expr *adjust_rowcompare_for_index(RowCompareExpr *clause,
 							IndexOptInfo *index,
 							int indexcol,

src/include/pg_config_manual.h

@@ -22,7 +22,7 @@
 /*
  * Maximum number of arguments to a function.
  *
- * The minimum value is 9 (index cost estimation uses 9-argument functions).
+ * The minimum value is 8 (GIN indexes use 8-argument support functions).
  * The maximum possible value is around 600 (limited by index tuple size in
  * pg_proc's index; BLCKSZ larger than 8K would allow more).  Values larger
  * than needed will waste memory and processing time, but do not directly