Changed a rule to be cost based in test_if_cheaper_ordering

- Simplified test by setting read_time=DBL_MAX at start of loop if
  FORCE INDEX is used
- No need to test for 'group by' as the cost compare should handle it.
- Only one test change where index scan was replaced with table scan
 (correct)

mysql-test/suite/innodb/r/innodb_mysql.result

@@ -1193,7 +1193,7 @@ INSERT INTO t1 SELECT  a + 32, MOD(a + 32, 20), 1 FROM t1;
 INSERT INTO t1 SELECT  a + 64, MOD(a + 64, 20), 1 FROM t1;
 EXPLAIN SELECT b, SUM(c) FROM t1 GROUP BY b;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
-1	SIMPLE	t1	index	NULL	b	5	NULL	128	
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	128	Using temporary; Using filesort
 EXPLAIN SELECT SQL_BIG_RESULT b, SUM(c) FROM t1 GROUP BY b;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	128	Using filesort

sql/sql_select.cc

@@ -30550,6 +30550,13 @@ test_if_cheaper_ordering(const JOIN_TAB *tab, ORDER *order, TABLE *table,
       add("fanout", fanout);
   }
 
+  /*
+    Force using an index for sorting if there was no ref key
+    and FORCE INDEX was used.
+  */
+  if (table->force_index && ref_key < 0)
+    read_time= DBL_MAX;
+
   Json_writer_array possible_keys(thd,"possible_keys");
   for (nr=0; nr < table->s->keys ; nr++)
   {
@@ -30702,8 +30709,7 @@ test_if_cheaper_ordering(const JOIN_TAB *tab, ORDER *order, TABLE *table,
             there is either a group by or a FORCE_INDEX
           - If the new cost is better than read_time
         */
-        if (((table->force_index || group) && best_key < 0 && ref_key < 0) ||
-            range_cost < read_time)
+        if (range_cost < read_time)
         {
           read_time= range_cost;
           possible_key.add("chosen", true);