Change cost for REF to take into account cost for 1 extra key read_next

The main difference in code path between EQ_REF and REF is that for
REF we have to do an extra read_next on the index to check that there
is no more matching rows.

Before this patch we added a preference of EQ_REF by ensuring that REF
would always estimate to find at least 2 rows.

This patch adds the cost of the extra key read_next to REF access and
removes the code that limited REF to at least 2 rows. For some queries
this can have a big effect as the total estimated rows will be halved
for each REF table with 1 rows.

multi_range cost calculations are also changed to take into account
the difference between EQ_REF and REF.

The effect of the patch to the test suite:
- About 80 test case changed
- Almost all changes where for EXPLAIN where estimated rows for REF
  where changed from 2 to 1.
- A few test cases using explain extended had a change of 'filtered'.
  This is because of the estimated rows are now closer to the
  calculated selectivity.
- A very few test had a change of table order.
  This is because the change of estimated rows from 2 to 1 or the small
  cost change for REF
  (main.subselect_sj_jcl6, main.group_by, main.dervied_cond_pushdown,
  main.distinct, main.join_nested, main.order_by, main.join_cache)
- No key statistics and the estimated rows are now smaller which cased
  estimated filtering to be lower.
  (main.subselect_sj_mat)
- The number of total rows are halved.
  (main.derived_cond_pushdown)
- Plans with 1 row changed to use RANGE instead of REF.
  (main.group_min_max)
- ALL changed to REF
  (main.key_diff)
- Key changed from ref + index_only to PRIMARY key for InnoDB, as
  OPTIMIZER_ROW_LOOKUP_COST + OPTIMIZER_ROW_NEXT_FIND_COST is smaller than
  OPTIMIZER_KEY_LOOKUP_COST + OPTIMIZER_KEY_NEXT_FIND_COST.
  (main.join_outer_innodb)
- Cost changes printouts
  (main.opt_trace*)
- Result order change
  (innodb_gis.rtree)

mysql-test/main/derived_split_innodb.result

@@ -20,7 +20,7 @@ EXPLAIN SELECT t1.n1 FROM t1, (SELECT n1, n2 FROM t1 WHERE c1 = 'a' GROUP BY n1)
 WHERE t.n1 = t1.n1 AND t.n2 = t1.n2 AND c1 = 'a' GROUP BY n1;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t1	ref	c1,n1_c1_n2	c1	1	const	2	Using index condition; Using where; Using temporary; Using filesort
-1	PRIMARY	<derived2>	ref	key0	key0	8	test.t1.n1,test.t1.n2	2	
+1	PRIMARY	<derived2>	ref	key0	key0	8	test.t1.n1,test.t1.n2	1	
 2	LATERAL DERIVED	t1	ref	c1,n1_c1_n2	n1_c1_n2	4	test.t1.n1	1	Using where; Using index
 SELECT t1.n1 FROM t1, (SELECT n1, n2 FROM t1 WHERE c1 = 'a' GROUP BY n1) as t
 WHERE t.n1 = t1.n1 AND t.n2 = t1.n2 AND c1 = 'a' GROUP BY n1;
@@ -49,7 +49,7 @@ t2
 WHERE t2.id2=t.id2;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	1	Using where
-1	PRIMARY	<derived2>	ref	key0	key0	5	test.t2.id2	2	
+1	PRIMARY	<derived2>	ref	key0	key0	5	test.t2.id2	1	
 2	DERIVED	t3	ALL	NULL	NULL	NULL	NULL	1	Using where; Using temporary; Using filesort
 2	DERIVED	t1	eq_ref	PRIMARY,id2	PRIMARY	4	test.t3.i3	1	
 2	DERIVED	t2	ALL	NULL	NULL	NULL	NULL	1	Using where; Using join buffer (flat, BNL join)
@@ -99,7 +99,7 @@ ON t2.id=t.id
 WHERE t2.id < 3;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	range	PRIMARY	PRIMARY	4	NULL	2	Using where
-1	PRIMARY	<derived2>	ref	key0	key0	5	test.t2.id	2	
+1	PRIMARY	<derived2>	ref	key0	key0	5	test.t2.id	1	
 2	LATERAL DERIVED	t1	eq_ref	PRIMARY	PRIMARY	4	test.t2.id	1	
 set join_cache_level=default;
 DROP TABLE t1,t2;
@@ -162,7 +162,7 @@ WHERE
 t1.a = dt.a;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t1	index	a,a_2	a_2	10	NULL	6	Using where; Using index
-1	PRIMARY	<derived3>	ref	key0	key0	5	test.t1.a	2	
+1	PRIMARY	<derived3>	ref	key0	key0	5	test.t1.a	1	
 3	DERIVED	t1_inner	index	NULL	a_2	10	NULL	6	Using where; Using index
 3	DERIVED	t2_inner	ref	c	c	5	test.t1_inner.b	1	Using index
 set statement optimizer_switch='split_materialized=on' for  EXPLAIN
@@ -174,7 +174,7 @@ WHERE
 t1.a = dt.a;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t1	index	a,a_2	a_2	10	NULL	6	Using where; Using index
-1	PRIMARY	<derived3>	ref	key0	key0	5	test.t1.a	2	
+1	PRIMARY	<derived3>	ref	key0	key0	5	test.t1.a	1	
 3	DERIVED	t1_inner	index	a,a_2	a_2	10	NULL	6	Using where; Using index
 3	DERIVED	t2_inner	ref	c	c	5	test.t1_inner.b	1	Using index
 DROP TABLE t1, t2;
@@ -209,7 +209,7 @@ where t1.id = dt.id and t1.itemid = dt.itemid and t2.id=t1.itemid;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	1	
 1	PRIMARY	t1	ref	idx	idx	4	test.t2.id	1	
-1	PRIMARY	<derived2>	ref	key0	key0	9	test.t2.id,test.t1.id	2	
+1	PRIMARY	<derived2>	ref	key0	key0	9	test.t2.id,test.t1.id	1	
 2	DERIVED	t3	ref	idx1,idx2	idx1	4	const	5	Using where; Using index
 select t1.id, t1.itemid, dt.id, t2.id
 from t1,
@@ -228,7 +228,7 @@ where t1.id = dt.id and t1.itemid = dt.itemid and t2.id=t1.itemid;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	1	
 1	PRIMARY	t1	ref	idx	idx	4	test.t2.id	1	
-1	PRIMARY	<derived2>	ref	key0	key0	9	test.t2.id,test.t1.id	2	
+1	PRIMARY	<derived2>	ref	key0	key0	9	test.t2.id,test.t1.id	1	
 2	DERIVED	t3	ref	idx1	idx1	4	const	5	Using where; Using index
 select t1.id, t1.itemid, dt.id, t2.id
 from t1,
@@ -273,7 +273,7 @@ on t3.a=t.a and t3.c=t.c
 where t3.b > 15;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t3	range	idx_b	idx_b	5	NULL	2	Using index condition; Using where
-1	PRIMARY	<derived2>	ref	key0	key0	133	test.t3.a,test.t3.c	2	
+1	PRIMARY	<derived2>	ref	key0	key0	133	test.t3.a,test.t3.c	1	
 2	LATERAL DERIVED	t4	ref	idx	idx	133	test.t3.a,test.t3.c	1	
 # ... and if one adds WITH ROLLUP, then LATERAL DERIVED is no longer used:
 explain select t3.a,t3.c,t.max,t.min