Repair breakage of aggregate FILTER option.

An aggregate's input expression(s) are not supposed to be evaluated
at all for a row where its FILTER test fails ... but commit 8ed3f11b
overlooked that requirement.  Reshuffle so that aggregates having a
filter clause evaluate their arguments separately from those without.
This still gets the benefit of doing only one ExecProject in the
common case of multiple Aggrefs, none of which have filters.

While at it, arrange for filter clauses to be included in the common
ExecProject evaluation, thus perhaps buying a little bit even when
there are filters.

Back-patch to v10 where the bug was introduced.

Discussion: https://postgr.es/m/30065.1508161354@sss.pgh.pa.us

src/include/nodes/execnodes.h

@@ -1830,10 +1830,8 @@ typedef struct AggState
 	int			num_hashes;
 	AggStatePerHash perhash;
 	AggStatePerGroup *hash_pergroup;	/* array of per-group pointers */
-	/* support for evaluation of agg inputs */
-	TupleTableSlot *evalslot;	/* slot for agg inputs */
-	ProjectionInfo *evalproj;	/* projection machinery */
-	TupleDesc	evaldesc;		/* descriptor of input tuples */
+	/* support for evaluation of agg input expressions: */
+	ProjectionInfo *combinedproj;	/* projection machinery */
 } AggState;
 
 /* ----------------

src/test/regress/expected/aggregates.out

@@ -1388,6 +1388,12 @@ select min(unique1) filter (where unique1 > 100) from tenk1;
  101
 (1 row)
 
+select sum(1/ten) filter (where ten > 0) from tenk1;
+ sum  
+------
+ 1000
+(1 row)
+
 select ten, sum(distinct four) filter (where four::text ~ '123') from onek a
 group by ten;
  ten | sum 

src/test/regress/sql/aggregates.sql

@@ -524,6 +524,8 @@ drop table bytea_test_table;
 
 select min(unique1) filter (where unique1 > 100) from tenk1;
 
+select sum(1/ten) filter (where ten > 0) from tenk1;
+
 select ten, sum(distinct four) filter (where four::text ~ '123') from onek a
 group by ten;