[MDEV-14210] Wrong result set from a query with HAVING containing a subquery with SF call. Created: 2017-10-30  Updated: 2023-12-05

Status: Stalled
Project: MariaDB Server
Component/s: Optimizer
Affects Version/s: 10.2, 10.3, 10.4, 10.5, 10.6, 10.7, 10.8
Fix Version/s: 10.3

Type: Bug Priority: Major
Reporter: Igor Babaev Assignee: Michael Widenius
Resolution: Unresolved Votes: 0
Labels: 11.0-sel


 Description   

The query with HAVING from the following test case

SET @@sql_mode = '';
 
 
 
CREATE TABLE series (
 
  val INT(10) UNSIGNED NOT NULL
 
);
 
INSERT INTO series VALUES(1);
 
CREATE TABLE seq_calls (
 
  c INT
 
);
 
INSERT INTO seq_calls VALUES(0);
 
 
 
DELIMITER |;
 
CREATE FUNCTION next_seq_value() RETURNS INT
 
BEGIN
 
 DECLARE next_val INT;
 
 SELECT val INTO next_val FROM series;
 
 UPDATE series SET val=mod(val + 1, 2);
 
 UPDATE seq_calls SET c=c+1;
 
 RETURN next_val;
 
END;
 
|
 
DELIMITER ;|
 
 
 
CREATE TABLE t1 (t INT, u INT, KEY(t));
 
INSERT INTO t1 VALUES(10, 10), (11, 11), (12, 12), (12, 13),(14, 15), (15, 16),
 
                     (16, 17), (17, 18);
 
ANALYZE TABLE t1;
 
 
 
SELECT * FROM series, seq_calls;
 
SELECT t, next_seq_value() r FROM t1 FORCE INDEX(t)
 
GROUP BY t HAVING (SELECT 1 FROM t1 WHERE r = 1 LIMIT 1)
 
ORDER BY t1.u;
 
 
 
DROP TABLE t1;
 
DROP FUNCTION next_seq_value;
 
DROP TABLE series, seq_calls;

returns a wrong result set:

MariaDB [test]> SELECT t, next_seq_value() r FROM t1 FORCE INDEX(t)
 
    -> GROUP BY t HAVING (SELECT 1 FROM t1 WHERE r = 1 LIMIT 1)
 
    -> ORDER BY t1.u;
 
+------+------+
 
| t    | r    |
 
+------+------+
 
|   10 |    1 |
 
|   11 |    1 |
 
|   12 |    1 |
 
|   14 |    1 |
 
|   15 |    1 |
 
|   16 |    1 |
 
|   17 |    1 |
 
+------+------+

The expected result set is:

MariaDB [test]> SELECT t, next_seq_value() r FROM t1 FORCE INDEX(t)
 
    -> GROUP BY t HAVING (SELECT 1 FROM t1 WHERE r = 1 LIMIT 1)
 
    -> ORDER BY t1.u;
 
+------+------+
 
| t    | r    |
 
+------+------+
 
|   10 |    1 |
 
|   12 |    1 |
 
|   15 |    1 |
 
|   17 |    1 |
 
+------+------+

This bug is a regression as 5.5,10.0,10.1 return the right result set

 Comments   
Comment by Alice Sherepa [ 2022-11-24 ]

Currently 10.3-10.10 (10.3 f4a1298f245f678badc8a5b5) returns the expected result, but without FORCE INDEX the result is different:

SELECT t, next_seq_value() r FROM t1 FORCE INDEX(t)
 
GROUP BY t HAVING (SELECT 1 FROM t1 WHERE r = 1 LIMIT 1)
 
ORDER BY t1.u;
 
t	r
 
10	1
 
12	1
 
15	1
 
17	1

without FORCE INDEX:

SELECT t, next_seq_value() r FROM t1 
GROUP BY t HAVING (SELECT 1 FROM t1 WHERE r = 1 LIMIT 1)
 
ORDER BY t1.u;
 
t	r
 
10	1
 
12	1
 
14	1
 
16	1

on bb-11.0 ebd10cdcd5f87743ff3 - there is a wrong (2nd) result in both cases

Generated at Thu Feb 08 08:11:47 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.