Tighten selection of equality and ordering operators for grouping

operations: make sure we use operators that are compatible, as determined
by a mergejoin link in pg_operator.  Also, add code to planner to ensure
we don't try to use hashed grouping when the grouping operators aren't
marked hashable.

Author: tglsfdc

src/backend/commands/analyze.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/commands/analyze.c,v 1.50 2002/11/13 00:39:46 momjian Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/commands/analyze.c,v 1.51 2002/11/29 21:39:10 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -402,10 +402,7 @@ examine_attribute(Relation onerel, int attnum)
 		return NULL;
 
 	/* If column has no "=" operator, we can't do much of anything */
-	func_operator = compatible_oper(makeList1(makeString("=")),
-									attr->atttypid,
-									attr->atttypid,
-									true);
+	func_operator = equality_oper(attr->atttypid, true);
 	if (func_operator != NULL)
 	{
 		oprrest = ((Form_pg_operator) GETSTRUCT(func_operator))->oprrest;
@@ -443,10 +440,7 @@ examine_attribute(Relation onerel, int attnum)
 		stats->attr->attstattarget = default_statistics_target;
 
 	/* Is there a "<" operator with suitable semantics? */
-	func_operator = compatible_oper(makeList1(makeString("<")),
-									attr->atttypid,
-									attr->atttypid,
-									true);
+	func_operator = ordering_oper(attr->atttypid, true);
 	if (func_operator != NULL)
 	{
 		oprrest = ((Form_pg_operator) GETSTRUCT(func_operator))->oprrest;

src/backend/executor/nodeAgg.c

@@ -45,7 +45,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/executor/nodeAgg.c,v 1.96 2002/11/19 23:21:57 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/executor/nodeAgg.c,v 1.97 2002/11/29 21:39:11 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -1321,14 +1321,9 @@ ExecInitAgg(Agg *node, EState *estate, Plan *parent)
 							&peraggstate->inputtypeLen,
 							&peraggstate->inputtypeByVal);
 
-			eq_function = compatible_oper_funcid(makeList1(makeString("=")),
-												 inputType, inputType,
-												 true);
-			if (!OidIsValid(eq_function))
-				elog(ERROR, "Unable to identify an equality operator for type %s",
-					 format_type_be(inputType));
+			eq_function = equality_oper_funcid(inputType);
 			fmgr_info(eq_function, &(peraggstate->equalfn));
-			peraggstate->sortOperator = any_ordering_op(inputType);
+			peraggstate->sortOperator = ordering_oper_opid(inputType);
 			peraggstate->sortstate = NULL;
 		}
 

src/backend/executor/nodeGroup.c

@@ -15,7 +15,7 @@
  *	  locate group boundaries.
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/executor/nodeGroup.c,v 1.49 2002/11/06 22:31:23 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/executor/nodeGroup.c,v 1.50 2002/11/29 21:39:11 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -353,11 +353,7 @@ execTuplesMatchPrepare(TupleDesc tupdesc,
 		Oid			typid = tupdesc->attrs[att - 1]->atttypid;
 		Oid			eq_function;
 
-		eq_function = compatible_oper_funcid(makeList1(makeString("=")),
-											 typid, typid, true);
-		if (!OidIsValid(eq_function))
-			elog(ERROR, "Unable to identify an equality operator for type %s",
-				 format_type_be(typid));
+		eq_function = equality_oper_funcid(typid);
 		fmgr_info(eq_function, &eqfunctions[i]);
 	}
 

src/backend/optimizer/plan/planner.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/optimizer/plan/planner.c,v 1.131 2002/11/26 03:01:58 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/optimizer/plan/planner.c,v 1.132 2002/11/29 21:39:11 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -17,6 +17,7 @@
 
 #include <limits.h>
 
+#include "catalog/pg_operator.h"
 #include "catalog/pg_type.h"
 #include "miscadmin.h"
 #include "nodes/makefuncs.h"
@@ -36,9 +37,11 @@
 #include "parser/analyze.h"
 #include "parser/parsetree.h"
 #include "parser/parse_expr.h"
+#include "parser/parse_oper.h"
 #include "rewrite/rewriteManip.h"
 #include "utils/lsyscache.h"
 #include "utils/selfuncs.h"
+#include "utils/syscache.h"
 
 
 /* Expression kind codes for preprocess_expression */
@@ -57,6 +60,7 @@ static Node *preprocess_expression(Query *parse, Node *expr, int kind);
 static void preprocess_qual_conditions(Query *parse, Node *jtnode);
 static Plan *inheritance_planner(Query *parse, List *inheritlist);
 static Plan *grouping_planner(Query *parse, double tuple_fraction);
+static bool hash_safe_grouping(Query *parse);
 static List *make_subplanTargetList(Query *parse, List *tlist,
 					   AttrNumber **groupColIdx);
 static Plan *make_groupsortplan(Query *parse,
@@ -1252,11 +1256,14 @@ grouping_planner(Query *parse, double tuple_fraction)
 			numGroups = (long) Min(dNumGroups, (double) LONG_MAX);
 
 			/*
+			 * Check can't-do-it conditions, including whether the grouping
+			 * operators are hashjoinable.
+			 *
 			 * Executor doesn't support hashed aggregation with DISTINCT
 			 * aggregates.  (Doing so would imply storing *all* the input
 			 * values in the hash table, which seems like a certain loser.)
 			 */
-			if (!enable_hashagg)
+			if (!enable_hashagg || !hash_safe_grouping(parse))
 				use_hashed_grouping = false;
 			else if (parse->hasAggs &&
 					 (contain_distinct_agg_clause((Node *) tlist) ||
@@ -1554,6 +1561,33 @@ grouping_planner(Query *parse, double tuple_fraction)
 	return result_plan;
 }
 
+/*
+ * hash_safe_grouping - are grouping operators hashable?
+ *
+ * We assume hashed aggregation will work if the datatype's equality operator
+ * is marked hashjoinable.
+ */
+static bool
+hash_safe_grouping(Query *parse)
+{
+	List	   *gl;
+
+	foreach(gl, parse->groupClause)
+	{
+		GroupClause *grpcl = (GroupClause *) lfirst(gl);
+		TargetEntry *tle = get_sortgroupclause_tle(grpcl, parse->targetList);
+		Operator	optup;
+		bool		oprcanhash;
+
+		optup = equality_oper(tle->resdom->restype, false);
+		oprcanhash = ((Form_pg_operator) GETSTRUCT(optup))->oprcanhash;
+		ReleaseSysCache(optup);
+		if (!oprcanhash)
+			return false;
+	}
+	return true;
+}
+
 /*---------------
  * make_subplanTargetList
  *	  Generate appropriate target list when grouping is required.

src/backend/parser/parse_clause.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/parser/parse_clause.c,v 1.99 2002/11/15 02:50:08 momjian Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/parser/parse_clause.c,v 1.100 2002/11/29 21:39:11 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -1128,8 +1128,7 @@ findTargetlistEntry(ParseState *pstate, Node *node, List *tlist, int clause)
 
 /*
  * transformGroupClause -
- *	  transform a Group By clause
- *
+ *	  transform a GROUP BY clause
  */
 List *
 transformGroupClause(ParseState *pstate, List *grouplist, List *targetlist)
@@ -1151,7 +1150,7 @@ transformGroupClause(ParseState *pstate, List *grouplist, List *targetlist)
 
 			grpcl->tleSortGroupRef = assignSortGroupRef(tle, targetlist);
 
-			grpcl->sortop = any_ordering_op(tle->resdom->restype);
+			grpcl->sortop = ordering_oper_opid(tle->resdom->restype);
 
 			glist = lappend(glist, grpcl);
 		}
@@ -1331,7 +1330,7 @@ addAllTargetsToSortList(List *sortlist, List *targetlist)
  * addTargetToSortList
  *		If the given targetlist entry isn't already in the ORDER BY list,
  *		add it to the end of the list, using the sortop with given name
- *		or any available sort operator if opname == NIL.
+ *		or the default sort operator if opname == NIL.
  *
  * Returns the updated ORDER BY list.
  */
@@ -1352,7 +1351,7 @@ addTargetToSortList(TargetEntry *tle, List *sortlist, List *targetlist,
 												  tle->resdom->restype,
 												  false);
 		else
-			sortcl->sortop = any_ordering_op(tle->resdom->restype);
+			sortcl->sortop = ordering_oper_opid(tle->resdom->restype);
 
 		sortlist = lappend(sortlist, sortcl);
 	}

src/backend/parser/parse_oper.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/parser/parse_oper.c,v 1.60 2002/09/18 21:35:22 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/parser/parse_oper.c,v 1.61 2002/11/29 21:39:11 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -130,22 +130,116 @@ LookupOperNameTypeNames(List *opername, TypeName *oprleft,
 	return operoid;
 }
 
+/*
+ * equality_oper - identify a suitable equality operator for a datatype
+ *
+ * On failure, return NULL if noError, else report a standard error
+ */
+Operator
+equality_oper(Oid argtype, bool noError)
+{
+	Operator	optup;
 
-/* Select an ordering operator for the given datatype */
-Oid
-any_ordering_op(Oid argtype)
+	/*
+	 * Look for an "=" operator for the datatype.  We require it to be
+	 * an exact or binary-compatible match, since most callers are not
+	 * prepared to cope with adding any run-time type coercion steps.
+	 */
+	optup = compatible_oper(makeList1(makeString("=")),
+							argtype, argtype, true);
+	if (optup != NULL)
+	{
+		/*
+		 * Only believe that it's equality if it's mergejoinable,
+		 * hashjoinable, or uses eqsel() as oprrest.
+		 */
+		Form_pg_operator pgopform = (Form_pg_operator) GETSTRUCT(optup);
+
+		if (OidIsValid(pgopform->oprlsortop) ||
+			pgopform->oprcanhash ||
+			pgopform->oprrest == F_EQSEL)
+			return optup;
+
+		ReleaseSysCache(optup);
+	}
+	if (!noError)
+		elog(ERROR, "Unable to identify an equality operator for type %s",
+			 format_type_be(argtype));
+	return NULL;
+}
+
+/*
+ * ordering_oper - identify a suitable sorting operator ("<") for a datatype
+ *
+ * On failure, return NULL if noError, else report a standard error
+ */
+Operator
+ordering_oper(Oid argtype, bool noError)
 {
-	Oid			order_opid;
+	Operator	optup;
 
-	order_opid = compatible_oper_opid(makeList1(makeString("<")),
-									  argtype, argtype, true);
-	if (!OidIsValid(order_opid))
-		elog(ERROR, "Unable to identify an ordering operator '%s' for type '%s'"
+	/*
+	 * Find the type's equality operator, and use its lsortop (it *must*
+	 * be mergejoinable).  We use this definition because for sorting and
+	 * grouping purposes, it's important that the equality and ordering
+	 * operators are consistent.
+	 */
+	optup = equality_oper(argtype, noError);
+	if (optup != NULL)
+	{
+		Oid		lsortop = ((Form_pg_operator) GETSTRUCT(optup))->oprlsortop;
+
+		ReleaseSysCache(optup);
+
+		if (OidIsValid(lsortop))
+		{
+			optup = SearchSysCache(OPEROID,
+								   ObjectIdGetDatum(lsortop),
+								   0, 0, 0);
+			if (optup != NULL)
+				return optup;
+		}
+	}
+	if (!noError)
+		elog(ERROR, "Unable to identify an ordering operator for type %s"
 			 "\n\tUse an explicit ordering operator or modify the query",
-			 "<", format_type_be(argtype));
-	return order_opid;
+			 format_type_be(argtype));
+	return NULL;
+}
+
+/*
+ * equality_oper_funcid - convenience routine for oprfuncid(equality_oper())
+ */
+Oid
+equality_oper_funcid(Oid argtype)
+{
+	Operator	optup;
+	Oid			result;
+
+	optup = equality_oper(argtype, false);
+	result = oprfuncid(optup);
+	ReleaseSysCache(optup);
+	return result;
+}
+
+/*
+ * ordering_oper_opid - convenience routine for oprid(ordering_oper())
+ *
+ * This was formerly called any_ordering_op()
+ */
+Oid
+ordering_oper_opid(Oid argtype)
+{
+	Operator	optup;
+	Oid			result;
+
+	optup = ordering_oper(argtype, false);
+	result = oprid(optup);
+	ReleaseSysCache(optup);
+	return result;
 }
 
+
 /* given operator tuple, return the operator OID */
 Oid
 oprid(Operator op)
@@ -731,28 +825,6 @@ compatible_oper_opid(List *op, Oid arg1, Oid arg2, bool noError)
 	return InvalidOid;
 }
 
-/* compatible_oper_funcid() -- get OID of a binary operator's function
- *
- * This is a convenience routine that extracts only the function OID
- * from the result of compatible_oper().  InvalidOid is returned if the
- * lookup fails and noError is true.
- */
-Oid
-compatible_oper_funcid(List *op, Oid arg1, Oid arg2, bool noError)
-{
-	Operator	optup;
-	Oid			result;
-
-	optup = compatible_oper(op, arg1, arg2, noError);
-	if (optup != NULL)
-	{
-		result = oprfuncid(optup);
-		ReleaseSysCache(optup);
-		return result;
-	}
-	return InvalidOid;
-}
-
 
 /* right_oper() -- search for a unary right operator (operator on right)
  * Given operator name and type of arg, return oper struct.