Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-5617

mysqld crashes when running a query with ONLY_FULL_GROUP_BY

Details

    • Bug
    • Status: Closed (View Workflow)
    • Critical
    • Resolution: Fixed
    • 5.5.35
    • 5.5.36, 10.0.9
    • None
    • None
    • Linux 64-bit, Gentoo package

    Description

      I tried to run this query:

      SELECT
          b.ekperif, s.short_description, s.imagesXrwma
      FROM
          (SELECT
              ps.synolo,
                  se.id_syndiasmosSeEnotita,
                  se.id_syndiasmos,
                  se.id_perifereiakiEnotita,
                  se.id_eklogikhPerifereia
          FROM
              (SELECT
              SUM(ps.psifoiSyndiasmou) AS synolo,
                  ps.id_syndiasmosSeEnotita
          FROM
              psifoiSyndiasmon AS ps
          WHERE
              ps.id_diexagogiEklogon = 3
          GROUP BY (ps.id_syndiasmosSeEnotita)) AS ps
          LEFT JOIN syndiasmosSeEnotita se ON ps.id_syndiasmosSeEnotita =
      se.id_syndiasmosSeEnotita
          WHERE
              se.active = 1) AS a
              INNER JOIN
          (SELECT
              MAX(t.synolo) AS maxsynolo,
                  t.id_eklogikhPerifereia AS ekperif
          FROM
              (SELECT
              ps.synolo,
                  se.id_syndiasmosSeEnotita,
                  se.id_syndiasmos,
                  se.id_perifereiakiEnotita,
                  se.id_eklogikhPerifereia
          FROM
              (SELECT
              SUM(ps.psifoiSyndiasmou) AS synolo,
                  ps.id_syndiasmosSeEnotita
          FROM
              psifoiSyndiasmon AS ps
          WHERE
              ps.id_diexagogiEklogon = 3
          GROUP BY (ps.id_syndiasmosSeEnotita)) AS ps
          LEFT JOIN syndiasmosSeEnotita se ON ps.id_syndiasmosSeEnotita =
      se.id_syndiasmosSeEnotita
          WHERE
              se.active = 1) AS t
          GROUP BY t.id_eklogikhPerifereia) AS b ON a.id_eklogikhPerifereia =
      b.ekperif
              LEFT JOIN
          syndiasmos s ON a.id_syndiasmos = s.id_syndiasmos
      WHERE
          a.synolo = b.maxsynolo;

      and the server crashed. Log:

      140205 15:58:42 Percona XtraDB (http://www.percona.com) 5.5.35-MariaDB-33.0 started; log sequence number 151241109
      140205 15:58:42 [Note] Event Scheduler: Loaded 0 events
      140205 15:58:42 [Note] /usr/sbin/mysqld: ready for connections.
      Version: '5.5.35-MariaDB-log'  socket: '/var/run/mysqld/mysqld.sock'  port: 0  Source distribution
      140205 17:24:25 [ERROR] mysqld got signal 11 ;
      This could be because you hit a bug. It is also possible that this binary
      or one of the libraries it was linked against is corrupt, improperly built,
      or misconfigured. This error can also be caused by malfunctioning hardware.
       
      To report this bug, see http://kb.askmonty.org/en/reporting-bugs
       
      We will try our best to scrape up some info that will hopefully help
      diagnose the problem, but since we have already crashed, 
      something is definitely wrong and this may fail.
       
      Server version: 5.5.35-MariaDB-log
      key_buffer_size=25165824
      read_buffer_size=131072
      max_used_connections=1
      max_threads=52
      thread_count=1
      It is possible that mysqld could use up to 
      key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 138638 K  bytes of memory
      Hope that's ok; if not, decrease some variables in the equation.
       
      Thread pointer: 0x0x7fb8acac1000
      Attempting backtrace. You can use the following information to find out
      where mysqld died. If you see no messages after this, something went
      terribly wrong...
      stack_bottom = 0x7fb8c88afed8 thread_stack 0x48000
      /usr/sbin/mysqld(my_print_stacktrace+0x26)[0x9ca0d6]
      /usr/sbin/mysqld(handle_fatal_signal+0x398)[0x69a458]
      /lib64/libpthread.so.0(+0xfbb0)[0x7fb8c9cc1bb0]
      /usr/sbin/mysqld(_ZN10Item_field10fix_fieldsEP3THDPP4Item+0x19e)[0x6b416e]
      /usr/sbin/mysqld(_ZN9Item_func10fix_fieldsEP3THDPP4Item+0x180)[0x6dddc0]
      /usr/sbin/mysqld(_Z11setup_condsP3THDP10TABLE_LISTR4ListIS1_EPP4Item+0x1e6)[0x550fa6]
      /usr/sbin/mysqld(_ZN4JOIN7prepareEPPP4ItemP10TABLE_LISTjS1_jP8st_orderbS7_S1_S7_P13st_select_lexP18st_select_lex_unit+0x394)[0x5bf8f4]
      /usr/sbin/mysqld(_ZN18st_select_lex_unit7prepareEP3THDP13select_resultm+0x95d)[0x5feced]
      /usr/sbin/mysqld(_Z21mysql_derived_prepareP3THDP3LEXP10TABLE_LIST+0x235)[0x56dc45]
      /usr/sbin/mysqld(_Z27mysql_handle_single_derivedP3LEXP10TABLE_LISTj+0xec)[0x56e92c]
      /usr/sbin/mysqld(_ZN13st_select_lex14handle_derivedEP3LEXj+0x3e)[0x580bce]
      /usr/sbin/mysqld(_ZN4JOIN7prepareEPPP4ItemP10TABLE_LISTjS1_jP8st_orderbS7_S1_S7_P13st_select_lexP18st_select_lex_unit+0xea)[0x5bf64a]
      /usr/sbin/mysqld(_Z12mysql_selectP3THDPPP4ItemP10TABLE_LISTjR4ListIS1_ES2_jP8st_orderSB_S2_SB_yP13select_resultP18st_select_lex_unitP13st_select_lex+0x8ec)[0x5c9d5c]
      /usr/sbin/mysqld(_Z13handle_selectP3THDP3LEXP13select_resultm+0x2cc)[0x5cfabc]
      /usr/sbin/mysqld[0x58388b]
      /usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x4a02)[0x58d272]
      /usr/sbin/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x119)[0x58f479]
      /usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0x135e)[0x59086e]
      /usr/sbin/mysqld(_Z24do_handle_one_connectionP3THD+0x173)[0x62da93]
      /usr/sbin/mysqld(handle_one_connection+0x4a)[0x62db6a]
      /lib64/libpthread.so.0(+0x7fda)[0x7fb8c9cb9fda]
      /lib64/libc.so.6(clone+0x6d)[0x7fb8c8db99ad]
       
      Trying to get some variables.
      Some pointers may be invalid and cause the dump to abort.
      Query (0x7fb8a1c14018): is an invalid pointer
      Connection ID (thread ID): 13
      Status: NOT_KILLED

      Attachments

        Issue Links

          Activity

            Are there any blocking bugs for 5.5.35 (that crash the server I mean)?

            There are several known crashing or potentially crashing bugs in 5.5.35, but none of them is a regression from 5.5.34 (like the one that you filed), and most of them are related to rather specific areas: virtual columns, spatial functions, prepared statements, dynamic change of certain system variables in concurrent environment, things like that.

            elenst Elena Stepanova added a comment - Are there any blocking bugs for 5.5.35 (that crash the server I mean)? There are several known crashing or potentially crashing bugs in 5.5.35, but none of them is a regression from 5.5.34 (like the one that you filed), and most of them are related to rather specific areas: virtual columns, spatial functions, prepared statements, dynamic change of certain system variables in concurrent environment, things like that.

            The problem was that in a few places, we still depended on that cached_table would have a value. I have now fixed this. Will run tests and push

            monty Michael Widenius added a comment - The problem was that in a few places, we still depended on that cached_table would have a value. I have now fixed this. Will run tests and push

            Fixed and pushed

            monty Michael Widenius added a comment - Fixed and pushed

            Thank you guys, you are great!

            How soon will 5.5.36 get released, do you have a release date?

            lourdas_v Vasilis Lourdas added a comment - Thank you guys, you are great! How soon will 5.5.36 get released, do you have a release date?

            Currently the planned release date is 2014-02-21. It can change later, but should not change too much.
            You can always find the latest known plan at the main page of our JIRA (https://mariadb.atlassian.net/secure/Dashboard.jspa), "Road Map" section.

            elenst Elena Stepanova added a comment - Currently the planned release date is 2014-02-21. It can change later, but should not change too much. You can always find the latest known plan at the main page of our JIRA ( https://mariadb.atlassian.net/secure/Dashboard.jspa ), "Road Map" section.

            People

              monty Michael Widenius
              lourdas_v Vasilis Lourdas
              Votes:
              0 Vote for this issue
              Watchers:
              6 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Git Integration

                  Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.