Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-6151

Wrong result (missing row) with LEFT JOIN, HAVING and ORDER BY

    XMLWordPrintable

    Details

      Description

      Test case:

      CREATE TABLE t1 (a INT) ENGINE=MyISAM;
      INSERT INTO t1 VALUES (3),(6);
       
      CREATE TABLE t2 (b INT) ENGINE=MyISAM;
      INSERT INTO t2 VALUES (8),(9);
       
      SELECT * FROM t1 LEFT JOIN t2 ON ( a = b ) HAVING a NOT IN ( SELECT 6 ) ORDER BY a ;
       
      DROP TABLE t1, t2;

      On current MariaDB 5.1, 5.2, 5.3, MySQL 5.1, 5.5 it returns the wrong result (empty set):

      a	b

      On current MariaDB 5.5, 10.0, MySQL 5.6, 57 it returns the expected result (1 row):

      a	b
      3	NULL

      Story:

      The bug was introduced in MySQL 5.1 (and merged up) by the following revision:

      revno: 4005
      revision-id: sergey.glukhov@oracle.com-20130507091058-x4li5jat6pvhop96
      parent: annamalai.gurusami@oracle.com-20130506105856-d5il8vpqcovru47g
      committer: Sergey Glukhov <sergey.glukhov@oracle.com>
      branch nick: mysql-5.1
      timestamp: Tue 2013-05-07 13:10:58 +0400
      message:
        Bug#16095534 CRASH: PREPARED STATEMENT CRASHES IN ITEM_BOOL_FUNC2::FIX_LENGTH_AND_DEC
        The problem happened due to broken left expression in Item_in_optimizer object.
        In case of the bug left expression is runtime created Item_outer_ref item which
        is deleted at the end of the statement and one of Item_in_optimizer arguments
        becomes bad when re-executed. The fix is to use real_item() instead of original
        left expression. Note: It feels a bit weird that after preparing, the field is
        directly part of the generated Item_func_eq, whereas in execution it is replaced
        with an Item_outer_ref wrapper object.

      Later the bug was fixed in MySQL 5.6 and up by the following revision:

       revno: 5849
      revision-id: chaithra.gopalareddy@oracle.com-20140228093141-v2sgort4nzof1ydg
      parent: georgi.kodinov@oracle.com-20140228084025-kijma5qd98ttna38
      committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
      branch nick: mysql-5.6
      timestamp: Fri 2014-02-28 15:01:41 +0530
      message:
        Bug#18014565: WRONG RESULT COMPUTATION USING ALL() AND GROUP BY

      The first change was merged into MySQL 5.5 and made it to MariaDB 5.5; but then the bug was fixed in MariaDB 5.5 by this revision:

       revno: 3895
      revision-id: sergii@pisem.net-20130913214200-0b5q0uanbj6mtti7
      parent: bar@mnogosearch.org-20130913104740-3nvxlwt6h9pxatqz
      fixes bug: https://mariadb.atlassian.net/browse/MDEV-5012
      committer: Sergei Golubchik <sergii@pisem.net>
      branch nick: 5.5
      timestamp: Fri 2013-09-13 23:42:00 +0200
      message:
        MDEV-5012 Server crashes in Item_ref::real_item on EXPLAIN with select subqueries or views, constant table, derived_merge+derived_with_keys
        
        revert incorrect change, merged from mysql-5.5

      So, it's not reproducible on MariaDB 5.5 and 10.0.

      Recently, MySQL 5.1 was merged into MariaDB 5.1/5.2/5.3. The faulty change was merged too, so the wrong result started appearing on these trees. But the fix for it only exists in MySQL 5.6 or MariaDB 5.5, so apparently the bug will stay in MySQL 5.1/5.5 forever, and in MariaDB 5.1-5.3 too unless we decide it's worth fixing.

      Please decide.

        Attachments

          Activity

            People

            Assignee:
            psergei Sergei Petrunia
            Reporter:
            elenst Elena Stepanova
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved:

                Git Integration