------------------------------------------------------------
revno: 8769
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Thu 2014-09-18 13:38:41 +0200
message:
  Follow patch to WL #6936, cset 8785 from trunk
------------------------------------------------------------
revno: 8768
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Thu 2014-09-18 13:34:08 +0200
message:
  Bug #19629959 M_I_DB: ROOT USER IS MISSING THE CREATE_TABLESPACE PRIVILEGE
  
    Merged cset 8876 from trunk
------------------------------------------------------------
revno: 8767
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Thu 2014-09-18 13:28:16 +0200
message:
  Bug #19621147 M_I_DB: EMPTY HELP TABLES
  
    Merged cset 8875 from trunk
------------------------------------------------------------
revno: 8766
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Thu 2014-09-11 13:09:09 +0200
message:
  Bug #19545350 MYSQL_INSTALL_DB --HELP MESSAGE PROBLEMS
  
    Merge of cset 8803 from trunk
------------------------------------------------------------
revno: 8765
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Wed 2014-09-10 14:12:59 +0200
message:
  Bug #19594823 ROOT RANDOM PASSWORD IS NOT GENERATED BY DEFAULT AS PART OF SOLARIS PKG INSTALL
  
    Remove --insecure from postinstall-solaris, it was a temporary fix
    This is a dummy empty commit, as fix has already been applied
------------------------------------------------------------
revno: 8764
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Wed 2014-09-10 14:10:00 +0200
message:
  BUG #19578256 --SSL SHOULD DEFAULT TO OFF FOR COMMUNITY EDITION
  
    Merge of cset 8810 from trunk
------------------------------------------------------------
revno: 8763
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Tue 2014-09-09 15:39:30 +0200
message:
  Bug #19498327 : M_I_DB FAILS WHEN BUILDING IN A SOURCE DIRECTORY
  
    Merge of cset 8786 from trunk
------------------------------------------------------------
revno: 8762
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Tue 2014-09-09 15:38:16 +0200
message:
  Disabling a test, will be undone next commit
------------------------------------------------------------
revno: 8761
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Tue 2014-09-09 15:28:46 +0200
message:
  Bug #19553266 MAIN.WL6219-UPGRADE ON PB2
  
    Merge of cset 8778 from trunk
------------------------------------------------------------
revno: 8760
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Tue 2014-09-09 15:26:54 +0200
message:
  Followup fix, cset 8774 from trunk
------------------------------------------------------------
revno: 8759
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Tue 2014-09-09 15:25:48 +0200
message:
  Followup fix, cset 8772 from trunk
------------------------------------------------------------
revno: 8758
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Tue 2014-09-09 15:24:10 +0200
message:
  Remove --insecure from postinstall-solaris, it was a temporary fix
------------------------------------------------------------
revno: 8757
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Mon 2014-09-08 16:16:00 +0200
message:
  Bug #19579105 SERVICE START FAILING WITH SLES11 5.7.5-COMMUNITY REPO PACKAGES
  
    Fixed by adding --rpm to mysql_install_db command
    Also some corrections enterprise -> commercial
------------------------------------------------------------
revno: 8756
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: rel-575
timestamp: Wed 2014-09-03 14:30:43 +0200
message:
  Use --loose to avoid test failing in server start
------------------------------------------------------------
revno: 8755
tags: clone-5.7.5-m15-build
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Tue 2014-09-02 09:56:28 +0200
message:
  Revert the following revisions:
  
  ------------------------------------------------------------
  revno: 8734
  committer: bin.x.su@oracle.com
  branch nick: mysql-trunk
  timestamp: Fri 2014-08-29 10:15:45 +0800
  message:
    Commit the missing test case result file for WL#6835.
  
  ------------------------------------------------------------
  revno: 8732 [merge]
  committer: Sunny Bains <Sunny.Bains@Oracle.Com>
  branch nick: trunk
  timestamp: Fri 2014-08-29 10:24:18 +1000
  message:
    WL#6835 - InnoDB: GCS Replication: Deterministic Deadlock Handling (High Prio Transactions in InnoDB)
    
    Introduce transaction priority. Transactions with a higher priority cannot
    be rolled back by transactions with a lower priority. A higher priority
    transaction will jump the lock wait queue and grab the record lock instead
    of waiting.
    
    This code is not currently visible to the users. However, there are debug
    tests that can exercise the code. It will probably require some additional
    work once it is used by GCS.
    
    rb#6036 Approved by Jimmy Yang.
------------------------------------------------------------
revno: 8754
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: 32bit-fail
timestamp: Mon 2014-09-01 14:04:04 +0200
message:
  Bug#19537790 INNODB_STATS_FETCH FAILS WITH WRONG CARDINALITY
               ESTIMATE ON 32 BIT PLATFORMS
      
  The innodb_stats_fetch test failed when running
  on 32 bit platforms due to an "off by one" cardinality number
  in the result from a query that read from the statitics table
  in information schema. This test failure is caused by WL#7339.
  
  The cardinality numbers that are retrieved from information schema
  is roughly calculated this way when the table is stored in InnoDB:
  
  1. InnoDB has the number of rows and the cardinality for the 
     index stored in the persistent statistics. In the failing
     case, InnoDB had 1000 as the number of rows and 3 as the 
     cardinality.
  2. InnoDB calculates the records per key value and stores this
     in the KEY object. This is calculated as 1000/3 and is thus
     333.333333.... when using the code from WL#7339 (before
     this worklog, the rec_per_key value was 166).
  3. When filling data into information schema, we re-calculate
     the cardinality number by using the records per key
     information (in sql_show.cc):
  
        double records= (show_table->file->stats.records /
                         key->records_per_key(j));
        table->field[9]->store((longlong) records, TRUE);
  
     in this case we first compute the cardinality to be
   
        records= 1000 / 333.3333... = 3.0000... or 2.9999999
  
     and then use the cast to get an integer value. On 64-bit
     platforms, the result of this was 3.00000 which was 
     casted to 3. On 32 bit platforms, the result was 2.999999
     which was casted to 2 before inserting it into the
     information schema table. (in the pre-wl7339 version, the
     calculated cardinality number was 6 for this case).
  
  This issue is caused by having converted to using a float numbers
  for the records per key estimate. So when re-calculating the
  cardinatily number in step 3 above, we can easily get a
  result that is just below the actual correct cardinality number
  and due to the use of cast, the number is always truncated.
  
  The suggested fix for this problem is to round the calculated
  cardinality number to the nearest integer value before 
  inserting it into the statistics table. This will both avoid the
  issue with different results on different platforms but it will
  also produce a more correct cardinality estimate.
  
  This change has caused a few other test result files to be
  re-recorded. The updated cardinality numbers are more correct
  than the previous.
------------------------------------------------------------
revno: 8753
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: compile-ps2
timestamp: Mon 2014-09-01 13:52:58 +0200
message:
  WL#7315 Optimizer cost model: main memory management of cost constants
    
  Follow-up patch: Fix compile failure in unit tests when compiling
  without performance schema. The fix is to remove the PSI_mutex_key
  object from being included in the unit tests. This is no
  longer needed since the unit test no longer needs to link in
  ha_resolve_by_name().
------------------------------------------------------------
revno: 8752
committer: Bharathy Satish <bharathy.x.satish@oracle.com>
branch nick: mysql-trunk-base
timestamp: Mon 2014-09-01 17:08:05 +0530
message:
  WL#7766: Deprecate EOF packet
  If SELECT query is cached and later if there exists a session
  state then the same SELECT query when run will not send OK packet
  with session state as the result sets is picked from cache. For
  now disabling deprecate_eof.test when query_cache is ON.
------------------------------------------------------------
revno: 8751 [merge]
committer: Akhil Mohan <akhil.mohan@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-09-01 16:43:29 +0530
message:
  BUG 19363801 Added Debian Ubuntu specific path for INFO files
    ------------------------------------------------------------
    revno: 3902.340.15
    committer: Akhil Mohan <akhil.mohan@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-09-01 16:39:42 +0530
    message:
      BUG 19363801 Added Debian Ubuntu specific path for INFO files
------------------------------------------------------------
revno: 8750
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-09-01 11:00:21 +0200
message:
  Added openssl_udf.so plugin to commercial packages
------------------------------------------------------------
revno: 8749 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-08-31 23:21:36 +0530
message:
  Bug#19145712 USER AFTER FREE / DOUBLE FREE ISSUE
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.340.14
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Sun 2014-08-31 23:00:06 +0530
    message:
      Bug#19145712 USER AFTER FREE / DOUBLE FREE ISSUE
      
      Fixing pb2 test script failure
------------------------------------------------------------
revno: 8748
committer: Gleb Shchepa <gleb.shchepa@oracle.com>
branch nick: trunk
timestamp: Fri 2014-08-29 16:24:04 +0400
message:
  WL#6263: Remove limited YEAR(2) support
  
  Minor after-commit fix
------------------------------------------------------------
revno: 8747
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Fri 2014-08-29 13:46:28 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Post-push fix: Add 3 #includes to fix Windows compile warnings.
------------------------------------------------------------
revno: 8746
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: 7316-trunk
timestamp: Fri 2014-08-29 13:13:33 +0200
message:
  WL#7316 Optimizer cost model: Command for online updating of
          cost model constants
              
  Implements the FLUSH OPTIMIZER_COSTS command. This command will
  reload the optimizer "cost constants" from the configuration
  tables into the "cost constant" cache. New sessions will use
  the updated "cost constants".
------------------------------------------------------------
revno: 8745
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk
timestamp: Fri 2014-08-29 13:29:53 +0400
message:
  Follow-up patch for WL#7828 & WL#8003.
  
  The problem was that if InnoDB failed to open a table within an attachable
  transaction, the server didn't call close_connection(). That led to a memory
  leak in that case.
  
  The quick fix is to fallback to InnoDB handlerton even if no tables were
  successfully opened.
------------------------------------------------------------
revno: 8744
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 14:48:37 +0530
message:
  Bug #19495721 REMOVE C-STYLE IB_LOGF FUNCTION, REPLACE USAGE WITH IB::INFO(),
  IB::WARN(), ETC
  
  Problem:
  
  There are now two ways of writing log messages - using the C-style function
  ib_logf() and using the C++ style classes ib::info, ib::warn, ib::error and
  ib::fatal.  
  
  Solution:
  
  Remove the C-style ib_logf(), so that everybody will start using the C++
  style classes.
  
  Note:
  
  This is part 2 of the patch.  The first part was done by rb#6442.  
  There will be subsequent patches to complete this bug.
  
  rb#6471 approved by Vasil.
------------------------------------------------------------
revno: 8743
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: wl7315-trunk
timestamp: Fri 2014-08-29 11:04:06 +0200
message:
  WL#7315 Optimizer cost model: main memory management of cost constants
  
  This worklog implements the following main components:
  
  1. A main memory representation of the "set of cost constants". This contains
     the cost constants read from the cost constant tables stored in the mysql
     database (see WL#7276). For cost constants that do not have entries in the
     cost constant tables, default values defined in the source code will be used.
        
  2. A cache for "cost constant sets". This is responsible for creating the set
     of cost constant, giving new sessions access to the latest versions of the
     cost constants, and for re-reading the cost constant tables in the case
     where these have been updated.
------------------------------------------------------------
revno: 8742 [merge]
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Fri 2014-08-29 16:39:57 +0800
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.340.13
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-5.6-bugfix1
    timestamp: Fri 2014-08-29 16:31:48 +0800
    message:
      BUG#18503734 - SEGV IN RBT_SEARCH(), FTS
      
      1. Check whether cached_stopword is NULL in fts_tokenizer_word_get,
      so that we can avoid segement fault. 2. Load stopwords if needed in
      fts_add_doc_by_id. Note: The root cause why cached_stopword is not
      initialized is still unfounded.
      
      rb#6462 approved by Jimmy.
------------------------------------------------------------
revno: 8741
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 11:20:34 +0300
message:
  Revert most of Bug #19458208 REMOVE REDUNDANCY FROM INNODB TEST FILE NAMES
  to avoid problems right before release clone-off.
------------------------------------------------------------
revno: 8740
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 11:02:14 +0300
message:
  Bug#19458208 73623: REMOVE REDUNDANCY FROM INNODB TEST FILE NAMES
  
  Most files were renamed by the following Bourne Again shell (bash)
  script that removes the test suite name prefix and replaces - with _
  in test names:
  
  for s in mysql-test/suite/*innodb*
  do
  	S="${s##*/}"
  	for t in "$s/t/$S"[-_]*.test
  	do
  		ST="${t%/t/*}"
  		St="${t#*/t/}"
  		TT="${t#*/t/$S?}"
  		T="${St%.test}"
  		U="$(echo "${TT%.test}"|tr - _)"
  		bzr mv $ST/t/{$T,$U}-master.opt
  		bzr mv $ST/r/{$T,$U}.result
  		bzr mv $ST/t/{$T,$U}.test
  	done
  done
  
  Some further cleanup was performed manually. Some suites contained
  tests that started with innodb_, not with the full suite name.
  
  Approved by Sunny Bains
------------------------------------------------------------
revno: 8739
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 16:04:23 +0900
message:
  Bug#19420253 : ASSERT \'S_LATCH == RW_LOCK_OWN(&INDEX->LOCK, RW_LOCK_S)\' AT ROW0INS.CC
  
  The assertion code was insufficient.
  Adjusted also for SX lock, not only for S lock.
  
  Approved by Jimmy by IM
------------------------------------------------------------
revno: 8738
committer: HaixLi < haixiang.li@oracle.com >
branch nick: mysql-trunk-wl6263-c6
timestamp: Fri 2014-08-29 08:42:27 +0200
message:
  WL#6263 Remove limited YEAR(2) support
  
  Description:
  ------------
  This WL will remove YEAR(2) cloumn. It supports:
  F1: Remove some old codes about YEAR(2).
  F2: Must not create YEAR(x!=4) column.
      Only support YEAR/YEAR(4) column.
  F3: Must not ALTER TABLE a column to YEAR(x!=4) column.
  F4: Run DML statements on existent legacy YEAR(X!=4) column will be given a
      error.
  F5: Run SHOW CREATE TABLE statement on existent legacy YEAR(X!=4) column will
      be given a warning.
  F6: Must not replicate CREATE YEAR(x!=4) column from 5.5 or 5.1 to this WL.  
------------------------------------------------------------
revno: 8737
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 09:18:04 +0300
message:
  Work around for Bug#19442959 DO NOT CALL EXIT() FROM INNODB
  
  The test innodb.log_file_name is creating an empty *.ibd file,
  prompting InnoDB to call exit() from an I/O handler thread.
  This will sometimes cause other threads to catch ut_ad(!m_freed)
  in ib0mutex.h in a mutex operation.
  
  We work around the problem by introducing a debug flag that will be set
  before calling exit().
  
  Also, remove some unreachable code and an obsolete comment.
  Approved by Sunny Bains on IM.
------------------------------------------------------------
revno: 8736
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 08:20:00 +0300
message:
  Follow-up to Bug#19514950 CLEAN UP TESTS THAT KILL OR RESTART THE SERVER
  Clean up an include file and re-record the results.
------------------------------------------------------------
revno: 8735
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 13:05:19 +0800
message:
  Fix these 2 test case failures on PB2.
------------------------------------------------------------
revno: 8734
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 10:15:45 +0800
message:
  Commit the missing test case result file for WL#6835.
------------------------------------------------------------
revno: 8733
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-08-29 09:31:40 +0800
message:
  Revert the patch of bug#19450143 because there is a regression after it.
------------------------------------------------------------
revno: 8732 [merge]
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Fri 2014-08-29 10:24:18 +1000
message:
  WL#6835 - InnoDB: GCS Replication: Deterministic Deadlock Handling (High Prio Transactions in InnoDB)
  
  Introduce transaction priority. Transactions with a higher priority cannot
  be rolled back by transactions with a lower priority. A higher priority
  transaction will jump the lock wait queue and grab the record lock instead
  of waiting.
  
  This code is not currently visible to the users. However, there are debug
  tests that can exercise the code. It will probably require some additional
  work once it is used by GCS.
  
  rb#6036 Approved by Jimmy Yang.
    ------------------------------------------------------------
    revno: 6753.1.131 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-29 05:07:13 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.130 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-29 03:08:48 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.129
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-28 21:24:31 +1000
    message:
      WL#6835 - Add a trx_t::onwer_id, it is the thread of the thread that started
      the transaction.
    ------------------------------------------------------------
    revno: 6753.1.128
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-28 14:43:34 +1000
    message:
      WL#6835 - Set the trx_t::killed_by thread ID atomically. Add an invariant that
      XA_PREPARED transactions cannot be rolled asynchronously.
    ------------------------------------------------------------
    revno: 6753.1.127 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-28 11:19:17 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.126
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-28 10:56:49 +1000
    message:
      WL#6835 - Move the trx_t::has_search_latch state changes to separate function.
      Only the thread that owns the AHI latch should unlock the latch in TrxInInnoDB.
      Some minor code formatting.
    ------------------------------------------------------------
    revno: 6753.1.125
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-27 15:42:37 +1000
    message:
      WL#6835 - Fix typo.
    ------------------------------------------------------------
    revno: 6753.1.124
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-27 14:33:06 +1000
    message:
      WL#6835 - Release the btr search latch in the destructor of TrxInInnoDB.
    ------------------------------------------------------------
    revno: 6753.1.123
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-27 14:31:18 +1000
    message:
      WL#6835 - Fix build issue on Windows. Inherited from trunk merge.
    ------------------------------------------------------------
    revno: 6753.1.122
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-27 10:39:02 +1000
    message:
      WL#6835 - Make the test deterministic.
    ------------------------------------------------------------
    revno: 6753.1.121 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-27 09:42:38 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.120 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 21:32:32 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.119
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 21:15:44 +1000
    message:
      WL#6835 - Release the btr search latch when entering InnoDB.
    ------------------------------------------------------------
    revno: 6753.1.118
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 20:58:12 +1000
    message:
      WL#6835 - Add a result file. Reset trx_t::state at innobase_commit() to
      TRX_STATE_NOT_STARTED if it was in state TRX_STATE_FORCED_ABORT.
    ------------------------------------------------------------
    revno: 6753.1.117
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 17:00:05 +1000
    message:
      WL#6835 - Fix whitespace.
    ------------------------------------------------------------
    revno: 6753.1.116
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 14:44:52 +1000
    message:
      WL#6835 - XA prepared transactions should not be rolled back. Re-enable the
      debug sync point to test rollback of successul update statements.
      
      Note: The commit should return an error but does not. The reap correctly
      returns the error. However, when we do a COMMIT the control never reaches
      InnoDB. The server returns success when it shouldn't.
    ------------------------------------------------------------
    revno: 6753.1.115 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 10:22:58 +1000
    message:
      WL#6835 - Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 6753.1.114 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-26 09:36:44 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.113 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-25 19:38:06 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.112
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-25 18:23:41 +1000
    message:
      WL#6835 - Handle the case where the high priority transactions owns the
      dict rw-lock in S mode during the kill blocking transaction phase.
    ------------------------------------------------------------
    revno: 6753.1.111
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-25 18:22:04 +1000
    message:
      WL#6835 - Skip transactions that are waiting in the record lock queue. We have
      to re-order the record lock queue if we are blocked by read-only transaction.
    ------------------------------------------------------------
    revno: 6753.1.110 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 16:02:07 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.109
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 15:00:04 +1000
    message:
      WL#6835 - Add assertions to check for potential race.
    ------------------------------------------------------------
    revno: 6753.1.108
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 13:54:53 +1000
    message:
      WL#6835 - Tidy up the rules, add an assertion.
    ------------------------------------------------------------
    revno: 6753.1.107
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 12:24:01 +1000
    message:
      WL#6835 - If the blocking transaction is a read-only transaction then we
      don't jump the queue but do a normal wait.
    ------------------------------------------------------------
    revno: 6753.1.106
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 12:19:55 +1000
    message:
      WL#6835 - Skip wait_for->trx transactions before acquiring the trx_t::mutex.
    ------------------------------------------------------------
    revno: 6753.1.105
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 12:02:13 +1000
    message:
      WL#6835 - Widen the trx mutex critical section. Not requird for correctness
      per se. We can narrow it down later.
    ------------------------------------------------------------
    revno: 6753.1.104
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 11:43:01 +1000
    message:
      WL#6835 - Tighten up the check for blocking read-only transaction.
    ------------------------------------------------------------
    revno: 6753.1.103
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 10:03:57 +1000
    message:
      WL#6835 - The assertion is not an invariant.
    ------------------------------------------------------------
    revno: 6753.1.102 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-22 09:43:02 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.101
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-21 20:13:16 +1000
    message:
      WL#6835 - Remove an assertion that is not an invariant.
    ------------------------------------------------------------
    revno: 6753.1.100
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-21 18:40:17 +1000
    message:
      WL#6835 - Fix the table row lock ref count logic.
    ------------------------------------------------------------
    revno: 6753.1.99 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-21 10:42:05 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.98
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-21 10:12:10 +1000
    message:
      WL#6835 - Avoid duplicates in the hit list. Improve the check for the head
      of the row lock queue. Increment the trx_t::version on commit too.
    ------------------------------------------------------------
    revno: 6753.1.97
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-20 21:25:00 +1000
    message:
      WL#6835 - Skip NL-AC-RO transactions. Add some assertions.
    ------------------------------------------------------------
    revno: 6753.1.96
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-20 19:52:52 +1000
    message:
      WL#6835 - Introduce trx_t::version. It is local to the trx_t instance. It is
      incrmented when the transaction is started. Before killing a blocking
      transaction in the trx_t::hit_list we check if it hasn't been reused to
      start a new transaction. We kill the transaction only if the version numbers
      match.
    ------------------------------------------------------------
    revno: 6753.1.95
    committer: Vinay Fisrekar <vinay.fisrekar@oracle.com>
    branch nick: mysql-trunk-wl6835
    timestamp: Wed 2014-08-20 10:29:55 +0530
    message:
      Fix testcase issue.
    ------------------------------------------------------------
    revno: 6753.1.94 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-20 13:30:51 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.93
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-20 13:08:27 +1000
    message:
      WL#6835 - There main fixes:
      
      1. Fix the test case, DEADLOCK is a full rollback. The COMMIT following the
         UPDATE is superfluous.
      
      2. Remove the blocking from the trx_t::hit_list if it is rolled back during
         the record lock enqueue phase
      
      3. When the trx_t::state == TRX_STATE_FORCED_ROLLBACK, return DB_FORCED_ABORT
         on COMMIT/ROLLBACK requests.
    ------------------------------------------------------------
    revno: 6753.1.92
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-19 15:07:24 +1000
    message:
      WL#6835 - The requesting transaction could have an existing weaker lock in
      the record queue.
    ------------------------------------------------------------
    revno: 6753.1.91
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-19 13:40:58 +1000
    message:
      WL#6835 - Add a point of no return check. Once we know that a transaction
      will commit or rollback we don't mark it for asynchronous rollback.
    ------------------------------------------------------------
    revno: 6753.1.90
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-19 12:41:17 +1000
    message:
      WL#6835 - Minor test clean up. Reset the tx priority on "real" transactions.
    ------------------------------------------------------------
    revno: 6753.1.89 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-19 09:21:29 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.88 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-18 13:06:42 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.87
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-18 12:07:18 +1000
    message:
      WL#6835 - Rename tests
    ------------------------------------------------------------
    revno: 6753.1.86
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Sun 2014-08-17 10:56:18 +1000
    message:
      WL#6835 - Don't rollback internal transactions that are blocking the high
      priority transaction.
    ------------------------------------------------------------
    revno: 6753.1.85
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-15 13:30:33 +1000
    message:
      WL#6835 - Instantiate the trx_t data members before calling trx_init().
    ------------------------------------------------------------
    revno: 6753.1.84 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-15 13:28:31 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.83
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-15 12:58:04 +1000
    message:
      WL#6835 - Check abort status in ha_innobase::unlock_row(). Add some comments,
      fix some debug assertions.
    ------------------------------------------------------------
    revno: 6753.1.82
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-14 18:44:13 +1000
    message:
      WL#6835 - Reset for implicit commit/rollback too.
    ------------------------------------------------------------
    revno: 6753.1.81
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-14 17:43:08 +1000
    message:
      WL#6835 - Reset the session transaction priority flag at commit/rollback.
    ------------------------------------------------------------
    revno: 6753.1.80
    committer: Vinay Fisrekar <vinay.fisrekar@oracle.com>
    branch nick: mysql-trunk-wl6835
    timestamp: Wed 2014-08-13 14:08:49 +0530
    message:
      Renamving test files according to naming convention.
    ------------------------------------------------------------
    revno: 6753.1.79 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-13 17:00:29 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.78
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-13 16:58:28 +1000
    message:
      WL#6835 - Add comments.
    ------------------------------------------------------------
    revno: 6753.1.77
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-08-13 16:36:57 +1000
    message:
      WL#6835 - Fix some assertions. The important change is that we can't guarantee
      that we will be granted a lock if we rollback the blocking transaction holding
      an S lock. Other transactions could be holding the S lock and they could be
      waiting on some other lock. We need to rollback all active transactions that
      have been granted that lock.
    ------------------------------------------------------------
    revno: 6753.1.76 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-11 07:14:28 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.75
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Sat 2014-08-09 10:48:24 +1000
    message:
      WL#6835 - If a transaction is already marked for termination skip it if
      it is blocking another high priority transaction.
    ------------------------------------------------------------
    revno: 6753.1.74
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Sat 2014-08-09 10:47:35 +1000
    message:
      WL#6835 - Don't skip transactions holding S locks ahead in the queue if they
      are waiting on some other lock. The need to be terminated too.
    ------------------------------------------------------------
    revno: 6753.1.73
    committer: Vinay Fisrekar <vinay.fisrekar@oracle.com>
    branch nick: mysql-trunk-wl6835
    timestamp: Fri 2014-08-08 17:56:16 +0530
    message:
      Add tests w.r.t QA plan.
    ------------------------------------------------------------
    revno: 6753.1.72 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-08 17:57:15 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.71 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-07 16:04:14 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.70
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-08-07 16:02:34 +1000
    message:
      WL#6835 - Background transactions that collect statistics don't have a MySQL
      session associated with them. They can block user transactions because they
      sample user tables. This case was not handled in the arbitration logic.
    ------------------------------------------------------------
    revno: 6753.1.69 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-05 14:14:04 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.68
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-08-05 14:13:31 +1000
    message:
      WL#6835 - Refacor the rec lock check
    ------------------------------------------------------------
    revno: 6753.1.67 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-08-04 07:32:24 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.66 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-08-01 08:39:07 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.65
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-07-31 13:56:57 +1000
    message:
      WL#635 - Non-functional changes
    ------------------------------------------------------------
    revno: 6753.1.64
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-07-31 11:57:33 +1000
    message:
      WL#6835 - Remove dead code.
    ------------------------------------------------------------
    revno: 6753.1.63 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-07-30 19:08:54 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.62
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-29 19:37:06 +1000
    message:
      WL#6835 - Remove the update_row_low() method. Not required anymore.
    ------------------------------------------------------------
    revno: 6753.1.61
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-29 19:26:46 +1000
    message:
      WL#6835 - Do not rollback RO transactions that are active.
    ------------------------------------------------------------
    revno: 6753.1.60
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-29 11:36:29 +1000
    message:
      WL#6835 - Non-functional changes related to previous push.
    ------------------------------------------------------------
    revno: 6753.1.59
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-29 10:54:32 +1000
    message:
      WL#6835 - Non-functional changes.
    ------------------------------------------------------------
    revno: 6753.1.58 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-29 07:29:58 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.57
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-07-28 07:25:47 +1000
    message:
      WL#6835 - Fix whitespace
    ------------------------------------------------------------
    revno: 6753.1.56 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-07-28 07:11:40 +1000
    message:
      WL#6835 -  Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.55 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-07-25 18:52:01 +1000
    message:
      WL#6835 - Test updates from Luis.
        ------------------------------------------------------------
        revno: 6753.3.1
        committer: Luis Soares <luis.soares@oracle.com>
        branch nick: mysql-trunk-wl6835
        timestamp: Tue 2014-07-22 01:47:38 +0100
        message:
          WL#6835
          
          Changes to test cases to be able to cover the functionality
          introduced in WL#6835, by both types of threads, slave
          threads and user threads.
    ------------------------------------------------------------
    revno: 6753.1.54 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-07-25 10:48:26 +1000
    message:
      WL#6835 - Merge from mysql-ytrunk
    ------------------------------------------------------------
    revno: 6753.1.53
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-22 16:08:03 +1000
    message:
      WL#6835 - Non-functional changes
    ------------------------------------------------------------
    revno: 6753.1.52
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-22 14:54:19 +1000
    message:
      WL#6835 - Non-functional change.
    ------------------------------------------------------------
    revno: 6753.1.51
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-22 14:50:50 +1000
    message:
      WL#6835 - Fix typo.
    ------------------------------------------------------------
    revno: 6753.1.50
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-22 14:46:57 +1000
    message:
      Fix a bug in the predicate lock. The predicate was not being set for the
      correct lock.
    ------------------------------------------------------------
    revno: 6753.1.49 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-22 07:35:12 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.48 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-22 07:34:12 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.47 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-07-16 19:47:02 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.46
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-07-02 10:55:08 +1000
    message:
      WL#6835 - Error code grouping. Add a proxy for ha_innobase::update_row(). We
      need to release the reference to the transaction and then wait.
    ------------------------------------------------------------
    revno: 6753.1.45
    committer: Nuno Carvalho <nuno.carvalho@oracle.com>
    branch nick: mysql-trunk-wl6835
    timestamp: Tue 2014-07-01 14:33:02 +0100
    message:
      WL#6835 - Fixed race condition on rpl_wl6835_slave.test
    ------------------------------------------------------------
    revno: 6753.1.44
    committer: Nuno Carvalho <nuno.carvalho@oracle.com>
    branch nick: mysql-trunk-wl6835
    timestamp: Tue 2014-07-01 10:09:48 +0100
    message:
      WL#6835 - Fixed tests outcome description.
    ------------------------------------------------------------
    revno: 6753.1.43
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-01 18:12:06 +1000
    message:
      WL#6835 - Remove debugging instrumentation.
    ------------------------------------------------------------
    revno: 6753.1.42
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-01 17:18:13 +1000
    message:
      WL#6835 - Update test to match record lock jump queue semantics.
    ------------------------------------------------------------
    revno: 6753.1.41 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-07-01 15:10:22 +1000
    message:
      WL#6835 - Patch from Nuno Carvalho
        ------------------------------------------------------------
        revno: 6753.2.2
        committer: Nuno Carvalho <nuno.carvalho@oracle.com>
        branch nick: mysql-trunk-wl6835
        timestamp: Mon 2014-06-30 17:05:24 +0100
        message:
          WL#6835 - Improved tests.
        ------------------------------------------------------------
        revno: 6753.2.1
        committer: Nuno Carvalho <nuno.carvalho@oracle.com>
        branch nick: mysql-trunk-wl6835
        timestamp: Mon 2014-06-30 17:02:59 +0100
        message:
          WL#6835 - Change thd_tx_priority and tx_priority from bool to int.
    ------------------------------------------------------------
    revno: 6753.1.40 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-06-30 11:49:49 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.39
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-27 23:50:07 +1000
    message:
      WL#6835 - Add test from Luis.
    ------------------------------------------------------------
    revno: 6753.1.38 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-27 23:47:54 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.37
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-27 23:38:40 +1000
    message:
      WL#6835 - Remoe the requirement that some of the trx_t data members state had
      to be kept after commit/rollback. Introduce a new transaction state to capture
      the same information, TRX_STATE_FORCED_ROLLBACK.
      
      Remove direct access to trx->state where possible, replace with a predicate:
      trx_is_started().
    ------------------------------------------------------------
    revno: 6753.1.36 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-27 00:24:55 +1000
    message:
      WL#6835 - Push WIP code.
    ------------------------------------------------------------
    revno: 6753.1.35 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-06-23 22:28:58 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.34
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-06-23 22:19:40 +1000
    message:
      WL#6835 - Non-functional change. Rename ha_innobase member data variables
      to m_*.
    ------------------------------------------------------------
    revno: 6753.1.33
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-06-23 18:23:56 +1000
    message:
      WL#6835 - Add test result files.
    ------------------------------------------------------------
    revno: 6753.1.32 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-06-23 17:45:00 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.31
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-06-23 17:38:24 +1000
    message:
      WL#6835 - Rollback running transactions asynchronously. Add tests from Nuno.
      Convert some ha_innobase member variables to the new m_* convention.
    ------------------------------------------------------------
    revno: 6753.1.30
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-06-17 19:51:26 +1000
    message:
      WL#6835 - Remove unused query state
    ------------------------------------------------------------
    revno: 6753.1.29 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-06-17 17:53:35 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.28
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-13 17:30:22 +1000
    message:
      WL#6835 - Make que_thr_t #defines into enums.
    ------------------------------------------------------------
    revno: 6753.1.27
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-13 16:30:24 +1000
    message:
      WL#6835 - Change query thread #defines to enums.
    ------------------------------------------------------------
    revno: 6753.1.26 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Fri 2014-06-13 16:09:41 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.25
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 22:25:35 +1000
    message:
      WL#6835 - Add a callback to arbitrate between transactions. This callback
      will return the victim transaction or NULL if it can't decide and wants the
      storage engine to decide.
    ------------------------------------------------------------
    revno: 6753.1.24
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 18:55:45 +1000
    message:
      WL#6835 - Remove typename.
    ------------------------------------------------------------
    revno: 6753.1.23
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 18:51:09 +1000
    message:
      WL#6835 - Convert the trx.lock.table_locks from ib_vector to std::vector.
    ------------------------------------------------------------
    revno: 6753.1.22
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 18:16:31 +1000
    message:
      WL#6835 - Minor code clean up.
    ------------------------------------------------------------
    revno: 6753.1.21
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 18:02:51 +1000
    message:
      WL#6835 - Rename enqueue() to add_to_waitq().
    ------------------------------------------------------------
    revno: 6753.1.20
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 17:06:07 +1000
    message:
      WL#6835 - Code cleanup.
    ------------------------------------------------------------
    revno: 6753.1.19
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 13:32:45 +1000
    message:
      WL#6835 - Move the trx mutex exit/enter close to where it is required.
    ------------------------------------------------------------
    revno: 6753.1.18
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 13:07:41 +1000
    message:
      WL#6835 - Move the lock priority code to a separate method.
    ------------------------------------------------------------
    revno: 6753.1.17 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-06-12 10:27:35 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.16
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-06-11 23:42:51 +1000
    message:
      WL#6835 - Try and rollback an idle transaction from within InnoDB. This is an
      experimental change.
    ------------------------------------------------------------
    revno: 6753.1.15 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-06-10 18:01:10 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.14 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-06-10 17:00:15 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.13 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-06-04 12:18:40 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.12
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-06-04 10:51:23 +1000
    message:
      WL#6835 - Rewrite the record lock and lock wait enqueue code. Allow a high
      priority transaction to rollback a waiting transaction that is holding a
      lock that the higher priority transaction wants.
    ------------------------------------------------------------
    revno: 6753.1.11
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Wed 2014-05-28 17:36:51 +1000
    message:
      WL#6835 - initialise tx_no_rollback in THD and set to false by default.
    ------------------------------------------------------------
    revno: 6753.1.10
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-05-27 13:20:40 +1000
    message:
      WL#6835 - Tie the trx to the THD tx_no_rollback variable.
    ------------------------------------------------------------
    revno: 6753.1.9
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-05-27 12:53:14 +1000
    message:
      WL#6835 - Do not rollback high priority transaction during deadlock checking.
    ------------------------------------------------------------
    revno: 6753.1.8
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Tue 2014-05-27 12:25:07 +1000
    message:
      WL#6835 - Add support for "START TRANSACTION HIGH_PRIORITY;" for testing.
    ------------------------------------------------------------
    revno: 6753.1.7
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-05-26 18:02:47 +1000
    message:
      WL#6835 - Don't select a transaction for rollback during deadlock checking that
      has the "no rollback" flag set.
    ------------------------------------------------------------
    revno: 6753.1.6
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-05-26 17:44:21 +1000
    message:
      WL#6835 - trx_can_rollback() is inlined.
    ------------------------------------------------------------
    revno: 6753.1.5
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-05-26 17:14:27 +1000
    message:
      WL#6835 - Non-functional change, code cleanup.
    ------------------------------------------------------------
    revno: 6753.1.4
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-05-26 16:29:24 +1000
    message:
      WL#6835- Remove UNIV_*LIKELY() hints.
    ------------------------------------------------------------
    revno: 6753.1.3 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-05-26 10:41:46 +1000
    message:
      WL#6835 - Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6753.1.2
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Mon 2014-05-26 10:40:52 +1000
    message:
      WL#6835 - Add new function trx_can_rollback
    ------------------------------------------------------------
    revno: 6753.1.1 [merge]
    committer: Sunny Bains <Sunny.Bains@Oracle.Com>
    branch nick: wl6835
    timestamp: Thu 2014-05-22 12:15:01 +1000
    message:
      WL#6835 merge from mysql-trunk
------------------------------------------------------------
revno: 8731
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Fri 2014-08-29 05:04:19 +1000
message:
  Bug#19525395 TTASFUTEXMUTEX DOESN'T EXPECTS BUT DOES NOT REQUEST ALIGNED M_LOCK_WORD
  
  Fix alignment of the lock word. Suggested fix from Davi Arnaut.
------------------------------------------------------------
revno: 8730 [merge]
committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
branch nick: mysql-trunk-wl7793
timestamp: Thu 2014-08-28 18:08:38 +0100
message:
  WL#7793: Add native support for syslog on unixoid platforms
  
  mysqld did not have native support for traditional unix-style
  syslog; it instead used a shell-script wrapper that would pipe
  the daemon's output to the syslog.  This implied certain limitations:
  - it was impossible to see the syslog status at runtime
  - it was impossible to change the syslog status at runtime
  - the shell script could interfere with log rotation
  - the setup would often result in log lines containing two timestamps
  
  a) add native syslog support for unixoid OSs.
  
     The following can be set a start-up as well as viewed
     and set at runtime via system variables (requiring the
     SUPER privilege).
  
     --log_syslog[=0|1]
       on/off switch, defaults to off
  
     --log_syslog_include_pid[=0|1]
       Include mysqld's process ID on each syslog line?
       Defaults to on.  (UNIX only)
  
     --log_syslog_facility=local5
       choose the syslog "facility". Well-known facilities
       may be selected by name.
       Defaults to "daemon".  (UNIX only)
  
     --log_syslog-tag=<string>
       If empty and by default, an ident of "mysqld" is used.
       If e.g. tag "abcd" is supplied, "mysqld-abcd" will be used.
       (The hyphen will be used automatically, the user need not
       supply it.)
  
  b) refactoring: mysys/my_syslog.cc and sql/log.cc had
     some duplication of effort with regard to the
     Windows EventLog.  This patch attempts to eliminate
     the duplicates from log.cc, and merge the remaining
     useful code into my_syslog.cc so that other binaries
     may have a chance to benefit from it.
  
  c) Support features of a) on Windows where possible.
  
  d) Update mysqld_safe to use features of a) instead of
     trying to emulate them in the wrapper script where
     possible, with as little change in syntax as possible.
  
  journalctl -f SYSLOG_IDENTIFIER=mysqld-meow -o json|
  jq '.SYSLOG_IDENTIFIER + " " + .SYSLOG_FACILITY + ": " + .MESSAGE' 2>/dev/null
    ------------------------------------------------------------
    revno: 8443.1.4
    committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
    branch nick: 57-11762739
    timestamp: Thu 2014-08-28 06:19:06 +0100
    message:
      WL#7793: Add native support for syslog on unixoid platforms
      
      * clean up
    ------------------------------------------------------------
    revno: 8443.1.3
    committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
    branch nick: 57-11762739
    timestamp: Mon 2014-08-18 06:29:24 +0100
    message:
      Bug#11762739: MYSQLD GENERAL LOG OUTPUT TO SYSLOG
      WL#7793: Add native support for syslog on unixoid platforms
      
      mysqld did not have native support for traditional unix-style
      syslog; it instead used a shell-script wrapper that would pipe
      the daemon's output to the syslog.  This implied certain limitations:
      - it was impossible to see the syslog status at runtime
      - it was impossible to change the syslog status at runtime
      - the shell script could interfere with log rotation
      - the setup would often result in log lines containing two timestamps
      
      a) add native syslog support for unixoid OSs.
      
         The following can be set a start-up as well as viewed
         and set at runtime via system variables (requiring the
         SUPER privilege).
      
         --log_syslog[=0|1]
           on/off switch, defaults to off
      
         --log_syslog_include_pid[=0|1]
           Include mysqld's process ID on each syslog line?
           Defaults to on.  (UNIX only)
      
         --log_syslog_facility=local5
           choose the syslog "facility". Well-known facilities
           may be selected by name.
           Defaults to "daemon".  (UNIX only)
      
         --log_syslog-tag=<string>
           If empty and by default, an ident of "mysqld" is used.
           If e.g. tag "abcd" is supplied, "mysqld-abcd" will be used.
           (The hyphen will be used automatically, the user need not
           supply it.)
      
      b) refactoring: mysys/my_syslog.cc and sql/log.cc had
         some duplication of effort with regard to the
         Windows EventLog.  This patch attempts to eliminate
         the duplicates from log.cc, and merge the remaining
         useful code into my_syslog.cc so that other binaries
         may have a chance to benefit from it.
      
      c) Support features of a) on Windows where possible.
      
      d) Update mysqld_safe to use features of a) instead of
         trying to emulate them in the wrapper script where
         possible, with as little change in syntax as possible.
      
      
      Now incorporates following additions after review:
      
      - mysql-test-run.pl disables syslog/EventLog
      - mysqld--help-win.result updated to reflect the above
      - my_syslog.c: EventID temporarily wasn't used after refactoring
      - mysqld.cc: syslog activated earlier in start-up, for more information
      
      Syslog query on systemd platforms:
      journalctl -f SYSLOG_IDENTIFIER=mysqld-mytag -o json|jq '.SYSLOG_IDENTIFIER + " " + .SYSLOG_FACILITY + ": " + .MESSAGE' 2>/dev/null
    ------------------------------------------------------------
    revno: 8443.1.2
    committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
    branch nick: 57-11762739
    timestamp: Fri 2014-08-08 04:15:47 +0100
    message:
      updating (c) year
    ------------------------------------------------------------
    revno: 8443.1.1
    committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
    branch nick: 57-11762739
    timestamp: Fri 2014-08-08 04:12:32 +0100
    message:
      Bug#11762739: MYSQLD GENERAL LOG OUTPUT TO SYSLOG
      WL#7793: Add native support for syslog on unixoid platforms
      
      mysqld did not have native support for traditional unix-style
      syslog; it instead used a shell-script wrapper that would pipe
      the daemon's output to the syslog.  This implied certain limitations:
      - it was impossible to see the syslog status at runtime
      - it was impossible to change the syslog status at runtime
      - the shell script could interfere with log rotation
      - the setup would often result in log lines containing two timestamps
      
      a) add native syslog support for unixoid OSs.
      
         The following can be set a start-up as well as viewed
         and set at runtime via system variables (requiring the
         SUPER privilege).
      
         --log_syslog[=0|1]
           on/off switch, defaults to off
      
         --log_syslog_include_pid[=0|1]
           Include mysqld's process ID on each syslog line?
           Defaults to on.  (UNIX only)
      
         --log_syslog_facility=local5
           choose the syslog "facility". Well-known facilities
           may be selected by name.
           Defaults to "daemon".  (UNIX only)
      
         --log_syslog-tag=<string>
           If empty and by default, an ident of "mysqld" is used.
           If e.g. tag "abcd" is supplied, "mysqld-abcd" will be used.
           (The hyphen will be used automatically, the user need not
           supply it.)
      
      b) refactoring: mysys/my_syslog.cc and sql/log.cc had
         some duplication of effort with regard to the
         Windows EventLog.  This patch attempts to eliminate
         the duplicates from log.cc, and merge the remaining
         useful code into my_syslog.cc so that other binaries
         may have a chance to benefit from it.
      
      c) Support features of a) on Windows where possible.
      
      d) Update mysqld_safe to use features of a) instead of
         trying to emulate them in the wrapper script where
         possible, with as little change in syntax as possible.
------------------------------------------------------------
revno: 8729
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 17:18:58 +0200
message:
  Follow-up to WL#7688 - add Joro's change to mysql-test-run.pl to make MTR
  tests run on Windows.
------------------------------------------------------------
revno: 8728
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 09:30:22 -0500
message:
  Fix build break on 32-bit Windows build.
------------------------------------------------------------
revno: 8727
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Thu 2014-08-28 15:53:20 +0200
message:
  Bug #19163915	INNODB: DUPLICATE RECORDS COULD EXIST WHEN SKIPPING
  			SORT FOR CLUSTER INDEX
  
  Post-push fix: Move assert + explicitly initialize mtuple_heap.
  
  Patch approved by marko over IM.
------------------------------------------------------------
revno: 8726
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 17:46:41 +0530
message:
  Bug #19163915	INNODB: DUPLICATE RECORDS COULD EXIST WHEN SKIPPING
  			SORT FOR CLUSTER INDEX
  Problem:
  Duplicate record could exist when there is no change in primary key
  during alter table operation. Duplicate record can exist between the buffers. Last
  mtuple in previous buffer can same as the first mtuple in the current buffer.
  
  Solution:
     Maintain and store the last mtuple in the buffer. Compare the last
  mtuple of the previous buffer with first mtuple of the current buffer. If both
  are equal then duplicate exist.
  
  	rb6194 approved by Marko, Shaohua
------------------------------------------------------------
revno: 8725
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: 32bit-fix
timestamp: Thu 2014-08-28 13:56:14 +0200
message:
  WL#7339 Use improved records per key estimate interface in optimizer
    
  Follow-up patch: The group_min_max test failed when running on 32 bit
  platforms in due to off by one "rows" estimates. This is caused by
  different results from doing floating point operations where we 
  previously used integer values for rec_per_key variables in the
  cost model for loose index scan. Adjusted the test to accept
  both values as the expected "rows" estimate.
------------------------------------------------------------
revno: 8724 [merge]
committer: vamsikrishna.bhagi@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 17:03:41 +0530
message:
  WL#7717 Expose the SSL library encryption functions to SQL
  
  Bazaar merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.340.12
    committer: vamsikrishna.bhagi@oracle.com
    branch nick: mysql-5.6
    timestamp: Thu 2014-08-28 16:57:34 +0530
    message:
      WL#7717 Expose the SSL library encryption functions to SQL
      
      Fixed the valgrind issues.
------------------------------------------------------------
revno: 8723
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 15:49:59 +0530
message:
  - Bug#19520482: INNODB_WL6501_CRASH_2/7 ARE FAILING ON DAILY-TRUNK WITH 8K
    DEBUG SIZE
  
    On truncate, if index is dropped and root-page is not re-allocated to other
    object before server crashes then page would be in free state.
  
    For checking lsn we need to pin root-page. To cover all kind of pages
    (including free) we should use BUF_GET_POSSIBLY_FREED vs BUF_GET.
  
    Approved by: Jimmy (rb#6495)
------------------------------------------------------------
revno: 8722
committer: Harin Vadodaria <harin.vadodaria@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 15:26:59 +0530
message:
  Bug#19521132 : SSL CERTS SHOULD BE AUTO GENERATED AFTER
                 SSL_START AND ONLY IF OPT_USE_SSL IS 1
  
  Description : Call to do_auto_cert_generation() is made
                before the call to ssl_start() and without
                checking value of opt_use_ssl. This may
                create issues while generating certificates
                and/or result into set of certificates which
                will never be used.
  
  Solution : Moved call to do_auto_cert_generation() after
             the call to ssl_start() and check opt_use_ssl
             status.
------------------------------------------------------------
revno: 8721 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 15:16:56 +0530
message:
  Bug#19145712 USER AFTER FREE / DOUBLE FREE ISSUE
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.340.11 [merge]
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-08-28 15:06:53 +0530
    message:
      Bug#19145712 USER AFTER FREE / DOUBLE FREE ISSUE
      
      Merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.598.36
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-08-28 14:29:54 +0530
        message:
                Bug#19145712 USER AFTER FREE / DOUBLE FREE ISSUE
                
                Problem: A corrupted header length in FORMAT_DESCRIPTION_LOG_EVENT
                can cause server to crash.
                Analysis: FORMAT_DESCRIPTION_EVENT will be considered invalid if
                header len is too small (i.e. below OLD_HEADER_LEN).
                
                Format_description_log_event:: Format_description_log_event(...)
                {
                  ...
                  if ((common_header_len=buf[ST_COMMON_HEADER_LEN_OFFSET]) < OLD_HEADER_LEN)
                    DBUG_VOID_RETURN; /* sanity check */
                  ...
                  post_header_len= my_memdup(...)
                }
                
                In that case Format_description_log_event constructor will return early,
                without allocating any memory for post_header_len. Thence this variable is
                left uninitialized and making server to crash when server is trying
                to free the uninitialized value.
                
                Fix: When Format_description_log_event constructor returns early, assign
                NULL to post_header_len.
------------------------------------------------------------
revno: 8720
committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 11:06:57 +0200
message:
  WL7688 Deprecate mysql_install_db and create a new program in C++
  
  The new C++ based mysql_install_db.cc will remove  the dependency on perl
  when the MySQL server is deploying.
------------------------------------------------------------
revno: 8719
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk-merge
timestamp: Thu 2014-08-28 10:45:45 +0200
message:
  WL#7444 GeoJson support for GIS
  
  This WL adds functions for parsing and generating GeoJSON documents into GIS data types:
    <geometry> = ST_GEOMFROMGEOJSON(<string>[, <options>[, <srid>]])
    <string> = ST_ASGEOJSON(<geometry>[, <maxdecimaldigits>[, <options>]])
  
  The functions uses the rapidjson library to both parse and generate JSON documents.
  Implementation limitations:
    - Only 2D geometries are supported.
    - The Feature and FeatureCollection objects are not supported, except that geometry objects are extracted from them.
    - The CRS support is limited to values that identify an SRID.
------------------------------------------------------------
revno: 8718
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk-merge
timestamp: Thu 2014-08-28 10:44:15 +0200
message:
  Modified parsing of DOUBLE values in rapidjson library to correctly parse numbers like "100.00000".
------------------------------------------------------------
revno: 8717
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk-merge
timestamp: Thu 2014-08-28 10:43:29 +0200
message:
  Changed print-method for DOUBLE in rapidjson library to behave like String::set_real().
------------------------------------------------------------
revno: 8716
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk-merge
timestamp: Thu 2014-08-28 10:42:35 +0200
message:
  Fix compiler errors and warning in rapidjson library:
    - Remove the use of UINT64_C and INT64_C (compiler error)
    - Change "-2147483648" to "-2147483648LL" (compiler warning)
------------------------------------------------------------
revno: 8715
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk-merge
timestamp: Thu 2014-08-28 10:41:41 +0200
message:
  Added the rapidjson library (clean, unmodified version).
------------------------------------------------------------
revno: 8714 [merge]
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk
timestamp: Thu 2014-08-28 12:22:39 +0400
message:
  Merge from 5.6.
    ------------------------------------------------------------
    revno: 3902.340.10 [merge]
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-08-28 12:20:22 +0400
    message:
      Merge from 5.5.
        ------------------------------------------------------------
        revno: 2875.598.35
        committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-08-28 12:16:53 +0400
        message:
          Update copyright text.
------------------------------------------------------------
revno: 8713
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-cleanup
timestamp: Wed 2014-08-20 09:20:57 +0200
message:
  Fixed build warnings
------------------------------------------------------------
revno: 8712
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 10:18:11 +0300
message:
  Bug#19514950 CLEAN UP TESTS THAT KILL OR RESTART THE SERVER
  Bug#19505636 INNODB.LOG_FILE_SIZE FAILS WITH WRONG ERRNO 2006
  
  There are a number of tests that are killing or restarting the server.
  We should try to use common infrastructure for these.
  
  SET DEBUG='d,crash_commit_before';
  should not be used for just killing the server.
  A normal kill will run on a non-debug binary as well.
  
  Valgrind should not be disabled when the server
  is being killed by the test script.
  
  There still seems to be a problem with DBUG_SUICIDE and Valgrind,
  so we will have to disable Valgrind when the server is killed by
  test instrumentation.
  
  We introduce some helper files:
  
  include/kill_mysqld.inc
  include/kill_and_restart_mysqld.inc
  
  We introduce $restart_parameters for the above, and for the existing
  
  include/start_mysqld.inc
  include/restart_mysqld.inc
  
  "Homebrew" restart has been replaced with the helpers as much as possible.
  
  Most use of shutdown_server has been replaced with use of one of the
  include files.
  
  rb#6485 approved by Satya Bodapati
------------------------------------------------------------
revno: 8711
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Thu 2014-08-28 08:45:28 +0200
message:
  Bug#19499701 GET RID OF DYNAMIC_ARRAY IN RPL_RLI_PDB
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8710
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 13:14:55 +0800
message:
  Fix the sporadic failure on PB2 of innodb_gis_point_basic, which is
  introduced by WL#6942.
  
  Approved by Jimmy over IM.
------------------------------------------------------------
revno: 8709
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 13:44:16 +0900
message:
  Bug#17554489 : UNNECESSARY OVERHEAD FROM PERSISTENT ADAPTIVE HASH INDEX LATCHES
  
  If INNODB_RW_LOCKS_USE_ATOMICS is enabled, rw_lock implementation is fast enough.
  So no need to keep btr_search_latch also when over 10000 times AHI searches per 1 transaction,
  because just might block the other AHI updates.
  
  Approved by Sunny in rb#6257
------------------------------------------------------------
revno: 8708
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 13:43:15 +0900
message:
  Some POWER specific optimizations
  
  Bug#18842925 : SET THREAD PRIORITY IN INNODB MUTEX SPINLOOP
  Like "pause" instruction for hyper-threading at Intel CPUs,
  POWER has special instructions only for hinting priority of hardware-threads.
  
  Bug#18814859 : "CACHE_LINE_SIZE IN INNODB SHOULD BE 128 ON POWER")
  Data arrangement is be better to be conscious about data-cache unit size of CPU, for more CPU cache consistency.
  Currently it is fixed to 64 bytes in InnoDB, but 128 bytes is better for POWER.
  
  Approved by Sunny in rb#6256
------------------------------------------------------------
revno: 8707
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-28 12:28:05 +0800
message:
  Fix Bug #19460824 - GIS TABLE, INSERT, ASSERT !MEMCMP(PAGE, TEMP_PAGE,
  PAGE_HEADER) 
  
  rb#6464 approved by Sunny Bains
------------------------------------------------------------
revno: 8706
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt3
timestamp: Wed 2014-08-27 20:01:35 +0200
message:
  n test changes necessary to support WL#7638
------------------------------------------------------------
revno: 8705 [merge]
committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-27 16:19:04 +0200
message:
  Null merging the changes from 5.5
    ------------------------------------------------------------
    revno: 3902.340.9 [merge]
    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-08-27 16:12:24 +0200
    message:
      Null merging changes from 5.5
        ------------------------------------------------------------
        revno: 2875.598.34
        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-08-27 16:06:50 +0200
        message:
          Changing the copyright headers a bit to be complaint with the standards
------------------------------------------------------------
revno: 8704
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: wl7276-trunk
timestamp: Wed 2014-08-27 16:07:29 +0200
message:
  WL#7276 Configuration data base for Optimizer Cost Model
        
  This worklogs adds two new tables to the "mysql" database for storing
  "cost constants" that will be used by the optimizer cost model. The
  two new tables are:
        
  - server_cost: This stores cost constants for server cost model.
  - engine_cost: For storing cost constants for the cost model
                 for storage engines.
  
  The tables are created by the mysql_system_tables.sql script that
  is run when bootstrapping and upgrading the MySQL server.
------------------------------------------------------------
revno: 8703
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-08-27 19:14:06 +0530
message:
  - Bug #16479309: FUNCTION LOCK_NUMBER_OF_ROWS_LOCKED MAY BE INEFFICIENT
  
    Locks held by a transaction are tracked using a bit-vector.
    For reporting, bit-vector is traversed to calculate number of locks held.
    This could be optimized by using a simple counter for tracking
  
    Approved: Sunny (rb#6478)
------------------------------------------------------------
revno: 8702 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Tue 2014-08-19 15:18:04 +0200
message:
  WL#7802 PERFORMANCE SCHEMA, BATCH TABLE IO
  
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 8146.1.31 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-19 15:15:37 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.30 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-19 14:34:20 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.29 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-08-18 18:45:47 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.28 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Sun 2014-08-17 18:20:23 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.27
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Thu 2014-08-21 17:01:06 +0200
    message:
      WL#7802 PERFORMANCE SCHEMA, BATCH TABLE IO
      
      In batch mode, delay the call to PSI_TABLE_CALL(start_table_io_wait)()
      to the first actual table io call.
      This avoids reporting table io with 0 rows, and uses the proper index.
    ------------------------------------------------------------
    revno: 8146.1.26 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Thu 2014-08-21 15:28:15 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.25 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-19 18:15:41 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.24
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-19 17:49:50 +0200
    message:
      Implemented code review comments
    ------------------------------------------------------------
    revno: 8146.1.23 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-19 17:32:57 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.22 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-08-18 15:21:11 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.21 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-08-11 18:12:09 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.20 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-08-11 00:10:19 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.19 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Fri 2014-08-08 22:09:57 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.18
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Fri 2014-08-08 21:42:39 +0200
    message:
      Fixed whitespace
      Fixed diff compared to trunk
      Improved test cases
    ------------------------------------------------------------
    revno: 8146.1.17
    committer: HaixLi < haixiang.li@oracle.com >
    branch nick: mysql-trunk-wl7802-0725
    timestamp: Fri 2014-08-08 05:00:53 -0400
    message:
      WL#7802 PERFORMANCE SCHEMA, BATCH TABLE IO
      
      Description:
      ------------
      This WL will reduce number of calls for PFS when MySQL do a single table scan 
      or join operation.  
      We set batch mode for PFS when do join operation in executor to implement
      reduce number of calls for PFS.
      This WL does not change any functionality of join operation.
      
      Notice:
      This version covers:
      1 Simplify the code by Roy's suggestion.
      
      Remaining problem:
      1 A bug that make the server crash.
        It occurs in Protocol::end_statement()->DBUG_ASSERT(0). For example:
          ./mtr i_main.bka
        It need to be resolved in next version. 
        Analyzing...
    ------------------------------------------------------------
    revno: 8146.1.16 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Wed 2014-08-06 17:54:41 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.15
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-05 18:18:21 +0200
    message:
      WL#7802 PERFORMANCE_SCHEMA BATCH TABLE IO
      
      Improved test cases
    ------------------------------------------------------------
    revno: 8146.1.14 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-08-05 17:00:10 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.13
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Sat 2014-08-02 02:08:16 +0200
    message:
      Fixed a build break when compiling without the performance schema
    ------------------------------------------------------------
    revno: 8146.1.12 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Sat 2014-08-02 00:55:58 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.11
    committer: HaixLi < haixiang.li@oracle.com >
    branch nick: mysql-trunk-wl7802-0725
    timestamp: Tue 2014-07-29 23:42:59 -0400
    message:
      WL#7802 PERFORMANCE SCHEMA, BATCH TABLE IO
      
      Description:
      ------------
      This WL will reduce number of calls for PFS when MySQL do join operation.  
      We set batch mode for PFS when do join operation in executor to implement
      reduce number of calls for PFS.
      This WL does not change any functionality of join operation.
      
      Notice:
      This version covers:
      1 Fix some assert.
      2 Resolve the confilt that merge WL#6042 to WL#7802.
    ------------------------------------------------------------
    revno: 8146.1.10 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-07-29 22:33:30 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.9 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-07-28 21:51:36 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
      
      Removed all changes in the optimizer done for WL 7802
      due to the refactoring introduced in mysql-trunk,
      which caused too many merge conflicts.
      
      The patch in the optimizer for WL 7802 needs to be revised,
      and applied again on the refactored code.
    ------------------------------------------------------------
    revno: 8146.1.8
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-07-28 19:44:13 +0200
    message:
      Added test case.
    ------------------------------------------------------------
    revno: 8146.1.7
    committer: HaixLi < haixiang.li@oracle.com >
    branch nick: mysql-trunk-wl7802-0725
    timestamp: Fri 2014-07-25 20:18:12 -0400
    message:
      WL#7802 PERFORMANCE SCHEMA, BATCH TABLE IO
      
      Description:
      ------------
      This WL will reduce number of calls for PFS when MySQL do join operation.  
      We set batch mode for PFS when do join operation in executor to implement
      reduce number of calls for PFS. 
      This WL does not change any functionality of join operation. 
      
      Notice:
      It is the first version.
      Next version will fix a assert and modify code by Roy's suggestion.
    ------------------------------------------------------------
    revno: 8146.1.6 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Fri 2014-07-25 10:42:33 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.5 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-07-21 17:03:22 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.4 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Mon 2014-07-07 16:21:40 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.3 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Tue 2014-06-24 23:14:44 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7802
    ------------------------------------------------------------
    revno: 8146.1.2
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Thu 2014-06-05 21:28:01 +0200
    message:
      WL#7802 PERFORMANCE SCHEMA, BATCH TABLE IO
      
      Added support to report table io events by batch,
      in the performance schema instrumentation.
    ------------------------------------------------------------
    revno: 8146.1.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7802
    timestamp: Thu 2014-06-05 13:49:34 +0200
    message:
      New branch
------------------------------------------------------------
revno: 8701 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Wed 2014-08-27 14:31:24 +0200
message:
  merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.340.8
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6
    timestamp: Wed 2014-08-27 11:26:14 +0200
    message:
      Bug#19506247 DTRACE SUPPORT DON'T BUILD IF COMMON BUILD -WP,OPTION WFORMAT=* IS USED
      
      In dtrace.cmake: change pattern matching to strip off all warnings, including -Werror=xxx-xxx
      but keep flags like -Wp,-D_FORTIFY_SOURCE=2
------------------------------------------------------------
revno: 8700
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-27 13:28:12 +0200
message:
  The following tests take too long in Valgrind. Only run them with --big-test.
  
  auth_sec.openssl_cert_generation       
  innodb_gis.rtree_create_inplace
  main.ctype_binary      
  main.ctype_uca 
  main.mysql_upgrade     
  main.partition 
  parts.part_exch_valid_key_innodb       
  parts.part_exch_valid_range_innodb     
  rpl.rpl_server_uuid
------------------------------------------------------------
revno: 8699 [merge]
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-27 12:35:37 +0200
message:
  Auto-merge WL#7159, WL#7828 & WL#8003 from mysql-trunk-wl7159-8.
    ------------------------------------------------------------
    revno: 8688.1.6
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: trunk-wl7159-8
    timestamp: Wed 2014-08-27 14:22:48 +0400
    message:
      WL#7159: Move time zone tables and help tables from MyISAM to transactional
      storage.
      
      Pushed on behalf of Dmitry Shulga.
      
      The following system tables are migrated to InnoDB by this patch:
        - help_category
        - help_keyword
        - help_relation
        - help_topic
        - time_zone
        - time_zone_leap_second
        - time_zone_name
        - time_zone_transition,
        - time_zone_transition_type
      
      Attachable transaction provided by previously pushed WLs (7828 and 8003) is
      started to be used by this patch.
    ------------------------------------------------------------
    revno: 8688.1.5
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: trunk-wl7159-8
    timestamp: Tue 2014-08-26 18:54:39 +0400
    message:
      WL#7159: Move time zone tables and help tables from MyISAM
      to transactional storage.
      
      Prerequisite refactoring patch: in-line open_and_lock_tables() call
      in open_nontrans_system_tables_for_read().
      
      The reasons for that are:
      
        - open_and_lock_tables() should now be used for all but
          the system tables. That allows better component layering and
          will be of use later.
      
        - open_and_lock_tables() code can be significantly simplified
          for opening system tables only (no handling of derived tables,
          no debug code).
      
        - it should be possible to assert that open_and_lock_tables()
          is used only for user tables;
      
        - it should be possible to assert that the system tables are
          opened used open_nontrans_system_tables_for_read().
      
      NOTE:
      
        - open_nontrans_system_tables_for_read() will be paired with
          open_trans_system_tables_for_read() to open system tables in InnoDB.
      
        - open_nontrans_system_tables_for_read() will eventually be gone
          (as soon as all system tables reside in InnoDB).
    ------------------------------------------------------------
    revno: 8688.1.4
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: trunk-wl7159-8
    timestamp: Tue 2014-08-26 18:52:06 +0400
    message:
      WL#7159: Move time zone tables and help tables from MyISAM
      to transactional storage.
      
      Prerequisite refactoring patch:
        - rename open_system_tables_for_read() to
          open_nontrans_system_tables_for_read();
        - rename close_system_tables() to close_nontrans_system_tables().
    ------------------------------------------------------------
    revno: 8688.1.3
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: trunk-wl7159-8
    timestamp: Tue 2014-08-26 18:46:34 +0400
    message:
      WL#7828: InnoDB: attachable transaction support.
      
      This is a cumulative patch from Satya Bodapati and Alexander Nozdrin.
      
      The patch introduces the InnoDB-side support (as opposed to the server-side
      support) of attachable transactions. Attachable transaction is a tiny
      subclass of a nested transaction. Attachable transactions are AC-RO-RC-NL
      transactions (auto-commit, read-only, read-committed, non-locking).
      
      The patch does the following:
        - Adds asserts to verify that the attachable transaction:
          - doesn't take locks,
          - is READ-COMMITTED,
          - is READ-ONLY,
          - is AUTOCOMMIT.
        - Adds test cases to ensure that InnoDB handles attachable
          transactions properly.
    ------------------------------------------------------------
    revno: 8688.1.2
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: trunk-wl7159-8
    timestamp: Tue 2014-08-26 18:30:20 +0400
    message:
      WL#8003: Server support for attachable transactions.
      
      The patch introduces the server-side support (as opposed to the InnoDB-side
      support) of attachable transactions. Attachable transaction is a tiny
      subclass of a nested transaction. Attachable transactions are AC-RO-RC-NL
      transactions (auto-commit, read-only, read-committed, non-locking).
      
      A new flag (HA_ATTACHABLE_TRX_COMPATIBLE) is added to the Handler API.
      Briefly, HA_ATTACHABLE_TRX_COMPATIBLE means that a SE is compatible with
      the attachable transaction requirements. That in turn means that the tables
      from the SE can be used within an attachable transaction. It does NOT mean
      that the SE supports attachable transactions per say.
      
      New public operations are introduced to THD to begin and end attachable
      transaction.
      
      Attachable transactions will be used later by to read from transactional
      system tables.
    ------------------------------------------------------------
    revno: 8688.1.1
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: trunk-wl7159-8
    timestamp: Tue 2014-08-26 17:54:39 +0400
    message:
      WL#8003: Server support for attachable transactions.
      
      Preliminary patch to make THD::m_transaction a pointer
      (std::auto_ptr<Transaction>). THD::m_transaction will be
      saved/replaced/restored later. Making it a pointer simplifies
      that.
------------------------------------------------------------
revno: 8698 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Tue 2014-08-19 11:43:20 +0200
message:
  WL#7445 PERFORMANCE SCHEMA: instrument SX-lock for rw_lock
  
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 7046.1.30 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-08-19 11:20:46 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.29 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-08-18 18:44:55 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.28 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Sun 2014-08-17 18:50:33 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.27
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Thu 2014-08-21 12:46:01 +0200
    message:
      WL#7445 PERFORMANCE SCHEMA: instrument SX-lock for rw_lock
      
      Added unit tests for code coverage,
      fixed TRY instrumentation.
    ------------------------------------------------------------
    revno: 7046.1.26 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Thu 2014-08-21 10:34:04 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.25 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-08-19 18:10:47 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.24 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-08-19 17:03:06 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.23 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-08-18 15:01:35 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.22 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-08-18 12:12:11 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.21 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-08-11 12:42:08 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.20 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Sun 2014-08-10 23:30:43 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.19 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Wed 2014-08-06 18:22:09 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.18 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-08-05 16:54:03 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.17 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Sat 2014-08-02 00:51:11 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.16 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-07-29 22:25:20 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.15
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-07-29 20:54:04 +0200
    message:
      Improved test robustness
    ------------------------------------------------------------
    revno: 7046.1.14 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-07-29 02:20:58 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.13
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-07-29 00:51:34 +0200
    message:
      Added test cases
    ------------------------------------------------------------
    revno: 7046.1.12
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-07-28 11:49:28 +0200
    message:
      WL#7445 PERFORMANCE_SCHEMA, SX-locks
      
      Implemented code review comments.
    ------------------------------------------------------------
    revno: 7046.1.11 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Fri 2014-07-25 11:58:18 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.10
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-07-21 16:24:31 +0200
    message:
      WL#7445 PERFORMANCE SCHEMA, SX locks
      
      Implement a different instrument name for SX locks
    ------------------------------------------------------------
    revno: 7046.1.9
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-07-21 11:44:46 +0200
    message:
      WL#7445 PERFORMANCE SCHEMA, Instrument SX-lock for rw locks
      
      Code cleanup
    ------------------------------------------------------------
    revno: 7046.1.8 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-07-21 10:28:46 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.7 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-07-07 16:35:18 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.6 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Tue 2014-06-24 23:31:16 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.5 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Mon 2014-06-02 13:56:01 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.4 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445-b
    timestamp: Wed 2014-04-30 16:40:34 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.3 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Wed 2014-01-15 01:57:41 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl7445
    ------------------------------------------------------------
    revno: 7046.1.2
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Wed 2014-01-15 00:16:03 +0100
    message:
      WL#7445 PERFORMANCE SCHEMA: instrument SX-lock for rw_lock
    ------------------------------------------------------------
    revno: 7046.1.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7445
    timestamp: Fri 2013-11-29 13:12:56 +0100
    message:
      New branch
------------------------------------------------------------
revno: 8697
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: wl7339-trunk
timestamp: Wed 2014-08-27 09:25:41 +0200
message:
  WL#7339 Use improved records per key estimate interface in optimizer
        
  This worklog implements the needed changes in the optimizer
  and server to start using the new "records per key" interface that
  was implmeneted in WL#7338. This interface allows storage engines
  to provide "records per key" values using float values instead of
  integer values. Using this new interface instead of using the
  currently used rec_per_key values gives more correct record estimates
  for ref access against an indexed column.
------------------------------------------------------------
revno: 8696
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-27 11:50:58 +0800
message:
  Bug #19473313 ASSERT ALWAYS TRUE DUE TO LIMITED RANGE OF DATA TYPE GIVES COMPILATION ERROR.
  The assert is always true, so remove it. 
------------------------------------------------------------
revno: 8695
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-27 11:32:23 +0800
message:
  Bug #19498383 MERGE_LARGE_TESTS UNIT TEST IS FAILING ON PB2
  Fix buffer free space management bug, allocate extra space for WKB buffer and fill it with defined pattern bytes.
------------------------------------------------------------
revno: 8694
committer: HaixLi < haixiang.li@oracle.com >
branch nick: mysql-trunk-bug-18066518-commit
timestamp: Wed 2014-08-27 04:06:07 +0200
message:
  Bug #18066518: THE COST VALUE IS A NEGATIVE NUMBER FOR MERGE ENGINE TABLE
  Bug #18486293: ASSERTION FAILED: KEYS >= 0.0 IN
                 COST_MODEL_TABLE::KEY_COMPARE_COST      
  
  Description:
  ------------
  The cost numbers should not be negative.
  The cost calculation code for loose index scan in cost_group_min_max()
  does not handle the case where the table is empty.
  When table_records == 0, the log(0) produces -INFINITY.
  So we should avoid calling log(0).
  
  Fix:
  ----
  When MySQL calculate the cost, we ban it to be a negative number.
  Fixed it by avoiding use log(0) when table_records == 0.
  
  Test case added.
------------------------------------------------------------
revno: 8693
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Wed 2014-08-27 09:54:40 +0800
message:
  Fix Bug #19474851 - INSERT INTO GIS TABLE, ASSERT ASSERTION: (ULINT)
  REC == OFFSETS[2] 
        
  rb://6445 approved by Sunny Bains
------------------------------------------------------------
revno: 8692 [merge]
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 16:32:11 -0500
message:
  merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.340.7
    committer: kevin.lewis@oracle.com
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-26 16:25:59 -0500
    message:
      Backport a change to i_innodb.innodb_bug16066351 that was 
      pushed to mysql-trunk in -r6708 on 2013-10-16.  This change 
      prevents this test from trying to run on MacOS10.8 and 
      any other OS that does not have case sensitive file names.
------------------------------------------------------------
revno: 8691
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 12:48:07 -0500
message:
  Fix Solaris build break again
------------------------------------------------------------
revno: 8690
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 11:53:24 -0500
message:
  Cleanup of previous push due to a build break on Solaris.
------------------------------------------------------------
revno: 8689
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 10:37:03 -0500
message:
  Non-functional changes culled from the patch for wl 6206 found in rb4869.
  
  Approved by Vasil in RB6174
------------------------------------------------------------
revno: 8688
committer: Norvald H. Ryeng <norvald.ryeng@oracle.com>
branch nick: mysql-trunk-17833261
timestamp: Tue 2014-08-26 14:01:00 +0200
message:
  Bug#17833261 FOUND_ROWS(): DIFFERENT RESULTS FOR LOCAL LIMIT WITH AND
  WITHOUT OFFSET IN UNION
  
  Problem: A UNION ALL query with SQL_CALC_FOUND_ROWS and a limit with
  offset for one query block reports the wrong number of found rows.
  
  In select_union_direct::send_eof(), the number of rows is aggregated
  by reading THD::limit_found_rows inc after each query block
  execution. However, this number is the sum of the limit (or number of
  rows, if smaller than the limit) and the offset. This number is
  correct for non-union queries, but in union queries the actual number
  of rows returned to the user is needed.
  
  Fix: Subtract the offset from limit_found_rows in
  select_union_direct::send_eof().
------------------------------------------------------------
revno: 8687
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Tue 2014-08-26 13:44:54 +0200
message:
  Follow-up WL#5869: Optimizer cost: use Cost_estimate in range optimizer
  
  Fixed one compiling error without opt.trace.
------------------------------------------------------------
revno: 8686
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: mysql-trunk-valgrind
timestamp: Tue 2014-08-26 15:33:39 +0400
message:
  Fix for bug #19295157: VALGRIND FAILURE IN THE AUDIT_LOG TESTS ON DAILY-TRUNK
  
  Free unused memory.
------------------------------------------------------------
revno: 8685
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 16:38:37 +0530
message:
  Bug #19495721 REMOVE C-STYLE IB_LOGF FUNCTION, REPLACE USAGE WITH IB::INFO(),
  IB::WARN(), ETC
  
  Problem:
  
  There are now two ways of writing log messages - using the C-style function
  ib_logf() and using the C++ style classes ib::info, ib::warn, ib::error and
  ib::fatal.  
  
  Solution:
  
  Remove the C-style ib_logf(), so that everybody will start using the C++
  style classes.
  
  Note:
  
  This patch replaces ib_logf() with the new C++ style logging technique for
  btr, buf, data and dict subsystems.  There will be subsequent patches for
  other subsystems. 
  
  rb#6442 approved by Vasil
------------------------------------------------------------
revno: 8684 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 12:14:58 +0200
message:
  Add rpm-sles directory for sles repo rpm builds
    ------------------------------------------------------------
    revno: 8614.1.2
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.7.5-m15-sles-release
    timestamp: Fri 2014-08-22 13:02:39 +0200
    message:
      Bug#19481375: Service restart failing with sles repo packages once native pkgs are installed
    ------------------------------------------------------------
    revno: 8614.1.1
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.7.5-m15-sles-release
    timestamp: Wed 2014-08-20 09:27:17 +0200
    message:
      Added rpm-sles directory for SLES repo rpm build
------------------------------------------------------------
revno: 8683 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 11:53:56 +0200
message:
  Merge from 5.6 => trunk - Rename enterprise repo packages to commercial
    ------------------------------------------------------------
    revno: 3902.340.6 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-26 11:45:55 +0200
    message:
      Merge from 5.5 => 5.6 Rename enterprise repo packages to commercial
        ------------------------------------------------------------
        revno: 2875.598.33
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-08-26 11:40:54 +0200
        message:
          Rename enterprise repo packages to commercial
------------------------------------------------------------
revno: 8682
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Tue 2014-08-26 10:48:47 +0200
message:
  WL#2489: Better ONLY_FULL_GROUP_BY mode
  
  Fix result file changes after push to trunk.
------------------------------------------------------------
revno: 8681
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 11:36:59 +0300
message:
  Pass --core-file when starting mysqld in order to catch crashes such as
  Bug#19505636 INNODB.LOG_FILE_SIZE FAILS WITH WRONG ERRNO 2006
  Approved by Jimmy Yang.
------------------------------------------------------------
revno: 8680
committer: Shivji Jha <shivji.jha@oracle.com>
branch nick: duplicate-server-id
timestamp: Tue 2014-08-26 12:04:33 +0530
message:
  Bug#18192817: IGNORE_SERVER_IDS IS SHOWING DUPLICATE
                SERVER ID FOR LAST VALUE
  
  Problem:
  
  The list of insert_id is stored in an array.
  Before inserting in the array, there is a binary
  search to find if the same server_id already 
  exists in the list. The issue is that the list
  is not sorted. Note that binary search can ONLY
  be applied on a sorted list.
  
  As a result the list can sometimes contain a value
  but not be spotted by the binary search logic.
  In such a case, the value is again added to the list.
  
  Fix:
  
  The bug was fixed by the patch for Bug#18920203.
  The patch for this bug implemented insertion-sort
  in Prealloced_array, and used insert_unique() to
  keep the array sorted at all times.
  
  The current patch only adds a test to verify that
  we dont have duplicate values in the list at any time.
------------------------------------------------------------
revno: 8679 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 11:47:18 +0530
message:
  Bug#19504805: COPY DEFAULT.WEEKLY FILE OF THREAD_POOL AND
                AUDIT_LOG TO FILE FROM 5.5 FROM 5.6
  
  Null merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.340.5 [merge]
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-26 11:39:11 +0530
    message:
      Bug#19504805: COPY DEFAULT.WEEKLY FILE OF THREAD_POOL AND
                    AUDIT_LOG TO FILE FROM 5.5 FROM 5.6
      
      Null merge from 5.5 to 5.6.
        ------------------------------------------------------------
        revno: 2875.598.32
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-08-26 11:37:01 +0530
        message:
          Bug#19504805: COPY DEFAULT.WEEKLY FILE OF THREAD_POOL AND
                        AUDIT_LOG TO FILE FROM 5.5 FROM 5.6
          
          On pb2, for 5.5 weekly tests were not running for thread_pool
          and audit_log modules. Issue here is, default.weekly is not
          created for these modules in 5.5. As part of this patch,
          copying default.weekly file for these modules in 5.5 from
          5.6.
------------------------------------------------------------
revno: 8678
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 08:52:13 +0300
message:
  Make a global function static.
------------------------------------------------------------
revno: 8677
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 08:51:53 +0300
message:
  Correct the documentation of some parameters.
------------------------------------------------------------
revno: 8676
committer: Bharathy Satish <bharathy.x.satish@oracle.com>
branch nick: mysql-trunk-new
timestamp: Tue 2014-08-26 10:45:04 +0530
message:
  WL#7766: Deprecate the EOF packet
                          
  Recent extensions to OK packet (reference WL#4797 and WL#6885) contains
  more additional information which might be needed to clients and connectors
  when EOF packet is sent. Since it would be redundant to apply the same
  extensions in EOF packet, this WL will take care of handling the changes to
  deprecate EOF packet and send OK packet wherever needed.
  
  Ok packet contains all the information which is present in EOF packet too.
  Contents of EOF packet include packet marker, server status and warning count.
  This information can also be extracted from OK packet along with more 
  additional information if needed.
                    
  With this WL EOF packet will no longer be used with result sets.
------------------------------------------------------------
revno: 8675 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-26 09:41:47 +0800
message:
  A follow-up push for bug#18477009, the test case is unstable on PB2.   
  Just remove it away.
    ------------------------------------------------------------
    revno: 3902.340.4
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-26 09:39:29 +0800
    message:
      A follow-up push for bug#18477009, the test case is unstable on PB2.
      Just remove it away.
------------------------------------------------------------
revno: 8674
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Tue 2014-08-26 10:17:15 +1000
message:
  Undo remove of:
  
  === removed file 'packaging/rpm-fedora/mysql-5.6.16-mysql-install.patch'
  === removed file 'packaging/rpm-oel/mysql-5.6.16-mysql-install.patch'
  === removed file 'scripts/mysql_install_db.pl.in'
  === removed file 'scripts/mysql_install_db.sh'
  
  They were removed in : georgi.kodinov@oracle.com-20140825075132-02yp6yypnfyit3sb
------------------------------------------------------------
revno: 8673
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Tue 2014-08-26 09:31:27 +1000
message:
  Revert jon.hauglid@oracle.com-20140825083117-lr7uj623qcaud3jb
  Revert georgi.kodinov@oracle.com-20140825075132-02yp6yypnfyit3sb
  
  Requested by Anitha because it breaks RQG and JET tests.
------------------------------------------------------------
revno: 8672
committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
branch nick: 57-19463867
timestamp: Mon 2014-08-25 19:24:16 +0100
message:
  Bug#19463877: I_MAIN.AUDIT_GENERAL_LOG TEST FAILS WITH RESULT CONTENT MISMATCH
  Bug#19463867: MAIN.AUDIT_PLUGIN FAILS WITH RESULT CONTENT MISMATCH
  
  Hotfix to address pushbuild failures:
  
  Fix for 12368204 changed logging behaviour, but audit_plugin.test was
  updated only with respect to normal protocol, not others such as
  --ps-protocol, so it failed in those modes. Hotfix updates the results
  expected by the test to match those produced by the server.
  
  Fix for 12368204 added audit_general_log.test which was set up only
  with respect to normal mode; hotfix adds support for all other modes,
  and aligns expected values with those produced.
  
  This fix addresses the immediate pb failure and updates the "multi
  mode deficiency" in the tests/results; any change to the actual code
  and behaviour introduced by 12368204 can then happen in that framework.
------------------------------------------------------------
revno: 8671
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Mon 2014-08-25 17:09:58 +0200
message:
  Bug#19488750 REMOVE BOOST_INCLUDE_DIR FROM INCLUDE_DIRECTORIES
  
  Having BOOST_INCLUDE_DIR in INCLUDE_DIRECTORIES
  means boost is visible to all source files.
  
  Add the path as a source file property instead,
  so we can maintain an explicit list of sources using boost.
------------------------------------------------------------
revno: 8670 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 16:27:22 +0200
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.340.3 [merge]
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-08-25 16:26:06 +0200
    message:
      Null merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.31
        committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-08-25 16:25:07 +0200
        message:
          Bug #19471516 SERVER CRASHES WHEN EXECUTING ALTER TABLE ADD FOREIGN KEY
          
          Problem:
          
          We maintain two rb trees in each dict_table_t.  The foreign_rbt must be in
          sync with foreign_list.  The referenced_rbt must be in sync with
          referenced_list. There is one function which checks this consistency and it
          failed, resulting in an assert failure.
          
          The root cause of the problem was identified that the search order was 
          lost in the referenced_rbt.  This is because while renaming the table,
          we didn't not refresh this referenced_rbt.
          
          Solution:
          
          When a foreign key is renamed, we must delete and re-insert into both
          foreign_rbt and referenced_rbt.  
          
          rb#6412 approved by Jimmy.
------------------------------------------------------------
revno: 8669 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 16:20:16 +0200
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.340.2
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-08-25 16:19:16 +0200
    message:
      Bug #19471516 SERVER CRASHES WHEN EXECUTING ALTER TABLE ADD FOREIGN KEY
      
      Problem:
      
      We maintain foreign keys sorted by foreign->id in an std::set object.  If
      the foreign->id is to be changed, then the foreign key object must be 
      removed from std::set and re-inserted.  This is necessary to maintain the
      search order in the std::set object.
      
      The root cause of the problem was identified that the search order was 
      lost in the referenced_set.  This is because while renaming the table,
      we didn't not refresh this referenced_set.
      
      Solution:
      
      When a foreign key is renamed, we must delete and re-insert into both
      foreign_set and referenced_set.  
      
      rb#6420 approved by Jimmy.
------------------------------------------------------------
revno: 8668
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Mon 2014-08-25 15:32:57 +0200
message:
  Bug#19459193 GET RID OF DYNAMIC_ARRAY IN CIRCULAR_BUFFER_QUEUE
  
  Post-push fix: -Werror build broken in optimized mode
  
  gcc 4.4 was warning: error: 'group.st_slave_job_group::ts'
  solution: use value-initialization of two stack-allocated objects of type Slave_job_group
  
  gcc 4.7 gives warning warning: variable ?ret? set but not used [-Wunused-but-set-variable]
  solution: annotate it with __attribute__((unused))
------------------------------------------------------------
revno: 8667
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-tls
timestamp: Mon 2014-08-25 15:26:57 +0200
message:
  Bug#19498446: REFACTOR THREAD-LOCAL STORAGE IMPLEMENTATION
  
  This patch moves the thread-local storage implementation to
  a separate file, removes dead code and changes the implementation
  from macros to static inline functions.
  
  The patch also removes serveral unneeded fields from
  st_my_thread_var, reducing memory usage and reducing
  thread-local storage usage.
  
  No chages in behavior.
------------------------------------------------------------
revno: 8666
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Mon 2014-08-25 14:26:21 +0200
message:
  WL#2489: Better ONLY_FULL_GROUP_BY mode
  
  Fix result file changes after push to trunk.
------------------------------------------------------------
revno: 8665 [merge]
committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 13:20:49 +0200
message:
  Empty version change upmerge
    ------------------------------------------------------------
    revno: 3902.340.1 [merge]
    author: murthy.narkedimilli@oracle.com
    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-08-25 13:14:23 +0200
    message:
      Raise version number after cloning 5.6.21
        ------------------------------------------------------------
        revno: 2875.598.30
        author: murthy.narkedimilli@oracle.com
        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-08-25 13:04:23 +0200
        message:
          Raise version number after cloning 5.5.40
------------------------------------------------------------
revno: 8664
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 12:15:25 +0300
message:
  Non-functional change: remove "using std::min;" and use
  "std::min(a, b);" instead of "min(a, b);"
------------------------------------------------------------
revno: 8663
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 12:14:01 +0300
message:
  Fix Bug#19488149 REMOVE A GOTO IN FIL_MUTEX_ENTER_AND_PREPARE_FOR_IO()
  
  Remove a goto in fil_mutex_enter_and_prepare_for_io(), convert:
  
  OLD CODE:
  retry:
  	A;
  	if (B) {
  		C;
  		goto retry;
  	}
  	D;
  	goto retry;
  
  NEW CODE:
  	for (;;) {
  		A;
  		if (B) {
  			C;
  			continue;
  		}
  		D;
  	}
  
  Approved by:	Sunny (rb:6427)
------------------------------------------------------------
revno: 8662
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 12:13:07 +0300
message:
  Remove a goto in fil_mutex_enter_and_prepare_for_io(), convert:
  
  OLD CODE:
  close_more:
  	A;
  	if (B) {
  		goto close_more;
  	}
  
  NEW CODE:
  	do {
  		A;
  	} while (B);
  
  Approved by:	Sunny (via IM)
------------------------------------------------------------
revno: 8661
committer: viswanatham gudipati <viswanatham.gudipati@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 14:44:11 +0530
message:
  Rmoved extra folder conf from the cc file
------------------------------------------------------------
revno: 8660
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-plain
timestamp: Mon 2014-08-25 10:57:15 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Post-push fix: Fix broken build on Solaris with InnoDB memcached and
  without performance schema.
------------------------------------------------------------
revno: 8659
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Mon 2014-08-25 10:31:17 +0200
message:
  Fix broken build: Add missing #includes
------------------------------------------------------------
revno: 8658 [merge]
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: mysql-trunk-wl7688
timestamp: Mon 2014-08-25 10:51:32 +0300
message:
  merge with trunk
    ------------------------------------------------------------
    revno: 8641.1.2
    committer: Georgi Kodinov <georgi.kodinov@oracle.com>
    branch nick: mysql-trunk-wl7688
    timestamp: Mon 2014-08-25 10:47:31 +0300
    message:
      made m_i_db.test experimental
    ------------------------------------------------------------
    revno: 8641.1.1 [merge]
    committer: Georgi Kodinov <georgi.kodinov@oracle.com>
    branch nick: merge-mysql-trunk-wl7688
    timestamp: Fri 2014-08-22 11:21:54 +0300
    message:
      merge from trunk
        ------------------------------------------------------------
        revno: 8640.1.1 [merge]
        committer: Georgi Kodinov <georgi.kodinov@oracle.com>
        branch nick: merge-mysql-trunk-wl7688
        timestamp: Fri 2014-08-22 11:17:31 +0300
        message:
          merge to trunk
            ------------------------------------------------------------
            revno: 8299.1.24
            committer: Sunanda Menon<sunanda.menon@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-08-21 11:59:44 +0200
            message:
              #reverting the postinstall changes for 7688
            ------------------------------------------------------------
            revno: 8299.1.23
            committer: Sunanda Menon<sunanda.menon@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Wed 2014-08-20 14:04:32 +0200
            message:
              #Testing for mysql user as siwtch to mysql is needed for Solaris in chroot
            ------------------------------------------------------------
            revno: 8299.1.22
            committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Mon 2014-08-18 15:57:18 +0200
            message:
              * Attempt to fix crash on solaris.
            ------------------------------------------------------------
            revno: 8299.1.21
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Fri 2014-08-15 14:13:52 +0300
            message:
              Addendum to WL#7688. Added the server's output to the failure message for diagnostic reasons
            ------------------------------------------------------------
            revno: 8299.1.20
            committer: Sunanda Menon<sunanda.menon@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Tue 2014-08-12 07:53:38 +0200
            message:
              mysql, test dirs are not to be created.
            ------------------------------------------------------------
            revno: 8299.1.19
            committer: Harin Vadodaria <harin.vadodaria@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-08-07 16:07:43 +0530
            message:
              WL#7688 : Deprecate mysql_install_db and
              	  create a new program in C++
              
              Description : create_initial_db.cmake.in relies
                            on mysql_system_tables_data.sql to
                            have information about user account.
                            Since new mysql_install_db
                            generates "create user" on its own,
                            user creation part was removed from
                            mysql_system_tables_data.sql. This
                            resulted in windows installations
                            having no data in mysql.user table.
              
              Solution : Added file mysql_user_data_win.sql
                         and modified create_initial_db.cmake.in
                         to process it.
            ------------------------------------------------------------
            revno: 8299.1.18
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-07-31 14:48:58 +0300
            message:
              fixed a wrong size passed to aes_decrypt
            ------------------------------------------------------------
            revno: 8299.1.17
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-07-31 12:13:40 +0300
            message:
              WL#7688 Addendum
              
              * Fix the encrypted file reader to handle
              endianness properly. 
              * Used named constants for non-0/1 
              numbers.
            ------------------------------------------------------------
            revno: 8299.1.16
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Wed 2014-07-30 10:51:39 +0300
            message:
              Fix for bug #19192826 and bug #19192849
              
              * Added $basedir/sbin to the list of directories to search for mysqld
              * added $basedir/share/mysql/english to the list of directories to search for language
              * Addressed a failure to set the group id reported by Sunanda
            ------------------------------------------------------------
            revno: 8299.1.15
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Fri 2014-07-25 13:52:19 +0300
            message:
              WL#7688 addendum 
              
              Fix a problem when compiling with Sun Studio:
              The initialization of the Log class it was passing 
              a pointer to the (yet uninitialized) member m_buffer
              to the base class initializer.
              Fixed by passing a NULL pointer to the constructor
              and in the constructor body (where m_buffer is already
              initialized) calling the init() function of the base class.
              
              Also fixed a warning of an "out" local variable 
              overshadowing a global by remaming it to sout.
            ------------------------------------------------------------
            revno: 8299.1.14 [merge]
            committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Mon 2014-07-21 10:44:09 +0200
            message:
              fixed: setuid is set prematurely and causes the password file to be generated in the wrong directory. The password file needs to be created before the user changes euid.
              fixed: attempt to set egid to pw_gid and not pw_uid.
                ------------------------------------------------------------
                revno: 7766.4.2
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-07-21 10:16:49 +0200
                message:
                  fixed: setuid is set prematurely and causes the password file to be generated in the wrong directory. The password file needs to be created before the user changes euid.
                  fixed: attempt to set egid to pw_gid and not pw_uid.
            ------------------------------------------------------------
            revno: 8299.1.13 [merge]
            committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Fri 2014-07-18 13:17:06 +0200
            message:
              Bug#18914499 SERVICE START FAILING WITH YUM REPO RPMS FAILING WITH CAN'T LOCATE MYSQLD ERROR
              
              Problem: The mysqld process is started using execve which in later kernels have been patch to prevent it from changing uid/gid. Even if mysqld is started with euid=0 it will fail to downgrade its euid.
              
              Fix: Instead this must be done by the parent process. In addition the mysqld process should never attempt to change its euid so we can't pass the --user option.
                ------------------------------------------------------------
                revno: 7766.4.1
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Fri 2014-07-18 13:12:00 +0200
                message:
                  Bug#18914499 SERVICE START FAILING WITH YUM REPO RPMS FAILING WITH CAN'T LOCATE MYSQLD ERROR
                  
                  Problem: The mysqld process is started using execve which in later kernels have been patch to prevent it from changing uid/gid. Even if mysqld is started with euid=0 it will fail to downgrade its euid.
                  
                  Fix: Instead this must be done by the parent process. In addition the mysqld process should never attempt to change its euid so we can't pass the --user option.
            ------------------------------------------------------------
            revno: 8299.1.12
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-07-10 11:48:30 +0300
            message:
              Fixed a crash on solaris 10 due to a wrong out of bounds substr() call.
            ------------------------------------------------------------
            revno: 8299.1.11
            committer: Sunanda Menon<sunanda.menon@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-07-10 06:09:20 +0200
            message:
              Trying to fix Solaris PKGVER issues
            ------------------------------------------------------------
            revno: 8299.1.10
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Tue 2014-07-08 15:49:50 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.9
            committer: Sayantan Dutta <sayantan.dutta@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-07-03 18:17:49 +0530
            message:
              BUG #19149727 - INTRODUCE A VARIABLE  POINTING TO THE BINARY MYSQL_INSTALL_DB
            ------------------------------------------------------------
            revno: 8299.1.8
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Tue 2014-07-01 15:44:49 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.7
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Sat 2014-06-28 10:14:23 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.6
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Sat 2014-06-28 00:16:36 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.5
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Fri 2014-06-27 22:10:22 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.4
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Fri 2014-06-27 16:54:43 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.3
            committer: Harin Vadodaria <harin.vadodaria@oracle.com>
            branch nick: wl7688
            timestamp: Fri 2014-06-27 15:34:18 +0530
            message:
              wl#7688 : Deprecate mysql_install_db and
              	 create a new program in C++
              
              Description : A non-empty data directory causes
                            mysql_install_db to fail.
              
              Solution :    Removing creation of directory for
                            test db.
            ------------------------------------------------------------
            revno: 8299.1.2
            committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Fri 2014-06-27 14:30:24 +0530
            message:
              WL#7688   : Deprecate mysql_install_db and create a new program in C++
              Issue     : To automate testing of the feature mysql_install_db
              Solution  : Added new mtr scripts for mysql_install_db
            ------------------------------------------------------------
            revno: 8299.1.1 [merge]
            committer: Georgi Kodinov <georgi.kodinov@oracle.com>
            branch nick: mysql-trunk-wl7688
            timestamp: Thu 2014-06-26 12:31:32 +0300
            message:
              wl#7688 merged with trunk
                ------------------------------------------------------------
                revno: 7766.1.42
                committer: Georgi Kodinov <georgi.kodinov@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-25 18:51:19 +0300
                message:
                  WL#7688: Addendum 4
                  
                  * fixed compilation problem with Sun Sudio due to
                  a wrong header file order. 
                  * Added an extern "C" for the callback expected by 
                  a C function
                  * Added an empty line at the end of a header file infix_ostream_it.h
                ------------------------------------------------------------
                revno: 7766.1.41
                committer: Georgi Kodinov <georgi.kodinov@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-25 16:36:58 +0300
                message:
                  WL#7688 addendum #3 : 
                  
                  * moved the generated header files out of the source
                  dir and into the build dir to support out-of-source builds.
                  * fixed the dependecies graph of the generated files 
                  and mysql_install_db
                ------------------------------------------------------------
                revno: 7766.1.40
                committer: Georgi Kodinov <georgi.kodinov@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-25 14:35:04 +0300
                message:
                  WL#7688 Addendum #2: Fix for a solaris compilation error.
                  
                  Made a local copy of the C++11 std::all_of that's not supported 
                  by all of the supported compilers.
                ------------------------------------------------------------
                revno: 7766.1.39
                committer: Georgi Kodinov <georgi.kodinov@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-06-24 18:14:12 +0300
                message:
                  WL#7688 addendum #1: Make unit tests conditional with the same condition as the code.
                ------------------------------------------------------------
                revno: 7766.1.38
                committer: Georgi Kodinov <georgi.kodinov@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-18 15:38:18 +0300
                message:
                  Bug #18939073: PARAMETERS ARE NOT PROPERLY PARSED IN MYSQLD
                  
                  A conservative fix for this bug. 
                  The problem was that m_i_db was still adding --no-defaults even if 
                  --defaults-extra-file was specified when constructing the mysqld 
                  command line.
                  Fixed by the same measure as for --defaults-file : by not adding 
                  --no-defaults.
                ------------------------------------------------------------
                revno: 7766.1.37
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-06-12 15:15:16 +0200
                message:
                  fixed: missing variable initialization in escape_string
                  fixed: missing dellocation in decrypt_login_cnf_file
                ------------------------------------------------------------
                revno: 7766.1.36
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-06-12 10:46:14 +0200
                message:
                  fixed: Improved on unittest for parse_cnf_file
                ------------------------------------------------------------
                revno: 7766.1.35
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-06-12 10:09:42 +0200
                message:
                  fixed: Added INSTALL_MYSQLSHAREDIR to default search paths
                  fixed: missing 'else'-statements in Path class allowed for suboptimal execution
                  fixed: use stat() instead of my_stat() to simplify dependencies.
                ------------------------------------------------------------
                revno: 7766.1.34
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-11 14:52:38 +0200
                message:
                  Bug#18956787 LOGIN DOES NOT GOES INTO SANDBOX MODE IF M_I_DB READS USERID FROM CONFIG EDITOR
                  
                  fixed: Changed meaning of login-path to mean the category in an encrypted cnf file rather than the location of this file.
                  fixed: Added new option 'login-file' to mean the location of an encrypted cnf file with login credentials.
                ------------------------------------------------------------
                revno: 7766.1.33
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-11 11:13:10 +0200
                message:
                  fixed: Reverted coding mistakes.
                ------------------------------------------------------------
                revno: 7766.1.32
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-06-10 17:45:40 +0200
                message:
                  fixed: Added missing path separator
                ------------------------------------------------------------
                revno: 7766.1.31
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-06-10 15:53:16 +0200
                message:
                  fixed: Using @default_prefix@
                ------------------------------------------------------------
                revno: 7766.1.30
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-06-10 11:57:41 +0200
                message:
                  fixed: Imported various variables from CMake to support different deployment layouts
                  fixed: now searching in INSTALL_BINDIR and INSTALL_SBIN dir for mysqld-binary per default
                ------------------------------------------------------------
                revno: 7766.1.29
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-06-09 18:26:18 +0200
                message:
                  fixed: Add basic support for delimiters
                ------------------------------------------------------------
                revno: 7766.1.28 [merge]
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-04 13:19:06 +0200
                message:
                  automerge
                    ------------------------------------------------------------
                    revno: 7766.3.1 [merge]
                    committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                    branch nick: mysql-trunk-wl7688
                    timestamp: Mon 2014-06-02 18:25:44 +0200
                    message:
                      automerge
                        ------------------------------------------------------------
                        revno: 7766.2.10
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-trunk-wl7688
                        timestamp: Wed 2014-05-28 14:31:10 +0200
                        message:
                          Revert the files in CMakeLists.txt + Removed the entries in regulare spec file
                        ------------------------------------------------------------
                        revno: 7766.2.9
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-trunk-wl7688
                        timestamp: Wed 2014-05-28 11:48:26 +0200
                        message:
                          Added missing files in spec file
                        ------------------------------------------------------------
                        revno: 7766.2.8
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-trunk-wl7688
                        timestamp: Wed 2014-05-28 09:03:30 +0200
                        message:
                          Applied Terje patch to wl7688
                        ------------------------------------------------------------
                        revno: 7766.2.7
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-trunk-wl7688
                        timestamp: Tue 2014-05-27 04:50:22 +0200
                        message:
                          Remove mysqlhotcopy from regular spec file to resolve build failures
                        ------------------------------------------------------------
                        revno: 7766.2.6
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-trunk-wl7688
                        timestamp: Tue 2014-05-27 02:10:09 +0200
                        message:
                          Remove mysqlhotcopy to resolve build failures
                        ------------------------------------------------------------
                        revno: 7766.2.5
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-trunk-wl7688
                        timestamp: Mon 2014-05-26 16:19:30 +0200
                        message:
                          Removed mysqlhotcopy from spec file
                ------------------------------------------------------------
                revno: 7766.1.27
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-06-04 13:17:44 +0200
                message:
                  * fixed: --login-path credentials always take precedence over admin-user and admin-host
                ------------------------------------------------------------
                revno: 7766.1.26 [merge]
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-06-02 17:50:21 +0200
                message:
                  * merge
                    ------------------------------------------------------------
                    revno: 7766.2.4
                    committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                    branch nick: mysql-trunk-wl7688
                    timestamp: Mon 2014-05-26 14:29:18 +0200
                    message:
                      *fixed: removed legacy m_i_db scripts
                    ------------------------------------------------------------
                    revno: 7766.2.3
                    committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                    branch nick: mysql-trunk-wl7688
                    timestamp: Thu 2014-05-15 15:37:14 +0200
                    message:
                      *fixed: my_mkdir() depends on umask restrictions set in my_init() hence it can't be used unless server global policies change. Using mkdir() instead.
                    ------------------------------------------------------------
                    revno: 7766.2.2 [merge]
                    committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                    branch nick: mysql-trunk-wl7688
                    timestamp: Wed 2014-05-14 11:26:35 +0200
                    message:
                      *merge trunk -> wl7688
                    ------------------------------------------------------------
                    revno: 7766.2.1
                    committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                    branch nick: mysql-trunk-wl7688
                    timestamp: Wed 2014-05-14 11:20:34 +0200
                    message:
                      * missing ostream header
                ------------------------------------------------------------
                revno: 7766.1.25
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-06-02 17:44:40 +0200
                message:
                  * fixed: search for mysqld in cmake defined target directories
                  * fixed: Allow for multi-comments in extra SQL file
                ------------------------------------------------------------
                revno: 7766.1.24
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-05-12 09:34:30 +0200
                message:
                  * omitted genpriv target from non-unix platforms
                  * fixed: missing closedir after opendir
                ------------------------------------------------------------
                revno: 7766.1.23
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-05-08 15:34:45 +0200
                message:
                  * path.h: moved function bodies from h file to cc file.
                  * code beautifications
                ------------------------------------------------------------
                revno: 7766.1.22
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-05-08 12:10:23 +0200
                message:
                  * added missing file
                ------------------------------------------------------------
                revno: 7766.1.21
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-05-08 10:43:03 +0200
                message:
                  * added --default-extra-file option
                  * fixed behavior of --default-file
                  * Introduced infix iterator generator
                ------------------------------------------------------------
                revno: 7766.1.20
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-05-08 10:43:03 +0200
                message:
                  * added --default-extra-file option
                  * fixed behavior of --default-file
                  * Introduced infix iterator generator
                ------------------------------------------------------------
                revno: 7766.1.19
                committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-05-07 17:12:18 +0200
                message:
                  WL#7826 Deprecate and remove mysql_zap and mysql_waitpid
                ------------------------------------------------------------
                revno: 7766.1.18
                committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-05-07 10:30:40 +0200
                message:
                  Merged cset 7950 from trunk
                ------------------------------------------------------------
                revno: 7766.1.17
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-05-06 13:36:12 +0200
                message:
                  * fixed: extra sql used MyISAM as default engine
                ------------------------------------------------------------
                revno: 7766.1.16
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-05-06 11:39:15 +0200
                message:
                  * fixed: log messages in nicer columns
                  * fixed: escape_string() now desgined for move paradigm
                  * fixed: extra-sql-file now correctly iterates a text file
                ------------------------------------------------------------
                revno: 7766.1.15
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-05-05 19:39:35 +0200
                message:
                  * fixed: don't try to build on Windows yet.
                ------------------------------------------------------------
                revno: 7766.1.14
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-05-05 19:21:43 +0200
                message:
                  * Added --version option
                  * Introduced Log class to eliminate some reoccurring patterns
                  * Fixed: admin_user / admin_host pair now works with proxies_priv table
                  * mysql_system_tables_data.sql is now empty, but kept for future needs.
                ------------------------------------------------------------
                revno: 7766.1.13
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-04-28 17:48:22 +0200
                message:
                  * fixed: use comp_sql to generate bootstrap sql2cc.cc
                  * fixed: client/sql_commands_system_data.h and client/sql_commands_system_tables.h are now generated using cmake
                  * fixed: comp_sql: failed to escape the '\' character
                  * removed all remaining traces of the test db and default user from mysql_system_tables_data.sql
                ------------------------------------------------------------
                revno: 7766.1.12
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Fri 2014-04-25 15:48:55 +0200
                message:
                  *fixed: The '.' should be a legal host name character.
                ------------------------------------------------------------
                revno: 7766.1.11
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Fri 2014-04-25 15:30:25 +0200
                message:
                  *Refactor: Path::qpath() shouldn't do syscalls. Move this to Path::exists() instead.
                ------------------------------------------------------------
                revno: 7766.1.10
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Fri 2014-04-25 14:23:30 +0200
                message:
                  * Fixed: Don't expire password if password cred comes from .login.cnf
                  * Fixed: Allow for --admin-user and --admin-host options to override --login-path credentials.
                  * Fixed: System sometimes crashed if login-path wasn't a qualified path.
                ------------------------------------------------------------
                revno: 7766.1.9
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Fri 2014-04-25 12:43:16 +0200
                message:
                  * Fixed regression: return value of get_admin_credentials() wasn't checked for success, only for failure.
                ------------------------------------------------------------
                revno: 7766.1.8
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-04-24 16:18:12 +0200
                message:
                  * fixed regresssion: --insecure option failed
                  * errno wasn't reset properly after error checks
                ------------------------------------------------------------
                revno: 7766.1.7
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-04-24 14:48:53 +0200
                message:
                  * fixed regression: if language dir was explicitly set the search path was set incorrectly.
                ------------------------------------------------------------
                revno: 7766.1.6
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-04-24 12:30:46 +0200
                message:
                  * Simplied Path class (no need for shared buffers)
                ------------------------------------------------------------
                revno: 7766.1.5
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Thu 2014-04-24 11:45:27 +0200
                message:
                  * Fixed: Use the target datadir if it is empty
                  * Fixed: Default location for .mysql_secret is the home dir of the euid
                ------------------------------------------------------------
                revno: 7766.1.4
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Wed 2014-04-23 17:36:41 +0200
                message:
                  * Datadir now created with the euid owner if specified and running with sufficient priv.
                  * Disabled default signal handler for SIGPIPE
                  * Adjusted process handling for stability. Non-blocking mode seems to cause saturation and EPIPE failure. Using blocking mode seems to work better.
                  * Language dir is not identified by the mandatory "english" subdirectory
                  * Fixed File_exist functor so that path isn't modified if the target isn't found
                  * introduced proper escaping of strings when creating sql
                  * Created defines for constants
                  * moved Access_privileges and option parser code to auth_utils.cc
                  * moved Path helper class to path.h
                  * Added unittest for auth_utils and Path helper class
                ------------------------------------------------------------
                revno: 7766.1.3
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-04-15 02:29:34 +0200
                message:
                  * Added missing files.
                ------------------------------------------------------------
                revno: 7766.1.2
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Tue 2014-04-15 01:01:49 +0200
                message:
                  * Incremental improvement: Dropped popen() in favour of execve()
                ------------------------------------------------------------
                revno: 7766.1.1
                committer: Kristofer Pettersson <kristofer.pettersson@oracle.com>
                branch nick: mysql-trunk-wl7688
                timestamp: Mon 2014-04-07 16:29:49 +0200
                message:
                  WL#7688 Deprecate mysql_install_db and create a new program in C++
                  
                  Preliminary  commit.
                  No Windows testing.
------------------------------------------------------------
revno: 8657 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-25 08:58:11 +0200
message:
  Bug #19471516 SERVER CRASHES WHEN EXECUTING ALTER TABLE ADD FOREIGN KEY
  
  Problem:
  
  We maintain foreign keys sorted by foreign->id in an std::set object.  If
  the foreign->id is to be changed, then the foreign key object must be 
  removed from std::set and re-inserted.  This is necessary to maintain the
  search order in the std::set object.
  
  The root cause of the problem was identified that the search order was 
  lost in the referenced_set.  This is because while renaming the table,
  we didn't not refresh this referenced_set.
  
  Solution:
  
  When a foreign key is renamed, we must delete and re-insert into both
  foreign_set and referenced_set.  
  
  rb#6420 approved by Jimmy.
    ------------------------------------------------------------
    revno: 3902.338.82
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-08-25 07:22:40 +0200
    message:
      Bug #19471516 SERVER CRASHES WHEN EXECUTING ALTER TABLE ADD FOREIGN KEY
      
      Problem:
      
      We maintain foreign keys sorted by foreign->id in an std::set object.  If
      the foreign->id is to be changed, then the foreign key object must be 
      removed from std::set and re-inserted.  This is necessary to maintain the
      search order in the std::set object.
      
      The root cause of the problem was identified that the search order was 
      lost in the referenced_set.  This is because while renaming the table,
      we didn't not refresh this referenced_set.
      
      Solution:
      
      When a foreign key is renamed, we must delete and re-insert into both
      foreign_set and referenced_set.  
      
      rb#6420 approved by Jimmy.
------------------------------------------------------------
revno: 8656 [merge]
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Mon 2014-08-25 10:06:22 +0800
message:
  WL#7236 
  Reimplemented area(), centroid(), distance() using Boost.Geometry; 
  Implemented convexhull() using Boost.Geometry.
  
  Bug #19351967 CENTROID() AND COVEXHULL() ACCEPTS GEOMETRYCOLLECTION CONTAINING INVALID POLYGON
  Bug #19304320 CENTROID() AND CONVEXHULL() FUNCTIONS RETURN NULL WITH VALID GEOMETRY INPUT
  Bug #19142227 ENVELOPE() SPATIAL ANALYSIS FUNCTION ACCEPTS INVALID POLYGONS AS THE ARGUMENT
  Bug #19133043 REPRESENTATION OF THE SAME POINT GEOMETRY ISSUE WITH SPATIAL ANALYSIS FUNCTIONS
  Bug #19204199 DISTANCE() AND ENVELOPE() FUNCTIONS RETURN WRONG RESULT
  Bug #18919820  CONVEXHULL() FUNCTION PRODUCES INVALID POLYGON AS THE RESULT
  Bug #18957010  CONVEXHULL() FUNCTION RETURNS WRONG RESULT
  Bug #18972363   CENTROID() FUNCTION RETURNS INCORRECT RESULT
  Bug #17371854 ENVELOPE() FUNCTION RESULT INFLUENCED BY THE 1ST ARGUMENT PASSED
  Bug #18935403 SPATIAL ANALYSIS FUNCTIONS NOT RETURNING CONSISTENT RESULTS
  Bug #18911154 CENTROID() RETURNING ERROR MESSAGE FOR VALID GEOMETRY-COLLECTION GEOMETRY
    ------------------------------------------------------------
    revno: 8116.1.52
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-08-21 19:33:36 +0800
    message:
      Add more unit tests to promote coverage and fixed a few bugs caught by them.
    ------------------------------------------------------------
    revno: 8116.1.51 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-08-21 14:49:32 +0800
    message:
      merge trunk rev 8630.
    ------------------------------------------------------------
    revno: 8116.1.50 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-08-20 20:07:49 +0800
    message:
      merge
        ------------------------------------------------------------
        revno: 8116.3.2
        committer: Pavan Naik<pavan.naik@oracle.com>
        branch nick: mysql-trunk-wl7236
        timestamp: Sun 2014-08-17 17:49:10 +0530
        message:
          WL#7236 : Spatial Analysis Functions
          
          Removed unused/unwanted lines from the tests. 
        ------------------------------------------------------------
        revno: 8116.3.1
        committer: Pavan Naik<pavan.naik@oracle.com>
        branch nick: mysql-trunk-wl7236
        timestamp: Sun 2014-08-17 11:46:01 +0530
        message:
          WL#7236 : Spatial Analysis Functions
          
          Added and modified the test cases in the existing tests.
    ------------------------------------------------------------
    revno: 8116.1.49 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-08-20 20:04:07 +0800
    message:
      merge trunk rev 8622.
    ------------------------------------------------------------
    revno: 8116.1.48 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-08-14 10:29:18 +0800
    message:
      merge to trunk rev 8579.
    ------------------------------------------------------------
    revno: 8116.1.47
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-08-14 10:12:11 +0800
    message:
      minor fix
    ------------------------------------------------------------
    revno: 8116.1.46
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-08-06 16:56:49 +0800
    message:
      review comment fixes.
    ------------------------------------------------------------
    revno: 8116.1.45
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-08-06 14:16:24 +0800
    message:
      review comment fixes.
    ------------------------------------------------------------
    revno: 8116.1.44 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-08-04 14:35:22 +0800
    message:
      merge to trunk rev 8515
    ------------------------------------------------------------
    revno: 8116.1.43
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-08-04 14:29:44 +0800
    message:
      Bug #19351967 CENTROID() AND COVEXHULL() ACCEPTS GEOMETRYCOLLECTION CONTAINING INVALID POLYGON
      Check for ring validity for polygons of geometry collections given to centroid, but not for convexhull and envelope, because in the latter two gis functions,
      we don't rely on polygon ring being valid, we can compute the logically correct result for any polygons.
      And also for envelope we don't do so for performance since they will be used by rtree index code of storage engines.
    ------------------------------------------------------------
    revno: 8116.1.42
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-07-30 15:13:16 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Added and modified test cases in the existing tests.
    ------------------------------------------------------------
    revno: 8116.1.41
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-07-30 15:58:21 +0800
    message:
      Bug #19304320 CENTROID() AND CONVEXHULL() FUNCTIONS RETURN NULL WITH VALID GEOMETRY INPUT
      Continue scanning WKB data after passing an empty geometry collection.
    ------------------------------------------------------------
    revno: 8116.1.40
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-07-28 15:06:21 +0800
    message:
      minor compiling fix
    ------------------------------------------------------------
    revno: 8116.1.39
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-07-25 20:52:21 +0800
    message:
      minor compiler fix
    ------------------------------------------------------------
    revno: 8116.1.38
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-07-24 21:59:24 +0800
    message:
      minor change.
    ------------------------------------------------------------
    revno: 8116.1.37
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-07-23 12:17:37 +0800
    message:
      st_distance performance optimization: given a geometry collection,
      compact its points/multipoints into one multipoint object and linestrings/multilinestrings into one multilinestring object
      before calling BG::distance for better performance, because BG::distance has O(NlogN) complexity for calculating distance of such objects.
    ------------------------------------------------------------
    revno: 8116.1.36
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-07-18 12:44:26 +0800
    message:
      set st_distance's maybe_null=true because either operand can be an empty 
      geometry collection and the distance would be unknown; other minor changes.
    ------------------------------------------------------------
    revno: 8116.1.35
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-07-15 11:45:40 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Added test cases for bugs fixed.
    ------------------------------------------------------------
    revno: 8116.1.34
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-07-15 12:13:24 +0800
    message:
      minor code refactoring.
    ------------------------------------------------------------
    revno: 8116.1.33
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-07-14 20:09:55 +0800
    message:
       
      Bug #19142227 ENVELOPE() SPATIAL ANALYSIS FUNCTION ACCEPTS INVALID POLYGONS AS THE ARGUMENT
      Check polygon validity in envelope().
    ------------------------------------------------------------
    revno: 8116.1.32
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-07-14 20:08:40 +0800
    message:
      Bug #19133043 REPRESENTATION OF THE SAME POINT GEOMETRY ISSUE WITH SPATIAL ANALYSIS FUNCTIONS
      When output to WKT, make coordinates of -0 as 0.
    ------------------------------------------------------------
    revno: 8116.1.31
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-07-14 18:14:14 +0800
    message:
      Bug #19204199 DISTANCE() AND ENVELOPE() FUNCTIONS RETURN WRONG RESULT
      Compute distance and envelope correctly given empty geometry collection operands.
      For distance(), if at least one operand is an empty collection, the result is NULL; if an operand is a collection containing empty collections and other valid geometries, the empty collections components are ignored; For envelope(), if a component of a geometry collection is empty, it's ignored, and if a collection is a hirarchy (tree) of empty collections, it's envelope is NULL.
    ------------------------------------------------------------
    revno: 8116.1.30 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-07-14 16:07:31 +0800
    message:
      merge trunk rev 8427.
    ------------------------------------------------------------
    revno: 8116.1.29
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-07-04 10:50:29 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Added test cases for fixed bugs in the tests.
    ------------------------------------------------------------
    revno: 8116.1.28
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-07-02 15:35:15 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Added test cases for Sapatial Analysis functions with and without "ST_" prefix.
    ------------------------------------------------------------
    revno: 8116.1.27 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-07-02 14:51:55 +0800
    message:
      merge to trunk rev 8345.
    ------------------------------------------------------------
    revno: 8116.1.26 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-06-27 16:52:51 +0800
    message:
      merge
        ------------------------------------------------------------
        revno: 8116.2.1
        committer: Pavan Naik<pavan.naik@oracle.com>
        branch nick: mysql-trunk-wl7236
        timestamp: Thu 2014-06-26 12:54:30 +0530
        message:
          WL#7236 : Spatial Analysis Functions
          
          Added new test cases for Centroid() and ConvexHull() functions.
    ------------------------------------------------------------
    revno: 8116.1.25 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-06-27 16:41:08 +0800
    message:
      merge to trunk rev 8306.
    ------------------------------------------------------------
    revno: 8116.1.24
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-06-27 16:38:46 +0800
    message:
      minor fix
    ------------------------------------------------------------
    revno: 8116.1.23
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-06-25 16:54:14 +0800
    message:
      Normalize accumulated multipolygon's ring order before sending it to BG when computing geometry collection centroid.
    ------------------------------------------------------------
    revno: 8116.1.22
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Sat 2014-06-21 15:33:21 +0800
    message:
      resolve/merge test and update result
    ------------------------------------------------------------
    revno: 8116.1.21 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Sat 2014-06-21 15:18:50 +0800
    message:
      merge with trunk rev 8263 complete
    ------------------------------------------------------------
    revno: 8116.1.20
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Sat 2014-06-21 15:09:50 +0800
    message:
      merge to trunk rev 8263
    ------------------------------------------------------------
    revno: 8116.1.19
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Sat 2014-06-21 15:03:01 +0800
    message:
      Bug #18972363   CENTROID() FUNCTION RETURNS INCORRECT RESULT
      Compute centroid for geometry collection using geometries of biggest dimension, 
      group components into a multipolygon, if it's not empty, use it to compute centroid, 
      otherwise filter and build a multilinestring and compute centroid, if it's empty, 
      filter and build a multipoint and compute centroid.
      
      Bug #18919820  CONVEXHULL() FUNCTION PRODUCES INVALID POLYGON AS THE RESULT
      Bug #18957010  CONVEXHULL() FUNCTION RETURNS WRONG RESULT
      The two bugs are caused by the same BG bug, the fix is to add our own patched
      hull_graham_andrew.hpp file
    ------------------------------------------------------------
    revno: 8116.1.18
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-06-17 16:38:29 +0800
    message:
      Bug #17371854 ENVELOPE() FUNCTION RESULT INFLUENCED BY THE 1ST ARGUMENT PASSED
      Make uniform 0 values, store -0 as 0.
      Some other minor fixes.
    ------------------------------------------------------------
    revno: 8116.1.17
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Sat 2014-06-14 11:32:08 +0800
    message:
      A memory leak fix.
    ------------------------------------------------------------
    revno: 8116.1.16
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-06-13 12:27:58 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Removed test cases that were failing on linux 32-bit machine because of overflow issue.
    ------------------------------------------------------------
    revno: 8116.1.15 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-06-13 14:16:56 +0800
    message:
      merge trunk rev 8197
    ------------------------------------------------------------
    revno: 8116.1.14
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-06-13 11:49:46 +0800
    message:
      Make sure Inplace_vector member functions don't throw exceptions.
    ------------------------------------------------------------
    revno: 8116.1.13
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-06-12 13:27:49 +0800
    message:
      split and simplify bg_distance.
    ------------------------------------------------------------
    revno: 8116.1.12
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-06-11 20:46:23 +0800
    message:
      Review comment fixes;
    ------------------------------------------------------------
    revno: 8116.1.11
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-gis
    timestamp: Wed 2014-06-11 11:48:02 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Developed new RQG grammar for spatial analysis functions.
    ------------------------------------------------------------
    revno: 8116.1.10
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Wed 2014-06-11 10:14:58 +0530
    message:
      WL#7236 : Spatial Analysis Functions
            
      Added test cases to the existing tests under "gis" test suite.
    ------------------------------------------------------------
    revno: 8116.1.9
    committer: Pavan Naik<pavan.naik@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-06-10 12:09:29 +0530
    message:
      WL#7236 : Spatial Analysis Functions
      
      Developed new MTR tests for the following spatial analysis functions.
      
      1) Area(geometry)
      2) Centroid(geometry)
      3) ConvexHull(geometry)
      4) Envelope(geometry)
      5) ST_Distance(geometry1,geometry2)
    ------------------------------------------------------------
    revno: 8116.1.8
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-06-09 17:31:54 +0800
    message:
      Bug #18935403 SPATIAL ANALYSIS FUNCTIONS NOT RETURNING CONSISTENT RESULTS
      Take empty geometry collections into account, be prepared for no-results result.
    ------------------------------------------------------------
    revno: 8116.1.7 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-06-09 11:42:57 +0800
    message:
      merge trunk rev 8154
    ------------------------------------------------------------
    revno: 8116.1.6
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Mon 2014-06-09 10:46:15 +0800
    message:
      Don't make st_area positive if it's negative because we always normalize polygon ring orientation, if we get a negative value, it means the polygon's inner ring is larger than its outer ring, and the area has to be negative.
    ------------------------------------------------------------
    revno: 8116.1.5
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Thu 2014-06-05 17:06:33 +0800
    message:
      Bug #18911154 CENTROID() RETURNING ERROR MESSAGE FOR VALID GEOMETRY-COLLECTION GEOMETRY
      Another unrecorded crash bug.
      1. Fixed a memory issue for convexhull(point)
      2. Fixed a WKB scanner bug which caused bug#18911154
      3. Be more robust in linear hull detection for convexhull, avoid returning invalid polygons for geometries that have linear hulls.
    ------------------------------------------------------------
    revno: 8116.1.4
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-06-03 15:07:19 +0800
    message:
      Minor bug fix.
    ------------------------------------------------------------
    revno: 8116.1.3 [merge]
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-06-03 14:03:42 +0800
    message:
      merge trunk rev 8127
    ------------------------------------------------------------
    revno: 8116.1.2
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Tue 2014-06-03 11:37:23 +0800
    message:
      Remove changes of WL#7541 and WL#7420, now this tree only contains changes of WL#7236.
    ------------------------------------------------------------
    revno: 8116.1.1
    committer: David Zhao <david.zhao@oracle.com>
    branch nick: mysql-trunk-wl7236
    timestamp: Fri 2014-05-30 16:50:09 +0800
    message:
      
      WL7236 Spatial Analysis Functions
      WL7541 GIS MBR spatial operations enhancement
      WL7420 Geometry Collection Support
      Adding all latest code for above 3 work logs.
------------------------------------------------------------
revno: 8655
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Sun 2014-08-24 12:59:53 +0200
message:
  WL#2489: Better ONLY_FULL_GROUP_BY mode
  
  Complete patch. See sql/aggregate_check.h or the WL for design.
    
  Moreover, this contains changes (not written in the WL page):
  
  1) Item_ident::alias_name_used used to convey the two pieces of
     information below; it's removed and replaced with:
  
     * m_alias_of_expr tells if this is a reference, through an alias, to a
       SELECTed expression which has "AS alias"
     * to know if this is a reference to a column of an aliased table, we
       consult table->alias_name_used
  
  2) only_full_group_by checks are now done after simplify_joins();
     the latter function may convert outer joins to inner joins, and
     inner joins allow more functional dependency detection;
     so with this change we can allow more queries.
  
  3) moved JOIN::outer_join to SELECT_LEX::outer_join; indeed, this is
     constant through executions, and we now need it for FD detection so
     computing this variable at optimization time was too late.
  
  4) in trunk we eliminate DISTINCT if all group expressions are in the
     select list, but we do it at optimization stage; in this WL we do
     it at preparation stage. This makes only_full_group_by
     distinct-related checks to allow more queries, like
  
       SELECT distinct a, min(b) FROM t1 group by a order by count(*)-count(*);
  
     (which otherwise would yield the new error
     ER_AGGREGATE_IN_ORDER_NOT_SELECT): and it allows the optimizer in
     general to choose better plans (see the change in group_min_max.result).
  
  5) a new class Bool3 for holding true/false/unknown values.
------------------------------------------------------------
revno: 8654 [merge]
committer: Harin Vadodaria <harin.vadodaria@oracle.com>
branch nick: trunk
timestamp: Sat 2014-08-23 09:13:04 +0530
message:
  Bug#19370676 : YASSL PRE-AUTH BUFFER OVERFLOW WHEN CLIENT
                 LIES ABOUT SUITE_LEN_
                 and
  Bug#19355577 : YASSL PRE-AUTH BUFFER OVERFLOW WHEN CLIENT
                 LIES ABOUT COMP_LEN_
  
  Description : merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.338.81 [merge]
    tags: clone-5.6.21-build
    committer: Harin Vadodaria <harin.vadodaria@oracle.com>
    branch nick: 5.6
    timestamp: Sat 2014-08-23 09:04:14 +0530
    message:
      Bug#19370676 : YASSL PRE-AUTH BUFFER OVERFLOW WHEN CLIENT
                     LIES ABOUT SUITE_LEN_
                     and
      Bug#19355577 : YASSL PRE-AUTH BUFFER OVERFLOW WHEN CLIENT
                     LIES ABOUT COMP_LEN_
      
      Description : Merge from 5.5 to 5.6
        ------------------------------------------------------------
        revno: 2875.598.29
        tags: clone-5.5.40-build
        committer: Harin Vadodaria <harin.vadodaria@oracle.com>
        branch nick: 5.5
        timestamp: Sat 2014-08-23 08:59:03 +0530
        message:
          Bug#19370676 : YASSL PRE-AUTH BUFFER OVERFLOW WHEN CLIENT
                         LIES ABOUT SUITE_LEN_
                         and
          Bug#19355577 : YASSL PRE-AUTH BUFFER OVERFLOW WHEN CLIENT
                         LIES ABOUT COMP_LEN_
          
          Description : Updating yaSSL to version 2.3.4.
------------------------------------------------------------
revno: 8653
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Fri 2014-08-22 17:37:11 +0200
message:
  Bug#19459193 GET RID OF DYNAMIC_ARRAY IN CIRCULAR_BUFFER_QUEUE
  
  Post-push fix: gcc 4.4 needs some help doing template deduction.
------------------------------------------------------------
revno: 8652
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Fri 2014-08-22 16:36:59 +0200
message:
  Bug#19459193 GET RID OF DYNAMIC_ARRAY IN CIRCULAR_BUFFER_QUEUE
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
  
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8651
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-likely
timestamp: Fri 2014-08-22 15:44:02 +0200
message:
  Bug #18242233 LIKELY() AND UNLIKELY() CONFLICT WITH BOOST
  
  With the current #define's we are unable to use Boost atomics/atomics/lockfree modules.
  Solution: convert to inline functions for C++
  keep #defines in C to minimize size of patch.
------------------------------------------------------------
revno: 8650
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk-CLEAN
timestamp: Fri 2014-08-22 15:35:40 +0200
message:
  WL#7928 Geohash encoding and decoding functions
  
  Post-push fix: Removed ">= 0" ASSERT on a char variable, which caused compilation failure on ARM.
------------------------------------------------------------
revno: 8649
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: test-trunk
timestamp: Fri 2014-08-22 14:07:41 +0200
message:
  bug#14041454: INNODB.INNODB_BUG-13628249 TEST FAILS SPORADICALLY ON PB2
  
  Post-push fix for failing compile of cluster on windows
  due to minidump code change.
------------------------------------------------------------
revno: 8648 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-08-22 19:56:04 +0800
message:
  Null merge from 5.6
    ------------------------------------------------------------
    revno: 3902.338.80
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Fri 2014-08-22 19:54:02 +0800
    message:
      Revert the push(rev:6120) of
      Bug#19450143 - BUF_READ_PAGE_ASYNC CALLS BUF_READ_PAGE_LOW WITH SYNC=TRUE,
      since some test case failure needs to be fixed.
------------------------------------------------------------
revno: 8647
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-plain
timestamp: Fri 2014-08-22 13:01:54 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Patch #5: This patch does the following:
  - Removes HAVE_CRYPT_H, HAVE_CXXABI_H, HAVE_SYNCH_H, HAVE_SYS_MALLOC_H,
    HAVE_SYS_PRCTL_H, HAVE_ASM_TERMBITS_H, HAVE_TERMCAP_H, HAVE_UTIME_H,
    HAVE_SYS_UTIME_H. The header they check for is not needed.
  - Removes the inclusion of several system header files that is not needed.
  - Moves inclusion of serveral system header files from header files to impl files.
  - Removes serveral unneeded extern "C" blocks.
  - Uses HAVE_xxx checks consistently when including system headers.
  - Removes NEED_EXPLICITY_SYNC_DIR - only used in one file. Replaced with __linux__
  - Changes many timespec macros to inline functions, removes dead code.
  - Adds a missing header guard.
------------------------------------------------------------
revno: 8646
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: base-trunk
timestamp: Fri 2014-08-22 14:13:30 +0530
message:
  WL# 7583 - THD transitional string refactoring.
  This patch does transitional refactoring of string
  types in the THD class. This will ease on the migration to
  the new strings framework. A number of places in the
  server uses char* instead of const char*. A gradual
  transformation of the server code is required to use
  const char* and use const references to pass objects
  around functions. This will enable the smooth transition
  of the server code to use the new string framework classes.
      
  This changeset is responsible changing the argument of mysql_rm_db
  to a const LEX_CSTRING reference.
------------------------------------------------------------
revno: 8645
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: base-trunk
timestamp: Fri 2014-08-22 14:07:31 +0530
message:
  WL# 7583 - THD transitional string refactoring.
  This patch does transitional refactoring of string
  types in the THD class. This will ease on the migration to
  the new strings framework. A number of places in the
  server uses char* instead of const char*. A gradual
  transformation of the server code is required to use
  const char* and use const references to pass objects
  around functions. This will enable the smooth transition
  of the server code to use the new string framework classes.
      
  This changeset is responsible for changing the type of
  db name of the Statement class from a raw char* to
  LEX_CSTRING. This results in set of cascade changes across
  several files. The following data members are modified as
  part of this patch:
        1. sp_name
             LEX_CSTRING m_db
        2. Alter_table_ctx
            const char *db
            const char *table_name
            const char *alias
            const char *new_db
            const char *new_alias
        3. Foreign_key
            LEX_CSTRING ref_db
            LEX_CSTRING ref_table
        4. Statement
            LEX_CSTRING m_db
        5. Table_ident
            LEX_CSTRING db
            LEX_CSTRING table
        6. TABLE_LIST
            const char *db, *table_name, *alias
            LEX_CSTRING view_db
            LEX_CSTRING view_name
        7. Trigger
            LEX_CSTRING m_db_name
            LEX_CSTRING m_subject_table_name
        8. user_var_entry
        LEX_CSTRING m_catalog
      
  The functions which are modified as part of this patch include:
        1. create_embedded_thd
        2. check_embedded_connection
        3. acl_getroot
        4. check_routine_access
        5. acl_authenticate
        6. check_access
        7. mysql_table_grant
        8. get_table_grant
        9. acl_notify_htons
        10. log_loaded_block
        11. decide_logging_format
        12. net_after_header_psi
        13. Event_job_data::execute
        14. Events::update_event
        15. Events::drop_schema_events
        16. filesort
        17. write_schema_op_to_binlog
        18. injectApplyStatusWriteRow
        19. ndb_binlog_thread_func
        20. handler::ha_external_lock
        21. Item_field::fix_fields
        23. Item_func_sp::itemize
        24. Item_func_database::val_str
        25. File_query_log::write_slow
        26. set_thd_db
        27. Query_log_event::Query_log_event
        28. Query_log_event::do_apply_event
        29. Load_log_event::do_apply_event
        30. Log_event::get_db
        31. handle_slave_io
        32. handle_slave_sql
        33. Current_schema_tracker::store
        34. db_load_routine
        35. lock_db_routines
        36. sp_drop_db_routines
        37. sp_load_for_information_schema
        38. sp_head::execute
        39. sp_head::execute_trigger
        40. sp_instr_stmt::exec_core
        41. mysql_admin_table
        42. Alter_table_ctx::Alter_table_ctx
        43. MYSQL_AUDIT_NOTIFY_CONNECTION_CONNECT (macro)
        44. MYSQL_AUDIT_NOTIFY_CONNECTION_CHANGE_USER (macro)
        45. list_open_tables
        46. open_table
        47. make_cache_key
        48. Query_cache::invalidate
        49. THD::~THD
        50. thd_binlog_filter_ok
        51. db
        52. set_db
        53. reset_db
        54. copy_db_to
        55. Table_ident::Table_ident
        56. Table_ident::change_db
        57. mysql_open_cursor
        58. mysql_change_db_impl
        59. write_to_binlog
        60. get_default_db_collation
        61. mysql_create_db
        62. mysql_alter_db
        63. mysql_rm_db
        64. mysql_change_db_impl
        65. backup_current_db_name
        66. mysql_change_db
        67. mysql_opt_change_db
        68. mysql_upgrade_db
        69. mysql_derived_prepare
        70. Sql_cmd_handler_open::execute
        71. db_is_default_db
        72. mysql_load
        73. write_execute_load_query_log_event
        74. mysql_load
        75. all_tables_not_ok
        76. dispatch_command
        77. mysql_execute_command
        78. mysql_parse
        79. add_table_to_list
        80. plugin_load
        81. Prepared_statement::set_db
        82. Prepared_statement::prepare
        83. Prepared_statement::reprepare
        84. Prepared_statement::swap_prepared_statement
        85. Prepared_statement::execute
        86. mysql_rename_tables
        87. do_rename
        88. close_cached_connection_tables
        89. mysqld_show_create_db
        90. store_create_info
        91. view_store_create_info
        92. List_process_list::operator()
        93. Fill_process_list::operator()
        94. make_table_list
        95. fill_schema_table_by_open
        96. get_all_tables
        97. make_schema_select
        98. get_trigger_table
        99. initialize_information_schema_acl
        100. mysql_rm_table_no_locks
        101. mysql_alter_table
        102. add_table_for_trigger
        103. udf_init
        104. mysql_create_view
        105. mysql_make_view
        106. mysql_drop_view
        107. get_table_category
        108. open_table_def
        109. TABLE_LIST::prepare_view_securety_context
        110. TABLE_LIST::prepare_security
        111. init_one_table
        112. new_nested_join
        113. TABLE_LIST::get_db_name
        114. TABLE_LIST::get_table_name
        115. Table_trigger_dispatcher::Table_trigger_dispatcher
        116. Trigger::create_from_dd
        117. Trigger::execute
        118. Trigger::parse
        119. Trigger::get_db_name
        120. Trigger::get_subject_table_name
        121. Trigger_creation_ctx::create
        122. Trigger_loader::load_triggers
        123. Trigger_loader::drop_all_triggers
        124. my_tz_init
        125. initialize_performance_schema_acl
        126. ParserTest::parse
        127. ACL_internal_schema_registry::register_schema
        128. mysql_routine_grant
        129. Log_to_csv_event_handler::log_slow
        130. sp_exist_routines
        131. create_file
        132. Statement::Statement
        133. end_connection
        134. prepare_new_connection_state
        135. mysql_ha_hash_get_key
        136. init_lex_with_single_table
        137. innobase_get_foreign_key_info
        138. ha_myisammrg::append_create_info
        139. initialize_peformance_schema_acl.
  A new helper function to_lex_cstring(const char*) has been
  introduced which converts a const char* null-terminated 
  string to LEX_CSTRING type.
------------------------------------------------------------
revno: 8644
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: base-trunk
timestamp: Fri 2014-08-22 14:06:53 +0530
message:
  WL# 7583 - THD transitional string refactoring.
  This patch does transitional refactoring of string types
  in the THD class. This will ease on the migration to the 
  new strings framework. A number of places in the server
  uses char* instead of const char*. A gradual transformation
  of the server code is required to use const char* and use
  const references to pass objects around functions. This
  will enable the smooth transition of the server code to 
  use the string framework classes.
              
  This changeset is responsible for changing the name of
  the prepared statement to type LEX_CSTRING and makes it 
  a priavate member. Also adhoc functions to_lex_cstring
  and to_lex_string have been introduced in sql_string.h. 
  The following data members are modified as part
  of this patch.
     1. Prepared_statement
     	LEX_CSTRING m_name.
     2. struct LEX
     	LEX_CSTRING prepared_stmt_name.
  The functions which are modified as part of this patch include:
     1. to_lex_cstring
     2. to_lex_string
     3. Prepared_statement::statement
     4. get_stmt_name_hash_key
     5. Prepared_statement::set_name
     6. Prepared_statement::name
     7. Prepared_statement::find_by_name
     8. Prepared_statement::set_name
     9. mysql_sql_stmt_prepare
     10. mysql_sql_stmt_execute
     11. mysql_sql_stmt_close
     12. Prepared_statement::reprepare
     13. Prepared_statement::swap_prepared_statement
------------------------------------------------------------
revno: 8643
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: base-trunk
timestamp: Fri 2014-08-22 14:03:52 +0530
message:
  WL# 7583 - THD transitional string refactoring.
  This patch does transitional refactoring of string types
  in the THD class. This will ease on the migration to the
  new strings framework. A number of places in the server
  uses char* instead of const char*. A gradual transformation
  of the server code is required to use const char* and use
  const references to pass objects around functions. This
  will enable the smooth transition of the server code to use
  the new string framework classes.
              
  This changeset is responsible for changing the catalog
  member of the THD class to LEX_CSTRING. The following 
  data members are modified as part of this patch:
        1. THD (sql/sql_class.h)
             LEX_CSTRING m_catalog
  
  The functions which are modified as part of this patch include: 
        1. Query_log_event::Query_log_event
        2. Query_log_event::do_apply_event
        3. Load_log_event::do_apply_event
        4. handle_slave_sql
        5. THD::THD
------------------------------------------------------------
revno: 8642
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: base-trunk
timestamp: Fri 2014-08-22 14:01:14 +0530
message:
  WL# 7583 - THD transitional string refactoring.
  This patch does transitional refactoring of string
  types in the THD. This will ease on the migration to
  the new strings framework. A number of places in the
  server uses char* instead of const char*. A gradual
  transformation of the server code is required to use
  const char* and use const references to pass objects
  around functions. This will enable the smooth transition
  of the server code to use the new string framework classes.
    
  The following data members are modified as part of this
  patch:
      1. QUERY_PROFILE (sql/sql_profile.h)
          LEX_STRING m_query_source
      2. PROF_MEASUREMENT (sql/sql_profile.h)
            const char *function
            const char *file
            const char *status
          
  The functions which are modified as part of this patch include:
      1. QUERY_PROFILE::QUERY_PROFILE
      2. QUERY_PROFILE::~QUERY_PROFILE
      3. QUERY_PROFILE::set_query_source
      4. PROFILING::finish_current_query
      5. PROFILING::show_profiles
------------------------------------------------------------
revno: 8641
committer: Pavan Naik<pavan.naik@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-22 13:38:01 +0530
message:
  Disable i_main.audit_general_log.test 
------------------------------------------------------------
revno: 8640
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Fri 2014-08-22 09:31:46 +0200
message:
  Bug#19187034: ADD COMPILER VERSION CHECKS TO CMAKE
  
  Post-push fix: Rename CMake variable used to hold result of SunStudio
  version check to avoid conflict with variable used in mysql_version.cmake
------------------------------------------------------------
revno: 8639
committer: Pavan Naik<pavan.naik@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-22 11:54:55 +0530
message:
  Disable main.audit_plugin (BUG#19463867) and i_main.audit_general_log (BUG#19463877) tests, since they are failing frequently on PB2.
  
  Make rpl.rpl_gtid_server_sighup experimental (BUG#19470658).
  
  Approved by Anitha
------------------------------------------------------------
revno: 8638
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-22 07:34:36 +0200
message:
  Make innodb.log_file experimental on Windows, cf. Bug#19363615 
  
  Approved by Anitha over IM.
------------------------------------------------------------
revno: 8637 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Thu 2014-08-21 17:24:48 +0200
message:
  Bug#18928848 II. MALLOC OF UNINITIALIZED MEMORY SIZE
  
  merge 5.6 => 5.7
    ------------------------------------------------------------
    revno: 3902.338.79 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Thu 2014-08-21 17:05:57 +0200
    message:
      Bug#18928848 II. MALLOC OF UNINITIALIZED MEMORY SIZE
      
      Merge 5.5=> 5.6
      fix merge conflicts
        ------------------------------------------------------------
        revno: 2875.598.28
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-08-21 16:42:04 +0200
        message:
          Bug#18928848 II. MALLOC OF UNINITIALIZED MEMORY SIZE
          
          Several string functions have optimizations for constant
          sub-expressions which lead to setting max_length == 0.
          
          For subqueries, where we need a temporary table to holde the result,
          we need to ensure that we use a VARCHAR(0) column rather than a
          CHAR(0) column when such expressions take part in grouping.
          With CHAR(0) end_update() may write garbage into the next field.
------------------------------------------------------------
revno: 8636 [merge]
committer: vamsikrishna.bhagi@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-21 20:00:53 +0530
message:
  WL#7717 Expose the SSL library encryption functions to SQL
  
  merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.78
    committer: vamsikrishna.bhagi@oracle.com
    branch nick: mysql-5.6
    timestamp: Thu 2014-08-21 19:42:31 +0530
    message:
      WL#7717 Expose the SSL library encryption functions to SQL
      
      SQL application developers may benefit from using encryption
      and other cryptographic algorithms available in the
      SSL library. We currently provide SQL functions to a limited
      subset of these (DES and AES symmetric encryption to be
      precise). This worklog is to extend that list through a SQL
      user defined functions plugin.
      
      Algorithms supported: RSA, DSA, DH.
------------------------------------------------------------
revno: 8635
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Thu 2014-08-21 13:44:35 +0200
message:
  WL#7540: Move properties from TABLE to TABLE_LIST: table map and number
  
  Post-push fix: Missing updates in NDB code.
------------------------------------------------------------
revno: 8634
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Thu 2014-08-21 12:07:55 +0200
message:
  WL#7540: Move properties from TABLE to TABLE_LIST: table map and number
  
  Move table number and table map from class TABLE to class TABLE_LIST
  and make them private. Introduce getter functions tableno() and map()
  and setter function set_tableno().
  
  Add table_ref pointer to class JOIN_TAB.
  
  Add table_ref pointer to class Item_field.
  
  Replace field "table" in class Item_func_match with "table_ref".
  
  Replace Field pointer in opt_sum_query() with Item_field pointer.
  
  Replace TABLE pointer in Key_use with TABLE_LIST pointer.
  
  Delete the unused field TABLE::used_fields.
  
  Delete the unused field TABLE::used_keys.
  
  Delete the unused field THD::current_tablenr.
------------------------------------------------------------
revno: 8633
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-11764618
timestamp: Thu 2014-08-21 13:19:34 +0400
message:
  Test for bug #11764618 "DEADLOCK WHEN DDL UNDER LOCK TABLES WRITE,
  READ + PREPARE".
  
  This patch adds only test coverage for this bug. The bug itself
  was fixed as a side-effect of WL#6671 "Improve scalability by not
  using thr_lock.c locks for InnoDB tables".
------------------------------------------------------------
revno: 8632
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug19187034
timestamp: Thu 2014-08-21 10:44:24 +0200
message:
  Bug#19187034: ADD COMPILER VERSION CHECKS TO CMAKE
  
  Add CMake checks for minimum version of supported compilers:
    gcc (Linux, Solaris): 4.4
    Sun Studio (Solaris client lib): 12u2
    clang (Mac OS X, FreeBSD): 3.3
  
  A similar check for Visual Studio 2013 already exists.
  
  The patch also adds a new CMake option -DFORCE_UNSPPORTED_COMPILER
  which if set disables the compiler version check.
------------------------------------------------------------
revno: 8631
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-21 13:13:06 +0530
message:
  - WL#6965: Truncate UNDO logs.
    - Explicitly namespace qualified for find function (accidentally missed)
      This is need for code to compile on Sun Studio.
------------------------------------------------------------
revno: 8630
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 19:20:17 +0300
message:
  Followup to vasil.dimov@oracle.com-20140819130324-popsb4fncjc4db71:
  
  Re-throw the exception if an exception is thrown from some of the
  constructor calls when creating an array of objects in
  ut_allocator::new_array(). That is the standard behavior of:
  
  	foo_t*	foo = new foo_t[n];
  	or
  	foo_t*	foo = new(std::nothrow) foo_t[n];
  
  (std::nothrow prevents an exception only if the raw allocation of the memory
  chunk fails, not if some of the constructor calls throws an exception).
------------------------------------------------------------
revno: 8629
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-21 10:06:48 +0530
message:
  - WL#6965: Truncate UNDO logs.
    Updated name of test-cases as per the Test-Case Naming Convention.
------------------------------------------------------------
revno: 8628
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-08-21 10:38:33 +0800
message:
  Bug#19450143 - BUF_READ_PAGE_ASYNC CALLS BUF_READ_PAGE_LOW WITH SYNC=TRUE
  
  A follow-up patch, to fix the test case failure
  
  Approved by Sunny over IM.
------------------------------------------------------------
revno: 8627
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 20:22:39 +0200
message:
  bug#18900198: SHUTDOWN_SERVER FAILS WITH ERROR -2 IN SOME TESTS
  
  Fix for innodb.innodb_bug60196 which had a too small timeout,
  which failed sometimes on slow platforms.
  
  Fixed by removing the hardcoded 10 sec shutdown timeout.
  (The default is 60 s)
  
  Missed the second shutdown_server :(
------------------------------------------------------------
revno: 8626
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 18:55:31 +0200
message:
  Bug #19465696 MYSQL_UPGRADE --HELP IMPROVEMENTS AFTER REFACTORING
  
  After refactoring of mysql_upgrade there are descriptions of options that are out of date. --help option can print default values as well as special options.
  
  Fixed compilation issue.
------------------------------------------------------------
revno: 8625
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 17:09:02 +0200
message:
  Bug #19465696 MYSQL_UPGRADE --HELP IMPROVEMENTS AFTER REFACTORING
  
  After refactoring of mysql_upgrade there are descriptions of options that are out of date. --help option can print default values as well as special options.
  
  Added options sorting by long name, --help first. Updated options descriptions. Printing special options and default values in help.
------------------------------------------------------------
revno: 8624
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 15:50:06 +0300
message:
  Correct incorrect comments abuot innobase_flush_logs.
  
  The function has never made an InnoDB redo log checkpoint,
  which would involve flushing data pages from the buffer pool
  to the file system. It has only ever flushed the redo log buffer
  to the redo log files.
  
  The actual InnoDB function call has been changed a few times
  since the introduction of InnoDB to MySQL in 2000, but the semantics
  never changed.
  
  Approved by Vasil Dimov
------------------------------------------------------------
revno: 8623
committer: Sven Sandberg <sven.sandberg@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-20 14:02:55 +0200
message:
  Bug#19459836: STRANGE ERROR MESSAGE FOR INCOMPATIBLE GTID OPTIONS FOR MYSQLBINLOG
  Background:
  Some combinations of options for mysqlbinlog are invalid.
  Those combinations give various error messages.
  Problem:
  Some of these error messages were ungrammatical, unnecessarily complex,
  or partially wrong.
  Fix:
  Corrected the messages.
------------------------------------------------------------
revno: 8622
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 16:30:57 +0530
message:
  - WL#6965: Truncate UNDO logs.
    Follow-up checkin. Increasing the timeout so that TCs passes on all
    machine (including slow one)
------------------------------------------------------------
revno: 8621
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 12:50:55 +0300
message:
  Followup to vasil.dimov@oracle.com-20140819130324-popsb4fncjc4db71:
  
  Update a forgotten .result file (it needs --big-test to run)
------------------------------------------------------------
revno: 8620
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: b14041454-t
timestamp: Fri 2014-08-08 15:29:51 +0200
message:
  bug#14041454: INNODB.INNODB_BUG-13628249 TEST FAILS SPORADICALLY ON PB2
  
  Added abort if shutdown_server failed in mysqltest.
  
  On windows abort is simulated by creating a minidump file
  if core_file variable is ON. Also included more information
  in the minidump file to help crash analysis on Windows.
  
  Also some minor typing errors fixed in mysqltest.
------------------------------------------------------------
revno: 8619
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 11:49:21 +0300
message:
  Bug#19471743 REMOVE UNUSED CODE RELATED TO BINLOG INFO STORED IN THE INNODB
  TRX_SYS PAGE
  
  Approved by Vasil Dimov.
------------------------------------------------------------
revno: 8618 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 10:05:25 +0200
message:
  Merge from 5.6 => trunk Add my.cnf.d directory for EL7 regular rpm build
    ------------------------------------------------------------
    revno: 3902.338.77 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-08-20 09:55:19 +0200
    message:
      Merge from 5.5. => 5.6  Add my.cnf.d directory for EL7 regular rpm build
        ------------------------------------------------------------
        revno: 2875.598.27 [merge]
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-08-20 09:46:38 +0200
        message:
          Add my.cnf.d to regular rpm for EL7 build
            ------------------------------------------------------------
            revno: 2875.624.7
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Tue 2014-08-12 19:37:49 +0200
            message:
              Corrected typo
            ------------------------------------------------------------
            revno: 2875.624.6
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Tue 2014-08-12 18:55:05 +0200
            message:
              Experimental testing
            ------------------------------------------------------------
            revno: 2875.624.5
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Tue 2014-08-12 18:26:46 +0200
            message:
              Experimental testing for patch
            ------------------------------------------------------------
            revno: 2875.624.4
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Tue 2014-08-12 16:53:31 +0200
            message:
              Added my.cnf.d directory, removed mysql-5.5-libmysqlclient-symbols.patch
            ------------------------------------------------------------
            revno: 2875.624.3
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Tue 2014-08-12 14:32:16 +0200
            message:
              Add patch mysql-5.5-libmysqlclient-symbols.patch for el7
------------------------------------------------------------
revno: 8617
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 10:51:56 +0300
message:
  Remove an obsolete reference to UNIV_LOG_DEBUG.
------------------------------------------------------------
revno: 8616
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 10:51:38 +0300
message:
  Do not unnecessary skip a test.
------------------------------------------------------------
revno: 8615 [merge]
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 13:17:04 +0530
message:
  - WL#6965: Truncate UNDO logs.
    WL aims at truncating undo logs online (without halting server operation).
    
    1. Identify the undo tablespace to truncate based on size threshold.
    2. Take it offline by marking all the associated rsegs as not available
       for allocation.
    3. Wait for rseg to get free (let the active transaction complete and rseg
       data qualify for purge.)
    4. Truncate UNDO tablespace.
    5. Restore the associated rseg active for allocation.
  
    Approved by: Sunny, Shaohua (rb#5680)
    ------------------------------------------------------------
    revno: 8056.1.123
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-08-20 10:04:22 +0530
    message:
      - WL#6965: follow-up fix post mysql-trunk merge
    ------------------------------------------------------------
    revno: 8056.1.122 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-08-20 09:44:41 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.121 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-08-19 08:25:36 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.120
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-08-18 13:01:23 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Recorded test-cases post update of variable name.
    ------------------------------------------------------------
    revno: 8056.1.119 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-08-18 09:36:36 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.118 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-08-13 08:54:10 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.117 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-08-12 10:09:13 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.116 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-08-11 09:22:12 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.115 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-08-08 09:50:19 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.114 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-08-06 09:00:22 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.113
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-08-04 11:07:50 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Renamed innodb_undo_truncate to innodb_undo_log_truncate
          based on review comments to clarify the context of truncate is undo log.
    ------------------------------------------------------------
    revno: 8056.1.112 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-08-04 09:34:39 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.111 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-23 07:15:21 +0530
    message:
      - WL#6965: merge fropm mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.110 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-17 12:41:28 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.109
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-17 12:39:29 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Addressed review comments.
    ------------------------------------------------------------
    revno: 8056.1.108
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-17 12:36:41 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Addressed review comments.
    ------------------------------------------------------------
    revno: 8056.1.107 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-17 07:58:24 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.106 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-16 16:02:47 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.105
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-16 15:46:47 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Addressed review comments.
    ------------------------------------------------------------
    revno: 8056.1.104
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-16 15:35:25 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added test case to cover boundary value scenario for
          innodb_max_undo_log_size.
    ------------------------------------------------------------
    revno: 8056.1.103
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-16 15:24:04 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - increase timeout considering windows slower run.
    ------------------------------------------------------------
    revno: 8056.1.102
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-16 15:13:47 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Increased the limit of innodb_max_undo_log_size from LONGLONG_MAX to
          ULONGLONG_MAX.
    ------------------------------------------------------------
    revno: 8056.1.101 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-16 14:16:00 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.100 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-07-15 14:08:54 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.99 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-07-14 09:47:20 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.98 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-11 08:27:43 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.97 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-10 16:29:57 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.96
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-04 10:33:41 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Trailing space fix.
    ------------------------------------------------------------
    revno: 8056.1.95
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-04 09:56:28 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Corrected name of test-case. Missing suffix _basic for sys_var test-case.
    ------------------------------------------------------------
    revno: 8056.1.94 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-04 09:50:22 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.93
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-04 09:27:20 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Part-3: Addressing review comments.
    ------------------------------------------------------------
    revno: 8056.1.92
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-04 08:22:02 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Post name change and introduction of new variable to control truncate
          of undo logs failed to update sys_vars test cases. Fixed it now.
    ------------------------------------------------------------
    revno: 8056.1.91
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-07-04 07:58:03 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Part-2: Addressing review comment. Revamp code to avoid over-use of class
          instead used namespace.
    ------------------------------------------------------------
    revno: 8056.1.90
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-03 15:30:01 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Part-1 of review comments. Mainly about naming convention change.
    ------------------------------------------------------------
    revno: 8056.1.89
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-03 14:41:15 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Introduce a new variable name innodb_undo_truncate=0/1.
        - If on/1 only then we continue to truncate undo tablespace.
    ------------------------------------------------------------
    revno: 8056.1.88
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-03 14:08:42 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Change the name of from srv_undo_purge_lag to
          srv_purge_rseg_truncate_frequency in order to reflect what it
          exactly does.
      
        - Beside this there is also change in how this value is used.
          - On start undo_trunc cache the truncate frequency.
          - When tablespace is marked for truncate this undo_trunc cached
            value is set to 1 on completion it is restored back to global
            value.
          - Frequency of rseg truncate is
            min(undo_trunc cached frequency, global value)
      
          This ensures when tablespace is marked for truncate we acclerate
          rseg purge in turn truncate action and make the tablespace/rsegs
          active asap.
    ------------------------------------------------------------
    revno: 8056.1.87 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-07-03 08:22:32 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.86
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 11:25:52 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed a case where-in transaction are resurrected and rseg
          is directly assigned to them. In this case too we need to increment
          trx_ref_count.
    ------------------------------------------------------------
    revno: 8056.1.85 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 10:26:46 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.84
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 10:15:48 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixing code as per coding guidelines.
    ------------------------------------------------------------
    revno: 8056.1.83
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 10:02:20 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Removed trailing spaces.
    ------------------------------------------------------------
    revno: 8056.1.82
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 09:56:25 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Moved the common code to inc file for better maintenance.
    ------------------------------------------------------------
    revno: 8056.1.81 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 07:40:51 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.80
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-07-02 07:35:13 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Addressed review comment.
    ------------------------------------------------------------
    revno: 8056.1.79 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-07-01 14:39:53 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.78
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-07-01 14:38:56 +0530
    message:
      - WL#6965: Truncate UNDO logs.
      
        Addressed review comment.
    ------------------------------------------------------------
    revno: 8056.1.77
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-07-01 14:11:36 +0530
    message:
      - WL#6965: Truncate UNDO logs.
      
        Fixed a race condition. Let's understand with an example:
      
        user-thrd: select rseg, trx->rseg = rseg
        purge-thrd: rseg->skip_allocation = true, rseg->curr_size ==1 
        continue with truncate.
        user-thrd: allocate page from trx->rseg and write undo-record.
        purge-thrd: truncate the tablespace.
        Oops! we lost the undo record of active transaction.
      
        To avoid this scenario we introduced a trx_ref_count that
        will keep track of active trx for the assigned rseg.
        While checking if rseg is free for truncate we also ensure it
        doesn't have any active transaction.
    ------------------------------------------------------------
    revno: 8056.1.76
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-07-01 09:24:33 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Addressed review comments.
    ------------------------------------------------------------
    revno: 8056.1.75 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-07-01 08:32:55 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.74
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-30 09:17:49 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Scheduled innodb_undo as part of gcov run by adding the entry to
          default.release.in
    ------------------------------------------------------------
    revno: 8056.1.73
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-30 08:47:58 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Scheduled run of innodb_undo suite as part of mysql-trunk-stage.
    ------------------------------------------------------------
    revno: 8056.1.72 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-30 08:16:01 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.71
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-06-27 15:57:35 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Disable run in valgrind as the test-case relies on output that is timed.
    ------------------------------------------------------------
    revno: 8056.1.70 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-06-27 07:28:34 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.69
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-26 10:39:28 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Some code-cleanup.
    ------------------------------------------------------------
    revno: 8056.1.68 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-26 10:20:35 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.67
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-26 10:16:37 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Some code-cleanup.
    ------------------------------------------------------------
    revno: 8056.1.66 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-26 08:50:18 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.65 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-25 16:27:22 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.64
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-25 16:26:16 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Revert back the optimization of delayed truncate.
        - This has to be atomic operation and we start delaying truncate
          moment we mark tablespace for truncate but if server is shutdown
          before that then metadata can be in-consistent.
          Why inconsistent ?
          - Because we delay only free of segment but besides that there is
            freeing of the pages too. Tracking freeing of pages will add extra
            overhead than helping in general case.
    ------------------------------------------------------------
    revno: 8056.1.63
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-25 15:54:45 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - os_file_get_size now uses os_aio(OS_AIO_SYNC) so that it
          is compatible with all handle open for AIO too.
    ------------------------------------------------------------
    revno: 8056.1.62
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-24 16:44:25 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed the removed_logs bug.
    ------------------------------------------------------------
    revno: 8056.1.61
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-24 16:35:57 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added informative message to indicate truncate completion.
    ------------------------------------------------------------
    revno: 8056.1.60
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-24 16:24:47 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Corrected bug that tried using wrong filehandler for setting
          file size of truncated undo-tablespace.
    ------------------------------------------------------------
    revno: 8056.1.59
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-24 13:10:02 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - If trx-commit is adding undo record to cached segment rseg_history_len
          is incremented too. While truncating this incremnet adjustment has to
          be considered as normal purge will not purge the cached segment.
    ------------------------------------------------------------
    revno: 8056.1.58
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-24 10:06:46 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Advancing to the next rseg iterator improved.
    ------------------------------------------------------------
    revno: 8056.1.57 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-23 16:59:20 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.56
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-23 16:37:00 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - On truncate we set the file size to restore the default file size
          by filling all zero. If file is open in ASYNC MODE then it would not
          be compatible with os_file_set_size and so we need to re-open the same
          file in compatible mode for os_file_set_size. This instance is closed
          immediately. In meantime all new ops on the file are anyway barred
          as file is being truncated.
    ------------------------------------------------------------
    revno: 8056.1.55
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-23 15:08:22 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - On truncate if purge_sys->rseg points a rseg that has been truncated
          then advance the purge_sys->rseg_iter to point to next entry.
    ------------------------------------------------------------
    revno: 8056.1.54
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-23 09:25:31 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Adjusted trx_sys->rseg_history_len on truncate of undo tablespace
          We just keep track of removed logs while after we mark rseg for truncate
          as we use delayed truncated strategy.
    ------------------------------------------------------------
    revno: 8056.1.53 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-23 08:27:35 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.52
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-19 15:22:00 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - insert/update_undo_cached can have multiple elements which means size > 1
          We initially assumed it is limited to 1.
    ------------------------------------------------------------
    revno: 8056.1.51 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-19 12:11:02 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.50
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-19 09:30:48 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed an issue where-in rseg that is being truncated is added to purge
          queue. Before starting truncate we should remove all such rseg instances
          from purge queue.
        - Ignore processing of these rseg is other option but if same rseg get
          re-added immediately after truncate then invalid rseg instance will
          be projected with wrong information.
      
          let's understand this with timeline graph
          t1 -> rseg-1 ... rseg-2
          t2 -> rseg-2 undo-tablespace gets truncated.
          t3 -> rseg-2 is now being processed but last_page_no = FIL_NULL.
          Issue. Solved by removing rseg-2 from the list before truncate.
    ------------------------------------------------------------
    revno: 8056.1.49
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-18 11:03:49 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added safety assert.
    ------------------------------------------------------------
    revno: 8056.1.48
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-18 09:29:19 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed decrement of history size as we now used optimized way
          to free the purge segment if the segment is due for truncate.
    ------------------------------------------------------------
    revno: 8056.1.47 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-18 08:27:29 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.46
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-17 15:59:30 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added safety assert.
        - Once undo tablespace is selected for truncate then switch the
          srv_undo_purge_lag to 1 so that it can be undo purge frequency
          increases helping in accelerating the tablespace truncate.
    ------------------------------------------------------------
    revno: 8056.1.45
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-17 12:35:41 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed situation where-in rseg is added to history list but corresponding
          undo-tablespace is truncated. Note: rseg anyway doesn't have valid data
          else undo-tablespace would never qualify for truncate.
    ------------------------------------------------------------
    revno: 8056.1.44
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-17 11:23:06 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed threshold calculation to base on srv_page_size instead
          of UNIV_PAGE_SIZE_DEF.
    ------------------------------------------------------------
    revno: 8056.1.43 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-16 10:01:13 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.42
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-06-13 15:16:11 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Reverting the test-push that tried to truncate size to UNIV_PAGE_SIZE
          instead of 0.
    ------------------------------------------------------------
    revno: 8056.1.41 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-06-13 10:59:54 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.40
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-06-13 10:41:15 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Test push that will try to set the truncate size to UNIV_PAGE_SIZE
          instead of 0 as probably setting it to 0 is causing issues with
          Write post truncate on Windows.
    ------------------------------------------------------------
    revno: 8056.1.39
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-12 16:23:13 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Use of aligned buffer for writing to DDL log file
          as demanded by Windows for non-buffered file viz. LOG_FILE.
    ------------------------------------------------------------
    revno: 8056.1.38
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-12 11:30:45 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed trailing spaces.
    ------------------------------------------------------------
    revno: 8056.1.37
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-12 11:28:31 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed trailing spaces.
    ------------------------------------------------------------
    revno: 8056.1.36 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-12 11:12:19 +0530
    message:
      - W#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.35 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-12 09:29:30 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.34
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-11 11:34:52 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed debug mode error check to avoid interference with release build.
    ------------------------------------------------------------
    revno: 8056.1.33 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-11 09:25:33 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.32 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-06-05 08:38:43 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.31 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-06-04 13:39:04 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.30 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-06-03 10:51:57 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.29 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-06-02 16:51:01 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.28 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-05-29 09:12:29 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.27
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-05-29 09:06:36 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Avoid redo logging while fixing undo tablespace truncate.
    ------------------------------------------------------------
    revno: 8056.1.26
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-05-29 08:16:48 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Code cleanup.
        - Fixed condition where-in ddl log will not be removed only after truncate
          of UNDO log is complete (including header initialization).
    ------------------------------------------------------------
    revno: 8056.1.25
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-28 16:14:50 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Another flavor of multiple-client test-case with default rollback segments.
    ------------------------------------------------------------
    revno: 8056.1.24
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-28 16:11:01 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added test-case to cover multiple clients + multiple undo tablespace
    ------------------------------------------------------------
    revno: 8056.1.23
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-28 14:38:41 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added Test-Case to test crash scenario for truncate of undo tablespace.
    ------------------------------------------------------------
    revno: 8056.1.22
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 18:29:01 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed the issue where-in size of rseg could be 3 if both
          update and insert segments are cached.
    ------------------------------------------------------------
    revno: 8056.1.21
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 17:46:46 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Removed some unneeded action as the tablespace is being created new.
    ------------------------------------------------------------
    revno: 8056.1.20 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 14:45:52 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.19
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 14:16:15 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added needed error condition.
    ------------------------------------------------------------
    revno: 8056.1.18
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 11:04:07 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added new configuration variable related sys_var test.
        - Based on discussion adjusting the default value of
          innodb_max_undo_log_size = 1GB.
    ------------------------------------------------------------
    revno: 8056.1.17
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 10:40:07 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Introduced srv_undo_purge_lag that controls rate at which undo logs
          should be purged. Value = N suggest purge undo logs on every Nth iteration
          of purge invocation.
    ------------------------------------------------------------
    revno: 8056.1.16
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Tue 2014-05-27 09:27:47 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - If UNDO tablespace is scheduled for truncate and rseg undo segments
          qualifies for free operation then no point in doing so as anyway
          segment will not be re-used for new booting transaction instead
          we would free it in due-course.
    ------------------------------------------------------------
    revno: 8056.1.15
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-05-26 12:20:07 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - On crash, if the ddl log exist and it has magic number written
          that indicates successful truncate completion. Remove the file
          and continue operation w/o any fixup.
    ------------------------------------------------------------
    revno: 8056.1.14 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-05-26 09:18:29 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.13
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Mon 2014-05-26 08:44:31 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Removed innodb_undo suite from default suite as it needs specialized
          parameters to execute.
        - This suite is now being run with specialized parameters.
    ------------------------------------------------------------
    revno: 8056.1.12
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-05-23 16:15:58 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added entries to daily and weekly run to schedule run of innodb_undo suite
          with undo-tablespace configured
      
        Approved by: Anitha (over IM)
    ------------------------------------------------------------
    revno: 8056.1.11
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-05-23 15:11:06 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added special push enviornment that will create undo tablespace
          during bootstrap.
    ------------------------------------------------------------
    revno: 8056.1.10 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-05-23 13:32:04 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.9
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-05-23 12:39:05 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added logic to fix up the truncate if things crashes. 
    ------------------------------------------------------------
    revno: 8056.1.8
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Fri 2014-05-23 11:15:42 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Removed dbug_crash point as this part of the code can't be tested using
          traditional explicit crash point.
        - Random crash is best.
    ------------------------------------------------------------
    revno: 8056.1.7
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-05-22 16:51:19 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added Test-Case to verify the truncate action.
    ------------------------------------------------------------
    revno: 8056.1.6
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-05-22 14:56:37 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed the naming convention for ddl_log.
        - Fixed the redo logging truned-off issue for truncate ddl_log
    ------------------------------------------------------------
    revno: 8056.1.5
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Thu 2014-05-22 14:30:30 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added logic to write DDL_LOG so that crash during truncate can be
          taken care off.
    ------------------------------------------------------------
    revno: 8056.1.4 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-21 14:52:45 +0530
    message:
      - WL#6965: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 8056.1.3
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-21 14:51:14 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Fixed issue that mis-handled the next scan id setting.
    ------------------------------------------------------------
    revno: 8056.1.2
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-21 14:00:18 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Flush buffer pages and removed them from buffer pool before
          truncating the tablepsace.
    ------------------------------------------------------------
    revno: 8056.1.1
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl6965
    timestamp: Wed 2014-05-21 11:43:49 +0530
    message:
      - WL#6965: Truncate UNDO logs.
        - Added API to truncate tablespace to needed size.
        - Avoid allocation of rseg if tablespace it resides in is marked for truncate.
        - Added logic to mark/select the UNDO tablespace to truncate and then
          truncate it once it has been verified that all the rsegs residing in it are
          free.
------------------------------------------------------------
revno: 8614
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-20 05:27:43 +0200
message:
  Bug#19274972: RESULT DIFF SEEN FOR QUERIES WITH DISTINCT AND ORDER BY
  It's a bug introduced by wl6711 on unique_constraint implementation.
  
  When unique_constriant is used, Hash value is treated to compare whether
  two tuples are equal or not. However, Hash function can't assure the
  uniqueness for different tuples. So we need to compare the content if
  there is the same Hash value. During compare the content, there is an
  error for comparing length in function cmp_field_value.
------------------------------------------------------------
revno: 8613 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-08-20 10:10:01 +0800
message:
  Bug#19450143 - BUF_READ_PAGE_ASYNC CALLS BUF_READ_PAGE_LOW WITH SYNC=TRUE
  
  We should pass sync=false so that we use the asynchronous aio.
  
  Approved by Sunny over IM.
    ------------------------------------------------------------
    revno: 3902.338.76
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Wed 2014-08-20 09:44:01 +0800
    message:
      Bug#19450143 - BUF_READ_PAGE_ASYNC CALLS BUF_READ_PAGE_LOW WITH SYNC=TRUE
      
      We should pass sync=false so that we use the asynchronous aio.
      
      Approved by Sunny over IM.
------------------------------------------------------------
revno: 8612
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt2
timestamp: Tue 2014-08-19 23:45:13 +0200
message:
  n tests: code necessary to test WL#6835;
    part 1 of fix to bug#18975974;
    updates to bug signatures and other small updates
------------------------------------------------------------
revno: 8611
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-cleanup
timestamp: Tue 2014-08-19 18:08:59 +0200
message:
  Merge cleanup
------------------------------------------------------------
revno: 8610
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-cleanup
timestamp: Tue 2014-08-19 16:25:49 +0200
message:
  WL#7777 Performance schema memory instrumentation for innodb
  
  Post merge fix, reduce memory consumption by default.
------------------------------------------------------------
revno: 8609
committer: Shivji Jha <shivji.jha@oracle.com>
branch nick: 7817
timestamp: Tue 2014-08-19 19:39:31 +0530
message:
  WL#7817- RPL Monitoring: Move status variables to                               
           replication P_S tables                                               
                                                                                  
  Added more fields to replication P_S tables.
  Following variables that were a part of
  SHOW STATUS have now been added:                                                 
                                                                                                                      
  Show status like 'Slave_last_heartbeat';                                  
  Show status like 'Slave_received_heartbeats';                                   
  show status like 'Slave_heartbeat_period';
  Show status like 'Slave_retried_transactions';
------------------------------------------------------------
revno: 8608 [merge]
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-19 16:03:24 +0300
message:
  Merge mysql-trunk-wl7777 -> mysql-trunk
  
  WL#7777 Integrate PFS memory instrumentation with InnoDB
  
  Account all memory allocations in InnoDB via PFS using the interface
  provided by "WL#3249 PERFORMANCE SCHEMA, Instrument memory usage".
  
  Approved by:	Yasufumi, Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.112 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-19 16:02:05 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.111 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-19 08:43:25 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.110
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-08-18 20:42:58 +0300
    message:
      Remove unused ut_bh* functions and types.
    ------------------------------------------------------------
    revno: 8114.1.109
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-08-18 20:08:35 +0300
    message:
      Remove the unused functions:
      ib_ut_allocator_create()
      ib_ut_allocator_free()
      ib_ut_malloc()
      ib_ut_resize()
      ib_ut_free()
    ------------------------------------------------------------
    revno: 8114.1.108
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-08-18 19:47:10 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Remove automatic PFS memory accounting event names for files that
      do not contain calls to malloc-like functions.
      
      Those were harmless but would always show up with count_alloc == 0 in
      PFS tables until somebody adds a malloc-like call in those files. This
      upsets QA (Tarique Saleem) and causes unnecessary long discussion
      around itself.
      
      Now, when those are removed, if somebody adds a malloc-like call without
      an explicit key to some of those files, the accounting will appear under
      'memory/innodb/other'.
    ------------------------------------------------------------
    revno: 8114.1.107
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-08-14 09:02:54 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Don't fill up bg stats' recalc_pool with zeroes (recalc_pool_t::value_type())
      when creating it because that will be treated later like valid table_ids and
      will be processed, wasting lots of time.
      
      Instead use the reserve() method for pre-allocating some entries,
      like before.
    ------------------------------------------------------------
    revno: 8114.1.106 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-08-14 08:05:35 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.105
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-08-14 08:01:21 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Update threads_innodb.result after registering two more
      InnoDB threads with PFS.
    ------------------------------------------------------------
    revno: 8114.1.104
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-08-14 08:00:58 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Instrument all memory allocated by std::* containers by providing
      ut_allocator to them.
    ------------------------------------------------------------
    revno: 8114.1.103
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-08-13 21:14:03 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move OUT_OF_MEMORY_MSG from ut0mem.h to ut0new.h since it is only used
      in ut0new.h.
    ------------------------------------------------------------
    revno: 8114.1.102
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-08-13 17:17:06 +0300
    message:
      Remove unused "struct heap_alloc_t".
    ------------------------------------------------------------
    revno: 8114.1.101
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-08-13 16:56:10 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Avoid compilation error:
      
      /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/ut0new.h:217:7: error: cannot define the
            implicit copy assignment operator for 'ut_allocator<std::_Rb_tree_node<dict_foreign_t *> >', because
            non-static const member 'm_key' can't use copy assignment operator
      
      when ut_allocator is used in a constructs like:
      
      typedef std::set<
              dict_foreign_t*,
              dict_foreign_compare,
              ut_allocator<dict_foreign_t*> > dict_foreign_set;
    ------------------------------------------------------------
    revno: 8114.1.100 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 12:11:41 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.99 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 12:11:03 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.98 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 12:02:32 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.97
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 11:59:01 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      PFS instrument the background statistics' recalc_pool.
    ------------------------------------------------------------
    revno: 8114.1.96
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 11:57:03 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Register the dict_stats thread for PFS accounting.
    ------------------------------------------------------------
    revno: 8114.1.95
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 11:15:50 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Register the buf_dump thread for PFS accounting.
    ------------------------------------------------------------
    revno: 8114.1.94
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-08-12 11:14:01 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Non-functional change: sort alphabetically all_innodb_threads[]
      and mysql_pfs_key_t keys.
    ------------------------------------------------------------
    revno: 8114.1.93 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-08 16:32:37 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.92
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-08 16:29:05 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Handle the count_alloc==0 for 'memory/innodb/other' check when PFS is disabled
      in mysql-test/suite/innodb/include/check_pfs_mem_other_is_zero.inc.
    ------------------------------------------------------------
    revno: 8114.1.91
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-08 15:31:46 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation of ut0new-t.c unit test when PFS is disabled.
    ------------------------------------------------------------
    revno: 8114.1.90 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-08 14:17:34 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.89
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-08 14:04:46 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Extend the ut0new unit test with some edge cases and also
      swap the arguments to EXPECT_EQ(). It should be:
      EXPECT_EQ(expected, actual)
    ------------------------------------------------------------
    revno: 8114.1.88 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-08-06 20:18:52 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.87 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-01 18:21:25 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.86
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-08-01 13:15:37 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Adjust mysql-test/r/mysqld--help-win.result after its max value was changed
      from 250 to 1024.
    ------------------------------------------------------------
    revno: 8114.1.85
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-31 11:07:38 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Avoid dynamic heap memory allocation when reading a directory under Windows.
      This messed with the return of GetLastError().
      
      Submitted by:	Mattias (via email) + some cosmetic changes by Vasil
    ------------------------------------------------------------
    revno: 8114.1.84 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 17:32:13 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.83 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 17:00:39 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.82
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 15:16:17 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Update perfschema .result test files after the addition of one more event.
    ------------------------------------------------------------
    revno: 8114.1.81
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 14:57:13 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Add os0event to the list of automatic event_names. UT_NEW_NOKEY() is
      being called from os0event.cc.
      
      Also hook a check that the number of allocations under
      'memory/innodb/other' is zero after 4 of the biggest InnoDB tests.
    ------------------------------------------------------------
    revno: 8114.1.80
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 13:25:40 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Followup to vasil.dimov@oracle.com-20140725165803-3g8u629t2ya3pc9p
      remove an assert that key is not PSI_NOT_INSTRUMENTED because when
      PFS is disabled at runtime then all keys are PSI_NOT_INSTRUMENTED.
    ------------------------------------------------------------
    revno: 8114.1.79
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 13:01:14 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Adjust an assignment to m_key in a constructor of ut_allocator.
      If other.m_key is PSI_NOT_INSTRUMENTED, then m_key would have been assigned
      mem_key_std.
      
      Spotted by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.78
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-30 12:37:54 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move a member variable ut_allocator::m_max_retries to a global
      variable to avoid repetitions of "60" when initializing it.
    ------------------------------------------------------------
    revno: 8114.1.77 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-29 15:24:02 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.76 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-29 15:23:40 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.75
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-29 14:54:53 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Add clarifying comments inside ut_allocator::reallocate().
    ------------------------------------------------------------
    revno: 8114.1.74
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-28 17:32:53 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation when UNIV_PFS_MEMORY is not defined.
    ------------------------------------------------------------
    revno: 8114.1.73
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-28 17:11:34 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Change argument type of deallocate_large() from void* to pointer.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.72
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-28 11:07:55 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      In ut_allocator::new_array() handle the case when the first constructor
      we call throws an exception.
      
      Spotted by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.71
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 20:15:46 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Join two subsequent "if" conditions with identical body.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.70
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 19:58:03 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Add an assert that all callers of ut_allocator::allocate_large() have
      set the memory accounting key when constructing the ut_allocator object.
    ------------------------------------------------------------
    revno: 8114.1.69
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 19:44:21 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Elaborate the comments for ut_new_pfx_t and allocate_large().
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.68
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 19:33:36 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move ut_new_basename_noext() from ut0new to ut0ut and rename it
      to ut_basename_noext() since it is a generic function, not related
      to ut0new.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.67
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 18:55:17 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Rename the auto_names[] variable to auto_event_names[].
    ------------------------------------------------------------
    revno: 8114.1.66
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 16:51:16 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move some code from ut0new.h to ut0new.cc
    ------------------------------------------------------------
    revno: 8114.1.65 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 16:41:39 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.64 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 16:37:16 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.63 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 16:34:01 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.62
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-25 12:23:24 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation under Windows - #include <algorithm>, needed
      for std::min.
    ------------------------------------------------------------
    revno: 8114.1.61
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-24 14:18:50 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Refactor the alloc-retry loop so that we do not sleep after the last
      unsuccessful alloc attempt (assuming alloc keeps failing).
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.60
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-24 13:23:32 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Join variable declaration and assignment.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.59
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-24 12:38:31 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Change argument type of ut_allocator::deallocate() from 'void*' to 'pointer'.
      This is also as per the standard:
      http://en.cppreference.com/w/cpp/memory/allocator/deallocate
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.58
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-24 12:07:14 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Isolate the code that extracts the basename of a file without the
      extension into a dedicated function.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.57
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-24 11:19:19 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Make the global variable mem_keys_auto private/static within ut0new.cc.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.56 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-23 18:54:27 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.55 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-23 18:19:18 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.54 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-23 18:02:05 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.53
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 22:18:03 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Do not use UT_NEW() and UT_DELETE() for managing handler_ctx
      because that object may sometimes be freed by the upper layer using
      Sql_alloc's "delete", which is different from UT_DELETE().
    ------------------------------------------------------------
    revno: 8114.1.52
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 16:09:03 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Make the assignment operator of ut_allocator private.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.51
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 14:22:57 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Change the return type of ut_allocator::reallocate() to 'pointer'.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.50
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 14:13:51 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Add unit tests for the code in ut0new.h
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.49
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 11:43:32 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Also count sizeof(ut_new_pfx_t) when estimating the maximum allocatable
      memory by ut_allocator when UNIV_PFS_MEMORY is defined.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.48
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 11:35:22 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Extend a comment.
    ------------------------------------------------------------
    revno: 8114.1.47
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 11:08:10 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation with UNIV_PFS_MEMORY disabled.
      
      In file included from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/mem0mem.ic:26:0,
                       from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/mem0mem.h:479,
                       from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/dict0mem.h:33,
                       from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/dict0dict.h:32,
                       from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/dict0crea.h:31,
                       from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/handler/handler0alter.cc:31:
      /bzrroot/server/mysql-trunk-wl7777/storage/innobase/handler/handler0alter.cc: In member function 'virtual bool ha_innobase::prepare_inplace_alter_table(TABLE*, Alter_inplace_info*)':
      /bzrroot/server/mysql-trunk-wl7777/storage/innobase/handler/handler0alter.cc:4150:8: error: no matching function for call to 'ha_innobase_inplace_ctx::operator new(sizetype, const std::nothrow_t&)'
             0));
              ^
      /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/ut0new.h:1030:47: note: in definition of macro 'UT_NEW_NOKEY'
       #define UT_NEW_NOKEY(expr)  new(std::nothrow) expr
                                                     ^
      /bzrroot/server/mysql-trunk-wl7777/storage/innobase/handler/handler0alter.cc:4150:8: note: candidates are:
             0));
              ^
      /bzrroot/server/mysql-trunk-wl7777/storage/innobase/include/ut0new.h:1030:47: note: in definition of macro 'UT_NEW_NOKEY'
       #define UT_NEW_NOKEY(expr)  new(std::nothrow) expr
                                                     ^
      In file included from /bzrroot/server/mysql-trunk-wl7777/sql/sql_list.h:18:0,
                       from /bzrroot/server/mysql-trunk-wl7777/sql/sql_plugin.h:46,
                       from /bzrroot/server/mysql-trunk-wl7777/sql/structs.h:23,
                       from /bzrroot/server/mysql-trunk-wl7777/sql/unireg.h:160,
                       from /bzrroot/server/mysql-trunk-wl7777/sql/log.h:19,
                       from /bzrroot/server/mysql-trunk-wl7777/storage/innobase/handler/handler0alter.cc:27:
      /bzrroot/server/mysql-trunk-wl7777/sql/sql_alloc.h:29:16: note: static void* Sql_alloc::operator new(size_t)
         static void *operator new(size_t size) throw ()
                      ^
      /bzrroot/server/mysql-trunk-wl7777/sql/sql_alloc.h:29:16: note:   candidate expects 1 argument, 2 provided
      /bzrroot/server/mysql-trunk-wl7777/sql/sql_alloc.h:39:16: note: static void* Sql_alloc::operator new(size_t, MEM_ROOT*)
         static void *operator new(size_t size, MEM_ROOT *mem_root) throw ()
                      ^
      /bzrroot/server/mysql-trunk-wl7777/sql/sql_alloc.h:39:16: note:   no known conversion for argument 2 from 'const std::nothrow_t' to 'MEM_ROOT* {aka st_mem_root*}'
    ------------------------------------------------------------
    revno: 8114.1.46
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 10:40:01 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move the defintion of the type mem_keys_auto_t and the variable
      mem_keys_auto inside UNIV_PFS_MEMORY, they are not needed if UNIV_PFS_MEMORY
      is not defined.
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.45
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-22 10:28:51 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Make ut_allocator::m_key const
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.44
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-21 16:48:09 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Extend some comments
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.43
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-21 12:33:59 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      * Clarify comment
      
      * Assert that std::map::insert() has succeeded in inserting a new entry
      
      * Increment "retries" in the malloc()-retry loop only on failure
      
      Suggested by:	Annamalai (rb:5845)
    ------------------------------------------------------------
    revno: 8114.1.42 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-21 09:52:42 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.41
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Sat 2014-07-19 10:01:19 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation under Windows - all InnoDB .cc files must include univ.i
    ------------------------------------------------------------
    revno: 8114.1.40
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-18 18:02:56 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Avoid using the GNU C++ extension ({ ... }) by adding one more parameter
      to ut_allocator::allocate(), telling it whether to throw an exception on
      error or to return NULL. If we know that ut_allocator::allocate() will
      not throw an exception, then some simplifications can be done and the usage
      of ({ ... }) avoided.
    ------------------------------------------------------------
    revno: 8114.1.39
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-18 14:05:52 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Followup to:
      vasil.dimov@oracle.com-20140711135727-3jtss2fq6c7diyhb
      adjust mtr result files after lifting the limits of
      performance_schema_max_memory_classes.
    ------------------------------------------------------------
    revno: 8114.1.38
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-17 18:24:08 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation failure:
      .../ut0new.h:707: error: invalid static_cast from type 'size_t*' to type 'ulint*'
    ------------------------------------------------------------
    revno: 8114.1.37
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-17 15:46:10 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Use UT_DELETE_ARRAY() instead of delete[] to free the memory occupied by
      ReadView::ids_t::m_ptr, which is allocated with UT_NEW_NOKEY() in
      ReadView::ids_t::reserve().
    ------------------------------------------------------------
    revno: 8114.1.36
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-17 15:37:31 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Followup to:
      vasil.dimov@oracle.com-20140711135727-3jtss2fq6c7diyhb
      adjust mtr result files after lifting the limits of
      performance_schema_max_memory_classes.
    ------------------------------------------------------------
    revno: 8114.1.35
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-07-17 15:26:17 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation failure:
      .../ut0new.h:707: error: invalid conversion from 'size_t*' to 'ulint*'
      .../ut0new.h:707: error:   initializing argument 1 of 'void* os_mem_alloc_large(ulint*)'
    ------------------------------------------------------------
    revno: 8114.1.34 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-16 19:10:13 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.33
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-16 19:07:02 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Instrument all "new" calls by changing:
      
      new expr -> UT_NEW_NOKEY(expr)
      new type[n] -> UT_NEW_ARRAY_NOKEY(type, n)
      delete ptr -> UT_DELETE(ptr)
      delete[] ptr -> UT_DELETE_ARRAY(ptr)
    ------------------------------------------------------------
    revno: 8114.1.32
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-16 13:15:15 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Change ut_malloc() to take also a PFS key argument.
      Introduce ut_malloc_nokey() as a legacy fallback.
      
      Replace ut_malloc() with ut_malloc_nokey() all over the code.
      
      Same for ut_zalloc().
    ------------------------------------------------------------
    revno: 8114.1.31 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-16 11:38:06 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.30
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-15 09:42:23 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation without PFS enabled.
    ------------------------------------------------------------
    revno: 8114.1.29 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-14 16:38:01 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.28
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-07-14 16:33:53 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Add tracing for memory allocated by os_mem_alloc_large().
    ------------------------------------------------------------
    revno: 8114.1.27
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-11 16:59:08 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Add all file names to the list of auto-names (accounted
      under the file name, instead of a manually specified name)
      from which either ut_malloc() or "new" is called.
    ------------------------------------------------------------
    revno: 8114.1.26
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-11 16:57:27 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Increase performance_schema_max_memory_classes:
      max from 256 to 1024
      default from 250 to 1024
      
      InnoDB will add about 100-200 new entries.
      
      Approved by:	Marc Alff (via email)
    ------------------------------------------------------------
    revno: 8114.1.25
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-11 13:12:08 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Whitespace fixup
    ------------------------------------------------------------
    revno: 8114.1.24
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-07-11 13:10:48 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move the initialization of memory tracing methods before any calling
      any functions that could allocate memory inside InnoDB.
    ------------------------------------------------------------
    revno: 8114.1.23 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:23:36 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.22 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:22:36 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
      
      Resolved trivial conflict due to
      marko.makela@oracle.com-20140703081921-1zughv1o6fr7y6ou
    ------------------------------------------------------------
    revno: 8114.1.21 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:20:44 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.20 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:20:20 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.19 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:19:45 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.18 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:18:48 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.17 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-07-09 18:17:53 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.16 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-01 22:43:21 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.15
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-01 22:39:42 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Move the PFS keys/names that are used in auto tracing (when the caller
      does not provide a key, then we automatically pick one based on the file
      name of the caller) into a separate array to ease management and to
      speed up search (std::map is used instead of a linear array).
    ------------------------------------------------------------
    revno: 8114.1.14
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-07-01 12:31:25 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Extend a comment
    ------------------------------------------------------------
    revno: 8114.1.13
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-06-30 15:17:42 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Implement ut_allocator::reallocator() and use it for ut_realloc().
    ------------------------------------------------------------
    revno: 8114.1.12
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Mon 2014-06-30 11:10:33 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Replace ut_realloc() with a performance schema aware mechanism.
      For now it just does free()+malloc().
      
      Redirect malloc()/realloc()/free() calls from fts0blex.cc and
      fts0tlex.cc to ut_malloc()/ut_realloc()/ut_free().
      
      Manage st_innobase_share::idx_trans_tbl::index_mapping with
      ut_realloc()/ut_free() instead of my_realloc/my_free().
    ------------------------------------------------------------
    revno: 8114.1.11
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Fri 2014-06-27 18:20:21 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix a UT_NEW()/delete mismatch.
    ------------------------------------------------------------
    revno: 8114.1.10 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-06-26 11:42:50 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.9
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-06-26 11:39:48 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Fix compilation without PFS and add a comment on how to use the
      methods on top of ut0new.h
    ------------------------------------------------------------
    revno: 8114.1.8
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-06-25 15:01:32 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Replace ut_malloc(), ut_zalloc() and ut_free() with macros that
      call the corresponding ut_allocator methods.
      
      Fix a malloc()/ut_free() inconsistency introduced in
      marko.makela@oracle.com-20140402200417-qe02djjtf4ro9s15 that uses
      strdup(3) (which uses malloc(3)) to allocate a string in the
      constructor of Datafile but uses ut_free() to destroy the string
      in the destructor.
      
      Replace all ::ut_free() occurences with ut_free() because when PFS is
      disabled at compile time, then our ut_free() macro will expand to ::free().
      
      Replace all usages of strdup(3) with mem_strdup() and use ut_free() to free
      the string.
    ------------------------------------------------------------
    revno: 8114.1.7
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-06-24 12:01:22 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Simplify the code - line number and function name of the caller are not
      used, so remove them as parameters to all methods where they are present.
    ------------------------------------------------------------
    revno: 8114.1.6 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Tue 2014-06-24 11:37:25 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.5 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-06-11 14:21:18 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.4
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-06-11 13:22:45 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Handle ut_allocator::allocate() throwing exceptions from within UT_NEW().
    ------------------------------------------------------------
    revno: 8114.1.3
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Wed 2014-06-11 13:15:48 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Introduce an allocator class to be used for tracing memory allocated
      by std::* containers.
      
      Also move the whole implementation of ut0new in ut0new.h to ease
      maintenance of the mem_key_* variables.
    ------------------------------------------------------------
    revno: 8114.1.2 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Sat 2014-05-31 11:49:08 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl7777
    ------------------------------------------------------------
    revno: 8114.1.1
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl7777
    timestamp: Thu 2014-05-29 20:13:29 +0300
    message:
      WL#7777 Integrate PFS memory instrumentation with InnoDB
      
      Start with a prototype for accounting new-like allocations.
      
      Usage of the provided accounting methods is:
      
      new int;       --replacedby--> UT_NEW(int);
      new Foo;       --replacedby--> UT_NEW(Foo);
      new Foo(args); --replacedby--> UT_NEW(Foo(args));
      delete x;      --replacedby--> UT_DELETE(x);
      new Foo[8];    --replacedby--> UT_NEW_ARRAY(Foo, 8);
      delete[] arr;  --replacedby--> UT_DELETE_ARRAY(arr);
      
      Output in performance_schema looks like this:
      
      mysql> SELECT * FROM memory_summary_global_by_event_name WHERE event_name LIKE '%/innodb/%'\G
      *************************** 1. row ***************************
                        EVENT_NAME: memory/innodb/btr0btr
                       COUNT_ALLOC: 80
                        COUNT_FREE: 80
         SUM_NUMBER_OF_BYTES_ALLOC: 1675
          SUM_NUMBER_OF_BYTES_FREE: 1675
                    LOW_COUNT_USED: 0
                CURRENT_COUNT_USED: 0
                   HIGH_COUNT_USED: 1
          LOW_NUMBER_OF_BYTES_USED: 0
      CURRENT_NUMBER_OF_BYTES_USED: 0
         HIGH_NUMBER_OF_BYTES_USED: 25
      *************************** 2. row ***************************
                        EVENT_NAME: memory/innodb/dict0dict
                       COUNT_ALLOC: 3
                        COUNT_FREE: 0
         SUM_NUMBER_OF_BYTES_ALLOC: 624
          SUM_NUMBER_OF_BYTES_FREE: 0
                    LOW_COUNT_USED: 0
                CURRENT_COUNT_USED: 3
                   HIGH_COUNT_USED: 3
          LOW_NUMBER_OF_BYTES_USED: 0
      CURRENT_NUMBER_OF_BYTES_USED: 624
         HIGH_NUMBER_OF_BYTES_USED: 624
      ...
      *************************** 6. row ***************************
                        EVENT_NAME: memory/innodb/other
                       COUNT_ALLOC: 0
                        COUNT_FREE: 0
         SUM_NUMBER_OF_BYTES_ALLOC: 0
          SUM_NUMBER_OF_BYTES_FREE: 0
                    LOW_COUNT_USED: 0
                CURRENT_COUNT_USED: 0
                   HIGH_COUNT_USED: 0
          LOW_NUMBER_OF_BYTES_USED: 0
      CURRENT_NUMBER_OF_BYTES_USED: 0
         HIGH_NUMBER_OF_BYTES_USED: 0
      6 rows in set (0.00 sec)
------------------------------------------------------------
revno: 8607
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug18920075
timestamp: Tue 2014-08-19 12:15:51 +0200
message:
  Bug#18920075: INNOBASE_MYSQL_PREPARE_PRINT_ARBITRARY_THD BROKEN SINCE 5.0.87
  
  Pre-requisite patch: Get rid of the Statement class.
  
  Move old Statement members to THD and/or Prepared_statement.
  Move old Statement backup/restore logic to sql_prepare.cc.
  Rewrite Statement_map - only Preppared_statement is stored there.
  Remove Query_arena::type(), not needed.
  
  This makes it easier to reason about the copy of query strings
  between Prepared_statement and THD, which is the cause of the bug.
------------------------------------------------------------
revno: 8606
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk
timestamp: Tue 2014-08-19 10:39:08 +0200
message:
  WL#7293: Make sure that when 32-bit connection-IDs roll over
  
  Post-push fix: remove gtest warning:
  [WARNING] gtest-death-test.cc:789:: 
  Death tests use fork(), which is unsafe particularly in a threaded context. 
  For this test, Google Test couldn't detect the number of threads.
------------------------------------------------------------
revno: 8605
committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-19 14:03:48 +0530
message:
  Bug#18509896: 0.00000000042 MOD 10 = 0.42000000000
        
  Problem:
  Mod operation on a decimal with leading zeroes gives
  wrong results.
        
  Analysis:
  In do_div_mod while calculting the destination buffer's length 
  of integer part, precision is presumed to be more than
  the length of fraction always. This presumption is proven false
  when leading zeroes are present in the fractional part as 
  precision is calculated after stripping the leading zeroes.
        
  As a result, when intg0 should have a negative value so that
  the destination buffer would be filled with zeroes later, it is
  right now having a positive value. Hence the change in result.
        
  Solution:
  Calculate the intg0 taking into consideration that prec1
  can be less than frac1.
------------------------------------------------------------
revno: 8604
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk
timestamp: Tue 2014-08-19 10:30:42 +0200
message:
  Bug#18899869 SHA2(X, NULL) YIELDS SHA2(X, 0) INSTEAD OF NULL
  Bug#72856    SHA2(x, NULL) yields SHA2(x, 0) instead of NULL
  
  The documentation for SHA2(str, hash_length) says:
  If either argument is NULL or the hash length is not one of the permitted values, 
  the return value is NULL.
  
  This was not the case for hash_length == NULL or for 
  hash_length == <some illegal value, which becomes legal when cast to uint>
------------------------------------------------------------
revno: 8603
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-19 09:59:06 +0200
message:
  wl6647: Changes (beautification) regarding the code review (rb 5318).
------------------------------------------------------------
revno: 8602 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-08-19 15:11:33 +0800
message:
  Bug#18477009 - INACCURATE HANDLING OF SRV_ACTIVITY_COUNT
  
  We call srv_active_wake_master_thread() directly and one of the places is
  innobase_commit (and prepare as well). This call not only wakes up the
  master thread but also increments the srv_activity_count which tells
  the page_cleaner that server is not idle. That's no what we expect.
  
  We should call srv_active_wake_master_thread() only after the commitment
  of a write trx, but not read-only trx, or after a rollback. This patch also
  changes some call of srv_active_wake_master_thread() to
  ib_wake_master_thread().
  
  Original patch is provided by Inaam.
  
  rb#5909, approved by Jimmy.
    ------------------------------------------------------------
    revno: 3902.338.75
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-19 15:10:06 +0800
    message:
      Bug#18477009 - INACCURATE HANDLING OF SRV_ACTIVITY_COUNT
      
      We call srv_active_wake_master_thread() directly and one of the places is
      innobase_commit (and prepare as well). This call not only wakes up the
      master thread but also increments the srv_activity_count which tells
      the page_cleaner that server is not idle. That's no what we expect.
      
      We should call srv_active_wake_master_thread() only after the commitment
      of a write trx, but not read-only trx, or after a rollback. This patch also
      changes some call of srv_active_wake_master_thread() to
      ib_wake_master_thread().
      
      Original patch is provided by Inaam.
      
      rb#5909, approved by Jimmy.
------------------------------------------------------------
revno: 8601 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug17450876_mysql-trunk
timestamp: Tue 2014-08-19 10:11:39 +0530
message:
  Bug#17450876:REPLICATION STOP WITH "ERROR IN XID_LOG_EVENT:
  COMMIT COULD NOT BE COMPLETED"
  
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.74
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug17450876_mysql-5.6
    timestamp: Tue 2014-08-19 09:57:08 +0530
    message:
      Bug#17450876:REPLICATION STOP WITH "ERROR IN XID_LOG_EVENT:
      COMMIT COULD NOT BE COMPLETED"
      
      Problem:
      ========
      When a SQL thread which is waiting for commit lock is killed
      and restarted it causes a transaction to be skipped on slave.
      
      Analysis:
      ========
      when SQL thread is at a state where a DML is waiting for MDL
      commit lock if SQL thread is killed then position are getting
      updated in memory. i.e in the existing design positions are
      flushed before the actual commit because of this rli object
      will have its positions updated but the transaction is yet
      to be committed.  When the SQL thread is restarted it reads
      position from the rli object and hence the last transaction
      gets skipped on slave.
      
      Fix:
      ===
      When SQL thread is killed at a stage where it is waiting for
      commit lock, the commit fails and an error is reported back
      saying "Commit could not be completed and Query execution
      was interrupted".  As part of fix SQL threads positions that
      existed before the commit are persisted and they are
      restored back on error.
      
      Similar symptoms exist in case of MTS as well. In MTS
      "The slave coordinator and worker threads are stopped,
      possibly leaving data in inconsistent state" error is
      reported. In MTS a bitmap is maintained for successful
      commits. This bit map is cleared on error and the old
      positions are retrieved from the checkpoint which points to
      old positions.
------------------------------------------------------------
revno: 8600
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-19 06:21:01 +0200
message:
  Bug #18883885 INNODB: LATCHING ORDER VIOLATION WHILE INSERTING TO BLOB DATA
  
  Problem:
  
  In the latch order checking code, the following assert failed in SyncDebug::lock():
  
                          ut_a(latches->empty()
                               || latch->m_level == SYNC_LEVEL_VARYING
                               || latch->m_level == SYNC_NO_ORDER_CHECK
                               || latches->back()->m_level == SYNC_NO_ORDER_CHECK
                               || latches->back()->m_level == SYNC_LEVEL_VARYING
                               || latches->back()->m_level >= latch->m_level);
  
  
  Solution:
  
  There were three issues that we identified:
  
  1.  For intrinsic tables, there is no need to take index->lock.  We now avoid
      taking index->lock for intrinsic tables.
  2.  When a SX lock is recursively taken, sync_check_relock() must be called
      instead of sync_check_lock().
  3.  If we take an index lock (SYNC_INDEX_TREE), and then the space
      latch (SYNC_FSP) and we recursively take index lock (SYNC_INDEX_TREE), then
      this is not violating the latch order.
  
  
  rb#6364 approved by Sunny
------------------------------------------------------------
revno: 8599
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Tue 2014-08-19 10:34:58 +0800
message:
  Bug #18127749 OPTIMIZER SHOULD THROW AN ERROR WHEN USER USES COMPARISION OPERATORS ON GIS DATA
  In arithmetic and comparison operators, as well as math functions and
  aggregate functions that operate on numerical data, don't allow operands
  being geometries. At prepare time, check whether the operand's field type
  is MYSQL_TYPE_GEOMETRY, if so, report error.
  Doing so isn't sufficient if all operands are variables, because we don't
  store a variable's value's specific field type.
------------------------------------------------------------
revno: 8598 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push2
timestamp: Mon 2014-08-18 14:49:50 +0200
message:
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 8589.1.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug19419463
    timestamp: Thu 2014-08-14 21:37:12 +0200
    message:
      Bug#19419463 ASSERT FROM->M_BYTE_COUNT == 0 AT SQL_DIGEST_STORAGE::COPY IN SQL_DIGEST.H
      
      Before this bug, a select from table
      performance_schema.events_statements_current
      could cause a failed assert in debug mode.
      
      The assert did fail under load, when reading the digest of a query
      that is concurrently starting.
      
      Note that tables
      - performance_schema.events_statements_history
      - performance_schema.events_statements_history_long
      are not affected by this bug, as only reading from a live query
      can expose the race condition found.
      
      The root cause is that the copy is doing a dirty read by design,
      but the code is not robust to handle data changing unexpectedly.
      
      Fixed sql_digest_storage::copy to be more robust.
------------------------------------------------------------
revno: 8597 [merge]
committer: Sven Sandberg <sven.sandberg@oracle.com>
branch nick: trunk
timestamp: Mon 2014-08-18 14:34:43 +0200
message:
  Merged BUG#18385953 from 5.6 to trunk. No conflicts.
    ------------------------------------------------------------
    revno: 3902.338.73
    committer: Sven Sandberg <sven.sandberg@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-08-18 14:29:06 +0200
    message:
      BUG#18385953 - RPL_GTID_STRESS_FAILOVER FAILS WITH "ERROR IN SYNC_WITH_MASTER.INC"
      
      This test failed sporadically in the cleanup code.
      The cleanup code contained this:
      
        for each server:
          DROP DATABASE
          --source include/stop_slave.inc
          CHANGE MASTER TO MASTER_AUTO_POSITION = 0;
          --source include/start_slave.inc
      
      The problem is that CHANGE MASTER will drop the relay logs.  If the IO
      thread was ahead of the SQL thread at this point, received and
      not-yet-executed transactions in the relay log would be lost.  Since it
      does not use the auto-position protocol when it does the
      start_slave.inc, replication would resume after the lost transaction
      rather than retransmit it.
      
      This caused two types of failures:
       1. The IO thread could be stopped in the middle of a transaction, when
          the SQL thread had processed only up to the last complete transaction.
          Here all transactions consist of two events: Gtid followed by Query.
          Thus, the last received transaction was a Gtid and after
          start_slave.inc the slave would receive a Query. The slave SQL thread
          would then see a Query without Gtid. This looks like an anonymous
          transaction, which is not allowed when GTID_MODE = ON. So the SQL thread
          would stop with error 1782: "Error '@@SESSION.GTID_NEXT cannot be set to
          ANONYMOUS when @@GLOBAL.GTID_MODE = ON.' on query. Default database: 
          'db_3'. Query: 'DROP DATABASE db_3'".
       2. The IO thread could be stopped between transactions, when the SQL
          thread had processed only up to the second-last transaction.
          Then the entire transaction would get lost. Later, in rpl_sync.inc,
          the GTID would not be received at all by the slave, so the sync would
          fail with a timeout.
      
      Fixed by dropping the databases and syncing all servers before stopping
      all the slave threads and executing CHANGE MASTER.
------------------------------------------------------------
revno: 8596
committer: Harin Vadodaria <harin.vadodaria@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-18 17:18:24 +0530
message:
  WL#7699 : Auto-enable SSL for MySQL Enterprise Server
  
  Description : This worklog introduces following
                features/functionality for MySQL
                Enterprise Server.
                1. Two new boolean options:
                auto_generate_certs
                sha256_password_auto_generate_rsa_keys
                Both options are ON by default.
                2. If set to ON, these options check
                for SSL artifacts and RSA key pair
                respectively and generate required
                keys/certificates if they are absent.
------------------------------------------------------------
revno: 8595
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: bug19424197-trunk
timestamp: Mon 2014-08-18 16:39:56 +0530
message:
  Bug#19424197 : MYSQLSLAP --SQL-MODE OPTION OF OUT ORDER IN --HELP
                 MESSAGE
  
  --sql-mode option was added by WL#7764 to mysqlslap tool. It was
  added out of order in --help message. Fixed --sql-mode option position
  to appear in lexical order in the --help message.
------------------------------------------------------------
revno: 8594
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Mon 2014-08-18 11:09:27 +0200
message:
  Bug#19179338 GET RID OF DYNAMIC_ARRAY IN RELAY_LOG_INFO AND DEFERRED_LOG_EVENTS
  Bug#19191355 PB2 A NUMBER OF RPL TESTS CRASHED THE SERVER IN METHOD LOG_EVENT::APPLY_EVENT
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Also: fix a bug introduced by
  WL#6964 MTS: Support SLAVE_TRANSACTION_RETRIES in MTS mode 
  The curr_group_da array now contains Slave_job_item structs,
  rather than Log_event pointers. 
  One part of that refactoring was missed, and we were still inserting
  pointers into the array.
  
  Also: make code more type-safe by changing
  Slave_job_item::data from void* to Log_event*
------------------------------------------------------------
revno: 8593
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-18 12:02:15 +0530
message:
  BUG#17286856 - SOCKET FILE OF A RUNNING MYSQLD INSTANCE CAN
                 BE TAKEN OVER BY ANOTHER ONE.
  PROBLEM DESCRIPTION AND FIX:
  Multiple instances of mysqld can be started specifying the
  same unix socket socket file. When one instance of mysqld is
  shutdown, the unix socket file goes away.
  The fix is to use a lock file for each unique socket file
  name. The lock file contains the pid of the running mysqld
  instance it is attached to. If another instance with the
  same socket filename comes up, the lock file is checked
  and the pid of the process is read and it is checked to
  see if it is running via calling kill with signal 0. If so
  the current instance is not allowed to come up. Thus this
  prevents multiple of mysqld running on the same unix socket
  file. The lock pathname is the name of the unix socket file
  pathname with .lock appended, which by default is /tmp/mysql.
  sock.lock. 
------------------------------------------------------------
revno: 8592
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Fri 2014-08-15 13:58:37 +0200
message:
  Bug#19060850: SQL_YACC.CC REGENERATED TWICE IN WINDOWS BUILD, CAN LEAD TO BUILD BREAK
  
  Run Bison separately for libmysqld so that independent copies of
  sql_yacc.h/sql_yacc.cc are generated. This avoids the Windows
  build break because of libmysqld compilation of sql_yacc.cc while
  it is being generated.
------------------------------------------------------------
revno: 8591
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-15 09:25:55 +0400
message:
  Fixed audit_general_log test failure on Windows.
------------------------------------------------------------
revno: 8590
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-08-15 09:51:12 +0800
message:
  WL#6942 InnoDB GIS: Store GIS POINT datatype as fixed length column than BLOB
  
  We can store POINT as fixed length data(DATA_POINT) or var length data
  (DATA_VAR_POINT). A new session variable called innodb_optimize_point_storage
  is introduced to control the format of POINT we want.
  
  rb#3839, approved by Jimmy and Allen
------------------------------------------------------------
revno: 8589 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 17:49:37 +0100
message:
  Merge of BUG#19311260 patch from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.72
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-08-14 17:44:55 +0100
    message:
      Bug#19311260 ASSERTION IN MTS WHEN WORKER IS KILLED WHILE WAITING FOR
                   COMMIT LOCK
      
      Problem:
      =======
      
      With MTS enabled, when a worker is killed while waiting for commit
      lock, it is causing a debug assert to fail.
      
      Analysis:
      ========
      
      When an error is reported through my_error() function call, the assert
      expects 'thd->get_stmt_da()->m_status' diagnostic variable to be
      'DA_EMPTY', but  when a worker reaches 
      Xid_log_event::do_apply_event_worker, thd->get_stmt_da()->m_status has
      'DA_OK' as value.
      
      In STS version of Xid_log_event applier, because of the above
      explained, the THD object needs to be reset so that the errors, flags
      and diagnostic variables that are set by the previous command are
      cleared.
      
      Fix:
      ===
      
      Added code into MTS version of Xid_log_event "do_apply_event_worker" to
      reset THD for the next command the same way it is done for STS version
      "do_apply_event".
------------------------------------------------------------
revno: 8588
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-gtest
timestamp: Thu 2014-08-14 15:57:56 +0200
message:
  Bug#19303858 -DMAX_INDEXES != 64 CAUSES BUILD FAILURE IN OPT_RANGE-T UNIT TEST
  Bug#73407 -DMAX_INDEXES != 64 causes build failure in opt_range-t unit test
  
  Use the key_map typedef rather than Bitmap<64>
  
  Also:
  nullable-t did not compile with -DMERGE_UNITTESTS=0, nullable.h needs my_dbug.h
  Run dos2unix on nullable-t.cc, and move 'using' inside namespace.
------------------------------------------------------------
revno: 8587
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: test-trunk
timestamp: Thu 2014-08-14 11:41:12 +0200
message:
  bug#18900198: SHUTDOWN_SERVER FAILS WITH ERROR -2 IN SOME TESTS
  
  Fix for innodb.innodb_bug60196 which had a too small timeout,
  which failed sometimes on slow platforms.
  
  Fixed by removing the hardcoded 10 sec shutdown timeout.
  (The default is 60 s)
  
  Approved by Marko on IM.
------------------------------------------------------------
revno: 8586
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Thu 2014-08-14 17:18:21 +0800
message:
  BUG#19433999: TABLESPACE FREE SPACE INFO CORRUPTED - FREEING PAGE ALREADY MARKED FREE
  
  The root cause is that redo logging is skipped for freeing a page in bulk load.
  
  Patch approved by Marko on IM.
------------------------------------------------------------
revno: 8585
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 11:02:07 +0200
message:
  Bug #19435077 MYSQL_UPGRADE MUST HAVE --SECURE-AUTH OPTION DEPRECATED
  
  WL#7308 introduces rewritten mysql_upgrade. Before completion of WL#7308 the WL#8006 was pushed, which makes --secure-auth option in all clients deprecated.
  
  --secure-auth option is made deprecated.
------------------------------------------------------------
revno: 8584
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 10:49:15 +0200
message:
  Bug #19267051 ASSERT FK_TABLES.EMPTY() INNOBASE_UPDATE_FOREIGN_CACHE
  
  Problem:
  
  In the function innobase_update_foreign_cache(), we assert that a call to
  dict_load_foreigns() will successfully complete without needed to load any
  further tables.  But this seems to fail under certain conditions.
  
  Solution:
  
  After the call to dict_load_foreigns(), load all the associated tables to
  ensure that all foreign keys are in the cache.
  
  rb#6292 approved by Jimmy.
------------------------------------------------------------
revno: 8583
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 09:54:52 +0200
message:
  Bug#19142753 ASSERT: MODE !=LOCK_X || LOCK_TABLE_HAS(THR_GET_TRX(THR), INDEX->TABLE, LOCK_IX)
  
  Post-push fix: Use DBUG_RETURN in ha_partition::records() to match DBUG_ENTER.
------------------------------------------------------------
revno: 8582
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 09:14:51 +0300
message:
  Make a test experimental due to frequent failures.
------------------------------------------------------------
revno: 8581
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 08:29:23 +0300
message:
  Second try: Add assertions to help catch Bug#19316315 earlier.
  The first try had misplaced one of the assertions.
------------------------------------------------------------
revno: 8580
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-08-14 08:26:28 +0300
message:
  Remove the experimental status of some tests.
------------------------------------------------------------
revno: 8579
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-deadcode
timestamp: Wed 2014-08-13 18:25:49 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Patch #4: This patch does the following:
  - Removes HAVE_PATHS_H since paths.h is not needed
  - Removes unused HAVE_FP_EXCEPT
  - Moves some compiler related code from my_global.h to my_compiler.h
  - Reorganizes #includes in my_global.h, removes some duplicates
  - Removes definition of _XOPEN_SOURCE, no longer needed
  - Replaces BACKSLASH_MBTAIL and BASKSLASH_MBTAIL (typo...) with _WIN32
  - Removes HAVE_CHSIZE, NO_OPEN_3, SOLARIS and NO_CLIENT_LONGLONG, never defined 
  - Removes workaround for ancient FreeBSD bug, no longer relevant
  - Removes pthread_attr_setstacksize check, function always exists
  - HAVE_FEDISABLEEXCEPT is always true on FreeBSD
  - Removes THD::signals, never used
  - Removes THD::query_start_used, set but never read 
  - Removes checks for old versions of Sun Studio
  - DTrace workarounds for old versions of GCC.
------------------------------------------------------------
revno: 8578
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-winwarn2
timestamp: Wed 2014-08-13 18:05:01 +0200
message:
  Bug#11752665: WINDOWS SERVER X64: SO MANY COMPILER WARNINGS
  
  Patch #11:
  
  This patch reduces the compiler warning count on Win64 by an additional
  80, as reported by VS2013. The patch also removes a number of unnecessary casts.
------------------------------------------------------------
revno: 8577
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-13 17:34:10 +0200
message:
  reverted due to too many test failures:
  Add assertions to help catch Bug#19316315 earlier.
  Approved by Vasil Dimov on IM.
------------------------------------------------------------
revno: 8576
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-13 16:51:11 +0300
message:
  Add assertions to help catch Bug#19316315 earlier.
  Approved by Vasil Dimov on IM.
------------------------------------------------------------
revno: 8575
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug19078013
timestamp: Wed 2014-08-13 15:21:16 +0200
message:
  Bug#19078013: AVOID USING PUSH_WARNING_PRINTF() WITH NO PRINTF ARGUMENTS
  
  This patch adds __attribute__((format(printf... to the push_warning_printf()
  declaration so that any use of this function without any printf argments
  is detected at compile time. In such cases, push_warning() should be 
  used instead.
  
  Note that since the format string is not known at compile time, we cannot 
  detect wrong number or type of arguments.
  
  The patch also changes a number of push_warning_printf() calls to 
  push_warning() where this possibility is now detected by the compiler.
------------------------------------------------------------
revno: 8574
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Wed 2014-08-13 18:29:36 +0800
message:
  Fix Bug #19430876 - INNODB_GIS.INNODB_GIS_RTREE_CONCURRENT_SRCH FAILS SPORADICALLY
  ON TRUNK  
  
  Approved by Marko
------------------------------------------------------------
revno: 8573
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-13 11:50:22 +0300
message:
  Bug#19229231 INNODB.LOG_FILE_NAME FAILS SPORADICALLY ON PB2
  or revisiting
  Bug#11763660 PLEASE ASSERT INSTEAD OF CALLING EXIT(1)
  
  On crash recovery, InnoDB was calling exit() when noticing corruption
  or inconsistency. This call triggered a debug assertion failure
  ut_ad(!m_freed) in a I/O handler thread that was trying to acquire a
  mutex.
  
  A better solution is to return an error to the caller of
  innobase_init() and letting a proper shutdown happen.
  
  innodb.log_corruption: Adjust the excepted message. We no longer
  trigger a crash (assertion failure) when noticing a corrupted log
  record.
  
  fil_op_replay_rename(): Return whether the operation was successfully
  processed. The operation can fail if both the old and new name exist,
  and the old name is for the requested tablespace_id.
  
  recv_sys_t::found_corrupt_fs: A new field, to be set when
  inconsistency between the redo log and the file system is detected and
  manual intervention is needed.
  
  fil_name_process(): Set recv_sys->found_corrupt_fs when needed.
  
  fil_name_parse(): When finding invalid contents inside a MLOG_FILE_*
  record, let the caller know the presumed length of the record.  When
  failing to replay a MLOG_FILE_RENAME operation, set
  recv_sys->found_corrupt_fs.
  
  recv_report_corrupt_log(): Return whether processing should continue.
  (If innodb_force_recovery is set, it may continue.)
  
  recv_parse_log_recs(): Return true also if processing was aborted due
  to corruption. This will let the caller stop the processing.
  
  recv_scan_log_recs(): Check for recv_sys->found_corrupt_fs.
  
  recv_group_scan_log_recs(): If corruption was noticed, return false.
  
  recv_recovery_from_checkpoint_start(): Return an error to the caller
  if recv_sys->found_corrupt_log or recv_sys->found_corrupt_fs is set.
  Add a missing log_mutex_exit() call to apparently uncovered error
  handling code.
  
  innobase_start_or_create_for_mysql(): Remove a check if
  pthread_mutex_trylock() works, and the associated exit() call.
  
  rb#6341 approved by Jimmy Yang
------------------------------------------------------------
revno: 8572
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Wed 2014-08-13 14:48:16 +0800
message:
  Fix Bug #19434871 - RTR_PAGE_CAL_MBR() MISUSES DBL_MIN
  
  
  Approved by Sunny Bains
------------------------------------------------------------
revno: 8571
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 18:50:52 +0300
message:
  Fix Bug#18663997 INNODB: TOO MANY FILES STAY OPEN
  
  Flushing is not done for FIL_TYPE_TEMPORARY files (at the beginning
  of fil_flush()):
  
          if (space == NULL
              || space->purpose == FIL_TYPE_TEMPORARY
              || space->stop_new_ops
              || space->is_being_truncated) {
                  mutex_exit(&fil_system->mutex);
  
                  return;
          }
  
  but FIL_TYPE_TEMPORARY and FIL_TYPE_IMPORT were considered equivalent
  and thus flushing was also not done for FIL_TYPE_IMPORT.
  
  This lead to the problem that modification counter > 0 and
  flush counter == 0 for to-be-imported tablespaces and it was left like
  that after the tablespace got converted to normal FIL_TYPE_TABLESPACE
  in fil_space_set_imported().
  
  Thus, when flushing do not skip the to-be-imported tablespaces.
  
  Approved by:	Marko (rb:6318)
------------------------------------------------------------
revno: 8570
committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-13 10:02:43 +0530
message:
  Bug#12368204: AUDIT PLUGIN API: NO MYSQL_AUDIT_GENERAL_LOG NOTIFICATIONS
                                  WITH GENERAL L
        
        ISSUE: Audit notification for MYSQL_AUDIT_GENERAL_LOG events is not
               generated without general_log enabled. Currently to write 
               a audit_log plugin which logs MYSQL_AUDIT_GENERAL_LOG events
               we have to enable the general log also.
               To log general_log events server checks for general_log handlers. 
               If no handler is found it does provide any notification to audit_log
               plugin.
        
        SOLUTION: Server should provide notification to audit_log plugin regardless
                  of any handlers (general_log enabled) present in the list.
                  Moving out the audit_log function from the check and making sure
                  its always called when a general_log event occurs.
------------------------------------------------------------
revno: 8569
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug19428231
timestamp: Tue 2014-08-12 15:15:32 +0200
message:
  Bug#19428231: WAITING FOR FSYNC REPORTS WRONG WAIT TYPE
  
  Before doing fsync, we report a wait event to currently registered listeners.
  This wait allows e.g. the thread pool to start another thread.
  
  The problem was that the wait type reported for fsync was THD_WAIT_TABLE_LOCK
  and not THD_WAIT_SYNC. Currently this did not have any harmful consequences
  other than possibly wrong I_S stats for the thread pool plugin.
------------------------------------------------------------
revno: 8568
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 15:46:15 +0300
message:
  Enable core files in a test to analyze PB2 failures.
------------------------------------------------------------
revno: 8567 [merge]
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 17:29:36 +0530
message:
  Bug #11755818 : LIKE DOESN'T MATCH WHEN CP932_BIN/SJIS_BIN
                  COLLATIONS ARE USED.
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.71 [merge]
    committer: mithun <mithun.c.y@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-12 17:24:15 +0530
    message:
      Bug #11755818 : LIKE DOESN'T MATCH WHEN CP932_BIN/SJIS_BIN
                      COLLATIONS ARE USED.
      
      Merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.598.26
        committer: mithun <mithun.c.y@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-08-12 17:16:51 +0530
        message:
          Bug #11755818 : LIKE DOESN'T MATCH WHEN CP932_BIN/SJIS_BIN
                          COLLATIONS ARE USED.
          
          ISSUE :
          -------
          Code points of HALF WIDTH KATAKANA in SJIS/CP932 range from
          A1 to DF. In function my_wildcmp_mb_bin_impl while comparing
          such single byte code points, there is a code which compares
          signed character with unsigned character. Because of this,
          comparisons of two same code points representing a HALF
          WIDTH KATAKANA character always fails.
          
          Solution:
          ---------
          A code point of HALF WIDTH KATAKANA at-least need 8 bits.
          Promoting the variable from uchar to int will fix the issue.
------------------------------------------------------------
revno: 8566
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 13:48:18 +0300
message:
  Bug#19233510 [ERROR] INNODB: IGNORING THE REDO LOG DUE TO MISSING
  MLOG_CHECKPOINT
  
  InnoDB crash recovery expects the redo log to be empty after the
  latest checkpoint, or to contain a MLOG_CHECKPOINT marker.
  
  Sometimes, a MLOG_CHECKPOINT marker was incorrectly being omitted when
  performing a log checkpoint.
  
  We must only omit the marker when performing a shutdown, to avoid
  hitting an infinite loop.
  
  This bug used to mostly occur in tests that invoke extra checkpoints,
  such as those related to WL#6501 (TRUNCATE TABLE) and WL#7277
  (bulk load for index creation).
  
  rb#6330 approved by Jimmy Yang
------------------------------------------------------------
revno: 8565
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 15:26:28 +0530
message:
  Bug#18095449 REPLICATE_WILD_DO_TABLE AND REPLICATE_WILD_IGNORE_TABLE
  ACCEPTING INVALID VALUES
  
  Problem: REPLICATE_WILD_DO_TABLE AND REPLICATE_WILD_IGNORE_TABLE
  filter values should be in the format of 'db_pattern.table_pattern'.
  Invalid values (value without a dot ('.') in it) are also getting accepted
  which is not correct.
  
  Analysis: A dot is used as a separator in the table filter list values.
  If a filter value is found without a '.' in it, it is ignored and 
  will not be accepted. But in the server, before the logic reaches 
  to this check, Server is clearning the existing values. In this situation,
  the existing (old) values are getting cleared and new values are not
  getting set.
  Eg: Replicate_wild_do_table= db1.t%
  If Change replication filter Replicate_wild_do_table=('t2%') is executed
  db1.t% is cleared and 't2%' is not accepted and also the command is
  not throwing any error.
  
  Fix: Change the Wild_[do|ignore]_table syntax at parser level
  not to allow the pattern if it does not contain a dot(.).
  Now we are introducing new tokenizer "qualified_table_as_string"
  which will treat any string without 'dot' as invalid and will
  throw error "ER_INVALID_RPL_WILD_TABLE_FILTER_PATTERN".
------------------------------------------------------------
revno: 8564
committer: Pavan Naik<pavan.naik@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 15:20:47 +0530
message:
  Skipping mysql_no_login.test on embedded server
------------------------------------------------------------
revno: 8563 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-12 14:44:10 +0530
message:
  Bug#19169765 PB2 RPL_GTID_SERVER_SIGHUP.TEST FAILING SPORADICALLY
  WITH RESULT MISMATCH
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.70
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-08-12 14:42:00 +0530
    message:
      Bug#19169765 PB2 RPL_GTID_SERVER_SIGHUP.TEST FAILING SPORADICALLY WITH RESULT MISMATCH
      
      Problem: rpl_gtid_server_sighup.test is failing sporadically with result mismatch.
      A rotate event is getting generated in the relay logs sporadically.
      
      Analysis: The test script tests sighup signal (flush logs) behaviour 
      on both master and slave i.e., the test runs in two iteration
      first iteration on master and second iteration on slave. At the end
      of the each iteration, test restarts server to make sure that
      newly generated binary log/relay log does not cause any issues.
      After the first iteration's restart, we must be sure that the IO
      thread has connected again with the master that has just restarted,
      or else the results of the test case may vary.
      
      Fix: A sync with slave is added after end of the first iteration.
------------------------------------------------------------
revno: 8562
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Mon 2014-08-11 15:53:23 +0200
message:
  Bug#19001781: ADD SUPPORT FOR CMAKE 3 
  
  Post-push fix: CMP00042 is needed for OS X.
------------------------------------------------------------
revno: 8561 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-11 13:07:10 +0100
message:
  Merge of BUG#17326020 patch from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.69
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-08-11 13:02:10 +0100
    message:
      BUG#17326020 ASSERTION ON SLAVE AFTER STOP/START SLAVE
                   USING MTS+GTID REPLICATION
      
      Problem:
      =======
      
      When the IO thread reconnects to a master using GTID auto positioning
      while in the middle of a transaction, it will left the partial
      transaction on the relaylog and will fully retrieve the same
      transaction again.
      
      If the slave is configured to use MTS, it will hit an assert "!worker"
      once reaching the ROTATE_LOG_EVENT send by the master in the IO thread
      reconnection.
      
      Analysis:
      ========
      
      Once a slave with MTS reaches the beginning of a group (by an
      GTID_LOG_EVENT or QUERY(BEGIN)), it will expect to feed the same
      worker with events until reaching the end of the transaction. No
      events that should be applied synchronously (by the MTS coordinator
      with all workers waiting for jobs) can be applied while MTS is feeding
      a worker with events. The assertion exists to stop the server execution
      in such situations.
      
      The correct behavior of the slave once knowing that a transaction was
      left in the middle and will not finish (as it will be applied again
      from the beginning later) is to abort this transaction.
      
      STS SQL thread knows how to rollback the incomplete transaction, but
      MTS doesn't know how to do it yet.
      
      Fix:
      ===
      
      The SQL slave will now check if it is going to apply a synchronous
      ROTATE_LOG_EVENT sent by the master during GTID auto negotiation after
      IO thread reconnection.
      
      Before applying this ROTATE_LOG_EVENT, the SQL slave will check also if
      it is in the middle of a group. If it is, it will queue to the current
      worker a QUERY(ROLLBACK) event to make the worker gracefully finish its
      work and, only then, will let the MTS coordinator to apply the
      ROTATE_LOG_EVENT in synchronous mode.
      
      @ sql/rpl_slave.cc
      
      Added code into exec_relay_log_event() to inject a QUERY(ROLLBACK)
      event if necessary to make the current worker gracefully finish its
      job before applying an event that needs synchronous MTS execution.
------------------------------------------------------------
revno: 8560
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-11 10:43:11 +0300
message:
  Bug#19330255 WL#7142 - CRASH DURING ALTER TABLE LEADS TO
  DATA DICTIONARY INCONSISTENCY
  
  The server crashes on a SELECT because of space id mismatch. The
  mismatch happens if the server crashes during an ALTER TABLE.
  
  There are actually two cases of inconsistency, and three fixes needed
  for the InnoDB problems.
  
  We have dictionary data (tablespace or table name) in 3 places:
  
  (a) The *.frm file is for the old table definition.
  (b) The InnoDB data dictionary is for the new table definition.
  (c) The file system did not rename the tablespace files yet.
  
  In this fix, we will not care if the *.frm file is in sync with the
  InnoDB data dictionary and file system. We will concentrate on the
  mismatch between (b) and (c).
  
  Two scenarios have been mentioned in this bug report. The simpler one
  first:
  
  1. The changes to SYS_TABLES were committed, and MLOG_FILE_RENAME2
  records were written in a single mini-transaction commit.
  The files were not yet renamed in the file system.
  2a. The server is killed, without making a log checkpoint.
  3a. The server refuses to start up, because replaying MLOG_FILE_RENAME2
  fails.
  
  I failed to repeat this myself. I repeated step 3a with a saved
  dataset. The problem seems to be that MLOG_FILE_RENAME2 replay is
  incorrectly being skipped when there is no page-redo log or
  MLOG_FILE_NAME record for the old name of the tablespace.
  
  FIX#1: Recover the id-to-name mapping also from MLOG_FILE_RENAME2
  records when scanning the redo log. It is not necessary to write
  MLOG_FILE_NAME records in addition to MLOG_FILE_RENAME2 records for
  renaming tablespace files.
  
  The scenario in the original Description involves a log checkpoint:
  1. The changes to SYS_TABLES were committed, and MLOG_FILE_RENAME2
  records were written in a single mini-transaction commit.
  2. A log checkpoint and a server kill was injected.
  3. Crash recovery will see no records (other than the MLOG_CHECKPOINT).
  4. dict_check_tablespaces_and_store_max_id() will emit a message about
  a non-found table #sql-ib22*.
  5. A mismatch is triggering the assertion failure.
  
  In my test, at step 4 the SYS_TABLES root page (0:8) contains these 3
  records right before the page supremum:
  * delete-marked (committed) name=#sql-ib21* record, with space=10.
  * name=#sql-ib22*, space=9.
  * name=t1, space=10.
  space=10 is the rebuilt table (#sql-ib21*.ibd in the file system).
  space=9 is the old table (t1.ibd in the file system).
  
  The function dict_check_tablespaces_and_store_max_id() will enter
  t1.ibd with space_id=10 into the fil_system cache without noticing
  that t1.ibd contains space_id=9, because it invokes
  fil_open_single_table_tablespace() with validate=false.
  
  In MySQL 5.6, the space_id from all *.ibd files are being read when
  the redo log checkpoint LSN disagrees with the FIL_PAGE_FILE_FLUSH_LSN
  in the system tablespace. This field is only updated during a clean
  shutdown, after performing the final log checkpoint.
  
  FIX#2: dict_check_tablespaces_and_store_max_id() should pass
  validate=true to fil_open_single_table_tablespace() when a non-clean
  shutdown is detected, forcing the first page of each *.ibd file to be
  read. (We do not want to slow down startup after a normal shutdown.)
  
  With FIX#2, the SELECT would fail to find the table. This would
  introduce a regression, because before WL#7142, a copy of the table
  was accessible after recovery.
  
  FIX#3: Maintain a list of MLOG_FILE_RENAME2 records that have been
  written to the redo log, but not performed yet in the file system.
  When performing a checkpoint, re-emit these records to the redo
  log. In this way, a mismatch between (b) and (c) should be impossible.
  
  fil_name_process(): Refactored from fil_name_parse(). Adds an item to
  the id-to-filename mapping.
  
  fil_name_parse(): Parses and applies a MLOG_FILE_NAME,
  MLOG_FILE_DELETE or MLOG_FILE_RENAME2 record. This implements FIX#1.
  
  fil_name_write_rename(): A wrapper function for writing
  MLOG_FILE_RENAME2 records.
  
  fil_op_replay_rename(): Apply MLOG_FILE_RENAME2 records. Replaces
  fil_op_log_parse_or_replay(), whose logic was moved to fil_name_parse().
  
  fil_tablespace_exists_in_mem(): Return fil_space_t* instead of bool.
  
  dict_check_tablespaces_and_store_max_id(): Add the parameter
  "validate" to implement FIX#2.
  
  log_sys->append_on_checkpoint: Extra log records to append in case of
  a checkpoint. Needed for FIX#3.
  
  log_append_on_checkpoint(): New function, to update
  log_sys->append_on_checkpoint.
  
  mtr_write_log(): New function, to append mtr_buf_t to the redo log.
  
  fil_names_clear(): Append the data from log_sys->append_on_checkpoint
  if needed.
  
  ha_innobase::commit_inplace_alter_table(): Add any MLOG_FILE_RENAME2
  records to log_sys->append_on_checkpoint(), and remove them once the
  files have been renamed in the file system.
  
  mtr_buf_copy_t: A helper functor for copying a mini-transaction log.
  
  rb#6282 approved by Jimmy Yang
------------------------------------------------------------
revno: 8559
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk
timestamp: Mon 2014-08-11 11:23:16 +0400
message:
  WL#7976: Deprecate skip-innodb in 5.6, remove in 5.7.
  
  Follow-up patch to remove rpl_myisam_recovery.test.
  
  The point of rpl_myisam_recovery.test was to check that even if the
  procedure to the binary log and at the same time recover prepared
  transaction was engaged when total_ha_2pc <= 1 there would not be any
  assert raised inside ha_recover.
  
  Since WL#7976 there is no way to start the server without InnoDB. Thus,
  there can be no case when total_ha_2pc <= 1.
  
  This removal has been discussed with RPL & InnoDB teams.
------------------------------------------------------------
revno: 8558
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-11 15:17:10 +0800
message:
  WL#6964 MTS: Support SLAVE_TRANSACTION_RETRIES in MTS mode
  
  Removed the unnecessary statement 'SELECT @@slave_parallel_type' which
  would have different result and cause a test failure if LOGICAL_CLOCK
  was set as the default through mysqld option.
------------------------------------------------------------
revno: 8557 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug17879675_mysql-trunk
timestamp: Mon 2014-08-11 12:26:06 +0530
message:
  Bug#17879675:SHUTDOWN HANG IF SEMISYNC IS ENABLED AND SLAVE
  ABORT
  
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.68
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug17879675_mysql-5.6
    timestamp: Mon 2014-08-11 12:06:08 +0530
    message:
      Bug#17879675:SHUTDOWN HANG IF SEMISYNC IS ENABLED AND SLAVE
      ABORT
      
      Problem:
      ========
      Step 1. enable semisync and set a big value for
      rpl_semi_sync_master_timeout.
      
      Step 2. kill io thread of slave, so the threads on master
      will keep on WAITING ACK FROM SLAVE.
      
      Step 3. shutdown the master ---hang
      
      Analysis:
      ========
      On master when semisync is enabled the transaction thread
      will wait for an ack from the slave in a conditional timed
      wait. Since the IO thread is already killed and on master
      a huge timeout value is set transaction thread will continue
      to wait for an ack for a long time. At this time if shutdown
      command is issued dump thread will try to go down. But dump
      thread when goes down it has a check if user has enabled
      rpl_semi_sync_master_wait_no_slave or not. If it is enabled
      dump thread will go down without switching off semi sync.
      Transaction thread will continue to wait for an ack even
      when slave count drops to zero. Because of this shutdown
      also will wait for the transaction thread to go down and
      shutdown hangs.
      
      Fix:
      ===
      At the time of shutdown if dump thread finds that there no
      semi sync slaves that are connected to master it will switch
      off semisync even when rpl_semi_sync_master_wait_no_slave
      is set to on. i.e if shutdown finds no slaves are there to
      provide ack it will not honour
      rpl_semi_sync_master_wait_no_slave and it will switch off
      semisync. The transaction thread when woken up realises the
      semi sync is switched off it will not wait any more for ack
      it will continue and shutdown will be complete. When semi
      sync is forcefully shutdown at the time of waiting for an
      ack a warning is written to error log to convey to users
      that it was forceful shutdown.
      
      There can be a scenario where dump thread is already gone
      and transaction thread still waits for an ack. Hence a
      similar fix is implemented for the waiting transaction
      thread. I.e at the time of shutdown the thread gets a
      wakeup signal and it checks if shutdown is in progress
      if no slaves are connected then it will switch off the
      semi sync.
------------------------------------------------------------
revno: 8556
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-11 11:35:05 +0530
message:
  BUG#17787664 - ASSERT `! IS_SET()` IN DIAGNOSTICS_AREA::SET_OK_STATUS
                 ON ER_SUBQUERY_NO_1_ROW
  
  
  ANALYSIS :
  
  The SELECT subquery in the below query triggers an assert on
  a debug build.
  
  DELETE  FROM OUTR1.* USING t1 AS OUTR1
      LEFT OUTER JOIN t2 AS OUTR2  ON ( OUTR1.a = OUTR2.b )
  WHERE OUTR1.a < ( SELECT t3.c  from t3 where 1 XOR OUTR2.b IS NOT NULL);
  
  SELECT subquery returns two records which is an incorrect
  input for the comparison operator('<'). Hence the execution of
  the query sets the status of the diagnostic area to DA_ERROR
  due to the error 'ER_SUBQUERY_NO_1_ROW'. The query
  execution continues due to a missing error check in the
  function 'evaluate_null_complemented_join_record()'.
  While trying to set the status of the diagnostic area to
  DA_OK, the assert is triggered since the error condition
  is already set in the diagnostic area for the query.
  
  In case of a release build, an appropriate error
  'ER_SUBQUERY_NO_1_ROW' is reported.
  
  FIX:
  
  Added error check in evaluate_null_complemented_join_record().
------------------------------------------------------------
revno: 8555
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-11 09:59:05 +0800
message:
  WL#6964 MTS: Support SLAVE_TRANSACTION_RETRIES in MTS mode
  
  Postfix:
  Fixed the valgrind warning that trans_retries was not initialized.
  And a possible memory leak on the allocated Log_event if it was a
  Rotate_log_event or other types of event didn't belong to the transaction.
------------------------------------------------------------
revno: 8554 [merge]
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-08-10 10:48:26 +0800
message:
  Auto Merge WL#6964
    ------------------------------------------------------------
    revno: 8552.1.2
    committer: Libing Song <libing.song@oracle.com>
    branch nick: mysql-trunk
    timestamp: Sat 2014-08-09 10:18:51 +0800
    message:
      Bug#19282301 WL6964: SLAVE HAS EXECUTED ALL THE TRXNS STILL ERROR 1205 REMAINS
      
      SHOW SLAVE STATUS reported the tempoary error to users when the worker is still
      retrying the transaction. That was because workers stored any error(
      including temporary error) into slave status.
            
      After this patch, workers will not store tempoary error into slave status.
    ------------------------------------------------------------
    revno: 8552.1.1
    committer: Libing Song <libing.song@oracle.com>
    branch nick: mysql-trunk
    timestamp: Sat 2014-08-09 10:16:20 +0800
    message:
      WL#6964  MTS: Support SLAVE_TRANSACTION_RETRIES in MTS mode
      
      DESCRIPTION
      ===========
      This patch implemented slave_transaction_retries feature on multi-threaded slave.
      Workers can retry the transactions being applied automatically if any temporary error
      is encountered.
      
      Basically, the behaviour of retrying transaction on multi-threaded slave is same
      to the behavior of retrying transaction on singled-threaded slave. Definitions of
      temporary error are same, and sleep policies are same too.
      
      DESIGN
      ======
      - How To Retry
        The worker reads the transaction's event from relay logs and applies them.
        But it just reads and applies the events from the first one to the one on which
        temporary error happens. The following events are still in job queue, so they
        are not read from relay logs.
      
      - How To Know Events' Positions of The Transaction
        When coordinator dispatches each event to a worker, it stores the event's position
        (relay log name and offset) into Slave_job_item structure together with the event.
        To minimize the performance impact, relay log name is not stored as a string. As
        you know, relay log's extension is an auto-increment number. So we just store the
        number as a integer into Slave_job_item.
      
        Workers can construct the real relay log name by using relay_log_basename server
        variable and the number.
        NOTE: Workers may construct a wrong relay log name if users have changed
              relay log base name and events of the transactions are in the relay log with
              old basename. In this case, slave will stop with an error.
------------------------------------------------------------
revno: 8553 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-08-09 07:38:55 +0200
message:
  Merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.338.67
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Sat 2014-08-09 07:35:41 +0200
    message:
      Bug#19295893 : ASSERT ON SYNC_THREAD_LEVELS_G(ARRAY, SYNC_IBUF_BITMAP IN IBUF_BITMAP_GET_MAP_PA
      
      btr_insert_on_non_leaf_level() might get SYNC_IBUF_BITMAP level page temporally for child temporally mini-transaction.
      So the caller of btr_insert_on_non_leaf_level() should not have any SYNC_IBUF_BITMAP level pages.
      
      But btr_insert_into_right_sibling() might call btr_insert_on_non_leaf_level() after ibuf_update_free_bits_zip().
      ibuf_update_free_bits_zip() uses the parent mini-transaction and has SYNC_IBUF_BITMAP level page until committed the mini-transaction.
      
      ibuf_update_free_bits_zip() should be at the last of btr_insert_into_right_sibling().
      
      This problem is only for compressed table.
      
      Approved by Sunny in rb#6253
------------------------------------------------------------
revno: 8552
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-08 13:53:52 +0200
message:
  WL#7308: Refactor mysql_upgrade
  
  mysql_upgrade should not call other applications by command line (as it makes password being visible).
  
  This tool was refactored to call mysql queries by MYSQL C API. To use mysqlcheck it was first divided into library and client that parses options, creates connection and passes parsed parameters to library. Mysql_upgrade uses this mysqlcheck library. Also C++ wrappers were created on parsing options to encapsulate program options groups. Classes for common groups of options were implemented on that. Abstract client tool program class was implemented. All this new functionalities were used in mysql_upgrade. A C++ wrapper to mysqlcheck functionality was created and used in mysql_upgrade.
------------------------------------------------------------
revno: 8551
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk-wl7764-itch
timestamp: Fri 2014-08-08 14:19:21 +0530
message:
  WL#7764: Make STRICT mode for transactional SEs default in 5.7
  
  Addressing post push test failures.
------------------------------------------------------------
revno: 8550
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-08 16:30:50 +0800
message:
  Fix Bug #19365351 - ASSERT IN PURGE, RTREE: LOCK->LOCK_WORD <= -0X20000000
  
  rb://6281 approved by Yasufumi
------------------------------------------------------------
revno: 8549
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-08 09:33:35 +0200
message:
  Provide mysql_no_login.so plugin
------------------------------------------------------------
revno: 8548
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-08 07:44:26 +0200
message:
  Fix RPM package build after WL#7726.
  
  Patch approved by Bala over IM.
------------------------------------------------------------
revno: 8547 [merge]
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-08 10:44:18 +0800
message:
  BUG#16741603: MYSQLD SCANS ALL BINARY LOGS ON CRASH RECOVERY
  
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.66
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-08-08 10:12:49 +0800
    message:
      BUG#16741603: MYSQLD SCANS ALL BINARY LOGS ON CRASH RECOVERY
      
      Mysql server iterates backwards through binary logs, looking for
      the last binary log that contains a Previous_gtids_log_event for
      gathering the set of gtid_executed, and iterates forwards through
      binary logs, looking for the first binary log that contains both
      Previous_gtids_log_event and gtid_log_event for gathering the set
      of gtid_purged on crash recovery. Mysql server also iterates
      forwards through binary logs, looking for the first binary log
      that contains both Previous_gtids_log_event and gtid_log_event
      for gathering the set of gtid_purged when purging binary logs.
      This may take very long time if the mysql server has many binary
      logs and almost all of them are out of filesystem cache.
      
      To fix the problem, introduce an option
      'simplified-binlog-gtid-recovery', whose default value is false.
      If the option is enabled, we do this: In the first scan (where
      it iterates over binary logs from the newest to the oldest):
      If the last binary log does not contain any GTID event, do not
      read any more binary logs, and set GTID_EXECUTED = '' and
      GTID_PURGED = ''. Otherwise, initialize GTID_EXECUTED as usual.
      Then, in the second scan (where it iterates from the oldest to
      the newest): If the first binary log does not contain any GTID
      event, do not read any more binary logs, and set GTID_PURGED = ''.
------------------------------------------------------------
revno: 8546
committer: Todd Farmer <todd.farmer@oracle.com>
branch nick: mysql-trunk-wl7726-merge
timestamp: Thu 2014-08-07 15:00:13 -0700
message:
  Fixing WL#7726 push for 5.7 API.
------------------------------------------------------------
revno: 8545
committer: Todd Farmer <todd.farmer@oracle.com>
branch nick: mysql-trunk-wl7726-merge
timestamp: Thu 2014-08-07 13:53:21 -0700
message:
  Merging WL#7726 - mysql_no_auth plugin.
------------------------------------------------------------
revno: 8544
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk
timestamp: Thu 2014-08-07 12:24:26 +0200
message:
  post-push fix for Bug#19336348 DEBUG CRASH AT SETUP_SEMIJOIN_DUPS_ELIMINATION IN SQL/SQL_SELECT.CC
  we have to free memory.
------------------------------------------------------------
revno: 8543
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk-wl7764-itch
timestamp: Thu 2014-08-07 16:09:35 +0530
message:
  WL#7764: Make STRICT mode for transactional SEs default in 5.7
  
  Addressing post push test failures.
------------------------------------------------------------
revno: 8542
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-06 12:56:26 +0300
message:
  Whitespace and style fixup.
------------------------------------------------------------
revno: 8541
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-06 16:29:11 +0200
message:
  Bug#19336348 DEBUG CRASH AT SETUP_SEMIJOIN_DUPS_ELIMINATION IN SQL/SQL_SELECT.CC
  the assertion failure was introduced by wl#6042; but even before the
  WL the chosen plan was buggy - semijoin loose scan was using table
  scan which gives unsorted rows.
  The cause is: pre-WL, when we notice that a QUICK cannot be used for
  sj-loosescan, we drop this QUICK, accidentally reverting to table
  scan; pre-WL it was in an "else" branch which was not covered:
          if (tab->select && tab->select->quick)
          {
            if (tab->select->quick->index == pos->loosescan_key)
              tab->select->quick->need_sorted_output();
            else
              tab->select->set_quick(NULL); <<<<<<<<<<
  In the WL, the inner if() was changed to an assertion.
  Dropping the QUICK in setup_semijoin_dup_etc was wrong, because at
  this stage the QUICK is already considered the final access method.
  We must indeed drop QUICK, but must do it earlier, at a stage where
  QUICK is still "the best QUICK _if we want to use a QUICK_ "; then
  other options are still open, and the optimizer will consider ref or
  index scan for sj-loosescan.
------------------------------------------------------------
revno: 8540
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: vg
timestamp: Tue 2014-08-05 18:04:44 +0200
message:
  review comment from Tor for wl#6042
------------------------------------------------------------
revno: 8539
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: vg
timestamp: Tue 2014-08-05 10:48:41 +0200
message:
  signature of some functions changed in wl#6042, so suppression has to change too
------------------------------------------------------------
revno: 8538
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-06 15:55:55 +0200
message:
  WL5869: Optimizer cost: use Cost_estimate in range optimizer
    Introduced some changes for other bug fix by mistake. Rollbacked them.
------------------------------------------------------------
revno: 8537
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-06 13:44:25 +0200
message:
  WL#5869: Optimizer cost: use Cost_estimate in range optimizer
------------------------------------------------------------
revno: 8536
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk-wl7764-itch
timestamp: Wed 2014-08-06 16:47:22 +0530
message:
  WL#7764: Make STRICT mode for transactional SEs default in 5.7
  
  Change default SQL mode from 'NO_ENGINE_SUBSTITUTION'
  to 'NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES'.
  
  Introduce command line option --sql_mode to mysqlslap tool.
  This facilitates to emulate client load in user defined SQL mode.
------------------------------------------------------------
revno: 8535 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-06 15:37:30 +0530
message:
  Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
  	merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.65
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-08-06 15:34:47 +0530
    message:
      Bug #18734396	INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
      
      Analysis:
      	When an InnoDB in-place ALTER fails, it leaves behind both a
      temporary filename like "#sql-ibtid" where tid represents the table ID of
      the table being altered) in its data dictionary. This makes future in-place
      ALTERs of the same table impossible because the temporary table is named
      only with the table ID.
      
      Solution:
      	This patch is adding more uniqueness to the temporary file name. It
      creates a temporary tablename like
      "#sql-ibtid-inc" where 
               tid = the table ID
               inc = static global number that is initialized to a random
      distributed 32-bit number using ut_time() and ut_crc32().It is then 
      incremented atomically for each temporary file name assigned.
      
      	rb5580 Approved by Kevin, Marko
------------------------------------------------------------
revno: 8534 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-06 11:39:14 +0200
message:
  Merge from 5.6 => trunk
  
  - Provide mysql-compat-server dependencies
    ------------------------------------------------------------
    revno: 3902.338.64 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-08-06 11:16:54 +0200
    message:
      Merge from 5.5 => 5.6 
      
      - Provide mysql-compat-server dependencies
      - Updated Release version for 5.6.21
        ------------------------------------------------------------
        revno: 2875.598.25 [merge]
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-08-06 09:56:37 +0200
        message:
          - Merge from mysql-5.5.39-ol7-release branch
          - Reverted version variable 
            ------------------------------------------------------------
            revno: 2875.624.2
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Mon 2014-08-04 15:56:19 +0200
            message:
              Updated for el7 regular rpms
            ------------------------------------------------------------
            revno: 2875.624.1
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-ol7-release
            timestamp: Thu 2014-07-24 11:37:40 +0200
            message:
              Bug#19223915 Provide mysql-compat-server dependencies
------------------------------------------------------------
revno: 8533
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-06 14:17:27 +0530
message:
  Bug #16204823 INNODB WASTES 62 OUT OF EVERY 16384 PAGES
  
  Reverting my previous patch (rb#5559) because of regressions.
------------------------------------------------------------
revno: 8532
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-08-06 10:33:56 +0200
message:
  Followup for Bug#18805275 fix to fix innodb.log_file problem at Windows.
  
  All AIO requests are waited to be finished even if aborting init process of InnoDB.
------------------------------------------------------------
revno: 8531 [merge]
committer: bin.x.su@oracle.com
branch nick: trunk
timestamp: Wed 2014-08-06 09:56:11 +0800
message:
  Remove unstable test case innodb_bug18942294, approved by Jimmy over IM.
    ------------------------------------------------------------
    revno: 3902.338.63 [merge]
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Wed 2014-08-06 09:54:52 +0800
    message:
      Remove unstable test case innodb_bug18942294, approved by Jimmy over IM.
        ------------------------------------------------------------
        revno: 2875.598.24
        committer: bin.x.su@oracle.com
        branch nick: mysql-5.5
        timestamp: Wed 2014-08-06 09:51:20 +0800
        message:
          Remove unstable test case innodb_bug18942294, approved by Jimmy over IM.
------------------------------------------------------------
revno: 8530
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Wed 2014-08-06 03:33:47 +0200
message:
  Fixed bug#19318293: SWITCHING TO INNODB AS DEFAULT SE (FOR INTRINSIC TABLE)
                      IS CAUSING CORE-DUMP ON
  In wl6711, if we want Innodb as the internal tmp table engine, we must assure
  Innodb plugin is loaded always. So in order to fix such an issue, I added the
  option to load Innodb plugin always if temp table engine is Innodb. However,
  recently, new push(revno:8473) made Innodb always loaded which results in a
  conflict.
------------------------------------------------------------
revno: 8529
committer: bin.x.su@oracle.com
branch nick: trunk
timestamp: Wed 2014-08-06 09:28:51 +0800
message:
  BUG#19297656 - REDUCE TWO MEMSET TO ONE MEMSET CALL FOR UPD_CREATE CALL.
  
  In upd_create, we call mem_heap_zalloc twice to allocate memory for upd_t.
  In fact, we can call it just once for the whole memory, which we can divide
  into 2 pieces and assign them to different pointer.
  
  rb#6147, approved by Jimmy.
------------------------------------------------------------
revno: 8528
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt2
timestamp: Tue 2014-08-05 17:31:32 +0200
message:
  Fix bug#18550977:
    N TEST FRAMEWORK IN MYSQL-TRUNK DOES NOT SUPPORT OUT-OF-SOURCE BUILDS
    Enables n testing in an arbitrary build or install directory.
    The directory to test is specified through new options --build
    to ibtest_ctd.pl. --install, as well as --build-directory and
    --install-directory are synonyms to --build.
    It does not have to be built on the same MySQL source tree that
    ibtest_ctd.pl itself comes from.
    All test data is collected in mysql-test/var under the build directory,
    similar to what mtr does.
    It is possible now to run simultaneous tests over different
    build and install location using using ibtest_ctd.pl from
    a single MySQL source tree.
    For the case of testing a build directory, the original source
    directory must be available.
  n/n_ct/ibtest_ctd.pl:
    add new option --build to specify MySQL build directory.
    It is supported only in MySQL source tree environment.
  n/n.pm:
    n_mst_env(): add optional build directory parameter
    n_mst_setup_nroot(): add optional parameter for 'source' N_ROOT
      from where ibtest_ctd.pl comes from.
  n/n: always use N_ROOT from the environment, if it is there
  n/n.dtn:
    support simultaneous tests with the same ibtest_ctd.pl,
    in MySQL source tree environmen:
      - do not resolve symbolic links in N_ROOT to set N_Absroot
      - change definitions of some processgroups to account for the fact
        that the sandbox name in MySQL source tree environment
        is always 'mysql'.
------------------------------------------------------------
revno: 8527
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-05 17:25:57 +0530
message:
  Bug #19077964 ASSERT PAGE_SIZE.EQUALS_TO SPACE_PAGE_SIZE
  BTR_COPY_BLOB_PREFIX
  
  Problem:
  
  The function row_build_index_entry_low() takes a dtuple_t object ('row') and
  dict_index_t object ('index') as input and returns a new dtuple_t object
  ('entry') as output.  The dtuple_t object 'row' that is given as input might
  have been constructed from a different dict_index_t object (!= index).  So
  when accessing the externally stored data of the given 'row' we need to make
  use of the correct index object.
  
  Solution:
  
  Store the page size information in the associated row_ext_t object.
  
  rb#6086 approved by Vasil and Jimmy.
------------------------------------------------------------
revno: 8526
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-08-05 11:48:15 +0200
message:
  Followup for Bug#18805275 fix.
  
  The rest problem for innodb.log_file test for Windows still exists.
  Adjusting innodb.log_file and add debug output for debug build to abtain accurate reason.
------------------------------------------------------------
revno: 8525
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18937504cpuflat
timestamp: Mon 2014-08-04 16:43:00 +0200
message:
  Bug#18937504 ARITHMETIC WITH UTF16 STRING WITH A INVALID UTF16 STRING FLATLINE CPU
  
  Handle illegal string input in
   - my_strtoll10_mb2()
   - my_strnncollsp_utf16()
   - my_strnncollsp_utf16_bin()
------------------------------------------------------------
revno: 8524
committer: Tiago Jorge <tiago.jorge@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-04 11:56:34 +0100
message:
  Bug#17867478 BINLOG.BINLOG_SERVER_START_OPTIONS FAILING
               DUE TO ERROR IN SERVER LOGS
  
  Problem:
  
  An error in the logs is causing the test to fail
  
  Analysis:
  
  If one has a closer look at the issue has always been there.
  But since mysqld was waiting for an error, the test didn't
  failed when mysqld crashed because a the missing property.
  
  But i went further on the analysis and the test was not
  testing what was coded, which was
  "if the param 'datadir' is empty then we should switch
  its value to point to the actual 'datadir'"
  
  This possibly corrects also other bugs.
  
  Fix:
  
  Rewrite the mysql test to check if the modification of
  the variable value suceeded only on the context of
  rpl_make_log_name.
------------------------------------------------------------
revno: 8523
committer: Vamsikrishna Bhagi <vamsikrishna.bhagi@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-08-04 11:39:52 +0530
message:
  Problem: mysql_secure_installation_ssl.test fails on
           windows.
  
  Solution: mysql_secure_installation is not usable on
            windows. Hence disabling the test for it.
------------------------------------------------------------
revno: 8522
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-08-03 11:08:23 +0200
message:
  Followup for Bug#18805275 fix.
  
  Fixed problem for innodb.log_file test for Windows.
  Avoids to generate mysqld-debug.dmp during the test.
  IO thread can exit always when SRV_SHUTDOWN_EXIT_THREADS, if page cleaner already exited.
  
  ------------------------------------------------------------
  revno: 8489
  committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
  branch nick: mysql-trunk
  timestamp: Sat 2014-07-26 03:22:05 +0900
  message:
    Bug #18805275 : PAGE CLEANER SHOULD BE MULTI-THREADED ALSO FOR RECOVERY AND SHUTDOWN
  
    From 5.7, page cleaner is multi-threaded for performance scalability.
    But it is not used during shutdown and recovery phases.
  
    It should be multi-threaded during shutdown and recovery phases for their
    optimal performance.
  
    * The previous problem was fixed. (releasing the event objects was too early)
  
    Approved by Sunny in rb#5465
  ------------------------------------------------------------
------------------------------------------------------------
revno: 8521 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18710853
timestamp: Sat 2014-08-02 13:33:42 +0530
message:
  Merge from 5.6 to trunk.
  
  Fix in trunk
  -------------
  Innodb now passes the key 'dbname/tablename' in its canonical format to 
  the query cache interface. Query cache then converts the string to a 
  format 'dbname\0tablename' in its non-canonical format before the lookup
  for invalidating the query cache for the table specified the key.
    ------------------------------------------------------------
    revno: 3902.338.62
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18710853
    timestamp: Sat 2014-08-02 13:21:08 +0530
    message:
      BUG#18710853: QUERY CACHE NOT INVALIDATED ON CASCADE DELETE 
                    IF DB NAME HAS SPECIAL SYMBOLS 
      
      Analysis      
      --------
      The query cache is not invalidated for a table when the CASCADE
      DELETE/UPDATE referential constraint is specified and the
      database name or table name contains special characters.
            
      InnoDB triggers invalidation of the query cache while performing
      the check for CASCADE DELETE/UPDATE referential constraint. InnoDB
      passes the key in the format of 'dbname\0tablename' to the query cache
      interface where the database name and table name are in the canonical
      format(encoded-format for special characters). The key used by the query
      cache interface is 'dbname\0tablename' in its non-canonical format. 
      The lookup performed for query cache invalidation fails for the condition
      specified above due to the mismatch in the key.
            
      Hence the records fetched with the query cache and without the query
      cache differs.
            
      Fix
      ---
         
      Innodb now passes the key 'dbname\0tablename' in its non-canonical format to 
      the query cache interface. Thus the query cache look up succeeds and the
      query is invalidated.
------------------------------------------------------------
revno: 8520 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Fri 2014-08-01 17:12:37 +0530
message:
  Bug #18415196 MYSQL_UPGRADE DUPLICATE KEY ERROR FOR MYSQL.USER FOR 5.5.35+, 5.6.15+, 5.7.3+ 
  
  Merging from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.61 [merge]
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Fri 2014-08-01 17:11:15 +0530
    message:
      Bug #18415196 MYSQL_UPGRADE DUPLICATE KEY ERROR FOR MYSQL.USER FOR 5.5.35+, 5.6.15+, 5.7.3+ 
      
      Merging from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.23
        committer: Venkata Sidagam <venkata.sidagam@oracle.com>
        branch nick: 5.5
        timestamp: Fri 2014-08-01 17:09:55 +0530
        message:
          Bug #18415196 MYSQL_UPGRADE DUPLICATE KEY ERROR FOR MYSQL.USER FOR 5.5.35+, 5.6.15+, 5.7.3+
          
          Follow-up patch. Removed unwanted code.
------------------------------------------------------------
revno: 8519 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Fri 2014-08-01 14:25:13 +0530
message:
  Bug #18415196 MYSQL_UPGRADE DUPLICATE KEY ERROR FOR MYSQL.USER FOR 5.5.35+, 5.6.15+, 5.7.3+
  
  Merging from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.60 [merge]
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Fri 2014-08-01 14:22:04 +0530
    message:
      Bug #18415196 MYSQL_UPGRADE DUPLICATE KEY ERROR FOR MYSQL.USER FOR 5.5.35+, 5.6.15+, 5.7.3+
      
      Merging from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.22
        committer: Venkata Sidagam <venkata.sidagam@oracle.com>
        branch nick: 5.5
        timestamp: Fri 2014-08-01 14:18:28 +0530
        message:
          Bug #18415196 MYSQL_UPGRADE DUPLICATE KEY ERROR FOR MYSQL.USER FOR 5.5.35+, 5.6.15+, 5.7.3+
          
          Description: mysql_upgrade fails with below error, 
          when there are duplicate entries(like 'root'@'LOCALHOST'
          and 'root'@'localhost') in mysql.user table.
          ERROR 1062 (23000) at line 1140: Duplicate entry 'localhost-root' for key 'PRIMARY'
          FATAL ERROR: Upgrade failed
          
          Analysis: As part of the bug 12917151 fix we are 
          making all the hostnames as lower case hostnames.
          So, this has been done by mysql_upgrade.
          In case of above mentioned duplicate entries 
          mysql_upgrade tries to change hostname to lowercase.
          Since there is already 'root'@'localhost' exists.
          it is failing with "duplicate entry" error.
          
          Fix: Since its a valid error failure. We are 
          making the error more verbose. So, that user will
          delete the duplicate errors manually.
          Along with existing error we are printing below
          error as well.
          ERROR 1644 (45000) at line 1153: Multiple accounts exist for @user_name, @host_name that differ only in Host lettercase; remove all except one of them
------------------------------------------------------------
revno: 8518 [merge]
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-01 13:58:38 +0530
message:
  Bug #11752369 : MYSQLDUMP SHOULD CONVERT GEOMETRY COLUMNS TO
                  HEX WITH --HEX-BLOB OPTION
  
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.59
    committer: mithun <mithun.c.y@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-08-01 13:56:47 +0530
    message:
      Bug #11752369 : MYSQLDUMP SHOULD CONVERT GEOMETRY COLUMNS TO
                      HEX WITH --HEX-BLOB OPTION
      
      ISSUE :
      -------
      If mysqldump is executed with --hex-blob option, all the
      blob type data should be converted to hex value. The
      geometry data are also blob, and their values should be
      dumped in hexadecimal form. But mysqldump dumps geometry
      data in binary form.
      
      Fix :
      -----
      If value being dumped is of geometry type and --hex-blob is
      set then mysqldump will dump such values in hexadecimal form.
------------------------------------------------------------
revno: 8517 [merge]
committer: Akhil Mohan <akhil.mohan@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-08-01 09:21:11 +0200
message:
  Upmerge of the 5.6.20 build
    ------------------------------------------------------------
    revno: 3902.338.58 [merge]
    author: akhil.mohan@oracle.com
    committer: Akhil Mohan <akhil.mohan@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-08-01 09:04:12 +0200
    message:
      Merge from mysql-5.6.20-release
        ------------------------------------------------------------
        revno: 3902.339.12
        tags: mysql-5.6.20
        committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Fri 2014-07-18 21:03:31 +0530
        message:
          WL#7219: Fixing failures in pb2
        ------------------------------------------------------------
        revno: 3902.339.11
        committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Fri 2014-07-18 17:41:16 +0530
        message:
          Bug#19244772:BACKPORT BUG #18923691 TO MYSQL-5.6
        ------------------------------------------------------------
        revno: 3902.339.10
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Fri 2014-07-18 16:30:44 +0530
        message:
          Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
          SHOW PROCESSLIST, SHOW BINLOGS
          
          Post push fix (removing try-catch)
        ------------------------------------------------------------
        revno: 3902.339.9
        committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Thu 2014-07-17 19:26:40 +0530
        message:
          WL#7219: Pushing it to mysql-5.6.20-release branch
        ------------------------------------------------------------
        revno: 3902.339.8
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Tue 2014-07-15 16:14:30 +0530
        message:
          Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO
                        DTRACE PROBLEMS
          
          Reverting back patch for this.
        ------------------------------------------------------------
        revno: 3902.339.7
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Tue 2014-07-15 13:19:03 +0530
        message:
          Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO
                        DTRACE PROBLEMS
          
          Follow up patch. Adding compile definitions was resulting in build
          break on some platforms. Reverting compile definitions appended
          to CFLAGS.
        ------------------------------------------------------------
        revno: 3902.339.6
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Mon 2014-07-14 14:23:44 +0530
        message:
          Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO
                        DTRACE PROBLEMS
          
          Issue here is, in function "DTRACE_INSTRUMENT"(dtrace.cmake), 
          CMAKE_C_FLAGS are stored in environment variable CFLAGS. Value of
          this variable is used by dtrace tool. But currently we were 
          passing C FLAGS as "-Wall -Wextra -Wformat-security -Wvla
          -Wwrite-strings -Wdeclaration-after-statement -Werror" also which
          is not necessary. Code to set environment variable CFLAGS is 
          added to pass flags as -m32 or -m64.
          
          To fix this issue, changed code to strip flags starting with
          "-W" and passing other flags. Also appended COMPILE_DEFINITIONS
          (as -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64) to the CFLAGS.
        ------------------------------------------------------------
        revno: 3902.339.5
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Thu 2014-07-10 13:11:43 +0200
        message:
          Bug#19172145 - Remove dtrace dependencies
        ------------------------------------------------------------
        revno: 3902.339.4
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Tue 2014-07-08 14:08:15 +0200
        message:
           Bug#19155121 - Remove perl(GD) and dtrace dependencies and bench fix
        ------------------------------------------------------------
        revno: 3902.339.3
        committer: Laasya Moduludu <laasya.moduludu@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Tue 2014-07-08 11:29:20 +0200
        message:
          patch for 18779944
------------------------------------------------------------
revno: 8516 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: trunk_new
timestamp: Fri 2014-08-01 11:13:10 +0530
message:
  WL#7219: Fixing failures in pb2
  
  Null merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.338.57
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6_new
    timestamp: Fri 2014-08-01 10:51:31 +0530
    message:
      WL#7219: Fixing failures in pb2
      
      Pushing pb2 issue fixed by "Ashish Agarwal<ashish.y.agarwal@oracle.com>"
      on 5.6.20-release branch to 5.6 and trunk.
------------------------------------------------------------
revno: 8515 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Thu 2014-07-31 21:54:18 +0200
message:
  WL#6884 PERFORMANCE SCHEMA, USER VARIABLES
  
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 6953.2.36 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-07-31 20:33:27 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.35 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-07-31 18:09:14 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.34
    committer: Tarique Saleem <tarique.saleem@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-07-31 19:37:00 +0530
    message:
      Added tests around user_variable interaction with procedures and triggers
    ------------------------------------------------------------
    revno: 6953.2.33 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2014-07-29 22:31:13 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.32 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2014-07-29 01:56:04 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.31
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Sun 2014-07-27 22:51:38 +0200
    message:
      Fixed test robustness (not executed in embedded)
    ------------------------------------------------------------
    revno: 6953.2.30
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-07-25 20:33:55 +0200
    message:
      Added missing locks to protect user variables on thread disconnect
    ------------------------------------------------------------
    revno: 6953.2.29
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-07-25 20:01:06 +0200
    message:
      Added missing locks on user variables
    ------------------------------------------------------------
    revno: 6953.2.28
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-07-25 19:49:38 +0200
    message:
      Code cleanup
    ------------------------------------------------------------
    revno: 6953.2.27 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-07-25 17:42:54 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.26
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-07-25 12:13:57 +0200
    message:
      Improved test robustness
    ------------------------------------------------------------
    revno: 6953.2.25
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-07-25 11:48:07 +0200
    message:
      Bug#19266549 - CRASH AT USER_VAR_ENTRY::VAL_STR IN ITEM_FUNC.CC
      
      Enforced that access to user variables, or user variables values,
      is protected with THD::LOCK_thd_data.
      
      This is needed to prevent race conditions between:
      - a monitoring thread that inspects a target thread user variable
      - the target thread that changes user variables and/or values
    ------------------------------------------------------------
    revno: 6953.2.24
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-07-24 01:02:56 +0200
    message:
      Implement code review comments (use Prealloced_array instead of std::vector)
    ------------------------------------------------------------
    revno: 6953.2.23
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Wed 2014-07-23 23:13:01 +0200
    message:
      Implemented code review comments
    ------------------------------------------------------------
    revno: 6953.2.22 [merge]
    committer: Christopher Powers <chris.powers@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Wed 2014-07-23 11:14:33 -0500
    message:
      merge from mysql-trunk
    ------------------------------------------------------------
    revno: 6953.2.21
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-07-17 10:50:10 +0200
    message:
      WL#6884 PERFORMANCE SCHEMA, USER VARIABLES
      
      Implemented code review comments.
      
      Bug#19193827 DIFFERENCE IN LENGTH OF THE VARIABLE_VALUE IN USER_VARIABLES_BY_THREAD TABLE
      
      Changed the VARIABLE_VALUE column from BLOB to LONGBLOB.
    ------------------------------------------------------------
    revno: 6953.2.20 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Wed 2014-07-16 18:23:19 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.19 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Mon 2014-07-07 17:33:33 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.18 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Wed 2014-07-02 23:20:10 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.17 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2014-06-24 23:19:00 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.16 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Mon 2014-06-23 21:22:34 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.15
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Mon 2014-06-23 15:54:01 +0200
    message:
      Improve test coverage
    ------------------------------------------------------------
    revno: 6953.2.14 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-06-19 13:59:27 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.13
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-06-19 13:57:35 +0200
    message:
      Adjusted test script for 5.7
    ------------------------------------------------------------
    revno: 6953.2.12 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-06-19 09:59:15 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 6953.2.11
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-06-12 02:36:57 +0200
    message:
      WL#6884 PERFORMANCE SCHEMA, USER VARIABLES
      
      Code cleanup
    ------------------------------------------------------------
    revno: 6953.2.10 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2014-06-10 16:10:39 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.9
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2014-06-10 14:38:16 +0200
    message:
      WL#6884 PERFORMANCE SCHEMA, USER VARIABLES
      
      Enforce a length limit on a user variable name
    ------------------------------------------------------------
    revno: 6953.2.8
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-04-11 11:38:07 +0200
    message:
      work in progress
    ------------------------------------------------------------
    revno: 6953.2.7 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2014-04-10 10:52:16 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.6 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Fri 2014-01-03 17:19:46 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.5 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2013-12-17 13:00:29 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.4 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Tue 2013-12-17 12:08:51 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.3
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2013-11-28 00:57:09 +0100
    message:
      WL#6884 PERFORMANCE SCHEMA, USER VARIABLES
      
      Work in progress
    ------------------------------------------------------------
    revno: 6953.2.2 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Mon 2013-11-25 16:05:58 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl6884
    ------------------------------------------------------------
    revno: 6953.2.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl6884
    timestamp: Thu 2013-11-14 00:08:16 +0100
    message:
      New branch
------------------------------------------------------------
revno: 8514
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-31 20:58:04 +0530
message:
  Bug #16204823 INNODB WASTES 62 OUT OF EVERY 16384 PAGES
  
  Problem:
  
  One extent descriptor page contains information about page allocations in N
  number of extents. If page size is 16K (the default), then N = 256. Of these
  extents 1 or more is used for fragment pages and such extents are of the state
  XDES_FREE_FRAG or XDES_FULL_FRAG.  The remaining extents (XDES_FSEG) will
  belong to any one segment.
  
  Any segment can make use of the pages from the fragment extent.  But each
  segment can use only 32 of these fragment pages.  Because of this most of the
  pages of these fragment extents are wasted by InnoDB.
  
  Solution:
  
  Make it possible for a segment to make use of these fragment extent.  Introduce
  the new XDES_FSEG_FRAG state for the extent.  
  
  rb#5559 approved by Yasufumi.
------------------------------------------------------------
revno: 8513 [merge]
committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-31 12:40:24 +0200
message:
  Upmerge of the 5.5.39 build
    ------------------------------------------------------------
    revno: 3902.338.56 [merge]
    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-07-31 12:33:38 +0200
    message:
      Upmerge of the 5.5.39 build
        ------------------------------------------------------------
        revno: 2875.598.21 [merge]
        author: 
        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-07-31 12:30:05 +0200
        message:
          Merge from mysql-5.5.39-release
            ------------------------------------------------------------
            revno: 2875.618.7
            tags: mysql-5.5.39
            committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
            branch nick: revert_patch
            timestamp: Sat 2014-07-19 11:24:21 +0530
            message:
              WL#7219: Reverting the wl#7219 patch in mysql-5.5.39-release branch 
            ------------------------------------------------------------
            revno: 2875.618.6
            committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
            branch nick: mysql-5.5.39-release
            timestamp: Fri 2014-07-18 20:55:52 +0530
            message:
              WL#7219: Fixing PB2 failures.
            ------------------------------------------------------------
            revno: 2875.618.5
            committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
            branch nick: mysql-5.5.39-release
            timestamp: Thu 2014-07-17 19:21:56 +0530
            message:
              WL#7219: Pushing it to release 5.5.39-release branch
------------------------------------------------------------
revno: 8512 [merge]
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B18384260-trunk
timestamp: Thu 2014-07-31 13:01:54 +0300
message:
  merge
    ------------------------------------------------------------
    revno: 3902.338.55 [merge]
    committer: Georgi Kodinov <georgi.kodinov@oracle.com>
    branch nick: B18384260-5.6
    timestamp: Thu 2014-07-31 12:58:24 +0300
    message:
      merge
        ------------------------------------------------------------
        revno: 2875.598.20
        committer: Georgi Kodinov <georgi.kodinov@oracle.com>
        branch nick: B18384260-5.5
        timestamp: Thu 2014-07-31 12:52:49 +0300
        message:
          Bug #18384260: MULTIPLE SECURITY ISSUES IN CERTIFICATE VALIDATION
          
          the 5.5 version of the fix.
          Added a call to X509_verify_cert_error_string() into the client certificate
          verification code.
------------------------------------------------------------
revno: 8511
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-31 11:18:11 +0200
message:
  Followup for Bug#18805275 fix.
  
  Fixed problem for innodb.log_file test for Windows.
  There are race between io threads and page cleaners to exit only when abort initialize InnoDB.
  Especially, Windows native AIO is weak for the race.
  IO threads should not exit during page cleaners active, in accurate.
  
  ------------------------------------------------------------
  revno: 8489
  committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
  branch nick: mysql-trunk
  timestamp: Sat 2014-07-26 03:22:05 +0900
  message:
    Bug #18805275 : PAGE CLEANER SHOULD BE MULTI-THREADED ALSO FOR RECOVERY AND SHUTDOWN
  
    From 5.7, page cleaner is multi-threaded for performance scalability.
    But it is not used during shutdown and recovery phases.
  
    It should be multi-threaded during shutdown and recovery phases for their
    optimal performance.
  
    * The previous problem was fixed. (releasing the event objects was too early)
  
    Approved by Sunny in rb#5465
  ------------------------------------------------------------
------------------------------------------------------------
revno: 8510
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-6042
timestamp: Thu 2014-07-31 10:45:37 +0200
message:
  fix for lowercase issue in pb2
------------------------------------------------------------
revno: 8509
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Wed 2014-07-30 19:06:12 +1000
message:
  WL#8023 - Follow up fix.
  
  1. Change ha_innobase::prebuilt -> ha_innobase::m_prebuilt
  2. Remove enum prefix
  3. Remove the nonnull attribute from function declarations.
------------------------------------------------------------
revno: 8508 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-30 09:38:52 +0800
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.54
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-30 09:35:07 +0800
    message:
      fix Bug #19172013 - MEMCACHED DAEMON CRASHES AND RESTARTS ON SECOND ADD
      OPERATION 
                  
      rb://6067 approved by Sunny Bains
------------------------------------------------------------
revno: 8507
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-30 09:20:53 +0800
message:
  Bug #18963676 	MISSED TO HANDLE OOM ERROR WHILE ADDING A GTID INTO GTID SET
  
  In Gtid_state::update_on_commit(), missed to handle 'out of memory'
  error while adding GTID into gtid set, which is risky.
  
  To the 'out of memory' error, we log the error and terminate the
  server after several tries. And update affected functions.
------------------------------------------------------------
revno: 8506
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-30 03:04:07 +0200
message:
  Fixed syntax for one of the disabled test lines.
------------------------------------------------------------
revno: 8505
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt2
timestamp: Tue 2014-07-29 19:35:35 +0200
message:
  post-push to the fix to bug#19184930:
    as part of the fix, the entire n test tree under
    internal/nroot/tests, was lost.
    Now it is restored.
------------------------------------------------------------
revno: 8504
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-07-29 09:33:00 -0600
message:
  Fix a few compiler warnings that have been introduced to mysql-trunk
  in the last couple months, seen with MSVC 2013.
  
  Approved by Sunny in RB#6153
------------------------------------------------------------
revno: 8503
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: test-trunk
timestamp: Tue 2014-07-29 16:04:43 +0200
message:
  WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
  
  post-push test fix.
  
  performance schema digest is affected by new tokens in sql_yacc.yy/lex.h.
------------------------------------------------------------
revno: 8502 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-29 19:06:27 +0530
message:
  Bug#18860225 HASH_SCAN SEEMS BROKEN: CAN'T FIND RECORD IN 'T1' ERROR_CODE: 1032
  
  Fixing windows build failure
  (Merging fix from mysql-5.6)
    ------------------------------------------------------------
    revno: 3902.338.53
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-29 19:04:33 +0530
    message:
      Bug#18860225 HASH_SCAN SEEMS BROKEN: CAN'T FIND RECORD IN 'T1' ERROR_CODE: 1032
      
      Fixing windows build failure
------------------------------------------------------------
revno: 8501 [merge]
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: wl5630
timestamp: Tue 2014-07-29 12:43:39 +0200
message:
  WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
  
  Merge into updated mysql-trunk
    ------------------------------------------------------------
    revno: 8460.1.1 [merge]
    committer: Mattias Jonsson <mattias.jonsson@oracle.com>
    branch nick: wl5630
    timestamp: Mon 2014-07-21 23:36:03 +0200
    message:
      WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
      
      Merge into updated mysql-trunk
        ------------------------------------------------------------
        revno: 8454.1.2
        committer: Mattias Jonsson <mattias.jonsson@oracle.com>
        branch nick: wl5630
        timestamp: Mon 2014-07-21 23:28:21 +0200
        message:
          WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
          
          Fix for previous patch for bug#19075764.
          There where a case where m_new_file was NULL
          and m_added_file was set, which caused dereferning
          the NULL pointer.
        ------------------------------------------------------------
        revno: 8454.1.1 [merge]
        committer: Mattias Jonsson <mattias.jonsson@oracle.com>
        branch nick: wl5630
        timestamp: Sat 2014-07-19 23:44:49 +0200
        message:
          WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
          
          Merge into updated mysql-trunk
            ------------------------------------------------------------
            revno: 8057.1.6
            committer: Mattias Jonsson <mattias.jonsson@oracle.com>
            branch nick: wl5630
            timestamp: Thu 2014-07-17 11:21:49 +0200
            message:
              Bug#19075764: ALTER... REBUILD WITH MISPLACED ROWS IS CRASHING IN MYISAM.
              
              Problem was bad error handling. When copy_partitions failed it called
              cleanup_new_partition() which closed the added partitions. Later when
              handle_alter_part_error was called, it would wait for all open table instances
              which also called handler::extra(HA_EXTRA_FORCE_REOPEN).
              That call would then try to reopen the added partitions that was
              closed in cleanup_new_partition.
              
              By removing the m_added_file handlers from m_new_file this
              problem was avoided.
              
              Test changes will follow.
            ------------------------------------------------------------
            revno: 8057.1.5
            committer: horst.hunger@oracle.com
            branch nick: mysql-trunk-wl5630
            timestamp: Wed 2014-07-02 20:02:38 +0200
            message:
              WL#5630: QA sign off tests for mtr.
            ------------------------------------------------------------
            revno: 8057.1.4
            committer: horst.hunger@oracle.com
            branch nick: mysql-trunk-wl5630
            timestamp: Fri 2014-06-27 08:45:41 +0200
            message:
              Tests for wl#5630
            ------------------------------------------------------------
            revno: 8057.1.3
            committer: Mattias Jonsson <mattias.jonsson@oracle.com>
            branch nick: wl5630
            timestamp: Wed 2014-05-28 12:39:49 +0200
            message:
              WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
              
              Fixed a bug with REBUILD/REORGANIZE PARTITION when
              the partition had misplaced rows.
            ------------------------------------------------------------
            revno: 8057.1.2
            committer: Mattias Jonsson <mattias.jonsson@oracle.com>
            branch nick: wl5630
            timestamp: Mon 2014-05-26 12:59:06 +0200
            message:
              WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
              
              Updated according to review comments in rb#5467.
            ------------------------------------------------------------
            revno: 8057.1.1
            committer: Mattias Jonsson <mattias.jonsson@oracle.com>
            branch nick: wl5630
            timestamp: Wed 2014-05-21 11:57:05 +0200
            message:
              WL#5630: Allow WITHOUT VALIDATION on EXCHANGE PARTITION
              
              Adds the possibility to skip validation of rows when EXCHANGING a table into a partition.
              
              Adds WITHOUT and VALIDATION as reserved words.
------------------------------------------------------------
revno: 8500
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-6042
timestamp: Tue 2014-07-29 11:55:16 +0200
message:
  Fix for Bug#19307777 ASSERTION `QEP_TAB->CONDITION() == QEP_TAB->CONDITION_OPTIM()' FAILED:
  qep_tab's condition_optim() is set in make_tmp_tables_info().
  In this testcase, JOIN::optimize leaves early with "impossible HAVING", no call
  to make_tmp_tables_info; so qep_tab's condition_optim() is left unset (NULL pointer).
  Then in JOIN::exec(), prepare_result() calls do_fill_table()
  which has this new assertion:
  
    /*
      We pass a condition, which can be used to do less file manipulations (for
      example, WHERE TABLE_SCHEMA='test' allows to open only directory 'test',
      not other database directories). Filling schema tables is done before
      QEP_TAB::sort_table() (=filesort, for ORDER BY), so we can trust
      that condition() is complete, has not been zeroed by filesort:
    */
    DBUG_ASSERT(qep_tab->condition() == qep_tab->condition_optim());
    ... code which uses condition() ...
  (Before the WL, this same code was using:
  "filesort && filesort->select ? filesort->select->cond : condition()" ).
  
  So it fails, because condition() is Item_equal while condition_optim() is still
  unset (NULL).
  Fix: no matter how we leave JOIN::optimize() we must set condition_optim()
  (except in error paths), this is here achieved by extending set_plan_state()
  (which is logical, as condition_optim() is also used by EXPLAIN).
  With this patch, there are less places where *optim variables are updated,
  which is better for maintenance.
  A comment is added for having_for_explain.
------------------------------------------------------------
revno: 8499
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-29 17:32:48 +0800
message:
  Fix Bug #19304815 - FIX INNODB_GIS_ROW_FORMAT TEST FAILURE 
  
  rb://6150 approved by Sunny Bains
------------------------------------------------------------
revno: 8498
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-6042
timestamp: Tue 2014-07-29 10:48:03 +0200
message:
  fix for build failure
------------------------------------------------------------
revno: 8497
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-29 06:44:46 +0200
message:
  Moving test from experimenatl to dsiabled list since the test fails consistently
------------------------------------------------------------
revno: 8496
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-6042
timestamp: Mon 2014-07-28 20:53:13 +0200
message:
  fix for compilation error, adding back "=0" like before wl#6042.
------------------------------------------------------------
revno: 8495
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-6042
timestamp: Mon 2014-07-28 15:29:25 +0200
message:
  wl#6042 "split JOIN_TAB into optimizer and executor part".
  The "executor part" is a new class QEP_TAB. JOIN_TABs are dropped at end
  of JOIN::optimize. EXPLAIN and JOIN::exec use only QEP_TABs. So does the
  construction of the AQP.
  QEP_TAB and JOIN_TAB share certain members, put in a class QEP_shared.
  SQL_SELECT is removed, most of it goes into QEP_shared or QEP_TAB.
  The memcpy of all JOIN_TABs at end of get_best_combination() is gone;
  instead, we keep the old JOIN_TABs for a bit longer, but access them
  in the final plan's order thanks to best_ref[] pointers.
  This refactoring fixes
  Bug#18921626 DEBUG CRASH IN PLAN_CHANGE_WATCHDOG::~PLAN_CHANGE_WATCHDOG AT SQL_OPTIMIZER.CC
  and Bug#18535226 DEBUG CRASH ON QUICK_RANGE_SELECT::RESET.
------------------------------------------------------------
revno: 8494 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-28 18:43:44 +0530
message:
  Bug#18860225 HASH_SCAN SEEMS BROKEN: CAN'T FIND RECORD IN 'T1' ERROR_CODE: 1032
  
  Merge fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.52
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-28 18:41:22 +0530
    message:
      Bug#18860225 HASH_SCAN SEEMS BROKEN: CAN'T FIND RECORD IN 'T1' ERROR_CODE: 1032
      
      Problem: In RBR, on slave setting slave_search_algorithm to HASH_SCAN causes "Can't find
      record" error even though that record exists in storage layer.
      
      Analysis: When slave's slave_search_algorithm is set to Hash_scan, it prepares
      a unique key list in add_key_to_distinct_keyset() for all the rows in a
      particular Row_event.
      For eg: Lets say a table 't1(c1,c2, key(c2))' contains 4 tuples
      (1,10), (2,20), (3,10), (4,40). And lets say on Master a delete
      is executed to delete all these four tuples. Received Rows_log_event on slave
      contains all these 4 tuples. Slave should prepare a unique key list(i.e., 10,20,40).
      Later, the same unique key list will be used to retrieve all tuples associated
      to each key in the list from storage engine. In the old code, there was a problem
      while preparing this unique key list in  add_key_to_distinct_keyset. Before
      inserting an element in to the unique list, instead of searching the entire list whether an
      element is there or not, it was comparing only with the last inserted element.
      This can create problem when the processing list is not a sorted array.
      For eg: While processing 10,20,10,40 list, add_key_to_distinct_keyset() list
      thinks that the third 10 is unique as it compares with last inserted key which
      is 20. Hence unique key list prepared will be 10,20,10,40 instead of 10,20,40.
      
      Later in Rows_log_event::do_apply_event, we retrieve all the tuples with the first
      key in the list (10) and will apply the event on those tuples (which is delete in
      this example). When the logic retrieves '10' again from wrongly formed unique list,
      it tries to find the tuples from storage engine using the key value '10'. Thus
      storage engine returns 'ER_RECORD_NOT_FOUND' error.
      
      Fix: Correcting add_key_to_distinct_keyset() logic to maintain unique key list
      i.e., instead of comparing the newly inserted value with only last inserted key,
      now the list (unique list) will be maintained by std::set. To compare two keys,
      We have a compare class which will be supplied to std::set. This compare class
      compares two given keys using key_cmp2 function which is added in key.cc file.
------------------------------------------------------------
revno: 8493 [merge]
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-28 11:43:24 +0400
message:
  Merge from mysql-5.6.
    ------------------------------------------------------------
    revno: 3902.338.51 [merge]
    committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-28 11:36:07 +0400
    message:
      Merge from mysql-5.5.
        ------------------------------------------------------------
        revno: 2875.598.19
        committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-07-28 11:19:19 +0400
        message:
          Fix for bug#19184973: PB2 AUDIT_LOG_ROTATE_VALID TIMED OUT
          
          Call cond_signal(COND_data) in alog_close() under protection of
          the same mutex used with cond_wait(COND_data) in the writer thread.
          Otherwise, the condition might be signalled just before
          cond_wait(COND_data) which leads to infinite wait.
------------------------------------------------------------
revno: 8492
committer: Amit Bhattacharya<amit.bhattacharya@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-28 11:36:37 +0530
message:
  Fixed the costmodel_planchange test failures
------------------------------------------------------------
revno: 8491
committer: Amit Bhattacharya<amit.bhattacharya@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-28 11:35:37 +0530
message:
  Fixed the costmodel_planchange test failures
------------------------------------------------------------
revno: 8490 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-07-26 22:21:44 +0900
message:
  Merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.338.50
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Sat 2014-07-26 22:17:23 +0900
    message:
      Bug#18631496 : MULTIPLE PURGE THREADS - CORRUPTION - ERROR IN SEC INDEX ENTRY UPDATE
      
      Though seems very rare case, purge process might attempt to delete secondary index record which is not marked as deleted.
      
      If not marked as deleted, the record must not be deleted. Because the deleting causes the data corruption (inconsistency at the secondary index).
      
      This fix should avoid ROW_NOT_FOUND at row_upd_sec_index_entry() reported at Bug#18631496 which causes data corruption.
      
      Approved by Krunal in rb#6062
------------------------------------------------------------
revno: 8489
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-07-26 03:22:05 +0900
message:
  Bug #18805275 : PAGE CLEANER SHOULD BE MULTI-THREADED ALSO FOR RECOVERY AND SHUTDOWN
  
  From 5.7, page cleaner is multi-threaded for performance scalability.
  But it is not used during shutdown and recovery phases.
  
  It should be multi-threaded during shutdown and recovery phases for their
  optimal performance.
  
  * The previous problem was fixed. (releasing the event objects was too early)
  
  Approved by Sunny in rb#5465
------------------------------------------------------------
revno: 8488
committer: Satya Bodapati <satya.bodapati@oracle.com>
branch nick: mysql-trunk-plain
timestamp: Fri 2014-07-25 18:16:54 +0530
message:
  Fix typo which blocks innodb build with UNIV_ZIP_DEBUG
  
  Approved by Vasil on IM.
------------------------------------------------------------
revno: 8487
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-25 10:02:57 +0530
message:
  Bug#19129559 -  pb2 main.max_statement_time.test crashing the
                  server in weekly trunk.
  
  Analysis:
  ---------
  This test is failing because few reasons,
  
  1. Issue mentioned in the bug report: (Started failing on Window
     from some time 21 June)
     
    qld-debug.exe!timer_notify_thread_func()[win_timers.c:78]
    7ff74fe2b42e    mysqld-debug.exe!pfs_spawn_thread()[pfs.cc:2074]
    7ff74f92d306    mysqld-debug.exe!pthread_start()[my_winthread.c:73]
    7ff74ff964f5    mysqld-debug.exe!_callthreadstartex()[threadex.c:376]
    7ff74ff96747    mysqld-debug.exe!_threadstartex()[threadex.c:359]
    7ff86d0015dd    KERNEL32.DLL!BaseThreadInitThunk()
    7ff86f6343d1    ntdll.dll!RtlUserThreadStart()
  
    => This issue is seen on Windows 64 bit machines. In function 
       "timer_notify_thread_func" pointer value of timer object is
       stored in ULONG variable then type casted to pointer of 
       my_timer_t. This works well in 32 bit machine and 64 bit machine
       until address crosses 2^32 value. 
       Windows 64 is of LLP64 model. So ULONG variable is of 32 bit 
       and pointer is of 64 bit. Because of this pointer value was 
       getting corrupted when 64 bit value is stored in 32 bit ULONG
       type variable. Accessing this memory was causing an issue on
       Windows 64.
  
       To fix this issue, changed type of variable from ULONG to
       ULONG_PTR.
  
  2.  (Sporadic failure)
      mysqltest: At line 472: query 'SELECT CONVERT(VARIABLE_VALUE, UNSIGNED) INTO @qc_hits
                                     FROM INFORMATION_SCHEMA.GLOBAL_STATUS
                                     WHERE VARIABLE_NAME = 'Qcache_hits''
      failed: 1909: Query execution was interrupted, max_statement_time
                    exceeded
  
     => Issue here is, max_statement_time is set 50 milliseconds
        and assumption was "SELECT CONVERT(VARIABLE_NAME,...." 
        returns withing 50 millisecond. But on some slow platforms
        it took more than that so query terminated.
  
        To fix this issues, Changed session max_statement_time to 0. 
  
  
  3.  (Sporadic failure)
     mysqltest: At line 119: query 'SELECT * from t1' succeeded -
               should have failed with errno 1909...
  
     =>Issue here is, max_statement_time at session level is set
       to 2 milliseconds and query "SELECT * FROM..." is executed.
       Here, expectation is SELECT will take more than 2 milliseconds
       but on some platforms it returned before 2 millisecond.
       So it is not timed out.
  
       To fix this issues, changed query to select *, sleep(0.5) FROM t1;
       Now sleep(0.5) is executed for each row. This will make query
       execution time greater than 2 microsecond.
  
  4.  (Sporadic failure)
      SET @@SESSION.max_statement_time = 950;
      SELECT SLEEP(1);
       SLEEP(1)
       -1
       +0
  
     => Test case 10, checks timer precision. In this test case, 
        query calls sleep(1) function and for that timer with
        different timeout values are set. And expectation is, if
        timer is set for less than 1 sec then sleep should be 
        interrupted.
  
        But, max_statement_time is a soft hint for queries, on loaded
        machines or slow machines time taken notify thread after
        timer expiration would be little bit more. Because of this
        sleep() is not interrupted.
  
        We can reduce statement execution timeout value, but test
        case to check timer precision will be incorrect. Not 
        changing test case will result in sporadic test case
        failure. Removed this test case to avoid sporadic failure.
  
  
  5. (sporadic failure)
     @@ -151,7 +151,7 @@
     ERROR HY000: Query execution was interrupted, max_statement_time exceeded
     SELECT @a;
      @a
      -1
     +NULL
    
     => Here we set sessions level max_statement_time to 2 and
        execute stored function having "SELECT SLEEP(2) INTO @a".
        After SELECT query interruption value of variable @a is 
        printed. But some times, the SELECT is interrupted before
        populating variable. 
  
        Actually test case is verified by checking error reported. 
        We need not have to print variable value here. To avoid this
        failure, removed "SELECT @a" statement.
------------------------------------------------------------
revno: 8486
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Fri 2014-07-25 10:36:37 +0800
message:
  Bug#19164038 - INNODB: CLUSTER INDEX BUILD BY BULK LOAD WHEN SKIP SORT
  SHOULD RELEASE LATCHES
  
  When we do table rebuild and we can skip sort for cluster index build
  (rb#4788), if there is a spatial index, we need the cluster index BtrBulk
  object to release latches once it finishes a batch of insert, because in
  spatial index build, we may call 'log_free_check'.
  
  rb://6087 approved by Jimmy.
------------------------------------------------------------
revno: 8485
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk
timestamp: Thu 2014-07-24 21:44:34 +0400
message:
  WL#7976: Deprecate skip-innodb in 5.6, remove in 5.7.
  
  Follow-up patch to fix new test failures:
  
    - perfschema.no_threads failed on Windows only.
  
      The problem was that the number of P_S-thread-instances
      has increased when InnoDB was enabled. The test limited
      that number by 10, which was not enough on Windows.
  
      The fix is to increase the limit to 20.
  
    - rpl_myisam_recovery
  
      skip-innodb option was missed.
------------------------------------------------------------
revno: 8484 [merge]
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: wl8006-merge-trunk
timestamp: Thu 2014-07-24 16:51:00 +0300
message:
  merge WL#8006 to trunk
    ------------------------------------------------------------
    revno: 8424.1.1
    committer: Georgi Kodinov <georgi.kodinov@oracle.com>
    branch nick: wl8006-trunk
    timestamp: Thu 2014-07-24 10:56:30 +0300
    message:
      WL#8006: remove mysql_old_password
      
      * Marked server-side --secure-auth as a "zombie" option: can't set it to anything but 1. Throws a warning when used.
      * Marked client-side --secure-auth option as a "zombie" option. Returns a deprecation warning when used.And returns an error is anything but 1 is specified.
      * Removed mysql_old_passwords server plugin
      * Removed mysql_old_passwords libmysql plugin
      * old_passwords server side option does not allow setting to 1. So it's only 0 and 2
      * Removed the OLD_PASSWORD() SQL function.
      * Multiple internal special cases clean up from the server.
      * Cleanup of the test cases testing old_password
      * Update of the relevant test cases to cover the new functionality.
      * Created a macro for the deprecation warning in the client tools. 
      * Explictly marked --secure-auth as zombie in the help text too.
      * Removed all the calls to mysql_options(MYSQL_SECURE_AUTH) from the client tools.
      * Added a check to mysql_options(MYSQL_SECURE_AUTH) to generate an error if value is anything but 1
      * Removed the SKIPSECUREAUTH mysqltest connection option.
      * Cleaned up trailing space. 
      * Removed references to old_plugin from scripts/
------------------------------------------------------------
revno: 8483
committer: Erik Froseth <erik.froseth@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-24 14:18:54 +0200
message:
  WL#7928 Geohash encoding and decoding functions
  
  Implementation of four new functions for encoding and decoding geohash data: ST_GEOHASH, ST_LONGFROMGEOHASH, ST_LATFROMGEOHASH and ST_POINTFROMGEOHASH. These functions will make it possible for applications to import and export data from MySQL using the geohash format. It will also make it possible to index and search for geographic data in one-dimensional indexes.
  
  New tests for verifying these functions are also added.
------------------------------------------------------------
revno: 8482
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-24 19:20:49 +0800
message:
  Bug#18410442 CREATE A BASE CLASS FOR GTID_TABLE_ACCESS_CONTEXT AND RPL_INFO_TABLE_ACCESS
  
  Post-fix: Rename the class 'Table_access' with 'System_table_access',
  and declare the class 'System_table_access' in a new replication file
  named 'rpl_table_access.h' and define it in rpl_table_access.cc.
------------------------------------------------------------
revno: 8481 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-24 16:32:53 +0530
message:
  Bug#18684222  CURRENT_THD CALLS CAN BE MINIMIZED IN
  ACCESS::CREATE_THD()/DROP_THD FUNCTIONS
  
  Merge fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.49
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6-release
    timestamp: Thu 2014-07-24 15:36:54 +0530
    message:
      Bug#18684222  CURRENT_THD CALLS CAN BE MINIMIZED IN
      ACCESS::CREATE_THD()/DROP_THD FUNCTIONS
      
      Problem: Current_thd is costly operation and it is getting
      called in Rpl_info_table_access:create_thd/drop_thd functions
      many times and it can be avoided.
      
      Analysis: 
      In case of RPL_INFO_REPOSITORY=TABLE, Server try to open and
      close repository table. In the process of opening and closing
      the table, there are few calls to costly macro current_thd
      which can be avoided by changing the logic little.
      
      Fix: Replacing save_current_thd logic with 'thd_created'
      (bool) variable.
------------------------------------------------------------
revno: 8480
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-24 19:04:04 +0900
message:
  Bug#19196052 : SHOW ENGINE INNODB STATUS CAUSES CRASH
  
  trx_print_low() should not call innobase_mysql_print_thd(), if the transaction is not active.
  The thd for it might not exist already in mysqld.
  
  Approved by Sunny by IM
------------------------------------------------------------
revno: 8479
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-24 19:03:14 +0900
message:
  Bug#19183264 : ONLINE BUFFER POOL RESIZE MESSAGE USES WRONG PRINTF SPECIFIER ON WINDOWS
  
  printf for ulint which might be over 4G should use ULINTPF, should not use %lu directly for WIN64.
  WL6117 uses %lu for buffer pool size and chunk size.
  They should be fixed with ULINTPF.
  
  Approved by Vasil in rb#6010
------------------------------------------------------------
revno: 8478
committer: Satya Bodapati <satya.bodapati@oracle.com>
branch nick: mysql-trunk-plain
timestamp: Thu 2014-07-24 13:27:53 +0530
message:
  Problem:
  ========
  The trx returned from the trx pool is not in clean state.
  This is verified by adding assert ut_ad(trx->will_lock ==0)
  in trx create function.
  
  1) trx->will_lock is incremented after trx_commit
  
  i)   ha_commit_trans
  ii)  close_thread_tables->mysql_unlock_tables->unlock_external->
       ha_external_lock(.., F_UNLCK)->ha_innobase::external_lock(.. F_UNLCK)
  iii) in ha_innobase::external_lock() : this increments trx->will_lock
          if (!trx_is_started(trx)
              && (prebuilt->select_lock_type != LOCK_NONE
                  || prebuilt->stored_select_lock_type != LOCK_NONE)) {
  
                  ++trx->will_lock;
          }
  
  2)  in Functions like ha_innobase::create(), ha_innobase::delete_table(),
      etc, trx->will_lock is incremented but
      the decrement may or may not happen.
  
     i)   If functions called by ha_innobase::create() do a
          trx_commit()/rollback, trx->will_lock is reset, there is
          no problem in this case
     ii)  Sometimes(incase of failure) the trx is not started by
          ha_innobase::create()/rollback(), trx state remains as
          TRX_STATE_NOT_STARTED.
     iii) The following trx commit or rollback will not reset
          trx->will_lock because the trx state is TRX_STATE_NOT_STARTED
  
  Fix:
  ====
  1) This is solved by not incrementing the trx->will_lock counter when we
     are unlocking.
  2) This is solved by resetting trx->will_lock counter when the
     trx->state remains as TRX_STATE_NOT_STARTED
  
  Approved by Sunny. rb#6068
------------------------------------------------------------
revno: 8477 [merge]
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-07-23 16:56:29 +0200
message:
  Merge from working tree.
    ------------------------------------------------------------
    revno: 8456.1.1
    committer: horst.hunger@oracle.com
    branch nick: mysql-trunk-mats-maintenance
    timestamp: Wed 2014-07-23 16:43:50 +0200
    message:
      Increased coverage to again 69%. Put suite into default.push to run it on stage.Updated README with instructions for maintenance.
------------------------------------------------------------
revno: 8476
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-23 15:04:13 +0300
message:
  Fix Bug#14502665 INNODB.INNODB_16K TEST FAILS DUE TO DIFFERENT ORDERING
  IN RESULT OUTPUT
  
  Make the output from a SELECT query deterministic by adding ORDER BY.
  
  Approved by:	Sunny (via IM)
------------------------------------------------------------
revno: 8475 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Wed 2014-07-23 18:52:35 +0800
message:
  merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.48
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-23 18:28:17 +0800
    message:
      Fix Bug 16869435 - IN-PLACE ALTER TABLE FOR TABLE WITH BROKEN FK CAUSES ASSERT
      
      rb://5653 approved by Marko
------------------------------------------------------------
revno: 8474 [merge]
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-23 16:14:48 +0800
message:
  Bug#19172212	InnoDB Memcached Plugin "gets" command
  
  Merge from 5.6.
    ------------------------------------------------------------
    revno: 3902.338.47
    committer: Allen lai <zheng.lai@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-23 16:10:01 +0800
    message:
      Bug#19172212	InnoDB Memcached Plugin "gets" command
      
      We don't support multi-gets yet. This will be addressed in worklog 6650.
      We block multi-gets for now. Just return an error message to client
      instead of returning wrong results.
      
      Approved by Jimmy on IM.
------------------------------------------------------------
revno: 8473
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk-wl7976-6
timestamp: Wed 2014-07-23 09:04:03 +0400
message:
  WL#7976: Deprecate skip-innodb in 5.6, remove in 5.7.
  
  This patch marks the InnoDB plugin as mandatory, meaning that
  there is no way to disable it from the configuration (including
  command line options).
  
  Added a test case to ensure that there is no way to disable InnoDB.
------------------------------------------------------------
revno: 8472
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk-wl7976-6
timestamp: Wed 2014-07-23 09:02:56 +0400
message:
  WL#7976: Deprecate skip-innodb in 5.6, remove in 5.7
  
  Updated tests to remove all options which disable InnoDB
  (skip-innodb, loose-skip-innodb, innodb=OFF, loose-innodb).
  
  Updated tests to remove the 'innodb' option which explicitly
  enable InnoDB.
  
  Updated MTR: indicate that InnoDB is always enabled.
------------------------------------------------------------
revno: 8471 [merge]
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk-wl7976-6
timestamp: Wed 2014-07-23 09:01:28 +0400
message:
  Null-merge from 5.6. The patch for 5.7 will be different.
    ------------------------------------------------------------
    revno: 3902.338.46
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: 5.6-wl7976-2
    timestamp: Wed 2014-07-23 08:53:48 +0400
    message:
      WL#7976: Deprecate skip-innodb in 5.6, remove in 5.7.
      
      The patch adds a deprecation warning to all forms of the 'skip-innodb'
      option used to disable or enable InnoDB.
------------------------------------------------------------
revno: 8470
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-23 02:41:11 +0200
message:
  Making unstable tests experimental
------------------------------------------------------------
revno: 8469
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk-wl7293-3
timestamp: Tue 2014-07-22 16:48:41 +0400
message:
  WL#7293: Make sure that when 32-bit connection-IDs roll over, still in-use
  IDs are not re-issued
  
  Committed on behalf of Jon Olav Hauglid <jon.hauglid@oracle.com>.
  
  With this patch the currently used thread IDs are maintained in a sorted
  list so that it can be checked that each new connection is assigned an ID
  which is not currently in use. This prevents reuse of active IDs in cases
  where the 32 bit counter wraps around. The counter is kept at 32 bit as it
  is part of the client protocol.
  
  The patch makes code more consistent by using the my_pthread_id typedef
  instead of various uint/ulong variants.
  
  The patch also changes the thread_created counter to be incremented using
  atomics rather than protected by a separate mutex. The list of active THDs
  is changed from map to a sorted array.
  
  Combined this gives no reduced connect/disconnect performance even with the
  added processing required to assign unique IDs.
------------------------------------------------------------
revno: 8468
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-22 18:06:46 +0800
message:
  Fix Bug 19204759 - DAEMON_MEMCACHED_OPTION NOT SHOWN CORRECTLY
  
  rb://6072 approved by Sunny Bains
------------------------------------------------------------
revno: 8467 [merge]
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Tue 2014-07-22 09:58:18 +0200
message:
  Null-merge for backporting.
    ------------------------------------------------------------
    revno: 3902.338.45
    committer: Benny.Wang<Benny.Wang@oracle.com>
    branch nick: mysql5.6
    timestamp: Tue 2014-07-22 09:36:13 +0200
    message:
      Backported from trunk.
      BUG#18447874: WRONG RESULT COMING FROM SEMI-JOIN
      BUG#18223655: ASSERTION FAILED: (INT)IDX >= 0 && IDX < PARENT_JOIN->TABLES
      If the left-hand expression(LHE) of IN predicate is a scalar subquery, E.g
      select  1 from `b`  where (select 1 from  d ) in (select `d` from `c`).
      In current version, during transforming In-predicate into semi-join,
      the select list of LHE and RHE will be extracted to construct a new equivalent
      expression as the where_cond to be added to Outer-query. As for the example,
      IN-predicate (select 1 from  d ) in (select `d` from `c`) will make a new
      equal-condition like 1='d'. However, such a transformation is correct only if
      table 'd' has one row. If it has an empty resultset, such a transformation is
      incorrect. Moreover, by executing the scalar query to see whether the resultset
      is empty or not is not allowed during resolve.
      
      Actually, scalar query is parsed into an Item_cache object So in current patch,
      the where_cond will be constructed as the followings, Item_cache<(select 1 from
      d )> = Item_row('d'). During execution, Item_cache will be filled by executing
      the scalar query, which will make the query get the right result.
------------------------------------------------------------
revno: 8466
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: b19051306-trunk
timestamp: Tue 2014-07-22 11:31:05 +0400
message:
  Fix for bug #19051306: EXTRACTVALUE: STACK OVERFLOW IN MY_XPATH_PARSE_UNARYEXPR 
  WITH MANY MINUSES
  
  Stack usage should be checked in my_xpath_parse_UnaryExpr() to avoid overflow.
------------------------------------------------------------
revno: 8465
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-22 08:57:46 +0200
message:
  The test was added to disabled.def file in another commit. Fails consistently and hence disabled is a better state
------------------------------------------------------------
revno: 8464
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: b19056196-trunk
timestamp: Tue 2014-07-22 10:43:51 +0400
message:
  Fix for bug#19056196: CRASH FROM XML NUMBER FUNCTION
  
  Wrong XPATH "number" function's number of parameters leads to the crash. 
------------------------------------------------------------
revno: 8463
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-22 08:03:23 +0200
message:
  Making tests that fail frequently disabled/experimental
------------------------------------------------------------
revno: 8462
committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-22 11:21:03 +0530
message:
  making rpl.rpl_server_uuid experimental on windows
------------------------------------------------------------
revno: 8461
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Tue 2014-07-22 02:45:27 +0200
message:
  BUG#18447874: WRONG RESULT COMING FROM SEMI-JOIN
  BUG#18223655: ASSERTION FAILED: (INT)IDX >= 0 && IDX < PARENT_JOIN->TABLES
  If the left-hand expression(LHE) of IN predicate is a scalar subquery, E.g
  select  1 from `b`  where (select 1 from  d ) in (select `d` from `c`).
  In current version, during transforming In-predicate into semi-join,
  the select list of LHE and RHE will be extracted to construct a new equivalent
  expression as the where_cond to be added to Outer-query. As for the example,
  IN-predicate (select 1 from  d ) in (select `d` from `c`) will make a new
  equal-condition like 1='d'. However, such a transformation is correct only if
  table 'd' has one row. If it has an empty resultset, such a transformation is
  incorrect. Moreover, by executing the scalar query to see whether the resultset
  is empty or not is not allowed during resolve.
  
  Actually, scalar query is parsed into an Item_cache object So in current patch,
  the where_cond will be constructed as the followings, Item_cache<(select 1 from
  d )> = Item_row('d'). During execution, Item_cache will be filled by executing
  the scalar query, which will make the query get the right result.
------------------------------------------------------------
revno: 8460
committer: Manish Kumar<manish.4.kumar@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-21 22:20:01 +0530
message:
  WL#7796 - WAIT_FOR_EXECUTED_GTID_SET
  
  This worklog aims at adding a new sql function WAIT_FOR_EXECUTED_GTID_SET which 
  make the current syncing option for the slave with master independent of the 
  slave threads. 
  
  The use of the new function will be :
  
  WAIT_FOR_EXECUTED_GTID_SET(GTID_SET [, TIMEOUT])
  
  Wait until all the transactions whose global transaction identifiers are being 
  specified, are contained in gtid_set.
  
  ie. After this function returns with success, it must hold that 
      GTID_SUBSET(GTID_SET, @@global.gtid_executed) returns true.
------------------------------------------------------------
revno: 8459 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-17512527
timestamp: Mon 2014-07-21 21:48:42 +0530
message:
  Merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.338.44
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-17512527
    timestamp: Mon 2014-07-21 21:21:15 +0530
    message:
      BUG#17512527: LIST HANDLING INCORRECT IN MYSQL_PRUNE_STMT_LIST()
      
      Analysis:
      ---------
      Invalid memory access maybe observed when using prepared statements if:
      a) The mysql client connection is lost after statement preparation
         is complete and
      b) There is at least one statement which is in initialized state but
         not prepared yet.
      
      When the client detects a closed connection, it calls end_server()
      to shutdown the connection. As part of the clean up, the
      mysql_prune_stmt_list() removes the statements which has transitioned
      beyond the initialized state and retains only the statements which
      are in a initialized state. During this processing, the initialized
      statements are moved from 'mysql->stmts' to a temporary 'pruned_list'.
      When moving the first 'INIT_DONE' element to the pruned_list,
      'element->next' is set to NULL. Hence the rest of the list is never
      traversed and the statements which have transitioned beyond the
      initialized state are never invalidated.
      
      When the mysql_stmt_close() is called for the statement which is not
      invalidated; the statements list is updated in order to remove the
      statement. This would end up accessing freed memory(freed by the
      mysql_stmt_close() for a previous statement in the list).
      
      Fix:
      ---
      mysql_prune_stmt_list() called list_add() incorrectly to create a
      temporary list. The use case of list_add() is to add a single
      element to the front of the doubly linked list.
      mysql_prune_stmt_list() called list_add() by passing an entire
      list as the 'element'.
      
      mysql_prune_stmt_list() now uses list_delete() to remove the
      statement which has transitioned beyond the initialized phase.
      Thus the statement list would contain only elements where the
      the state of the statement is initialized.
      
      Note: Run the test with valgrind-mysqltest and leak-check=full
      option to see the invalid memory access.
------------------------------------------------------------
revno: 8458
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Mon 2014-07-21 22:17:16 +1000
message:
  WL#8023 - Follow up fix. Destroy the old vector allocator when a transaction
  is freed. Call the destructor for trx->lock.table_locks() when the transaction
  is destroyed.
------------------------------------------------------------
revno: 8457
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Mon 2014-07-21 18:42:42 +1000
message:
  WL#8023 - Refactoring for WL#6835
  
  Add a stub to track whether the transaction is executing inside InnoDB code.
  
  /** Track if a transaction is executing inside InnoDB code */
  class TrxInInnoDB {
  public:
  	TrxInInnoDB(trx_t* trx) { }
  	~TrxInInnoDB() { }
  	static bool is_aborted() { return(false); }
  };
  
  is_aborted() : is hard coded to return false and declared static so that
  	       the new code paths are *never* executed.
  
  This class will be fleshed out in WL#6835.
  
  Cleanup ha_innodb.cc and ha_innodb.h. Bring it in line with our coding
  guidelines. Remove "nonnull" attributes from the code. These are dangerous
  and provide no benefit whatsoever. Minor code reformatting.
  
  rb#6048 Approved by Jimmy Yang and Allen.
------------------------------------------------------------
revno: 8456
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-07-21 14:57:10 +0800
message:
  Bug#19233075 - MY_MBCHARLEN_PTR() SHOULD RETURN 0 INSTEAD OF ASSERT FOR
  INVALID GB18030 CODE
  
  my_mbcharlen_ptr() had better return 0 to indicate the input gb18030 code
  is invalid, like some my_mbcharlen does(e.g. utf8), so that it won't crash
  the server when it receives invalid code.
  
  rb#6065, approved by Ramil
------------------------------------------------------------
revno: 8455 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Mon 2014-07-21 11:41:22 +0530
message:
  Bug #17297324 GLIBC DOUBLE FREE OR CORRUPTION WHEN KILLING CLIENT; CTRL+C 
  
  Merging from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.43 [merge]
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-07-21 11:31:02 +0530
    message:
      Bug #17297324 GLIBC DOUBLE FREE OR CORRUPTION WHEN KILLING CLIENT; CTRL+C 
      
      Merging from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.18
        committer: Venkata Sidagam <venkata.sidagam@oracle.com>
        branch nick: 5.5
        timestamp: Mon 2014-07-21 11:26:50 +0530
        message:
          Bug #17297324 GLIBC DOUBLE FREE OR CORRUPTION WHEN KILLING CLIENT; CTRL+C 
          
          Description: Sometimes when killing the mysql command line client with
          KILL -2(SIGINT), mysql client core dumps as a result of a double free or
          corruption.
          
          Analysis: When we run the mysql client in command line mode it will goes
          to mysql_end() and frees many data structures. At the same time (i.e
          after some data structures are freed), if we give "KILL -2" signal then
          the signal will be handled with function handle_kill_signal() and as
          part of it will again calls mysql_end() and goes with free() to the
          already freed data structure for batch_readline_end() function, which
          causes core dump.
          
          Fix: Ignoring SIGQUIT and SIGINT signals when cleanup process starts.
          This will help in resolving the double free issues, which occurs 
          in case the signal handler function is started in between of the 
          clean up function.
          For 5.6 we need to ignore SIGHUP also.
------------------------------------------------------------
revno: 8454
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Fri 2014-07-18 15:14:31 +0200
message:
  Fix Windows cluster build break after recent merge from custer tree to trunk.
------------------------------------------------------------
revno: 8453 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug19244772_mysql-trunk
timestamp: Fri 2014-07-18 17:31:44 +0530
message:
  Bug#19244772:BACKPORT BUG #18923691 TO MYSQL-5.6
  
  Null merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.42
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug19244772_mysql-5.6
    timestamp: Fri 2014-07-18 17:18:17 +0530
    message:
      Bug#19244772:BACKPORT BUG #18923691 TO MYSQL-5.6
------------------------------------------------------------
revno: 8452 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-18 16:39:08 +0530
message:
  Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
  SHOW PROCESSLIST, SHOW BINLOGS
        
  Null-merge
    ------------------------------------------------------------
    revno: 3902.338.41
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-07-18 16:36:31 +0530
    message:
      Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
      SHOW PROCESSLIST, SHOW BINLOGS
            
      Post push fix (removing try-catch)
------------------------------------------------------------
revno: 8451
committer: Sergey Glukhov <sergey.glukhov@oracle.com>
branch nick: mysql-trunk-bugs
timestamp: Fri 2014-07-18 12:53:52 +0400
message:
  Fix for Bug#18924341 CRASH IN TEST_IF_SKIP_SORT_ORDER, GROUP BY MATCH AGAINST DESC
  Regression from WL#7123 Additional query optimization for Fulltext Search.
  Tetsing of tab->join->simple_order is insufficient to check
  the presence of ORDER BY clause. simple_order can be true even
  if there is no ORDER BY.
  Additonal check for presence of ORDER BY is added.
------------------------------------------------------------
revno: 8450
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-18 13:31:23 +0530
message:
  Increasing the timeout value for Expect .
------------------------------------------------------------
revno: 8449
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-07-18 14:06:56 +0800
message:
  BUG 19192364 - ONE_FLUSHED DEFINED BUT NOT USED
  
  Remove this unused variable. Approved by Sunny over IM.
------------------------------------------------------------
revno: 8448 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk-push
timestamp: Fri 2014-07-18 05:15:24 +0200
message:
  Merge of BUG#18472603 patch from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.40
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-07-18 04:35:26 +0200
    message:
      BUG#18472603 SQL THREAD COMMITS PARTIAL TRANSACTION WITH GTID AND
                   AUTO POSITIONING ENABLED
      
      Problem:
      =======
      
      When using GTID with auto positioning, if the IO thread restarts it
      might be stopped in the middle of a transaction and will ask the master
      for the whole transaction again.
      
      When the SQL thread reads this stream of events, it should rollback
      the partial transaction before starting the complete one, but it is
      only rolling back the previous GTID set, and setting the new GTID,
      letting the partial transaction untouched.
      
      The effect of this problem is that the SQL thread is committing the
      partial transaction left on the relay log before IO thread restart
      once it reaches the "BEGIN" of the transaction right after the partial
      transaction, consuming the GTID.
      
      When trying to commit the complete transaction, the SQL thread is
      stopping with ER_GTID_NEXT_TYPE_UNDEFINED_GROUP, because the GTID was
      already consumed by the partial transaction.
      
      By restating the SQL thread, it will try to apply the complete
      transaction again from the beginning. Unfortunately, SQL thread will
      silently skip the complete transaction as the GTID is already logged.
      
      
      Analysis:
      ========
      
      There were two issues related to the reported problem: one related to
      master's dump thread and how it clears the "created" variable of it's
      FD event before sending it to the slave; other related to SQL thread
      rolling back partial transactions relying only on master's FD event
      "created" variable.
      
      Master's dump thread clearing FD's "created" variable
      -----------------------------------------------------
      
      When not using GTID protocol, the dump thread sends the original
      master's binlog FD to the slave with it's original "created" variable
      set only if the slave asked for FD event position (4).
      
      If the slave asked for a position greater than 4, the dump thread
      assumes that it must clear the "created" variable before sending the
      FD event to the slave.
      
      So, the slave will only have this "created" set in master's FD event
      if the master has just (re)started.
      
      Assuming that if the master sent partial transaction before
      "crashing/recovering" and it also lost temporary tables, the "created"
      variable set in master's FD event will instruct the slave do two tasks:
      
      a) Rollback current transaction (if there is one);
      b) Drop the temporary tables;
      
      For GTID protocol, the master was always sending the original FD of the
      binlog. So, if a slave reconnected asking for transactions in the first
      master binlog file, it was receiving an FD event with the "created"
      variable set, making slave to rollback the current transaction (not a
      problem here, as the whole transaction would appear again in the relay
      log) and dropping the temporary tables (this was a problem).
      
      SQL thread rolling back partial transactions
      --------------------------------------------
      
      SQL thread relied on master's FD "created" variable to rollback partial
      transactions left on the relay log by a master crash, but there was no
      mechanism to do the same for partial transactions left on the relay log
      by the GTID protocol.
      
      The "created" flag on master's FD seemed to be an effective way but
      worked only when the master was in the first binlog after it's startup.
      
      After the first binlog rotation, and before restarting the server
      again, all master FD's will not have the "created" variable set. So, a
      slave reconnecting and asking for transactions to any master binlog
      file but the first didn't rolled back the partial transactions left
      by GTID protocol.
      
      The only effective way to SQL thread know that the IO thread has left
      partial transactions in the relay log when using GTID protocol is by
      verifying if the SQL thread has a partial transaction when applying a
      Gtid_log_event. In this case, the SQL thread should rollback the whole
      partial transaction, and not only the GTID.
      
      
      Fix:
      ===
      
      Master's dump thread must clear "created" variable on the FD event
      before sending it to slave if the slave already has the first
      transaction in the masters binlog file.
      
      SQL thread cannot rely only in master's FD event to rollback partial
      transactions left by GTID protocol on relay log. It must also rely on
      Gtid_log_events being applied before committing transactions. 
      
      @ sql/binlog.cc / sql/binlog.h
      
      Changed read_gtids_from_binlog() and find_first_log_not_in_gtid_set()
      to return the first GTID found on the binlog file if asked.
      
      @ sql/rpl_master.cc
      
      Added code to mysql_binlog_send() to obtain the first GTID of the
      binlog file to be sent to slave, and to verify if slave already has
      this first GTID before sending the FD event, cleaning the "created"
      variable in this case.
      
      @ sql/rpl_slave.cc
      
      Added a debug point to help test case design.
      
      @ sql/log_event.cc
      
      Added code to make the Gtid_log_event to rollback the incomplete
      transaction before rolling back the GTID.
------------------------------------------------------------
revno: 8447 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk-push
timestamp: Fri 2014-07-18 03:58:27 +0200
message:
  Merge of BUG#18652178 and BUG#18306199 patch from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.39
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-07-18 03:49:01 +0200
    message:
      BUG#18652178 STOP SQL_THREAD, START SQL_THREAD CAUSES A TRX TO LOG
                   WITH A DIFFERENT GTID
      BUG#18306199 START SLAVE UNTIL MASTER_LOG_POS=MIDDLE-OF-GTID-EVENT
                   STOPS BEFORE TRANSACTION
      Problem:
      =======
      
      Slave loses track of GTID-header group boundaries when the group spans
      across multiple relay log files.
      
      This means that when the transaction is retried, or if you stop the SQL
      thread in the middle of a transaction after some Rotates (considering
      that the transaction/group was spanned into multiple relay log files),
      the Gtid_log_event will be silently skipped on slave, and the
      transaction will be logged with a slave's GTID.
      
      Also, when using "START SLAVE UNTIL MASTER_LOG_POS = x;", if "x" is in
      the middle of a transaction, the server is supposed to complete
      the transaction. This works fine when GTIDs are disabled. However,
      when GTIDs are enabled, if "x" is in the middle of the Gtid_log_event,
      it will stop before the transaction, not considering the
      Gtid_log_event as the beginning of the transaction.
      
      Analysis:
      ========
      
      When Rotate events are applied at the slave, the SQL thread verifies
      if it is not inside of a transaction to update SQL thread position.
      For STS, the SQL thread uses the Relay_log_info::is_in_group()
      function to determine if it is inside of a transaction.
      
      It was found a problem in Relay_log_info::is_in_group(). It wasn't
      considering a Gtid_log_event as the beginning of a group/transaction.
      
      Because of this problem, the SQL thread was updating its position
      when applying a Rotate_log_event that immediately follows a
      Gtid_log_event (i.e., in the middle of a transaction but outside
      BEGIN...COMMIT).
      
      This should not happen, as SQL thread position should not be updated
      in the middle of a transaction so it could retry (or re-apply) the
      transaction from the beginning in the case of failures because of
      InnoDB deadlock or because the transaction's execution time exceeded
      InnoDB's innodb_lock_wait_timeout (or in the case of a request to
      stop the SQL thread). 
      
      For the problem with "START SLAVE UNTIL MASTER...", the SQL thread
      uses the Relay_log_info::is_until_satisfied() function to verify if
      the until condition is satisfied. In this function, when the until
      condition is UNTIL_MASTER_POS, the SQL thread will use the current
      transaction position if in the middle of a transaction or the current
      event position if not in the middle of a transaction. This
      verification wasn't considering an Gtid_log_event as the beginning of
      a transaction.
      
      Fix:
      ===
      
      Made Relay_log_info::is_in_group() to check if the thread has a GTID
      set to the current transaction. If so, it will return true, stating
      that it is already in a group.
      
      Made Relay_log_info::is_until_satisfied() to use the is_in_group()
      function to verify if the SQL thread is in the middle of a transaction.
      With this change, is_until_satisfied() will consider the
      Gtid_log_event as part of the transaction.
------------------------------------------------------------
revno: 8446
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-18 02:13:09 +0200
message:
  Created default.weekly.basic collection for commercial plugins. Without that commercial tests were not running on tier 2 and tier 3 platforms
------------------------------------------------------------
revno: 8445 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-17 17:04:37 +0530
message:
  Bug#18731211  Duplicate slave server_uuid in replication
  forum gives no specific error message
  
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.38
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-07-17 16:56:55 +0530
    message:
      Bug#18731211  Duplicate slave server_uuid in replication 
      forum gives no specific error message
      
      Problem & Analysis: 
      ===================
      When a new slave joins to a replication forum with uuid which
      is exactly same as one of the existing slave's uuid, then master is 
      confused between old slave and the newly joined slave. When com_binlog_send
      request comes from newly joined slave, it thinks that old slave's dump
      thread is zombie and kills it. After some time, old slave tries to reconnect
      to the master and sends com_binlog_send request to Master. Then Master
      thinks that newly joined slave's dump thread is zombie and kills it.
      This ping-pong reconnections between Master->old_slave and Master->new_slave
      happens continously. This is known expected behaviour but the
      error message in old_slave and new_slave's error log file is not clear
      enough to identify this situation.
      
      Information in error log file was:
      "received end packet from server, apparent master shutdown" which
      does not say anything about server_uuid misconfiguration in the
      replication forum.
      
      Fix: Changing the error message which clearly mentions all the possible
      causes which resulted this disconnection.
      
      New information in error log file is: 
      "Slave: received end packet from server due to dump thread being killed on master.
      Dump threads are killed for example during master shutdown,
      explicitly by a user, or when the master receives a binlog send
      request from a duplicate server UUID <server_uuid>. "
      
      And also a NOTE is added on master's error log file with the following
      information when log_warnings is greater than 1.
      "While initializing dump thread for slave with UUID <slave_uuid>,
      found a zombie dump thread with the same UUID.
      Master is killing the zombie dump thread."
------------------------------------------------------------
revno: 8444
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Thu 2014-07-17 12:56:22 +0200
message:
  Bug#19143857: GET RID OF DYNAMIC_ARRAY IN RPL_FILTER
  
  Post-push fix: Don't call plugin_unlock_list if the array
  is empty.
------------------------------------------------------------
revno: 8443
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-17 09:28:57 +0200
message:
  Fix the copyright header in some sql/ndb* files.
------------------------------------------------------------
revno: 8442
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-17 09:21:08 +0200
message:
  Fix copyright headers in recently added tests.
------------------------------------------------------------
revno: 8441 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-17 11:50:14 +0530
message:
  Bug#14757009: WHEN THE GENERAL_LOG IS A SOCKET AND THE READER
                GOES AWAY, MYSQL QUITS WORKING.
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.338.37 [merge]
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-07-17 11:27:03 +0530
    message:
      Bug#14757009: WHEN THE GENERAL_LOG IS A SOCKET AND THE READER
                    GOES AWAY, MYSQL QUITS WORKING.
      
      Merging from 5.5 to 5.6.
        ------------------------------------------------------------
        revno: 2875.598.17
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-07-17 11:21:18 +0530
        message:
          Bug#14757009: WHEN THE GENERAL_LOG IS A SOCKET AND THE READER
                        GOES AWAY, MYSQL QUITS WORKING.
          
          Analysis:
          -----------------
          Issue in this bug and in bug 11907705 is, the socket file or
          fifo file is set for general log at command line while starting
          the server. But currently, only regular file can be set for the 
          general log. Instead of reporting any error, the provided files
          are opened for writing and continued. Because of this issues
          mentioned in the bug reports are seen.
          
          As mentioned, only when any non-regular file is set for general
          log at command line while starting the server, these issues are
          seen. If general log file is set to non-regular file from CLI
          using system variable general_log_file then error is reported.
          
          These issues can also be faced with slow query log file, if it is
          set to non-regular file.
          
          Fix:
          -----------------
          Currently while starting the server if we fail to open log file
          then we report an error, disable logging to file and continue.
          To fix issue reported code is modified to check whether file
          is regular file or not before opening it. If file is not a 
          regular file then error is logged to error log and logging to 
          file is disabled.
------------------------------------------------------------
revno: 8440
committer: Tarique Saleem <tarique.saleem@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-17 01:03:13 +0530
message:
  i_main.costmodel_planchange disabled on mysql-trunk due to many failures
------------------------------------------------------------
revno: 8439 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18110156
timestamp: Wed 2014-07-16 23:29:12 +0530
message:
  Merge from mysql-5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.338.36
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18110156
    timestamp: Wed 2014-07-16 22:19:12 +0530
    message:
      BUG#18110156: RECREATE+ANALYZE OPTIMIZE TABLE T AND ONLINE
                    ALTER TABLE T MAY DEADLOCK
      
      Analysis:
      --------
      OPTIMIZE TABLE and ONLINE ALTER TABLE on the same table
      using INNODB storage engine may cause a deadlock.
      
      OPTIMIZE TABLE for INNODB tables is mapped to recreate
      and analyze of the table. The deadlock occurs because
      InnoDB:
      
      1) Tries to support "online" ALTER TABLE which at some
         point tries to upgrade SU metadata lock to X lock,
         while holding TL_READ thr_lock lock.
      2) The analyze phase of the OPTIMIZE TABLE statement,
         acquires SW metadata lock (which is compatible with
         SU lock) and then tries to acquire "strong" TL_WRITE
         lock, which is not compatible with TL_READ.
         I.e. there is a discrepancy between "weak" MDL and
         "strong" thr_lock requested.
      
      The problem does not occur for other statements and other
      phases of OPTIMIZE TABLE, since they either acquire strong
      metadata locks which are not compatible with SU lock or
      downgrade the request to acquire TL_WRITE, to request for
      TL_WRITE_ALLOW_WRITE lock, which is compatible with TL_READ.
      
      
      Fix:
      ----
      For the normal ANALYZE TABLE operation, the locks requested
      are MDL_SHARED_READ and TL_READ_NO_INSERT locks.
      In case of OPTIMIZE TABLE operation, after recreate, the
      table is closed and locks are released. The table is re-opened
      and locks are re-acquired for the analyze phase. Since the
      analyze operation does not involve changing the data or
      metadata, a MDL_SHARED_READ and TL_READ_NO_INSERT locks
      are requested instead of MDL_SHARED_WRITE and TL_WRITE
      locks(same as the normal ANALYZE TABLE operation).
      This prevents the deadlock reported.
------------------------------------------------------------
revno: 8438 [merge]
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-16 11:12:48 +0100
message:
  BUG#16409385
  
  Automerge into latest mysql-trunk.
    ------------------------------------------------------------
    revno: 8415.1.2
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Wed 2014-07-16 11:10:31 +0100
    message:
      BUG#16409385
      
      Addressed review comments.
    ------------------------------------------------------------
    revno: 8415.1.1
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Thu 2014-07-10 12:44:21 +0100
    message:
      BUG#16409385: SEVERAL RPL TESTS FAIL ON SOLARIS WITH TIMEOUT FOR 'CHECK WARNING'
      
      The check-testcase stage is timing out for these two test cases
      on PB2. The tests output too many spurious warnings related
      to restarting the slave threads.
      
      To fix this, we suppress warnings from the error log for 
      rpl_typeconv and rpl_show_slave_status_nonblocking. These
      warnings are not the focus of the test cases anyway. This reduces
      the size of the error log and makes the tests less vulnerable to
      fail on the check-testcase stage. Note that, error entries are 
      still logged.
      
      Also, increased the timeout for rpl_typeconv as well.
------------------------------------------------------------
revno: 8437 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 21:56:10 +0530
message:
  Bug#18675348  RPL_PARALLEL_SECONDS_BEHIND_MASTER VALGRIND TEST IS FAILING
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.35
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-15 21:51:03 +0530
    message:
      Bug #18675348  RPL_PARALLEL_SECONDS_BEHIND_MASTER VALGRIND TEST IS FAILING
      
      Problem: In Pb2 valgrind runs, rpl_parallel_seconds_behind_master is failing.
      
      Analysis: Test script is written under the assumption that while calculating
      seconds behind_master(SBM), one of the workers should be waiting on 't2' lock.
      It is achieved through executing 'LOCK TABLE t2' directly on slave before
      'INSERT INTO t2' query (which was executed on Master) is executed by the
      slave worker. But the problem is some times slave worker is completing
      the execution of 'INSERT INTO t2' before the test script executes
      'LOCK TABLE t2'. Hence the SBM value is becoming zero instead of
      the expected 3 seconds and test logic is waiting for SBM to become 
      greater than 3 and it is getting timed out.
      
      Fix: Changing the logic in such way that 'LOCK TABLE t2' will always
      be executed before slave worker sees 'INSERT INTO t2' query.
------------------------------------------------------------
revno: 8436
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18675237
timestamp: Tue 2014-07-15 21:49:23 +0530
message:
  BUG#18675237- NOW() : MISSING FRACTIONAL SECONDS FROM A VIEW
  
  Analysis:
  --------
  
  Creating a view using the time function NOW(6) registers
  the view query definition as NOW() in the FRM file for
  the view. Thus the fractional seconds part is missing
  in the view definition.
  
  CREATE VIEW v1 AS SELECT NOW(6) as fld1;
  
  Is transformed as:
  CREATE ALGORITHM=UNDEFINED DEFINER=`root`@`localhost`
  SQL SECURITY DEFINER VIEW `v1` AS select now() AS `fld1`
  
  While registering the query definition in the FRM file,
  the 'decimals'(fractional seconds) was ignored. Hence NOW(6)
  was printed as NOW() in the FRM file, thus missing the
  fractional seconds.
  
  Fix:
  ---
  While registering the view query definition in the FRM file
  which uses temporal functions like NOW(), CURTIME() and
  SYSDATE(), print the 'decimals'(fractional seconds part)
  value, if it is set.
------------------------------------------------------------
revno: 8435
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 22:26:18 +0800
message:
  Bug #19140865 	PB2 INNODB_ZIP.INNODB_4K AND INNODB_ZIP.INNODB_8K TEST FAILS FOR RESULT MISMATCH
  
  PB2 innodb_zip.innodb_4k and innodb_zip.innodb_8k test fails for result mismatch.
  
  The result mismatch is introduced after 'gtid_executed' table is
  introduced in wl#6559. We missed to update it. So fix to update it.
------------------------------------------------------------
revno: 8434 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 16:28:16 +0530
message:
  Bug #18770469 OUT OF MEMORY ON 5.6 SLAVE USING RBR WITH DATETIME WHEN TABLE CREATED IN 5.5.
  
  Post push fix (merge from mysql-5.6)
  (Removing newly added THD::alloc() member. It is not required)
    ------------------------------------------------------------
    revno: 3902.338.34
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-15 16:26:07 +0530
    message:
      Bug#18770469 OUT OF MEMORY ON 5.6 SLAVE USING RBR WITH DATETIME WHEN TABLE CREATED IN 5.5.
      
      Post push fix
      (Removing newly added THD::alloc() member. It is not required)
------------------------------------------------------------
revno: 8433 [merge]
committer: Aditya A <aditya.a@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 14:33:39 +0530
message:
  Bug #17780517 5.6 PARTITIONS USE MUCH MORE MEMORY THAN 5.1
  
  Post push fix. Compilation warning in gcov build.
  [Merge from 5.6]
    ------------------------------------------------------------
    revno: 3902.338.33
    committer: Aditya A <aditya.a@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-15 14:30:32 +0530
    message:
      Bug #17780517 5.6 PARTITIONS USE MUCH MORE MEMORY THAN 5.1
      
      Post push fix. Compilation warning in gcov build.
------------------------------------------------------------
revno: 8432 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 13:32:20 +0530
message:
  Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO
                DTRACE PROBLEMS
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.338.32
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-15 13:30:57 +0530
    message:
      Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO
                    DTRACE PROBLEMS
      
      Follow up patch. Adding compile definitions was resulting in build
      break on some platforms. Reverting compile definitions appended
      to CFLAGS.
------------------------------------------------------------
revno: 8431
committer: Vamsikrishna Bhagi <vamsikrishna.bhagi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 11:33:30 +0530
message:
  Bug #19127636 MYSQL_SECURE_INSTALLATION FAILS TO RUN IF
                THE ROOT PLUGIN IS SHA256_PASSWORD
  
  Problem: When the root user uses sha256_password,
           mysql_secure_installation fails to connect to the
           server after reading the password from mysql_secret
           file.
  
  Solution: mysql_secure_installation initially tries to
            connect with a blank password and then reads
            mysql_secret file to fetch the password. If the
            login fails with blank password, all the options
            in mysql handle are reset and hence the ssl
            options too.
  
            A step is added in this patch to set the
            connection options in mysql handle after the login
            fails with blank password.
------------------------------------------------------------
revno: 8430 [merge]
committer: Aditya A <aditya.a@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-15 10:28:19 +0530
message:
  Bug #17780517 5.6 PARTITIONS USE MUCH MORE MEMORY THAN 5.1
  
  [ Merge from 5.6 ]
    ------------------------------------------------------------
    revno: 3902.338.31
    committer: Aditya A <aditya.a@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-15 10:23:36 +0530
    message:
      Bug #17780517 5.6 PARTITIONS USE MUCH MORE MEMORY THAN 5.1
      
      PROBLEM
      -------
      More memory usage was seen due to increase in size of
      ha_innobase object . The two character arrays 
      srch_key_val*  in ha_innobase class each contribute 
      approximately 6Kbytes ,so when large databases with 
      many partitions are used ,memory usage increases.
      
      FIX
      ---
      MySQL stores the INTs datatypes in little endian format
      where as innodb stores them in big endian format.The 
      srch_key_val* arrays are only needed when we are 
      converting INTs datatypes in the search key from little 
      endian format to big endian format. So we allocate 
      memory equivalent to number of INTs in each index of the 
      table ,which will reduce memory footprint. Also we have
      moved the allocation of the memory to ha_innobase::open()
      so that the memory is only allocated during opening of 
      table handler.  
      
      [Approved by Mattiasj and Jimmy #rb5565]
------------------------------------------------------------
revno: 8429
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-14 11:24:14 +0200
message:
  Bug #18756233 ASSERT !HAS_DONE_RESERVATION, FSEG_ALLOC_FREE_PAGE_LOW(),
  BTR_STORE_BIG_REC_EXTE
  
  Post push fix.  The test case uses innodb_buf_flush_list_now which is
  a debug only variable.  So adding have_debug.inc in the test case.
  
  approved by Jimmy over IM.
------------------------------------------------------------
revno: 8428 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-14 13:48:30 +0530
message:
  Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO 
                DTRACE PROBLEMS.
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.338.30
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-14 13:39:02 +0530
    message:
      Bug#19180531: UNABLE TO COMPILE MYSQL-TRUNK WITH DEBUG DUE TO
                    DTRACE PROBLEMS
      
      Issue here is, in function "DTRACE_INSTRUMENT"(dtrace.cmake), 
      CMAKE_C_FLAGS are stored in environment variable CFLAGS. Value of
      this variable is used by dtrace tool. But currently we were 
      passing C FLAGS as "-Wall -Wextra -Wformat-security -Wvla
      -Wwrite-strings -Wdeclaration-after-statement -Werror" also which
      is not necessary. Code to set environment variable CFLAGS is 
      added to pass flags as -m32 or -m64.
      
      To fix this issue, changed code to strip flags starting with
      "-W" and passing other flags. Also appended COMPILE_DEFINITIONS
      (as -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64) to the CFLAGS.
------------------------------------------------------------
revno: 8427 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-14 15:09:15 +0900
message:
  Merge 5.6 => trunk
  
  Approved by Marko in rb#5857
    ------------------------------------------------------------
    revno: 3902.338.29
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-14 15:06:43 +0900
    message:
      This is the second attempt to fix the bug
      
      Bug#15923864 : Bug#67718 : INNODB DRASTICALLY UNDER-FILLS PAGES IN CERTAIN CONDITIONS
      
      InnoDB should try to insert to the next page before split the page, if the insert record for split_and_insert is last of the page.
      
      Otherwise, the (reverse) sequential inserts to the middle of the index might cause 1 record per 1 page.
      
      -------
      The previous attempt had the bug and fixed already at this push.
      
      Bug#19045042 : ASSERT MACH_READ_FROM_8()==INDEX->ID RECV_RECOVERY_IS_ON() IS_INSIDE_IBUF()
      The new function btr_insert_into_right_sibling() misses to treat ibuf bitmap for the page which the record inserted.
      
      -------
      Approved by Marko in rb#5856
------------------------------------------------------------
revno: 8426
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-14 10:21:01 +0800
message:
  Fix Bug 16869435 - IN-PLACE ALTER TABLE FOR TABLE WITH BROKEN FK CAUSES
  ASSERT 
  
  rb://5653 approaved by Marko
------------------------------------------------------------
revno: 8425
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-review
timestamp: Fri 2014-07-11 10:28:37 +0200
message:
  Bug#19179338 GET RID OF DYNAMIC_ARRAY IN RELAY_LOG_INFO AND DEFERRED_LOG_EVENTS
  
  Some tests are failing when run with:
   --mysqld=--slave-parallel-workers=4 --mysqld=--slave-transaction-retries=0 --mysqld=--slave-parallel-type=database 
  
  rpl.rpl_row_corruption rpl.rpl_row_idempotency rpl.rpl_rotate_row_trans ...
  
  Revert by doing: bzr merge -r 8421..8420 .
------------------------------------------------------------
revno: 8424 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug17453826_mysql-trunk
timestamp: Fri 2014-07-11 11:25:24 +0530
message:
  Bug#17453826:ASSERTION ERROR WHEN SETTING FUTURE BINLOG
  FILE/POS WITH SEMISYNC
  
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.28
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug17453826_mysql-5.6
    timestamp: Fri 2014-07-11 11:20:54 +0530
    message:
      Bug#17453826:ASSERTION ERROR WHEN SETTING FUTURE BINLOG
      FILE/POS WITH SEMISYNC
      
      Fixing a post push test issue.
------------------------------------------------------------
revno: 8423
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt
timestamp: Thu 2014-07-10 18:31:24 +0200
message:
  Fix bug#19184930 
  Revert those changes from michael.izioumtchenko@oracle.com-20140709144138-qon47txx7u0c0tm0 that were committed unintentionally
------------------------------------------------------------
revno: 8422
committer: magnus.blaudd@oracle.com
branch nick: trunk
timestamp: Thu 2014-07-10 16:49:02 +0200
message:
  WL#6815 Adapt MySQL Cluster to 5.7
  
   - rewrite ha_ndbcluster::records() to return the number of records
    in the new "num_rows" argument.
   - use the return value from update_stats() as the return value
    from records(), it should in most cases by a MySQL mapped
    error code.
------------------------------------------------------------
revno: 8421
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Thu 2014-07-10 16:15:40 +0200
message:
  Bug#19179338 GET RID OF DYNAMIC_ARRAY IN RELAY_LOG_INFO AND DEFERRED_LOG_EVENTS
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
  
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8420 [merge]
committer: magnus.blaudd@oracle.com
branch nick: trunk
timestamp: Thu 2014-07-10 14:32:17 +0200
message:
  Merge trunk-cluster -> trunk
    ------------------------------------------------------------
    revno: 8417.1.1 [merge]
    committer: magnus.blaudd@oracle.com
    branch nick: merge_trunk-cluster_2_trunk
    timestamp: Thu 2014-07-10 12:46:02 +0200
    message:
      Merge trunk-cluster -> trunk
       
       - cluster as of ndb-7.3.6
        ------------------------------------------------------------
        revno: 2876.563.88
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Thu 2014-07-10 10:22:44 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - fix Windows compilation problem
           - replace another strcasecmp() -> native_strcasecmp()
        ------------------------------------------------------------
        revno: 2876.563.87
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Tue 2014-07-08 14:33:41 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - new order of tables in join circumvents the scan pruning
             sicne first (or head) pushed join is not a scan.
           - caused by WL#6635's new condition_fanout_filter
           - fix by turning condition_fanout_filter off to use old QEP.
        ------------------------------------------------------------
        revno: 2876.563.86
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Tue 2014-07-08 11:38:42 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - fix new order of tables in join involving multiple engines
           - new QEP uses eq_ref on the myisam table instead of table scan
           - slight change in the Note describing why the join can't be
             pushed to MyISAM...
           - caused by WL#6635's new condition_fanout_filter
           - tested to turn condition_fanout_filter off and the old QEP is used.
           - added some comments in the queries for markup
        ------------------------------------------------------------
        revno: 2876.563.85
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Tue 2014-07-08 10:45:38 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - fix new order of tables in join
           - everything else the same
           - caused by WL#6635's new condition_fanout_filter 
           - query is still not pushed -> ok
           - tested to turn condition_fanout_filter off and the old QEP is used.
           - added some comments in the queries for markup
        ------------------------------------------------------------
        revno: 2876.563.84
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Thu 2014-07-03 10:56:05 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
                            
           - fourth update of  .result files with EXPLAINs showing the 'partitions'
              used since the default number of partitions has been changed from 2 to 8.
           - this is only detected in Cluster based on 5.7 since EXPLAIN has been changed to output
             information about partitions by default
        ------------------------------------------------------------
        revno: 2876.563.83
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Wed 2014-07-02 15:13:57 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
                      
            - third update of  .result files with EXPLAINs showing the 'partitions'
              used since the default number of partitions has been changed from 2 to 8.
            - this is only detected in Cluster based on 5.7 since EXPLAIN has been changed to output
               information about partitions by default
        ------------------------------------------------------------
        revno: 2876.563.82
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Wed 2014-07-02 13:42:22 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
                
                 - second update of  .result files with EXPLAINs showing the 'partitions'
                   used since the default number of partitions has been changed from 2 to 8.
                 - this is only detected in Cluster based on 5.7 since EXPLAIN has been changed to output
                  information about partitions by default
        ------------------------------------------------------------
        revno: 2876.563.81
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Wed 2014-07-02 12:17:52 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
          
           - add the new warnings produced by INSERT IGNORE to the .result file
        ------------------------------------------------------------
        revno: 2876.563.80
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Wed 2014-07-02 11:49:51 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - update .result files with EXPLAINs showing the 'partitions'
             used since the default number of partitions has been changed from 2 to 8.
           - this is only detected in Cluster based on 5.7 since EXPLAIN has been changed to output
            information about partitions by default
        ------------------------------------------------------------
        revno: 2876.563.79
        committer: magnus.blaudd@oracle.com
        branch nick: merge_74_2_trunk-cluster-stage
        timestamp: Wed 2014-07-02 10:25:56 +0200
        message:
          Traditional record of perfschema.pfs_upgrade*
        ------------------------------------------------------------
        revno: 2876.563.78 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_74_2_trunk-cluster-stage
        timestamp: Wed 2014-07-02 10:17:11 +0200
        message:
          Merge 7.4 -> trunk-cluster
          
           - just after 7.3.7 tagged
            ------------------------------------------------------------
            revno: 2876.577.211 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Fri 2014-06-27 15:56:45 +0200
            message:
              merge release tagging -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.179 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Fri 2014-06-27 15:51:59 +0200
                message:
                  bump version to 7.3.7
                    ------------------------------------------------------------
                    revno: 2875.612.141 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Fri 2014-06-27 15:48:15 +0200
                    message:
                      bump version to 7.2.18
                        ------------------------------------------------------------
                        revno: 2555.843.71
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Fri 2014-06-27 15:45:51 +0200
                        message:
                          bumb version to 7.1.33
            ------------------------------------------------------------
            revno: 2876.577.210 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-26 16:10:34 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.178 [merge]
                tags: clone-mysql-5.6.19-ndb-7.3.6-src-build
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-26 16:09:38 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.140 [merge]
                    tags: clone-mysql-5.5.38-ndb-7.2.17-src-build
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-26 16:08:24 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.70
                        tags: clone-mysql-5.1.73-ndb-7.1.32-src-build
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-26 16:07:15 +0200
                        message:
                          The AutoTest, testBitfield does not correctly word align
                          the length of a default bit values. 
                          
                          The length of the default value of a bit column, need to be word aligned.
                          Incorrect alignment seems to have caused several (undetected) coredumps
                          from testBitfield, like in:
                          
                          Thread 1 (Thread 0x7f159a9bf720 (LWP 869)):
                          #0  0x0000000000431975 in u_32_value (this=0x7fffa6765988, pRow=0xb5d600) at /data0/ndbdev/autotest/build/clone-mysql-5.5-cluster-7.2-2014-06-23.24869/storage/ndb/test/src/HugoCalculator.cpp:299
                          #1  HugoCalculator::verifyRowValues (this=0x7fffa6765988, pRow=0xb5d600) at /data0/ndbdev/autotest/build/clone-mysql-5.5-cluster-7.2-2014-06-23.24869/storage/ndb/test/src/HugoCalculator.cpp:299
                          #2  0x000000000041f528 in HugoTransactions::pkReadRecords (this=0x7fffa6765930, pNdb=0xac27c0, records=1000, batch=13, lm=NdbOperation::LM_Read, _rand=0) at /data0/ndbdev/autotest/build/clone-mysql-5.5-cluster-7.2-2014-06-23.24869/storage/ndb/test/src/HugoTransactions.cpp:1078
                          #3  0x000000000040c23a in transactions (pNdb=0xac27c0, tab=<value optimized out>) at /data0/ndbdev/autotest/build/clone-mysql-5.5-cluster-7.2-2014-06-23.24869/storage/ndb/test/ndbapi/testBitfield.cpp:187
                          #4  0x000000000040dc68 in main (argc=0, argv=0xab6808) at /data0/ndbdev/autotest/build/clone-mysql-5.5-cluster-7.2-2014-06-23.24869/storage/ndb/test/ndbapi/testBitfield.cpp:99
                          
                          
                          Failures resulting in core dumps seems to be incorrectly trapped p? AutoTest: 
                          The coredumping test seems to be incorrectly recorded as PASSED, and the coredump
                          is then related to the next failing tests - Like in:
                          
                          http://ndbmaster.se.oracle.com/~autotest/report.pl?id=23419
                          
                          Where we find:
                          
                          - testBitfield 	PASSED
                            .... ~50 tests later:
                          - testSRBank -n SR -l 300 -r 15 T1 	FAILED(TEST) 	Y
                          
                          The failed testSRBank has attached a coredump from testBitfield!
                          
                          This fix doesn't fix the incorrect reporting from AutoTest.
                          However, it likely fix the coredump itself by correctly alligning
                          the bit-type columns.
                          
                          NOTE: We also find in ha_ndbcluster.cc, which has 
                                similar code for handling bit columns:
                          
                          static int create_ndb_column(THD *thd,
                                                       NDBCOL &col,
                                                       Field *field,
                                                       HA_CREATE_INFO *create_info
                          
                          .......
                          
                                      /* For bit columns, default length is rounded up to 
                                         **nearest word**, ensuring all data sent    <<<< HERE!
                                      */.
                                      Uint32 defaultLen = field_used_length(field);
                                      if(field->type() == MYSQL_TYPE_BIT)
                                        defaultLen = ((defaultLen + 3) /4) * 4;      <<<<<<<< and HERE !
                                      col.setDefaultValue(buf, defaultLen);
                          
                          Which is the correct way to handle length of default bit values
            ------------------------------------------------------------
            revno: 2876.577.209 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-26 18:19:48 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.177 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-26 18:14:13 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.139 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-26 18:10:12 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.69
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-26 18:07:30 +0530
                        message:
                          Bug#18044717: INCORRECT LCP STATE CAUSES AN INTERNAL ERROR
                          
                          An autotest failure was caused by the optimisation to skip metadata 
                          updates while removing a node if the LCP status is LCP_STATUS_IDLE. 
                          The failing test was 'testNodeRestart -n RestartNodeDuringLCP T6'. 
                          Removed the optimisation to fix the autotest failure.
            ------------------------------------------------------------
            revno: 2876.577.208 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-26 10:36:20 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.176 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-26 10:35:23 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.138 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-26 10:32:18 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.68
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-26 10:29:44 +0200
                        message:
                          Decrease some AutoTest max-time(out) values back to theirs
                          original settings from 1-2 months back.
                          
                          A lots of timeout were increased back then as there were
                          several testfailures due to a too low timeout value.
                          
                          However, some of these tests also timeouts due to restart
                          failures which still has to be analyzed / fixed. So these test
                          will timeout independent of how high we set the timeout.
                          
                          As the original timeout then seems to be sufficient, we
                          decrease it back in order to shorten the AutoTest run length.
            ------------------------------------------------------------
            revno: 2876.577.207 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-26 09:55:12 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.175 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-26 09:51:36 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.137 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-26 09:48:43 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.67
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-26 09:42:07 +0200
                        message:
                          Part 2 with fix for 'Node Restart' part of Bug#19053226 
                                
                          AUTOTEST FAIL: TESTSRBANK -N MIX -L 300 -R 15 T1
                                
                          This will also fix the AutoTest 'testSRBank -n NR ...'
                          However, there are still more failures in the 'Mix'
                          part of the test where we fails to start after a Node Restart
                          followed by a System Restart before the NR has completed.
                          ... will likely open another bug report to track that issue.
                          
                          The root cause for the failing 'Node Restart' (this fix) was
                          that the different 'Validate' methods in the Bank class 
                          did not differentiate between permanent (NDBT_FAILED)
                          and temporary (NDBT_TEMPORARY) failures. Thus, a node restart
                          during validation of Bank consistency caused Bank::performMakeGLs()
                          to return NDBT_FAILED if it were in the first part where it check
                          consistency of the 'GL'
                          
                          This fix catch temporary errors from ::startTransaction(),
                          ::execute() and ::nextResult(), and handle these by 
                          skipping the Validate part for this ::performMakeGLs round.
            ------------------------------------------------------------
            revno: 2876.577.206 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-06-25 12:52:44 +0200
            message:
              Merge 7.3 -> 7-4
                ------------------------------------------------------------
                revno: 2876.579.174 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-06-25 12:51:19 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.136 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-06-25 12:49:26 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.66
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-06-25 12:47:31 +0200
                        message:
                          Fix for 'System Restart'?part of Bug#19053226 
                          
                          AUTOTEST FAIL: TESTSRBANK -N MIX -L 300 -R 15 T1
                          
                          This will also fix the AutoTest 'testSRBank -n SR ...'
                          However, the 'Node Restart' part of the 'Mix'
                          will not be fixed by this patch ... still working on that.
                          
                          The comment for the these tests states: 
                          
                          " Test that a consistent bank is restored after **graceful shutdown**"
                          
                          However, the restart_cluster() was called with a default value
                          of the 'abort = true' argument, which implied a *non-gracefull shutdown*.
                          
                          This resulted in that in cases where the first System Restart happened
                          before the first GCP had completed, we would restore empty Bank tables in lack
                          of a REDO log containg the initial fill of these tables. Thus, the
                          later verification of the table contents after restart failed.
            ------------------------------------------------------------
            revno: 2876.577.205 [merge]
            committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-06-23 14:55:19 +0530
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.173 [merge]
                committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-06-23 14:49:08 +0530
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.135 [merge]
                    committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-06-23 14:46:19 +0530
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.65
                        committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-06-23 14:42:29 +0530
                        message:
                          BUG#18283369: Read/Write between two tables with identical char fields but
                                        different encoding
                          
                          When the value read from a fixed width char column with an utf8 charset is 
                          written to a column of same type and length but with a latin1 charset, it fails.
                          
                          The actual issue is, reading from a fixed width char column with an utf8 charset
                          returns the stored data with extra spaces, padded during its insertion. 
                          The fix is to trim the value before returning it. 
                          
                          This patch also fixes the issue that causes 'Data length too long' error during
                          the insertion of valid utf8 characters of 2 bytes and more. 
                          This is due to the padding of blank characters to the data, before encoding it.
                          The fix is to do the padding after encoding, rather than doing it beforehand.
                          
                          NdbRecordImpl.java
                            The result string is now trimmed before returning
                            	
                          Utility.java
                            Overloaded the padString to handle ByteBuffer class
                            Padding blank bytes to the data is moved after encoding
                            	
                          schema.sql
                            added a new charsetswedishutf8 table
                            	
                          Added testcases to verify the bug
            ------------------------------------------------------------
            revno: 2876.577.204 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-06-23 10:07:32 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.172 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-06-23 10:04:44 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.134 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-06-23 10:02:42 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.64
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-06-23 10:01:22 +0200
                        message:
                          Increased max-time for the AutoTest 'testDict -n SchemaTrans -l 1' 
                          as it failed due to timeout on some platforms / servers (ndb07)
            ------------------------------------------------------------
            revno: 2876.577.203 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18825966-74
            timestamp: Sat 2014-06-21 12:19:00 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.171
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18825966-73
                timestamp: Sat 2014-06-21 11:54:52 +0300
                message:
                  bug#18825966 bugfix.diff
                  UTIL_EXECUTE expects real attrids
                ------------------------------------------------------------
                revno: 2876.579.170
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18825966-73
                timestamp: Sat 2014-06-21 11:53:55 +0300
                message:
                  bug#18825966 bugtest-x1.diff
                  post-review add test case
                ------------------------------------------------------------
                revno: 2876.579.169
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18825966-73
                timestamp: Sat 2014-06-21 11:53:08 +0300
                message:
                  bug#18825966 bugtest.diff
                  bug test (fails before bug fix)
            ------------------------------------------------------------
            revno: 2876.577.202 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18824753-74
            timestamp: Fri 2014-06-20 17:16:17 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.168
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18824753-73
                timestamp: Fri 2014-06-20 16:10:10 +0300
                message:
                  bug#18824753 bugfix-x1.diff
                  post-review comment
                ------------------------------------------------------------
                revno: 2876.579.167
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18824753-73
                timestamp: Fri 2014-06-20 16:06:51 +0300
                message:
                  bug#18824753 bugfix.diff
                  bug fix (enforce FK name format NN/MM/name)
                ------------------------------------------------------------
                revno: 2876.579.166
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18824753-73
                timestamp: Fri 2014-06-20 16:03:57 +0300
                message:
                  bug#18824753 bugtest-x1.diff
                  post-review add test
                ------------------------------------------------------------
                revno: 2876.579.165
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18824753-73
                timestamp: Fri 2014-06-20 16:03:18 +0300
                message:
                  bug#18824753 bugtest.diff
                  bug test (fails before bug fix)
            ------------------------------------------------------------
            revno: 2876.577.201 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18766430-74
            timestamp: Fri 2014-06-20 12:36:30 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.164 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18766430-73
                timestamp: Fri 2014-06-20 10:40:53 +0300
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.133 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug18766430-72
                    timestamp: Fri 2014-06-20 10:19:57 +0300
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.843.63
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug18766430-71
                        timestamp: Fri 2014-06-20 10:09:40 +0300
                        message:
                          bug#18766430 bugfix-x1.diff
                          post-review add comment
                        ------------------------------------------------------------
                        revno: 2555.843.62
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug18766430-71
                        timestamp: Fri 2014-06-20 10:08:35 +0300
                        message:
                          bug#18766430 bugfix.diff
                          early check for duplicate drop within tx
            ------------------------------------------------------------
            revno: 2876.577.200 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-19 09:43:35 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.163 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-19 09:41:32 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.132 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-19 09:39:48 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.61
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-19 09:21:10 +0200
                        message:
                          Fix for bug#18993037
                          
                            NODE RESTARTS WHILE ANOTHER NODE IS ALREADY STARTING MAY CRASH 'MASTER'
                                
                          This was a regression introduced by the fix for bug#16007980
                          That fix added the following code to Qmgr::failReportLab()
                          
                          +  /**
                          +   * If any node is starting now (c_start.startNode != 0)
                          +   *   sendPrepFailReq to that too
                          +   */
                          +  if (c_start.m_startNode != 0)
                          +  {
                          +    jam();
                          +    cfailedNodes[cnoFailedNodes++] = c_start.m_startNode;
                          +    c_start.reset();
                          +  }
                                
                          This was to ensure that the (local) knowledge of 'startNode'
                          was preparing to start, should not be lost in the case that it
                          re-failed before it had reconnected.
                          
                          However, this also introduced a race condition where we either 
                          could, or could not, be notified through our normal 'channels'
                          about the re-fail of this node. Thus, the above code could
                          insert a duplicate of an already known failure in cfailedNodes[].
                          
                          Later, this 'list of nodes' was converted into a BitMask
                          used in the PREP_FAILREQ-signal, and converted back into
                          a 'list of nodes' by Qmgr::execPREP_FAILREQ(). During this
                          BitMask conversion, the duplicated NodeId was eliminated.
                          However, **the 'noOfNodes' count is kept unchanged**.
                          Thus we end up with a materialized 'list of nodes' where
                          the size is of-by-one, and the last list item contains garbage.
                          
                          This patch replaces the Qmgr internal usage of node lists with
                          NodeBitMask. This natively takes care of duplicate removal, and
                          simplifies the conversion to/from signal formats which use
                          similar bitmaps.
                          
                          Furthermore, this patch also fixes an issue where 'c_start.m_startNode' 
                          could be reset to early. It should not be reset until 
                          Qmgr::execPREP_FAILREQ() has received a signal informing
                          about the failure of the actual 'startNode'.
            ------------------------------------------------------------
            revno: 2876.577.199 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18759252-74
            timestamp: Tue 2014-06-17 22:04:29 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.162 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18759252-73
                timestamp: Tue 2014-06-17 21:41:23 +0300
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.131 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug18759252-72
                    timestamp: Tue 2014-06-17 21:33:25 +0300
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.843.60
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug18759252-71
                        timestamp: Tue 2014-06-17 21:04:52 +0300
                        message:
                          bug#18759252 testdict-x1.diff
                          post-review
                        ------------------------------------------------------------
                        revno: 2555.843.59
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug18759252-71
                        timestamp: Tue 2014-06-17 21:04:13 +0300
                        message:
                          bug#18759252 testdict.diff
                          filter out irrelevant objects from dict list
            ------------------------------------------------------------
            revno: 2876.577.198 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-06-17 08:34:34 +0200
            message:
              Null merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.161
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-06-17 08:30:14 +0200
                message:
                  Fix for bug#18704052
                  
                  REMOVE NOISY DEBUG LOG LEFTOVER FROM DBSPJ::EXECSTTOR()
                  
                  Code cleanup which removes two ndbout to ndbd.log which
                  were part of early development stages and never intended
                  as part of the Release Built.
                  
                  This had already been removed in 7.4, but was left in 7.3
                  for some reasons.
            ------------------------------------------------------------
            revno: 2876.577.197
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7858
            timestamp: Mon 2014-06-16 14:04:06 +0200
            message:
              WL#7858: Remove most of the use of environment variables in production builds, protected by #ifdef NDB_USE_GET_ENV in most cases
            ------------------------------------------------------------
            revno: 2876.577.196 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18183143-74
            timestamp: Sun 2014-06-15 15:01:34 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.160 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18183143-73
                timestamp: Sun 2014-06-15 14:54:17 +0300
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.130 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug18183143-72
                    timestamp: Sun 2014-06-15 14:41:12 +0300
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.843.58
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug18183143-71
                        timestamp: Sun 2014-06-15 14:24:02 +0300
                        message:
                          bug#18183143 aliasing.diff
                          avoid pointer aliasing which was optimized away
                          version: gcc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-3)
            ------------------------------------------------------------
            revno: 2876.577.195 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-12 15:59:08 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.159 [merge]
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-12 15:55:23 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.129 [merge]
                    committer: saroj tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-12 15:52:22 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.57
                        committer: saroj tripathy <saroj.tripathy@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-12 15:49:06 +0530
                        message:
                          disable testcases mysqlhotcopy* in 7.1 branch
            ------------------------------------------------------------
            revno: 2876.577.194 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-06-05 14:49:31 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.158 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2014-06-05 14:47:43 +0200
                message:
                  Merge 7.2-ndbapi-examples -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.623.7
                    committer: magnus.blaudd@oracle.com
                    branch nick: 5.5-cluster-7.2-ndbapi-examples
                    timestamp: Thu 2014-06-05 14:08:13 +0200
                    message:
                      ndbapi examples
                       - only build ndbapi examples by default when using compile-cluster on !windows
                    ------------------------------------------------------------
                    revno: 2875.623.6
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-ndbapi-examples
                    timestamp: Tue 2014-02-11 13:09:07 +0100
                    message:
                      ndbapi examples
                       - remove unistd.h from Windows.
                       - emulat sleep with Sleep
                    ------------------------------------------------------------
                    revno: 2875.623.5
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-ndbapi-examples
                    timestamp: Tue 2014-02-11 10:16:52 +0100
                    message:
                      ndbapi examples
                       - use winsock2.h instead of windows.h
                    ------------------------------------------------------------
                    revno: 2875.623.4
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-ndbapi-examples
                    timestamp: Mon 2014-02-10 14:37:33 +0100
                    message:
                      ndbapi examples
                      
                       - rename ndb_rpl_ndbapi_multi to ndb_rpl_ndbapi-examples
                    ------------------------------------------------------------
                    revno: 2875.623.3
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-ndbapi-examples
                    timestamp: Mon 2014-02-10 14:04:30 +0100
                    message:
                      ndbapi examples
                      
                       - fix also the second .test running ndbapi example programs in a replicated setup
                       - remove duplicate line terminator
                      -  change to use echo in favor of "exec echo"
                       - setup path to log file
                       - adapt to new name(s) of binaries
                    ------------------------------------------------------------
                    revno: 2875.623.2
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-ndbapi-examples
                    timestamp: Mon 2014-02-10 10:40:19 +0100
                    message:
                      ndbapi examples
                       - need to include stddef.h for offsetof()
                    ------------------------------------------------------------
                    revno: 2875.623.1 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-ndbapi-examples
                    timestamp: Fri 2014-02-07 15:00:11 +0100
                    message:
                      Merge
                        ------------------------------------------------------------
                        revno: 2875.622.14
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Fri 2014-02-07 14:56:31 +0100
                        message:
                          ndbapi examples
                          
                           - fix warning about string constant to 'char* for putenv
                           - add commetns explaining what dbug and signal trace code does
                        ------------------------------------------------------------
                        revno: 2875.622.13
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Fri 2014-02-07 14:46:44 +0100
                        message:
                          ndbapi examples
                          
                           -don't run test with embedded since propgrams need to connect to a running MySQL Server
                        ------------------------------------------------------------
                        revno: 2875.622.12
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Fri 2014-02-07 14:45:29 +0100
                        message:
                          ndbapi examples
                           
                          - need to include <windows.h> before <mysql.h> on Windows since SOCKET is used in 
                            datatypes and functions.
                        ------------------------------------------------------------
                        revno: 2875.622.11
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Thu 2014-02-06 14:57:20 +0100
                        message:
                          ndbapi examples
                                
                                 - fix another warning about ambigous else branch by adding
                                  braces
                        ------------------------------------------------------------
                        revno: 2875.622.10
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Thu 2014-02-06 14:51:46 +0100
                        message:
                          ndbapi examples
                          
                           - fix warning(s) about ambigous else branch by adding
                            braces
                        ------------------------------------------------------------
                        revno: 2875.622.9
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Thu 2014-02-06 14:42:59 +0100
                        message:
                          ndbapi examples
                          
                           - fix typo where "connectstring2" was not used for the second connection
                           - detected by "unused" warning
                        ------------------------------------------------------------
                        revno: 2875.622.8
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples-in-source
                        timestamp: Thu 2014-02-06 14:05:43 +0100
                        message:
                          ndbapi examples
                           - prefix the build binaries with ndb_ to avoid name clash with
                            the directory names when building in source
                           - also the mgmapi_ programs was in need of a ndb_ prefix
                        ------------------------------------------------------------
                        revno: 2875.622.7
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Wed 2014-02-05 16:17:31 +0100
                        message:
                          ndbapi_examples test
                           
                          - remove usage of "exec echo" in favor of plain "echo"
                        ------------------------------------------------------------
                        revno: 2875.622.6
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Wed 2014-02-05 16:13:43 +0100
                        message:
                          ndbapi-examples test
                          
                           - move the defitiontion of where log file ends up from mtr to test
                        ------------------------------------------------------------
                        revno: 2875.622.5
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Wed 2014-02-05 15:41:10 +0100
                        message:
                          ndbapi-examples test
                           - make mtr find ndbapi-examples
                           - rename ndbapi.test -> ndbapi-examples.test
                           - Fix paths to binaries(they are not compiled in the top level example dir)
                           - Fix syntax error with let, using both -- and ;
                           - Remove useless "drop database" of things never created
                           - Remove "use" of database never created
                           - Add drop of database ndb_examples which is created by the examples.
                        ------------------------------------------------------------
                        revno: 2875.622.4
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Wed 2014-02-05 15:08:51 +0100
                        message:
                          ndbapi-examples
                          
                           - Add ndbapi-examples/CMakeLists.txt which builds all examples
                           - Add option WITH_NDBAPI_EXAMPLES to control if the examples
                            should be built or not, default is to not build
                           - Turn on build of ndbapi-examples by default in compile-cluster
                           - Remove old Makefile(which was not automake'd but rather hand crafted)
                        ------------------------------------------------------------
                        revno: 2875.622.3
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Fri 2013-12-20 13:36:42 +0100
                        message:
                          ndbapi-examples
                           - include stddef.h for those programs which uses 'offsetof'
                        ------------------------------------------------------------
                        revno: 2875.622.2
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Fri 2013-12-20 13:35:42 +0100
                        message:
                          ndbapi-examples
                           - remov eold comment describing outdated LD_LIBRRARY_PATH
                        ------------------------------------------------------------
                        revno: 2875.622.1
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-ndbapi-examples
                        timestamp: Fri 2013-12-20 12:40:05 +0100
                        message:
                          ndbapi-examples
                           - change ndbapi_simple_index to use ndb_examples as database like all other examples
            ------------------------------------------------------------
            revno: 2876.577.193 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-06-05 12:30:03 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.157 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2014-06-05 12:29:34 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.128 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Thu 2014-06-05 12:28:54 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.56 [merge]
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Thu 2014-06-05 12:26:52 +0200
                        message:
                          Merge 7.1-wl6815 -> 7.1
                            ------------------------------------------------------------
                            revno: 2555.847.3
                            committer: magnus.blaudd@oracle.com
                            branch nick: 7.1
                            timestamp: Thu 2014-06-05 12:11:49 +0200
                            message:
                              WL#6815 Adapt MySQL Cluster to 5.7
                              
                              - Usage of -Wextra produces another 3 "ordered comparison of pointer with integer zero [-Wextra]"
                                 warnings. This time comparing the abort() function with > 0. Presumably abort
                                 has been a local variable in the function which rename/remove went undetected.
                               - Intention seem to be to only check the "should trans be aborted" condition when
                                 the abortPercent variable is set.
                               - Fix by replacing the usage of abort to abortPercent
                            ------------------------------------------------------------
                            revno: 2555.847.2
                            committer: magnus.blaudd@oracle.com
                            branch nick: 7.1
                            timestamp: Thu 2014-06-05 12:08:58 +0200
                            message:
                              WL#6815 Adapt MySQL Cluster to 5.7
                               - Usage of -Wextra produces 3 "ordered comparison of pointer with integer zero [-Wextra]"
                                 warnings about comparing return pointer from function against  < 0. The intention
                                 is to check if function returned NULL.
                               - Code has been there since "initial revision" and only affects test programs.
                            ------------------------------------------------------------
                            revno: 2555.847.1
                            committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                            branch nick: mysql-5.1-telco-7.1
                            timestamp: Mon 2014-03-10 10:59:58 +0100
                            message:
                              removed the extra blank line from the header
            ------------------------------------------------------------
            revno: 2876.577.192 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-06-05 01:03:17 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.156 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-06-05 00:54:50 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.127 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-06-05 00:40:14 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.55
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-06-05 00:17:48 +0100
                        message:
                          Bug #18465747     NDB REPLICATION : MONITOR FOR UNEXPECTED EPOCH SKIPPING 
                          
                          Fix compiler warning
            ------------------------------------------------------------
            revno: 2876.577.191 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-06-04 17:29:20 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.155 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-06-04 17:23:52 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.126 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-06-04 17:09:16 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.54
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-06-04 16:55:09 +0100
                        message:
                          Bug #18465747     NDB REPLICATION : MONITOR FOR UNEXPECTED EPOCH SKIPPING 
                          
                          This patch modifies the Ndb replication slave layer code to monitor for
                          epoch skipping behaviour.
                          
                          Specifically, there have been issues where the generic replication layer
                          slave retry-on-temp-error code has not functioned correctly, resulting
                          in an epoch transaction encountering a temp error being skipped entirely
                          instead of being retried.
                          
                          Retry-on-temp error is critical to replication correctness, and is 
                          explicitly used to have multi-pass-apply when using transactional
                          conflict detection.
                          
                          To avoid this situation recurring in future, the Ndb slave code is
                          modified here to check that every epoch which is started (identified
                          by an ndb_apply_status write_row event) is completed before a 
                          new epoch is started.
                          
                          The exception to the rule occurs when the Slave SQL thread is stopped 
                          and restarted (and hence a CHANGE MASTER could occur).
                          
                          This gives some protection against replication layer errors, and 
                          avoids data corruption / harder to debug downstream + later symptoms.
                          
                          This can be considered an extension of the existing check for epoch
                          decline.
            ------------------------------------------------------------
            revno: 2876.577.190 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Wed 2014-06-04 16:49:46 +0200
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.154 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Wed 2014-06-04 16:48:26 +0200
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.125
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Wed 2014-06-04 16:46:57 +0200
                    message:
                      Bug #18889568	 OUT OF SOURCE BUILDS CHANGES SOURCE DIR FOR CLUSTERJ AND NDBINFO_SQL
                      Bug #72843	out of source builds changes source dir for clusterj and ndbinfo_sql
                      
                      For clusterj, CMake files are changed so that manifest.mf files,
                      pom.xml files, and mvn_install_ndbjtie.sh are created in build dir
                      instead of source dir.
                      
                      ndbinfo.sql are configured as output to build dir instead of
                      source dir.  The file are already created in build dir.
            ------------------------------------------------------------
            revno: 2876.577.189 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-06-04 14:53:37 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.153 [merge]
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-06-04 14:50:34 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.124
                    committer: saroj tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-06-04 14:48:24 +0530
                    message:
                      reduced the test case ndb_big.rqg_spj run for 7.2 branch ONLY
            ------------------------------------------------------------
            revno: 2876.577.188 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-06-04 09:30:08 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.152 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-06-04 09:20:28 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.123 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-06-04 09:17:48 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.53
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-06-04 08:59:10 +0200
                        message:
                          Fix for: 'testNodeRestart -n Bug16772'?failure.
                          
                          In case a NDB_FAILCONF is delayed, or arrives out of order, 
                          restart of the failed node should await cleanup of nodestatus
                          on other nodes.
                          
                          The test expected the failed node to boot in 'no start'
                          mode, but stall before start phase 3 was reached. However,
                          this behaviour changed long ago (in 2006) such that the
                          failed node will now not be restarted at all until 
                          NDB_FAILCONF has been handled on all nodes.
                          
                          Test case was adjusted to reflect this new behavior.
            ------------------------------------------------------------
            revno: 2876.577.187 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-06-03 17:55:28 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.151 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-06-03 16:37:00 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.122 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-06-03 16:21:30 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.52
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-06-03 16:01:32 +0100
                        message:
                          Bug #18459944 	NDB REPLICATION : NDB$EPOCH_TRANS DELETE-DELETE-INSERT IS NON-DETERMINISTIC
                          
                          With NDB$EPOCH_TRANS, delete-delete conflicts were handled like
                          e.g. update-update conflicts, with the primary rejecting the transaction
                          and dependents, and realigning the secondary.
                          
                          This meant that the behaviour w.r.t. following operations on the row(s)
                          having delete-delete conflicts was dependent on whether they were in
                          the same or a different epoch.
                          
                          In the same epoch, they were also considered in-conflict (as the row
                          in question was 'refreshed' and therefore 'toxic' to following operations).
                          
                          In different epochs, they were not considered in conflict.
                          
                          This behaviour was generally regarded as bad for NDB$EPOCH, and so the
                          same decision is made for NDB$EPOCH_TRANS.
                          
                          This patch extends ndb_rpl_conflict_epoch_trans to test delete-delete
                          with NDB$EPOCH and encapsulate the expected result.
                          
                          One notable enhancement is that randomness is used to have the 
                          following secondary operations more commonly in a separate epoch.
                          
                          The transactional conflict handling is modified so that a delete-delete
                          conflict *alone* is not sufficient to cause a transaction to be
                          considered in-conflict.
            ------------------------------------------------------------
            revno: 2876.577.186 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-06-03 13:11:46 +0200
            message:
              NULL merge
                ------------------------------------------------------------
                revno: 2876.579.150 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-06-03 13:10:23 +0200
                message:
                  NULL merge
                    ------------------------------------------------------------
                    revno: 2875.612.121 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-06-03 13:09:05 +0200
                    message:
                      NULL merge
                        ------------------------------------------------------------
                        revno: 2555.843.51
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-06-03 13:05:12 +0200
                        message:
                          Remove use of not_parallel.inc in version where it doesn't exist
            ------------------------------------------------------------
            revno: 2876.577.185 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-06-03 00:16:18 +0200
            message:
              null merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.149 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-06-03 00:15:28 +0200
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.120
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2014-06-03 00:11:53 +0200
                    message:
                      Adjusted disabled tests in i_main after merge of 5.5.30
            ------------------------------------------------------------
            revno: 2876.577.184 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-06-02 13:25:32 +0200
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.148
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-06-02 13:22:55 +0200
                message:
                  Cherrypick from mysql-5.6
                  
                    revno: 5960
                    revision-id: tor.didriksen@oracle.com-20140523105024-zidi66prp5p27kbg
                    parent: venkatesh.duggirala@oracle.com-20140523051255-xfzbokhi1hdwiqvl
                    committer: Tor Didriksen <tor.didriksen@oracle.com>
                    branch nick: 5.6-cmake
                    timestamp: Fri 2014-05-23 12:50:24 +0200
                    message:
                      Bug#18605389 PROVIDE OPTION TO LINK AGAINST LIBCSTD INSTEAD OF STLPORT4 ON SOLARIS 10+
                      Bug#72352    Provide option to link against libCstd instead of stlport4 on Solaris 10+
                  
                      Patch for 5.6
                      This patch enables the use of Cstd when building with SunStudio.
                      It only works for client code, as the server depends on C++98
                      Usage:
                      cmake -DWITHOUT_SERVER=1 -DSUNPRO_CXX_LIBRARY=Cstd
                    
                      Also backport:
                      Bug#14367046: PROBLEM BUILDING CLIENT ONLY
            ------------------------------------------------------------
            revno: 2876.577.183 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-06-02 12:57:19 +0200
            message:
              merge 7.3 -> 7.4 (Merge of 5.6.19)
                ------------------------------------------------------------
                revno: 2876.579.147 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-06-02 12:54:53 +0200
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.119 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Mon 2014-06-02 12:53:20 +0200
                    message:
                      Merge 5.5.38 into 7.2.17
                ------------------------------------------------------------
                revno: 2876.579.146 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-06-02 12:51:44 +0200
                message:
                  Merge 5.6.19 into 7.3.6
            ------------------------------------------------------------
            revno: 2876.577.182
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-06-02 11:12:55 +0100
            message:
              Bug #18874849 	NDB : ADD 7.4+ TO NDB VERSION COMPATIBILITY MATRIX
              
              Add 7.4 to version matrix
            ------------------------------------------------------------
            revno: 2876.577.181 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-06-02 10:13:58 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.145 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-06-02 10:12:25 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.118 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-06-02 10:10:35 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.50
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-06-02 10:08:35 +0200
                        message:
                          'testSystemRestart -n SR_FULLDB T6' is really slow on 
                          some test rigs (ndb07), and need even long timeout
                          to be able to complete. 
                          
                          Increase from 2.5h to 4h
            ------------------------------------------------------------
            revno: 2876.577.180 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-06-02 09:34:35 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.144 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-06-02 09:33:18 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.117 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-06-02 09:32:02 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.49
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-06-02 09:30:57 +0200
                        message:
                          The AutoTest run of 'CreateTableWhenDbIsFull' failed due to
                          the DB not being full after all.
                          
                          This could happen as HugoTransactions::fillTable() gave up
                          when the first datanode reached 'DbIsFull'. However, there
                          could still be free DataMemory on other datanodes, such that
                          the later 'Create Table' succeeded in inserting a row 
                          on a different (non-full) datanode.
                          
                          This patch fix ::fillTable() such that it tries more
                          eagerly to fill DataMemory on all nodes.
            ------------------------------------------------------------
            revno: 2876.577.179 [merge]
            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
            branch nick: 7.4
            timestamp: Fri 2014-05-30 19:53:07 +0200
            message:
              Null merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.143
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Thu 2014-05-29 15:22:42 +0200
                message:
                  Bug#18845822 - DEADLOCK AND TIMEOUT IN BINLOG INJECTOR DUE TO WAITING FOR ITSELF
            ------------------------------------------------------------
            revno: 2876.577.178 [merge]
            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
            branch nick: 7.4
            timestamp: Thu 2014-05-29 13:47:57 +0200
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.593.1
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Thu 2014-05-29 13:42:05 +0200
                message:
                  Bug#18845822 - DEADLOCK AND TIMEOUT IN BINLOG INJECTOR DUE TO WAITING FOR ITSELF
            ------------------------------------------------------------
            revno: 2876.577.177 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-05-29 16:55:45 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.142 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-05-29 16:47:51 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.116 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-05-29 16:41:47 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.48
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-05-29 16:37:59 +0530
                        message:
                          Bug#18044717: INCORRECT LCP STATE CAUSES AN INTERNAL ERROR
                          
                          An LCP is initialized by setting the global LCP state to
                          LCP_INIT_TABLES, then looping over all the ndb tables and
                          setting the table-specific LCP status to TLS_ACTIVE. If any
                          table is not ready for LCP, the process continues using a
                          CONTINUEB signal. This is done to break up the work across
                          multiple signals to maintain the responsiveness of the
                          signal processing thread doing the work.
                          
                          After all the tables are updated, the global LCP status
                          is set to LCP_STATUS_ACTIVE. Therefore, during the init
                          procedure, some tables may have a status of TLS_ACTIVE while
                          the global LCP status is LCP_INIT_TABLES.
                          
                          While handling node failure of the master during an LCP, the
                          LCP takeover code checks the global LCP status and, if the
                          status is LCP_INIT_TABLES, changes it to LCP_STATUS_IDLE. This
                          results in a case where the global LCP status is LCP_STATUS_IDLE
                          while some tables may have a state of TLS_ACTIVE.
                          
                          However, there is an ndbrequire in the node-failure handler
                          which states that the lcp status cannot be LCP_STATUS_IDLE while
                          any table has a status of TLS_ACTIVE. This ndbrequire fails and
                          results in the core.
                          
                          Removed the ndbrequire to fix this bug. Also added an optimisation
                          to skip reverting metadata updates while removing a node from a
                          table if the LCP status is LCP_INIT_TABLES. The revert is not
                          required in this case since no metadata updates have taken
                          place yet.
            ------------------------------------------------------------
            revno: 2876.577.176 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4-bug18496153
            timestamp: Thu 2014-05-29 00:14:14 +0200
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.141 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-bug18496153
                timestamp: Wed 2014-05-28 23:57:45 +0200
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.115
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2-bug18496153
                    timestamp: Wed 2014-05-28 23:37:28 +0200
                    message:
                      ndb - added printers for signal ALLOC_NODEID_XXX in signal logs and trace files.
                    ------------------------------------------------------------
                    revno: 2875.612.114
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2-bug18496153
                    timestamp: Wed 2014-05-28 23:36:11 +0200
                    message:
                      ndb - extend DUMP 1
                      
                      DUMP 1      as before - dump phases for node id 1..48 (MAX_NDB_NODES-1)
                      DUMP 1 0   dump phases for node id 1..255 (MAX_NODES-1)
                      DUMP 1 x   dump phases for node id 1..x
                    ------------------------------------------------------------
                    revno: 2875.612.113
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2-bug18496153
                    timestamp: Wed 2014-05-28 23:34:32 +0200
                    message:
                      ndb_suma_handover.test
                      
                      For a restarting node in start phase 101 SUMA waits for all
                      subscribing nodes to connect.
                      
                      This test discards one CONNECT_REP from an API node during restart
                      of one node.
                      
                      Without preceding fix the test will fail (timeout).
                    ------------------------------------------------------------
                    revno: 2875.612.112
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2-bug18496153
                    timestamp: Wed 2014-05-28 23:33:08 +0200
                    message:
                      Bug #18599198 NDB : LIMIT TIME SPENT BY SUMA WAITING FOR API CONNECTIONS IN PHASE 101
                      
                      This patch adds a limit for how long Suma will wait for all
                      subscribers to connect before continuing with suma handover during
                      start phase 101.
                      
                      The root couse to why some subscribers does not connect to
                      restarted data nodes is not found nor fixed.
                      
                      A new configuration parameter RestartSubscriberConnectTimeout for
                      data nodes are added.  This timeout specifies how long Suma will
                      wait in phase 101 for all subscribing API nodes to connect.
                      
                      Those API nodes that have not connected in time will be reported
                      failed.
                      
                      And eventually when all those API nodes have failed, SUMA will
                      continue handover.
            ------------------------------------------------------------
            revno: 2876.577.175 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-05-28 15:42:06 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.140 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-05-28 15:41:04 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.111 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-05-28 15:39:35 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.47
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-05-28 15:38:17 +0200
                        message:
                          The AutoTest 'testSRBank -n Mix -l 300 -r 15 T1' did an 'abort()'
                          instead of returning NDBT_FAILED when the test failed.
                          
                          This prevents the FINALIZER() from being run, this causing
                          leftovers in the database which in turns caused *other tests*
                          to fail when they found the leftovers.
            ------------------------------------------------------------
            revno: 2876.577.174 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-05-28 11:17:00 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.139 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-05-28 11:11:49 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.110 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-05-28 11:06:54 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.46
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-05-28 11:05:26 +0200
                        message:
                          There are still some AutoTest failures due to timeout values 
                          set too low (Increased several TO-values ~week ago)
            ------------------------------------------------------------
            revno: 2876.577.173 [merge]
            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
            branch nick: 7.4
            timestamp: Tue 2014-05-27 22:33:39 +0200
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.138 [merge]
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Tue 2014-05-27 22:28:14 +0200
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.109 [merge]
                    committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                    branch nick: 7.2
                    timestamp: Tue 2014-05-27 22:23:58 +0200
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.45
                        committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                        branch nick: 7.1
                        timestamp: Tue 2014-05-27 22:22:01 +0200
                        message:
                          Bug#18716991 - INCONSISTENT EVENT DATA IS NOT REMOVED FROM EVENT QUEUE CAUSING CONSUMPTION STOP
            ------------------------------------------------------------
            revno: 2876.577.172 [merge]
            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
            branch nick: 7.4
            timestamp: Mon 2014-05-26 22:51:15 +0200
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.137 [merge]
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Mon 2014-05-26 22:47:41 +0200
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.108 [merge]
                    committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                    branch nick: 7.2
                    timestamp: Mon 2014-05-26 22:43:04 +0200
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.44
                        committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                        branch nick: 7.1
                        timestamp: Mon 2014-05-26 22:37:17 +0200
                        message:
                          bug#18816827 - REFACTOR EVENT BUFFER CODE TO ENHANCE READABILITY
            ------------------------------------------------------------
            revno: 2876.577.171
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4-aa
            timestamp: Fri 2014-05-23 13:08:13 +0100
            message:
              WL7639 Ndb Active Active Manageability improvements
              
              Fix too-short storage epoch values.
            ------------------------------------------------------------
            revno: 2876.577.170
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4-aa
            timestamp: Thu 2014-05-22 22:31:03 +0100
            message:
              WL7639 Ndb Active Active Manageability improvements
              
              Add ndb_slave_last_conflict_epoch status variable
              
              The ndb_slave_last_conflict_epoch status variable records the last slave
              epoch which included conflict handling activity.
              
              It is set when a MySQLD slave is applying binlogs to a cluster, and
              detects a conflict.
              
              It is set once the conflict causing transaction has been committed into
              the slave cluster.
              
              It is set to the slave cluster epoch assigned to the committed conflict 
              causing transaction.
              
              The intention is that in an asymmetric active-active setup (e.g. 
              NDB$EPOCH), this will indicate when the PRIMARY cluster last experienced
              a conflict.
              
              If this variable is <= the MaximumReplicatedEpoch then from the Primary 
              cluster's point of view there are no in-flight realignment operations, 
              and therefore no outstanding realignment dependencies.
              
              This can be a valuable state to be aware of when considering role
              switchovers or failure handling steps required.
              
              A new testcase is added to verify the behaviour of this new status
              variable.
            ------------------------------------------------------------
            revno: 2876.577.169
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4-aa
            timestamp: Thu 2014-05-22 22:21:39 +0100
            message:
              WL7639 Ndb Active Active Manageability improvements
              
              Implement status variables exposing last commit epochs for the server
              and each session.
              
              New status variables Ndb_last_commit_epoch_server and 
              Ndb_last_commit_epoch_session are added to the MySQL server, allowing
              SQL layer code to determine e.g.
                - What was the epoch of the last transaction committed by this server
                  (Same info as embedded in SHOW ENGINE NDB STATUS output)
                - What was the epoch of the last transaction committed by this session
              
              The session variant particularly is useful when determining when various
              post-commit activities have completed on a transaction.  e.g. it could be 
              used to determine k-safety round a replication ring by comparing with
              the Max_replicated_epoch, or if a latest-disk-durable epoch were exposed, it
              could be used to determine when a transaction becomes disk-durable.
              
              A new testcase is added to verify the behaviour of the variables.
              
               
            ------------------------------------------------------------
            revno: 2876.577.168
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4-aa
            timestamp: Thu 2014-05-22 22:09:51 +0100
            message:
              WL7639 Ndb Active Active Manageability improvements
              
              Implementation of ndb_slave_conflict_role control variable.
              
              This control variable allows the slave role in a future
              asymmetric conflict detection algorithm to be modified online.
              
              No conflict algorithm supports this control yet.
              
              The roles are : NONE, SECONDARY, PRIMARY, PASS.
              
              The role can only be changed when the slave SQL thread is not running.
            ------------------------------------------------------------
            revno: 2876.577.167 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-05-22 14:11:37 +0200
            message:
              BUG#18533982, added test case, merge
                ------------------------------------------------------------
                revno: 2876.579.136 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-05-22 14:10:40 +0200
                message:
                  BUG#18533982, added test case, merge
                    ------------------------------------------------------------
                    revno: 2875.612.107 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-05-22 14:09:39 +0200
                    message:
                      BUG#18533982, added test case, merge
                        ------------------------------------------------------------
                        revno: 2555.843.43
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-05-22 14:08:43 +0200
                        message:
                          BUG#18533982, added test case
            ------------------------------------------------------------
            revno: 2876.577.166 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-05-22 14:07:16 +0200
            message:
              BUG#18533982: Fixes missing logPartPtr init and added test case, merge
                ------------------------------------------------------------
                revno: 2876.579.135 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-05-22 14:05:19 +0200
                message:
                  BUG#18533982: Fixes missing logPartPtr init and added test case, merge
                    ------------------------------------------------------------
                    revno: 2875.612.106 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-05-22 14:03:12 +0200
                    message:
                      BUG#18533982: Fixes missing logPartPtr init and added test case
                        ------------------------------------------------------------
                        revno: 2555.843.42
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-05-22 14:01:27 +0200
                        message:
                          BUG#18533982: Fixes missing logPartPtr init and added test case
            ------------------------------------------------------------
            revno: 2876.577.165 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Thu 2014-05-22 12:04:17 +0200
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.134
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Thu 2014-05-22 11:57:37 +0200
                message:
                  WL#7774 Make MTR sync_slave_sql_with_master.inc handle NDB
                  
                  Adjust previous commit for worklog to align with actual push into
                  MySQL Server.
                  
                  Removing needless 'connection slave' right after including
                  sync_slave_sql_with_master.inc
            ------------------------------------------------------------
            revno: 2876.577.164 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-20 17:34:27 +0200
            message:
              BUG#18533982 fix merge
                ------------------------------------------------------------
                revno: 2876.579.133 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-05-20 17:33:23 +0200
                message:
                  BUG#18533982 fix merge
                    ------------------------------------------------------------
                    revno: 2875.612.105 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-05-20 17:32:05 +0200
                    message:
                      BUG#18533982 fix merge
                        ------------------------------------------------------------
                        revno: 2555.843.41
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-05-20 17:30:50 +0200
                        message:
                          BUG#18533982 fix
            ------------------------------------------------------------
            revno: 2876.577.163 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-20 16:12:44 +0200
            message:
              Merge 7.3->7.4.
                ------------------------------------------------------------
                revno: 2876.579.132 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-05-20 15:26:16 +0200
                message:
                  Merge 7.2->7.3.
                    ------------------------------------------------------------
                    revno: 2875.612.104 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-05-20 15:18:06 +0200
                    message:
                      Merge 7.1->7.2.
                        ------------------------------------------------------------
                        revno: 2555.843.40
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-05-20 14:59:44 +0200
                        message:
                          This is an update of the commit below. The call to markEndOfSigExec() is moved from SimulatedBlock::executeFunction() to the place where the scheduler calls executeFunction() as this improves readability and simplifies merge with7.4. (In 7.4, executeFunction() has already been split into executeFunction() and executeFunction_async(), and the orginal patch would create yet another version of each of these.)
                          
                          --------------------------
                          #At file:///net/atum17/export/home/tmp/jw159207/mysql/repo/mysql-5.1-telco-7.1/ based on revid:mikael.ronstrom@oracle.com-20140517184852-4y8six09z1fhs37e
                          
                           5008 Jan Wedvik	2014-05-19
                                This commit implements WL#7841 (Mark start of signal execution in jam trace). Bit 31 in each JamEvent is used for tagging the last JamEvent before starting to execute a new signal form the signal queue. The trace below is an example of how the trace will now look. Each new signal is marked with '---> signal'. Observe that only signals taken from the signal queues are marked in this way, not those from EXECUTE_DIRECT. This is to be consistent with the signal trace that appears after the jam trace, since EXECUTE_DIRECT signals do not appear there either.
                                
                                
                                ---> signal
                                DbtcMain.cpp         04536 04543 04543 04566 04673 04675 04716 04797 04914 
                                                     04941 05038 
                                DblqhMain.cpp        03552 
                                DbtcMain.cpp         05007 05009 05016 05016 05016 05016 05016 05024 
                                DbtupBuffer.cpp      00035 
                                Ndbfs.cpp            01550 01383 01550 01383 
                                ---> signal
                                DbtcMain.cpp         06682 06709 
                                DbdihMain.cpp        09497 09506 
                                DbtcMain.cpp         05188 05264 05310 05501 05038 05446 
                                DblqhMain.cpp        03552 
                                DbtcMain.cpp         05007 05009 05016 05016 05016 05016 05018 05016 
                                DbtupBuffer.cpp      00035 
                                Ndbfs.cpp            01550 01383 01550 01383 
                                ---> signal
                                DblqhMain.cpp        03392 03424 
                                ---> signal
                                DblqhMain.cpp        07934 08291 
                                DbtupCommit.cpp      00072 00089 
                                DblqhMain.cpp        08351 08374 08377 08409 08299 08341 08427 08476 
                                DbtupCommit.cpp      00684 00735 00747 00858 
                                DbtupTrigger.cpp     00904 00907 
                                DbtupCommit.cpp      00881 00325 00469 00480 00905 
                                DbaccMain.cpp        02245 04166 04124 04134 04143 02287 
                                DblqhMain.cpp        08559 08613 08666 03962 03585 03552 03556 03570 
                                DbtcMain.cpp         05007 
                                DbtupBuffer.cpp      00035 
                                Ndbfs.cpp            01550 01383 
                                ---> signal
                                DbtcMain.cpp         04125 
                                ---> signal
                                DbtcMain.cpp         05576 05678 05687 05789 05799 05310 05890 05038 05849 
                                DblqhMain.cpp        03552 
                                DbtcMain.cpp         05007 05009 05016 05016 05016 05016 05018 05016 
                                DbtupBuffer.cpp      00035 
                                Ndbfs.cpp            01550 01383 
                                
                                --------------- Signal ----------------
                                r.bn: 245 "DBTC", r.proc: 4, r.sigId: 2295700 gsn: 342 "PACKED_SIGNAL" prio: 1
                                s.bn: 247 "DBLQH", s.proc: 4, s.sigId: 2295699 length: 3 trace: 1 #sec: 0 fragInf: 0
                                Signal data: H'2007a929 H'00002039 H'00700500 
                                --------- Begin Packed Signals --------
                                --------------- Signal ----------------
                                r.bn: 245 "DBTC", length: 3 "COMMITTED"
                                Signal data: H'2007a929 H'00002039 H'00700500 
                                --------- End Packed Signals ----------
                                --------------- Signal ----------------
                                r.bn: 247 "DBLQH", r.proc: 4, r.sigId: 2295699 gsn: 342 "PACKED_SIGNAL" prio: 1
                                s.bn: 247 "DBLQH", s.proc: 3, s.sigId: -1 length: 5 trace: 1 #sec: 0 fragInf: 0
                                Signal data: H'0000000b H'0000000e H'00002039 H'00700500 H'00000000 
                                --------- Begin Packed Signals --------
                                --------------- Signal ----------------
                                r.bn: 247 "DBLQH", length: 5 "COMMIT"
                                Signal data: H'0000000b H'0000000e H'00002039 H'00700500 H'00000000 
                                --------- End Packed Signals ----------
                                --------------- Signal ----------------
                                r.bn: 245 "DBTC", r.proc: 4, r.sigId: 2295698 gsn: 19 "TC_COMMITREQ" prio: 1
                                s.bn: 32775 "API", s.proc: 5, s.sigId: 0 length: 3 trace: 1 #sec: 0 fragInf: 0
                                 H'00000022 H'00002039 H'00700500
                                --------------- Signal ----------------
                                r.bn: 245 "DBTC", r.proc: 4, r.sigId: 2295697 gsn: 342 "PACKED_SIGNAL" prio: 1
                                s.bn: 247 "DBLQH", s.proc: 3, s.sigId: -1 length: 7 trace: 1 #sec: 0 fragInf: 0
                                Signal data: H'40000018 H'0007a929 H'00f70003 H'00000000 H'00002039 H'00700500 H'00000000 
                                --------- Begin Packed Signals --------
                                --------------- Signal ----------------
                                r.bn: 245 "DBTC", length: 7 "LQHKEYCONF"
                                Signal data: H'40000018 H'0007a929 H'00f70003 H'00000000 H'00002039 H'00700500 H'00000000 
                                --------- End Packed Signals ----------
                                --------------- Signal ----------------
                                r.bn: 247 "DBLQH", r.proc: 4, r.sigId: 2295696 gsn: 342 "PACKED_SIGNAL" prio: 1
                                s.bn: 245 "DBTC", s.proc: 3, s.sigId: -1 length: 3 trace: 1 #sec: 0 fragInf: 0
                                Signal data: H'50000000 H'00002038 H'00700500 
                                --------- Begin Packed Signals --------
                                --------------- Signal ----------------
                                r.bn: 247 "DBLQH", length: 2 "REMOVE_MARKER"
                                Signal data: H'00002038 H'00700500 
                                --------- End Packed Signals ----------
            ------------------------------------------------------------
            revno: 2876.577.162 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-20 14:43:21 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.131 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-05-20 14:41:35 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.103 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-05-20 14:34:41 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.39
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-05-20 14:25:46 +0200
                        message:
                          Increase timeout for several Autotest which has 'FAILED(TO)'
                          and where the expected runtime is too close to max-time
            ------------------------------------------------------------
            revno: 2876.577.161 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-05-19 15:44:32 +0200
            message:
              WL#7762: merge fix
                ------------------------------------------------------------
                revno: 2876.579.130 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-05-19 15:43:39 +0200
                message:
                  WL#7762: merge fix
                    ------------------------------------------------------------
                    revno: 2875.612.102
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-05-19 15:42:10 +0200
                    message:
                      WL#7762: Removed result file for ndb_multi_tc_takeover and moved test to ndb_big test suite
            ------------------------------------------------------------
            revno: 2876.577.160 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-05-19 11:11:26 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.129 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-05-19 11:10:07 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.101 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-05-19 11:08:52 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.38
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-05-19 11:07:35 +0200
                        message:
                          Fix for failing AutoTest described in bug#18777940
                          
                            AUTOTEST FAILURE IN 'TESTDICT -N CREATETABLEWHENDBISFULL T6'
                          
                          DataMemory is no longer allocated as part of a create table,
                          but instead allocated on demand when rows are inserted. Thus
                          creation of an (empty) table should not fail even if 
                          DataMemory is exhausted.
                          
                          Instead we should expect an error when the first row is
                          inserted into the table.
            ------------------------------------------------------------
            revno: 2876.577.159 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-05-19 10:56:21 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.128 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-05-19 10:55:24 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.100 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-05-19 10:54:08 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.37
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-05-19 10:52:40 +0200
                        message:
                          Fix for autotest failure in: 'testNdbApi -n MaxOperations'
                          
                          This test has been failing for *every day* since 2008 with the following
                          problems:
                          
                          1)
                            Got error '1217 Out of operation records in local data manager ..'
                            instead of the expected error: 
                            '233 Out of operation records in transaction coordinator'
                          
                            This was caused by only generating pkReadRecord-ops for
                            row #1 in the test, thus all LQHKEYREQs ended up on the same
                            datanode and thus exhausted the 'MaxNoOfLocalOperations'
                            in that node.
                          
                            Fixed by pseudo randomly creating rowNo in the range 1..256
                          
                          2)
                            As all pkReadRecords were executed in a giant execute batch,
                            the Transporter got overloaded for some of the larger tables.
                            (ERROR: 1218 Send Buffers overloaded in NDB kernel)
                          
                            Fixed by sending an 'execute_NoCommit()' after every 1000'th row.
                            Note: The no-commit keeps the operation records in TC which the
                            test try to exhaust.
                          
                            As previously, a execute_Commit() is still done after the last
                            row.
                          
                          An optimization is also introduced in HugoOperations.cpp, where we
                          use the getValue() variant taking a 'Column*' arg. instead of the
                          'char* name' argument of the same column. Debugging with this testcase
                          revealed that the test client spent significant time doing strcmp
                          in getValue.
            ------------------------------------------------------------
            revno: 2876.577.158 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-05-19 10:10:56 +0200
            message:
              Merge 7.3->7.4.
                ------------------------------------------------------------
                revno: 2876.579.127 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-05-19 09:37:26 +0200
                message:
                  Merge 7.2->7.3.
                    ------------------------------------------------------------
                    revno: 2875.612.99 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-05-19 09:28:56 +0200
                    message:
                      Merge 7.1->7.2.
                        ------------------------------------------------------------
                        revno: 2555.843.36
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-05-19 09:22:35 +0200
                        message:
                          This commit implements WL#7841 (Mark start of signal execution in jam trace). Bit 31 in each JamEvent is used for tagging the last JamEvent before starting to execute a new signal form the signal queue. The trace below is an example of how the trace will now look. Each new signal is marked with '---> signal'. Observe that only signals taken from the signal queues are marked in this way, not those from EXECUTE_DIRECT. This is to be consistent with the signal trace that appears after the jam trace, since EXECUTE_DIRECT signals do not appear there either.
                          
                          
                          ---> signal
                          DbtcMain.cpp         04536 04543 04543 04566 04673 04675 04716 04797 04914 
                                               04941 05038 
                          DblqhMain.cpp        03552 
                          DbtcMain.cpp         05007 05009 05016 05016 05016 05016 05016 05024 
                          DbtupBuffer.cpp      00035 
                          Ndbfs.cpp            01550 01383 01550 01383 
                          ---> signal
                          DbtcMain.cpp         06682 06709 
                          DbdihMain.cpp        09497 09506 
                          DbtcMain.cpp         05188 05264 05310 05501 05038 05446 
                          DblqhMain.cpp        03552 
                          DbtcMain.cpp         05007 05009 05016 05016 05016 05016 05018 05016 
                          DbtupBuffer.cpp      00035 
                          Ndbfs.cpp            01550 01383 01550 01383 
                          ---> signal
                          DblqhMain.cpp        03392 03424 
                          ---> signal
                          DblqhMain.cpp        07934 08291 
                          DbtupCommit.cpp      00072 00089 
                          DblqhMain.cpp        08351 08374 08377 08409 08299 08341 08427 08476 
                          DbtupCommit.cpp      00684 00735 00747 00858 
                          DbtupTrigger.cpp     00904 00907 
                          DbtupCommit.cpp      00881 00325 00469 00480 00905 
                          DbaccMain.cpp        02245 04166 04124 04134 04143 02287 
                          DblqhMain.cpp        08559 08613 08666 03962 03585 03552 03556 03570 
                          DbtcMain.cpp         05007 
                          DbtupBuffer.cpp      00035 
                          Ndbfs.cpp            01550 01383 
                          ---> signal
                          DbtcMain.cpp         04125 
                          ---> signal
                          DbtcMain.cpp         05576 05678 05687 05789 05799 05310 05890 05038 05849 
                          DblqhMain.cpp        03552 
                          DbtcMain.cpp         05007 05009 05016 05016 05016 05016 05018 05016 
                          DbtupBuffer.cpp      00035 
                          Ndbfs.cpp            01550 01383 
                          
                          --------------- Signal ----------------
                          r.bn: 245 "DBTC", r.proc: 4, r.sigId: 2295700 gsn: 342 "PACKED_SIGNAL" prio: 1
                          s.bn: 247 "DBLQH", s.proc: 4, s.sigId: 2295699 length: 3 trace: 1 #sec: 0 fragInf: 0
                          Signal data: H'2007a929 H'00002039 H'00700500 
                          --------- Begin Packed Signals --------
                          --------------- Signal ----------------
                          r.bn: 245 "DBTC", length: 3 "COMMITTED"
                          Signal data: H'2007a929 H'00002039 H'00700500 
                          --------- End Packed Signals ----------
                          --------------- Signal ----------------
                          r.bn: 247 "DBLQH", r.proc: 4, r.sigId: 2295699 gsn: 342 "PACKED_SIGNAL" prio: 1
                          s.bn: 247 "DBLQH", s.proc: 3, s.sigId: -1 length: 5 trace: 1 #sec: 0 fragInf: 0
                          Signal data: H'0000000b H'0000000e H'00002039 H'00700500 H'00000000 
                          --------- Begin Packed Signals --------
                          --------------- Signal ----------------
                          r.bn: 247 "DBLQH", length: 5 "COMMIT"
                          Signal data: H'0000000b H'0000000e H'00002039 H'00700500 H'00000000 
                          --------- End Packed Signals ----------
                          --------------- Signal ----------------
                          r.bn: 245 "DBTC", r.proc: 4, r.sigId: 2295698 gsn: 19 "TC_COMMITREQ" prio: 1
                          s.bn: 32775 "API", s.proc: 5, s.sigId: 0 length: 3 trace: 1 #sec: 0 fragInf: 0
                           H'00000022 H'00002039 H'00700500
                          --------------- Signal ----------------
                          r.bn: 245 "DBTC", r.proc: 4, r.sigId: 2295697 gsn: 342 "PACKED_SIGNAL" prio: 1
                          s.bn: 247 "DBLQH", s.proc: 3, s.sigId: -1 length: 7 trace: 1 #sec: 0 fragInf: 0
                          Signal data: H'40000018 H'0007a929 H'00f70003 H'00000000 H'00002039 H'00700500 H'00000000 
                          --------- Begin Packed Signals --------
                          --------------- Signal ----------------
                          r.bn: 245 "DBTC", length: 7 "LQHKEYCONF"
                          Signal data: H'40000018 H'0007a929 H'00f70003 H'00000000 H'00002039 H'00700500 H'00000000 
                          --------- End Packed Signals ----------
                          --------------- Signal ----------------
                          r.bn: 247 "DBLQH", r.proc: 4, r.sigId: 2295696 gsn: 342 "PACKED_SIGNAL" prio: 1
                          s.bn: 245 "DBTC", s.proc: 3, s.sigId: -1 length: 3 trace: 1 #sec: 0 fragInf: 0
                          Signal data: H'50000000 H'00002038 H'00700500 
                          --------- Begin Packed Signals --------
                          --------------- Signal ----------------
                          r.bn: 247 "DBLQH", length: 2 "REMOVE_MARKER"
                          Signal data: H'00002038 H'00700500 
                          --------- End Packed Signals ----------
            ------------------------------------------------------------
            revno: 2876.577.157 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18662582-74
            timestamp: Sun 2014-05-18 21:14:25 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.126
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18662582-73
                timestamp: Sun 2014-05-18 14:50:26 +0300
                message:
                  bug#18662582 cleanup1.diff
                  remove obsolete check from inplace drop fk
                ------------------------------------------------------------
                revno: 2876.579.125
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18662582-73
                timestamp: Sun 2014-05-18 14:48:22 +0300
                message:
                  bug#18662582 bugfix2.diff
                  fix bug for alter copy drop fk
                ------------------------------------------------------------
                revno: 2876.579.124
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18662582-73
                timestamp: Sun 2014-05-18 14:47:43 +0300
                message:
                  bug#18662582 bugtest2b.diff
                  test bug for alter copy drop fk
                ------------------------------------------------------------
                revno: 2876.579.123
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18662582-73
                timestamp: Sun 2014-05-18 14:47:07 +0300
                message:
                  bug#18662582 bugfix1.diff
                  fix bug for alter inplace drop fk
                ------------------------------------------------------------
                revno: 2876.579.122
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18662582-73
                timestamp: Sun 2014-05-18 14:45:58 +0300
                message:
                  bug#18662582 bugtest1b.diff
                  test bug for alter inplace drop fk
            ------------------------------------------------------------
            revno: 2876.577.156 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Sat 2014-05-17 20:56:02 +0200
            message:
              Merge BUG#18533982
                ------------------------------------------------------------
                revno: 2876.579.121 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Sat 2014-05-17 20:51:19 +0200
                message:
                  Merge BUG#18533982
                    ------------------------------------------------------------
                    revno: 2875.612.98 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Sat 2014-05-17 20:50:11 +0200
                    message:
                      Merge BUG#18533982
                        ------------------------------------------------------------
                        revno: 2555.843.35
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Sat 2014-05-17 20:48:52 +0200
                        message:
                          BUG#18533982: Fix deadlock in abort handling with queueing REDO logs, ensure we can abort also when in REDO log queue
            ------------------------------------------------------------
            revno: 2876.577.155 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Sat 2014-05-17 19:48:42 +0200
            message:
              WL#7762, minor type fix
                ------------------------------------------------------------
                revno: 2876.579.120 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Sat 2014-05-17 19:47:24 +0200
                message:
                  Merged WL#7762 fix
                    ------------------------------------------------------------
                    revno: 2875.612.97
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Sat 2014-05-17 19:46:13 +0200
                    message:
                      WL#7762, minor type fix
            ------------------------------------------------------------
            revno: 2876.577.154 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Sat 2014-05-17 17:58:56 +0200
            message:
              Merge WL#7762
                ------------------------------------------------------------
                revno: 2876.579.119 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Sat 2014-05-17 17:56:19 +0200
                message:
                  Merge WL#7762
                    ------------------------------------------------------------
                    revno: 2875.612.96
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Sat 2014-05-17 17:53:17 +0200
                    message:
                      WL#7762: Fix multi-TC takeover handling
            ------------------------------------------------------------
            revno: 2876.577.153 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Fri 2014-05-16 14:33:49 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.118 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3-bug11762867
                timestamp: Fri 2014-05-16 14:23:33 +0200
                message:
                  Bug#11762867 DEFAULT VALUES FOR INITLOGFILEGROUP CAUSE ERROR 2350: 'INVALID CONFIGURAT
                  - Add testcase for the new error message
                  - Add testcase for starting cluster with InitialLogFileGroup with undo_buffer_size greater than
                   available  SharedGlobalMemory. Check that memory was allocated for undo buffer
                    and that the created log file group is usable.
                   - Slight change of error message to indicate that user can choose between reducing the undo_buffer_size
                     or increase SharedGlobalMemory
                   - Slight change of logic in the code which (pre)allocates the memory to only peek at
                     the undo_buffer_size value in the InitialLogFileGroup spec. This avoids duplicating all the logic from
                    NdbCntrMain.cpp and should make the behaviour more distinct.
                    ------------------------------------------------------------
                    revno: 2876.592.1
                    committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                    branch nick: 7.3
                    timestamp: Fri 2014-05-09 14:45:25 +0200
                    message:
                      Bug#11762867 - DEFAULT VALUES FOR INITLOGFILEGROUP CAUSE ERROR 2350: 'INVALID CONFIGURAT
                      
                      Problem:
                      
                      The config.ini specifies that a log file group should be created
                      during initial start of the cluster with the help of
                      InitialLogFileGroupSpec. When no undo_buffer_size=<size> is specified,
                      the default value for undo_buffer_size will be used.
                      
                       When cluster starts initial with default values, it fails to allocate
                       undo buffer of the requested size from the SharedGlobalMemory
                       pool. This is very likely in 7.0-1, where defaults for SGA is 20M and
                       undo_buffer_size is 64M.
                      
                       Suggested solution:
                      
                      1) At initial startup, allocate the memory for the undo_buffer_size
                      specified in the InitialLogFileGroupSpec or the default, in addition
                      to the sga.
                      
                      2)Improve the error message printed when ndbd detects that it
                      cannot allocate the undo buffer.
            ------------------------------------------------------------
            revno: 2876.577.152 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Fri 2014-05-16 13:56:37 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.117 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2014-05-16 13:49:25 +0200
                message:
                  Merge
            ------------------------------------------------------------
            revno: 2876.577.151 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Fri 2014-05-16 13:39:54 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.591.1 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2014-05-16 13:22:22 +0200
                message:
                  Merge patch for bug17018500 into 7.3
                    ------------------------------------------------------------
                    revno: 2876.590.1
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.3-bug17018500
                    timestamp: Mon 2014-04-07 15:35:05 +0200
                    message:
                      Bug#17018500 BOGUS DEFINER VALUE FOR NDBINFO VIEWS REPORTED IN INFORMATION_SCHEMA.VIEW
                       - Properly quote teh definer value in the views cretaed for ndbinfo
                       - Add query which checks that all ndbinfo views got expected definer value
            ------------------------------------------------------------
            revno: 2876.577.150
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-05-16 08:33:00 +0200
            message:
              This commit fixes two errors in the new regression test in the commit below:
              
              --------------------------------------------------------------------------
              #At file:///net/atum17/export/home/tmp/jw159207/mysql/repo/mysql-5.6-cluster-7.4/ based on revid:ole.john.aske@oracle.com-20140514124556-04t9xmxcwpndzi0x
              
               4333 Jan Wedvik	2014-05-14
                    This commit implements WL#7375 (Ndbinfo: Per-fragment memory usage reporting). It provides a new view, 'ndbinfo.memory_per_fragment', that gives an overview of memory used per fragment replica, including fixed-element-size pages, var-element-size pages, rows, fixed-element-free-slots, var-element-free-bytes, hash index memory usage etc.
                    
                    In addition, the ndbinfo.ndb$pools has been extended with new rows for L2PMap memory usage.
                    
                    Finally, there are MTR regression test cases for the new functionality.
            ------------------------------------------------------------
            revno: 2876.577.149 [merge]
            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
            branch nick: 7.4
            timestamp: Wed 2014-05-14 15:40:48 +0200
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.116 [merge]
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Wed 2014-05-14 15:24:49 +0200
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.95 [merge]
                    committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                    branch nick: 7.2
                    timestamp: Wed 2014-05-14 15:19:05 +0200
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.34
                        committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                        branch nick: 7.1
                        timestamp: Wed 2014-05-14 14:49:41 +0200
                        message:
                          Bug#18419907 - EASE DEBUGGING OF EVENT BUFFER BY ADDING AN IDENTIFIER TO ITS PRINT OUTS
            ------------------------------------------------------------
            revno: 2876.577.148
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-05-14 15:35:09 +0200
            message:
              This commit implements WL#7375 (Ndbinfo: Per-fragment memory usage reporting). It provides a new view, 'ndbinfo.memory_per_fragment', that gives an overview of memory used per fragment replica, including fixed-element-size pages, var-element-size pages, rows, fixed-element-free-slots, var-element-free-bytes, hash index memory usage etc.
              
              In addition, the ndbinfo.ndb$pools has been extended with new rows for L2PMap memory usage.
              
              Finally, there are MTR regression test cases for the new functionality.
            ------------------------------------------------------------
            revno: 2876.577.147 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-05-14 14:45:56 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.115 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-05-14 14:39:11 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.94 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-05-14 14:31:46 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.33
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-05-14 14:28:08 +0200
                        message:
                          Temporary workaround fix for bug#18759252
                          
                          AUTOTEST REGRESSION CAUSED BY NEW SYSTEM TABLES ADDED BY INDEX STATISTICS
                          
                          Force testDict to ignore the two system tables added by
                          index statistics. 
                          Hopefulle we will later be able to revert this patch when
                          we have a propper fix for this bug
            ------------------------------------------------------------
            revno: 2876.577.146 [merge]
            committer: Martin Zaun <martin.zaun@oracle.com>
            branch nick: ndb-7.4
            timestamp: Tue 2014-05-13 14:58:57 -0700
            message:
              merge 73->74
                ------------------------------------------------------------
                revno: 2876.579.114 [merge]
                committer: Martin Zaun <martin.zaun@oracle.com>
                branch nick: ndb-7.3
                timestamp: Tue 2014-05-13 14:57:30 -0700
                message:
                  merge 72->73
                    ------------------------------------------------------------
                    revno: 2875.612.93 [merge]
                    committer: Martin Zaun <martin.zaun@oracle.com>
                    branch nick: ndb-7.2
                    timestamp: Tue 2014-05-13 14:56:06 -0700
                    message:
                      merge 71->72
                        ------------------------------------------------------------
                        revno: 2555.843.32 [merge]
                        committer: Martin Zaun <martin.zaun@oracle.com>
                        branch nick: ndb-7.1
                        timestamp: Tue 2014-05-13 14:54:38 -0700
                        message:
                          merge from 7.1-crund
                            ------------------------------------------------------------
                            revno: 2555.846.1
                            committer: Martin Zaun <martin.zaun@oracle.com>
                            branch nick: ndb-7.1-crund
                            timestamp: Tue 2014-05-13 14:38:55 -0700
                            message:
                              crund - fixed minor bugs, updated usage documentation, code cleanup
            ------------------------------------------------------------
            revno: 2876.577.145
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-13 14:43:52 +0200
            message:
              Fix build break on Solaris in previous commit:
              
              #At file:///net/atum17/export/home/tmp/jw159207/mysql/repo/mysql-5.6-cluster-7.4/ based on revid:maitrayi.sabaratnam@oracle.com-20140509153539-ogxgpf9v85oyu14d
              
               4329 Jan Wedvik	2014-05-13
                    This commit is a fix for Bug#18717583 (UNINITIALIZED FRAGRECORD::ACCFRAGPTR USED IN LOG_FRAGMENT_COPIED() CRASHES NODE). The fix ensures that accFragptr is initialized to RNIL. It also adds a check for accFragptr==RNIL to Dblqh::log_fragment_copied(), so that it does not try to access nonexistent ACC fragments. (The latter would happen if the table was an ordered index, since ordered indexes do not have hash indexes.) Finally ndbd_malloc() is changed such that it initializes allocated bytes to 0xfb, to make it easier to detect use of uninitialized memory.
              
                  modified:
                    storage/ndb/src/kernel/blocks/dblqh/DblqhMain.cpp
                    storage/ndb/src/kernel/vm/ndbd_malloc.cpp
                    storage/ndb/src/kernel/vm/ndbd_malloc.hpp
            ------------------------------------------------------------
            revno: 2876.577.144
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-13 13:30:21 +0200
            message:
              This commit is a fix for Bug#18717583 (UNINITIALIZED FRAGRECORD::ACCFRAGPTR USED IN LOG_FRAGMENT_COPIED() CRASHES NODE). The fix ensures that accFragptr is initialized to RNIL. It also adds a check for accFragptr==RNIL to Dblqh::log_fragment_copied(), so that it does not try to access nonexistent ACC fragments. (The latter would happen if the table was an ordered index, since ordered indexes do not have hash indexes.) Finally ndbd_malloc() is changed such that it initializes allocated bytes to 0xfb, to make it easier to detect use of uninitialized memory.
            ------------------------------------------------------------
            revno: 2876.577.143 [merge]
            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
            branch nick: 7.4
            timestamp: Fri 2014-05-09 17:35:39 +0200
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.113 [merge]
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Fri 2014-05-09 17:24:57 +0200
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.92 [merge]
                    committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                    branch nick: 7.2
                    timestamp: Fri 2014-05-09 17:00:06 +0200
                    message:
                      Merge from 7.1
                        ------------------------------------------------------------
                        revno: 2555.843.31
                        committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                        branch nick: 7.1
                        timestamp: Fri 2014-05-09 16:51:32 +0200
                        message:
                          Bug#18703871 - POLLEVENTS RETURN 1 WHEN CALLED WITH NOWAIT, EVEN IF THERE IS NO EVENT AVAILABLE
                          
                          This contradicts the documented behaviour. Fix makes polEvent to return 0 when called with no wait and no event data found in the event queue.
            ------------------------------------------------------------
            revno: 2876.577.142
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-05-08 12:10:49 +0200
            message:
              Addendum patch for WL#7532 which introduced a build break 
              if compiled with VM_TRACE_TIME
            ------------------------------------------------------------
            revno: 2876.577.141 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-05-07 10:18:19 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.112
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-05-07 10:17:31 +0200
                message:
                  Reduce rqg_spj.test from running 60mins to 30mins.
                  Also reduce the timeout from 3h to 1h.
                  
                  (Will timeout in the very rare case that an 'incomputable' 
                  join was created randomly, or in case of error)
            ------------------------------------------------------------
            revno: 2876.577.140 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-05-07 09:33:41 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.111
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-05-07 09:31:50 +0200
                message:
                  Fix the failing ndb_big_addnode.test which started failing after
                  increasing ThreadConfig to 4 LDM threads.
                  
                  Similar changes has already been performed on the similar 
                  'non-big' ndb_addnode.test
            ------------------------------------------------------------
            revno: 2876.577.139 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-06 15:49:11 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.110 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-05-06 15:35:54 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.91
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-05-06 15:27:28 +0100
                    message:
                            Bug #18455971 	CLUSTER CRASH CAUSED BY ERROR: 2301 ILLEGAL SIGNAL RECEIVED (GSN 40 NOT ADDED)
                            
                            During data node failure handling, the taking-over TC will gather all the
                            known state for the failed TC instance(s) transactions, and determine
                            a state for each transaction - committed or aborted.
                            
                            Once this has been determined, it will inform the API node(s) involved so that
                            they can report the status accurately to clients.
                            
                            This is done by the taking over TC instance sending TCKEY_FAILREF or TCKEY_FAILCONF
                            signals to the API nodes.
                            
                            However it is possible that the taking over TC instance does not have
                            a direct connection to the API node.  Perhaps a connection was not
                            successfully made at the point of the node failure.
                            
                            In this case, a best-effort attempt to deliver the signal is made
                            by *routing* it via another node.  A node in the same nodegroup as the
                            failing TC is chosen, and a GSN_TCKEY_FAILREFCONF_R signal is sent
                            to TC block instance 0 in that node.
                            
                            With multi-TC, the TC block instance 0 will be an instance of the TC
                            proxy block.
                            
                            The TC proxy block does not have a signal handler for TCKEY_FAILREFCONF_R,
                            and so will crash.
                            
                            This patch adds a handler to the TC proxy block.  The behaviour is to 
                            forward the signal to one of the local TC worker instances, which will
                            behave as today (forward on to the Api if possible).
                            
                            A general 'forward signal' and 'round robin forward signal' mechanism
                            is added to the LocalProxy superclass.
                      
                            Additionally, some bugs with EXECUTE_DIRECT into DIH to find a nodegroup peer
                            were fixed.
            ------------------------------------------------------------
            revno: 2876.577.138 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-05-06 09:14:22 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.109
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-05-06 09:13:38 +0200
                message:
                  Reduction of 'MaxNoOfConcurrentScans' in 'default_ndbd.cnf' 
                  caused ndb_bushy_joins.test to fail due to 'out of scans'.
                  
                  Increased this config variable in the test specific config file.
                  Also removes the [ENV] setting as this is now included in the
                  new default config.
            ------------------------------------------------------------
            revno: 2876.577.137 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-05-05 15:10:20 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.108
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-05-05 14:45:27 +0200
                message:
                  Change default config for MTR tests such that the ndbmtd deamon
                  memory footprint is reduced from ~900M to ~400M.
            ------------------------------------------------------------
            revno: 2876.577.136 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-05-05 14:27:03 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.107 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-05-05 14:23:56 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.90 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-05-05 14:20:03 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.30
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-05-05 14:16:13 +0530
                        message:
                          Bug#17882305 "#SQL-*" TABLES SHOULD BE SKIPPED BY NDB_RESTORE
                          
                          Added an option --exclude-intermediate-tables, default TRUE, to 
                          ndb_restore. When this is set to TRUE, ndb_restore skips restoring 
                          tables whose names are prefixed with "#sql", thus skipping intermediate 
                          tables created during ALTER TABLE operations.
                          
                          Added a unit test to autotest to create an NDB table with a 
                          "#sql"-prefixed name and take a backup. This backup is then restored 
                          to a database which does not contain the "#sql"-prefixed table.
            ------------------------------------------------------------
            revno: 2876.577.135 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-04-30 12:32:59 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.106
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-04-30 12:32:08 +0200
                message:
                  ndb_backup_rate.test checks that we are able to recover normal
                  operation after a 'Redo log full' situation.
                  
                  This is done by configuring a slow 'DiskCheckpointSpeed=1M', and
                  letting the test program produce lots of logging by inserting and
                  deleting rows while waiting for a 'Redo log full' error.
                  
                  However, when running with debug compiled binaries, the insert + delete
                  performance could be so slow that we are unable to fill the redo log.
                  Thus this test often failed in the 'mix-debug' part of the test.
                  
                  This fix disable the test when run with debug binaries.
                  NOTE: It was already disabled for Valgrind for the same reasons.
            ------------------------------------------------------------
            revno: 2876.577.134 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-04-30 09:11:24 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.105
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-04-30 09:10:14 +0200
                message:
                  After the default_ndbd.cnf was changed to run with 4 LDM threads,
                  clusterj.test has been failing due to insufficient IndexMemory.
                  
                  This patch introduce clusterj.cnf which overrides the default
                  IndexMemory config for clusterj.test to 16M which has been
                  (experimentally) determined to be sufficient.
            ------------------------------------------------------------
            revno: 2876.577.133 [merge]
            committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-04-29 17:39:24 +0200
            message:
              Upmerge of the 7.1.31 build
                ------------------------------------------------------------
                revno: 2876.579.104 [merge]
                committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-04-29 17:34:42 +0200
                message:
                  Upmerge of the 7.1.31 build
                    ------------------------------------------------------------
                    revno: 2875.612.89 [merge]
                    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-04-29 17:28:58 +0200
                    message:
                      Upmerge of the 7.1.31 build
                        ------------------------------------------------------------
                        revno: 2555.843.29 [merge]
                        author: murthy.narkedimilli@oracle.com
                        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-04-29 17:22:18 +0200
                        message:
                          Merge from mysql-cluster-7.1.31-release
                            ------------------------------------------------------------
                            revno: 2555.845.2
                            tags: mysql-cluster-7.1.31
                            committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                            branch nick: mysql-cluster-7.1.31-release
                            timestamp: Tue 2014-04-01 17:56:22 +0200
                            message:
                              Applying the patch for bug 18426180 : Adding the files
                            ------------------------------------------------------------
                            revno: 2555.845.1
                            committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                            branch nick: mysql-cluster-7.1.31-release
                            timestamp: Tue 2014-04-01 17:14:32 +0200
                            message:
                              Applying the patch for bug 18426180 which needs to be part of 7.1.31 build.
            ------------------------------------------------------------
            revno: 2876.577.132
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7847
            timestamp: Tue 2014-04-29 15:30:15 +0200
            message:
              WL#7847: Ensure scanFragState and tcConnectstate are in same state always in free list
            ------------------------------------------------------------
            revno: 2876.577.131 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-04-29 13:34:06 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.103
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-04-29 13:33:08 +0200
                message:
                  Adjust 'initial size' of tablespace used as part of ndb_restore_conv_more.test.
                  
                  Required by introducing more LDM threads in the ThreadConfig.
            ------------------------------------------------------------
            revno: 2876.577.130 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug16895311-74
            timestamp: Mon 2014-04-28 20:12:15 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.102 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug16895311-73
                timestamp: Mon 2014-04-28 19:53:51 +0300
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.88 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug16895311-72
                    timestamp: Mon 2014-04-28 19:22:57 +0300
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.843.28
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug16895311-71
                        timestamp: Mon 2014-04-28 14:40:32 +0300
                        message:
                          bug#16895311 bug16895311.diff
                          checkUpdateOfPrimaryKey: account for xfrm size
                        ------------------------------------------------------------
                        revno: 2555.843.27
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug16895311-71
                        timestamp: Mon 2014-04-28 14:39:52 +0300
                        message:
                          bug#16895311 testnr.diff
                          test NR with long UTF8 PK
            ------------------------------------------------------------
            revno: 2876.577.129 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-04-28 13:29:49 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.101
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-04-28 13:28:12 +0200
                message:
                  Removed the MTR test ndb_mt_recv.test:
                  
                  This test was introduced in order to have a minimum of testing
                  with multiple receiver threads in the ThreadConfig. The testload
                  itself for this test, was the ndb_basic.test
                  
                  As multiple receiver threads now has been made part of default_ndbd.cnf,
                  this special variant of ndb_basic.test is obsolete.
            ------------------------------------------------------------
            revno: 2876.577.128 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-04-25 15:51:59 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.100
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2014-04-25 15:50:48 +0200
                message:
                  As part of preparing the MTR test suite for running with 4 LDM threads,
                  the size of the ts_1 tablespace in ndb_addnode.test was increased
                  from 4M to 12M. This was tested, and wound sufficient on Linux and
                  x86 Solaris.
                  
                  However, in Sparc Solaris, we observe that we get a 't2 table is full'
                  even with this setting.
                  
                  This fix increase the tablespace size to 16M, which is 4 times
                  the original size used with a single LDM thread.
            ------------------------------------------------------------
            revno: 2876.577.127 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-04-25 11:35:32 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.99
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2014-04-25 11:33:54 +0200
                message:
                  ndb_index_stat_partition.test failed on some platforms after increasing number
                  of LDM threads to 4. This was due to index statistics only being sampled 
                  from one of the LDM's. Thus increasing #LDM's, also decreased the size of the
                  sample, which gave more variations in the statistics.
                  
                  As ndb_index_stat_partition.test required the estimated number of rows
                  to within 45% of the actuall number of rows, more variations in the 
                  statistics caused this test to fail.
                  
                  This fix doubles the number of rows in the test table in
                  order to give better statistcs estimates when number of 
                  LDMs are increased.
            ------------------------------------------------------------
            revno: 2876.577.126 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-04-23 13:07:01 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.98
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-04-23 13:00:35 +0200
                message:
                  Index statistics will only sample rows from a single fragments pr. datanodes.
                  With few rows in the test tables, we might get rather huge variations in the
                  statistics, somethimes causing an unexpected query plan to be choosen.
                  
                  ndb_index.test contains a testcase depending on index_merge
                  being used as access methods. This was not always used due to 
                  reasons described above.
                  
                  This fix increases number of rows used by this testcase, such that the
                  index statistics will be more stable.
            ------------------------------------------------------------
            revno: 2876.577.125 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-04-22 10:57:43 +0200
            message:
              Merge 7.3->7.4.
                ------------------------------------------------------------
                revno: 2876.579.97 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-04-22 10:47:47 +0200
                message:
                  Merge 7.2->7.3.
                    ------------------------------------------------------------
                    revno: 2875.612.87 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-04-22 10:38:21 +0200
                    message:
                      Merge 7.1->7.2.
                        ------------------------------------------------------------
                        revno: 2555.843.26
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-04-22 10:31:10 +0200
                        message:
                          This commit is a fix for Bug#18550318 (HOST REBOOT CAUSES CRASH,
                          INTERNAL ERROR, DBDIH (LINE: 18314) 0X00000002). This bug is a
                          regression caused by the fix for Bug#14220269 (CLUSTER WIDE SHUTDOWN
                          POINTER TOO LARGE IN DBDIH (LINE: 9290) (TCKEYREQ)). That fix added 
                          a new dump handler using a dump code that was already in use (7019). 
                          This meant that 'dump 7019' would execute two different handlers with 
                          different semantics and different interpretation of the parameters. 
                          'Dump 7019' is called automatically from Qmgr::checkStartInterface() 
                          if node failure handling takes more than a minute. This triggered 
                          an ndbrequire in the new dump handler. 
                          
                          The fix consists in allocating a new dump code for the new dump
                          handler (7024). A DumpStateType enum literal for 7019 is also defined, to
                          make similar errors less likely.
            ------------------------------------------------------------
            revno: 2876.577.124 [merge]
            committer: Sreedhar S <sreedhar.sreedhargadda@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-04-09 09:09:03 +0200
            message:
              Upmerge of the 7.2.16 build
                ------------------------------------------------------------
                revno: 2876.579.96 [merge]
                committer: Sreedhar S <sreedhar.sreedhargadda@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-04-09 09:01:37 +0200
                message:
                  Upmerge of the 7.2.16 build
                    ------------------------------------------------------------
                    revno: 2875.612.86 [merge]
                    author: sreedhar.sreedhargadda@oracle.com
                    committer: Sreedhar S <sreedhar.sreedhargadda@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-04-09 07:39:16 +0200
                    message:
                      Merge from mysql-cluster-7.2.16-release
                        ------------------------------------------------------------
                        revno: 2875.621.1
                        tags: mysql-cluster-7.2.16
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: cluster-7.2.16-release
                        timestamp: Tue 2014-04-01 16:37:12 +0200
                        message:
                          ndb - cherrypicked fix
                          
                            revision-id: mauritz.sundell@oracle.com-20140401081334-nrve3gyqwgf2b6vx
                            parent: frazer.clement@oracle.com-20140331161153-5b4yccimejdo13v9
                            parent: mauritz.sundell@oracle.com-20140401081054-quhto0gfak217e1n
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-7.2
                            timestamp: Tue 2014-04-01 10:13:34 +0200
                            message:
                              merge 7.1 -> 7.2
                                ------------------------------------------------------------
                                revno: 2555.840.20
                                revision-id: mauritz.sundell@oracle.com-20140401081054-quhto0gfak217e1n
                                parent: frazer.clement@oracle.com-20140331155509-nscusf07z7kjwp47
                                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                                branch nick: mysql-7.1
                                timestamp: Tue 2014-04-01 10:10:54 +0200
                                message:
                                  ndb - regression: corrupt message detected with tcp checksum activated
                                  Bug #18426180    ENHANCED DETECTION AND DUMP OF CORRUPT MESSAGES AND SIGNALS RECEIVED
            ------------------------------------------------------------
            revno: 2876.577.123 [merge]
            committer: MySQL Build Team <mysql-build@oss.oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-04-07 14:17:02 +0200
            message:
              Upmerge of the 7.3.5 build
                ------------------------------------------------------------
                revno: 2876.579.95 [merge]
                author: balasubramanian.kandasamy@oracle.com
                committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-04-07 14:03:36 +0200
                message:
                  Merge from mysql-cluster-7.3.5-release
                    ------------------------------------------------------------
                    revno: 2876.589.1
                    tags: mysql-cluster-7.3.5
                    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                    branch nick: mysql-cluster-7.3.5-release
                    timestamp: Tue 2014-04-01 17:42:30 +0200
                    message:
                      Applied patch for Bug#18426180 - ENHANCED DETECTION AND DUMP OF CORRUPT MESSAGES AND SIGNALS RECEIVED
                      
                      ndb - regression: corrupt message detected with tcp checksum activated
            ------------------------------------------------------------
            revno: 2876.577.122 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-04-04 18:13:27 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.94 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2014-04-04 17:59:08 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.85 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2014-04-04 17:49:32 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.25
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Fri 2014-04-04 17:45:57 +0100
                        message:
                          Bug #18518037 	NDB : MT RECEIVER DOES NOT PARTITION [DIS]CONNECT WORK CORRECTLY
                          
                          Fix avoids all Trpman instances (receiver threads) executing OPEN_COMORD,
                          ENABLE_COMREQ and CLOSE_COMREQ.
                          
                          Also fixes trpman ROUTE_ORD processing so that the correct Trpman instance
                          is chosen when routing 'from' the trpman instance for a specific remote
                          connection.
            ------------------------------------------------------------
            revno: 2876.577.121 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-04-04 10:18:10 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.93
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2014-04-04 10:14:41 +0200
                message:
                  Fix for Enhancement Request Bug#18490331
                  
                    DEFAULT CONFIG FOR MTR TESTS SHOULD BE MORE MULTITHREADED
                  
                  Set the default config for NDB datanodes such that we actually
                  runs the test with ndbmtd *and* multiple worker threads.
            ------------------------------------------------------------
            revno: 2876.577.120 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-04-03 22:54:49 +0200
            message:
              Merge fix 2 for BUG#18449222
                ------------------------------------------------------------
                revno: 2876.579.92 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-04-03 22:53:50 +0200
                message:
                  Merge fix 2 for BUG#18449222
                    ------------------------------------------------------------
                    revno: 2875.612.84 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-04-03 22:51:27 +0200
                    message:
                      Merge fix 2 for BUG#18449222
                        ------------------------------------------------------------
                        revno: 2555.843.24
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-04-03 22:48:26 +0200
                        message:
                          Fix 2 for BUG#18449222
            ------------------------------------------------------------
            revno: 2876.577.119 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-04-03 22:41:04 +0200
            message:
              Merge fix for BUG#18449222
                ------------------------------------------------------------
                revno: 2876.579.91 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-04-03 22:39:56 +0200
                message:
                  Merge fix for BUG#18449222
                    ------------------------------------------------------------
                    revno: 2875.612.83 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-04-03 22:38:20 +0200
                    message:
                      Merge fix for BUG#18449222
                        ------------------------------------------------------------
                        revno: 2555.843.23
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-04-03 22:36:58 +0200
                        message:
                          Fix for BUG#18449222 push
            ------------------------------------------------------------
            revno: 2876.577.118 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_bug18449222_74
            timestamp: Thu 2014-04-03 15:10:41 +0200
            message:
              Merge BUG18449222
                ------------------------------------------------------------
                revno: 2876.579.90 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: push_bug18449222_73
                timestamp: Thu 2014-04-03 14:52:42 +0200
                message:
                  Merge BUG18449222
                    ------------------------------------------------------------
                    revno: 2875.612.82 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: push_bug18449222_72
                    timestamp: Thu 2014-04-03 14:51:13 +0200
                    message:
                      Merge BUG18449222
                        ------------------------------------------------------------
                        revno: 2555.843.22
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: push_bug18449222_71
                        timestamp: Thu 2014-04-03 14:49:33 +0200
                        message:
                          BUG#18449222: Timeout waiting for local DIH caused crash, removed ndbrequire
            ------------------------------------------------------------
            revno: 2876.577.117
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-04-03 13:45:39 +0200
            message:
              Bug #16505832 A FEW MINOR DBUG PROBLEMS....
              
               - add missing DBUG_RETURN in functions which started with DBUG_ENTER
            ------------------------------------------------------------
            revno: 2876.577.116 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-04-03 11:13:35 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.89 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2014-04-03 11:12:47 +0200
                message:
                  Merge
                    ------------------------------------------------------------
                    revno: 2876.588.1
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.3
                    timestamp: Thu 2014-04-03 11:05:12 +0200
                    message:
                      Remove temp. hack for bug#36431(aka. bug#11748453) since ndb_update_no_read now works
                      fine also for embedded. Closing the bug.
            ------------------------------------------------------------
            revno: 2876.577.115 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-04-02 13:17:56 +0200
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.88
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-04-02 13:16:14 +0200
                message:
                  Fix for bug#18490173:
                  
                  TEST RESULTS SHOULD NOT DEPEND ON NUMBER OF TC / LDM THREADS IN CONFIG
                  
                  Several MTR tests are modified such that the expected *.result 
                  file is independent of how many LDM / TC threads are configured, 
                  and on whether the ndbd or ndbmtd is used.
                  
                  Most changes are related to default 'partiotions' being a function
                  of number of LDM's. Thus explicit 'partiotions <n>' spec has been 
                  added to several create table statements. 
                  
                  A few places the test itself tests that the expected number of
                  default partitions are created. In these places we have 
                  included 'ThreadConfig=ldm={count=1}' in the *.cnf file.
                  
                  Also had to increase setting of some config variable in order to 
                  account for possible more partitions being created with bigger
                  configs.
                  
                  Also fixed a testcase for bug#58280 (ndb_index.test), which
                  required that 'index_merge' access method was used to test
                  aganst a previous defect. (Added more non-matching rows to
                  test table to make 'index_merge' more favourable)
                  
                  In the join_pushdown test, I have now removed verification of
                  some 'info_counters'. These has in the past caused several
                  issues due to failing test, and maintaining these are IMO not
                  worth the effort anymore.
            ------------------------------------------------------------
            revno: 2876.577.114 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-04-01 14:34:51 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.87 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-04-01 14:31:48 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.81 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-04-01 14:20:28 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.21
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-04-01 14:18:04 +0100
                        message:
                          Bug # 17461576 	NDB REPLICATION : STOP SLAVE WHEN MASTER EPOCH DECLINES
                          
                          Avoid master switch over testcase resulting in irrelevant epoch decline warnings.
            ------------------------------------------------------------
            revno: 2876.577.113 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-04-01 13:57:35 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.86
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-04-01 13:56:03 +0100
                message:
                  Cherry pick fix for BUG#18482854 RPL : ROTATE_LOG_EVENT INCORRECTLY ADVANCES GROUP_RELAY_LOG_POS IN A GROUP
            ------------------------------------------------------------
            revno: 2876.577.112 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-04-01 10:24:25 +0200
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.85 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-04-01 10:18:09 +0200
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.80 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2014-04-01 10:13:34 +0200
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.20
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Tue 2014-04-01 10:10:54 +0200
                        message:
                          ndb - regression: corrupt message detected with tcp checksum activated
                          Bug #18426180    ENHANCED DETECTION AND DUMP OF CORRUPT MESSAGES AND SIGNALS RECEIVED
            ------------------------------------------------------------
            revno: 2876.577.111 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-04-01 10:20:32 +0200
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.587.1 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-03-31 22:48:41 +0200
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.620.1 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Mon 2014-03-31 22:47:32 +0200
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.844.1
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Mon 2014-03-31 22:44:56 +0200
                        message:
                          ndb - regression: corrupt message detected with tcp checksum activated
                          Bug #18426180    ENHANCED DETECTION AND DUMP OF CORRUPT MESSAGES AND SIGNALS RECEIVED
            ------------------------------------------------------------
            revno: 2876.577.110 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-04-01 00:50:15 +0100
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.84 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-04-01 00:13:55 +0100
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.79 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-03-31 17:11:53 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.19
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-03-31 16:55:09 +0100
                        message:
                                Bug #17461625 	NDB REPLICATION : IMPROVE NDB_BINLOG_INDEX PROBLEM REPORTING
                                
                                Occasionally issues are seen writing to the mysql.ndb_binlog_index table, 
                                generally error 121 (duplicate key).
                                
                                One theory is that the --ndb-log-orig option is involved in some way so that the
                                same epoch key values are inserted twice for some reason.
                                
                                To help test this theory, the logging of insert failures is improved.
                                
                                This should shine some light on the failures when they occur in the field.
                                
                                A new testcase is added which tests the operation of these warnings using
                                error insertion.
            ------------------------------------------------------------
            revno: 2876.577.109 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-03-31 10:49:42 +0200
            message:
              null merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.83 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-03-31 10:45:28 +0200
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.78 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Mon 2014-03-31 10:42:42 +0200
                    message:
                      null merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.18
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Mon 2014-03-31 10:39:00 +0200
                        message:
                          bump version to 7.1.32
                    ------------------------------------------------------------
                    revno: 2875.612.77
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Mon 2014-03-31 10:40:11 +0200
                    message:
                      bump version to 7.2.17
                ------------------------------------------------------------
                revno: 2876.579.82
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-03-31 10:41:12 +0200
                message:
                  bump version to 7.3.6
            ------------------------------------------------------------
            revno: 2876.577.108 [merge]
            tags: clone-mysql-5.6.17-ndb-7.4.0-src-build
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Sat 2014-03-29 16:08:00 +0100
            message:
              null merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.81
                tags: clone-mysql-5.6.17-ndb-7.3.5-src-build
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Sat 2014-03-29 16:06:59 +0100
                message:
                  ndb - null commit to trigger push build
            ------------------------------------------------------------
            revno: 2876.577.107 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Fri 2014-03-28 12:33:25 +0100
            message:
              null merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.80 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Fri 2014-03-28 12:31:07 +0100
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.76 [merge]
                    tags: clone-mysql-5.5.37-ndb-7.2.16-src-build
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Fri 2014-03-28 12:28:28 +0100
                    message:
                      merge 5.5.37 into 7.2.16 (from 5.5 tree)
                    ------------------------------------------------------------
                    revno: 2875.612.75
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Fri 2014-03-28 12:27:18 +0100
                    message:
                      merge 5.5.37 into 7.2.16 (from release tree)
            ------------------------------------------------------------
            revno: 2876.577.106 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Fri 2014-03-28 12:19:42 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.79 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Fri 2014-03-28 12:16:16 +0100
                message:
                  merge 5.6.17 into 7.3.5 with tag
                ------------------------------------------------------------
                revno: 2876.579.78
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Fri 2014-03-28 12:13:09 +0100
                message:
                  merge 5.6.17 into 7.3.5
            ------------------------------------------------------------
            revno: 2876.577.105 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Thu 2014-03-27 21:03:55 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.77
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Thu 2014-03-27 20:11:51 +0100
                message:
                  WL#7774 Make MTR sync_slave_sql_with_master.inc handle NDB
                  
                  The sync_slave_with_master command in mysqltest is patched in
                  MySQL Cluster to make sure that the current epoch is binlogged
                  entirely before getting master binlog position.
                  Bug #18375920 NDB TESTING : SYNC_SLAVE_WITH_MASTER CAN TIMEOUT AND CONTINUE
                  
                  sync_slave_with_master are more and more being replaced by
                  include/sync_slave_sql_with_master.inc, for example in coming
                  WL#7205 Make the MTR rpl suite GTID_MODE agnostic
                  
                  This patch imitates the functionality added to sync_slave_with_master
                  in sync_slave_sql_with_master.inc by adding a file
                  wait_for_ndb_committed_to_binlog.inc with logic that first look at
                  latest_trans_epoch in SHOW ENGINE NDB STATUS and then waits until
                  latest_handled_binlog_epoch catch up.
                  
                  Changes are only done to mysql-test subsystem.
            ------------------------------------------------------------
            revno: 2876.577.104 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Thu 2014-03-27 17:15:36 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2876.579.76 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Thu 2014-03-27 16:47:09 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.74 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Thu 2014-03-27 16:43:23 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.17
                        tags: clone-mysql-5.1.73-ndb-7.1.31-src-build
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Thu 2014-03-27 16:33:49 +0100
                        message:
                          Bug#18436558 ALTER TABLE ADD COLUMN FAIL TO COPY HASHMAP TO NEW TABLE DEFINITION
                          
                          For ALTER TABLE there table schema are changed without change of
                          partitioning, the new table definition will not copy the hashmap from
                          the old defintion but use the current default hashmap - but will not
                          reorganize data according to the new hashmap!
                          
                          This makes some rows inaccessible via primary key lookup if the current
                          default hashmap is not compatible with the hashmap of the old definition.
                          
                          Checks are added both in kernel and in ndbapi that an alter table request will not
                          contain a hashmap change without also trigger reorganisation.
                          
                          Also the hashmap will now be copied when copying a new table definition from an old.
                          
                          Test ndb_addnode will fail if one do not apply the part of patch that changes
                          NdbTableImpl::assign().
            ------------------------------------------------------------
            revno: 2876.577.103 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-03-27 14:13:39 +0100
            message:
              Merge 7.3->7.4.
                ------------------------------------------------------------
                revno: 2876.579.75 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-03-27 14:09:04 +0100
                message:
                  Merge 7.2->7.3.
                    ------------------------------------------------------------
                    revno: 2875.612.73 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-03-27 14:04:05 +0100
                    message:
                      Merge 7.1->7.2.
                        ------------------------------------------------------------
                        revno: 2555.843.16
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-03-27 14:00:13 +0100
                        message:
                          This if a followup to the commit below. That commit would not compile with g++ 4.2.1. This one hopefully will.
                          ------------------------------------------------------
                          
                          #At file:///net/atum17/export/home/tmp/jw159207/mysql/repo/mysql-5.1-telco-7.1/ based on revid:frazer.clement@oracle.com-20140326005112-xb8a0shvx5jiwkv1
                          
                           4985 Jan Wedvik	2014-03-26
                                This commit implements a subset of WL#6562 ("Improve LongMessageBuffer stats and shortage behaviour"). It does three things:
                                
                                1) If the LMB pool is emptied, it prints an informative message in the ndb_X_out.log file, that explains what happened and what might be done about it.
                                
                                2) It increases the default size of the pool from 4MB to 64MB.
                                
                                3) It makes the LMB pool status visible through ndbinfo.
            ------------------------------------------------------------
            revno: 2876.577.102 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-03-26 15:38:34 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.74 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Wed 2014-03-26 15:17:42 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.72
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-03-26 15:14:02 +0100
                    message:
                      Bug#18366909 MGMD BINARY CONFIG LOOSES SERVERPORT AND OTHER SETTINGS WITH TYPE KEY_INTERNAL
                      
                       - follow up fix which moves the two new config parameters up a notch to
                        avoid conflict with already used number 634
                    ------------------------------------------------------------
                    revno: 2875.612.71 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-03-26 14:38:59 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.15
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Wed 2014-03-26 14:26:36 +0100
                        message:
                          Use mtr.add_suppression(<pattern>) instead of direct access to the mtr.test_suppression table. This
                          should improve the tests resiliency against future changes in the test driver framework.
                    ------------------------------------------------------------
                    revno: 2875.612.70 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-03-26 14:32:15 +0100
                    message:
                      Merge
                        ------------------------------------------------------------
                        revno: 2875.619.2
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Fri 2014-03-14 13:42:44 +0100
                        message:
                          Bug#18403670 VERIFY THAT ENUM VALUE IN BINARY CONFIG ARE CONSECUTIVELY ASCENDING
                        ------------------------------------------------------------
                        revno: 2875.619.1
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Fri 2014-03-14 13:18:53 +0100
                        message:
                          Bug#18366909 MGMD BINARY CONFIG LOOSES SERVERPORT AND OTHER SETTINGS WITH TYPE KEY_INTERNAL
                          
                           - Neither ServerPort or TcpBind_INADDR_ANY ended up in the binary config since they
                            was marked as internal. This was probably done intentionally since both these settings are
                            only used to calculate values which end up in other places of the config. But it's hard to
                            reverse engineer and thus it's better to add both the original value and the calculated value
                            to the binary config. There are more options marked as internal but they are either deprecated or
                            undocumented.
                           - Fix by giving the two options a proper value and ass testcase to show the values with ndb_config.
            ------------------------------------------------------------
            revno: 2876.577.101 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-26 12:59:21 +0100
            message:
              Merge 7.3->7.4.
                ------------------------------------------------------------
                revno: 2876.579.73 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-26 12:54:01 +0100
                message:
                  Merge 7.2->7.3.
                    ------------------------------------------------------------
                    revno: 2875.612.69 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-03-26 12:49:23 +0100
                    message:
                      Merge 7.1->7.2.
                        ------------------------------------------------------------
                        revno: 2555.843.14
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-03-26 12:47:16 +0100
                        message:
                          This is a followup to the commit below. The new test cases needed 
                          'order by node_id' to be deterministic.
                          
                          ---------------------------------------------------------------------------
                          
                          #At file:///net/atum17/export/home/tmp/jw159207/mysql/repo/mysql-5.1-telco-7.1/ based on revid:frazer.clement@oracle.com-20140326005112-xb8a0shvx5jiwkv1
                          
                           4985 Jan Wedvik	2014-03-26
                                This commit implements a subset of WL#6562 ("Improve LongMessageBuffer stats and shortage behaviour"). It does three things:
                                
                                1) If the LMB pool is emptied, it prints an informative message in the ndb_X_out.log file, that explains what happened and what might be done about it.
                                
                                2) It increases the default size of the pool from 4MB to 64MB.
                                
                                3) It makes the LMB pool status visible through ndbinfo.
            ------------------------------------------------------------
            revno: 2876.577.100 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-26 09:43:21 +0100
            message:
              Merge 7.3->7.4.
                ------------------------------------------------------------
                revno: 2876.579.72 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-26 09:35:54 +0100
                message:
                  Merge 7.2->7.3.
                    ------------------------------------------------------------
                    revno: 2875.612.68 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-03-26 09:25:27 +0100
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.13
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-03-26 09:18:05 +0100
                        message:
                          This commit implements a subset of WL#6562 ("Improve LongMessageBuffer stats and shortage behaviour"). It does three things:
                          
                          1) If the LMB pool is emptied, it prints an informative message in the ndb_X_out.log file, that explains what happened and what might be done about it.
                          
                          2) It increases the default size of the pool from 4MB to 64MB.
                          
                          3) It makes the LMB pool status visible through ndbinfo.
            ------------------------------------------------------------
            revno: 2876.577.99 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-26 01:32:44 +0000
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.71 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-26 01:22:19 +0000
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.67 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-03-26 01:05:54 +0000
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.843.12
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-03-26 00:51:12 +0000
                        message:
                          Bug #17461576 	NDB REPLICATION : STOP SLAVE WHEN MASTER EPOCH DECLINES
                                
                          Fix which monitors the epochs received from the immediately 
                          upstream master, determined by server_id column == server_id metadata.
                                
                          Specifically, an observed epoch decline will have different results 
                          depending on the state of the Slave SQL thread.
                                
                            1.  Just after a RESET SLAVE statement
                                No effect
                            2.  Just after a START SLAVE statement
                                A warning is produced
                            3.  None of the above.
                                The Slave SQL thread will stop.
                                
                          The intention with 1) is to enable warning free setting.
                          The intention with 2) is to warn when a Slave is being positioned on 
                          a previously applied epoch.
                          The intention with 3) is to stop the Slave when there is some system
                          malfunction resulting in the re-application of an existing epoch.
                                
                          Testing is performed using error insertion on the upstream master.
                                
                          A new testcase, ndb_rpl_slave_replay is added to the ndb_rpl suite.
            ------------------------------------------------------------
            revno: 2876.577.98 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-03-25 07:41:05 +0100
            message:
              merge 7.3 -> 7.4 (revert bad merge of clone-5.6.17)
                ------------------------------------------------------------
                revno: 2876.579.70 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-03-25 07:38:44 +0100
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.66
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2014-03-25 07:34:43 +0100
                    message:
                      revert bad clone-5.5.27 into 7.2.16
                ------------------------------------------------------------
                revno: 2876.579.69
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-03-25 07:36:49 +0100
                message:
                  revert bad merge merge clone-5.6.17 into 7.3.5
            ------------------------------------------------------------
            revno: 2876.577.97 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-03-25 00:49:30 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.68 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-03-25 00:47:25 +0100
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.65 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2014-03-25 00:44:47 +0100
                    message:
                      merge clone-5.5.27 into 7.2.16 ignoring changes in ndb files
                ------------------------------------------------------------
                revno: 2876.579.67 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-03-25 00:40:42 +0100
                message:
                  merge clone-5.6.17 into 7.3.5 ignoring changes in ndb files
------------------------------------------------------------
revno: 8419
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk-clean
timestamp: Thu 2014-07-10 18:56:59 +0700
message:
  This path is a follow-up for patch for bug#19022040 (SPORADICAL FAILURE IN THE
  TC_LOG_MMAP UNIT TEST).
  
  This patch fixes the bug in proper cleanup of TC_Log_MMap_thread on object destroying.
------------------------------------------------------------
revno: 8418
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug16268055update
timestamp: Thu 2014-07-10 10:46:28 +0200
message:
  Bug#19142753 ASSERT: MODE !=LOCK_X || LOCK_TABLE_HAS(THR_GET_TRX(THR), INDEX->TABLE, LOCK_IX)
  
  WL6742-Improve InnoDB SELECT COUNT(*) performance by using handler::records()
  introduced new functionality in InnoDB.
  
  handler::records can now fail in ways never seen by the server before.
  
  This improved things on the InnoDB side:
  kevin.lewis@oracle.com-20130626194505-muocxuem7bd2wiw8
    bug#16802288 - FAILING ASSERTION: PREBUILT->SQL_STAT_START ||
    TRX->STATE == TRX_STATE_ACTIVE
    
    Add error handling for DB_DEADLOCK and DB_LOCK_TABLE_FULL to
    handler::records() for COUNT(*).  Also rollback transaction when
    handler::records receive DB_DEADLOCK, DB_LOCK_TABLE_FULL or
    DB_LOCK_WAIT_TIMEOUT.
  
  However: thd_mark_transaction_to_rollback() isn't enough.
  The optimizer needs to abort immediately for the cases mentioned above.
  
  Fix: improve error handling for all calls to handler::records()
  Add a new public function ha_records() which does the error checking,
  make handler::records() protected.
------------------------------------------------------------
revno: 8417
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Thu 2014-07-10 12:07:52 +0800
message:
  Fix Bug #19063836 - ASSERT !DICT_INDEX_IS_SPATIAL(INDEX), CAN NOT MOVE A
  GEOMETRY COLUMN 
  
  rb://5932 approved by Marko
------------------------------------------------------------
revno: 8416 [merge]
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 16:01:35 +0100
message:
  BUG#18844897
  
  Automerged into latest mysql-trunk.
    ------------------------------------------------------------
    revno: 8301.1.4
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Fri 2014-07-04 16:46:34 +0100
    message:
      BUG#18844897
      
      Fixed some instability in the test case.
    ------------------------------------------------------------
    revno: 8301.1.3
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Fri 2014-07-04 15:18:33 +0100
    message:
      BUG#18844897
      
      Removed trailing whitespaces.
    ------------------------------------------------------------
    revno: 8301.1.2
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Fri 2014-07-04 15:06:53 +0100
    message:
      BUG#18844897
      
      Addressing review comments.
    ------------------------------------------------------------
    revno: 8301.1.1
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Thu 2014-07-03 10:35:38 +0100
    message:
      BUG#18844897:  ASSERT AT INFO.DRY_RUN
      
      A test case configured with --skip-innodb and that checks whether the
      slave server is able to restart and replication is able to either
      restart or step into a degraded mode when using slave_master_info and
      slave_relay_log_info tables on InnoDB, was failing with two
      crashes. One during the shutdown and another during recovery on server
      restart. The former, even though not entirely related to replication,
      was triggering the latter. Since the former crash was sporadic, so was
      the subsequent one. This was noticed in that test, because the test
      stops and restarts the slave server several times.
      
      The core of the problem was that the shutdown crash was causing the
      server not to close the binary log properly. Then, on server restart,
      while opening the the binary log, the server would find that it had
      not been closed properly. Therefore, it would engage the log for
      recovery and roll forward or roll back the pending prepared
      transactions in the engine. This results in the server scanning the
      binary log, collect xids - in this case none, since the tests was
      running with MyISAM (remeber --skip-innodb?) - and call ha_recover.
      But, inside, this function would run into an assert that checks
      whether there are sufficient parties to resume the pending prepared
      phases of the 2PC. Since the engine involved was MyISAM and no xid had
      been collected and ha_recover was called regardless, this assert would
      be triggered.
      
      This patch fixes this issue by checking if there are enough 2PC parties 
      involved before calling into ha_recover.
------------------------------------------------------------
revno: 8415
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt5
timestamp: Wed 2014-07-09 16:46:13 +0200
message:
  Bug#18637443 USABILITY IMPROVEMENTS TO N TESTS 
  
  * allow passing workload options such as --blob, directly to ibtest_ctd.pl, without bracketing them with --swlopt
    so instead of --swlopt --fts=0,--adipm=50 it is possible to just say --fts 0 --adipm 50
  * --swlopt is kept but deprecated
  * implement option where ibtest_ctd.pl seed value will determine
    seed values for all invocations of ibtest_ctd_w.pl during the test. This new option is --gseeds,
    enabled by default. For now the old code where workload seeds have been determined by the workload itself,
    is still supported.
  * eliminate ntest/tests/ibtest* workloads. ibtest_ctd_w.pl is robust enough now to be single workload. 
    It has been a while 
    since I saw a bug coming from ibtests. Remove ibtest_ctd.pl option --ibworkload.
    Keep --wob but make it non functional and deprecate it.
  * remove most files in ntest/tests/n/n_ct/rbs. They are not essential for testing.
  
  Note: both ibtests and everything under rbs will still be available in the development n tree under svn.
  They will disappear only from the version distributed under mysql-trunk/internal/nroot
  
  * more consistent pattern of RNG use in ibtest_ctd_w.pl and ibtest_ctd.pl which should help in better failure reproducibility
    when random seed replay is used:
    roll(N) < numeric-option-value and|or boolean-option-value
  * smaller improvements and bug fixes: 
  ** sa.scn: improvements in failure to known bug reconciliation
  ** ibtest_ctd.pl: 'mysqld startup failed message was seen in obtest_ctd.log even for a successful startup'
  * rein.pl: better output, update to changed interface to ibtest_ctd.pl and ibtest_ctd_w.pl
  * simpler interface to 'n rein', the script that recovers random seed
    and other similar information from n test incident directory
  * n.pm: more robust getlogdir(), allow values like '2' standing for incident_ndb.2
  
  rb://5462 approved by Krunal Bauskar
------------------------------------------------------------
revno: 8414
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt5
timestamp: Wed 2014-07-09 16:41:38 +0200
message:
  Bug#18637443 USABILITY IMPROVEMENTS TO N TESTS 
  
  * allow passing workload options such as --blob, directly to ibtest_ctd.pl, without bracketing them with --swlopt
    so instead of --swlopt --fts=0,--adipm=50 it is possible to just say --fts 0 --adipm 50
  * --swlopt is kept but deprecated
  * implement option where ibtest_ctd.pl seed value will determine
    seed values for all invocations of ibtest_ctd_w.pl during the test. This new option is --gseeds,
    enabled by default. For now the old code where workload seeds have been determined by the workload itself,
    is still supported.
  * eliminate ntest/tests/ibtest* workloads. ibtest_ctd_w.pl is robust enough now to be single workload. 
    It has been a while 
    since I saw a bug coming from ibtests. Remove ibtest_ctd.pl option --ibworkload.
    Keep --wob but make it non functional and deprecate it.
  * remove most files in ntest/tests/n/n_ct/rbs. They are not essential for testing.
  
  Note: both ibtests and everything under rbs will still be available in the development n tree under svn.
  They will disappear only from the version distributed under mysql-trunk/internal/nroot
  
  * more consistent pattern of RNG use in ibtest_ctd_w.pl and ibtest_ctd.pl which should help in better failure reproducibility
    when random seed replay is used:
    roll(N) < numeric-option-value and|or boolean-option-value
  * smaller improvements and bug fixes: 
  ** sa.scn: improvements in failure to known bug reconciliation
  ** ibtest_ctd.pl: 'mysqld startup failed message was seen in obtest_ctd.log even for a successful startup'
  * rein.pl: better output, update to changed interface to ibtest_ctd.pl and ibtest_ctd_w.pl
  * simpler interface to 'n rein', the script that recovers random seed
    and other similar information from n test incident directory
  * n.pm: more robust getlogdir(), allow values like '2' standing for incident_ndb.2
  
  rb://5462 approved by Krunal Bauskar
------------------------------------------------------------
revno: 8413
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk-bug18497612
timestamp: Wed 2014-07-09 17:30:15 +0530
message:
  Bug#18497612 - INCORRECT VIEW BEHAVIOR WHEN SEVERAL CHARACTER SETS
                 ARE USED
  
  Description:
  
  There are two issues as follows
  a)main.ddl_i18n_utf8.test and main.ddl_i18n_koi8r.test are skipped
    when tried with mtr.
  b)As test cases were skipped, 5.7 has got regression with collations.
  
  Analysis:
  
  a) Due to change in 'show collation' command behavior, new records
     returned by 'SHOW COLLATIONS' started to mismatch what's stored
     within r/have_xxx.require. This caused the tests to skip.
  
  b) Regression was developed by the fix of the Bug#13520710.
     Fix made sure that while writing view statement, client character
     set is used. Regression arised in the case the view statement has
     different character set than client character set.
  
  FIX :
  
  a) Changed r/have_xxx.require files with the latest result of
     'SHOW COLLATIONS' statement. Test cases will not be skipped now.
     As the test case was disabled for a long time, result file got
     outdated. Updated result file.
  b) Fixed code to use the character set if provided in the statement.
------------------------------------------------------------
revno: 8412 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 12:59:05 +0200
message:
  Merge 5.6 => trunk Bug#19172145 - Remove perl(GD) and dtrace dependencies and bench fix
    ------------------------------------------------------------
    revno: 3902.338.27 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-09 12:49:55 +0200
    message:
      Merge 5.5 => 5.6 Bug#19172145 - Remove perl(GD) and dtrace dependencies and bench fix
        ------------------------------------------------------------
        revno: 2875.598.16 [merge]
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-07-09 12:39:19 +0200
        message:
          Bug#19172145 - Remove perl(GD) and dtrace dependencies and bench fix
            ------------------------------------------------------------
            revno: 2875.618.4
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.39-release
            timestamp: Tue 2014-07-08 13:55:42 +0200
            message:
              Bug#19155121 - Remove perl(GD) and dtrace dependencies and bench fix
            ------------------------------------------------------------
            revno: 2875.618.3
            committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
            branch nick: mysql-5.5.39-release
            timestamp: Tue 2014-07-08 11:13:37 +0200
            message:
              Applying patch for bug 18779944
            ------------------------------------------------------------
            revno: 2875.618.2
            committer: Bjorn Munch <bjorn.munch@oracle.com>
            branch nick: rel-5539
            timestamp: Wed 2014-07-02 09:38:43 +0200
            message:
              Unconditionally disable dtrace for rpm, barfs on Oracle dtrace
            ------------------------------------------------------------
            revno: 2875.618.1
            committer: Bjorn Munch <bjorn.munch@oracle.com>
            branch nick: rel-5539
            timestamp: Tue 2014-07-01 15:19:30 +0200
            message:
              Unconditionally disable dtrace for rpm-oel, barfs on Oracle dtrace
------------------------------------------------------------
revno: 8411
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 15:10:35 +0530
message:
  Bug#19140536 - MAIN.IGNORE_STRICT FAILS SPORADICALLY WITH
                  RESULT MISMATCH
  
  DESCRIPTION:
  
  main.ignore_strict.test fails sporadically with result content
  mistach. One extra warning comes in case of failure with
  multi table delete statement.
  
  FIX:
  
  Test fails due to non deterministic nature of multi table delete
  statement. The test case is to check the effect of IGNORE on error
  code ER_SUBQUERY_NO_1_ROW. Changed multi table delete to single table
  delete statement to get deterministic behavior in test case.
------------------------------------------------------------
revno: 8410
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 11:34:34 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Post-push fix: CMake doesn't seem to like multiple definitions in ADD_DEFINITIONS
  if they are quoted.
------------------------------------------------------------
revno: 8409
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 14:35:41 +0530
message:
  Bug #18756233 ASSERT !HAS_DONE_RESERVATION, FSEG_ALLOC_FREE_PAGE_LOW(),
  BTR_STORE_BIG_REC_EXTE
  
  Post push fix.  The test case is failing on weekly-trunk.  This patch
  attempts to fix it.
  
  approved by Jimmy over IM.
------------------------------------------------------------
revno: 8408 [merge]
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: main-tr
timestamp: Wed 2014-07-09 10:19:50 +0200
message:
  Null merge of 19149091, not relevant for 5.6+
    ------------------------------------------------------------
    revno: 3902.338.26 [merge]
    committer: Bjorn Munch <bjorn.munch@oracle.com>
    branch nick: main-56
    timestamp: Wed 2014-07-09 10:15:30 +0200
    message:
      Null merge of 19149091, not relevant for 5.6
        ------------------------------------------------------------
        revno: 2875.598.15
        committer: Bjorn Munch <bjorn.munch@oracle.com>
        branch nick: main-55
        timestamp: Wed 2014-07-09 10:11:38 +0200
        message:
          Bug #19149091 5.5 BUILD BREAKS ON LINUX IF SUN DTRACE IS INSTALLED
          
            Add some code adapted from 5.6 to check for "real" DTrace. If found,
            and system is Linux, we simply set DTRACE to OFF. Otherwise no change.
          
            Build will still break if one tries to manually set DTRACE to ON.
------------------------------------------------------------
revno: 8407 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 15:47:14 +0800
message:
  Re-write the test case for bug#18942294 to make it stable on PB2.
    ------------------------------------------------------------
    revno: 3902.338.25 [merge]
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-09 15:43:11 +0800
    message:
      Re-write the test case for bug#18942294 to make it stable on PB2.
        ------------------------------------------------------------
        revno: 2875.598.14
        committer: bin.x.su@oracle.com
        branch nick: mysql-5.5
        timestamp: Wed 2014-07-09 15:41:13 +0800
        message:
          Re-write the test case for bug#18942294 to make it stable on PB2.
------------------------------------------------------------
revno: 8406
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk5
timestamp: Wed 2014-07-09 15:08:08 +0800
message:
  Fix Bug #19126466 - ADD FIL_PAGE_RTREE PAGE TYPE
  
  rb://5916 approved by Sunny Bains
------------------------------------------------------------
revno: 8405
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 15:33:20 +0900
message:
  Revert the following change, because the regression was innocent about the Bug#11755438:Bug#47213 fix now.
  
  ------------------------------------------------------------
  revno: 8279 [merge]
  committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
  branch nick: mysql-trunk
  timestamp: Wed 2014-06-25 05:40:00 +0200
  message:
    Disable memory barrier only for Intel CPU, because performance regression was observed at some conditions for Intel CPU.
    follow up for Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
------------------------------------------------------------
revno: 8404 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 11:40:09 +0530
message:
  WL#7894 -  Improve dtrace support on Oracle Linux 6
  
  Null merge from 5.6
    ------------------------------------------------------------
    revno: 3902.338.24
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-09 11:36:53 +0530
    message:
      WL#7894 -  Improve dtrace support on Oracle Linux 6
       
      Follow up patch to remove shebang line from dtrace.d
      file to avoid dependency on /usr/sbin/dtrace.
      Also changed dtrace.d file to 644 as script file
      as executed using dtrace command now in dynamic_tracing.test.
------------------------------------------------------------
revno: 8403
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 11:29:04 +0530
message:
  WL#7894 -  Improve dtrace support on Oracle Linux 6
  
  Merging from 5.6 to trunk.
------------------------------------------------------------
revno: 8402
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-09 13:18:34 +0900
message:
  move i_innodb.innodb_buffer_pool_resize* tests to public
------------------------------------------------------------
revno: 8401
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Wed 2014-07-09 05:19:51 +0200
message:
  WL#6711: This file is for big test. If --valgrind is used for regression
  test, the test will last a long time which will result in Timeout for
  PB2 test. So I skip such a file test if in --valgrind mode.
------------------------------------------------------------
revno: 8400
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 22:00:56 +0200
message:
  Innodb test case times out in Valgrind runs - disable it in Valgrind
------------------------------------------------------------
revno: 8399
committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
branch nick: 57-999
timestamp: Tue 2014-07-08 16:09:29 +0100
message:
  update (c) banner
------------------------------------------------------------
revno: 8398
committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
branch nick: 57-999
timestamp: Tue 2014-07-08 15:57:49 +0100
message:
  Bug#16953758: PREPARED STATEMENT IS WRITTEN TO GENERAL QUERY LOG AFTER ITS EXECUTION IS FINISH
  
  When general-logging, the "Execute" line for prepared
  statements was always written after that execution.
  
  The Execute line will be written before execution, same
  as for other statements, for consistency, and so if
  something crashes, we'll be able to more easily tell
  what it was.
  
  audit_plugin.test: Since general-logging (and therefore
  the call to audit plugins) happens before execution now,
  the SHOW counter will be one higher now (it "sees itself",
  so to say). Expected behaviour, test adjusted.
------------------------------------------------------------
revno: 8397
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B18184793-trunk
timestamp: Fri 2014-07-04 17:00:09 +0300
message:
  Bug #18184793: Inexact memmove() call in my_net_read()
  
  Removed the reference to an always 0 variable from 
  a memmove() call and replaced with a DBUG_ASSERT.
  Addressed the review comments.
------------------------------------------------------------
revno: 8396
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Tue 2014-07-08 13:26:56 +0200
message:
  Bug#19157369 GET RID OF DYNAMIC_ARRAY IN RELAY_LOG_INFO::WORKERS
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
              
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8395
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 16:55:09 +0800
message:
  Fixed test case innodb_gis-rt_precise sporadically fail on Solaris.
  
  Approved by Jimmy on IM.
------------------------------------------------------------
revno: 8394
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 11:37:46 +0300
message:
  Skip a test on non-debug server.
------------------------------------------------------------
revno: 8393
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 11:37:28 +0300
message:
  Remove an extra declaration.
------------------------------------------------------------
revno: 8392 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 15:19:35 +0800
message:
  BUG#18942294 - SEGV IN DICT_FIND_TABLE_BY_SPACE TRYING TO MARK SPACE CORRUPT
  IN RECOVERY
  
  During redo log processing, the data dictionary is not available. We should
  check it in dict_find_table_by_space() to prevent SEGV error.
  
  rb#5678, approved by Jimmy.
    ------------------------------------------------------------
    revno: 3902.338.23 [merge]
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-08 15:18:03 +0800
    message:
      BUG#18942294 - SEGV IN DICT_FIND_TABLE_BY_SPACE TRYING TO MARK SPACE CORRUPT
      IN RECOVERY
      
      During redo log processing, the data dictionary is not available. We should
      check it in dict_find_table_by_space() to prevent SEGV error.
      
      rb#5678, approved by Jimmy.
        ------------------------------------------------------------
        revno: 2875.598.13
        committer: bin.x.su@oracle.com
        branch nick: mysql-5.5
        timestamp: Tue 2014-07-08 15:10:15 +0800
        message:
          BUG#18942294 - SEGV IN DICT_FIND_TABLE_BY_SPACE TRYING TO MARK SPACE CORRUPT
          IN RECOVERY
          
          During redo log processing, the data dictionary is not available. We should
          check it in dict_find_table_by_space() to prevent SEGV error.
          
          rb#5678, approved by Jimmy.
------------------------------------------------------------
revno: 8391
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 09:43:00 +0300
message:
  Move the declaration of fil_space_t from fil0fil.cc to fil0fil.h.
  
  Approved by Yasufumi Kinoshita on IM.
------------------------------------------------------------
revno: 8390
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 07:09:02 +0200
message:
  Make
  
   innodb.blob_page_reserve
   innodb.log_file_size_checkpoint
   sys_vars.log_slow_admin_statements_func
  
  no_valgrind_without_big since they take more than 5 mins to run.
------------------------------------------------------------
revno: 8389
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 13:50:12 +0900
message:
  mark i_innodb.innodb_buffer_pool_resize_with_chunks and i_innodb.innodb_buffer_pool_resize as big-tests
------------------------------------------------------------
revno: 8388
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-08 11:16:39 +0800
message:
  Bug#18410442 CREATE A BASE CLASS FOR GTID_TABLE_ACCESS_CONTEXT AND RPL_INFO_TABLE_ACCESS
  
  Create a base class for Gtid_table_access_context and Rpl_info_table_access clases. ~
------------------------------------------------------------
revno: 8387 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-07 23:47:55 +0530
message:
  Bug #18770469 OUT OF MEMORY ON 5.6 SLAVE USING RBR WITH DATETIME WHEN TABLE
  CREATED IN 5.5.
  
  Null merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.22
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-07 23:46:29 +0530
    message:
      Bug#18770469 OUT OF MEMORY ON 5.6 SLAVE USING RBR WITH DATETIME WHEN TABLE
      CREATED IN 5.5.
      
      My earlier patch by mistake changed mysqlaccess.conf file permissions.
      Reverting it back.
------------------------------------------------------------
revno: 8386 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-07 23:38:15 +0530
message:
  Bug#18770469 OUT OF MEMORY ON 5.6 SLAVE USING RBR WITH DATETIME WHEN TABLE
  CREATED IN 5.5.
  Merge fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.21
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-07 23:25:21 +0530
    message:
      Bug #18770469 OUT OF MEMORY ON 5.6 SLAVE USING RBR WITH DATETIME WHEN TABLE
      CREATED IN 5.5.
      
      Problem: In RBR, running a long transaction, which has big number or rows
      log events, causes Out of Memory(OOM) error on slave when Slave's table
      structure is not compatible with Master's table structure.
      
      Even though the reported issue is between 5.5->5.6 replication
      and with DATATIME column, the problem can be seen in 5.6->5.6 replication
      and with any data type column when Slave's table structure is not
      compatible with Master's table structure i.e., when Slave server finds
      both Master and Slave table's are not compatible with each other and
      decides to create a virtual temporary table to do the row conversion
      at replication layer. This situation happens when some one alters the
      table on Slave by connecting directly to Slave server.
      
      Analysis:
      
      In the above explained scenario, when slave is creating a virtual
      temporary table, the logic is using thd's mem_root to create field
      and table structure. This temporary table is required only till
      the end of Rows_log_event scope but Slave's SQL thread mem_root is
      getting freed only at the end of the transaction. There are some
      exceptions that mem_root is getting freed in some other places in
      the code. Eg: After every Query_log_event::do_apply_event(),
      free_root is called on thread's mem_root. 
      
      Sample events:
        master-bin.000001 536  Query 1 608 BEGIN
        master-bin.000001 608 Table_map 1 696 table_id: 81 (test.t1)
        master-bin.000001 696 Write_rows  1 900 table_id: 81 flags: STMT_END_F
        master-bin.000001 900 Table_map 1 988 table_id: 81 (test.t1)
        master-bin.000001 988 Write_rows  1 1192  table_id: 81 flags: STMT_END_F
        ....
        master-bin.000001 29808 Xid 1 29839 COMMIT /* xid=167 */
      
      Lets say table is 't1' and it's structure is different from Master
      to Slave. And for every Rows_log_event, it creates temporary table
      and uses some 'X' amount of memory from thd's mem_root.
      If a transaction involves 'Y' number of Rows_log_events, then by
      the time of the end of the transaction, server would have used
      'XY' amount of memory from thd's mem_root. This can cause OOM
      when 'Y' is increasing. The root cause of the problem is server
      is holding the memory of temporary table even though it's scope
      is ended after applying the Rows_log_event, hence mem_root usage
      is growing up.
      
      Unlike Query_log_event::do_apply_event(), Rows_log_event::do_apply_event
      can be called from a regular thread (not only SQL thread), hence
      free_root(thd->mem_root,...) cannot be called at the end of
      Rows_log_event::do_apply_event.
      
      Fix: m_event_mem_root ( a special mem root) is added to Log_event
      class and it will be initialized in constructor and freed in
      destructor. If there is a memory requirement while applying
      any type of events whose scope should be till applying the
      event, it can be allocated from this special mem root.
      In the above situation, while creating the temporary table
      the memory needed is allocated from this special mem root
      which will be freed in ~Log_event().
------------------------------------------------------------
revno: 8385 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Mon 2014-07-07 14:27:13 +0200
message:
  merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.338.20 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Mon 2014-07-07 14:23:49 +0200
    message:
      merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.598.12
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5-merge
        timestamp: Mon 2014-07-07 12:05:30 +0200
        message:
          Bug#18935421 RPAD DIES WITH CERTAIN PADSTR INTPUTS....
          
          For rpad() and lpad(): verify that the padding string is well-formed.
------------------------------------------------------------
revno: 8384 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-07 17:02:33 +0530
message:
  Bug #17657223 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	- Null merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.19
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-07-07 16:59:04 +0530
    message:
      Bug #17657223 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
      	- Reverting the patch due to high risk involved in GA versions.
------------------------------------------------------------
revno: 8383
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk_work
timestamp: Mon 2014-07-07 20:14:44 +0900
message:
  WL#6117 InnoDB: Resize the InnoDB Buffer Pool Online
  
  - 'innodb_buffer_pool_size' is changed to 'Dynamic Variable' (The Default value is still 128M. not changed.)
  
  - able to monitor progress of resizing buffer pool (by global status 'innodb_buffer_pool_resize_status' or server error log)
  
  - 'innodb_buffer_pool_chunk_size' is a new global read-only variable, default value is 128M.
  
  The 'innodb_buffer_pool_chunk_size' unit is 1M. And the actual buffer pool size is always aligned to 'innodb_buffer_pool_chunk_size' * 'innodb_buffer_pool_instances' granularity.
  
  (rb#4225 : Approved by Vasil and Sunny)
  
  
  This push also contains following bug fixes:
  
  Bug#19001299 : SEGV WHILE INITIALIZING BUFFER POOL
  Bug#19069698 : ASSERT BTR_SEARCH_ENABLED IN PURGE AFTER SET GLOBAL INNODB_ADAPTIVE_HASH_INDEX=0
------------------------------------------------------------
revno: 8382
committer: Sayantan Dutta <sayantan.dutta@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-07 15:13:58 +0530
message:
  Bug #19056454 - RPL_SEMI_SYNC FAILS DUE TO ERROR FOUND IN LOG, DESPITE BEING SUPPRESSED
        replaced () with .. to match them in regexp
------------------------------------------------------------
revno: 8381
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Mon 2014-07-07 17:25:19 +0800
message:
  BUG#19149400 - ASSERT MTR->IS_ACTIVE() MYSQL_INPLACE_ALTER_TABLE() 
  
  It's a regression of Fix Bug#18674219 - IMPLEMENT ALTER TABLE...
  ALGORITHM=INPLACE FOR INNODB GIS INDEX CREATION.
  
  rb://5940 approved by Jimmy.
------------------------------------------------------------
revno: 8380
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Mon 2014-07-07 10:36:11 +0200
message:
  WL#6711: Made the result file of internal_tmp_disk_storage_engine.test stable.
------------------------------------------------------------
revno: 8379
committer: Sayantan Dutta <sayantan.dutta@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-07-07 12:02:18 +0530
message:
  Bug #18274766 - RESULT FILE LOST FOR FAILING TEST
------------------------------------------------------------
revno: 8378
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: test-trunk
timestamp: Sun 2014-07-06 02:31:54 +0200
message:
  Revert of REBUILD SUBPARTITIONS from bug#14028340.
  
  Subpartitions was actually not rebuilt, only partitions
  was rebuilt, due to underlying functions in alter partitions.
  
  So this disables the non-functional REBUILD of named
  subpartitions.
  
  post-push fix. Also fixes bug#19075411.
------------------------------------------------------------
revno: 8377
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Sat 2014-07-05 04:31:06 +0200
message:
  WL#6711: Support InnoDB as additional storage engine for tmp table
------------------------------------------------------------
revno: 8376
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 16:02:59 +0200
message:
  Fix main.file_contents so that it works on el7. Based on a suggestion by Terje R.
------------------------------------------------------------
revno: 8375
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-deadcode
timestamp: Fri 2014-07-04 14:10:59 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Patch #3: This patch does the following:
  - Remove the HAVE_WVLA check since all used versions of GCC/Clang now supports -Wvla
  - Remove HAVE_GETCURRENTPROCESSORNUMBER as it is true on all supported Windows versions
  - Remove workaround for old Intellisense bug
  - Unconditionally define S_IROTH/S_IFIFO on Windows rather than doing a check with hardcoded result
  - Use SIGTERM/SIGINT explicitly on Windows rather than define SIGQUIT/SIGPIPE to them
  - Use isnan() directly on Windows, remove unneeded check 
  - Typedef ssize_t on Windows rather than doing check with hardcoded result 
  - Remove HAVE_SCHED_H and HAVE_PTHREAD_YIELD_ZERO_ARG checks, results not needed anymore
  - Remove HAVE_MMAP, always true
  - Remove HAVE_SCHED_YIELD, always true on non-Windows 
  - No point in testing size of char, defined to 1 in standard 
  - Remove inline check, explicitly define for Visual Studio in C mode instead
  - Combine separate C/C++ checks for builtin_unreachable as we now always use the same compiler for C/C++
  - Remove HAVE_ACCESS, was never defined. Update code to work like intended.
  - Set SOCKET_SIZE_TYPE, size_socket and sigset_t and mode_t using typedef rather than CMake #define 
  - Remove unused FN_NO_CASE_SENCE, ulong_to_double() and table_case_convert().
  - Remove use of non-standard pthread_yield(), use POSIX standard sched_yield() instead
  - Remove unneeded thread pool config.h.cmake, add missing WindowsCache.cmake entry
  - Rewrite InnoDB CMake futex check to avoid having to do unset/remove_definitions
------------------------------------------------------------
revno: 8374
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Fri 2014-07-04 14:07:00 +0200
message:
  Bug#19143857 GET RID OF DYNAMIC_ARRAY IN RPL_FILTER
  
  Post-push fix
  gcc 4.4.6 with -Werror -O3 said:
  prealloced_array.h:226: error: 'job_worker.st_slave_job_group::ts' may be used uninitialized in this function
  
  when we did push_back on an uninitialized object.
  Fix: use value-initialization on the job_worker object.
  
  Also fix some warnings from gcc 4.7.2 in optimized mode.
------------------------------------------------------------
revno: 8373
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 17:47:27 +0800
message:
  Fix bug#18977303 - ABORT IN LOCK_MOVE_REORGANIZE_PAGE(), GIS
      bug#18994059 ASSERT PAGE_ZIP_VALIDATE(PAGE_ZIP, PAGE_ALIGN(REC), NULL),
                   PURGE, GIS
  
  bug#18977303 is caused by rtr_split_page_move_rec_list() called by
  rtr_page_split_and_insert() is not moving lock records as it splits the page.
  
  bug#18994059 is similar to another old bug# 18772498. We just need to do the
  same change in function rtr_page_copy_rec_list_end_no_locks.
  
  rb://5833, rb://5836 approved by Jimmy.
------------------------------------------------------------
revno: 8372
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Fri 2014-07-04 10:32:10 +0200
message:
  Bug#19143857 GET RID OF DYNAMIC_ARRAY IN RPL_FILTER
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
            
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8371
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 13:50:09 +0530
message:
  Bug #18756233 ASSERT !HAS_DONE_RESERVATION, FSEG_ALLOC_FREE_PAGE_LOW(),
  BTR_STORE_BIG_REC_EXTE
  
  Problem:
  
  When blob data is stored many pages are allocated from the table space.
  But these allocations are being done without doing reservation.  All
  page allocations via btr_page_alloc_low() must be done after doing
  reservation.
  
  Solution:
  
  Reserve the required number of pages in the table space for blob storage
  before doing the actual page allocations.
  
  rb#5879 approved by Marko
------------------------------------------------------------
revno: 8370
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 15:23:37 +0800
message:
  Followup fix to wl#7277: bulk load.
  
  1. Add "--big-test" option to innodb_bulk_create_index*.test
  and add a new case innodb_bulk_create_index_small.test;
  2. Remove macro "BULK_LOAD_PFS_PRINT".
------------------------------------------------------------
revno: 8369
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug18757964_mysql-trunk
timestamp: Fri 2014-07-04 12:43:02 +0530
message:
  Bug#18757964:RPL.RPL_CIRCULAR_FOR_4_HOSTS FAILS WITH RESULT
  MISMATCH NON DETERMINISTIC RESULTS
  
  Problem:
  ========
  rpl_circular_for_4_hosts  failing sporadically in pb2
  regression runs. Test lead to non deterministic results in
  case of lower performance. If the lock of the table will be
  later than the inert in next session then the result may
  differ from the expected one.
  
  Analysis:
  ========
  This bug is clearly reproducible in windows with MTS
  and --slave-parallel-type=logical_clock.
  Test script has circular replication like 1->2->3->4->1.
  As part of test they do following two steps.
  
  1) Insert value 6 in the autoinc column on server_3 (and
  prevent it from replicating further using
  SQL_SLAVE_SKIP_COUNTER on server_4). Due to the
  auto_increment_offset setting, the autoinc value 6 is
  normally generated on server_2. When we later insert a row
  on server_2,we thus cause a duplicate key error on server_3.
  
  2) Reconfigure the topology to like this 1->2->4->1,2->3.
  On server 4 change master is executed by taking Server 3's
  Exec_master_log_pos and Master_Log_File. On Server 3
  Exec master position changes as per the order in which
  workers execute the following parallel transactions.
  
  use `test`; INSERT INTO t1(b,c) VALUES('B',2)
  use `mtr`; INSERT INTO test_suppressions (pattern) VALUES
  ( NAME_CONST('pattern',_latin1'Slave SQL.*Duplicate entry
  
  Example:
  Normal case:
  "Server 3 pos_c is ------3639"
  "Server 3 file_c is -----slave-bin.000001"
  
  slave-bin.000001 3554 Query  4 3639 COMMIT
  slave-bin.000001 3639 Query  2 3732 BEGIN
  slave-bin.000001 3732 Intvar 2 3764 INSERT_ID=6
  slave-bin.000001 3764 Query  2 3876 use `test`; INSERT
  INTO t1(b,c) VALUES('B',2)
  slave-bin.000001 3876 Query  2 3961 COMMIT
  slave-bin.000001 3961 Query  3 4047 BEGIN
  slave-bin.000001 4047 Query  3 4294 use `mtr`; INSERT INTO
  test_suppressions (pattern) VALUES ( NAME_CONST
  
  Error case:
  "Server 3 pos_c is ------4372"
  "Server 3 file_c is -----slave-bin.000001"
  
  slave-bin.000001 3554 Query  4 3639 COMMIT
  slave-bin.000001 3639 Query  2 3732 BEGIN
  slave-bin.000001 3732 Intvar 2 3764 INSERT_ID=6
  slave-bin.000001 3764 Query  2 3876 use `test`; INSERT
  INTO t1(b,c) VALUES('B',2)
  slave-bin.000001 3876 Query 2 3961  COMMIT
  slave-bin.000001 3961 Query 3 4047  BEGIN
  slave-bin.000001 4047 Query 3 4294  use `mtr`; INSERT INTO
  test_suppressions (pattern) VALUES ( NAME_CONST('pattern',
  _latin1'Slave SQL.*Duplicate entry .6. for key .PRIMARY.*
  Error_code: 1062' COLLATE 'latin1_swedish_ci'))
  slave-bin.000001 4294 Query 3 4372  COMMIT
  
  Because of this incorrect position in change master
  INSERT B gets missed in 4 & 1.
  
  Fix:
  ===
  Since the main intention is to generate duplicate key
  error on Server 3 to make test script more deterministic
  moved the mtr add suppression statement to a place bit
  ahead where we do a proper sync up of all servers.
------------------------------------------------------------
revno: 8368
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 14:49:56 +0800
message:
  Bug #19068390 	VALGRIND FAILURE IN WEEKLY TESTING OF RPL.RPL_CHECK_GTID
        
  rpl.rpl_check_gtid failed with valgrind. Since we missed to delete
  the created THD after failed to create a compression thread, which
  causes a memory leak.
        
  To fix the problem, delete the created THD after failed to create
  a compression thread.
------------------------------------------------------------
revno: 8367
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk-bug19022040
timestamp: Fri 2014-07-04 12:48:36 +0700
message:
  This is a patch for bug#19022040 (SPORADICAL FAILURE IN THE
  TC_LOG_MMAP UNIT TEST)
   
  The test tc_log_mmap started sporadically failed after patch for Bug#18913949
  has been applied.
  
  The reason for sporadicall failure was that we construct several instances of
  TC_Log_MMap_thread that ran concurrently and used the same
  instance of TCLogMMapTest. This means that they also used the same THD,
  the same Transaction_ctx and the same Transaction_ctx::m_mem_root.
  Therefore it was possible to have a call for free_root for the some THD's
  memroot while this memroot is used in another thread. It could lead to
  test failure.
------------------------------------------------------------
revno: 8366
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 06:01:03 +0300
message:
  Bug#19061440 CMAKE -DINNODB_PAGE_ATOMIC_REF_COUNT:BOOL=OFF IS BROKEN
  
  The configuration option
  cmake -DINNODB_PAGE_ATOMIC_REF_COUNT:BOOL=OFF
  
  is broken ever since WL#7682 was merged.
  
  It is also useless since WL#7655, because MySQL 5.7 can no longer
  be built on systems that do not support atomic memory access primitives.
  
  This patch removes the option. It should be a non-functional change,
  because the undocumented build-time option was ON by default.
  
  rb#5844 approved by Yasufumi Kinoshita
------------------------------------------------------------
revno: 8365 [merge]
committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-07-04 03:45:02 +0530
message:
  WL#7219: Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.18 [merge]
    committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-07-04 03:38:23 +0530
    message:
      WL#7219: Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.11
        committer: Ashish Agarwal<ashish.y.agarwal@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-07-04 03:29:34 +0530
        message:
          WL#7219: Implement audit filter
------------------------------------------------------------
revno: 8364
committer: Raghav Kapoor <raghav.kapoor@oracle.com>
branch nick: mysql-trunk-bug-17487701
timestamp: Thu 2014-07-03 18:00:14 +0530
message:
  BUG#17487701 - VIEW WITH WITH CHECK OPTION ALLOWS INVALID REPLACE
  BUG#18286777 - REPLACE SUCCEEDS TO UPDATE JOIN VIEWS
  
  BACKGROUND:
  - BUG#17487701:
  The row being replaced in a view might conflict with one or
  more rows in the base table, some of which might not be
  accessible by the view.
  The buggy behaviour was that REPLACE deleted such rows (conflicting
  rows inaccessible by the view). 
  
  - BUG#18286777:
  The DELETE statement is not allowed on views created using 
  joins from the base tables according to The Manual.
  Also, The Manual states that REPLACE is DELETE and INSERT.
  Since the DELETE statement is not allowed on join views,
  the REPLACE statement should not be allowed either. 
  
  FIX:
  Regarding fix for Bug#17487701, the row which we are trying
  to REPLACE is checked if it is present in the View or not.
  If the row which we are trying to REPLACE is found in the view,
  REPLACE is allowed to proceed.
  Otherwise, If the row is not found in the view and we are
  trying to REPLACE it, we throw an error saying that
  REPLACE cannot be executed as it requires to delete rows 
  which are not in the view.
  Regarding fix for Bug#18286777, before executing REPLACE,
  we check if a view is a JOIN view or multitable view, we
  reject the REPLACE statement.
  
  NOTE TO DOCUMENTATION:
  As noted above, the REPLACE statement might result in deletion
  of more than one row before a new row is inserted. That happens
  if the table contains multiple unique indexes and the new row
  duplicates values for different existing rows in different
  unique indexes.
  
  The REPLACE statement will throw an error if the row being
  deleted is not accessible in the view. The thing is that the
  REPLACE statement might encounter inaccessible row after the
  statement has already deleted a few accessible rows.
  Raising an error will trigger a rollback of the current
  transaction, so transactional storage engines (InnoDB, ...)
  are able to rollback intermediate changes made by REPLACE.
  Non-transactional storage engines (MyISAM, ...) however
  are not able to rollback intermediate changes, so REPLACE
  might leave the database in an inconsistent state.
  
  Also The patch is backward incompatible: in the past REPLACE
  might succeed for join view, after the patch REPLACE is always
  rejected for such views.
------------------------------------------------------------
revno: 8363 [merge]
committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-03 14:30:13 +0530
message:
  Merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.338.17 [merge]
    committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-07-03 14:17:01 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.10
        committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-07-03 14:12:02 +0530
        message:
          Bug#18469276: MOD FOR SMALL DECIMALS FAILS
                
          Problem:
          If leading zeroes of fractional part of a decimal
          number exceeds 45, mod operation on the same fails.
                
          Analysis:
          Currently there is a miscalcultion of fractional
          part for very small decimals in do_div_mod.
                
          For ex:
          For 0.000(45 times).....3
          length of the integer part becomes -5 (for a length of one,
          buffer can hold 9 digits. Since number of zeroes are 45, integer
          part becomes 5) and it is negative because of the leading
          zeroes present in the fractional part.
          Fractional part is the number of digits present after the
          point which is 46 and therefore rounded off to the nearest 9
          multiple which is 54. So the length of the resulting fractional
          part becomes 6.
                
          Because of this, the combined length of integer part and fractional
          part exceeds the max length allocated which is 9 and thereby failing.
                
          Solution:
          In case of negative integer value, it indicates there are
          leading zeroes in fractional part. As a result stop1 pointer 
          should be set not just based on frac0 but also intg0. This is
          because the detination buffer will be filled with 0's for the length
          of intg0.
------------------------------------------------------------
revno: 8362
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-wl1159
timestamp: Thu 2014-07-03 12:24:39 +0400
message:
  Follow-up for WL#1159 "Allow multiple locks in GET_LOCK()".
  
  Adjusted MDL_map::find() code to circumvent false "'pinned'
  may be used uninitialized in this function" warning emitted
  by GCC 4.4 compiler.
------------------------------------------------------------
revno: 8361
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-03 11:21:25 +0300
message:
  Fix a warning about comparison between signed and unsigned.
------------------------------------------------------------
revno: 8360
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-03 11:19:21 +0300
message:
  Bug#19145637 IMPROVE ERROR HANDLING ON CRASH RECOVERY
  
  The mysql-test coverage on crash recovery error handling is spotty.
  There is no test with a corrupted redo log.
  
  Clean up and add some tests. Clean up some error handling and diagnostics.
  Rename InnoDB redo log related tests to innodb.log_*.
  
  rb#5913 approved by Jimmy Yang
------------------------------------------------------------
revno: 8359 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-03 10:58:03 +0530
message:
  Bug #17657223 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.338.16
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-07-03 10:56:54 +0530
    message:
      Bug #17657223 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
      	- Changing the test case to avoid pb2 failure.
------------------------------------------------------------
revno: 8358 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-03 10:31:19 +0530
message:
  Merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.338.15 [merge]
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-07-03 10:18:01 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.9
        committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-07-03 10:13:29 +0530
        message:
          Bug #19140907 DUPLICATES IN UNIQUE SECONDARY INDEX BECAUSE OF FIX OF BUG#68021
          
          Problem:
          
          When a unique secondary index is scanned for duplicate checking, gap locks
          were not taken if the transaction had isolation level <= READ COMMITTED. 
          This change was done while fixing Bug #16133801 UNEXPLAINABLE INNODB UNIQUE
          INDEX LOCKS ON DELETE + INSERT WITH SAME VALUES (rb#2035). Because of this
          the duplicate check logic failed, and resulted in duplicate values in unique
          secondary index.
          
          Solution:
          
          When a unique secondary index is scanned for duplicate checking, gap locks
          must be taken irrespective of the transaction isolation level.  This is
          achieved by reverting rb#2035.
          
          rb#5910 approved by Jimmy
------------------------------------------------------------
revno: 8357 [merge]
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-07-03 10:33:21 +0800
message:
  wl#7277: InnoDB: Bulk Load for Create Index
  
  We build index from bottom up other than by insert rows repeatedly, so that
  we avoid B-tree split and merge operations.
  
  We introduce a new global configuration variable "innodb_fill_factor". It
  determines percentage of space on page to be filled with data, reserving
  the remainder on each page as free space for future growth. Its valid range
  is 10 ~ 100.
  
  rb://4237 approved by Krunal & Satya.
    ------------------------------------------------------------
    revno: 7613.1.38 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-07-03 10:00:06 +0800
    message:
      Merge from mysql-trunk r8356
    ------------------------------------------------------------
    revno: 7613.1.37 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Mon 2014-06-30 13:28:00 +0800
    message:
      Merge from mysql-trunk r8323
    ------------------------------------------------------------
    revno: 7613.1.36 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-06-27 12:27:30 +0800
    message:
      Merge from mysql-trunk r8305
    ------------------------------------------------------------
    revno: 7613.1.35 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-06-24 11:13:44 +0800
    message:
      Merge from mysql-trunk r8267
    ------------------------------------------------------------
    revno: 7613.1.34
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-06-24 10:41:49 +0800
    message:
      Change test case names
    ------------------------------------------------------------
    revno: 7613.1.33
    committer: Vinay Fisrekar <vinay.fisrekar@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-06-20 15:25:36 +0530
    message:
      Adding and Formatted existing testcases for wl7277
    ------------------------------------------------------------
    revno: 7613.1.32 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-06-20 12:35:32 +0800
    message:
      Merge from mysql-trunk r8243
    ------------------------------------------------------------
    revno: 7613.1.31 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-06-18 10:58:59 +0800
    message:
      Merge from mysql-trunk r8210
    ------------------------------------------------------------
    revno: 7613.1.30 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-06-10 16:53:20 +0800
    message:
      Merge from mysql-trunk r8158
    ------------------------------------------------------------
    revno: 7613.1.29
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-06-05 09:50:51 +0800
    message:
      Address Satya's review comments
    ------------------------------------------------------------
    revno: 7613.1.28
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-06-04 11:26:30 +0800
    message:
      Address Satya and Sunny's review comments
    ------------------------------------------------------------
    revno: 7613.1.27
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-05-21 12:12:16 +0800
    message:
      Address Krunal's review comments
    ------------------------------------------------------------
    revno: 7613.1.26
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-05-08 14:04:19 +0800
    message:
      Address Krunal's review comment about test case
    ------------------------------------------------------------
    revno: 7613.1.25
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-05-07 20:28:45 +0800
    message:
      Address Krunal's review comments
    ------------------------------------------------------------
    revno: 7613.1.24
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Mon 2014-04-28 21:34:17 +0800
    message:
      Fix fts assert failure found by ima
    ------------------------------------------------------------
    revno: 7613.1.23
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-04-24 18:57:05 +0800
    message:
      Merge fix of bug#18544317 from mysql-trunk
    ------------------------------------------------------------
    revno: 7613.1.22
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-04-22 22:18:52 +0800
    message:
      Fix btr validate assert failure found by ima
    ------------------------------------------------------------
    revno: 7613.1.21
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-04-18 18:01:25 +0800
    message:
      Fix page split assert failure found by ima
    ------------------------------------------------------------
    revno: 7613.1.20
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-04-17 15:53:31 +0800
    message:
      Pass is_redo_skipped to fts index build
    ------------------------------------------------------------
    revno: 7613.1.19
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-04-15 11:22:20 +0800
    message:
      Fix build error in pb2
    ------------------------------------------------------------
    revno: 7613.1.18
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Mon 2014-04-14 19:16:34 +0800
    message:
      Support fulltext index bulk load
    ------------------------------------------------------------
    revno: 7613.1.17
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Mon 2014-04-14 17:44:24 +0800
    message:
      Fix fsp_space_modify_check assert failure found by ima
    ------------------------------------------------------------
    revno: 7613.1.16 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-04-11 20:49:30 +0800
    message:
      Merge from mysql-trunk r7853
    ------------------------------------------------------------
    revno: 7613.1.15
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-04-11 10:42:20 +0800
    message:
      Fix issue: deep recursion in bulk load page commit found by ima
    ------------------------------------------------------------
    revno: 7613.1.14
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-04-10 16:25:55 +0800
    message:
      Change innodb_bulk_load_fill_factor to innodb_fill_factor
    ------------------------------------------------------------
    revno: 7613.1.13
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-04-09 20:33:52 +0800
    message:
      1. Remove innobase_enable_bulk_load;
      2. Enable fill factor for non-leaf pages.
    ------------------------------------------------------------
    revno: 7613.1.12 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-04-09 17:32:56 +0800
    message:
      1.Merge from mysql-trunk r7837
      2.Disable page allocation log for rebuilding table with FILE_PER_TABLE.
    ------------------------------------------------------------
    revno: 7613.1.11
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-04-08 20:14:03 +0800
    message:
      Remove row threshold and flush threshold parameters
    ------------------------------------------------------------
    revno: 7613.1.10
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-04-08 15:25:46 +0800
    message:
      Disable redo logging and set checkpoint at the end of bulk load
    ------------------------------------------------------------
    revno: 7613.1.9 [merge]
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Mon 2014-03-31 16:52:46 +0800
    message:
      Merge from mysql-trunk r7792
    ------------------------------------------------------------
    revno: 7613.1.8
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-03-27 11:50:43 +0800
    message:
      Remove m_heap_no in PageBulk
    ------------------------------------------------------------
    revno: 7613.1.7
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-03-26 23:12:02 +0800
    message:
      Fix wl7277.test failure in pb2
    ------------------------------------------------------------
    revno: 7613.1.6
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-03-26 15:13:20 +0800
    message:
      1. Add new test cases.
      2. Fix crash recovery issues.
      3. Fix test failures in pb2.
    ------------------------------------------------------------
    revno: 7613.1.5
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Tue 2014-03-18 18:48:37 +0800
    message:
      Fix crash issue found by ima
    ------------------------------------------------------------
    revno: 7613.1.4
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Mon 2014-03-17 11:34:18 +0800
    message:
      Fix blob log issue
    ------------------------------------------------------------
    revno: 7613.1.3
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Fri 2014-03-14 19:46:24 +0800
    message:
      Set ibuf bitmap when finishing a page
    ------------------------------------------------------------
    revno: 7613.1.2
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Wed 2014-03-12 18:24:54 +0800
    message:
      Refactory using c++/class and run test suite innodb_zip
    ------------------------------------------------------------
    revno: 7613.1.1
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-trunk-wl7277
    timestamp: Thu 2014-03-06 13:06:30 +0800
    message:
      Test suite innodb,i_innodb,innodb_fts passed,
      Except innodb.innodb_information_schema_buffer.
------------------------------------------------------------
revno: 8356
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-wl1159
timestamp: Wed 2014-07-02 22:07:50 +0400
message:
  WL#1159 "Allow multiple locks in GET_LOCK()".
  
  This patch is based on contribution from Konstantin Osipov.
  
  The goal of this patch is to allow user to acquire multiple
  user-level locks in the same connection via serie of calls to
  GET_LOCK() function. Also it makes possible to take the same
  lock twice in connection (i.e. locks became recursive).
  GET_LOCK() no longer implicitly releases the previous lock held
  by the connection. To release all locks in connection one can
  use new RELEASE_ALL_LOCKS() function.
  
  The above is achieved by deleting all old code which implements
  user-level locks, and forwarding user-level lock requests to the
  metadata locking subsystem. A new metadata type was introduced
  for user-level locks - "user". Instances of "USER" locks are
  mutually exclusive.
  
  Possible deadlocks between multiple locks taken in different order,
  as well as between user-level locks and metadata/waits for table
  flushes are detected and resolved using the MDL deadlock detector. 
  Waits for user-level locks are preferred as victim over waits
  for locks typically acquired by DDL, but waits for locks typically
  acquired by DML are preferred over waits for user-level locks.
  
  Wait for user-level lock which is aborted due to deadlock is
  reported as ER_USER_LOCK_DEADLOCK error. No transaction rollback
  happens in this case.
  
  MDL subsystem was extended to support the old behavior when waits
  for user-level locks are automagically aborted if connection which
  requested them disconnects. However now this situation is handled
  similarly to KILL statement rather than to lock wait timeout.
  
  Since IS_USED_LOCK() function returns id of connection which owns
  the lock the MDL API had to be extended with capability to query
  lock owner.
  
  Note that with old implementation when a wait for user-level lock was
  aborted due to query or connection being killed GET_LOCK() function
  always returned NULL and statement using this function might have
  succeeded or failed with ER_QUERY_INTERRUPTED error, depending if
  statement tried to do anything else after calling GET_LOCK().
  With new implementation statement which called GET_LOCK() and was
  killed will always fail with ER_QUERY_INTERRUPTED due to slightly
  different KILL error handling by MDL subsystem.
  
  Since MDL subsystem imposes limits on the length of key used to
  identify objects new implementation introduces limit on user-level
  lock name length. New limit is 64 characters.
  ER_USER_LOCK_WRONG_NAME error is emitted when one of functions
  accepting user-level lock name as argument gets name which is
  longer than 64 character. Also behavior is changed to return the
  same error for NULL or empty ('') lock name.
  Also new implementation will always convert lock name to utf8 from
  its original charset and perform case-insensitive comparison.
  
  Existing test coverage for user-level lock functionality has been
  extended and moved from main.func_misc to main.user_lock test.
  
  New rpl.rpl_user_lock test covers replication of statements using
  user-level lock functions.
  
  Some other existing tests had to be adjusted to the new user-lock
  behavior including rpl.rpl_err_ignoredtable test. The latter was
  additionally fixed to really test what it was intended to test.
  
  Tests in perfschema test suite which used mutex used in user-level
  lock implementation to test aggregation of wait events were changed
  to use different mutex. Tests which used user-level locks to test
  aggregation for memory events we adjusted to new implementation
  memory usage. Coverage for new MDL namespace was added to
  perfschema.mdl_func test.
  
  Unit tests covering MDL API/code changes were added.
------------------------------------------------------------
revno: 8355
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 18:01:06 +0300
message:
  Fix for a compilation failure on windows.
------------------------------------------------------------
revno: 8354
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-plain
timestamp: Wed 2014-07-02 14:46:35 +0200
message:
  Bug#11752665: WINDOWS SERVER X64: SO MANY COMPILER WARNINGS
  
  Patch #10:
  
  This patch reduces the compiler warning count on Win64 by an additional
  230, as reported by VS2013.
------------------------------------------------------------
revno: 8353 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 17:47:25 +0530
message:
  Bug #17657223 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.338.14
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-02 17:45:29 +0530
    message:
      Bug #17657223 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
      
      Analysis:
      	Temporary file usage during alter table operation can be avoided
      during clustered index rebuild when new primary key follow existing
      pk order. So it will reduce the file usage during clustered index rebuild.
      Delaying the temporary file creation and log file creation for alter table
      operation will lead to avoid the file creation for smaller tables.
      
      	Approved by marko, shaohua (rb-4788)
------------------------------------------------------------
revno: 8352 [merge]
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 15:05:10 +0530
message:
  Bug#17873011 NO DEPRECATION WARNING FOR THREAD_CONCURRENCY
  
  Null Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.13 [merge]
    committer: Arun Kuruvila <arun.kuruvila@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-02 14:54:18 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.8
        committer: Arun Kuruvila <arun.kuruvila@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-07-02 14:52:52 +0530
        message:
          Bug#17873011 NO DEPRECATION WARNING FOR THREAD_CONCURRENCY
          
          Description:
          THREAD_CONCURRENCY is deprecated and there is no 
          deprecation warning message while setting this variable
          while starting the server.
          
          Analysis:
          This variable is specific to Solaris 8 and earlier systems
          and is ignored on all other platforms. But since many 
          customers, who uses other than Solaris, still has this 
          variable in their configuration file, it is important to
          have a deprecation warning.
          
          Fix:
          THREAD_CONCURRENCY deprecation warning message is added.
------------------------------------------------------------
revno: 8351
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Wed 2014-07-02 11:29:31 +0200
message:
  Bug#19134997 GET RID OF DYNAMIC_ARRAY IN RPL_GTID
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
          
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8350 [merge]
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 11:02:50 +0200
message:
  BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
  
  Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
  
  Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
  Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
    ------------------------------------------------------------
    revno: 3902.338.12 [merge]
    committer: Marcin Babij <marcin.babij@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-07-02 10:46:32 +0200
    message:
      BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
      
      Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
      
      Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
      Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
        ------------------------------------------------------------
        revno: 2875.598.7
        committer: Marcin Babij <marcin.babij@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-07-02 10:45:22 +0200
        message:
          BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
          
          Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
          
          Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
          Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
------------------------------------------------------------
revno: 8349
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 14:23:36 +0530
message:
  Bug #19141017 : PB2 MAIN.GRANT_EXPLAIN_NON_SELECT.TEST FAILING ON PB2 FOR EXPLAIN PLAN
------------------------------------------------------------
revno: 8348
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 10:26:42 +0300
message:
  WL#7990 Repurpose FIL_PAGE_FLUSH_LSN
  
  The field FIL_PAGE_FLUSH_LSN is consulted during InnoDB startup.
  
  At startup, InnoDB reads the FIL_PAGE_FLUSH_LSN from the first page of
  each file in the InnoDB system tablespace. If there are multiple
  files, the minimum and maximum LSN can differ. These numbers are
  passed to InnoDB startup.
  
  Having the number in other files than the first file of the InnoDB
  system tablespace is not providing much additional value. It is
  conflicting with other use of the field, such as on InnoDB R-tree
  index pages.
  
  The FIL_PAGE_FLUSH_LSN was also being written to InnoDB undo
  tablespace files, even though the fields are not going to be consulted
  on crash recovery.
  
  This worklog will stop writing FIL_PAGE_FLUSH_LSN to other files
  than the first file of the InnoDB system tablespace (page number 0:0).
  
  fil_write_flushed_lsn(): A new function, merged from
  fil_write_lsn_and_arch_no_to_file() and
  fil_write_flushed_lsn_to_data_files().
  Only write to the first page of the system tablespace (page 0:0),
  and invoke fil_flush_file_spaces(FIL_TYPE_TABLESPACE) afterwards.
  
  fil_open_single_table_tablespace(): Remove output of LSN, because it
  was only valid for the system tablespace and the undo tablespaces, not
  user tablespaces.
  
  Datafile::m_flushed_lsn: Remove.
  
  Datafile::validate_first_page(),
  SysTablespace::read_lsn_and_check_flags():
  Add an output parameter for flushed_lsn.
  
  SysTablespace::read_lsn_and_check_flags(): Only read the first file.
  
  SysTablespace::open_or_create(): Replace the output parameters
  min_lsn,max_lsn with flush_lsn.
  
  Tablespace::m_min_flushed_lsn: Remove.
  Tablespace::m_max_flushed_lsn: Remove.
  
  recv_recovery_from_checkpoint_start(): Replace the two parameters
  min_flushed_lsn, max_flushed_lsn with one: flush_lsn.
  
  innobase_start_or_create_for_mysql(): Remove a message that would
  detect FIL_PAGE_FILE_FLUSH_LSN mismatch between multiple files of the
  system tablespace.
  
  rb#5882 approved by Jimmy Yang
------------------------------------------------------------
revno: 8347
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-07-02 10:26:09 +0300
message:
  Remove tests that should have been removed as part of reverting a fix in
  marko.makela@oracle.com-20140701084724-6q9ks858j6fxhj32
------------------------------------------------------------
revno: 8346 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug17453826_mysql-trunk
timestamp: Wed 2014-07-02 11:43:45 +0530
message:
  Bug#17453826:ASSERTION ERROR WHEN SETTING FUTURE BINLOG
  FILE/POS WITH SEMISYNC
  
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.11
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug17453826_mysql-5.6
    timestamp: Wed 2014-07-02 11:34:11 +0530
    message:
      Bug#17453826:ASSERTION ERROR WHEN SETTING FUTURE BINLOG
      FILE/POS WITH SEMISYNC
      
      Problem:
      ========
      When DMLs are in progress on the master stopping a slave and
      setting ahead binlog name/pos will cause an assert on the
      master.
      
      Analysis:
      ========
      Trx1 is waiting for ack on certain position meanwhile change
      master command is issued from the slave to master with a
      future log file name and position. Upon receipt master's
      dump thread assumes that it has received ack till the future
      log file and position and clears the active transaction list
      and sets the reply_file_name_ and reply_file_pos_ to future
      file name and pos. since reply file name and position are
      ahead of Trx1's current position Trx1 completes
      successfully. Execute a new DML Trx2. During commit Trx2
      waits for an ack on some position but the reply file name
      and position are already pointing to future file name and
      position so this current transaction thinks that it has
      already got an ack and no need to wait and proceeds further.
      But the active transaction has not actually received any ack
      the transaction node is not cleared from active transaction
      list. During the exit of commit operation there is an assert
      which expects active transaction node to be cleared from the
      transaction list which fails this leads to an assert.
      
      Fix:
      ===
      Ideally when a request for change master is received the
      request file name and position should be validated for
      correctness and only after the validation the transmit_start
      hook should be fired.  Hence moved the transmit_start hook
      position to appropriate place where it is placed post the
      validation of file name and position. Post fix when a future
      log file name and position is specified
      ER_MASTER_FATAL_ERROR_READING_BINLOG error will be returned
      to slave.
      
      Note: During fixing this bug one more bug was identified.
      
      Rpl_semi_sync_master_clients this parameter can become
      negative at certain conditions causing UNINSTALL PLUGIN
      rpl_semi_sync_master to fail.
      
      'Rpl_semi_sync_master_clients' client is incremented when
      dump thread initiates transmission and when a new slave
      connection is established. This value is decremented when
      dump thread exits and slave gets disconnected. On a certain
      condition if IO thread stops due to an error and dump thread
      exited before "transmit_start" hook dump thread will still
      try to call "transmit_stop" and this will cause
      Rpl_semi_sync_master_clients= -1 which is invalid state.
      Hence as part of the fix a new flag is introduced such that
      "transmit_stop" will be called only when "transmit_start" is
      called.
------------------------------------------------------------
revno: 8345
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Wed 2014-07-02 12:40:57 +0800
message:
  Bug #18900932 ERROR MESSAGE ISN'T SENT TO CLIENT WHEN SERVER REPORTS ONE
  
  When invalid geometry data of WKT/WKB/GEOMETRY format is detected,
  do my_error to report the error to client, rather than only setting
  null_value to true. We used to do only the latter action, so the geometry
  functions returned NULL as result if given invalid geometry data, and
  the fact that the geometry data is invalid was silently ignored, which
  wasn't good. And now we always do error report in GIS function items, 
  i.e. in Item_xxx::val_xxx functions, to be in a clear context. Also we
  now only set null_value to true on error if maybe_null is true, reason as below:
  
  MySQL internally (in Item::send function) requires that if 
  Item_xxx::val_str returns NULL, its 'null_value' member
  must be set to true; It also requires that only Items with maybe_null=1 can 
  set null_value=true. So we can only set null_value to true and return NULL if
  maybe_null is true, otherwise we have to return an empty String and leave 
  null_value as false on error. for other val_xxx() functions we also have to
  return default values on error and conditionally set null_value to true.
  
  We now define common Item::error_xxx functions which does proper Item status
  setting and produce on-error return values according to above rules. 
  They should be called to get out of the Item_xxx::val_xxx function after 
  an error condition is detected and reported with my_error. 
  Now all GIS functions call them uniformly.
  
  Assign SQLSTATE error number for ER_GIS_INVALID_DATA error: 22023.
------------------------------------------------------------
revno: 8344
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: trunk
timestamp: Tue 2014-07-01 22:35:35 +0200
message:
  Fix for Bug#19133176 PB2 FILTER_SINGLE_COL_IDX_BIG UNSTABLE
                       - RESULT MISMATCH
  
  Wrong rows estimate in explain. Added ANALYZE TABLE to the
  three tables that are used in the query in order to get
  more stable statistics from the storage engine.
------------------------------------------------------------
revno: 8343 [merge]
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-01 17:02:41 +0100
message:
  BUG#14135002
  
  Empty merge.
    ------------------------------------------------------------
    revno: 3902.338.10 [merge]
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-5.6-push
    timestamp: Tue 2014-07-01 17:00:07 +0100
    message:
      BUG#14135002
      
      Automerged into latest mysql-5.6.
------------------------------------------------------------
revno: 8342
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-deadcode
timestamp: Tue 2014-07-01 15:28:42 +0200
message:
  Bug#18404381: REMOVE UNNEEDE CMAKE CHECKS AND #IFDEFS IN 5.7.5
  
  Patch #2: This patch does the following:
  - Enforces Visual Studio 2013 as the minimum version of Visual Studio,
    in accordance with our supported platform list for 5.7
  
  - Removes the unused CYBOZU and BACKUP_TEST CMake options
  - Removes the following CMake checks which were always false for
    our supported platforms: HAVE_TERMBITS_H, HAVE_NDIR_H, HAVE_SELECT_H,
    HAVE_SYS_NDIR_H, HAVE_SYSENT_H, HAVE_RDTSCLL
  - Removes the following CMake checks which were always true for
    our supported platforms: HAVE_SYS_TYPES_H, HAVE_FCNTL_H, HAVE_FENV_H,
    HAVE_INTTYPES_H, HAVE_STDINT_H, HAVE_SYS_STAT_H, HAVE_RINT,
    HAVE_FESETROUND
  - Removes the following CMake checks where the results are not needed:
    HAVE_SYS_DIR_H, HAVE_DLERROR, HAVE_ASM_MSR_H, HAVE_DECL_MADVISE
  
  - Moves the defintion of NOGDI from my_global.h to the Windows CMake
    file so its defintion no longer depends on including my_global.h
  - Moves the defintion of __EXTENSIONS, _POSIX_THREAD_SEMANTICS and
    _REENTRANT from my_global.h to the Solaris CMake file so their
    defintion no longer depends on including my_global.h
  - Moves the HAVE_SOLARIS_STYLE_GETHOST CMake check to the Solaris
    CMake file so that it is not run on all platforms
  
  - Removes use of #_lint
  - Removes use of TARGET_OS_LINUX, use __linux__ instead
  - Removes use of HAVE_LONG_LONG, always true
  - Removes various __GNUC__ version checks which are no longer relevant
  - Changes remaining use of DBUG_ON to !DBUG_OFF
  
  - Removes setting of _GNU_SOURCE from my_global.h, already set by CMake
  - Removes setting of _THREAD_SAFE from my_global.h, not needed
  - Removes declaration of madvise from my_global.h, not used
  - Removes duplicate declaration of closesocket from my_global.h
  - Removes STACK_DIRECTION check from my_global.h, already done by CMake
  - Removes various defintions from my_global, use limits.h instead
  
  - Removes many unused declarations from my_sys.h and my_static.h
  - Removes various definitions of __attribute__, use my_compiler.h instead
  
  - Removes the following files with only dead code: include/t_ctype.h,
    mysys/my_getncpus.c
  - Removes dead code from mysys/my_getsystime.c
  - Removes mysys/mf_sort.c and moves used code to bounded_queue unit test
  
  - Removes st_my_thread_var::cmp_length, not used
  - Removes st_my_thread_var::pthread_self, use pthread_self() instead
------------------------------------------------------------
revno: 8341 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-01 17:22:54 +0530
message:
    Bug#18258933 CONCATENATION OF MYSQLBINLOG OUTPUT DOES NOT
                 WORK WITH GTID
  
    merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.338.9
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-01 17:18:58 +0530
    message:
        Bug#18258933 CONCATENATION OF MYSQLBINLOG OUTPUT DOES NOT
                     WORK WITH GTID
      
        Problem:  When mysqlbinlog processes and concatenates
        multiple binary log files into one output file, this
        output file is not in useful state. When it is later
        used for point-in-time recovery, it is producing
        ER_GTID_NEXT_TYPE_UNDEFINED_GROUP "ERROR 1837: When
        @@SESSION.GTID_NEXT is set to a GTID, you must explicitly
        set it to a different value after a COMMIT or ROLLBACK.
        Please check GTID_NEXT variable manual page for detailed
        explanation. Current @@SESSION.GTID_NEXT is 'xyz'".
      
        Analysis: When mysqlbinlog processes a GTID enabled
        binary log, it outputs SET GTID_NEXT statements. As
        GTID_NEXT is set to 'undefined' when a transaction
        commits (or DDL commits implicitly), this will leave
        GTID_NEXT as 'undefined' when the server has processed
        all output from mysqlbinlog. Also, there can be some
        anonymous statements/transactions at the beginning of
        binary log (i.e., without SET GTID_NEXT statements).   
        When mysqlbinlog processes and concatenates two (or more)
        binary log files into one output file, the combination
        (GTID_NEXT is undefined at the end of first binary log
        and second binary log can contain anonymous transactions)
        will not work and will result in ERROR 1837
        (ER_GTID_NEXT_TYPE_UNDEFINED_GROUP).
       
        Fix: When mysqlbinlog processes second binary log's
             start_log_event, it can output
             "SET GTID_NEXT=AUTOMATIC" if required i.e., if
             previous binary log is leaving the GTID_NEXT state
             in 'undefined' state. In the old code (before this
             patch), when mysqlbinlog processes multiple binary
             log files from multiple command line arguments, it
             used a new PRINT_EVENT_INFO object for each file,
             so the state at the end of the previous file is lost.
             To solve this problem, this patch changes the logic
             to use the same PRINT_EVENT_INFO structure for all
             command line arguments. And to keep track of
             gtid_next state, we are introducing
             is_gtid_next_valid and is_gtid_next_set variables.
             At the beginning of mysqlbinlog process
             is_gtid_next_set is set to false and
             is_gtid_next_valid is set to true.
             When mysqlbinlog encounters a
               i) GTID_LOG_EVENT: it sets is_gtid_next_set to
                                  true and is_gtid_next_valid to
                                  true.
              ii) XID_LOG_EVENT
                   /end_group() : it sets is_gtid_next_valid to
                                  false if is_gtid_next_set to
                                  true.
             And when mysqlbinlog is processing start_log_event,
             if it finds gtid_next is not valid (i.e., undefined)
             then it produces "SET GTID_NEXT=AUTOMATIC" in the
             output file to avoid mentioned
             ER_GTID_NEXT_TYPE_UNDEFINED_GROUP error and it resets
             is_gtid_next_set to false and is_gtid_next_valid to
             true (the initial values). mysqlbinlog also produces
             'SET GTID_NEXT=AUTOMATIC' in the output file at the
             end of processing all the command line arguments if
             it finds GTID_NEXT is in invalid state.
      
             After this patch, if you are merging two or more
             binary log files, you will see two differences in
             the output when compared to the before the patch
      
             a) When mysqlbinlog start processing the second
                binary log file, if required (i.e., if the 
                previous binary log left gtid_next in a invalid
                state), it generates SET GTID_NEXT=AUTOMATIC.
             b) Before the patch, mysqlbinlog outputs session
                variable information for every binary log.
                Now that we will have only one print_event_info
                per run, this session variables value will be
                printed only once *unless there is a change*.
                Example of session variables:
                  SET @@session.pseudo_thread_id=999999999/*!*/;
                  SET @@session.foreign_key_checks=1
                  SET @@session.sql_mode=1073741824/*!*/;
                  SET @@session.auto_increment_increment=1,
                  SET @@session.auto_increment_offset=1/*!*/;
                  SET @@session.lc_time_names=0/*!*/;
                  SET @@session.collation_database=DEFAULT/*!*/;
------------------------------------------------------------
revno: 8340 [merge]
committer: Aditya A <aditya.a@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-01 16:59:50 +0530
message:
  Bug #18305270  BACKPORT BUG#18694052 FIX 
                 FOR ASSERTION `!M_ORDERED_REC_BUFFER' 
                 FAILED TO 5.6 
  
  [Null Merge fron 5.6]
    ------------------------------------------------------------
    revno: 3902.338.8
    committer: Aditya A <aditya.a@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-01 16:52:35 +0530
    message:
      Bug #18305270  BACKPORT BUG#18694052 FIX 
                     FOR ASSERTION `!M_ORDERED_REC_BUFFER' 
                     FAILED TO 5.6 
      
      PROBLEM
      -------
      Missed to remove record priority queue if 
      init_index failed for a partition which
      was causing the crash.
      
      FIX
      ---
      Remove priority queue if init_index fails
      for partition.
------------------------------------------------------------
revno: 8339
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Tue 2014-07-01 13:23:05 +0200
message:
  Bug#19125864 GET RID OF DYNAMIC_ARRAY IN SQL_PLUGIN.CC
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
        
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Also: remove 'struct' keyword from function signatures, not necessary in C++
------------------------------------------------------------
revno: 8338 [merge]
committer: magnus.blaudd@oracle.com
branch nick: trunk
timestamp: Tue 2014-07-01 12:02:09 +0200
message:
  Merge
    ------------------------------------------------------------
    revno: 8329.1.1 [merge]
    committer: magnus.blaudd@oracle.com
    branch nick: merge_trunk-cluster_2_trunk
    timestamp: Tue 2014-07-01 10:46:59 +0200
    message:
      Merge trunk-cluster -> trunk
      
       - revert MCP's
        ------------------------------------------------------------
        revno: 2876.563.77
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster-stage
        timestamp: Mon 2014-06-30 12:47:03 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - different 'rows' and 'filtered' from EXPLAIN in ndb_statistics0 and ndb_statistics1
           - fix by recording the new .result for now. Don't  mask the 'rows' and 'filtered' yet as
             it feels like the statistics tests should be able to get the stats correct  - eventually.
        ------------------------------------------------------------
        revno: 2876.563.76
        committer: magnus.blaudd@oracle.com
        branch nick: merge_trunk_2_trunk-cluster
        timestamp: Mon 2014-06-30 11:21:10 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - mask out more unstable 'rows' and 'filtered' columns from EXPLAIN
           - still a few result failures where the EXPLAIN differs in other ways. Those
             need more investigation to find the cause.
        ------------------------------------------------------------
        revno: 2876.563.75
        committer: magnus.blaudd@oracle.com
        branch nick: merge_trunk_2_trunk-cluster
        timestamp: Fri 2014-06-27 10:42:17 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
                
                 - While implementing 'WL7149 Remove storage_engine system variable in 5.7' all
                   usages of the 'storage_engine' system variable was changed to use 'default_storage_engine'
                   system variable.
                 - Fix one missing replace in ndb_load.result
        ------------------------------------------------------------
        revno: 2876.563.74
        committer: magnus.blaudd@oracle.com
        branch nick: merge_trunk_2_trunk-cluster
        timestamp: Fri 2014-06-27 10:40:34 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - After 'WL6614 Define and reimplement IGNORE' a warning is always printed
            for those rows whose errors are ignored.
           - Fix by recording the new warnings into .result files.
        ------------------------------------------------------------
        revno: 2876.563.73
        committer: magnus.blaudd@oracle.com
        branch nick: merge_trunk_2_trunk-cluster
        timestamp: Wed 2014-06-25 14:37:47 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - The function thd_query_string() was renamed to thd_query_unsafe()
        ------------------------------------------------------------
        revno: 2876.563.72 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_trunk_2_trunk-cluster
        timestamp: Tue 2014-06-24 15:49:46 +0200
        message:
          Merge trunk(8263) -> trunk-cluster
          
           - pthread_mutex and pthread_cond functions has been renamed to native_mutex and native_cond
             respectively, this is expected to cause some confusion and build problems.
           - marked changes in some of the packaging/Wix/ files with MCP tags since those had been
             reformatted and the patches had to be reapplied
        ------------------------------------------------------------
        revno: 2876.563.71
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2014-06-23 10:22:25 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - remove unnecessary check for platform, the variable CMAKE_SHARED_LIBRARY_C_FLAGS
            will only be set for the appropriate platforms
        ------------------------------------------------------------
        revno: 2876.563.70
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2014-06-23 10:16:57 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - add back accidentally removed line in preivous patch for fixing LINK_FLAG
        ------------------------------------------------------------
        revno: 2876.563.69
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-19 16:09:35 +0200
        message:
          Revert MCP for bug 17602208
          
           - already fixed upstream
        ------------------------------------------------------------
        revno: 2876.563.68
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-19 14:03:24 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - Problem which occurs when building with solaris/gcc 64 bit, he object files was compiled
             as 64bit but when linking with gcc the -m64 flag was missing and thus causing "Wrong ELF class" error
           - Fix the same way as upstream fix for "bug 16555106 FIX BROKEN BUILD WITH SOLARIS/GCC 64BIT MODE"
            i.e use the variable CMAKE_SHARED_LIBRARY_C_FLAGS which contains the flags that should be used when building
            shared libraries.
        ------------------------------------------------------------
        revno: 2876.563.67
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 13:41:28 +0200
        message:
          Cherrypick fix for pfs_noop-t from trunk
        ------------------------------------------------------------
        revno: 2876.563.66
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 12:40:09 +0200
        message:
          Cherrypick WL#6815 fixes from 7.4 -> trunk-cluster
        ------------------------------------------------------------
        revno: 2876.563.65
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 12:15:12 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
          - 5.7 outputs additional columns from EXPLAIN since EXTENDED and PARTITIONS have
             been turned on by default.
             1) This causes the 'rows' column to shift from position 9 to 10
             2) Also it introduces the new column 'filtered' which is to be
                considered as unreliable as 'rows'. Quoting comment in existing
                test case "experience has shown that 'rows' estimates may vary
                depending on OS, compiler and .. phase of moon."
           - Fix problem by increasing the column number in masking of 'rows' and at the
             same time add masking of the 'filtered' column.
           - NOTE! Should have done this the first time when EXPLAIN output was recorded
             for 5.7 but the problem went unnoticed at that time.
        ------------------------------------------------------------
        revno: 2876.563.64
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 12:13:54 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
          - The jties tests intentionally checks more or less all permutations of
              const result/parameter types in order to see what is supported
              by C++ vs. Java. Tthis causes warnings about "type qualifiers ignored
              on function return type" (which is turned on by -Wextra in maintainer.cmake).
              Avoid the superfluous warnings by turning them off.
        ------------------------------------------------------------
        revno: 2876.563.63
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 12:13:01 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - "type qualifiers ignored on function return type [-Wignored-qualifiers]"
           - caused by "const" on function return types which are always considered const
           - fix by removing the superfluous const
        ------------------------------------------------------------
        revno: 2876.563.62
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 11:53:24 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
           - Fix usage of char* with the endian independence macros from mysys
           - Following patch for "Bug #12590101 *KORR/*STORE MACROS IN MY_GLOBAL.H
             ASSUME THE ARGUMENT TO BE A CHAR POINT" the code detects by warning when a char
             buffer or char* are used with the endian independence macros from mysys.
             Following the logic in that patch usages of char and char* should be converted to uchar
             and uchar* in combination with these macros/inline functions thus avoiding the warning.
        ------------------------------------------------------------
        revno: 2876.563.61
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 11:52:17 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7  
           - The -Wdeclaration-after-statement has been made default
             for build of C files in 5.7. This flag turns on the warning
             "ISO C90 forbids mixed declarations and code" when code and
             declarations are mixed in a function. Most likely this intends
             to help with keeping the C code portable to compilers which
             does not yet support such constructs  - thus we should adapt
             all our code not to cause that warning.
           - Fix by changing the code in ndb_engine.c which uses
             said construct.
        ------------------------------------------------------------
        revno: 2876.563.60
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-06-05 11:41:50 +0200
        message:
          Adapt pushed joins to WL#6016
          
           1) The call to ha_make_pushed_join() was moved to a place where
             "semijoin dups elimination" had not yet been performed and
             thus the AQP code failed to detect that such part of the
             query should not be pushed. Fix by moving the call to ha_make_pushed_join()
             even later in the optimizer(although that code is in sql_select.cc).
             This is more or less the exact same place as in 5.6 based code.
          
           2) The join type is now determined earlier and serves in the current
              design as the "primary source of the type". Thus the AQP code
              will see join types of JT_RANGE and JT_INDEX_MERGE which can be treated
              the same way as JT_ALL. In long run the two new join types
              can each have separate implementations to avoid the checks to see
              if "quick" is used or not since that is now determined by the join type.
              - JT_RANGE/INDEX_MERGE have to have "quick" set
              - JT_ALL can have "quick" set
        ------------------------------------------------------------
        revno: 2876.563.59
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2014-05-12 10:24:50 +0200
        message:
          Adapt MySQL Cluster to 5.7
          
           - update .result file for an EXPLAIN which contains FORCE INDEX FOR GROUP BY
             since the FOR GROUP BY part of the query is now (sincve bug 17889511) printed
             to the warnings.
        ------------------------------------------------------------
        revno: 2876.563.58
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Thu 2014-04-17 13:32:44 +0200
        message:
          Cherrypick fix for removal of ATTRIBUTE_FORMAT from trunk which caused
          compilation failure on non gcc compilers
        ------------------------------------------------------------
        revno: 2876.563.57
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Thu 2014-04-17 13:28:40 +0200
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
           
            - adapt .result file for rpl_ndb_row_001 to upstream change
              in sourced .test file(changes made the mysql.user table
              are restored before test ends).
        ------------------------------------------------------------
        revno: 2876.563.56
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Thu 2014-04-17 13:23:34 +0200
        message:
          Cherrypick fix for WL#7774 Make MTR sync_slave_sql_with_master.inc handle
          NDB from 7.4 -> trunk-cluster
          
           - manually adapt few .result files to the new output
        ------------------------------------------------------------
        revno: 2876.563.55 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Wed 2014-04-16 11:51:49 +0200
        message:
          Merge 7.4(-r4281) -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.577.96 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-03-24 23:44:38 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.66 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-03-24 23:41:31 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.64 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Mon 2014-03-24 16:26:55 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.11
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Mon 2014-03-24 16:21:56 +0100
                        message:
                          ndb - Bug#18173037 BUILD FAILURE WHEN USING -WERROR WITH ARGUMENTS
                           
                          Don't strip away -Werror=xxx flags from CMAKE_C_FLAGS
        ------------------------------------------------------------
        revno: 2876.563.54 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Wed 2014-04-16 11:47:23 +0200
        message:
          Merge 7.4(-r4280) -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.577.95
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Mon 2014-03-24 12:24:24 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, handle ENUM fields specially to handledifferent endianess
            ------------------------------------------------------------
            revno: 2876.577.94
            tags: clone-mysql-cluster-7.4.0-labs-april-build
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Mon 2014-03-24 09:11:40 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, handle ENUM fields specially to handledifferent endianess
            ------------------------------------------------------------
            revno: 2876.577.93
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Mon 2014-03-24 08:42:15 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, added proper termination of empty error_details string
            ------------------------------------------------------------
            revno: 2876.577.92
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Mon 2014-03-24 08:28:44 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, added proper termination of empty error_details string
            ------------------------------------------------------------
            revno: 2876.577.91
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Fri 2014-03-21 17:21:22 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, adding flushing of logs to make test ndb_rpl_conflict_epoch_extended_exceptions_table more predictable
            ------------------------------------------------------------
            revno: 2876.577.90
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_74_wl7761
            timestamp: Fri 2014-03-21 17:13:25 +0100
            message:
              WL#7761: Make atrt work on Mac OS X
            ------------------------------------------------------------
            revno: 2876.577.89 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_74_wl7747
            timestamp: Fri 2014-03-21 14:32:50 +0100
            message:
              merge WL#7747, BUG#18417623
                ------------------------------------------------------------
                revno: 2876.579.65 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: push_73_wl7747
                timestamp: Fri 2014-03-21 14:31:39 +0100
                message:
                  merge WL#7747, BUG#18417623
                    ------------------------------------------------------------
                    revno: 2875.612.63 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: push_72_wl7747
                    timestamp: Fri 2014-03-21 14:29:25 +0100
                    message:
                      merge WL#7747, BUG#18417623
                        ------------------------------------------------------------
                        revno: 2555.843.10
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: push_71_wl7747
                        timestamp: Fri 2014-03-21 14:26:38 +0100
                        message:
                          WL#7747, fixes BUG#18417623, move generator of TIME_SIGNAL to own thread to avoid weird cases when OS schedules different threads differently
            ------------------------------------------------------------
            revno: 2876.577.88
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Fri 2014-03-21 11:37:10 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, removed compiler warning
            ------------------------------------------------------------
            revno: 2876.577.87
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Fri 2014-03-21 10:57:54 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table, fixed test failure when referencing new value in a DELETE_ROW operation
            ------------------------------------------------------------
            revno: 2876.577.86
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.4_active_active
            timestamp: Fri 2014-03-21 08:36:50 +0100
            message:
              WL#6146 Extend Ndb conflict exceptions table
              
              The original conflict table is defined like this:
              CREATE TABLE original_table$EX  (
                  server_id INT UNSIGNED,
                  master_server_id INT UNSIGNED,
                  master_epoch BIGINT UNSIGNED,
                  count INT UNSIGNED,
                  original_table_pk_columns,
                  [additional_columns,]
                  PRIMARY KEY(server_id, master_server_id, master_epoch, count)
              ) ENGINE=NDB;
              
              Requirements
              ------------
              
                1) Support backwards compatibility so that existing conflict detection users
              needn't change their table schema
                   1.1 Mandatory columns may have any names
                   1.2 Main table primary key columns may have any names
                   1.3 Additional columns sharing names with main table columns are treated as
              user columns.
              
                2) Support use of a reserved column name namespace for metacolumns, allowing
              an arbitrary subset of non-pk main table columns to be recorded
                   2.1 NDB$ prefix used for all mandatory (non main table / user data columns)
                   2.2 Existing mandatory columns prefixed with NDB$
                   2.3 The full user table PK needn't be recorded
                   2.4 Matching of main table columns to exceptions table columns to be on the
              basis of name and type.
                   2.5 Informative error message to be produces when column name/types mismatch
                   2.5.1 NDB$ prefixed column with invalid suffix
                   2.5.2 Column type not matching expectation
              
                3) Support recording rejected non-primary key after values in the exceptions
              table where available
                   3.1 None, some or all non-primary key after values may be stored.
                   3.2 WRITE_ROW events may not have all column after values due to
                        i) Logging updated only on the Master
                       ii) Master having fewer columns than the Slave
                   3.3 UPDATE_ROW events may not have all column after values due to
                        i) Logging updated only on the Master
                       ii) Master having fewer columns than the Slave
                   3.4 DELETE_ROW events have no after values
                   3.5 All non primary key main table columns must be NULLABLE or natively
              DEFAULTed in the Exceptions table
              
                4) Support recording optional conflicting operation type
                   4.1 e.g. NDB$OP_TYPE INT UNSIGNED NOT NULL
                   4.2 Current types : WRITE_ROW, UPDATE_ROW, DELETE_ROW
              
                5) Support recording optional conflict cause
                   5.1 e.g. NDB$CFT_CAUSE INT UNSIGNED NOT NULL
                   5.2 Current causes :
                       ROW_DOES_NOT_EXIST (UPDATE_ROW, WRITE_ROW)
                       ROW_ALREADY_EXISTS (WRITE_ROW)
                       DATA_IN_CONFLICT   (UPDATE_ROW, DELETE_ROW)
                       TRANS_IN_CONFLICT  (*)
                   5.3 Note that TRANS_IN_CONFLICT obscures per-row causation information, so
                       perhaps warn when a table with transactional conflict detection is
                       configured with a conflict cause column in the exceptions table
              
                6) Support recording optional originating transaction id info
                   6.1 e.g. NDB$ORIG_TRANSID BIGINT UNSIGNED NOT NULL
                   6.2 Note that this is most likely to be useful with transactional conflict
                       detection, so perhaps warn if it's configured for a non-transactional
                       algorithm
              
                7) Additional added functionality
                   7.1 Support for referencing old/new values of non-primary key columns
                     <column_name>$old and <column_name>$new (same as just writing <column_name>) columns
                    in the extended exceptions will reference old and new values in update operations. For
                   write operations the old value will be NULL and for delete operations the new value will be NULL.
                   Checks will be made that such columns will
                   have to be declared as nullable or with a default value (native ndb default) and cannot be part of the primary key.
                   7.2 NDB$OP_TYPE will be also aligned with ENUM('WRITE_ROW','UPDATE_ROW', 'DELETE_ROW') NOT NULL
                   7.3  NDB$CFT_CAUSE will also be aligned with
              ENUM('ROW_DOES_NOT_EXIST','ROW_ALREADY_EXISTS','DATA_IN_CONFLICT','TRANS_IN_CONFLICT') NOT NULL
        ------------------------------------------------------------
        revno: 2876.563.53 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Wed 2014-04-16 11:21:06 +0200
        message:
          Merge 7.4(-r4270) -> trunk-cluster
           - removed #ifdef HAVE_NDB_BINLOG from ndb_wait_for_binlog_injector() 
             mysqltest.cc since that define is only in the ha_ndb code
             starting with 5.7
            ------------------------------------------------------------
            revno: 2876.577.85
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Thu 2014-03-20 08:48:14 +0100
            message:
              ndb - corrected merge of: more consistency checks on received messages
              
              Bug #18426180    ENHANCED DETECTION AND DUMP OF CORRUPT MESSAGES AND SIGNALS RECEIVED
              
              update of readPtr has moved earlier in 7.4
            ------------------------------------------------------------
            revno: 2876.577.84 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-03-20 11:58:53 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.64 [merge]
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-03-20 11:56:28 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.62 [merge]
                    committer: saroj tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-03-20 11:52:53 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.9
                        committer: saroj tripathy <saroj.tripathy@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-03-20 11:36:08 +0530
                        message:
                          added extra # after the comment line
            ------------------------------------------------------------
            revno: 2876.577.83 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Wed 2014-03-19 16:24:31 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.63 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Wed 2014-03-19 16:22:00 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.61 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Wed 2014-03-19 16:16:47 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.8
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Wed 2014-03-19 16:11:08 +0100
                        message:
                          ndb - more consistency checks on received messages
                          
                          Bug #18426180    ENHANCED DETECTION AND DUMP OF CORRUPT MESSAGES AND SIGNALS RECEIVED
                          
                          Checks are added that verifiexs that message length match embedded signal size.
                          
                          Also when a consistency check fail, message or signal data are dumped in hex format.
                          
                          The hex dump are printed to stdout via ndbout.
                        ------------------------------------------------------------
                        revno: 2555.843.7
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Wed 2014-03-19 15:46:09 +0100
                        message:
                          ndb - dump all words in signal correct in trace file
                          
                          Bug #18419554	 SIGNAL TRACE TRASHED FOR SIGNALS WITH MORE THAN 18 WORDS (UNKNOWN SIGNAL TYPE)
            ------------------------------------------------------------
            revno: 2876.577.82 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-19 18:23:43 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.62
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-19 18:22:46 +0530
                message:
                  added one more missing test case in daily-basic
            ------------------------------------------------------------
            revno: 2876.577.81 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-19 18:02:19 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.61
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-19 18:01:17 +0530
                message:
                  modify 2 bug numbers
            ------------------------------------------------------------
            revno: 2876.577.80 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-19 17:21:31 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.60
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-19 17:20:32 +0530
                message:
                  added extra # after the comment line
            ------------------------------------------------------------
            revno: 2876.577.79 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-19 16:43:17 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.59 [merge]
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-19 16:42:35 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.60
                    committer: saroj tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-03-19 16:32:29 +0530
                    message:
                      removed extra commented test cases
            ------------------------------------------------------------
            revno: 2876.577.78 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-03-19 16:07:02 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.58 [merge]
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-03-19 15:50:38 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.59 [merge]
                    committer: saroj tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-03-19 15:34:05 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.6
                        committer: saroj tripathy <saroj.tripathy@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-03-19 15:29:50 +0530
                        message:
                          added extra # after the commented line
            ------------------------------------------------------------
            revno: 2876.577.77
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7507
            timestamp: Tue 2014-03-18 16:50:05 +0100
            message:
              WL#7507: Decrease waits in restore and LCPs, also added some extra jams and debug printouts found useful as part of debugging process
            ------------------------------------------------------------
            revno: 2876.577.76 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_74_bug18344004
            timestamp: Tue 2014-03-18 16:39:22 +0100
            message:
              Merge BUG#18344004
                ------------------------------------------------------------
                revno: 2876.579.57 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: push_73_bug18344004
                timestamp: Tue 2014-03-18 16:38:14 +0100
                message:
                  Merge BUG#18344004
                    ------------------------------------------------------------
                    revno: 2875.612.58 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: push_72_bug18344004
                    timestamp: Tue 2014-03-18 16:21:51 +0100
                    message:
                      Merge BUG#18344004
                        ------------------------------------------------------------
                        revno: 2555.843.5
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: push_71_bug18344004
                        timestamp: Tue 2014-03-18 16:19:08 +0100
                        message:
                          BUG#18344004: Fix mappings of NDB error codes using restricted error codes in the MySQL Server and fix error codes for BackupInProgress
            ------------------------------------------------------------
            revno: 2876.577.75 [merge]
            committer: saroj tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-18 20:32:50 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.56
                committer: saroj tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-03-18 20:31:01 +0530
                message:
                  changes to daily-basic and daily-devel file for autotest restructure
            ------------------------------------------------------------
            revno: 2876.577.74 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-18 17:36:46 +0530
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.55 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-03-18 17:33:49 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.57 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-03-18 17:28:21 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.4
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-03-18 17:21:58 +0530
                        message:
                          Bug #18196562: NDBAPI : RACE BETWEEN LATE TRANSID_AI AND API CLOSETRANSACTION()
                          
                          An NDBAPI application sends a scan query to a data node, which is 
                          processed by the TC. The TC forwards the LQHKEYREQ request to the 
                          appropriate LDM, and aborts the transaction if it does not receive 
                          a LQHKEYCONF response within the specified time limit. After the 
                          transaction is successfully aborted, the TC sends a TCROLLBACKREP 
                          to the NDBAPI client. The NDBAPI client processes the TCROLLBACKREP 
                          by cleaning up the Ndb objects associated with the transaction.
                          
                          The NDBAPI client receives the data which it has requested in the 
                          form of TRANSID_AI signals sent by the LDM. These TRANSID_AI signals 
                          are buffered for sending at the data node, and may be delivered to the 
                          NDBAPI client after a delay. On receiving a TRANSID_AI signal, the 
                          NDBAPI checks the transaction state and ID: if these are as expected, 
                          it processes the signal using the Ndb objects associated with that 
                          transaction.
                          
                          The bug occurs when all the following conditions are fulfilled:
                          
                          - The TC aborts a transaction due to delays and sends a TCROLLBACPREP 
                          to the NDBAPI client. Simultaneously, a TRANSID_AI which has been 
                          buffered for delivery at an LDM is delivered to the same NDBAPI client.
                          
                          - The NDBAPI client considers the transaction complete on receiving a 
                          TCROLLBACKREP and immediately closes the transaction.
                          
                          - The NDBAPI client has a separate receiver thread running concurrently 
                          with the transaction closing thread.
                          
                          - The reception of the late TRANSID_AI interleaves with the user 
                          thread's transaction close so that the TRANSID_AI processing passes the 
                          normal checks (TransId, 'magic number') before the closeTransaction 
                          resets the transaction state and invalidates the receiver.
                          
                          The receiver thread proceeds to continue working on the TRANSID_AI 
                          signal using the invalidated receiver. Since the receiver is already 
                          invalidated, its usage results in a core.
                          
                          Added a fix in the NDBAPI, so that the Ndb object cleanup done for 
                          TCROLLBACKREP now includes invalidation of the transaction ID. So, for 
                          a given transaction, any signal which is received after the 
                          TCROLLBACKREP arrives, does not pass the transaction ID check and is 
                          silently dropped. Duplicated the fix for TC_COMMITREF, TCROLLBACKREF, 
                          TCKEY_FAILCONF and TCKEY_FAILREF.
            ------------------------------------------------------------
            revno: 2876.577.73 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-18 11:38:07 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.54 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-03-18 11:34:02 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.56 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-03-18 11:27:36 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.3
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-03-18 10:49:42 +0100
                        message:
                          Fix for Bug#11746486 CLUSTER RELATED PARAMETERS CANNOT BE SET IN MY.CNF
                          
                          The config parameters ndb_index_stat_cache_entries and 
                          ndb_index_stat_update_freq has been deprecated, but was
                          not completely removed from the set of variables. 
                          
                          This fix removes these variables.
                          
                          NOTE: The bug report also mention multi_read_range and 
                          ndb_use_transactions. These will require a document update, see
                          bug report.
            ------------------------------------------------------------
            revno: 2876.577.72 [merge]
            committer: Saroj Tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-18 10:08:23 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.53
                committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-03-18 10:06:54 +0100
                message:
                  changes to daily-basic and daily-devel file for autotest restructure
            ------------------------------------------------------------
            revno: 2876.577.71
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-18 09:21:37 +0100
            message:
              This commit implements WL#7692: "Add Ptr<T> constructor to detect use of uninitialized variables".
              
              It adds a default constructor that sets all bytes in Ptr<T> variables to 0xff (for debug builds). Also, Ptr<T>:isNull() asserts for i<=RNIL, to detect if this method has been called on an object that was not properly initialized.
              
              To make the code compilable, it also make a few other changes:
              
              * Aggregate initialization is converted to constructor initialization, e.g. "Ptr<T> ptr = {0, RNIL};" becomes "Ptr<T> ptr(0, RNIL);".
              
              * Some Ptr<T> definitions have been moved up above a 'goto' statement, so that 'goto' does not skip initialization.
              
              * Some case-alternatives have been enclosed in curly brackets, to limit the scope of Ptr<T> variables declared there.
              
              * The LINT_SET_PTR macro has been removed because it used aggregate initialization of of Ptr and thus broke the  build for "linux-release". It was not used consequently and did not seem to serve any purpose. (If dummy initialization is needed, it is better to so via the Ptr constructor, see also BUG 18259286 - REMOVE LINT_INIT/UNINIT_VAR .)
            ------------------------------------------------------------
            revno: 2876.577.70 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-03-17 21:38:42 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.52
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-03-17 21:24:06 +0100
                message:
                  ndb - revert test configuration for testFK
                  
                  With the fix for
                  Bug #18082045	 TESTFK CREATEDROPWITHDATA NEED HIGH MAXNOOFCONCURRENTSCANS AND MAXNOOFLOCALSCANS
                  
                  the higher MaxNoOfConcurrentScans are no longer needed for testFK -n CreateDropWithData introduced in
                  
                  revision-id: mauritz.sundell@oracle.com-20140110163326-48yzitisfij7y26m
                  message:
                    ndb - test: testFK -n CreateDropWithData T1 needs higher MaxNoOfConcurrentScans
                    
                    Raising to MaxNoOfConcurrentScans = 300.
                    Experiments gave that ~257 is needed.
            ------------------------------------------------------------
            revno: 2876.577.69 [merge]
            committer: Saroj Tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-03-17 17:49:42 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.51 [merge]
                committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-03-17 17:47:40 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.55 [merge]
                    committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-03-17 17:42:13 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.2
                        committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-03-17 17:39:03 +0100
                        message:
                          changes to daily-basic and daily-devel file for autotest restructure
            ------------------------------------------------------------
            revno: 2876.577.68
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7514
            timestamp: Mon 2014-03-17 12:22:41 +0100
            message:
              WL#7514: Added lots of restart printouts, added large segments of new comments on restarts
            ------------------------------------------------------------
            revno: 2876.577.67
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_bug18345287
            timestamp: Sat 2014-03-15 09:39:17 +0100
            message:
              BUG#18345287 seize often uses .i value from seize also when unsuccessful, ensure all pools handle this correctly
            ------------------------------------------------------------
            revno: 2876.577.66
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: wl7509
            timestamp: Fri 2014-03-14 23:25:20 +0100
            message:
              WL#7509: Increase parallelism at same disk checkpoint speed to fully utilize all LDM threads in parallel when running LCPs
            ------------------------------------------------------------
            revno: 2876.577.65 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18082045-74
            timestamp: Fri 2014-03-14 20:44:02 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.50
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18082045-73
                timestamp: Fri 2014-03-14 20:27:50 +0200
                message:
                  bug#18082045 trigpend2-x1.diff
                  add some signal log printouts
                ------------------------------------------------------------
                revno: 2876.579.49
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18082045-73
                timestamp: Fri 2014-03-14 20:23:11 +0200
                message:
                  bug#18082045 trigpend2.diff
                  limit parallel FK cascading scans
                ------------------------------------------------------------
                revno: 2876.579.48
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18082045-73
                timestamp: Fri 2014-03-14 20:19:59 +0200
                message:
                  bug#18082045 mtrtest2.diff
                  mtr test (fails on too many TC scans)
            ------------------------------------------------------------
            revno: 2876.577.64 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-03-14 15:22:47 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.47
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2014-03-14 15:21:12 +0100
                message:
                  Fix for bug#16209645
                  
                    MTR NDB_WL946.PRE/POST.TEST CRASHES MYSQLD ON SOLARIS/SPARC
                  
                  Disable these test on big endian platforms.
                  
                  The test restore a backup image created on a little endian arch.
                  As the backed up FRM binary representation depend on
                  the platform arch, this can never be tested on big endian platforms
            ------------------------------------------------------------
            revno: 2876.577.63 [merge]
            committer: Saroj Tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-03-13 17:53:53 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.46 [merge]
                committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-03-13 17:52:29 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.54 [merge]
                    committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-03-13 17:48:18 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.843.1
                        committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-03-13 16:54:12 +0100
                        message:
                          removed blank line from daily-devel
            ------------------------------------------------------------
            revno: 2876.577.62 [merge]
            committer: Saroj Tripathy <saroj.tripathy@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-03-13 17:17:05 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.45 [merge]
                committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-03-13 17:13:59 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.53 [merge]
                    committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-03-13 17:07:54 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.172
                        committer: Saroj Tripathy <saroj.tripathy@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-03-13 16:24:58 +0100
                        message:
                          removed blank line from the daily-devel
            ------------------------------------------------------------
            revno: 2876.577.61 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-03-13 10:37:38 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.44
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-03-13 10:33:56 +0100
                message:
                  Fix for Bug#18390140
                  
                   	CODE CLEANUP: MAKE BITMASK CLASSES CONST-CORRECT
                  
                  Some members in the Bitmask* classes are not cosnt-correct.
                  Fix this - No functional change.
            ------------------------------------------------------------
            revno: 2876.577.60 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-11 17:21:48 +0000
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.43 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-03-11 17:09:51 +0000
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.52
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-03-11 16:59:41 +0000
                    message:
                      Bug #18375920 	NDB TESTING : SYNC_SLAVE_WITH_MASTER CAN TIMEOUT AND CON
                      TINUE
                      
                      Modify testcase + result to align with >= 7.2 changes :
                        - Default to v2 Binlog RBR events
                        - Default to ROW format, including for MyISAM changes.
            ------------------------------------------------------------
            revno: 2876.577.59 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-03-11 15:07:59 +0000
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.42 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-03-11 15:03:01 +0000
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.51 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-03-11 14:28:05 +0000
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.817.171
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-03-11 12:52:39 +0000
                        message:
                          Bug #18375920 	NDB TESTING : SYNC_SLAVE_WITH_MASTER CAN TIMEOUT AND CONTINUE
                          
                          This fix changes mysqltest's sync_slave_with_master 
                          ndb-binlog-injector-sync timeout from 30s to 150s, and
                          makes timeout a hard test failure.
                          
                          This causes the 'fishy' ndb_rpl_commit_afterflush testcase to fail, as
                          it relied on a timeout+pass behaviour from sync_slave_with_master.
                          
                          This testcase is removed and a new testcase is added to replace it:
                          ndb_binlog_flush_tables_with_read_lock.test.
                          
                          The new testcase verifies the basics of the behaviour, with 2 MySQLDs
                          attached to a cluster.  
            ------------------------------------------------------------
            revno: 2876.577.58
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7533_2
            timestamp: Fri 2014-03-07 16:23:54 +0100
            message:
              Removed warning removal not properly handled yet
            ------------------------------------------------------------
            revno: 2876.577.57
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7533_2
            timestamp: Fri 2014-03-07 16:22:50 +0100
            message:
              WL#7533: Part 2, Fix warnings
            ------------------------------------------------------------
            revno: 2876.577.56
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7533_3
            timestamp: Fri 2014-03-07 16:14:04 +0100
            message:
              WL#7533: Part 3: Upgrade asserts to requires where it makes no sense to continue
            ------------------------------------------------------------
            revno: 2876.577.55
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7533_1
            timestamp: Fri 2014-03-07 16:07:20 +0100
            message:
              WL#7533: Part 1: Turn asserts into require in test parts
            ------------------------------------------------------------
            revno: 2876.577.54 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Thu 2014-03-06 12:58:12 +0000
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.41 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-03-06 12:54:21 +0000
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.50 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2014-03-06 12:50:03 +0000
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.817.170 [merge]
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2014-03-06 12:44:20 +0000
                        message:
                          Merge 6.3->7.1
                            ------------------------------------------------------------
                            revno: 2555.836.11
                            committer: Frazer Clement <frazer.clement@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Thu 2014-03-06 12:25:09 +0000
                            message:
                              Bug #18321018 	NDB : MTR + PB2 LOGS DON'T INCLUDE DATA NODE FAILURE LOGS
                              
                              mysql-test-run is modified so that unexpected server failure will
                              result in collection and presentation of some data node error logs,
                              output logs and crash jam/signal traces.
                              
                              The main intention here is that crashes occurring in CluB/PB2 have more
                              context to begin debugging with.
                              
                              Depending on user experience + feedback, the volume of information 
                              produced can be adjusted
            ------------------------------------------------------------
            revno: 2876.577.53 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-03-05 10:18:54 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.40
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3-bug18350021
                timestamp: Wed 2014-03-05 10:04:40 +0100
                message:
                  Bug #18350021  REMOVE USAGE OF LINT_INIT FROM NDB
                   - Remove the LINT_INIT usage
            ------------------------------------------------------------
            revno: 2876.577.52
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7504
            timestamp: Tue 2014-03-04 12:12:46 +0100
            message:
              WL#7504: Multi-thread touch of memory for memory allocation to speed up restarts
            ------------------------------------------------------------
            revno: 2876.577.51 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-02-27 14:10:00 +0100
            message:
              Merge 7.4-seagull -> 7.4
                ------------------------------------------------------------
                revno: 2876.586.10
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Thu 2014-02-27 12:54:39 +0100
                message:
                  WL#7646 Ndb_component usage
                   - add missing parenteshis in comment
                   - remove misplaced comment
                ------------------------------------------------------------
                revno: 2876.586.9
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Thu 2014-02-27 10:04:37 +0100
                message:
                  WL#7646 Ndb_component usage
                  
                   - remove the newly introuduced dependency on my_attribute.h since that file is
                    being removed in the future.
                   - Also ATTRIBUTE_FORMAT is removed, thus resort to direct __attribute__ usage.
                    This is safe because __attribute__ is defined away for compilers which don't have it
                ------------------------------------------------------------
                revno: 2876.586.8
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-26 13:43:19 +0100
                message:
                  WL#7646
                  - Rename ndb_index_stat_waiter to client_waiting
                ------------------------------------------------------------
                revno: 2876.586.7
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-26 13:38:32 +0100
                message:
                  WL#7646 Wait for log message while waiting for stop
                  
                   - Adding log messages in appropriate places will make it clear which thread
                     (if any) is hanging during shutdown.
                   - Changing from log_verbose() -> log_info() to make important printouts
                     always show
                ------------------------------------------------------------
                revno: 2876.586.6
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-26 13:31:31 +0100
                message:
                  WL#7647 Ndb_component usage - thread safe wakeup
                   - Add abstract do_wakeup() function which subclasses must implement
                     in order to detect the stop as fast as possible. This involves
                     signaling conditions etc.
                   - Move the wakeup functionality of our three components into
                     do_wakeup() functions.
                ------------------------------------------------------------
                revno: 2876.586.5
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Tue 2014-02-25 15:34:52 +0100
                message:
                  WL#7646 Improve ndbcluster Ndb_component usage
                   - Returning an error from ndbcluster_init() will only
                     lead to a warning being printed by the MySQL Server
                     and the next plugin to be loaded. I.e the error
                     is more or less silently ignored.
                   - Handle the problem by printing an error to the log
                     and then exit the MySQL Server in case an error is detected
                     in ndbcluster_init(). Errors like this should be rare, but
                     better to have a disctinct behaviour than allowing 
                     the MySQL Server to continue start.
                   - Workaround is to start the server "anyway" is to use --skip-ndbcluster 
                   - Remove code which tries to cleanup failed ndbcluster_init
                ------------------------------------------------------------
                revno: 2876.586.4
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Tue 2014-02-25 14:52:17 +0100
                message:
                  WL#7646 Improve ndbcluster Ndb_component usage(binlog thread part)
                    - Rely solely on Ndb_component start/stop for Ndb_binlog_thread
                     startup and shutdown, this involves:
                    -- use is_stop_requested() to check if time to stop thread
                    -- implement stop() (just for having a place the put the comment about
                        not waking up)
                    -- remove "running" since that state is implemented in baseclass
                    -- remove "ndbcluster_binlog_terminating"
                ------------------------------------------------------------
                revno: 2876.586.3
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Tue 2014-02-25 14:50:37 +0100
                message:
                  WL#7646 Improve ndbcluster Ndb_component usage (util thread part)        
                   - Rely solely on Ndb_component start/stop for Ndb_util_thread
                     startup and shutdown, this involves:
                    -- use is_stop_requested() to check if time to stop thread
                    -- implement stop() which first wakes the thread from potential sleep
                       before continuing in baseclass stop function
                    -- remove "running" since that state is implemented in baseclass
                    -- remove "COND_ready" since the case where thread started up but "failed" is
                       unlikely and not handled properly anyway. (The two returns in early phase
                       of Ndb_util_thread::do_run() will be removed later when util thread get rid
                       of THD and Thd_ndb dependency.)
                    -- furthermore there are still a few case in all ndbcluster daemon threads where
                       they silently exit...
                ------------------------------------------------------------
                revno: 2876.586.2
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Tue 2014-02-25 14:48:52 +0100
                message:
                  WL#7646 Improve ndbcluster Ndb_component usage (index stat part)
                   - Rely solely on Ndb_component start/stop for Ndb_index_stat_thread
                     startup and shutdown, this involves:
                     -- use is_stop_requested() to check if time to stop thread
                     -- implement stop() which first wakes the thread from potential sleep
                        before continuing in baseclas stop function
                     -- implement wakeup() which tells the thread that there are work to do.
                     -- remove "running" since that state is implemented in baseclass
                     -- remove "COND_ready" since the case where thread starteed up but "failed" 
                        does not happen in the index_stat_thread. 
                     -- move "ndb_index_stat_waiter" global variable to Ndb_index_stat_thread
                   - Remove extern declaration of ndb_index_stat_thread instance from ha_ndbcluster.h
                     but add that same declaration to ndb_index_stat.cc since implementation still
                     uses it's own instance.
                ------------------------------------------------------------
                revno: 2876.586.1 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4
                timestamp: Tue 2014-02-25 14:01:16 +0100
                message:
                  Merge
                    ------------------------------------------------------------
                    revno: 2876.585.3
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.4-seagull
                    timestamp: Fri 2014-02-21 14:04:00 +0100
                    message:
                      ndb
                       - pass handlerton* directly to ndbcluster_binlog_init() instead of using global variable
                       - remove yet another circular dependcy on ndbcluster_hton
                    ------------------------------------------------------------
                    revno: 2876.585.2
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.4-seagull
                    timestamp: Thu 2014-02-20 14:35:36 +0100
                    message:
                      WL#7646 Improve ndbcluster Ndb_component usage
                      
                      - fix problem in ndb binlog thread looking at the wrong variable to determine
                         if it should shut down the thread
                    ------------------------------------------------------------
                    revno: 2876.585.1 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.4
                    timestamp: Thu 2014-02-20 11:55:02 +0100
                    message:
                      Merge 7.4-seagull -> 7.4
                        ------------------------------------------------------------
                        revno: 2876.584.10
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Thu 2014-02-20 11:35:41 +0100
                        message:
                          WL#7646 Improve ndbcluster Ndb_component usage
                          
                           - add ndb_binlog_thread.cc to the build
                        ------------------------------------------------------------
                        revno: 2876.584.9
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Thu 2014-02-20 11:24:08 +0100
                        message:
                          WL#7646 Improve ndbcluster Ndb_component usage
                           - Add proper DBUG_VOID_RETURN in do_run() functions
                        ------------------------------------------------------------
                        revno: 2876.584.8
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Thu 2014-02-20 11:17:08 +0100
                        message:
                          WL#7646 Improve ndbcluster Ndb_component usage
                                 - remove usage of instance variable name from member functions of Ndb_index_stat_thread
                        ------------------------------------------------------------
                        revno: 2876.584.7
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Thu 2014-02-20 11:13:31 +0100
                        message:
                          WL#7646 Improve ndbcluster Ndb_component usage
                          
                           - Add "name" to Ndb_component
                            - Add logging functionality to be used from each component which print out the
                              component name and desired log message to server log file
                            - Extend logging in the components making it possible to see what each component
                               is doing or waiting for - partiularly the startup/shutdown stages
                           - Add ndb_log_* logging functions, encapsulating the sql_print_* functions
                        ------------------------------------------------------------
                        revno: 2876.584.6
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Wed 2014-02-19 20:58:37 +0100
                        message:
                          WL#7646 Improve ndbcluster Ndb_component usage
                           - remove usage of instance variable name from member functions of Ndb_util_thread
                        ------------------------------------------------------------
                        revno: 2876.584.5
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Wed 2014-02-19 20:56:16 +0100
                        message:
                          WL#7646 Improve ndbcluster Ndb_component usage
                           - convert ndb_binlog_thread_func into a Ndb_component
                           - rename "ndb_binlog_thread_running" to "Ndb_binlog_thread:running"
                        ------------------------------------------------------------
                        revno: 2876.584.4
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.4-seagull
                        timestamp: Wed 2014-02-19 18:15:38 +0100
                        message:
                          WL#7571 Reduce dependency of THD in ndbcluster
                          
                           - remove unused THD arg to ha_ndbcluster::alter_frm()
            ------------------------------------------------------------
            revno: 2876.577.50 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-02-26 09:32:20 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.39 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-02-26 09:29:42 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.49 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-02-26 09:22:39 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.169
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-02-26 09:21:01 +0100
                        message:
                          Fix for bug#18235177 DEBUG BUILDS MAY FAIL IN NDBASSERT(FINDTRANSACTION() == ZNOT_FOUND)
                           
                          Dirty operations (non-transactional) are not inserted into the transactional
                          hash-list. When deleting such a transaction, there is an assert
                          'ndbassert(findTransaction() == ZNOT_FOUND) which occasionally fails
                                
                          This assert is too strict, as transactional uniqueness is only required 
                          for non-dirty operations. tcOpRec ids of dirty operations are reused
                          within TC (which assign them), and uniqueness for such operations
                          are not guaranteed, nor required.
            ------------------------------------------------------------
            revno: 2876.577.49 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Tue 2014-02-25 13:39:29 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.38
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Tue 2014-02-25 13:18:44 +0100
                message:
                  BUG#18304820 NDB BINLOG THREAD MY_NET_INIT CAUSES SMALL MEMORY LEAK
                   - release the memory allocated for "net created without vio" before releasing the THD
            ------------------------------------------------------------
            revno: 2876.577.48 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-02-25 08:49:04 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.37 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-02-25 08:48:05 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.48 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2014-02-25 08:46:32 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.168
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Tue 2014-02-25 08:44:40 +0100
                        message:
                          ndb - update memory usage of index memory on drop table
                          
                          Bug #18296810     INDEX MEMORY IS NOT FREEING EVEN IF TABLE IS DROPPED FROM THE DATABASE ..
                          
                          The global counters in Cmvmi used for reporting index memory usage are now updated on drop table.
            ------------------------------------------------------------
            revno: 2876.577.47 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-02-24 14:30:03 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.36 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-02-24 14:27:12 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.47 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-02-24 14:23:10 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.167 [merge]
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-02-24 14:21:41 +0100
                        message:
                          Merge 7.0 -> 7.1
                            ------------------------------------------------------------
                            revno: 2555.827.74
                            committer: Ole John Aske <ole.john.aske@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Mon 2014-02-24 14:20:23 +0100
                            message:
                              Fix for Bug#18293112
                              
                                NDB.HPP DEPENDS ON PORTLIB/NDBTICK.HPP THAT'S NOT INCLUDED IN BINARY DIST
                              
                              This patch fix a regression introduced by the refactoring part of
                              bug #17647637: 'Refactor of NdbTick as prep for real fix.'
                              
                              5252 Ole John Aske	2013-11-18
                              Part1 of fix for bug#17647637: Refactor of NdbTick as prep for real fix.
                              
                              Thix patch reverts the introduction of 'NDB_TICKS' and 'portlib/NdbTick.h'
                              in the API header files.
            ------------------------------------------------------------
            revno: 2876.577.46 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Fri 2014-02-21 00:59:52 +0000
            message:
              Merge 7.3->7.4
                ------------------------------------------------------------
                revno: 2876.579.35 [merge]
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2014-02-21 00:55:56 +0000
                message:
                  Merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.612.46 [merge]
                    committer: Frazer Clement <frazer.clement@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2014-02-21 00:44:05 +0000
                    message:
                      Merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.817.166 [merge]
                        committer: Frazer Clement <frazer.clement@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Fri 2014-02-21 00:38:36 +0000
                        message:
                          Merge 7.0->7.1
                            ------------------------------------------------------------
                            revno: 2555.827.73
                            committer: Frazer Clement <frazer.clement@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Fri 2014-02-21 00:20:39 +0000
                            message:
                              Bug #16693068     7.1.18 TO 7.1.22-27 UPGRADE NOT POSSIBLE IF LARGE NUMBER OF TABLES PRESENT
                              
                                The Dbdih::getInstanceKey() function is used to determine which LDM instance
                                is responsible for storing a table fragment, based on the table fragment's
                                log_part_id, which is set at table creation time, and must be stable between
                                initial restarts of a node.
                                    
                                The DbDih::getInstanceKey() function should return a value between 1 and 
                                NDBMT_MAX_BLOCK_INSTANCES-1 inclusive, identifying which LDM worker array entry
                                should be used for a given fragment.  Value 0 refers to the LDM Proxy instance
                                in ndbmtd, and so is not used.  
                                However getInstanceKey() has a bug where it can return numbers between 1 and 
                                NDBMT_MAX_BLOCK_INSTANCES inclusive.  This exceeds its range.
                                
                                Where it returns NDBMT_MAX_BLOCK_INSTANCES as a result, this causes an 
                                array out-of-bounds exception when mapping an instance key to a receiving
                                thread, and this can have various bad effects. 
                              
                                This bug is not generally observed as getInstanceKey() uses the log_part_id
                                to determine the instance number, and recent code changes keep log_part_ids
                                far below values which can return NDBMT_MAX_BLOCK_INSTANCES.
                              
                                Older code stored unlimited log part ids in the fragment definition,
                                and used modulo division by the number of LDM instances at 'runtime' 
                                to get the correct log part.
                              
                                More recent code stores log part ids modulo the current number of running
                                LDM instances at table creation time, and continues to determine fragment 
                                instance keys (LDM instance numbers) based on modulo division.  It's not
                                that clear exactly what problem was solved here, but it happens to hide the 
                                getInstanceKey() bug.
                              
                                The number of running LDM instances is always < NDBMT_MAX_BLOCK_INSTANCES -1,
                                so normal code does not hit the bug in getInstanceKey()
                              
                                During an upgrade we can load table definitions on disk from an older version
                                with non-modulo log part numbers.  This can expose the getInstanceKey() bug
                                which has various symptoms (hangs, crashes). 
                                
                                To solve this, this patch performs modulo division of log_part_ids by the running
                                node's number of LDM instances when loading a potentially 'old' (unbounded) fragment
                                definition from disk.  This extends the guarantee that fragment log_part_ids will
                                be < NDBMT_MAX_BLOCK_INSTANCES - 1 to cover the upgrade scenario so that the 
                                getInstanceKey() bug is not exposed.  This fix is aligned with the previous modification
                                which stores these 'modulo log part ids'.
                              
                                To further clarify the new situation, getInstanceKey() is 
                                modified to no longer perform modulo division on the 
                                log_part_id as it is no longer required.
                                Additionally, getInstanceKey() is modified to require that 
                                the resulting instance key is in range.
                                Further, existing code which sets a fragment's log_part_id 
                                is modified to require that the log part id is within range 
                                (0 to NDBMT_MAX_BLOCK_INSTANCES -2 inclusive).
                                A new define NDBMT_MAX_WORKER_INSTANCES is defined as 
                                NDBMT_MAX_BLOCK_INSTANCES - 1, in which terms the range of the 
                                log_part_id is 0 to NDBMT_MAX_WORKER_INSTANCES -1 inclusive.
                              
                                Performing %= num_ldm_instances on the log_part_id should be
                                safe for non-initial restarts, as those are only allowed when
                                the number of log parts is not changed.
                                
                                One downside of storing log parts in their 'modulo' form is that an
                                increase in the number of log parts cannot remap existing fragments
                                to the new log parts over an initial restart.
                                That problem was introduced by the original change to store modulo
                                log parts and is considered out of scope here.
                              
                                testUpgrade is extended to cover the scenario required here :
                                 - Many tables
                                 - Many fragments/table
                                 - System restart upgrade rather than node restart upgrade
                                    
                                An execution of this upgrade test is added to the upgrade-tests
                                suite for running in Autotest.
            ------------------------------------------------------------
            revno: 2876.577.45 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-19 22:17:14 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.34 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Wed 2014-02-19 22:16:38 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.45 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-02-19 22:15:56 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.165 [merge]
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Wed 2014-02-19 22:13:06 +0100
                        message:
                          Merge 7.1-list-tables-patch -> 7.1
            ------------------------------------------------------------
            revno: 2876.577.44 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-19 20:29:51 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.33 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Wed 2014-02-19 20:29:21 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.44 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-02-19 20:28:17 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.842.1
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Wed 2014-02-19 20:24:17 +0100
                        message:
                          Bug#18276327 NDB_SHOW_TABLES DOES NOT PRINT FULL ERROR MESSAGE WHEN CONNECT FAILS
                          
                           - Print the last error message from Ndb_cluster_connection when Ndb_cluster_connection::connect() fails
            ------------------------------------------------------------
            revno: 2876.577.43 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-19 15:53:10 +0100
            message:
              Merge 7.4-seagull -> 7.4
                ------------------------------------------------------------
                revno: 2876.584.3
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-19 15:51:05 +0100
                message:
                  WL#6815 Adapt MySQL Cluster to 5.7
                        
                    - remove usage of char buffer + llstr() to circumvent limitation of
                      not having %llu in DBUG print functions.
                   - The 5.7(actually 5.6) server supports %llu in my_snprintf and thus also DBUG which uses
                      my_snprintf
                ------------------------------------------------------------
                revno: 2876.584.2
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-19 15:28:07 +0100
                message:
                  WL#7571Reduce dependency of THD in ndbcluster
                  
                   - reduce ndbcluster_show_status_binlog() to return a string 
                     which the caller may print. This avoid the need for THD, stat_print etc
                     in the ndbcluster_show_status_binlog() function and it's declaration.
                ------------------------------------------------------------
                revno: 2876.584.1
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-19 14:17:22 +0100
                message:
                  WL#6815 Adapt MySQL Cluster to 5.7
                  
                   - remove usage of char buffer + llstr() to circumvent limitation of
                     not having %llu in print functions.
                   - The 5.7(actually 5.6) server supports %llu in my_snprintf and thus also DBUG which uses
                     my_snprintf
            ------------------------------------------------------------
            revno: 2876.577.42 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-19 12:21:11 +0100
            message:
              Merge 7.4-seagull -> 7.4
                ------------------------------------------------------------
                revno: 2876.583.2
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-19 12:19:20 +0100
                message:
                  WL#7570 Remove ifdefs which are not necessary since trunk has it all
                  
                   - removal of the 'closed_cached_tables' emulator function showed that the
                     have_lock argument to ndb_err() was unused.
                   - remove  the unused "have_lock" in ndb_err and calling functions
                ------------------------------------------------------------
                revno: 2876.583.1
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4-seagull
                timestamp: Wed 2014-02-19 11:46:33 +0100
                message:
                  WL#7570 Remove ifdefs which are not necessary since trunk has it all
                  
                   - ha_ndbcluster have only two use cases for close_cached_tables(), either it closes
                    alll tables or a specifc named table. The other parameters used are always the same.
                   - implement ndb_tdc_close_cached_table() and ndb_tdc_close_cached_tables() for interfacing
                    with MySQL Server's table defintion cache
                   - use the two new functions throughout ha_ndbcluster and ha_ndbinfo
                   - remove the 'close_cached_tables' backwards compatibilty function
            ------------------------------------------------------------
            revno: 2876.577.41 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-19 10:11:00 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.32 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Wed 2014-02-19 10:10:27 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.43
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-02-19 10:08:09 +0100
                    message:
                      Bug#18272631 NDBINFO SET OFFLINE UNNECESSARILY WAITS FOR ALL TABLES TO BE CLOSED
                       - don't wait for tables to be closed when switching into or out of offline mode.
            ------------------------------------------------------------
            revno: 2876.577.40 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Tue 2014-02-18 20:56:49 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.31 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Tue 2014-02-18 20:56:09 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.42
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2014-02-18 20:55:16 +0100
                    message:
                      ------------------------------------------------------------
                      revno: 4224
                      revision-id: mauritz.sundell@oracle.com-20140218152134-5dy6o9jhozcm95zj
                      parent: mauritz.sundell@oracle.com-20140217100854-0mgr9qzn8wdsf5hh
                      committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                      branch nick: mysql-7.4
                      timestamp: Tue 2014-02-18 16:21:34 +0100
                      message:
                        ndb - add support for cmake -DWITH_NDBMTD=0 again
                        
                        Bug #18267919	 CMAKE -DWITH_NDBMTD=0 ... HAVE NO EFFECT
                        
                        Do not build ndbmtd, ndbsched_mt, mt-send-t if WITH_NDBMTD=0.
            ------------------------------------------------------------
            revno: 2876.577.39
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: push_wl7532
            timestamp: Tue 2014-02-18 17:26:14 +0100
            message:
              WL#7532: Scan optimisations, asynch signals -> synch signals, use of local variables to assist compiler, improved checksum algorithm, prefetch of fixed part of rows, many new extensive comments, execute up to 3-4 rows before sending asycnh signal in LQH scanning, removal of lots of dead code, minor jam improvements, turned bit variables into 32-bit variables for higher speed
            ------------------------------------------------------------
            revno: 2876.577.38 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-02-17 11:08:54 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.30
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-02-17 11:03:41 +0100
                message:
                  revision-id: nisha.gopalakrishnan@oracle.com-20131114050427-535pjj7ek97qhss4
                  parent: shaohua.wang@oracle.com-20131114021621-wb2n9etjhxx3tvop
                  committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
                  branch nick: mysql-5.6-17246318
                  timestamp: Thu 2013-11-14 10:34:27 +0530
                  message:
                  ndb - test: re-recording results
                  
                  Re-recording results for ndb_native_default_support and
                  ndb_dd_restore_compat due to below change in server.
                  
                    BUG#17246318: ALTER TABLE SHOULD NOT ALLOW CREATION OF TABLES
                                  WITH BOTH 5.5 AND 5.6 TEMPORALS
                    
                    Analysis
                    --------
                    
                    'ALTER TABLE' allows creation of tables containing temporal
                    columns of both mysql-5.5 and mysql-5.6 format.
                    
                    Recreating tables having both 5.5 and 5.6 temporals becomes
                    tedious process when the metadata file(.frm) for the table
                    is unavailable. This would involve recreating the table having
                    the old temporal columns using 5.5 server instance and then
                    adding the 5.6 temporal columns using 5.6 server instance.
                    Also recreating tables having 5.5 temporals and 5.6 INNODB
                    FULLTEXT indexes has to follow the same process.
                    
                    Currently operations are supported on the old temporal columns.
                    The old temporal columns are upgraded to the new format only
                    when they are altered(like renaming the old temporal column)
                    else they are retained in the old format.
                    
                    In order to overcome the above mentioned tedious process
                    and use the space efficient new temporal format, the old
                    temporal types are upgraded for certain ALTER TABLE
                    operations listed in the fix info.
                    
                    Fix info:
                    ---------
                    
                    The columns of old temporal types of mysql-5.5 are upgraded to
                    mysql-5.6 format when ALTER TABLE requests ADD/CHANGE/MODIFY
                    COLUMN, ADD INDEX or FORCE operation.
                    
                    Since such conversion cannot be done using INPLACE algorithm,
                    the attempt to use ALGORITHM=INPLACE clause in such ALTER TABLE
                    for a table with mysql-5.5 temporals leads to an error.
                    
                    Also a 'NOTE' is reported to indicate the upgrade of the
                    old temporal columns to the new format under the above
                    condition.
            ------------------------------------------------------------
            revno: 2876.577.37 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Mon 2014-02-17 02:55:15 +0100
            message:
              merge 5.6.16 via 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.29 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-02-17 02:48:36 +0100
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.41 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Mon 2014-02-17 02:18:35 +0100
                    message:
                      merge 5.5.36
                ------------------------------------------------------------
                revno: 2876.579.28 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Mon 2014-02-17 02:44:45 +0100
                message:
                  merge 5.6.16
            ------------------------------------------------------------
            revno: 2876.577.36 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17916243-74
            timestamp: Sun 2014-02-16 15:21:41 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.27 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17916243-73
                timestamp: Sun 2014-02-16 15:05:56 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.40 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug17916243-72
                    timestamp: Sun 2014-02-16 14:45:41 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.817.164
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17916243-71
                        timestamp: Sun 2014-02-16 14:29:13 +0200
                        message:
                          bug#17916243 fix1.diff
                          dictSignal does not retry Busy if no errCodes
                          (there are other similar cases)
                        ------------------------------------------------------------
                        revno: 2555.817.163
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17916243-71
                        timestamp: Sun 2014-02-16 14:28:46 +0200
                        message:
                          bug#17916243 test2.diff
                          GET_TABINFOREQ test (errins enabled, fails)
                        ------------------------------------------------------------
                        revno: 2555.817.162
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17916243-71
                        timestamp: Sun 2014-02-16 14:28:07 +0200
                        message:
                          bug#17916243 test1.diff
                          GET_TABINFOREQ test (errins disabled, succeeds)
            ------------------------------------------------------------
            revno: 2876.577.35 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-02-12 18:50:29 -0800
            message:
              merge 7.3 -> 7.4 clusterj-tie/pom.xml.in
                ------------------------------------------------------------
                revno: 2876.579.26 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.6-cluster-7.3-merge
                timestamp: Wed 2014-02-12 18:43:46 -0800
                message:
                  merge 7.2 -> 7.3 clusterj-tie/pom.xml.in
                    ------------------------------------------------------------
                    revno: 2875.612.39 [merge]
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-02-12 18:36:12 -0800
                    message:
                      merge 7.1 -> 7.2 clusterj-tie/pom.xml.in
                        ------------------------------------------------------------
                        revno: 2555.817.161
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-02-12 18:31:51 -0800
                        message:
                          Remove bad check for ndbclient in clusterj-tie pom.xml
        ------------------------------------------------------------
        revno: 2876.563.52 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_7.4_2_trunk-cluster
        timestamp: Tue 2014-04-15 14:01:53 +0200
        message:
          Merge 7.4(-r4219) -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.577.34
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-12 09:47:47 +0100
            message:
              Remove ha_ndbcluster_push dependency on ndbcluster_hton
              
               - circular, ndbcluster_hton is the singleton instance
            ------------------------------------------------------------
            revno: 2876.577.33
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-12 09:45:49 +0100
            message:
              Remove missleading code pretending to handle states which never occurs
              
               - Neither isOverrun() or isConsistent() is implemented, they always return OK
               - The "generic printout" continue to be used
            ------------------------------------------------------------
            revno: 2876.577.32
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-12 09:31:42 +0100
            message:
              Move comment for "binlog main loop" to correct places
            ------------------------------------------------------------
            revno: 2876.577.31
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-12 09:29:13 +0100
            message:
              Remove unnecessary extra trailing semicolon
            ------------------------------------------------------------
            revno: 2876.577.30 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-02-11 15:57:14 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.25 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-02-11 15:54:50 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.38 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-02-11 15:52:22 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.160
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-02-11 15:50:40 +0100
                        message:
                          Fix for Bug#18229003 
                          
                            'MT-SCHEDULER: BLOCK THREADS MAY GET STUCK IN A MUTUAL WAIT-LOCK'
                          
                          The block-threads managed by the mt-scheduler, communicates by putting
                          its signals in a queue (also known as the job-buffer) which
                          is set up between all block-threads.
                          
                          This queue has a fixed max size, such that when the out-queue
                          is filled up, the worker thread has to wait for the consumer
                          to drain the queue.
                          
                          We have observed that in a highly loaded system, multiple
                          threads may end up in a circular wait-lock due to full out-buffers.
                          They are then mutually blocking each other from doing any useful
                          work, and the datanode will eventually be killed by a wathdog
                          timer and declared dead.
                          
                          This fix detects the situations where we are about to run into
                          a situation with circular wait lock. It will then give a
                          higher quota of the signal processing to the queues which
                          are highly loaded.
            ------------------------------------------------------------
            revno: 2876.577.29 [merge]
            committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-02-10 19:36:04 +0100
            message:
              Null merge 7.3-7.4
                ------------------------------------------------------------
                revno: 2876.579.24 [merge]
                committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-02-10 19:30:22 +0100
                message:
                  Null merge 7.2-7.3
                    ------------------------------------------------------------
                    revno: 2875.612.37 [merge]
                    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-02-10 19:20:35 +0100
                    message:
                      Null merging changes from 7.1
            ------------------------------------------------------------
            revno: 2876.577.28 [merge]
            committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-02-10 17:28:37 +0100
            message:
              Null merge from 7.3 - 7.4
                ------------------------------------------------------------
                revno: 2876.579.23 [merge]
                committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-02-10 17:16:14 +0100
                message:
                  Null merge from 7.2 - 7.3
                    ------------------------------------------------------------
                    revno: 2875.617.1 [merge]
                    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-02-10 17:05:13 +0100
                    message:
                      Null merge from 7.1 - 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.159 [merge]
                        author: murthy.narkedimilli@oracle.com
                        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2014-02-10 16:50:58 +0100
                        message:
                          Merge from mysql-cluster-7.1.30-release
                            ------------------------------------------------------------
                            revno: 2555.841.1
                            tags: mysql-cluster-7.1.30
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-cluster-7.1.30-release
                            timestamp: Thu 2014-01-23 14:28:57 +0100
                            message:
                              ndb - cherrypick: disable optimization of Ndbfs::readWriteRequest only for gcc 4.8 and up
                              
                              revision-id: mauritz.sundell@oracle.com-20140123120237-l5p6o1m3kg8apvu3
                              parent: mauritz.sundell@oracle.com-20140115081453-6cglx3zmar7ovdin
                              committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                              branch nick: mysql-6.3
                              timestamp: Thu 2014-01-23 13:02:37 +0100
                              message:
                                ndb - disable optimization of Ndbfs::readWriteRequest only for gcc 4.8 and up
                              
                                Bug #18122881  REVERT FIX FOR BUG #18055285 FOR LOWER VERSIONS (<4.8) OF GCC
                              
                                GCC_VERSION was not always defined before include of Ndbfs.cpp as in VoidFs.cpp
                                so also change to test __GNUC__ and __GNUC_MINOR__ that are predefined for gnu
                                compiler.
                              
                                note that for non gnu compiler __GNUC__ and __GNUC_MINOR__ will be treated as 0 in arithmetics
                                and so optimization not disabled.
            ------------------------------------------------------------
            revno: 2876.577.27 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17232212-74
            timestamp: Sun 2014-02-09 15:14:12 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.22
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17232212-73
                timestamp: Sun 2014-02-09 15:06:58 +0200
                message:
                  bug#17232212 fixsql2-x1.diff
                  changes after review
                ------------------------------------------------------------
                revno: 2876.579.21
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17232212-73
                timestamp: Sun 2014-02-09 15:06:11 +0200
                message:
                  bug#17232212 fixsql2.diff
                  sql level fixes
                ------------------------------------------------------------
                revno: 2876.579.20
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17232212-73
                timestamp: Sun 2014-02-09 15:05:29 +0200
                message:
                  bug#17232212 testsql.diff
                  MTR tests showing bugs
            ------------------------------------------------------------
            revno: 2876.577.26 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Sun 2014-02-09 12:00:08 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.19 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.6-cluster-7.3-merge
                timestamp: Sat 2014-02-08 13:16:25 -0800
                message:
                  Merge 7.2 -> 7.3 clusterj-test/pom.xml.in
                    ------------------------------------------------------------
                    revno: 2875.612.36 [merge]
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Sat 2014-02-08 13:06:31 -0800
                    message:
                      Merge 7.1 -> 7.2 clusterj-test/pom.xml.in
                        ------------------------------------------------------------
                        revno: 2555.817.158
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Sat 2014-02-08 13:01:56 -0800
                        message:
                          Fix clusterj/clusterj-test pom.xml.in for proper mvn suffix
            ------------------------------------------------------------
            revno: 2876.577.25 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-main-74
            timestamp: Thu 2014-02-06 19:25:38 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.18 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-main-73
                timestamp: Thu 2014-02-06 19:22:47 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.35 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-main-72
                    timestamp: Thu 2014-02-06 19:20:54 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.817.157
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-main-71
                        timestamp: Thu 2014-02-06 19:18:17 +0200
                        message:
                          main x1.diff
                          fix DBUG assert in setting NDB_ALIVE_NODE_ID
                ------------------------------------------------------------
                revno: 2876.579.17 [merge]
                author: hery.ramilison@oracle.com
                committer: Hery Ramilison <hery.ramilison@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2014-02-06 11:42:06 +0100
                message:
                  Merge from mysql-cluster-7.3.4-release
                    ------------------------------------------------------------
                    revno: 2876.582.1
                    tags: mysql-cluster-7.3.4
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-cluster-7.3.4-release
                    timestamp: Thu 2014-01-23 14:10:57 +0100
                    message:
                      ndb - cherrypick: disable optimization of Ndbfs::readWriteRequest only for gcc 4.8 and up
                      
                      revision-id: mauritz.sundell@oracle.com-20140123120237-l5p6o1m3kg8apvu3
                      parent: mauritz.sundell@oracle.com-20140115081453-6cglx3zmar7ovdin
                      committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                      branch nick: mysql-6.3
                      timestamp: Thu 2014-01-23 13:02:37 +0100
                      message:
                        ndb - disable optimization of Ndbfs::readWriteRequest only for gcc 4.8 and up
                         
                        Bug #18122881  REVERT FIX FOR BUG #18055285 FOR LOWER VERSIONS (<4.8) OF GCC
                      
                        GCC_VERSION was not always defined before include of Ndbfs.cpp as in VoidFs.cpp
                        so also change to test __GNUC__ and __GNUC_MINOR__ that are predefined for gnu
                        compiler.
                                
                        note that for non gnu compiler __GNUC__ and __GNUC_MINOR__ will be treated as 0 in arithmetics
                        and so optimization not disabled.
            ------------------------------------------------------------
            revno: 2876.577.24 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-wl6244-74
            timestamp: Wed 2014-02-05 23:57:22 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.16
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-wl6244-73
                timestamp: Wed 2014-02-05 23:41:44 +0200
                message:
                  wl#6244 errins2.diff
                  move old FK error inserts 8098,8099 to 8105,8106
                ------------------------------------------------------------
                revno: 2876.579.15
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-wl6244-73
                timestamp: Wed 2014-02-05 23:41:12 +0200
                message:
                  wl#6244 errins.diff
                  move old FK error inserts 8100,8101 to 8103,8104
            ------------------------------------------------------------
            revno: 2876.577.23 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-05 11:48:49 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.14 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Wed 2014-02-05 11:48:04 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.34 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-02-05 11:46:26 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.156 [merge]
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Wed 2014-02-05 11:38:17 +0100
                        message:
                          Merge
                    ------------------------------------------------------------
                    revno: 2875.612.33 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Wed 2014-02-05 11:39:34 +0100
                    message:
                      Merge
                ------------------------------------------------------------
                revno: 2876.579.13 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Wed 2014-02-05 11:40:50 +0100
                message:
                  Merge
            ------------------------------------------------------------
            revno: 2876.577.22 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Wed 2014-02-05 11:42:15 +0100
            message:
              Merge
                ------------------------------------------------------------
                revno: 2876.581.2 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4
                timestamp: Wed 2014-02-05 10:47:09 +0100
                message:
                  Merge 7.3 -> 7.4
                    ------------------------------------------------------------
                    revno: 2876.580.2 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.3
                    timestamp: Wed 2014-02-05 10:45:58 +0100
                    message:
                      Merge 7.2 -> 7.3
                        ------------------------------------------------------------
                        revno: 2875.616.2 [merge]
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Wed 2014-02-05 10:44:49 +0100
                        message:
                          Merge 7.1 -> 7.2
                            ------------------------------------------------------------
                            revno: 2555.840.1 [merge]
                            committer: magnus.blaudd@oracle.com
                            branch nick: 7.1
                            timestamp: Wed 2014-02-05 10:39:37 +0100
                            message:
                              Merge
                                ------------------------------------------------------------
                                revno: 2555.839.2
                                committer: magnus.blaudd@oracle.com
                                branch nick: 7.1-bug18165088
                                timestamp: Wed 2014-02-05 10:37:33 +0100
                                message:
                                  Mark runTestCreateLogEvent() as static
                                ------------------------------------------------------------
                                revno: 2555.839.1
                                committer: magnus.blaudd@oracle.com
                                branch nick: 7.1-bug18165088
                                timestamp: Tue 2014-02-04 10:03:50 +0100
                                message:
                                  Bug#18165088 FAILURE TO SETUP EVENT LISTENER DETECTED BY NDB_RESTORE_COMPAT_COMPRESSION
                                  
                                  Problem:
                                  - The ndb_mgmd processes the  "listen event" command, writes the reply "listen
                                  event" to m_output and then adds the new listener to the list of active
                                  listeners. But since m_output is buffered(using BufferedSockOutputStream)
                                  there is a probability that check_socket_listener() is called before the
                                  output is flushed. This caused the client to receive the string "<PING>" instead
                                  of the expected "listen event".
                                  - Actually it's also possible for a new log event (from Ndb_mgmd_event_service::log() ) to
                                    be written to the socket while pending reply was in the output buffer.
                                  
                                  Solution:
                                  - Flush the reply from buffer before adding the new event listener to the list
                                  of active listeners.
                                  - Unlock m_mutex while flushing to avoid holding the mutex while doing
                                  network IO. This is smilar to the other place where m_oputput is flushed and
                                  thus the same logic should apply here.
                ------------------------------------------------------------
                revno: 2876.581.1 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.4
                timestamp: Wed 2014-02-05 09:51:43 +0100
                message:
                  Merge 7.3 -> 7.4
                    ------------------------------------------------------------
                    revno: 2876.580.1 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.3
                    timestamp: Wed 2014-02-05 09:50:22 +0100
                    message:
                      Merge 7.2 -> 7.3
                        ------------------------------------------------------------
                        revno: 2875.616.1 [merge]
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Wed 2014-02-05 09:12:22 +0100
                        message:
                          Merge
                            ------------------------------------------------------------
                            revno: 2875.615.1 [merge]
                            committer: magnus.blaudd@oracle.com
                            branch nick: 7.2
                            timestamp: Tue 2014-02-04 13:49:33 +0100
                            message:
                              Bug#18096866 NDBAPI-EXAMPLES/NDBAPI_BLOB_NDBRECORD #INCLUDE <NDB_GLOBAL.H>
                               - merge in patch from Gustaf
                                ------------------------------------------------------------
                                revno: 2875.614.1
                                committer: Gustaf Thorslund <gustaf.thorslund@oracle.com>
                                branch nick: mysql-5.5-cluster-7.2-bug71409
                                timestamp: Fri 2014-01-17 16:11:25 +0100
                                message:
                                  Fix for Bug #71409 - ndbapi-examples/ndbapi_blob_ndbrecord #include <ndb_global.h>
            ------------------------------------------------------------
            revno: 2876.577.21 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Wed 2014-02-05 09:10:24 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.12 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2014-02-05 09:07:34 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.32 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2014-02-05 09:04:04 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.155
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2014-02-05 09:02:01 +0100
                        message:
                          Fix for Bug#18181656 	PSET_DESTROY() CALLED WITH INCORRECT ARGUMENT
                          
                          When compiled for Solaris, pset_destroy() was called with
                          a 'psetid_t*' argument, instead of a 'psetid_t'.
                          
                          This patch also removes some obsolete typecasts to 'void*'
                          and 'char*' when calling free(void*) and NdbMem_Free(void*)
            ------------------------------------------------------------
            revno: 2876.577.20 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-02-04 20:59:10 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.11 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-02-04 20:52:47 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.31 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-02-04 20:43:49 +0100
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.154
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-02-04 20:38:09 +0100
                        message:
                          ndb - make test ndb_big.ndb_big_addnode deterministic
                          
                          Bug #17749286	 TEST CASE NDB_BIG.NDB_BIG_ADDNODE IS FAILING IN CLUB MYSQL-5.5-CLUSTER-7.2
            ------------------------------------------------------------
            revno: 2876.577.19 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18069680-74
            timestamp: Tue 2014-02-04 18:12:30 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.579.10
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18069680-73
                timestamp: Tue 2014-02-04 18:03:13 +0200
                message:
                  bug#18069680 fixdrop.diff
                  dropTable() drop FKs avoiding error 21081/21082
                ------------------------------------------------------------
                revno: 2876.579.9
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug18069680-73
                timestamp: Tue 2014-02-04 18:02:12 +0200
                message:
                  bug#18069680 testdrop.diff
                  test dropTable() drops FKs (as child or parent)
            ------------------------------------------------------------
            revno: 2876.577.18 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-02-04 12:30:36 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.8 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-02-04 12:27:02 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.30 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-02-04 12:17:37 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.153
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-02-04 12:14:49 +0100
                        message:
                          Addendum patch for bug#18068206 COMPILER WARNING: "ERR" REDEFINED
            ------------------------------------------------------------
            revno: 2876.577.17 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-02-04 10:24:26 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.7 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-02-04 10:21:25 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.29 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2014-02-04 10:18:38 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.152
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Tue 2014-02-04 10:15:57 +0100
                        message:
                          Fix for bug#18068206 	COMPILER WARNING: "ERR" REDEFINED
                                
                          On Solaris platforms the definition of 'ERR' in NDBT_Error.hpp
                          redefined 'ERR' which was defined in the include file 
                          /usr/include/sys/regset.h
                                
                          This patch rename *ERR to NDB_*ERR to remove that name clash.
                                
                          NOTE: No functional changes, only a pure rename.
            ------------------------------------------------------------
            revno: 2876.577.16 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Tue 2014-02-04 09:39:56 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.6
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-02-04 09:37:51 +0100
                message:
                  Cherry picked fix for bug#18164798 to mysql-5.6-cluster-7.3
                        
                        INFORMATION_SCHEMA-BIG FAILS WHEN COMPILED WITH NDB
            ------------------------------------------------------------
            revno: 2876.577.15 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.4
            timestamp: Mon 2014-02-03 16:01:51 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.5 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2014-02-03 15:59:29 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.28
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2014-02-03 15:57:06 +0100
                    message:
                      Fix for Bug#18175080 INCORRECT 'LOCAL_RANGE_SCAN' REPORTED FROM NDB_JOIN_PUSHDOWN_*.TEST
                      
                      Remove nondetermenistic use of 'LOCAL_RANGE_SCANS_SENT' ndbinfo.counter
            ------------------------------------------------------------
            revno: 2876.577.14
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Mon 2014-02-03 13:17:18 +0100
            message:
              Also bump tree name to 5.6-cluster-7.4
            ------------------------------------------------------------
            revno: 2876.577.13 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Fri 2014-01-31 14:08:11 +0100
            message:
              merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.4
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Fri 2014-01-31 14:04:03 +0100
                message:
                  ndb - mark unstable test innodb_bug14712710 experimental
                  
                  Bug #18164719	 TEST I_INNODB.INNODB_BUG14712710 UNSTABLE IN PB2 FOR 5.6 
        ------------------------------------------------------------
        revno: 2876.563.51
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2014-04-14 15:30:56 +0200
        message:
          WL#6815   Adapt MySQL Cluster to 5.7
          
           - fix compilation error caused by 'thd_query_string' return
             type change.
        ------------------------------------------------------------
        revno: 2876.563.50 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2014-04-14 15:03:50 +0200
        message:
          Merge 5.7.4 -> trunk-cluster
        ------------------------------------------------------------
        revno: 2876.563.49 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2014-01-31 08:09:53 +0100
        message:
          Merge 7.4 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.577.12 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-01-30 18:04:20 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.3 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2014-01-30 18:03:51 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.27 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Thu 2014-01-30 18:02:02 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.151 [merge]
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Thu 2014-01-30 18:00:19 +0100
                        message:
                          Merge in rep_latency patch to 7.1
            ------------------------------------------------------------
            revno: 2876.577.11 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.4
            timestamp: Thu 2014-01-30 14:52:12 +0100
            message:
              Merge 7.3 -> 7.4
                ------------------------------------------------------------
                revno: 2876.579.2 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2014-01-30 14:51:41 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.612.26 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Thu 2014-01-30 14:50:56 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.838.1
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.1
                        timestamp: Thu 2014-01-30 14:41:56 +0100
                        message:
                          Bug#11755732 REP_LATENCY: USES NDBOUT BEFORE CALLING NDB_INIT()
                                 - Move ndb_init() first in program
                                 - Note! this is an internal test program
                ------------------------------------------------------------
                revno: 2876.579.1 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Tue 2014-01-28 14:09:39 +0100
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.612.25
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: 7.2.16-bugs
                    timestamp: Tue 2014-01-28 13:54:20 +0100
                    message:
                      Restore lost fix for memcache compiler warnings 
            ------------------------------------------------------------
            revno: 2876.577.10
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.4
            timestamp: Fri 2014-01-24 12:14:15 +0100
            message:
              initial version 7.4.0
            ------------------------------------------------------------
            revno: 2876.577.9 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2014-01-23 13:56:44 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.24 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2014-01-23 13:54:48 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.150 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Thu 2014-01-23 13:46:42 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.72 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Thu 2014-01-23 13:45:45 +0100
                        message:
                          merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.836.10
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Thu 2014-01-23 13:02:37 +0100
                            message:
                              ndb - disable optimization of Ndbfs::readWriteRequest only for gcc 4.8 and up
                              
                              Bug #18122881	 REVERT FIX FOR BUG #18055285 FOR LOWER VERSIONS (<4.8) OF GCC
                              
                              GCC_VERSION was not always defined before include of Ndbfs.cpp as in VoidFs.cpp
                              so also change to test __GNUC__ and __GNUC_MINOR__ that are predefined for gnu
                              compiler.
                              
                              note that for non gnu compiler __GNUC__ and __GNUC_MINOR__ will be treated as 0 in arithmetics
                              and so optimization not disabled.
            ------------------------------------------------------------
            revno: 2876.577.8 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2014-01-20 13:49:25 +0100
            message:
              Merge 7.3-bug18094360 -> 7.3
                ------------------------------------------------------------
                revno: 2876.578.4
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3-bug18094360
                timestamp: Fri 2014-01-17 15:23:46 +0100
                message:
                  Bug#18094360 COLUMN ORDER IN INDEX NOT CHECKED WHEN CREATING FK
                   - The column order checks are correct already but the error message returned differ
                    comapred to the legacy engine.
                   - Change the error message to be same as legacy engine.
                  - Note that ndbcluster allow creating foregin keys referencing unique indexes
                    in parent table with different column order than the child table. This is ok
                    as long as the unique index covers all columns wherea there will only be one
                    or zero rows regardless of column order.
                   - Add test showing that legacy and ndb have the same error code when non matching
                    parent are found
                ------------------------------------------------------------
                revno: 2876.578.3
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3-bug18094360
                timestamp: Fri 2014-01-17 15:11:46 +0100
                message:
                  ndb fk - add comment explaining that column order for unique index does not matter
                ------------------------------------------------------------
                revno: 2876.578.2
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3-bug18094360
                timestamp: Fri 2014-01-17 15:08:17 +0100
                message:
                  ndb fk cleanup - remove unnecessary else after return
                ------------------------------------------------------------
                revno: 2876.578.1
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3-bug18094360
                timestamp: Fri 2014-01-17 15:06:03 +0100
                message:
                  ndb fk cleanup
                   - Fold unnecessary goto into the only place where it's used.
            ------------------------------------------------------------
            revno: 2876.577.7 [merge]
            committer: John David Duncan <john.duncan@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2014-01-17 10:33:39 -0800
            message:
              merge
                ------------------------------------------------------------
                revno: 0.82.197
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2014-01-17 10:31:40 -0800
                message:
                  Fix bug that prevented clean async shutdown 
            ------------------------------------------------------------
            revno: 2876.577.6 [merge]
            committer: John David Duncan <john.duncan@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2014-01-16 22:25:52 -0800
            message:
              merge latest nodejs-adapter into cluster 7.3
                ------------------------------------------------------------
                revno: 0.82.196 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2014-01-16 22:22:19 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.101.1 [merge]
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Thu 2014-01-16 19:18:11 -0800
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 0.100.2
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: nodejs-adapter
                        timestamp: Tue 2014-01-14 19:11:24 -0800
                        message:
                          Improve error reporting for DefaultValuesTest
                        ------------------------------------------------------------
                        revno: 0.100.1
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: nodejs-adapter
                        timestamp: Tue 2014-01-14 18:59:23 -0800
                        message:
                          Improve error reporting for default values
                ------------------------------------------------------------
                revno: 0.82.195
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2014-01-16 22:20:10 -0800
                message:
                  bug#18080250 : nodejs adapter build fails in 7.4
                ------------------------------------------------------------
                revno: 0.82.194
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2014-01-16 21:50:33 -0800
                message:
                  fix preprocessor conditional
                ------------------------------------------------------------
                revno: 0.82.193
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2014-01-16 19:09:02 -0800
                message:
                  Refactor shutdown_flag in AsyncNdbContext to allow possibly more efficient implementations
                ------------------------------------------------------------
                revno: 0.82.192
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2014-01-14 15:13:40 -0800
                message:
                  Add is_xxx functions to debug logger; guard Promise.fulfill with is_detail
                  
                  unified_debug.js
                    add functions to allow guard code, e.g. udebug.is_detail()
                      is_urgent
                      is_notice
                      is_info
                      is_debug
                      is_detail 
                  
                  UserContext.js
                    Promise.fulfill guard creation of new Error object with is_detail
                ------------------------------------------------------------
                revno: 0.82.191
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2014-01-01 14:10:49 -0800
                message:
                  Bug 16980227 Autoincrement values ignored on insert
                  
                  DBTableHandler.js
                    add field autoIncFieldName to table handler
                    use autoIncFieldName for setAutoincrement method
                  
                  NdbAutoIncrement.js
                    check needAutoInc flag in operation to set autoincrement functionality
                  
                  NdbOperation.js
                    for insert operation, set needAutoInc if value for autoincrement column 
                      is not set in domain object
                ------------------------------------------------------------
                revno: 0.82.190 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-30 13:59:07 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.99.2
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-12-24 11:34:22 -0800
                    message:
                      Update autopk tests to detect adapter ignoring primary key field
                      
                      test/autoincrement/lib.js:
                        add parameter to test whether to verify id field
                      
                      test/autoincrement/AutoPKInsertTest.js
                        update key values for rows to insert
                        add duplicate primary key test
                    ------------------------------------------------------------
                    revno: 0.99.1
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Fri 2013-12-20 15:04:27 -0800
                    message:
                      Update Transaction documentation.
                ------------------------------------------------------------
                revno: 0.82.189
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-12-17 21:43:07 -0800
                message:
                  fix case-sensitive file name 
                ------------------------------------------------------------
                revno: 0.82.188
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-12-17 19:05:21 -0800
                message:
                  Only specialize over primitive types
                ------------------------------------------------------------
                revno: 0.82.187
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-12-17 11:12:59 -0800
                message:
                  Add UTF-8 Byte Order Mark at start of test/mysql56types/CharsetTest.js
                ------------------------------------------------------------
                revno: 0.82.186
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-16 23:06:59 -0800
                message:
                  Improve error handling when NDB connection fails
                ------------------------------------------------------------
                revno: 0.82.185
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-16 22:18:16 -0800
                message:
                  Shorten too-long strings in CharsetTest
                ------------------------------------------------------------
                revno: 0.82.184
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-16 20:44:05 -0800
                message:
                  Revise all docs to include promises.
                ------------------------------------------------------------
                revno: 0.82.183
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-16 18:56:18 -0800
                message:
                  Remove documentation file for "Context" and copy all of Context's docs
                  into both Session and Batch.
                ------------------------------------------------------------
                revno: 0.82.182
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-16 16:34:51 -0800
                message:
                  lint fixes
                ------------------------------------------------------------
                revno: 0.82.181 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-16 16:28:30 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.98.2
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Mon 2013-12-16 16:27:32 -0800
                    message:
                      Return true from synchronous tests
                    ------------------------------------------------------------
                    revno: 0.98.1
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Mon 2013-12-16 16:21:36 -0800
                    message:
                      Remove ConcurrentSubTest support from test driver and refactor the tests that used it.
                ------------------------------------------------------------
                revno: 0.82.180
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-12-12 13:14:57 -0800
                message:
                  Fix comment on MySQLConnectionPool; fix error return from new connection
                ------------------------------------------------------------
                revno: 0.82.179
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-12-12 13:06:46 -0800
                message:
                  Change default charset for connections to UTF8MB4
                  
                  MySQLConnectionPool.js
                    allow user to specify mysql_charset in connection properties
                    send the settings on each new connection to the database
                      SET character_set_client
                      SET character_set_connection
                      SET character_set_results
                  
                  mysql_service_provider.js
                    set the default to UTF8MB4 in getConnectionProperties
                ------------------------------------------------------------
                revno: 0.82.178
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-12-11 16:41:55 -0800
                message:
                  CharsetTest.js fail instead of hang on insert error
                ------------------------------------------------------------
                revno: 0.82.177
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-12-11 15:36:23 -0800
                message:
                  Improve user documentation for Errors
                  
                  API-documentation/Errors
                    Error must contain message, stack, sqlstate
                    Error may contain cause
                  
                  MySQLConnectionPool.js
                    Create a new Error object with a new message and stack for connection failures
                    Include the sqlstate and cause (underlying connection err)
                ------------------------------------------------------------
                revno: 0.82.176
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-12-11 15:01:45 -0800
                message:
                  Improve promises testing and error reporting
                  
                  Transaction.js
                    add sqlstate 25000 to transaction state errors
                  
                  MySQLConnectionPool.js
                    add sqlstate 08000 to connection failure errors
                  
                  PromisesTest.js
                    add checks that err.message and err.sqlstate are strings
                  
                  harness.js
                    improve comparisons to accommodate null and undefined arguments
                ------------------------------------------------------------
                revno: 0.82.175
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-12-03 22:34:22 -0800
                message:
                  Promises implementation
                ------------------------------------------------------------
                revno: 0.82.174
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-12-02 12:58:01 -0800
                message:
                  tweet.js: all operations now work; some use promises, some don't
                ------------------------------------------------------------
                revno: 0.82.173
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-30 19:01:58 -0800
                message:
                  work in progress: use promises in tweet.js
                ------------------------------------------------------------
                revno: 0.82.172
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2013-11-29 22:52:06 -0800
                message:
                  More transaction promises
                ------------------------------------------------------------
                revno: 0.82.171 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2013-11-29 12:43:28 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.97.2
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Fri 2013-11-29 08:51:02 -0800
                    message:
                      Add test for transaction commit promise
                ------------------------------------------------------------
                revno: 0.82.170 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2013-11-29 08:08:01 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.97.1 [merge]
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Fri 2013-11-29 08:00:11 -0800
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 0.96.5
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: nodejs-adapter
                        timestamp: Thu 2013-11-28 20:26:42 -0800
                        message:
                          Remove debugging from t_basic/PromisesTest
                        ------------------------------------------------------------
                        revno: 0.96.4
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: nodejs-adapter
                        timestamp: Thu 2013-11-28 20:12:50 -0800
                        message:
                          Add PromisesTest to t_basic
                        ------------------------------------------------------------
                        revno: 0.96.3
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: nodejs-adapter
                        timestamp: Wed 2013-11-27 23:14:43 -0800
                        message:
                          A few Promise cleanups from code review
                        ------------------------------------------------------------
                        revno: 0.96.2
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: nodejs-adapter
                        timestamp: Wed 2013-11-27 18:34:11 -0800
                        message:
                          Fix IllegalArgumentTest for promises
                ------------------------------------------------------------
                revno: 0.82.169
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2013-11-29 08:02:21 -0800
                message:
                  work in progress: use promises in tweet demo
                ------------------------------------------------------------
                revno: 0.82.168
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-27 17:56:04 -0800
                message:
                  Return promise from createQuery(), executeQuery(), and executeBatch()
                ------------------------------------------------------------
                revno: 0.82.167
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-27 17:09:22 -0800
                message:
                  api/Batch.js: return promise from UserContext
                ------------------------------------------------------------
                revno: 0.82.166
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-27 16:53:50 -0800
                message:
                  Return promises from api/Transaction
                  Active.commit(), Active.rollback(), and RollbackOnly.rollback() return the promise created by UserContext.
                  Other transactions don't.  I'm not sure this is the correct behavior. 
                ------------------------------------------------------------
                revno: 0.82.165 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-27 16:15:28 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.96.1
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Wed 2013-11-27 12:51:29 -0800
                    message:
                      Implement Promises/A+ Promise Resolution Procedure
                ------------------------------------------------------------
                revno: 0.82.164
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-27 16:14:54 -0800
                message:
                  tweet demo: work in progress to use promises
                ------------------------------------------------------------
                revno: 0.82.163 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-11-26 10:08:10 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.95.1
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter-npm
                    timestamp: Mon 2013-11-25 10:04:20 -0800
                    message:
                      update npm package.json info
                ------------------------------------------------------------
                revno: 0.82.162
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-11-26 10:07:02 -0800
                message:
                  tweet: work in progress on web server mode
                ------------------------------------------------------------
                revno: 0.82.161
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-11-25 22:50:46 -0800
                message:
                  tweet.js: refactor using promises
                ------------------------------------------------------------
                revno: 0.82.160
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-11-25 21:59:49 -0800
                message:
                  SessionFactory: return promises
                ------------------------------------------------------------
                revno: 0.82.159
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-11-21 22:14:16 -0800
                message:
                  Change BitMask.js after review
                ------------------------------------------------------------
                revno: 0.82.158
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-11-21 21:33:48 -0800
                message:
                  More work on IndexBounds
                  Currently tests q2,q7,q9 fail in multipartkeys
                ------------------------------------------------------------
                revno: 0.82.157
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-11-21 14:16:17 -0800
                message:
                  Always close NdbScanOperation
                ------------------------------------------------------------
                revno: 0.82.156
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-11-21 12:58:39 -0800
                message:
                  Redesign of index scans
                  For each node of the predicate, we store two bitmasks: a usedColumnMask and an equalColumnMask.
                  Each bit in the mask represents one column of the table.
                  The usedColumnMask is set by all comparator nodes, and "bubbles up" through AND, OR, and NOT.
                  The equalColumnMask is only set by QueryEq and only bubbles up through AND.
                  A Primary Key or Unique Index is usable for a query if the index's own mask AND the top level predicate's equalColumnMask == the index's mask.
                  An ordered index is (minimally) usable if the top level usedColumnMask has the bit set for the first index column.
                ------------------------------------------------------------
                revno: 0.82.155
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-11-21 12:52:46 -0800
                message:
                  Change some scan tests to expect index scan rather than table scan.
                ------------------------------------------------------------
                revno: 0.82.154
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-11-21 12:49:59 -0800
                message:
                  Retry scans that fail with TimeoutExpired error.
                  Close the NdbScanOperation, then define & execute a new one.
                ------------------------------------------------------------
                revno: 0.82.153
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-11-18 09:30:08 -0800
                message:
                  Improve debugging messages in Query.js
                ------------------------------------------------------------
                revno: 0.82.152 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 22:40:27 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.92.3
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Sun 2013-11-17 13:03:45 -0800
                    message:
                      Improve promises
                      
                      PromisesTest.js implement test cases for: 
                      UserContext.js
                        2.2.7.1 onFulfilled returns a value
                        2.2.7.1 onRejected returns a value
                        2.2.7.2 onFulfilled throws an error
                        2.2.7.2 onRejected throws an error
                        2.2.7.3 missing onFulfilled
                        2.2.7.4 missing onRejected
                ------------------------------------------------------------
                revno: 0.82.151
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 22:33:41 -0800
                message:
                  Change the default value for use_ndb_async_api back to false, 
                  due to unexpected impact on single-thread latency measurements.
                ------------------------------------------------------------
                revno: 0.82.150
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 22:32:55 -0800
                message:
                  lint fixes
                ------------------------------------------------------------
                revno: 0.82.149
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 21:25:34 -0800
                message:
                  fix wscript
                ------------------------------------------------------------
                revno: 0.82.148 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 21:22:59 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.94.11
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Mon 2013-11-11 22:05:08 -0800
                    message:
                      jscrund: use setImmediate() for smaller stacks
                    ------------------------------------------------------------
                    revno: 0.94.10
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Mon 2013-11-11 19:26:08 -0800
                    message:
                      jscrund:  If node is run as "node --expose-gc jscrund ..." then a full GC will be performed between tests.
                    ------------------------------------------------------------
                    revno: 0.94.9
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Fri 2013-11-08 13:04:59 -0800
                    message:
                      Improve error handling in DBOperationHelper
                    ------------------------------------------------------------
                    revno: 0.94.8
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Thu 2013-11-07 13:15:41 -0800
                    message:
                      We no longer wrap NdbOperation for JavaScript.
                      DBOperationHelper now handles a list of pending operations,
                      and returns a single wrapped PendingOperationSet object.
                      Internally the PendingOperationSet has the NdbOperations, and
                      exposes a getOperationError() method on them.
                      We also no longer wrap this NdbError unless its code is non-zero.
                    ------------------------------------------------------------
                    revno: 0.94.7
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Thu 2013-11-07 13:12:30 -0800
                    message:
                      jscrund: accept e.g. "-i100,200,400,800" to run test with varying batch sizes
                    ------------------------------------------------------------
                    revno: 0.94.6
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-11-05 22:24:32 -0800
                    message:
                      Refactor DBOperationHelper, step 6
                      Move logic from JavaScript to C++
                    ------------------------------------------------------------
                    revno: 0.94.5
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-11-05 21:51:10 -0800
                    message:
                      Refactor DBOperationHelper, step 5
                      Move NdbTransaction back out of helperSpec 
                    ------------------------------------------------------------
                    revno: 0.94.4
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-11-05 21:31:48 -0800
                    message:
                      Refactor DBOperationHelper, step 4
                      Move loop over operationList into prepareOperations()
                    ------------------------------------------------------------
                    revno: 0.94.3
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-11-05 19:20:19 -0800
                    message:
                      Refactor DBOperationHelper, step 3
                    ------------------------------------------------------------
                    revno: 0.94.2
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-11-05 19:16:16 -0800
                    message:
                      Refactor DBOperationHelper, step 2
                    ------------------------------------------------------------
                    revno: 0.94.1
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-11-05 18:46:25 -0800
                    message:
                      Refactor DBOperationHelper API 
                      (First step towards bulk creation of NdbOperations)
                ------------------------------------------------------------
                revno: 0.82.147
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 21:20:29 -0800
                message:
                  Refactor IndexBoundVisitor.consolidate() to use a closure over allBounds
                ------------------------------------------------------------
                revno: 0.82.146
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 21:17:32 -0800
                message:
                  Some work on handling timeouts during index scans.
                ------------------------------------------------------------
                revno: 0.82.145
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 17:20:31 -0800
                message:
                  Compatibility: on older node.js, use process.nextTick() for setImmediate()
                ------------------------------------------------------------
                revno: 0.82.144
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 10:46:06 -0800
                message:
                  Revert to getIndexBounds() returning an array of IndexBounds
                  and look again at IsNull / IsNotNull handling.
                ------------------------------------------------------------
                revno: 0.82.143 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 08:57:38 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.93.1
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Sun 2013-11-17 08:53:02 -0800
                    message:
                      Nulls sort low for index scans; let CandidateIndex treat IsNull and IsNotNull as bounds.
                ------------------------------------------------------------
                revno: 0.82.142 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-17 08:23:05 -0800
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.92.2
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Sat 2013-11-16 21:52:36 -0800
                    message:
                      Complete initial promises implementation
                      
                      Session.js
                        return promise for getTableMetadata, listTables, getMapping
                        return promise for find, load, persist, remove, update, save, close
                      
                      UserContext.js
                        return promise for listTables, getTableMetadata
                    ------------------------------------------------------------
                    revno: 0.92.1
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Sat 2013-11-16 21:42:57 -0800
                    message:
                      Implement promises for mysql-js
                      
                      mynode.js
                        return promise from openSession and connect
                      
                      UserContext.js
                        implement Promise contract per http://promises-aplus.github.io/promises-spec/
                        create a Promise for all UserContext uses
                        return promise from openSession and connect
                        return promise from find, persist, save, update, load, remove, and getMapping
                        return promise from commit, rollback, closeSession
                        update applyCallback to fulfill or reject promise before calling user callbacks
                          promise callbacks are called on next timer tick after user callbacks
                        fix dbConnectionPoolCreated_callback to always call back multiple
                          requestors for the same connection properties
                        need to evaluate executeBatch
                      
                      PromisesTest.js
                        correct order of parameters for test.ErrorIfNotEqual
                        test for promise being an object with a then method
                      
                      These test case failures need to be evaluated
                      [FAIL] t_basic IllegalArgumentTest.js testPersistNoArgumentNoCallback 
                        t11 persist with no arguments must fail.
                      [FAIL] t_basic IllegalArgumentTest.js testRemoveNoArgumentNoCallback
                        t12 remove with no arguments must fail.
                      [FAIL] t_basic IllegalArgumentTest.js testSaveNoArgumentNoCallback
                        t13 save with no argument must fail.
                      [FAIL] t_basic IllegalArgumentTest.js testUpdateNoArgumentNoCallback
                        t14 update with no arguments must fail.
                      [FAIL] t_basic IllegalArgumentTest.js testFindNoArgumentNoCallback
                        t15 find with no arguments must fail.
                ------------------------------------------------------------
                revno: 0.82.141
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-16 11:32:44 -0800
                message:
                  IndexBounds
                  NDB now often passes integraltypes scan tests but intermittently fails
                ------------------------------------------------------------
                revno: 0.82.140
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2013-11-15 22:00:03 -0800
                message:
                  IndexBounds: work in progress.
                ------------------------------------------------------------
                revno: 0.82.139
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-13 17:59:51 -0800
                message:
                  minor syntax fixes
                ------------------------------------------------------------
                revno: 0.82.138
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-11-12 13:36:01 -0800
                message:
                  Minor fixes to PromisesTest
                ------------------------------------------------------------
                revno: 0.82.137
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-06 12:06:40 -0800
                message:
                  More promises tests.
                ------------------------------------------------------------
                revno: 0.82.136
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-11-06 11:20:59 -0800
                message:
                  First tests for Promise API
                ------------------------------------------------------------
                revno: 0.82.135
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-03 14:47:04 -0800
                message:
                  jscrund SQL adapter: use real batching.
                  This uses the multipleStatements feature of node-mysql to send an entire batch at one time.
                ------------------------------------------------------------
                revno: 0.82.134
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sun 2013-11-03 14:32:06 -0800
                message:
                  JSCrund null backend: measure time to create buffers
                ------------------------------------------------------------
                revno: 0.82.133
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-02 20:13:02 -0700
                message:
                  JSCrund null adapter: measure system calls & NDBAPI calls
                ------------------------------------------------------------
                revno: 0.82.132
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-02 18:08:59 -0700
                message:
                  Change so UserContext constructor does not refer to arguments array.
                  According to http://s3.mrale.ph/nodecamp.eu/#57 matrializing arguments is expensive. 
                ------------------------------------------------------------
                revno: 0.82.131
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-02 15:42:38 -0700
                message:
                  In jscrund null adapter, add option to measure performance of udebug.log() messages
                ------------------------------------------------------------
                revno: 0.82.130
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-02 15:16:50 -0700
                message:
                  jscrund: organize handling of properties and allow properties from the command line
                ------------------------------------------------------------
                revno: 0.82.129
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Sat 2013-11-02 13:34:06 -0700
                message:
                  Move responsibility for default connection properties from the jscrund driver to the jscrund backends.
                ------------------------------------------------------------
                revno: 0.82.128
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-10-30 22:23:28 -0700
                message:
                  New "null" adapter for jscrund measures time for NOOP operations with callbacks.
                ------------------------------------------------------------
                revno: 0.82.127
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-10-30 14:43:43 -0700
                message:
                  Fix for compiler warnings on gcc 4.8.2
                ------------------------------------------------------------
                revno: 0.82.126
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-10-30 13:51:45 -0700
                message:
                  jscrund_dbspi now works properly with the jscrund changes merged from Martin
                ------------------------------------------------------------
                revno: 0.82.125
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-10-30 13:50:29 -0700
                message:
                  Change STATS logging from log_debug() to log_detail()
                ------------------------------------------------------------
                revno: 0.82.124
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-29 12:58:47 -0700
                message:
                  jscrund: fix shutdown
                ------------------------------------------------------------
                revno: 0.82.123
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-29 12:47:34 -0700
                message:
                  Adapt jscrund changes to run on very old version on Node
                ------------------------------------------------------------
                revno: 0.82.122 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-29 11:59:13 -0700
                message:
                  Merge
                    ------------------------------------------------------------
                    revno: 0.90.4 [merge]
                    committer: Martin Zaun <martin.zaun@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-10-08 15:38:33 -0700
                    message:
                      merge to nodejs-adapter
                        ------------------------------------------------------------
                        revno: 0.91.1
                        committer: Martin Zaun <martin.zaun@oracle.com>
                        branch nick: nodejs-adapter-test2
                        timestamp: Tue 2013-10-08 15:36:55 -0700
                        message:
                          jscrund: fixes and additions for error callbacks, cmdline options, writing of result file, object verification, additional table mapping.
                ------------------------------------------------------------
                revno: 0.82.121
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-29 11:47:45 -0700
                message:
                  jscrund / jscrund_dbspi minor formatting & style changes
                ------------------------------------------------------------
                revno: 0.82.120
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-10-28 20:22:28 -0700
                message:
                  New JSCrund implementation uses SPI directly
                ------------------------------------------------------------
                revno: 0.82.119
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-10-28 18:07:02 -0700
                message:
                  Disable debugging output from IndexBounds.js
                ------------------------------------------------------------
                revno: 0.82.118
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-10-28 18:03:36 -0700
                message:
                  Disable code leading to segfault in GC.
                ------------------------------------------------------------
                revno: 0.82.117
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-10-28 17:58:22 -0700
                message:
                  reorg: wscript fixes
                ------------------------------------------------------------
                revno: 0.82.116
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-10-28 17:53:27 -0700
                message:
                  Source reorg part 2: change build scripts (CMake,waf,gyp)
                ------------------------------------------------------------
                revno: 0.82.115
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-10-28 17:45:51 -0700
                message:
                  Reorganize source
                  Files from Adapter/impl/include and Adapter/impl/src are moved, 
                  sorted into common/{src,include} or ndb/{src,include}
                ------------------------------------------------------------
                revno: 0.82.114
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Thu 2013-10-24 14:59:20 -0700
                message:
                  Change stats counter from ["connect","async"] to just ["connect"] because
                  there is no longer a connectSync() api.
                ------------------------------------------------------------
                revno: 0.82.113 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-08 14:53:19 -0700
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 0.90.3 [merge]
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-10-08 14:52:42 -0700
                    message:
                      merge
                    ------------------------------------------------------------
                    revno: 0.90.2 [merge]
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-10-08 13:41:44 -0700
                    message:
                      merge
                    ------------------------------------------------------------
                    revno: 0.90.1
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: nodejs-adapter
                    timestamp: Tue 2013-10-08 13:39:48 -0700
                    message:
                      Fix bug where transaction.begin that succeeds will not call the callback
                ------------------------------------------------------------
                revno: 0.82.112
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-08 14:51:34 -0700
                message:
                  Fix in IndexBounds after code review
                ------------------------------------------------------------
                revno: 0.82.111
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-10-08 13:39:16 -0700
                message:
                  IndexBounds.js: work in progress
                ------------------------------------------------------------
                revno: 0.82.110
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-09-25 15:44:07 -0700
                message:
                  Fix so IndexBounds can store ranges for multiple indexes
                ------------------------------------------------------------
                revno: 0.82.109
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-09-25 13:57:34 -0700
                message:
                  Change getIndexBounds() API so it might be used to evaluate 
                  a candidate index.
                ------------------------------------------------------------
                revno: 0.82.108
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-09-25 12:20:58 -0700
                message:
                  All index scans in Ndb are now bounded. 
                  The current behavior is only correct for values that JavaScript can 
                  properly compare.  For other values (e.g. strings using non-unicode
                  collations) there is still work to be done so that IndexBounds.js
                  can get encoded values rather than native values.
                ------------------------------------------------------------
                revno: 0.82.107
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-09-25 11:07:42 -0700
                message:
                  lint-related fixes
                ------------------------------------------------------------
                revno: 0.82.106
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Wed 2013-09-25 09:30:49 -0700
                message:
                  Expose some errors when they're useful
                ------------------------------------------------------------
                revno: 0.82.105
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-09-24 09:29:32 -0700
                message:
                  More tests for multi-part keys 
                ------------------------------------------------------------
                revno: 0.82.104
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-09-23 20:10:19 -0700
                message:
                  New "multipartkeys" test suite
                ------------------------------------------------------------
                revno: 0.82.103
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-09-23 20:09:50 -0700
                message:
                  Some fixes for IS NULL / IS NOT NULL in NDB queries
                ------------------------------------------------------------
                revno: 0.82.102
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-09-23 12:00:00 -0700
                message:
                  Consolidate individual column bounds into IndexBounds
                ------------------------------------------------------------
                revno: 0.82.101
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Mon 2013-09-23 11:06:37 -0700
                message:
                  Fix docs error for indexMetadata
                ------------------------------------------------------------
                revno: 0.82.100
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Fri 2013-09-20 21:56:37 -0700
                message:
                  Work on IndexBounds; it can now create a BoundsHelper for the first column of the index.
                ------------------------------------------------------------
                revno: 0.82.99
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-09-17 13:09:30 -0700
                message:
                  Remove tests that pass string parameter as timestamp
                ------------------------------------------------------------
                revno: 0.82.98
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-09-17 11:47:08 -0700
                message:
                  Cleanup of connection pool and connection handshake
                    move driverTypeConverter from MySQLConnectionPool to MySQLConnection
                  
                  MySQLConnection:
                    add driverTypeConverter
                    avoid meaningless copy of driverTypeConverter into Operation objects
                  
                  MySQLConnectionPool:
                    remove driverTypeConverter
                ------------------------------------------------------------
                revno: 0.82.97
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: nodejs-adapter
                timestamp: Tue 2013-09-17 11:01:10 -0700
                message:
                  Remove conncetSync()
            ------------------------------------------------------------
            revno: 2876.577.5
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Thu 2014-01-16 15:33:22 +0100
            message:
              Remove unused and never working test program
            ------------------------------------------------------------
            revno: 2876.577.4 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Thu 2014-01-16 15:31:01 +0100
            message:
              Merge
            ------------------------------------------------------------
            revno: 2876.577.3
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2014-01-16 13:57:11 +0100
            message:
              Bug#14166590 DATA NODE CRASHES AT LINE 1388 IN DBTC: Increased next available error code in DBTC for error injection to 8105
            ------------------------------------------------------------
            revno: 2876.577.2
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2014-01-16 11:20:04 +0100
            message:
              Bug#14166590 DATA NODE CRASHES AT LINE 1388 IN DBTC: Fixed incorrect merge from 7.2, error codes (for error injection in test cases) 8100 and 8101 are already used in 7.3, changed them to 8103 and 8104
            ------------------------------------------------------------
            revno: 2876.577.1 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Wed 2014-01-15 09:27:04 +0100
            message:
              freeze 5.6.15-ndb-7.3.4, bump to 7.3.5
                ------------------------------------------------------------
                revno: 2875.612.23 [merge]
                tags: mysql-cluster-7.2.15
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Wed 2014-01-15 09:23:31 +0100
                message:
                  freeze 5.5.35-ndb-7.2.15, bump to 7.2.16
                    ------------------------------------------------------------
                    revno: 2555.817.149 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Wed 2014-01-15 09:21:29 +0100
                    message:
                      freeze 5.1.73-ndb-7.1.30, bump to 7.1.31
                        ------------------------------------------------------------
                        revno: 2555.827.71 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Wed 2014-01-15 09:16:52 +0100
                        message:
                          freeze 5.1.73-ndb-7.0.41, bump to 7.0.42
                            ------------------------------------------------------------
                            revno: 2555.836.9
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Wed 2014-01-15 09:14:53 +0100
                            message:
                              freeze 6.3.54, bump to 6.3.55
        ------------------------------------------------------------
        revno: 2876.563.48
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2014-01-15 09:15:03 +0100
        message:
          Kludge the nodejs adapters version check for NdbApi to default instead of error
          
           - Awaiting the proper fix from upstream
           - Bug filed
        ------------------------------------------------------------
        revno: 2876.563.47
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-14 16:28:23 +0100
        message:
          Revert unnecessary upstream diff(already gone from 7.3)
        ------------------------------------------------------------
        revno: 2876.563.46 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-14 16:27:08 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.139
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Tue 2014-01-14 16:25:40 +0100
            message:
              Revert unneccessary upstream diff
        ------------------------------------------------------------
        revno: 2876.563.45 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-14 16:16:20 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.138
            tags: clone-mysql-5.6.15-ndb-7.3.4-src-build
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2014-01-13 14:36:24 +0100
            message:
              Revert MCP_BUG17954277 patch
               - fixed in trunk 
               - workaround implemented on failing build machine
            ------------------------------------------------------------
            revno: 2876.565.137
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2014-01-13 14:03:09 +0100
            message:
              Remove small diff against upstream in func_str.test
            ------------------------------------------------------------
            revno: 2876.565.136
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2014-01-13 13:57:15 +0100
            message:
              Revert extra new line diff in mysql-test/CMakeLists.txt
            ------------------------------------------------------------
            revno: 2876.565.135
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2014-01-13 13:51:33 +0100
            message:
              Revert disabling of some innodb_* tests which previously failed in valgrind
              
               - Leak has since been  fixed upstream
            ------------------------------------------------------------
            revno: 2876.565.134
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2014-01-10 17:33:26 +0100
            message:
              ndb - test: testFK -n CreateDropWithData T1 needs higher MaxNoOfConcurrentScans
              
              Raising to MaxNoOfConcurrentScans = 300.
              Experiments gave that ~257 is needed.
            ------------------------------------------------------------
            revno: 2876.565.133 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2014-01-10 17:31:13 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.22 [merge]
                tags: clone-mysql-5.5.35-ndb-7.2.15-src-build
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Fri 2014-01-10 17:29:00 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.148 [merge]
                    tags: clone-mysql-5.1.73-ndb-7.1.30-src-build
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Fri 2014-01-10 17:26:14 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.70 [merge]
                        tags: clone-mysql-5.1.73-ndb-7.0.41-src-build
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Fri 2014-01-10 17:25:10 +0100
                        message:
                          merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.836.8
                            tags: clone-mysql-5.1.73-ndb-6.3.54-src-build
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Fri 2014-01-10 17:23:22 +0100
                            message:
                              ndb - suppress array-bounds warning in DbtuxMeta.cpp
                              
                              Bug #17985608  BUILD: NEW WARNINGS WITH GCC 4.8.2
            ------------------------------------------------------------
            revno: 2876.565.132
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18041636-73
            timestamp: Fri 2014-01-10 18:10:30 +0200
            message:
              bug#18041636 fixdict-x1.diff
              put jam()s in the switch
            ------------------------------------------------------------
            revno: 2876.565.131
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18041636-73
            timestamp: Fri 2014-01-10 18:09:31 +0200
            message:
              bug#18041636 fixdict.diff
              drop FK parse must not zap triggers
            ------------------------------------------------------------
            revno: 2876.565.130
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug18041636-73
            timestamp: Fri 2014-01-10 18:08:31 +0200
            message:
              bug#18041636 testdict2.diff
              testDict FK schema trans
        ------------------------------------------------------------
        revno: 2876.563.44 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2014-01-10 18:11:49 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.129 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2014-01-10 15:35:10 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.21 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Fri 2014-01-10 15:33:58 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.147
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.1
                    timestamp: Fri 2014-01-10 15:31:52 +0100
                    message:
                      Run ndb_show_tables with fixed nodedi to avoid stealing of mysqld alloced nodeid
                      
                       - fixes problem with ndb_bug17624736 where ndb_show_tables return error
                ------------------------------------------------------------
                revno: 2875.612.20
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Fri 2014-01-10 11:16:44 +0100
                message:
                  Make mtr.pl saved the log file created by ndb_ * tools
                  
                   - Change the location of log file for ndb_* tools from var/log to var/tmp in order
                     to get file saved after test failure.
            ------------------------------------------------------------
            revno: 2876.565.128 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2014-01-10 10:29:30 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.19 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2014-01-10 10:27:12 +0100
                message:
                  Merge 7.1 -> 7.2
                  
                  Also add memcached and jtie components of the patch which did
                  not exists prior to 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.146 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2014-01-10 10:16:54 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.69
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2014-01-10 10:14:37 +0100
                        message:
                          POTENTIALLY SERIOUS COMPILER WARNING: "_FILE_OFFSET_BITS" REDEFINED
                          
                          '_FILE_OFFSET_BITS' is defined in my_config.h to either 32 or 64 depending
                          on wheteher configure determines that 32 or 64 bits file offsets should be 
                          used by the filesystem calls. (Like 'offs' argumet to lseek())
                          
                          If not defined, and compiled as 32 bit binaries,
                          some OS'es assume that 32 bit offsets will be used and thus provides the 
                          default '#define _FILE_OFFSET_BITS 32'.
                          
                          If 'my_config.h' is included to late (or my_global.h / ndb_global.h
                          which include 'config'), '_FILE_OFFSET_BITS 32' will first be
                          defined, and then later redefined to '64' by my_config.h. This could result
                          in undefined behaviour by using a mix of normal, and the 'large' filesystem calls.
                          
                          In addition to silence this warning, this fix also seems to fix an
                          issue where several tests in the main suite fails with errno=29.
                          (ESPIPE 29 /* Illegal seek */ )
            ------------------------------------------------------------
            revno: 2876.565.127 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2014-01-09 22:36:26 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.18
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2014-01-09 22:33:59 +0100
                message:
                  ndb - hide array-bounds warnings for DbtuxMeta.cpp in club
                  
                  Bug #17985608	 BUILD: NEW WARNINGS WITH GCC 4.8.2
                  
                  warning indicates no bug, and could not rewrite code so that gcc accepted it without warning.
            ------------------------------------------------------------
            revno: 2876.565.126 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2014-01-09 22:32:39 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.17 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2014-01-09 22:30:38 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.145
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Thu 2014-01-09 22:26:55 +0100
                    message:
                      ndb - reverting patch
                      
                      revision-id: ole.john.aske@oracle.com-20140109145657-6ey2qmpqap0jmykc
                      parent: ole.john.aske@oracle.com-20140108150924-mabnupd65qe6gcvs
                      committer: Ole John Aske <ole.john.aske@oracle.com>
                      branch nick: mysql-5.1-telco-7.0
                      timestamp: Thu 2014-01-09 15:56:57 +0100
                      message:
                        Fix for bug#18053050  POTENTIALLY SERIOUS COMPILER WARNING: "_FILE_OFFSET_BITS" REDEFINED
            ------------------------------------------------------------
            revno: 2876.565.125 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2014-01-09 17:26:14 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.16 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2014-01-09 17:23:53 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.144 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Thu 2014-01-09 17:22:23 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.68 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Thu 2014-01-09 17:20:51 +0100
                        message:
                          merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.836.7
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Thu 2014-01-09 17:18:15 +0100
                            message:
                              ndb - disabling optimization for Ndbfs::readWriteRequest() for gcc 4.4 and up
                              
                              Temporary workaround for
                              Bug #18055285	 LOTS OF TESTS FAILS IN CLUB MADNESS WITH NEW GCC 4.8.2 -O3
            ------------------------------------------------------------
            revno: 2876.565.124 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2014-01-09 16:25:00 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.15 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2014-01-09 16:23:15 +0100
                message:
                  Merge 7.1 -> 7.2
                  
                  Also added the memcache and jtie part of this fix.
                    ------------------------------------------------------------
                    revno: 2555.817.143 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2014-01-09 15:58:31 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.837.1
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2014-01-09 15:56:57 +0100
                        message:
                          Fix for bug#18053050  POTENTIALLY SERIOUS COMPILER WARNING: "_FILE_OFFSET_BITS" REDEFINED
                          
                          '_FILE_OFFSET_BITS' is defined in my_global.h to either 32 or 64 depending
                          on wheteher configure determines that 32 or 64 bits file offsets should be 
                          supported by lseek(). If not defined, and compiled as 32 bit binaries,
                          some OS'es assume that 32 bit lseek() will be used, and thus provides the 
                          default '#define _FILE_OFFSET_BITS 32'.
                          
                          If 'my_global.h' is included to late, '_FILE_OFFSET_BITS 32' will first be
                          defined, and then later redefined to '64' by my_global.h. This could result
                          in part of the code using 32 bit lseek() while later part use 64 bit lseek().
                          
                          In addition to silence this warning, this fix also seems to fix an
                          issue where several tests in the main suite fails with errno=29
                          (ESPIPE 29 /* Illegal seek */ )
        ------------------------------------------------------------
        revno: 2876.563.43 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2014-01-10 18:05:38 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.123 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Thu 2014-01-09 15:43:40 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.14
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Thu 2014-01-09 15:41:38 +0100
                message:
                  Add --valgrind option to compile-cluster
                   - Use WITH_VALGRIND to enhance detection
                   - Set the HAVE_purify define to suppress som detected "errors"
            ------------------------------------------------------------
            revno: 2876.565.122 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2014-01-09 11:40:54 +0100
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.13 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2014-01-09 11:39:06 +0100
                message:
                  merge 7.1 -> 7.2
            ------------------------------------------------------------
            revno: 2876.565.121 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2014-01-09 11:34:55 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.613.1 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2014-01-09 11:31:30 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.142 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Thu 2014-01-09 11:29:50 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.67
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Thu 2014-01-09 11:24:30 +0100
                        message:
                          ndb - replace check_version_ge with check_version_new in mgmapi
                                
                          check_version_new() checks that a version is newer or same as some
                          version in passed list of versions with same major and minor version,
                          or is newer than all passed versions.
                          
                          Bug #17782773	 FUNCTION CHECK_VERSION_GE IN MGMAPI HAS NOT INTENDED BEHAVIOUR
                          
                          also solves
                          
                          Bug #17775071     7.2.14 CLUSTER CAN'T CONNECT TO MGMD RESTARTED ON 7.3.2 DURING UPGRADE
            ------------------------------------------------------------
            revno: 2876.565.120 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2014-01-08 16:28:40 +0100
            message:
              Null-Merge into 7.3
                ------------------------------------------------------------
                revno: 2875.612.12 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2014-01-08 16:26:09 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.141 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Wed 2014-01-08 16:10:20 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.66
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Wed 2014-01-08 16:09:24 +0100
                        message:
                          Backport of MTR change from 7.3.
                          
                          That extra 'SELECT LENGTH()' caused several 'audit_log' tests
                          to fail as an extra 'Query' was written to audit.log.
        ------------------------------------------------------------
        revno: 2876.563.42 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2014-01-10 18:03:40 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.119
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Wed 2014-01-08 09:42:33 +0100
            message:
              Revert minor merge diff versus 5.6.15
              
               - The check of number of error messages had been merged out
            ------------------------------------------------------------
            revno: 2876.565.118 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Wed 2014-01-08 03:35:21 +0100
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.11 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Wed 2014-01-08 03:34:15 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.140 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Wed 2014-01-08 03:33:25 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.65
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Wed 2014-01-08 03:31:47 +0100
                        message:
                          ndb - check alter flags without overlapping memcpy
                          
                          Bug #17985657     VALGRIND: PB2 NDB.NDB_AUTO_INCREMENT FAILS. OVERLAPPING MEMCPY IN HANDLER
            ------------------------------------------------------------
            revno: 2876.565.117 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Wed 2014-01-08 01:34:00 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.10 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Wed 2014-01-08 01:32:13 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.139 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Wed 2014-01-08 01:30:29 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.64
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Wed 2014-01-08 01:27:44 +0100
                        message:
                          ndb - removing some warnings in madness builds (gcc-4.8.2)
                          
                          Bug #17985608    BUILD: NEW WARNINGS WITH GCC 4.8.2
                          
                          No change in behaviour is expected, but that signal
                          ALTER_TAB_REQ actually will be zero filled for AlterTableReadWrite.
            ------------------------------------------------------------
            revno: 2876.565.116 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Tue 2014-01-07 16:30:04 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.9 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Tue 2014-01-07 16:29:24 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.138
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.1
                    timestamp: Tue 2014-01-07 14:54:57 +0100
                    message:
                      Add missing include/have_ndb.inc to two tests which requires NDB
                      
                       - missing include/have_ndb.inc caused mtr.pl worker thread to die when trying
                         to start ndb_mgmd. This went undetected by the main mtr.pl thread and
                         thus you get the rather cryptic "only X-1 out of X tests completed" at the
                        end of mtr.pl
        ------------------------------------------------------------
        revno: 2876.563.41
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-01-09 11:19:44 +0100
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
           - The output from EXPLAIN has been changed to always
             included EXTENDED and PARTITIONS information. Thus
             all .result files using EXPLAIN has to be updated
             with additional output. The result set
             contains an additional column called partitions
             and the EXPLAIN command always shows a warning describing
             the query further.
           - When using EXPLAIN EXTENDED or EXPLAIN PARTITIONS
             a deprectaion warning will be printed. Include the
             warning in the .result files.
        ------------------------------------------------------------
        revno: 2876.563.40 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2014-01-09 09:06:39 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.115 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2014-01-07 12:44:11 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.8 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2014-01-07 12:42:46 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.137 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2014-01-07 12:38:01 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.63
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2014-01-07 12:35:29 +0100
                        message:
                          Fix for bug#16620938  SOL 10/11 X64 32BIT FAILING TO BUILD NDBMTD
                          
                                 '#if defines' in mt-asm.h didn't check for 32bit x86 platforms '__i386'
                          
                          Inluded this as suggested in bug report by Laasya Moduludu.
                          
                          Also change configure and makefiles such that we always try to
                          compile 'WITH_NDBMTD' - As 'ndbmtd' is a mandatory part of our deliveries,
                          and failure to build this should *not* be ignored.
            ------------------------------------------------------------
            revno: 2876.565.114 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2014-01-07 12:22:35 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.7 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2014-01-07 12:20:37 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.136 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2014-01-07 12:18:28 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.62
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2014-01-07 12:17:39 +0100
                        message:
                          Fix for bug#18040921 SIGSEV WHEN CONSTRUCTING 'G_THR_REPOSITORY' IF COMPILED WITH -O3
                          
                          Also intended to fix failing 'Madness' tests in Club2.
                          
                          'g_thr_repository' was defined and constructed inside mt.cpp as a global data
                          member.
                          
                          It is generally adviced against constructing global data members as the order
                          of construction are somewhat undefined. We change this such that
                          g_thr_repository is a pointer to a 'struct thr_repository', and
                          allocate & construct this object as part of the init code instead.
            ------------------------------------------------------------
            revno: 2876.565.113 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2014-01-06 10:03:27 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.6 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2014-01-06 10:02:00 +0100
                message:
                  Merge 7.1->7.2
                    ------------------------------------------------------------
                    revno: 2555.817.135 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2014-01-06 09:54:56 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.61
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Mon 2014-01-06 09:53:38 +0100
                        message:
                          Fix for Bug#17857442: MULTITHREADED JOB SCHEDULER SHOULD HANDLE LEAPS IN CLOCK TIME
                          
                          We have seen customer bug reports, like bug 17475425, where it
                          looks like delayed signal processing from the 'time_queue' comes
                          completely out of control: There seems to be two scenarios:
                          
                           - Processing of delayes signals stops completely for a long periode,
                             or forever. This will cause issues like the disk write speed throttling
                             to completely halt any disk activity (Local checkpoints stop)
                          
                           - We see a burst of signals which should have been delayed relatively
                             to each other, executed withing the same clock second.
                          
                          Inspecting the mt-scheduler code, we find that this may happen if the
                          NdbTick clock moves forward or backwards in large leaps. When the time
                          moved backwards, it will stop the time_queue handling until the
                          same amount of clock time has passed again. A forward leap will force
                          the scheduler to simulate that the same amount of time passing by expiring
                          all time_queue event in the elapsed periode.
                          
                          Such timer leaps might happen if the platform does not support monotonic
                          timers, or we didn't implement the usage of these on the specific platform.
                          Windows used to be such a platform prior to fixing bug 17647637. However,
                          there are still such platforms, OSX seems to be one of these, and possible
                          some Linux variants. Furthermore there seems to be several OS/HW/VM bugs
                          related to monotonic timers not always being monotonic after all. And
                          there might be time leaps due to CPU starvation such that we are stuck
                          for a long time - So this has to be expected and handled somehow.
                          
                          Looking at the single threaded scheduler we find that it actually handle
                          such leaps - see ThreadConfig::scanTimeQueue(). This logic doesn't seem
                          to have been ported to the mt-version !! The single threaded handling
                          is to accept a backtick as the new current time. A forward leap of more
                          than 1500ms will be consumed by resetting the current time to 'now-1000'ms
                          and continue from there.
                          
                          This fix introduce the same handling in the MT-scheduler.
                          
                          Furthermore, we also found that the mt-scheduler allowed itself to
                          yield CPU, or wait for receiving signals, even if it has possibly
                          expired event to handle from the time_queue. 
                          (see: 'lagging_timers')
                          
                          We are not sure about whether this has any ill effects - However, 
                          it doesn't seem like a good idea to let possible expired signals
                          linger any longer than necessary - also fixed.
                          
                          Also removed a ndbout_c() which was a leftover from previous debugging.
            ------------------------------------------------------------
            revno: 2876.565.112 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2014-01-02 16:32:05 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.5
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2014-01-02 16:23:04 +0100
                message:
                  Added missing 'sorted_result' to testcase for bug#17845161
            ------------------------------------------------------------
            revno: 2876.565.111 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2013-12-20 11:16:20 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.4
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-12-20 11:15:19 +0100
                message:
                  The MTR test ndb_big.rqg_spj attempted tp drop the incorrect (non-existing) 
                  database after the test has finished.
            ------------------------------------------------------------
            revno: 2876.565.110 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-12-19 16:12:08 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.3 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-12-19 16:09:32 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.134 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-12-19 16:08:05 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.60
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-12-19 16:07:16 +0100
                        message:
                          Fix compiler warning in prev commit
            ------------------------------------------------------------
            revno: 2876.565.109 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-12-19 14:26:20 +0100
            message:
              Empty merge 7.2 -> 7.3 to resolve previous merge jam
                ------------------------------------------------------------
                revno: 2875.612.2 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-12-19 14:25:01 +0100
                message:
                  Got merge-commit-push jam with Magnus latest pushed.
                  
                  Merged in these from parent branch.
            ------------------------------------------------------------
            revno: 2876.565.108 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-12-19 14:10:17 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.612.1 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-12-19 14:05:46 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.133 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-12-19 14:00:16 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.59
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-12-19 13:59:01 +0100
                        message:
                          Fix for bug#17973819  NEED TO ADAPT TO MONOTONIC TIMERS NOT ALWAYS BEING MONOTONIC 
                          
                          Also a followup fix for bug#17647637, NDBTICK_CURRENTMICROSECOND() MUST BE MONOTONIC
                          
                          This fix handle that Monotonic timers seems to have bugs on several platforms
                          which might result in the monotonic clock doing small jumps back in time.
                          This is normally due to imperfect syncing of the clock between multiple 
                          CPU cores.
                          
                          Such small backticks are not really harmfull for our scheduler and watchdog
                          algorithms, so we make the backtick protection less strict in this patch.
                          However, we will still assert that the backtick is less than 10ms in the
                          NdbTick_Elapsed() calculation.
                          
                          We removed redundant checks for backticks in several 'timediff'
                          methods as such a check already exist in NdbTick_Elapsed() which is 
                          used in the same place.
                          
                          We also removed some assert('not backtick') where they are used 
                          in the same place where NdbTick_Elapsed() was called for the
                          same reason as above.
            ------------------------------------------------------------
            revno: 2876.565.107 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Thu 2013-12-19 13:51:34 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.60 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Thu 2013-12-19 13:50:06 +0100
                message:
                  Merge in fixes for bug#17672846 NDB API EXAMPLE PROGRAMS DO NOT COMPILE
                    ------------------------------------------------------------
                    revno: 2875.611.2
                    committer: Gustaf Thorslund <gustaf.thorslund@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2-bug70759
                    timestamp: Wed 2013-12-18 12:20:33 +0100
                    message:
                      Fixed linking errors as part of bug #17672846 / bug 70759
                    ------------------------------------------------------------
                    revno: 2875.611.1
                    committer: Gustaf Thorslund <gustaf.thorslund@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2-bug70759
                    timestamp: Wed 2013-12-18 11:33:20 +0100
                    message:
                      Fixed compilation errors for bug #17672846 / bug 70759
            ------------------------------------------------------------
            revno: 2876.565.106 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-12-19 09:23:56 +0100
            message:
              Merge 7.2->7.3.
                ------------------------------------------------------------
                revno: 2875.603.59 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-12-19 09:08:07 +0100
                message:
                  Merge 7.1->7.2.
                    ------------------------------------------------------------
                    revno: 2555.817.132 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-12-19 09:04:13 +0100
                    message:
                      Merge 7.0->7.1.
                        ------------------------------------------------------------
                        revno: 2555.827.58
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-12-19 09:01:13 +0100
                        message:
                          This commit is a fix for Bug#17971449 'ASSERT FOR JAMEVENT::VERIFYID() CRASHES DATA NODE ON WINDOWS'
                          
                          With Visual C++, __FILE__ is always in lowercase. Therefore we must use
                          strcasecmp() rather than strcmp() when cheking jamFileNames entries against
                          __FILE__, since jamFileNames contains mixed-case names.
                ------------------------------------------------------------
                revno: 2875.603.58 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-12-18 20:11:59 -0800
                message:
                  merge 7.1 => 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.131
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1-maven
                    timestamp: Wed 2013-12-18 20:06:46 -0800
                    message:
                      Disable failing clusterjpa test
            ------------------------------------------------------------
            revno: 2876.565.105 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Wed 2013-12-18 18:33:20 -0800
            message:
              merge 7.2 => 7.3 clusterj build maven pom.xml
                ------------------------------------------------------------
                revno: 2875.603.57 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-12-18 15:53:06 -0800
                message:
                  merge 7.1 => 7.2 clusterj build maven pom.xml
                    ------------------------------------------------------------
                    revno: 2555.817.130
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1-maven
                    timestamp: Wed 2013-12-18 15:19:30 -0800
                    message:
                      Fix clusterj CMakeLists.txt
                    ------------------------------------------------------------
                    revno: 2555.817.129
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1-maven
                    timestamp: Wed 2013-12-18 13:33:14 -0800
                    message:
                      Clusterj create mvn pom.xml files
                      
                      Change configure and cmake files to create mvn pom.xml files
                        with proper version numbers and optional suffix, e.g. -SNAPSHOT
            ------------------------------------------------------------
            revno: 2876.565.104 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Wed 2013-12-18 17:29:23 +0100
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.56 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Wed 2013-12-18 17:27:51 +0100
                message:
                  null merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.128 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Wed 2013-12-18 17:17:41 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.57 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Wed 2013-12-18 17:16:25 +0100
                        message:
                          merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.836.6
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Wed 2013-12-18 17:12:44 +0100
                            message:
                              ndb - disable innodb.innodb_bug11766634
                              
                              bug is reported (and fixed) for 5.6,
                              but not fixed for lower version.
                              
                              Bug #14748587     INNODB.INNODB_BUG11766634 FAILING ON DAILY-5.6
                            ------------------------------------------------------------
                            revno: 2555.836.5
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Wed 2013-12-18 16:15:59 +0100
                            message:
                              ndb - remove some test from default.daily
                              
                              ndb.ndb_share, ndb_rpl.ndb_rpl_slave_lsu, and
                              ndb_rpl.ndb_rpl_slave_lsu_anyval starts some
                              mysqld without log-bin enabled and fails with
                              
                              [ERROR] You need to use --log-bin to make --binlog-format work.
                              
                              for runs in default.daily that sets binlog-format.
                              
                              The error is changed to warning for 5.5/7.2 and up as fix of
                              Bug #11751903	 42928: BINLOG-FORMAT SETTING PREVENTS SERVER FROM START IF BINARY LOGGING IS DIS
            ------------------------------------------------------------
            revno: 2876.565.103 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Wed 2013-12-18 12:11:22 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.55 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Wed 2013-12-18 12:10:08 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.127
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Wed 2013-12-18 11:50:47 +0100
                    message:
                      ndb - remove use of grep in test NDB_BINLOG_DISCOVER_MULTI
                      
                      Bug #17756394	 TEST CASE NDB_BINLOG.NDB_BINLOG_DISCOVER_MULTI IS FAILING IN PB2 (CLUSTER-7.3)
                    ------------------------------------------------------------
                    revno: 2555.817.126 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Wed 2013-12-18 11:43:18 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.56
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Wed 2013-12-18 11:38:51 +0100
                        message:
                          ndb - re-record test result for main.ctype_cp932_binlog_row
                          
                          Test is not run without adding --mysqld=--binlog-format=row to mtr command line.
                          
                          log event Write_rows changed to Write_rows_v1 by 
                          
                          revision-id: frazer.clement@oracle.com-20110921101158-0js4hali1q25aa1e
                          timestamp: Wed 2011-09-21 11:11:58 +0100
                          message:
                            WL5353 : Implement v2 of Binlog row events
                        ------------------------------------------------------------
                        revno: 2555.827.55 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Wed 2013-12-18 11:37:08 +0100
                        message:
                          merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.836.4
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Wed 2013-12-18 10:43:10 +0100
                            message:
                              ndb - change error number in test result for i_main.valgrind and parts.partition_debug_sync_innodb
                              
                              Error ER_DEBUG_SYNC_TIMEOUT has number 1639 in 5.1 but 1640 in 6.3, 7.0, 7.1.
                            ------------------------------------------------------------
                            revno: 2555.836.3
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Wed 2013-12-18 10:36:38 +0100
                            message:
                              ndb - cherrypick fix for test i_main.bug12917164 from future 5.1.74
                              
                              ------------------------------------------------------------
                              revision-id: venkata.sidagam@oracle.com-20131104133257-5nwr8nadsjb9yq2c
                              parent: murthy.narkedimilli@oracle.com-20131104102937-kfibutjolbi4v8sr
                              committer: Venkata Sidagam <venkata.sidagam@oracle.com>
                              branch nick: 5.1
                              timestamp: Mon 2013-11-04 19:02:57 +0530
                              message:
                                Bug #12917164 DROP USER CAN'T DROP USERS WITH LEGACY 
                                    UPPER CASE HOST NAME ANYMORE
                                Fixing the testcase failure for mysql-5.1
            ------------------------------------------------------------
            revno: 2876.565.102 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Tue 2013-12-17 00:36:12 +0100
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.54 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Tue 2013-12-17 00:34:01 +0100
                message:
                  null merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.125 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Tue 2013-12-17 00:31:04 +0100
                    message:
                      null merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.54 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Tue 2013-12-17 00:30:30 +0100
                        message:
                          null merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.836.2 [merge]
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Tue 2013-12-17 00:26:53 +0100
                            message:
                              merge 5.1.73 to 6.3.54
            ------------------------------------------------------------
            revno: 2876.565.101 [merge]
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-12-16 17:13:05 +0100
            message:
              Merge from 7.2
                ------------------------------------------------------------
                revno: 2875.603.53 [merge]
                committer: Martin Skold <Martin.Skold@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-12-16 17:02:11 +0100
                message:
                  Merge from 7.1
                    ------------------------------------------------------------
                    revno: 2555.817.124 [merge]
                    committer: Martin Skold <Martin.Skold@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-12-16 16:57:33 +0100
                    message:
                      Merge from 7.0
                        ------------------------------------------------------------
                        revno: 2555.827.53 [merge]
                        committer: Martin Skold <Martin.Skold@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Mon 2013-12-16 16:55:19 +0100
                        message:
                          Merge from 6.3
                            ------------------------------------------------------------
                            revno: 2555.836.1
                            committer: Martin Skold <Martin.Skold@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Mon 2013-12-16 16:52:51 +0100
                            message:
                              Bug#14166590 DATA NODE CRASHES AT LINE 1388 IN DBTC: Added have_ndb_debug test to verify that error insert is supported for test case
            ------------------------------------------------------------
            revno: 2876.565.100
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2013-12-16 10:21:41 +0100
            message:
              Remove disabling of innodb-wl5980-windows since bug reported about unzip has been fixed
            ------------------------------------------------------------
            revno: 2876.565.99 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Mon 2013-12-16 10:17:05 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.52 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Mon 2013-12-16 10:11:34 +0100
                message:
                  Merge
                    ------------------------------------------------------------
                    revno: 2875.610.1
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Fri 2013-12-13 10:08:34 +0100
                    message:
                      Backport fix for properly detecting if warnings should be silenced
                      
                       - Use a unique variable name for each invocation of CHECK_CXX_COMPILER_FLAG
                      - Also removes usage of UNSET which is apparently not supported in older cmake versions
        ------------------------------------------------------------
        revno: 2876.563.39 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2014-01-08 12:04:57 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.98 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Sun 2013-12-15 00:12:33 +0100
            message:
              merge 5.6.15 into 7.3.4
                ------------------------------------------------------------
                revno: 2876.576.7 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sun 2013-12-15 00:10:51 +0100
                message:
                  null merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.603.51 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Sun 2013-12-15 00:08:21 +0100
                    message:
                      null merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.123 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Sun 2013-12-15 00:05:31 +0100
                        message:
                          merge 7.0 -> 7.1
                            ------------------------------------------------------------
                            revno: 2555.827.52 [merge]
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-7.0
                            timestamp: Sun 2013-12-15 00:03:32 +0100
                            message:
                              merge 6.3 -> 7.0
                                ------------------------------------------------------------
                                revno: 2555.760.88 [merge]
                                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                                branch nick: mysql-6.3
                                timestamp: Sun 2013-12-15 00:01:11 +0100
                                message:
                                  merge 5.1.73 into 6.3.54
                    ------------------------------------------------------------
                    revno: 2875.603.50 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Sat 2013-12-14 23:57:05 +0100
                    message:
                      merge 5.5.35 into 7.2.15
                    ------------------------------------------------------------
                    revno: 2875.603.49
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Sat 2013-12-14 23:53:58 +0100
                    message:
                      ndb - remove internal/mysql-test/suite/i_innodb/t/disabled.def*
                ------------------------------------------------------------
                revno: 2876.576.6
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sat 2013-12-14 23:49:15 +0100
                message:
                  ndb - add MCP for BUG#17954277 BUILD FAILS ON SOLARIS IF NO STL_LIBRARY_NAME FOUND
                ------------------------------------------------------------
                revno: 2876.576.5
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sat 2013-12-14 11:50:23 +0100
                message:
                  removed i_innodb/t/disabled.def
                ------------------------------------------------------------
                revno: 2876.576.4
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sat 2013-12-14 11:18:25 +0100
                message:
                  ndb - adopt server bugfix to maintained version of ha_ndbcluster.cc
                  
                  revision-id: jon.hauglid@oracle.com-20131030232243-vyz4pd5yihqlmnjx
                  parent: sivert.sorumgaard@oracle.com-20131030120943-g0i7td24pbfv1wjj
                  committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
                  branch nick: mysql-5.6-bug17696705
                  timestamp: Thu 2013-10-31 00:22:43 +0100
                  message:
                    Bug#17647863: MYSQL DOES NOT COMPILE ON OSX 10.9 GM 
                    
                    Rename test() macro to MY_TEST() to avoid conflict with libc++.
                ------------------------------------------------------------
                revno: 2876.576.3 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sat 2013-12-14 11:14:24 +0100
                message:
                  merge 5.6.15 into 7.3.4
                ------------------------------------------------------------
                revno: 2876.576.2
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sat 2013-12-14 10:53:12 +0100
                message:
                  ndb - remove i_innodb/t/disabled.def.moved
                ------------------------------------------------------------
                revno: 2876.576.1
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Sat 2013-12-14 10:42:55 +0100
                message:
                  ndb - added missing server tests
                  
                  i_main.func_gconcat
                  i_innodb.innodb_bug16244691
                  i_main.subquery_innodb
        ------------------------------------------------------------
        revno: 2876.563.38 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2014-01-08 11:09:09 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.97 [merge]
            committer: Martin Zaun <martin.zaun@oracle.com>
            branch nick: ndb-7.3
            timestamp: Fri 2013-12-13 14:42:32 -0800
            message:
              merge ndb 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.48 [merge]
                committer: Martin Zaun <martin.zaun@oracle.com>
                branch nick: ndb-7.2
                timestamp: Fri 2013-12-13 14:41:08 -0800
                message:
                  merge ndb 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.122 [merge]
                    committer: Martin Zaun <martin.zaun@oracle.com>
                    branch nick: ndb-7.1
                    timestamp: Fri 2013-12-13 14:37:12 -0800
                    message:
                      merge java build mavericks patch to 7.1
                        ------------------------------------------------------------
                        revno: 2555.835.1
                        committer: Martin Zaun <martin.zaun@oracle.com>
                        branch nick: ndb-7.1-test
                        timestamp: Thu 2013-12-12 10:47:20 -0800
                        message:
                          fix broken 7.1 java build on MacOS 10.9, macports
        ------------------------------------------------------------
        revno: 2876.563.37 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2014-01-08 11:05:59 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.96 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2013-12-13 09:20:08 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.47 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-12-13 09:17:04 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.121 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-12-13 09:15:18 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.51
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-12-13 08:59:20 +0100
                        message:
                          Patch for bug Bug#17947674, 
                          
                            DEFINTION OF UINT_MAX64 EVALUATES TO A SIGNED VALUE IN VISUAL STUDIO
                          
                          Also required as an addendum fix to bug#17647637 which introduced usage
                          of UINT_MAX64.
                          
                          It seems to be compiler dependent whether the definition of
                          UINT64_MAX64 actully is *unsigned* as required - In Visual Studio
                          it actually ends up as a signed value.
                          
                          This fix makes the UINT_MAX64 more explicit about being an 'unsigned int64'.
                          Also see how ULONGLONG_MAX is defined in my_global.h.
        ------------------------------------------------------------
        revno: 2876.563.36 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2014-01-08 10:00:03 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.95 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Thu 2013-12-12 09:20:29 -0800
            message:
              merge 7.2 => 7.3 disable flaky clusterjpa timestamp test
                ------------------------------------------------------------
                revno: 2875.603.46 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-12-12 09:12:33 -0800
                message:
                  merge 7.1 => 7.2 fix flaky clusterjpa timestamp test
                    ------------------------------------------------------------
                    revno: 2555.817.120
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-12-12 09:05:12 -0800
                    message:
                      Disable flaky timestamp test for clusterjpa
            ------------------------------------------------------------
            revno: 2876.565.94 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-12-11 13:28:08 +0100
            message:
              Merge 7.2->7.3
                ------------------------------------------------------------
                revno: 2875.603.45
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-12-11 11:22:36 +0100
                message:
                  Addendum to fix for Bug#17875885 "ONE MYSQL ACCESS SERVER IS HUNG".
                  
                  THD::~THD() will in some cases call Thd_ndb::~Thd_ndb() which used 'curret_thd', leading to the 
                  crash below (triggered by the ndb_dist_priv MTR test). This commit fixes that, by using Thd_ndb::m_thd
                  instead of 'current_thd'.
                  
                  Stack from core dump:
                  #0  0x00007fd46ddb7f8c in pthread_kill () from /lib/x86_64-linux-gnu/libpthread.so.0
                  #1  0x0000000000950360 in my_write_core (sig=11) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/mysys/stacktrace.c:433
                  #2  0x00000000007dd31c in handle_fatal_signal (sig=11) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/signal_handler.cc:247
                  #3  <signal handler called>
                  #4  Thd_ndb::~Thd_ndb (this=0x29f2b30, __in_chrg=<optimized out>) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/ha_ndbcluster.cc:1231
                  #5  0x0000000000a1d423 in Thd_ndb::release (thd_ndb=0x29f2b30) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/ndb_thd_ndb.cc:58
                  #6  0x00000000009db5e0 in ndbcluster_close_connection (hton=0x245bac0, thd=0x2a9c610) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/ha_ndbcluster.cc:11138
                  #7  0x00000000007decac in closecon_handlerton (thd=0x2a9c610, plugin=0x7fffed70a898, unused=0x0) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/handler.cc:725
                  #8  0x000000000068875a in plugin_foreach_with_mask (thd=0x2a9c610, func=0x7dec32 <closecon_handlerton(THD*, plugin_ref, void*)>, type=1, state_mask=4294967287, arg=0x0) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/sql_plugin.cc:2024
                  #9  0x00000000007decf8 in ha_close_connection (thd=0x2a9c610) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/handler.cc:739
                  #10 0x00000000006439b9 in THD::~THD (this=0x2a9c610, __in_chrg=<optimized out>) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/sql_class.cc:1396
                  #11 0x0000000000643c62 in THD::~THD (this=0x2a9c610, __in_chrg=<optimized out>) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/sql_class.cc:1427
                  #12 0x00000000009ddb3c in ndb_wait_setup_func_impl (max_wait=119) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/ha_ndbcluster.cc:11822
                  #13 0x00000000005dd3d9 in mysqld_main (argc=53, argv=0x248d908) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/mysqld.cc:4701
                  #14 0x00000000005d6674 in main (argc=9, argv=0x7fffed70ab08) at /export/home/tmp/jw159207/mysql/repo/mysql-5.5-cluster-7.2/sql/main.cc:25
        ------------------------------------------------------------
        revno: 2876.563.35 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2014-01-08 09:51:52 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.93 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-12-11 09:22:42 +0100
            message:
              Merge 7.2->7.3
                ------------------------------------------------------------
                revno: 2875.603.44
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-12-11 09:16:19 +0100
                message:
                  This is a fix for Bug#17875885 "ONE MYSQL ACCESS SERVER IS HUNG".
                  
                  The problem is that ndb_wait_setup_func_impl() leaves the THR_THD thread local storage key pointing
                  to a deleted THD object. This pointer is later dereferenced (via current_thd), resulting in undefined 
                  behavior.
                  
                  The fix consists in setting the THR_THD (and THR_MALLOC) key to NULL.
            ------------------------------------------------------------
            revno: 2876.565.92 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-12-10 16:09:20 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.43 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-12-10 16:06:58 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.119 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-12-10 16:02:31 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.50
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-12-10 16:01:51 +0100
                        message:
                          Followup fix for bug#17647637, NDBTICK_CURRENTMICROSECOND() MUST BE MONOTONIC
                          
                          Previous fix introduced a warning being printed from NdbTick_Init()
                          if we were unable to find a monotonic timer to be used for the 'ticks'.
                          OSX is one if the platforms where we have to fallback to the non-monotonic
                          gettimeofday() and thus prints this warning.
                          
                          We find from recent PB2 tests, that not only the datanode binaries
                          ndbd/dbbmtd, but also clients as ndb_mgm need timers, and thus get
                          this warning printed as part of the startup.
                          
                          As this is handled as part of the output from these clients, several
                          tests like 'ndb_mgm -e backup' fails with verification failures.
                          
                          This fix removes these warning, and instead introduce NdbTick_IsMonotonic().
                          This function is now used inside Watchdog.cpp to check the timer capabilities
                          and print a warning from here. As the Watchdog is only part of the ndb's,
                          this warning will go into the ndb.log file, and not to the terminal.
                          
                          Patch also adds some extra asserts to NdbTick to
                          ensure that NdbTick_Init() has been called prior to the other NdbTick
                          functions, and that the windows part of NdbTick_Init() returned sensible
                          values. This was added in order to smoke out some Pb2 failures being
                          observed on Win32/64.
            ------------------------------------------------------------
            revno: 2876.565.91 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-12-10 15:46:40 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.42 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-12-10 15:40:37 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.118 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-12-10 15:38:55 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.49
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-12-10 15:38:02 +0100
                        message:
                          Fix for Bug#17842035 LCP WATCHDOG SHOULD USE REAL CLOCKTIME
                          
                          Recommit of fix updated according to Frazers review comments
                          
                          The LCP watchdog implemented by Dblqh::checkLcpFragWatchdog is intended to be
                          called by the scheduler event mechanism every 10th s. Every time it is
                          called it increment its own 'pollCounter', assuming that 10s has passed.
                          
                          When this counts reaches a limit representing that either a 'WarnPeriod'
                          (default 20s), or 'Timeout Limit' (60s) has passed, either a LCP warning
                          is printed, or the datanode aborted due to 'SCAN_WATCHDOG_FAIL
                          
                          However, there are no guarantee that this timer fires at the
                          specified 10s interval.:
                          
                          - There might be CPU contention which prevents the job scheduler
                            from running. which results in a longer delay.
                          
                          - If the scheduler was 'late', as above, it will try to make up
                            for that by running the internal scheduler timer faster.
                            In some cases it might even cause a delayed event to fire
                            immediately.
                          
                          To overcome these limitation the watchdog has been refactored to
                          keep track of its own starttime, and calculate elapsed time by
                          reading the clock every time it is called.
                          
                          Furthermore, it will now absorbe a 'backtick' (not likely to occure)
                          by taking the backwards time as new 'current time' and calcule 'elapsed'
                          time for this round as '0'.
                          The ill effect of a forward leap, which possibly could expire the watchdog
                          timer immediately, is reduced by never calculating an elapsed time
                          longer than the requested 'delay' time of the watchdog timer.
                        ------------------------------------------------------------
                        revno: 2555.827.48
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-12-10 15:37:08 +0100
                        message:
                          Fix for bug#17647469
                            'GCP LAG' WATCHDOG (DBDIH::CHECKGCPSTOPLAB) SHOULD USE REAL CLOCK TIME
                          
                          Recommit of fix updated according to comments from Frazer
                          
                          The GCP watchdog implemented by Dbdih::checkGcpStopLab is intended to be
                          called by the scheduler event mechanism every 100th ms. Every time it is
                          called it increment its own 'invocation counter'. When these counts reaches
                          (n*10) == <timeout limit>, the GCP LAG warning is printed, or the 'crashSystem'
                          action taken.
                          
                          However, there are no guarantee that such timed events actually are called
                          after the specified delay:
                          
                           - There might be CPU contention which prevents the job scheduler
                             from running. which results in a longer delay.
                           - If the scheduler was 'late', as above, it will try to make
                             up for that by running the internal scheduler timer faster. 
                             In some cases that might even cause a delayed event to fire
                             immediately.
                           - The resolution of the OS timers may be to coarse to
                             reliable being able to hit near the 100ms mark.
                          
                          To overcome these limitation the watchdog has been refactored to
                          keep track of its own starttime, and calculate elapsed time by
                          reading the clock every time it is called.
                          
                          Furthermore, it will now absorbe a 'backtick' (not likely to occure)
                          by taking the backwards time as new 'current time' and calcule 'elapsed'
                          time for this round as '0'.
                          The ill effect of a forward leap, which possibly could expire the watchdog
                          timer immediately, is reduced by never calculating an elapsed time
                          longer than the requested 'delay' time of the watchdog timer.
            ------------------------------------------------------------
            revno: 2876.565.90 [merge]
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-12-10 15:05:47 +0100
            message:
              Merge from 7.2
                ------------------------------------------------------------
                revno: 2875.603.41 [merge]
                committer: Martin Skold <Martin.Skold@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-12-10 14:53:16 +0100
                message:
                  Merge from 7.1
                    ------------------------------------------------------------
                    revno: 2555.817.117 [merge]
                    committer: Martin Skold <Martin.Skold@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-12-10 14:48:35 +0100
                    message:
                      Merge from 7.0
                        ------------------------------------------------------------
                        revno: 2555.827.47 [merge]
                        committer: Martin Skold <Martin.Skold@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-12-10 14:20:54 +0100
                        message:
                          Merge from 6.3
                            ------------------------------------------------------------
                            revno: 2555.760.87
                            committer: Martin Skold <Martin.Skold@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Tue 2013-12-10 13:48:01 +0100
                            message:
                              Bug#14166590 DATA NODE CRASHES AT LINE 1388 IN DBTC: Corrected condition when it is safe to restart a transaction when receiving TCINDXREQ (unique index operation), added a elaborate test case
            ------------------------------------------------------------
            revno: 2876.565.89 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug16766493-73
            timestamp: Sat 2013-12-07 16:34:10 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.603.40 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug16766493-72
                timestamp: Sat 2013-12-07 16:19:04 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.116 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug16766493-71
                    timestamp: Sat 2013-12-07 16:03:26 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.46
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug16766493-70
                        timestamp: Thu 2013-12-05 14:13:51 +0200
                        message:
                          bug#16766493 cb-ack-x2.diff
                          add comment to process_log_buffer_waiters
                        ------------------------------------------------------------
                        revno: 2555.827.45
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug16766493-70
                        timestamp: Thu 2013-12-05 14:13:23 +0200
                        message:
                          bug#16766493 cb-ack-x1.diff
                          weaken an assert in base patch
                        ------------------------------------------------------------
                        revno: 2555.827.44
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug16766493-70
                        timestamp: Thu 2013-12-05 14:12:53 +0200
                        message:
                          bug#16766493 cb-ack.diff
                          serialize UNDO buffer commits via CALLBACK_ACK
                        ------------------------------------------------------------
                        revno: 2555.827.43
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug16766493-70
                        timestamp: Thu 2013-12-05 14:12:02 +0200
                        message:
                          bug#16766493 lgman-crash.diff
                          testNodeRestart test case
            ------------------------------------------------------------
            revno: 2876.565.88 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Fri 2013-12-06 14:04:08 -0800
            message:
              merge 7.1 => 7.2 bug 17885485
                ------------------------------------------------------------
                revno: 2875.603.39 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-12-06 13:55:31 -0800
                message:
                  merge 7.1 => 7.2 bug 17885485
                    ------------------------------------------------------------
                    revno: 2555.817.115
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-12-06 11:43:30 -0800
                    message:
                      iBug 17885485 - SETPARTITIONKEY() FAILS AFTER PREVIOUS TRANSACTION FAILS
                      
                      After a failed commit, the session transaction state becomes idle
                      but the cluster transaction is not cleaned up. A subsequent setPartitionKey
                      fails because there is still a cluster transaction.
                      
                      SessionImpl:
                        change commit to fail the transaction on errors
                        change executeCommit to call flush with the commit flag
                        change flush to allow the caller to executeCommit instead of executeNoCommit
                          followed by executeCommit
                          serendipity: this has a slight performance benefit
                      
                      TransactionErrorSetPartitionKeyTest
                        Add four tests for setPartitionKey after a failure to insert a duplicate record
                          commit: this is the original bug report
                          flush/commit: flush the failed change and then commit the transaction
                          flush/rollback: flush the failed change and then roll back the transaction
                          rollback: roll back the failed change
            ------------------------------------------------------------
            revno: 2876.565.87 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-12-05 10:23:54 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.38 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-12-05 09:32:58 +0100
                message:
                  Merge 7.1 -> 7.2
                  
                  Also added MTR testcase to ndb_join_pushdown_* as these tests are only relevant for >= 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.114 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-12-05 09:29:02 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.42
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-12-05 09:27:54 +0100
                        message:
                          Fix for bug#17845161 
                          
                            ERROR 1296: GOT ERROR 290 'CORRUPT KEY IN TC, UNABLE TO XFRM'; NDB_JOIN_PUSHDOWN
                          
                          A NULL value in a varchar column was used to construct a lookup key for 
                          its join-child operation. As a NULL value is never equal to any other value,
                          such a lookup could have been eliminated already at SPJ. However, in an
                          attempt to simplify the logic, we instead built a dummy 'empty KeyInfo' for
                          the NULL lookup. The intention was that such an empty keyInfo would never match
                          anything and always cause a KEYREF to be returned from LQH.
                          
                          However, it turned out that such an empty KeyInfo was not sufficiently when
                          the varchar required usage of key-xfrm to construct an uniform hashable key.
                          This caused the error message mentioned in bug subject.
                          
                          This fix use the fact that a NULL-key lookup will never find a matching row.
                          Instead of sending this useless KEYREQ to a LQH, SPJ will now produce the 
                          same REF-signals to the API as required by the protocoll.
            ------------------------------------------------------------
            revno: 2876.565.86 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-12-03 14:40:06 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.37
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster
                timestamp: Tue 2013-12-03 14:34:23 +0100
                message:
                  Change the SPJ RQG test such that InnoDB is used instead of myISAM as the 'baseline' 
                  which NDB results are compared against.
            ------------------------------------------------------------
            revno: 2876.565.85 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-11-26 16:59:38 +0530
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.36 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-11-26 16:52:18 +0530
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.113 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-11-26 16:50:04 +0530
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.41
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-11-26 16:44:10 +0530
                        message:
                          Bug 16656639: ABORT BACKUP DELAYS
                          
                          For backups with long-running fragment scans, a backup abort does
                          not take effect immediately. These fragment scans are not closed when
                          the abort backup signal is received, so the abort is delayed until
                          the fragment scans complete. The fragment scan status is checked
                          periodically, but the scan is closed only if a file error is found.
                          
                          Modified the scan status check to check the backup error status, as
                          well as the file error status. Since the backup abort sets a backup
                          error, the scan check detects the backup abort and closes the scan.
                          
                          Added a test case which delays fragment scans and starts a backup.
                          Following this, it aborts the backup and then verifies that the backup
                          is aborted immediately.
            ------------------------------------------------------------
            revno: 2876.565.84 [merge]
            committer: Martin Skold <Martin.Skold@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-11-25 10:40:20 +0100
            message:
              Merge from 7.2
                ------------------------------------------------------------
                revno: 2875.603.35 [merge]
                committer: Martin Skold <Martin.Skold@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-11-22 20:08:04 +0100
                message:
                  Merge from 7.1
                    ------------------------------------------------------------
                    revno: 2555.817.112 [merge]
                    committer: Martin Skold <Martin.Skold@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-11-22 20:06:55 +0100
                    message:
                      Merge from 7.0
                        ------------------------------------------------------------
                        revno: 2555.827.40 [merge]
                        committer: Martin Skold <Martin.Skold@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-11-22 20:06:11 +0100
                        message:
                          Merge from 6.3
                            ------------------------------------------------------------
                            revno: 2555.760.86
                            committer: Martin Skold <Martin.Skold@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Fri 2013-11-22 20:05:23 +0100
                            message:
                               Bug#17719439 THREAD INTERLEAVING IN NDB::INIT CAN CAUSE CRASH AT SIGNAL RECEPTION: test case not for embedded
                ------------------------------------------------------------
                revno: 2875.603.34 [merge]
                committer: Martin Skold <Martin.Skold@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-11-22 18:47:34 +0100
                message:
                  Merge from 7.1
                    ------------------------------------------------------------
                    revno: 2555.817.111 [merge]
                    committer: Martin Skold <Martin.Skold@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-11-22 18:42:10 +0100
                    message:
                      Merge from 7.0
                        ------------------------------------------------------------
                        revno: 2555.827.39 [merge]
                        committer: Martin Skold <Martin.Skold@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-11-22 17:05:14 +0100
                        message:
                          Merge from 6.3
                            ------------------------------------------------------------
                            revno: 2555.760.85
                            committer: Martin Skold <Martin.Skold@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Fri 2013-11-22 09:43:16 +0100
                            message:
                              Bug#17719439 THREAD INTERLEAVING IN NDB::INIT CAN CAUSE CRASH AT SIGNAL RECEPTION: Adding check at signal reception that Ndb object is properly initialized before signals are handled, added test case
            ------------------------------------------------------------
            revno: 2876.565.83 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2013-11-22 13:49:42 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.33 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-11-22 13:38:56 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.110 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-11-22 13:36:14 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.38
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-11-22 13:35:35 +0100
                        message:
                          Part2 of fix (the real fix) for bug#17647637: 
                           
                          'NDBTICK_CURRENTMICROSECOND() MUST BE MONOTONIC'
                            
                          Use 'QueryPerformanceCounters' on Windows instead of GetSystemTimeAsFileTime()
                          to implement the NdbTick timers. This will ensure that the Windows timers are 
                          monotonic.
                             
                          Also removes the 'need_monotonic' argument for NdbTick_Init() and
                          NdbCondition_initialize(). Now  we always tries to use a monotonic
                          timer, and print a warning if this was not possible.
                          
                          Furthermore, CLOCK_HIGHRES is introduced as an alternative for 
                          CLOCK_MONOTONIC if that is not available on the platform (Solaris 9).
                          
                          NdbTick.cpp was refactored such that there are not two platform dependent
                          versions of NdbTick_Init() and NdbTick_getCurrentTicks() any more.
            ------------------------------------------------------------
            revno: 2876.565.82 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2013-11-22 13:10:31 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.32 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-11-22 13:08:33 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.109 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-11-22 13:06:28 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.37
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-11-22 13:05:52 +0100
                        message:
                          Another addendum for for NdbTick refactoring.
                          
                          A few platforms complained about 'unknown symbol __cxa_guard_acquire'.
                          Frazer provided the fix in this patch
            ------------------------------------------------------------
            revno: 2876.565.81 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-11-21 14:28:35 +0100
            message:
              Merge 7.2->7.3
                ------------------------------------------------------------
                revno: 2875.603.31 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-11-21 14:26:17 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.108 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-11-21 14:23:40 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.36
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-11-21 14:22:36 +0100
                        message:
                          Addendum patch for the NdbTick refactoring.
                          NdbTick_CurrentMicrosecond() was removed as part of that patch. 
                          However, there were still some debug code lingering around which used that function.
                    ------------------------------------------------------------
                    revno: 2555.817.107
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-11-21 14:16:52 +0100
                    message:
                      Reverted last push as it was not upmerged as mandated in our merge / push procedures
                    ------------------------------------------------------------
                    revno: 2555.817.106
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Wed 2013-11-20 15:36:00 -0800
                    message:
                      Remove junit dependency from clusterj build
                      Clusterj now includes an implementation of junit
                      so it no longer needs to be tested as a dependency
            ------------------------------------------------------------
            revno: 2876.565.80 [merge]
            committer: Frazer Clement <frazer.clement@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-11-20 21:55:33 +0000
            message:
              Merge 7.2->7.3
                ------------------------------------------------------------
                revno: 2875.603.30
                committer: Frazer Clement <frazer.clement@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-11-20 21:45:17 +0000
                message:
                  Bug #17829101 	NDB : BAD MERGE IN SQL/SQL_CLASS.CC IN 7.2
                  
                  Fix compiler warning / error
            ------------------------------------------------------------
            revno: 2876.565.79 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-11-20 11:39:04 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.29 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-11-20 11:37:23 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.105 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Wed 2013-11-20 11:35:02 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.35
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Wed 2013-11-20 11:33:54 +0100
                        message:
                          Addendum patch for part1 of fix for bug#17647637.
                          
                          Changes according to Frazers comments / review.
            ------------------------------------------------------------
            revno: 2876.565.78 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-11-18 13:08:40 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.28 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-11-18 13:05:55 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.104 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-11-18 13:04:19 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.34
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Mon 2013-11-18 13:03:21 +0100
                        message:
                          Addendum to prev commit of Part1 of fix for bug#17647637:
                          
                          On platforms with high precision 'tick' HW, we hit assert in Duration::millisec().
                          Avoid this by alternative (less precision) MilliSec calc on these platforms.
            ------------------------------------------------------------
            revno: 2876.565.77 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-11-18 12:46:38 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.27 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-11-18 11:40:54 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.103 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-11-18 11:07:35 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.33
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Mon 2013-11-18 11:06:19 +0100
                        message:
                          Part1 of fix for bug#17647637: Refactor of NdbTick as prep for real fix
                          
                          My rational for this refactoring is that I want a safer interface
                          for calculating elapsed time and doing timeout calculation.
                          
                          The general pattern in this patch is that the common code sequence:
                          
                          Ex1.
                            NDB_TICKS start = NdbTick_CurrentMillisecond();
                            ... Do some work ...
                            NDB_TICKS end = NdbTicks_CurrentMillisecond();
                            Uint64 elapsed = (end - start);
                          
                          
                          Has been refactored into:
                          Ex2:
                            NDB_TICKS start = NdbTick_getCurrentTicks();
                            ... Do some work ...
                            NDB_TICKS end = NdbTicks_getCurrentTicks();
                            Uint64 elapsed = NdbTick_ElapsedMilliseconds(start,end);
                          
                          
                          As part of this, NDB_TICKS has been redefined into an opague type
                          with the max 'tick resolution' the underlying hardware can provide.
                          No assumption about the resolution of a single tick is exposed
                          through the interface.
                          
                          Thus, it is not any longer an int, and all interval calculations has
                          to go through the new NdbTick_Elapsed() function.
                          (There are also Micro, Nano and Second variants which all can be
                          calculated from the same NDB_TICKS).
                          Asserts has also been included into these functions such that
                          uninitialized 'ticks' are being asserted, and negative intervals
                          resulting from mixing start-end arguments(Bug#17739131)
                          or non-monotonic timers. (Bug#17647637)
                          
                          The problems I try to rectify in this patch are:
                          
                          1) NdbTick.h defined 'typedef Uint64 NDB_TICKS' which was the
                             type returned from NdbTick_CurrentMillisecond(),
                             NdbTick_CurrentMicrosecond, NdbTick_CurrentSecond and
                             NdbTick_CurrentNanosecond. This provided no type
                             safety, and was quite often mixed with using plain Uint64
                             such that it didn't really offer anything of value.
                          
                             Furthermore it did not protect against misuse as:
                          
                              start = NdbTick_CurrentMillisecond()
                              ... do some
                              end = NdbTick_CurrentMicrosecond; << Mix ms / us
                              elapsed = (end - start);
                          
                              or:
                          
                              elapsed = (start - end)  << bug#17739131
                          
                          2) Catch 'elapsed time' calculation based on uninitialized
                             timers - Several bugs had been filed and fixes as a
                             result of this
                          
                          3) Avoid precision loss due to timer rollover:
                             Depending on the underlying platform, the timers
                             might have resolution in the microsecond range.
                             By converting into milliseconds as we usually did, we get
                             less accuracy as this extra precision is truncated, considder:
                          
                              t1 = 1000 us -> truncates to 1ms
                              t2 = 1999 us -> truncates to 1ms
                              t3 = 2001 us -> truncates to 2ms
                              t4 = 2999 us -> truncates to 2ms
                          
                             So calculating intervals between these time ticks using the old
                             NdbTick_CurrentMillisecond(), we will get:
                          
                             (t2-t1) = 0ms ... Sort of correct as it is < 1ms
                             (t3-t2) = 1ms ... Not what you would expect !!!
                          
                          4) Get rid of 'struct MicroSecondTimer', NdbTick_getMicroTimer(),
                             NdbTick_getMicrosPassed() and NdbTick_getMillisecond().
                             This was just another timer mechanism which partly overlapped
                             the functionality provided by NDB_TICKS.
                             In places where both a Milli- and Microseconds timer was required,
                             it even caused that the system clock had to be read twice (ThreadConfig.cpp)
                             or we had to convert between timers represented either as a 'MicroSecondTimer'
                             or a NDB_TICKS.
            ------------------------------------------------------------
            revno: 2876.565.76 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-11-18 09:01:23 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.26 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-11-18 08:58:18 +0100
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.102 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-11-18 08:55:18 +0100
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.32
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Mon 2013-11-18 08:53:41 +0100
                        message:
                          Fix for bug#17739131 JOB SCHEDULER IN SINGLE THREADED NDBD DOES NOT 'YIELD' CPU AS INTENDED
                          
                          When the internal single threaded job scheduler is configured to use
                          'RealTimeScheduler=1', it is intended to temporary lower its scheduling
                          priority to 'normal' every 10ms. This will allow other, non realtime threads,
                          to get a chance to run, and thus avoid CPU starvation of these.
                                
                          However, the arguments to the function which calculated elapsed time
                          singe last yield where swapped. Thus, '0' was always returned as elapsed
                          time from this function, and the CPU never yielded as intended.
            ------------------------------------------------------------
            revno: 2876.565.75 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Sat 2013-11-16 22:55:34 -0800
            message:
              merge 7.2 -> 7.3 Improve reliability of query scan lock test
                ------------------------------------------------------------
                revno: 2875.603.25 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Sat 2013-11-16 22:39:06 -0800
                message:
                  merge 7.1 -> 7.2 Improve reliability of query scan lock test
                    ------------------------------------------------------------
                    revno: 2555.817.101
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-11-15 15:50:11 -0800
                    message:
                      Improve reliability of query scan lock test
            ------------------------------------------------------------
            revno: 2876.565.74 [merge]
            author: kent.boortz@oracle.com
            committer: Kent Boortz <kent.boortz@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Sat 2013-11-16 17:02:17 +0100
            message:
              Merge from mysql-cluster-7.3.3-release
                ------------------------------------------------------------
                revno: 2876.572.4
                tags: mysql-cluster-7.3.3
                committer: magnus.blaudd@oracle.com
                branch nick: cluster-7.3.3-release
                timestamp: Wed 2013-11-13 14:32:03 +0100
                message:
                  Cherrypick fix for "Removed non gpl file docs/mysql.info from community packages" into 7.3.3
                ------------------------------------------------------------
                revno: 2876.572.3
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-cluster-7.3.3-release
                timestamp: Tue 2013-11-12 10:34:11 +0100
                message:
                  BUG#17626525, ensure API_REGREQ is sent once and only once from myself to myself
                ------------------------------------------------------------
                revno: 2876.572.2
                committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                branch nick: mysql-cluster-7.3.3-release
                timestamp: Mon 2013-10-21 19:39:21 +0200
                message:
                  Updated Files list to resolved build error
        ------------------------------------------------------------
        revno: 2876.563.34 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 16:30:16 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.73
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-11-15 12:05:59 +0100
            message:
              Remove tentative patches for BUG 16877045 since they don't seem to work
            ------------------------------------------------------------
            revno: 2876.565.72
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-11-15 11:17:20 +0100
            message:
              Remove unexplainable diff in .bzr-mysql/ vs 5.6.14
            ------------------------------------------------------------
            revno: 2876.565.71 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-11-15 11:01:31 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.24 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Fri 2013-11-15 10:47:59 +0100
                message:
                  Merge 7.2-bug17659501 -> 7.2
                    ------------------------------------------------------------
                    revno: 2875.609.4
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-bug17659501
                    timestamp: Thu 2013-11-14 15:08:28 +0100
                    message:
                      Add small portabilty function 'ndb_localtime_r' which emulates localtime_r on Windows
                    ------------------------------------------------------------
                    revno: 2875.609.3
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-bug17659501
                    timestamp: Tue 2013-11-12 14:37:27 +0100
                    message:
                      Fix spelling error in comment
                    ------------------------------------------------------------
                    revno: 2875.609.2
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-bug17659501
                    timestamp: Tue 2013-11-12 14:30:28 +0100
                    message:
                      Make format_timestamp() safe by always filling the passed string with
                      some value even if localtime_r() fails
                    ------------------------------------------------------------
                    revno: 2875.609.1 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2-bug17659501
                    timestamp: Tue 2013-11-12 13:46:09 +0100
                    message:
                      Merge
                        ------------------------------------------------------------
                        revno: 2875.608.18
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Tue 2013-11-12 13:45:20 +0100
                        message:
                          Rename badly named variable "time" to "timestamp"
                        ------------------------------------------------------------
                        revno: 2875.608.17
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Tue 2013-11-12 11:22:16 +0100
                        message:
                          Remove usage of pointer value returned from localtime_r since the return value may potentially be NULL
                        ------------------------------------------------------------
                        revno: 2875.608.16
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Tue 2013-11-12 11:10:07 +0100
                        message:
                          Call localtime_r outside of strftime() since localtime_r may potentially return NULL.
                        ------------------------------------------------------------
                        revno: 2875.608.15
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 16:31:59 +0100
                        message:
                          Bug#17775772 POTENTIALLY UNINITIALIZED BUFFER IN NDBOUT::PRINT* PRINTOUT
                            - print nothing or just empty newline instead of uninitialized buffer
                            - catch error with an assert in debug compile
                        ------------------------------------------------------------
                        revno: 2875.608.14
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 16:17:56 +0100
                        message:
                          Bug#17775607 POTENTIALLY UNINITIALIZED BUFFER IN VNDBOUT_C PRINTOUT
                           - print empty newline instead of uninitialized buffer
                           - catch error with an assert in debug compile
                           - Mark vndbout_c as static and thus local to implementation
                        ------------------------------------------------------------
                        revno: 2875.608.13
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 16:00:50 +0100
                        message:
                          Replacfe one hard coded constant for buffer size with sizeof
                        ------------------------------------------------------------
                        revno: 2875.608.12
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 15:59:00 +0100
                        message:
                          Remove unthread safe usage of static stack variable  in NDBT_Test and instead use
                          the common timestamp function from Logger in combination with
                          stack buffer
                        ------------------------------------------------------------
                        revno: 2875.608.11
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 15:35:42 +0100
                        message:
                          ndb_waiter
                           - remove the unnecessary WIN32 define
                           - remove superfluous variables
                        ------------------------------------------------------------
                        revno: 2875.608.10
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 15:30:24 +0100
                        message:
                          ndb_waiter
                           - pass length of buffer as argument instead of magically guessing in function
                           - mark getTimeAsString as static
                        ------------------------------------------------------------
                        revno: 2875.608.9
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 15:29:01 +0100
                        message:
                          Move implementation of format_timestamp into Logger.cpp to
                           avoid exposing too much details
                        ------------------------------------------------------------
                        revno: 2875.608.8
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 15:18:01 +0100
                        message:
                          Change ndb_mgm to use new timestamp string function
                        ------------------------------------------------------------
                        revno: 2875.608.7
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 14:57:30 +0100
                        message:
                          Remove the function getTimeAsString and instead call Logger:_format_timestamp direct
                        ------------------------------------------------------------
                        revno: 2875.608.6
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 14:47:37 +0100
                        message:
                          Refactor "format timestamp" code into new static function in Logger to 
                          be used from different places
                        ------------------------------------------------------------
                        revno: 2875.608.5
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 14:32:01 +0100
                        message:
                          Remove the MAX_DATE_TIME_HEADER_LENGTH constant an instead use sizeof to determine how big the buffer for the resulting string is.
                        ------------------------------------------------------------
                        revno: 2875.608.4
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 14:03:09 +0100
                        message:
                          Refactor 'formatTimeStampString'
                           - buffer and lenght of buffer as argument instead of "static"
                           - funtion local to the implementation
                        ------------------------------------------------------------
                        revno: 2875.608.3
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 13:03:29 +0100
                        message:
                          Bug#17750252 USAGE OF LOCALTIME IS UNSAFE IN THREADED CODE
                           - usage of localtime is not safe since some platforms are using a static
                             buffer shared between all threads for storing the result of the conversion.
                           - replace localtime() with localtime_r()+stackbuf
                           - add call to tzset() in ndb_init() to initialize the time conversion information
                             once.  This loads information about current timezone as well as  during which periods
                             daylight saving calculations should be performed.
                        ------------------------------------------------------------
                        revno: 2875.608.2
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 10:44:12 +0100
                        message:
                          Fix memory leak in test programs
                           - save away pointer to the memory returned by 'ndb_log_defaults'
                            so it can be freed with 'ndb_free_defaults' before function returns
                        ------------------------------------------------------------
                        revno: 2875.608.1
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2-bug17659501
                        timestamp: Mon 2013-11-11 09:32:34 +0100
                        message:
                          Test for EventLogger
                           - check multi thread safety
                           - run in mtr
            ------------------------------------------------------------
            revno: 2876.565.70
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-11-15 10:45:33 +0100
            message:
              Backport MCP for bug 17526814 from trunk
            ------------------------------------------------------------
            revno: 2876.565.69
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-11-15 10:30:35 +0100
            message:
              Remove MCP_WL6244 before backporting the official fix from trunk
            ------------------------------------------------------------
            revno: 2876.565.68 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Thu 2013-11-14 21:22:53 +0100
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.23 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Thu 2013-11-14 21:22:04 +0100
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.100 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Thu 2013-11-14 21:20:54 +0100
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.31
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Thu 2013-11-14 21:10:03 +0100
                        message:
                          ndb - refactor: move definitions Local outside and after declaration of IntrusiveList<>
                          
                          Bug #17739717  BUILD WARNING: INVALID USE OF INCOMPLETE TYPE CLASS INTRUSIVELIST
                          
                          gcc 4.8.2 warned about invalid use of incomplete type
                        ------------------------------------------------------------
                        revno: 2555.827.30
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Thu 2013-11-14 21:08:03 +0100
                        message:
                          ndb - test: added missing explicit instance Vector<int>
                           
                          Bug #17739679	 BUILD ERROR: TESTOPERATIONS FAILS TO LINK FOR 7.2 USING GCC 4.8.2
            ------------------------------------------------------------
            revno: 2876.565.67 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Thu 2013-11-14 14:25:06 +0100
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.22 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Thu 2013-11-14 14:24:06 +0100
                message:
                  Merge
                    ------------------------------------------------------------
                    revno: 2875.607.2
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Thu 2013-11-14 14:15:27 +0100
                    message:
                      Bug#16810415 QUERYING NDBINFO.MEMORYUSAGE AS CONNECTION TO CLUSTER IS LOST, CRASHES MYSQLD
                       - apparently rnd_next() can be called also when m_scan_op is NULL. Check m_scan_op and bail out.
                    ------------------------------------------------------------
                    revno: 2875.607.1
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Thu 2013-11-14 13:13:59 +0100
                    message:
                      Bug#17796161 UNUSED DEFINE NDB_MGMAPI IN CONFIGINFO
                       - Remove unused NDB_MGMAPI define since ConfigInfo is no longer compiled
                      as part of mgmapi
            ------------------------------------------------------------
            revno: 2876.565.66 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Tue 2013-11-12 13:40:12 -0800
            message:
              merge have-java clusterj
                ------------------------------------------------------------
                revno: 2875.603.21 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-11-12 13:32:25 -0800
                message:
                  Fix merge of have_java clusterj
                    ------------------------------------------------------------
                    revno: 2555.817.99
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-11-12 12:35:22 -0800
                    message:
                      Put back weird jdk-64 java paths
                ------------------------------------------------------------
                revno: 2875.603.20 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-11-12 13:07:54 -0800
                message:
                  merge have-java for clusterj
                    ------------------------------------------------------------
                    revno: 2555.817.98
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-11-12 11:17:54 -0800
                    message:
                      Improve have_java for clusterj
                    ------------------------------------------------------------
                    revno: 2555.817.97
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-11-07 10:09:14 -0800
                    message:
                      Bug 17720897: build clusterj with java 1.6
        ------------------------------------------------------------
        revno: 2876.563.33 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 13:55:17 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.65 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: bug17626525
            timestamp: Tue 2013-11-12 10:03:13 +0100
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.574.9 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2013-11-08 15:33:37 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.603.19
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Fri 2013-11-08 15:22:56 +0100
                    message:
                      Cherrypick fix for "Removed non gpl file docs/mysql.info from community packages" from 5.5
                ------------------------------------------------------------
                revno: 2876.574.8
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2013-11-08 15:32:20 +0100
                message:
                  Cherrypick fix for "Removed non gpl file docs/mysql.info from community packages" from 5.6
                ------------------------------------------------------------
                revno: 2876.574.7 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2013-11-08 15:12:06 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.603.18 [merge]
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.2
                    timestamp: Fri 2013-11-08 15:08:38 +0100
                    message:
                      Merge
                        ------------------------------------------------------------
                        revno: 2875.602.5
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Fri 2013-11-08 11:59:33 +0100
                        message:
                          Fix minor typo in error messages from NdbApi
                           - s/abortet/aborted/
                        ------------------------------------------------------------
                        revno: 2875.602.4
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Mon 2013-10-28 15:21:43 +0100
                        message:
                          Mark function only used in Watchdog.cpp as static
                        ------------------------------------------------------------
                        revno: 2875.602.3
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Mon 2013-10-28 15:19:16 +0100
                        message:
                          Improve comments in Watchdog.cpp
                        ------------------------------------------------------------
                        revno: 2875.602.2
                        committer: magnus.blaudd@oracle.com
                        branch nick: 7.2
                        timestamp: Mon 2013-10-28 15:12:09 +0100
                        message:
                          Bug#17669652 UNUSED DEFINE SNPRINTF_RETURN_TRUNC
                           - Remove the now unused code
                           - the define used to be set by configure.in
                ------------------------------------------------------------
                revno: 2876.574.6 [merge]
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2013-11-08 15:10:26 +0100
                message:
                  Merge
                    ------------------------------------------------------------
                    revno: 2876.575.1
                    committer: magnus.blaudd@oracle.com
                    branch nick: 7.3
                    timestamp: Mon 2013-10-28 08:29:22 +0100
                    message:
                      Add comments describing the install_mcc and run_mcc targets
                ------------------------------------------------------------
                revno: 2876.574.5 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2013-11-07 15:07:17 +0530
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.603.17 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-11-07 15:05:48 +0530
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.96 [merge]
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Thu 2013-11-07 14:50:57 +0530
                        message:
                          merge 7.0 -> 7.1
                            ------------------------------------------------------------
                            revno: 2555.827.29
                            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Thu 2013-11-07 14:47:10 +0530
                            message:
                              Bug #17737447: NDB_RESTORE_MISC AND NDB_RESTORE_COMPAT_COMPRESSION 
                              FAIL INTERMITTENTLY
                              
                              The ndb_restore_misc mtr test contains a 'nowait' backup command
                              followed by a 'wait started' backup command, where both backups
                              error out since the backup ID is already used. The 'nowait' backup
                              exits before erroring out, so in some cases, the 'wait started'
                              backup starts to execute while the 'nowait' backup is still running.
                              This causes a test failure because the 'wait started' backup exits
                              with a 'backup already started' error instead of a 'file already
                              exists' error.
                              
                              Changed the order of backups in ndb_restore_misc so that the nowait
                              backup is executed last and cannot interfere with the other backups.
                              This also fixes ndb_restore_compat_compression, which fails because
                              it includes ndb_restore_misc.
                ------------------------------------------------------------
                revno: 2876.574.4 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.6-cluster-7.3-merge
                timestamp: Wed 2013-11-06 16:25:58 -0800
                message:
                  merge fix to bug 17200163 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.603.16 [merge]
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2013-11-06 16:09:32 -0800
                    message:
                      merge fix to bug 17200163 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.95
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2013-11-06 14:01:33 -0800
                        message:
                          Add model class to build
                        ------------------------------------------------------------
                        revno: 2555.817.94
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2013-11-06 13:57:48 -0800
                        message:
                          Add model class for Bug 17200163 test case
                        ------------------------------------------------------------
                        revno: 2555.817.93
                        committer: Craig L Russell <Craig.Russell@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2013-11-06 13:38:19 -0800
                        message:
                          Bug 17200163: Ordering only works with the first column of primary key
                          The root cause is handling multiple indices that have the same score but
                            only one of the indices is usable because of ordering. The unusable
                            index occupies the tree set slot and prevents the usable index from
                            being evaluated. The fix is to change the comparator to allow "duplicate"
                            candidate indices in the tree map of indices.
                          
                          AndPredicateImpl:
                            Allow BetweenPredicateImpl to be included in AND terms
                          
                          PredicateImpl:
                            Fix Comparator<CandidateIndexImpl> to allow "duplicates"
                              where the score is the same as other candidate indices
                          
                          schema.sql:
                            Add conversation_summary that exposes the duplicate index bug
                          
                          Test case to verify bug
                ------------------------------------------------------------
                revno: 2876.574.3 [merge]
                committer: Kent Boortz <kent.boortz@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2013-11-06 12:56:28 +0100
                message:
                  Upmerge of the mysql-cluster-7.2.14 build
                    ------------------------------------------------------------
                    revno: 2875.603.15 [merge]
                    author: 
                    committer: Kent Boortz <kent.boortz@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2013-11-06 12:10:38 +0100
                    message:
                      Merge from mysql-cluster-7.2.14-release
                        ------------------------------------------------------------
                        revno: 2875.604.3
                        tags: mysql-cluster-7.2.14
                        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                        branch nick: mysql-cluster-7.2.14-release
                        timestamp: Mon 2013-10-21 20:16:20 +0200
                        message:
                          Updated Files list to resolved build error
                ------------------------------------------------------------
                revno: 2876.574.2 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2013-11-06 10:17:42 +0100
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.603.14 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2013-11-06 10:16:23 +0100
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.92 [merge]
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Wed 2013-11-06 10:14:48 +0100
                        message:
                          Merge 7.0?-> 7.1
                            ------------------------------------------------------------
                            revno: 2555.827.28
                            committer: Ole John Aske <ole.john.aske@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Wed 2013-11-06 09:34:28 +0100
                            message:
                              Fix for bug#17739932 'TESTSUITETIMERS' ARE NOT ALWAYS STARTED/STOPPED, WHICH LEAVES THEM UNDEFINED.
                              
                              This fix moves the start/stop of the testSuiteTimers out of the if-else-branch,such that
                              the timers are also used in case the else-branch is taken
                            ------------------------------------------------------------
                            revno: 2555.827.27
                            committer: Ole John Aske <ole.john.aske@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Wed 2013-11-06 09:31:40 +0100
                            message:
                              Fix for bug#17738720:
                                    
                                ARBITRATION TIMEOUT NOT INITIALIZED -> FALSE WARN 'COULD NOT FIND AN ARBITRATOR'
                                    
                              During arbitrator selection, QMGR runs through the states:
                              (NULL, INIT, FIND, PREP1, PREP2, START, ... and some more)
                                    
                              Currently the arbitration timeout timer is not set until we reach the
                              PREP-states. However, there is a check for timeout already in the
                              FIND-state which may then read uninitialized timestamp.
                              This may cause false warnings:
                                    
                              'Could not find an arbitrator, cluster is not partition-safe'
                                    
                              This fix sets the timeout timer in the 'INIT' state.
                            ------------------------------------------------------------
                            revno: 2555.827.26
                            committer: Ole John Aske <ole.john.aske@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Wed 2013-11-06 09:21:19 +0100
                            message:
                              Fix for bug#17647213 'WARNING...GCP_COMMIT LAG <N> SECONDS' USE INCORRECT 'LAG COUNTER'
                              
                              If Dbdih::checkGcpStopLab() does not detect any progress in GCP_COMMIT,
                              it should print :
                              
                              'WARNING  -- Node ...: GCP Monitor: GCP_COMMIT lag <n> seconds'
                              
                              every 10th second. However, the incorrect 'cnt' variable is
                              used to calculate when multiples of 10seconds had passed. Thus
                              this warning may be printed with odd number of 'lag <n> seconds'.
                ------------------------------------------------------------
                revno: 2876.574.1 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Wed 2013-11-06 10:11:40 +0100
                message:
                  Cleanup of missing merges from 7.1 and up
                    ------------------------------------------------------------
                    revno: 2875.603.13 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Wed 2013-11-06 10:06:42 +0100
                    message:
                      Cleanup of missing merges from 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.91 [merge]
                        committer: Martin Zaun <martin.zaun@oracle.com>
                        branch nick: ndb-7.1
                        timestamp: Sun 2013-11-03 18:21:26 -0800
                        message:
                          merge from ndb-7.1-crund
                            ------------------------------------------------------------
                            revno: 2555.834.3
                            committer: Martin Zaun <martin.zaun@oracle.com>
                            branch nick: ndb-7.1-crund
                            timestamp: Sun 2013-11-03 18:13:53 -0800
                            message:
                              crund -- misc configs, properties, scripts updates
                            ------------------------------------------------------------
                            revno: 2555.834.2
                            committer: Martin Zaun <martin.zaun@oracle.com>
                            branch nick: ndb-7.1-crund
                            timestamp: Sun 2013-11-03 18:11:37 -0800
                            message:
                              crund -- added support for running multiple cpp  loads
                            ------------------------------------------------------------
                            revno: 2555.834.1
                            committer: Martin Zaun <martin.zaun@oracle.com>
                            branch nick: ndb-7.1-crund
                            timestamp: Sun 2013-11-03 18:10:11 -0800
                            message:
                              crund -- added support for running multiple java  loads
                    ------------------------------------------------------------
                    revno: 2875.603.12 [merge]
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Tue 2013-11-05 10:16:59 -0800
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2875.600.17
                        committer: John David Duncan <john.duncan@oracle.com>
                        branch nick: 7.2.16-bugs
                        timestamp: Fri 2013-10-25 09:32:32 -0700
                        message:
                          Uncommitted revno 4075 from 7.3-working and reapplied to 7.2.
                          This patch addresses some of the error handling issues that cause bug#14065851
            ------------------------------------------------------------
            revno: 2876.565.64
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: bug17626525
            timestamp: Fri 2013-11-08 19:57:49 +0100
            message:
              BUG#17626525: Ensuring we don't send more than one API_REGREQ to ourselves, doing so might crash the API
            ------------------------------------------------------------
            revno: 2876.565.63 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17624736-73
            timestamp: Fri 2013-11-01 06:54:09 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.603.11 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17624736-72
                timestamp: Fri 2013-11-01 06:49:15 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.90 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug17624736-71
                    timestamp: Fri 2013-11-01 06:46:56 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.25
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17624736-70
                        timestamp: Thu 2013-10-31 21:18:51 +0200
                        message:
                          bug#17624736 skip-indexes-x1.diff
                          controlled ndb restart
                ------------------------------------------------------------
                revno: 2875.603.10 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17624736-72
                timestamp: Thu 2013-10-31 07:07:50 +0200
                message:
                  merge
            ------------------------------------------------------------
            revno: 2876.565.62 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17624736-73
            timestamp: Thu 2013-10-31 04:47:51 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.573.3 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3
                timestamp: Wed 2013-10-30 15:44:51 +0100
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.606.1 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Wed 2013-10-30 15:35:40 +0100
                    message:
                      cmake - ensures at most one producer of sql_yacc.cc and sql_yacc.h
                        ------------------------------------------------------------
                        revno: 2875.605.2
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.2-bug16877045
                        timestamp: Wed 2013-10-30 15:29:39 +0100
                        message:
                          cmake - remove preexisting sql_yacc.cc and sql_yacc.h if bison is used
                          
                          For in-source builds using a source distribution including sql_yacc.cc
                          and sql_yacc.h, make could start using the output files from source
                          while bison generates new ones.
                          
                          This patch tries to fix bug#16877045
                        ------------------------------------------------------------
                        revno: 2875.605.1
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.2-bug16877045
                        timestamp: Wed 2013-10-30 15:18:54 +0100
                        message:
                          cmake - ensure only one rule produces sql_yacc.cc and sql_yacc.h
                          
                          If one builds out-of-source from a source distribution that includes
                          sql_yacc.cc and sql_yacc.h, there was two rules producing those in
                          the build tree.
                          
                          One copying the files from source and a second that run bison to generate them.
                          
                          This patch makes sures that files are copied only if bison was not used.
                          
                          This should eliminate the risk that some part of make uses the copied files
                          while bison generates new ones.
                ------------------------------------------------------------
                revno: 2876.573.2
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17591531-73
                timestamp: Wed 2013-10-30 10:12:10 +0200
                message:
                  comment
                ------------------------------------------------------------
                revno: 2876.573.1
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17591531-73
                timestamp: Tue 2013-10-29 23:20:38 +0200
                message:
                  bug#17591531 bug17591531.diff
                  prevent dropping index used by a FK
            ------------------------------------------------------------
            revno: 2876.565.61
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17624736-73
            timestamp: Wed 2013-10-30 22:59:27 +0200
            message:
              bug#17624736 skip-fks.diff
              option to skip fks at restart
            ------------------------------------------------------------
            revno: 2876.565.60 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17624736-73
            timestamp: Wed 2013-10-30 22:38:32 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.603.9 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17624736-72
                timestamp: Wed 2013-10-30 20:59:21 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.89 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug17624736-71
                    timestamp: Wed 2013-10-30 20:39:06 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.24
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17624736-70
                        timestamp: Wed 2013-10-30 12:30:09 +0200
                        message:
                          bug#17624736 skip-indexes.diff
                          option to skip indexes at restart
            ------------------------------------------------------------
            revno: 2876.565.59 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-10-28 14:27:07 +0530
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.8
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-10-28 14:12:43 +0530
                message:
                  Bug#17659171: NDB_ERROR_REPORTER FAILS ON RPM BUILDS
                  
                  The ndb_error_reporter mtr test was failing on RPM builds
                  because it could not locate the ndb_error_reporter file.
                  
                  Modified the script to handle the RPM install directory
                  structure while searching for ndb_error_reporter.
            ------------------------------------------------------------
            revno: 2876.565.58 [merge]
            committer: Pratik Patodi<pratik.patodi@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-10-23 11:35:57 +0530
            message:
              merge 7.2->7.3
                ------------------------------------------------------------
                revno: 2875.603.7 [merge]
                committer: Pratik Patodi<pratik.patodi@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-10-23 11:32:27 +0530
                message:
                  merge 7.1->7.2
                    ------------------------------------------------------------
                    revno: 2555.817.88 [merge]
                    committer: Pratik Patodi<pratik.patodi@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-10-22 16:56:32 +0530
                    message:
                      Merge 7.0->7.1
                        ------------------------------------------------------------
                        revno: 2555.827.23 [merge]
                        committer: Pratik Patodi<pratik.patodi@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-10-22 16:53:46 +0530
                        message:
                          Merge 6.3->7.0
                            ------------------------------------------------------------
                            revno: 2555.760.84
                            committer: Pratik Patodi<pratik.patodi@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Tue 2013-10-22 16:50:03 +0530
                            message:
                              Bug#11748591: NO HELP OUTPUT FOR NDBD_REDO_LOG_READER
                              Description: ndbd_redo_log_reader provides no help output, and invoking it with --help produces an error.
                              
                              Fix: Added the --help option to ndbd_redo_log_reader.
        ------------------------------------------------------------
        revno: 2876.563.32 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 13:41:08 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.57 [merge]
            committer: John David Duncan <john.duncan@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-10-22 14:14:57 -0700
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.603.6
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-10-22 11:20:03 -0700
                message:
                  Look for clusterj_jpa in $BINDIR/share/java/
        ------------------------------------------------------------
        revno: 2876.563.31 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 13:38:20 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.56 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Mon 2013-10-21 22:31:32 -0700
            message:
              merge 7.2 -> 7.3 clusterj-openjpa tests
                ------------------------------------------------------------
                revno: 2875.603.5 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-10-21 22:10:52 -0700
                message:
                  merge 7.1 -> 7.2 clusterj-openjpa tests
                    ------------------------------------------------------------
                    revno: 2555.817.87
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-10-21 18:49:27 -0700
                    message:
                      Add BytePK and ShortPK to enhancement
                    ------------------------------------------------------------
                    revno: 2555.817.86
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-10-21 17:46:21 -0700
                    message:
                      Add clusterjpa tests for byte (tinyint) and short (smallint) primary key
                      The functional changes to enable these types to be used as primary keys
                        was implemented in clusterj earlier.
        ------------------------------------------------------------
        revno: 2876.563.30 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 13:30:04 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.55 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Mon 2013-10-21 14:34:53 +0200
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.4 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Mon 2013-10-21 14:33:56 +0200
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.85 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Mon 2013-10-21 14:33:00 +0200
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.22 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Mon 2013-10-21 14:30:58 +0200
                        message:
                          ndb - fix leaking commit ack markers
                          
                          Bug #16944817	 LEAKING COMMIT ACK MARKER CAN LEAD TO SUBSEQUENT CRASHES
                            ------------------------------------------------------------
                            revno: 2555.833.3
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-7.0-bug16944817
                            timestamp: Mon 2013-10-21 14:12:03 +0200
                            message:
                              ndb - make ndbapi send TC_COMMIT_ACK whenever marker flag is set
                              
                              This patch makes sure that ndbapi send TC_COMMIT_ACK even
                              if corresponding ndb-object is gone, even if transaction is closed
                              in short: EVERYTIME a commit with marker flag set.
                            ------------------------------------------------------------
                            revno: 2555.833.2
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-7.0-bug16944817
                            timestamp: Mon 2013-10-21 13:38:56 +0200
                            message:
                              ndb - reuse lingering commit ack markers
                              
                              this patch makes Dbtc (more) resillient to ndbapi failing
                              to TC_COMMIT_ACK transactions during node-failure,
                              and accepts same marker being reported several times
                              (in different node failures) from DBLQH.
                            ------------------------------------------------------------
                            revno: 2555.833.1
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-7.0-bug16944817
                            timestamp: Mon 2013-10-21 13:27:43 +0200
                            message:
                              ndb - test: misuse ndapi, release ndb object before commit_conf/ref
                              
                              A test program that can NOT! run in debug
                              (cause assertions makes it fail...) that sends
                              transactions using async-api and then deletes
                              ndb-object without waiting for reply.
        ------------------------------------------------------------
        revno: 2876.563.29 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 13:27:12 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.54 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Mon 2013-10-21 14:25:30 +0200
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.3 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Mon 2013-10-21 14:24:33 +0200
                message:
                  merge 7.2.14-release to 7.2
                    ------------------------------------------------------------
                    revno: 2875.604.2
                    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                    branch nick: mysql-cluster-7.2.14-release
                    timestamp: Fri 2013-10-18 19:35:29 +0200
                    message:
                      Removing the man pages to resolve build install error
                    ------------------------------------------------------------
                    revno: 2875.604.1
                    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                    branch nick: mysql-cluster-7.2.14-release
                    timestamp: Fri 2013-10-18 18:13:14 +0200
                    message:
                      Applying spec file patch for Cluster release
            ------------------------------------------------------------
            revno: 2876.565.53 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Mon 2013-10-21 14:23:10 +0200
            message:
              merge 7.3.3-release to 7.3
                ------------------------------------------------------------
                revno: 2876.572.1
                committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                branch nick: mysql-cluster-7.3.3-release
                timestamp: Mon 2013-10-21 10:31:54 +0200
                message:
                  Updated spec file for Cluster release
            ------------------------------------------------------------
            revno: 2876.565.52 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2013-10-18 23:08:43 +0200
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.603.2
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Fri 2013-10-18 23:07:41 +0200
                message:
                  raise version to 5.5.34-ndb-7.2.15 again
                ------------------------------------------------------------
                revno: 2875.603.1
                committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-10-18 17:56:04 +0200
                message:
                  Updated spec file for 7.2 Cluster release
            ------------------------------------------------------------
            revno: 2876.565.51 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2013-10-18 23:05:46 +0200
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.156
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Fri 2013-10-18 23:03:57 +0200
                message:
                  raise version to 5.5.34-ndb-7.2.15 again
        ------------------------------------------------------------
        revno: 2876.563.28 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2014-01-07 13:04:43 +0100
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.50
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2013-10-18 23:04:44 +0200
            message:
              raise version to 5.6.14-ndb-7.3.4 again
            ------------------------------------------------------------
            revno: 2876.565.49 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2013-10-18 17:01:06 +0200
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.155 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Fri 2013-10-18 17:00:08 +0200
                message:
                  null merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.84 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Fri 2013-10-18 16:59:14 +0200
                    message:
                      null merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.21 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Fri 2013-10-18 16:58:21 +0200
                        message:
                          null merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.760.83
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Fri 2013-10-18 16:49:18 +0200
                            message:
                              raise version to 5.1.72-ndb-6.3.54
                        ------------------------------------------------------------
                        revno: 2555.827.20
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Fri 2013-10-18 16:51:06 +0200
                        message:
                          raise version to 5.1.72-ndb-7.0.41
                    ------------------------------------------------------------
                    revno: 2555.817.83
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Fri 2013-10-18 16:52:25 +0200
                    message:
                      raise version to 5.1.72-ndb-7.1.30
                ------------------------------------------------------------
                revno: 2875.571.154
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Fri 2013-10-18 16:53:58 +0200
                message:
                  raise version to 5.5.34-ndb-7.2.15
                ------------------------------------------------------------
                revno: 2875.571.153 [merge]
                committer: Martin Skold <Martin.Skold@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-10-18 11:44:43 +0200
                message:
                  Merge from 7.1
                    ------------------------------------------------------------
                    revno: 2555.817.82 [merge]
                    committer: Martin Skold <Martin.Skold@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-10-18 09:51:41 +0200
                    message:
                      Merge from 7.0
                        ------------------------------------------------------------
                        revno: 2555.827.19
                        committer: Martin Skold <Martin.Skold@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-10-18 09:46:05 +0200
                        message:
                           Bug#16890703 MYSQLD STUCK IN OPN TABLES ON SELECT FROM RESTORED TABLE WHERE DB IS NOT THERE: Added error check for missing database directory, added testcase
            ------------------------------------------------------------
            revno: 2876.565.48
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Fri 2013-10-18 16:55:06 +0200
            message:
              raise version to 5.6.14-ndb-7.3.4
            ------------------------------------------------------------
            revno: 2876.565.47
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-10-18 14:54:30 +0200
            message:
              MCP backport of fix for bug16877045
        ------------------------------------------------------------
        revno: 2876.563.27
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2013-12-20 10:39:14 +0100
        message:
          Cherrypick fix for correct number of planned tests in pfs_user-oom-t
        ------------------------------------------------------------
        revno: 2876.563.26
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2013-12-20 10:26:39 +0100
        message:
          Fix problem with EXPLAIN of mixedcase table
          
           - The EXTENDED output from EXPLAIN is dependent on lowercase setting
           - Add --replace_result so that the output from EXPLAIN with mixedcase table
             is converted to lowercase on all platforms
        ------------------------------------------------------------
        revno: 2876.563.25
        committer: magnus.blaudd@oracle.com
        branch nick: merge_5.7.3_2_trunk-cluster
        timestamp: Thu 2013-12-19 21:17:24 +0100
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
           - The output from EXPLAIN has been changed to always
             included EXTENDED and PARTITIONS information. Thus
             all .result files using EXPLAIN has to be updated
             with additional output. The result set
             contains an additional column called partitions
             and the EXPLAIN command always shows a warning describing
             the query further.
           - When using EXPLAIN EXTENDED or EXPLAIN PARTITIONS
             a deprectaion warning will be printed. Include the
             warning in the .result files.
           - Update a few counter variables in ndb_join_pushdown_* which
             has changed slightly.
        ------------------------------------------------------------
        revno: 2876.563.24 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: merge_5.7.3_2_trunk-cluster
        timestamp: Thu 2013-12-19 20:14:29 +0100
        message:
          Merge 5.7.3 -> trunk-cluster
        ------------------------------------------------------------
        revno: 2876.563.23
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2013-12-19 14:17:29 +0100
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
           - avoid nameclash which occurs when both "Innodb memcache plugin" and "Ndb memcached" was built. Both components added a memcached_utilties library.
           - Fix problem by prefixing ndb_ to the memcached_utilities library built by "Ndb memcached"
        ------------------------------------------------------------
        revno: 2876.563.22
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2013-12-19 14:12:20 +0100
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
            - stricter handling of which libevent library to use when building
              the memcached bundled with NDB
            - after this patch we never build or use libevent bundled with NDB
            - like comment below suggests, building WITH_LIBEVENT=system
              seems broken. We probably never build that way.
           - remove the libevent which was bundled in storage/ndb/memcache/extra/
        ------------------------------------------------------------
        revno: 2876.563.21
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Thu 2013-12-19 14:07:51 +0100
        message:
          Cherrypick merge fix for building libevent for other than Innodb memcached
        ------------------------------------------------------------
        revno: 2876.563.20
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2013-12-17 09:56:03 +0100
        message:
          WL#6815 Adapt MySQL Cluster to 5.7
          
           - Move librt link directive to where the memcached is
             built since bundled or system libevent does not autolink with rt
        ------------------------------------------------------------
        revno: 2876.563.19
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2013-10-22 16:10:22 +0200
        message:
          Fifth follow up fix for "Bug17258782 MASTER DIDN'T WRITE ERROR MESSAGE IN THE
          LOG FILE FOR INCIDENT"
                      
            - ignore the warnings produced by ndbcluster's binlog
              thread when the mysqld is restarted. These warnings
              are expected.
        ------------------------------------------------------------
        revno: 2876.563.18
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2013-10-21 11:07:57 +0200
        message:
          Fourth  follow up fix for "Bug17258782 MASTER DIDN'T WRITE ERROR MESSAGE IN THE LOG FILE FOR INCIDENT"
                
                 - ignore the warnings produced by ndbcluster's binlog thread when the mysqld is restarted. These warnings are expected.
        ------------------------------------------------------------
        revno: 2876.563.17
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2013-10-21 10:55:26 +0200
        message:
          Disable ndb_fk_addnode due to bug17400320
        ------------------------------------------------------------
        revno: 2876.563.16 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2013-10-18 14:29:54 +0200
        message:
          Merge 7.3 -> trunk-cluster
            ------------------------------------------------------------
            revno: 2876.565.46 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-10-18 12:17:20 +0200
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.602.1
                committer: magnus.blaudd@oracle.com
                branch nick: 7.2
                timestamp: Fri 2013-10-18 12:16:06 +0200
                message:
                  Cherrypick fix for removing zombie .THIS files
            ------------------------------------------------------------
            revno: 2876.565.45 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Fri 2013-10-18 11:21:00 +0200
            message:
              Merge
                ------------------------------------------------------------
                revno: 2876.571.1
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Fri 2013-10-11 14:20:32 +0200
                message:
                  Revert tentative patch for BUG#16877045
                   - it didn't work
            ------------------------------------------------------------
            revno: 2876.565.44 [merge]
            tags: clone-mysql-5.6.14-ndb-7.3.3-src-build
            committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-10-17 13:36:46 +0530
            message:
              Merge Patch for BUG#11746263 from 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.152 [merge]
                tags: clone-mysql-5.5.34-ndb-7.2.14-src-build
                committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-10-17 13:31:07 +0530
                message:
                  Merge Patch for BUG#11746263 from 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.81 [merge]
                    tags: clone-mysql-5.1.72-ndb-7.1.29-src-build, mysql-cluster-7.1.29
                    committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-10-17 12:56:15 +0530
                    message:
                      Merge Patch for BUG#11746263 from 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.18
                        tags: clone-mysql-5.1.72-ndb-7.0.40-src-build
                        committer: Lakshmi Narayanan Sreethar<lakshmi.narayanan.sreethar@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-10-17 12:51:08 +0530
                        message:
                          BUG#11746263 : REMOVING OR REPLACING THE 'MASTER'-WORD FROM SHOW OUTPUT
                          
                          Modified the output of 'ndb_mgm -e show' command to print '*' instead of the word 'Master'.
                          Also updated a test case to reflect the changes.
            ------------------------------------------------------------
            revno: 2876.565.43 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: 73
            timestamp: Thu 2013-10-17 13:24:15 +0530
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.151 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: 72
                timestamp: Thu 2013-10-17 13:20:44 +0530
                message:
                  null merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.80
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-10-15 10:29:51 -0700
                    message:
                      Re-enable clusterjpa tests that failed due to change in supported version of openjpa
            ------------------------------------------------------------
            revno: 2876.565.42 [merge]
            committer: Jan Wedvik <jan.wedvik@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-10-15 15:08:34 +0200
            message:
              Merge 7.2->7.3.
                ------------------------------------------------------------
                revno: 2875.571.150 [merge]
                committer: Jan Wedvik <jan.wedvik@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-10-15 14:58:44 +0200
                message:
                  Merge 7.1->7.2.
                    ------------------------------------------------------------
                    revno: 2555.817.79 [merge]
                    committer: Jan Wedvik <jan.wedvik@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Tue 2013-10-15 14:53:38 +0200
                    message:
                      Merge 7.0->7.1
                        ------------------------------------------------------------
                        revno: 2555.827.17
                        committer: Jan Wedvik <jan.wedvik@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Tue 2013-10-15 13:17:58 +0200
                        message:
                          This is a fix for Bug#16426805 (ON FAILURE OF MGMT NODE, STATUS DOESNT
                          GET UPDATED FOR MORE THAN 5 MIN). The problem was that the variable
                          holding the interval (in milliseconds) between MGMD to MGMD heart beat
                          messages was never assigned a suitable value. Instead it was set to 0
                          minus a small constant, such that it wrapped around to a value close
                          to 2^32. To fix this, a new config variable has been defined,
                          "HeartbeatIntervalMgmdMgmd", with a default value of 1500ms. An mgmd
                          process will then detect the death of another mgmd after about 6000ms.
                          
                          In addition this commit adds some extra comments and asserts, and removes
                          a useless destructor.
            ------------------------------------------------------------
            revno: 2876.565.41 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Tue 2013-10-15 14:33:32 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.571.149 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Tue 2013-10-15 14:27:00 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.78 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Mon 2013-10-14 19:10:52 +0200
                    message:
                      merge
                ------------------------------------------------------------
                revno: 2875.571.148 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Mon 2013-10-14 19:11:58 +0200
                message:
                  merge
            ------------------------------------------------------------
            revno: 2876.565.40 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Mon 2013-10-14 19:13:55 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.570.10 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Mon 2013-10-14 15:19:49 +0200
                message:
                  Merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.600.16 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Mon 2013-10-14 15:16:11 +0200
                    message:
                      Merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.830.11 [merge]
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.1
                        timestamp: Mon 2013-10-14 15:14:01 +0200
                        message:
                          Merge 7.0 -> 7.1
            ------------------------------------------------------------
            revno: 2876.565.39 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Mon 2013-10-14 19:06:52 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.571.147 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Mon 2013-10-14 19:05:41 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.77 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Mon 2013-10-14 19:01:49 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.16 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Mon 2013-10-14 19:01:10 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.10
                            committer: Ole John Aske <ole.john.aske@oracle.com>
                            branch nick: mysql-5.1-telco-7.0
                            timestamp: Mon 2013-10-14 15:12:48 +0200
                            message:
                              Added commented regarding bug#13870457 which we have decided to not fix 
                              as the relevant code is now unused (No spinlock usage in MT-scheduler anymore)
            ------------------------------------------------------------
            revno: 2876.565.38 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Mon 2013-10-14 11:02:31 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.571.146 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Mon 2013-10-14 11:00:54 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.76 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Mon 2013-10-14 10:59:19 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.15
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Mon 2013-10-14 10:57:27 +0200
                        message:
                          Fix spelling error
                        ------------------------------------------------------------
                        revno: 2555.827.14 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Mon 2013-10-14 09:34:17 +0200
                        message:
                          merge
                        ------------------------------------------------------------
                        revno: 2555.827.13 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Fri 2013-10-11 13:57:57 +0200
                        message:
                          merge
                    ------------------------------------------------------------
                    revno: 2555.817.75 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Mon 2013-10-14 09:34:43 +0200
                    message:
                      merge
                    ------------------------------------------------------------
                    revno: 2555.817.74 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Fri 2013-10-11 13:58:38 +0200
                    message:
                      merge
                ------------------------------------------------------------
                revno: 2875.571.145 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Mon 2013-10-14 09:35:21 +0200
                message:
                  merge
                ------------------------------------------------------------
                revno: 2875.571.144 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Fri 2013-10-11 14:10:03 +0200
                message:
                  merge
            ------------------------------------------------------------
            revno: 2876.565.37 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Mon 2013-10-14 09:37:50 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.570.9 [merge]
                committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                branch nick: 7.3
                timestamp: Sat 2013-10-12 01:33:41 +0200
                message:
                  merge 7.2->7.3
                    ------------------------------------------------------------
                    revno: 2875.600.15 [merge]
                    committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                    branch nick: 7.2
                    timestamp: Sat 2013-10-12 00:59:53 +0200
                    message:
                      merge 7.1->7.2
                        ------------------------------------------------------------
                        revno: 2555.830.10 [merge]
                        committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                        branch nick: 7.1
                        timestamp: Sat 2013-10-12 00:30:22 +0200
                        message:
                          merge 7.0->7.1
                            ------------------------------------------------------------
                            revno: 2555.829.9
                            committer: Maitrayi Sabaratnam <maitrayi.sabaratnam@oracle.com>
                            branch nick: 7.0
                            timestamp: Sat 2013-10-12 00:12:05 +0200
                            message:
                              bug#16329082 NDBEVENTIMPL::SETTABLE(NDBDICTIONARY::TABLE CONST&), SEGFAULTS ON NULL
                        ------------------------------------------------------------
                        revno: 2555.830.9 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 21:27:23 +0200
                        message:
                          merge
                ------------------------------------------------------------
                revno: 2876.570.8 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2013-10-11 21:24:19 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.14 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 21:23:14 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2875.601.1 [merge]
                        committer: Martin Skold <Martin.Skold@oracle.com>
                        branch nick: mysql-5.5-cluster-7.2
                        timestamp: Fri 2013-10-11 19:46:03 +0200
                        message:
                          Merge from 7.1
                            ------------------------------------------------------------
                            revno: 2555.832.1 [merge]
                            committer: Martin Skold <Martin.Skold@oracle.com>
                            branch nick: mysql-5.1-telco-7.1
                            timestamp: Fri 2013-10-11 19:41:35 +0200
                            message:
                              Merge from 7.0
                    ------------------------------------------------------------
                    revno: 2875.600.13 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 21:21:17 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.8 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 21:19:08 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.8
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 21:18:18 +0200
                            message:
                              Activate check of HAVE_SOLARIS_AFFINITY on CMake builds, had a missing space in the wrong place
                            ------------------------------------------------------------
                            revno: 2555.829.7 [merge]
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 21:10:24 +0200
                            message:
                              merge
                                ------------------------------------------------------------
                                revno: 2555.831.2
                                committer: Martin Skold <Martin.Skold@oracle.com>
                                branch nick: mysql-5.1-telco-7.0
                                timestamp: Fri 2013-10-11 19:40:05 +0200
                                message:
                                  Bug #17066009 HIGH MYSQLD PROCESS MEMORY USAGE.: Updated test result due to new variable
                                ------------------------------------------------------------
                                revno: 2555.831.1
                                committer: Martin Skold <Martin.Skold@oracle.com>
                                branch nick: mysql-5.1-telco-7.0
                                timestamp: Fri 2013-10-11 17:41:59 +0200
                                message:
                                  Bug #17066009 HIGH MYSQLD PROCESS MEMORY USAGE.: Adding new parameter for limiting the maximum size of the event buffer and terminate the mysqld if the size is exceeded. This will later be replaced with a less intrusive handling of when the buffer is exhaused by directly injecting a LOST_EVENT into the binlog (without terminating the mysqld). Terminating the mysqld will have the same effect since a LOST_EVENT will be injected as well when the mysqld is restarted.
                ------------------------------------------------------------
                revno: 2876.570.7 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2013-10-11 20:37:41 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.12 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 20:36:24 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.7 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 20:35:26 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.6
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 20:34:46 +0200
                            message:
                              More fixes of Solaris locking of cpusets
                ------------------------------------------------------------
                revno: 2876.570.6 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2013-10-11 15:44:16 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.11 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 15:42:59 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.6 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 15:41:50 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.5
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 15:39:35 +0200
                            message:
                              Fix of Solaris for cpusets
                    ------------------------------------------------------------
                    revno: 2875.600.10 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 15:00:55 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.5 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 14:59:33 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.4
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 14:58:48 +0200
                            message:
                              Separate max number of processor sets from constant indicating not part of cpuset
            ------------------------------------------------------------
            revno: 2876.565.36 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Fri 2013-10-11 14:14:30 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.570.5 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2013-10-11 13:35:54 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.9 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 13:34:23 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.4 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 13:33:04 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.3
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 13:32:03 +0200
                            message:
                              Fixes for Solaris and locking to CPUs for cpusets
                ------------------------------------------------------------
                revno: 2876.570.4 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Fri 2013-10-11 09:43:56 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.8 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-11 09:41:25 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.3 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mysql-cluster-7.1
                        timestamp: Fri 2013-10-11 09:39:45 +0200
                        message:
                          merge
                ------------------------------------------------------------
                revno: 2876.570.3 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: push_rt_73
                timestamp: Fri 2013-10-11 07:42:54 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.7 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: push_rt_72
                    timestamp: Fri 2013-10-11 07:32:53 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.830.2 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: push_rt_71
                        timestamp: Fri 2013-10-11 07:29:07 +0200
                        message:
                          merge
                        ------------------------------------------------------------
                        revno: 2555.830.1
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: push_rt_71
                        timestamp: Fri 2013-10-11 07:28:03 +0200
                        message:
                          Fix for bug#16961971, NDB : REALTIMESCHEDULER REGRESSION ON NDBMTD, introducing more options to ThreadConfig configuration variable for realtime, spintime and cpusets
                    ------------------------------------------------------------
                    revno: 2875.600.6
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: push_rt_72
                    timestamp: Fri 2013-10-11 07:29:50 +0200
                    message:
                      Fix for bug#16961971, NDB : REALTIMESCHEDULER REGRESSION ON NDBMTD, introducing more options to ThreadConfig configuration variable for realtime, spintime and cpusets
                ------------------------------------------------------------
                revno: 2876.570.2
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: push_rt_73
                timestamp: Fri 2013-10-11 07:42:00 +0200
                message:
                  Fix for bug#16961971, NDB : REALTIMESCHEDULER REGRESSION ON NDBMTD, introducing more options to ThreadConfig configuration variable for realtime, spintime and cpusets
                ------------------------------------------------------------
                revno: 2876.570.1 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: merge_72
                timestamp: Fri 2013-10-11 07:37:09 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2875.600.5
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-10-10 11:45:02 -0700
                    message:
                      Better error message for mysql bug#70403 (but may have portability problems)
                    ------------------------------------------------------------
                    revno: 2875.600.4
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-10-10 11:35:55 -0700
                    message:
                      Once more on bug#14650456 / CMake writes to source tree
                    ------------------------------------------------------------
                    revno: 2875.600.3
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-10-10 11:17:59 -0700
                    message:
                      Another pass at bug#14650456 -- "CMake makes illegal changes to source"
                    ------------------------------------------------------------
                    revno: 2875.600.2
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-10-10 10:14:14 -0700
                    message:
                      Tentative fix for bug#14650456
                    ------------------------------------------------------------
                    revno: 2875.600.1
                    committer: bernd.ocklin@oracle.com
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-10-10 15:29:56 +0200
                    message:
                      fix rpms by adding ndb_move_data
            ------------------------------------------------------------
            revno: 2876.565.35
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Fri 2013-10-11 13:49:06 +0200
            message:
              Fixed use of wrong API for locking CPUs
            ------------------------------------------------------------
            revno: 2876.565.34 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Fri 2013-10-11 12:24:43 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.571.143
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Fri 2013-10-11 12:23:28 +0200
                message:
                  Fixed extraneous endif-statement
            ------------------------------------------------------------
            revno: 2876.565.33 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Fri 2013-10-11 10:16:07 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.571.142 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Fri 2013-10-11 10:06:17 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.73 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Fri 2013-10-11 10:00:34 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.12
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Fri 2013-10-11 09:59:37 +0200
                        message:
                          Moved pending_signals-methods to proper place
                        ------------------------------------------------------------
                        revno: 2555.827.11 [merge]
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Fri 2013-10-11 09:55:31 +0200
                        message:
                          merge
                            ------------------------------------------------------------
                            revno: 2555.829.2
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: mysql-cluster-7.0
                            timestamp: Fri 2013-10-11 09:37:57 +0200
                            message:
                              Added files that got lost in translation
                            ------------------------------------------------------------
                            revno: 2555.829.1
                            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                            branch nick: push_rt_70
                            timestamp: Fri 2013-10-11 07:26:26 +0200
                            message:
                              Fix for bug#16961971, NDB : REALTIMESCHEDULER REGRESSION ON NDBMTD, introducing more options to ThreadConfig configuration variable for realtime, spintime and cpusets
            ------------------------------------------------------------
            revno: 2876.565.32 [merge]
            committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
            branch nick: mt_opt_73
            timestamp: Thu 2013-10-10 14:49:42 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2875.571.141 [merge]
                committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                branch nick: mt_opt_72
                timestamp: Thu 2013-10-10 14:30:48 +0200
                message:
                  merge
                    ------------------------------------------------------------
                    revno: 2555.817.72 [merge]
                    committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                    branch nick: mt_opt_71
                    timestamp: Thu 2013-10-10 14:25:36 +0200
                    message:
                      merge
                        ------------------------------------------------------------
                        revno: 2555.827.10
                        committer: Mikael Ronstrom <mikael.ronstrom@oracle.com>
                        branch nick: mt_opt_70
                        timestamp: Thu 2013-10-10 14:24:07 +0200
                        message:
                          Some more docs of mt.cpp and also some local optimisatons there
            ------------------------------------------------------------
            revno: 2876.565.31 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-10-10 13:48:58 +0530
            message:
              merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.140 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-10-10 13:43:05 +0530
                message:
                  merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.71 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-10-10 13:37:41 +0530
                    message:
                      merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.9
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Thu 2013-10-10 13:32:32 +0530
                        message:
                          Bug#11752837 NDB_MGM -E " START BACKUP WAIT STARTED" NOT WORKS AS DOCUMENTED
                          
                          When a 'start backup wait started' command was executed on the 
                          command-line using the '-e' option of ndb_mgm, the client did not 
                          exit until the backup was completed.
                          
                          Modified ndb_mgm for a 'wait started' backup so that it exits 
                          immediately after printing 'Backup started' instead of waiting for 
                          backup completion.
            ------------------------------------------------------------
            revno: 2876.565.30 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Tue 2013-10-08 21:46:07 +0200
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.569.2 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Tue 2013-10-08 21:42:49 +0200
                message:
                  merge 7.2 -> 7.3
                    ------------------------------------------------------------
                    revno: 2875.571.139 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.2
                    timestamp: Tue 2013-10-08 21:38:26 +0200
                    message:
                      merge 7.1 -> 7.2
                        ------------------------------------------------------------
                        revno: 2555.817.70 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.1
                        timestamp: Tue 2013-10-08 21:37:29 +0200
                        message:
                          merge 7.0 -> 7.1
                            ------------------------------------------------------------
                            revno: 2555.827.8
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-7.0
                            timestamp: Tue 2013-10-08 21:35:23 +0200
                            message:
                              ndb - add missing initialisation of reorg flag in dbtup
                              
                              Bug #17036744    ONLINE REORGANIZE FAIL WITH FOREIGN KEY IN USE
                              
                              Added missing initialisation of KeyReqStruct::m_reorg in
                              Dbtup::execFIRE_TRIG_REQ and Dbtup::execTUP_COMMITREQ
                              
                              Before this change, check_fire_trigger() and check_fire_reorg() had undefined
                              return values for some calls during reorganise (commit/complete).
                              
                              This implied that readTriggerInfo() was called or not called (from executeTrigger)
                              unpredictable under some circumstances.
                ------------------------------------------------------------
                revno: 2876.569.1
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.3-merge
                timestamp: Tue 2013-10-08 21:40:55 +0200
                message:
                  ndb - test: add test of reorganize on tables using foreign keys
            ------------------------------------------------------------
            revno: 2876.565.29 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Mon 2013-10-07 18:27:18 +0530
            message:
              merge patch for Bug #16630410
                ------------------------------------------------------------
                revno: 2875.571.138 [merge]
                committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Mon 2013-10-07 17:53:08 +0530
                message:
                  merge patch for Bug #16630410
                    ------------------------------------------------------------
                    revno: 2555.817.69 [merge]
                    committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Mon 2013-10-07 17:33:32 +0530
                    message:
                      merge patch for Bug #16630410
                        ------------------------------------------------------------
                        revno: 2555.827.7
                        committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Mon 2013-10-07 17:29:00 +0530
                        message:
                          Bug #16630410 NDB:LCP FRAGMENT SCAN WATCHDOG SHOULD BE CONFIGURABLE
                          
                          The LCP fragment scan watchdog periodically checks for lack of
                          progress in a fragment scan performed as part of a local checkpoint.
                          The watchdog was hard-coded to shut down the node if there was no
                          progress for 60 seconds.
                          
                          Added a configuration parameter LcpScanProgressTimeout to set the
                          maximum time for which the local checkpoint can be stalled before
                          the LCP fragment scan watchdog shuts down the node. A parameter
                          value of 0 disables the LCP fragment scan watchdog.
                          
                          Added a unit test to set the parameter value to 0 and verify that
                          the LCP fragment scan watchdog is disabled.
            ------------------------------------------------------------
            revno: 2876.565.28 [merge]
            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
            branch nick: 73
            timestamp: Mon 2013-10-07 18:21:26 +0530
            message:
              null merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.137 [merge]
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-10-04 08:24:09 -0700
                message:
                  local merge
                    ------------------------------------------------------------
                    revno: 2875.599.2
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Fri 2013-10-04 08:16:45 -0700
                    message:
                      Merge from 7.3 - fix ndb_memcache type_binary test
                    ------------------------------------------------------------
                    revno: 2875.599.1
                    committer: John David Duncan <john.duncan@oracle.com>
                    branch nick: mysql-5.5-cluster-7.2
                    timestamp: Thu 2013-10-03 21:16:35 -0700
                    message:
                      Merge fix for bug#17532443 from 7.3 to 7.2
            ------------------------------------------------------------
            revno: 2876.565.27
            committer: John David Duncan <john.duncan@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2013-10-04 07:47:07 -0700
            message:
              stop type_binary test from causing sporadic failures of type_char test
            ------------------------------------------------------------
            revno: 2876.565.26 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Fri 2013-10-04 14:18:43 +0200
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.136 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Fri 2013-10-04 14:16:07 +0200
                message:
                  Merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.68 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Fri 2013-10-04 14:13:37 +0200
                    message:
                      Merge 7.0 > 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.6
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Fri 2013-10-04 11:19:26 +0200
                        message:
                          Improved fix for bug#15907515 to improve type safety as suggested by Mauritz / Magnus.
                          
                          No functional changes 
                        ------------------------------------------------------------
                        revno: 2555.827.5 [merge]
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17235051-70
                        timestamp: Thu 2013-10-03 22:07:21 +0300
                        message:
                          merge
            ------------------------------------------------------------
            revno: 2876.565.25 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Thu 2013-10-03 23:12:50 -0700
            message:
              merge 7.2 -> 7.3 implement clusterj byte primary key
                ------------------------------------------------------------
                revno: 2875.571.135 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-10-03 23:06:14 -0700
                message:
                  merge 7.1 -> 7.2 implement clusterj byte primary key
                    ------------------------------------------------------------
                    revno: 2555.817.67
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-10-03 21:56:26 -0700
                    message:
                      Implement clusterj byte primary key
                      
                      AbstractDomainFieldHandlerImpl.java:
                        add object operation handlers for byte key
                      
                      DomainFieldHandlerImpl.java:
                        support tinyint mapped to byte for primary key columns
                      
                      DomainTypeHandlerImpl.java:
                        allow byte.class and Byte.class to be used for key column values
                      
                      PartitionKey.java:
                        support byte primary key
                      
                      clusterj-test/Makefile.am:
                        test case for byte primary key
                      
                      DynamicBytePKTest.java:
                        test that instances with byte pk can be inserted, deleted, and queried
                      
                      PartitionKeyImpl.java:
                        implement primary key byte columns
                    ------------------------------------------------------------
                    revno: 2555.817.66 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug17235051-71
                    timestamp: Thu 2013-10-03 21:41:10 +0300
                    message:
                      merge
                ------------------------------------------------------------
                revno: 2875.571.134
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-10-03 21:04:27 -0700
                message:
                  Merge fix for bug#17532442 from 7.3 to 7.2
            ------------------------------------------------------------
            revno: 2876.565.24 [merge]
            committer: John David Duncan <john.duncan@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Thu 2013-10-03 21:01:18 -0700
            message:
              Local merge
                ------------------------------------------------------------
                revno: 2876.568.1
                committer: John David Duncan <john.duncan@oracle.com>
                branch nick: bigfish-7.3
                timestamp: Thu 2013-10-03 20:50:59 -0700
                message:
                  Fix bug#17532443 -- improper storage of binary values 
            ------------------------------------------------------------
            revno: 2876.565.23 [merge]
            committer: Craig L Russell <Craig.Russell@oracle.com>
            branch nick: mysql-5.6-cluster-7.3-merge
            timestamp: Thu 2013-10-03 11:52:51 -0700
            message:
              merge 7.2 -> 7.3 implement clusterj short primary key
                ------------------------------------------------------------
                revno: 2875.571.133 [merge]
                committer: Craig L Russell <Craig.Russell@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Thu 2013-10-03 11:45:13 -0700
                message:
                  merge 7.1 -> 7.2 implement clusterj short primary key
                    ------------------------------------------------------------
                    revno: 2555.828.3
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-10-03 09:13:56 -0700
                    message:
                      Don't drop tables that weren't created
                    ------------------------------------------------------------
                    revno: 2555.828.2
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Thu 2013-10-03 07:32:51 -0700
                    message:
                      Try to avoid Got error 904 'Out of fragment records (increase MaxNoOfOrderedIndexes)' from NDB
                      Can't create table 'test.#sql-547b_16' (errno: 136)
                    ------------------------------------------------------------
                    revno: 2555.828.1
                    committer: Craig L Russell <Craig.Russell@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Wed 2013-10-02 20:27:28 -0700
                    message:
                      Implement clusterj short primary key
                      
                      mysql-test/suite/ndb/t/clusterj.test:
                        drop tables created during test
                      
                      AbstractDomainFieldHandlerImpl.java:
                        add object operation handlers for short key
                      
                      DomainFieldHandlerImpl.java:
                        support smallest mapped to short for primary key columns
                      
                      DomainTypeHandlerImpl.java:
                        allow short.class and Short.class to be used for key column values
                      
                      PartitionKey.java:
                        support short primary key
                      
                      clusterj-test/Makefile.am:
                        test case for short primary key
                      
                      AbstractClusterJModelTest.java:
                        allow graceful failure if test model class is not defined
                      
                      DynamicShortPKTest.java:
                        test that instances with short pk can be inserted, deleted, and queried
                      
                      schema.sql:
                        create tables with short pks
                      
                      PartitionKeyImpl.java:
                        implement primary key short columns
            ------------------------------------------------------------
            revno: 2876.565.22 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17235051-73
            timestamp: Thu 2013-10-03 21:19:35 +0300
            message:
              merge
                ------------------------------------------------------------
                revno: 2876.567.2
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.6-cluster-7.3
                timestamp: Thu 2013-10-03 15:30:40 +0200
                message:
                  Cherry pick fix for bug#16999886 into mysql-5.6-cluster-7.3
                ------------------------------------------------------------
                revno: 2876.567.1
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2013-10-03 14:10:37 +0200
                message:
                  Supress all warnings from imported memcache and libevent code
            ------------------------------------------------------------
            revno: 2876.565.21 [merge]
            committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
            branch nick: ms-bug17235051-73
            timestamp: Thu 2013-10-03 14:30:04 +0300
            message:
              merge to 7.3
                ------------------------------------------------------------
                revno: 2875.571.132 [merge]
                committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                branch nick: ms-bug17235051-72
                timestamp: Thu 2013-10-03 12:32:28 +0300
                message:
                  merge to 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.65 [merge]
                    committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                    branch nick: ms-bug17235051-71
                    timestamp: Thu 2013-10-03 11:11:52 +0300
                    message:
                      merge to 7.1
                        ------------------------------------------------------------
                        revno: 2555.827.4
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17235051-70
                        timestamp: Wed 2013-10-02 22:38:24 +0300
                        message:
                          bug#17235051 restore.diff
                          conversions char/binary and char/text
                        ------------------------------------------------------------
                        revno: 2555.827.3
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17235051-70
                        timestamp: Wed 2013-10-02 19:53:40 +0300
                        message:
                          bug#17235051 movedata.diff
                          ndb_move_data library and program
                        ------------------------------------------------------------
                        revno: 2555.827.2
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17235051-70
                        timestamp: Wed 2013-10-02 13:39:01 +0300
                        message:
                          bug#17235051 bug17395298.diff
                          fix ndb_restore NDB$BLOB if column id changed
                        ------------------------------------------------------------
                        revno: 2555.827.1
                        committer: Pekka Nousiainen <pekka.nousiainen@oracle.com>
                        branch nick: ms-bug17235051-70
                        timestamp: Wed 2013-10-02 09:47:52 +0300
                        message:
                          bug#17235051 bug17275798.diff
                          do not call convert for NULL in Log part
            ------------------------------------------------------------
            revno: 2876.565.20 [merge]
            committer: magnus.blaudd@oracle.com
            branch nick: 7.3
            timestamp: Wed 2013-10-02 14:31:54 +0200
            message:
              Merge in revert of unused MCP's
                ------------------------------------------------------------
                revno: 2876.562.15
                committer: magnus.blaudd@oracle.com
                branch nick: 7.3
                timestamp: Thu 2013-09-19 09:50:25 +0200
                message:
                  Move type_mcc_utils.cmake into mcc/ dir since that's the only place where it's used(if used at all)
            ------------------------------------------------------------
            revno: 2876.565.19
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-10-02 13:22:15 +0200
            message:
              Cherry pick fix for bug#16853897
              
              	NDB_JOIN_PUSHDOWN_BKA.TEST FAILS ON 'SOL10-SPARC'
            ------------------------------------------------------------
            revno: 2876.565.18 [merge]
            committer: Ole John Aske <ole.john.aske@oracle.com>
            branch nick: mysql-5.6-cluster-7.3
            timestamp: Wed 2013-10-02 10:12:29 +0200
            message:
              Merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.131 [merge]
                committer: Ole John Aske <ole.john.aske@oracle.com>
                branch nick: mysql-5.5-cluster-7.2
                timestamp: Wed 2013-10-02 10:07:59 +0200
                message:
                  Merge 7.1-> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.64 [merge]
                    committer: Ole John Aske <ole.john.aske@oracle.com>
                    branch nick: mysql-5.1-telco-7.1
                    timestamp: Wed 2013-10-02 10:05:14 +0200
                    message:
                      Merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.810.79
                        committer: Ole John Aske <ole.john.aske@oracle.com>
                        branch nick: mysql-5.1-telco-7.0
                        timestamp: Wed 2013-10-02 09:49:48 +0200
                        message:
                          Temp fix for bug#16961971, NDB : REALTIMESCHEDULER REGRESSION ON NDBMTD
                          
                          This fix disables RealTimeScheduler if the MT scheduler (ndbmtd) is used.
                          
                          A warning will be printed:
                          
                          "Configuration contains RealTimeScheduler setting, not currently 
                           supported with ndbmtd (bug# 16961971). Disabling."
            ------------------------------------------------------------
            revno: 2876.565.17 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Tue 2013-10-01 22:01:32 +0200
            message:
              empty merge 7.2 -> 7.3
                ------------------------------------------------------------
                revno: 2875.571.130 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Tue 2013-10-01 21:59:56 +0200
                message:
                  empty merge 7.1 -> 7.2
                    ------------------------------------------------------------
                    revno: 2555.817.63 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Tue 2013-10-01 21:58:48 +0200
                    message:
                      empty merge 7.0 -> 7.1
                        ------------------------------------------------------------
                        revno: 2555.810.78 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Tue 2013-10-01 21:57:28 +0200
                        message:
                          empty merge 6.3 -> 7.0
                            ------------------------------------------------------------
                            revno: 2555.760.82 [merge]
                            tags: clone-mysql-5.1.72-ndb-6.3.53-src-build
                            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                            branch nick: mysql-6.3
                            timestamp: Tue 2013-10-01 21:53:06 +0200
                            message:
                              merged 5.1.72 into 6.3.53
                            ------------------------------------------------------------
                            revno: 2555.760.81
                            committer: Priyanka Sangam <priyanka.sangam@oracle.com>
                            branch nick: mysql-5.1-telco-6.3
                            timestamp: Mon 2013-09-23 11:47:02 +0530
                            message:
                              Bug#11752792  NDB_ERROR_REPORTER DOES NOT INCLUDE CLUSTER LOG WHEN LOGDESTINATION IS US
                              
                              Backporting latest fixes for ndb_error_reporter and its mtr test to 6.3. 
                              
                              Bug#11764570 Parse log filename from LogDestination, defaulting to 
                              ndb_<nodeid>*
                              Bug#16765651 Add * to scp command to include all log files
                              Bug#16602002 Add --connection-timeout, --skip-nodegroup and --dry-scp 
                              options 
                              In 6.3, the nodegroups cannot be queried using ndb_config since the 
                              'Nodegroup' config parameter is not present. Therefore an 'ndb_mgm -e 
                              show' command is used to get the nodegroups for the --skip-nodegroup 
                              option.
                              Bug#11756666 Add --help
                        ------------------------------------------------------------
                        revno: 2555.810.77 [merge]
                        committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                        branch nick: mysql-7.0
                        timestamp: Tue 2013-10-01 21:53:58 +0200
                        message:
                          merged 5.1.72 into 7.0.40
                    ------------------------------------------------------------
                    revno: 2555.817.62 [merge]
                    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                    branch nick: mysql-7.1
                    timestamp: Tue 2013-10-01 21:54:32 +0200
                    message:
                      merged 5.1.72 into 7.1.29
                ------------------------------------------------------------
                revno: 2875.571.129 [merge]
                committer: Mauritz Sundell <mauritz.sundell@oracle.com>
                branch nick: mysql-7.2
                timestamp: Tue 2013-10-01 21:55:22 +0200
                message:
                  merged 5.5.34 into 7.2.14
            ------------------------------------------------------------
            revno: 2876.565.16 [merge]
            committer: Mauritz Sundell <mauritz.sundell@oracle.com>
            branch nick: mysql-7.3
            timestamp: Tue 2013-10-01 21:54:55 +0200
            message:
              merged 5.6.14 into 7.3.3
        ------------------------------------------------------------
        revno: 2876.563.15
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2013-10-16 12:23:00 +0200
        message:
          Third follow up fix for "Bug17258782 MASTER DIDN'T WRITE ERROR MESSAGE IN THE LOG FILE FOR INCIDENT"
          
           - ignore the warnings produced by ndbcluster's binlog thread when the mysqld is restarted. These warnings are expected.
        ------------------------------------------------------------
        revno: 2876.563.14
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2013-10-16 11:51:17 +0200
        message:
          Third follow up fix for "BUG11748566 - MYSQL SHOWS WRONG DEFAULT VALUE FOR PK IF NOT NULL NOT SPECIFIED"
        ------------------------------------------------------------
        revno: 2876.563.13
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2013-10-16 11:19:26 +0200
        message:
          Another follow up fix for "Bug17258782 MASTER DIDN'T WRITE ERROR MESSAGE IN THE LOG FILE FOR INCIDENT"
          
           - ignore the warnings produced by ndbcluster's binlog thread when the cluster is restarted. These warnings are expected.
        ------------------------------------------------------------
        revno: 2876.563.12
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2013-10-16 11:07:53 +0200
        message:
          Follow up fix for BUG11748566 - MYSQL SHOWS WRONG DEFAULT VALUE FOR PK IF
                         NOT NULL NOT SPECIFIED
          
           - fix lead to another bit being set in the "flags" and thus a small change in .result file
        ------------------------------------------------------------
        revno: 2876.563.11
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Wed 2013-10-16 10:55:50 +0200
        message:
          Don't use DEFAULT NULL primary key in test
          
           - evaluated as  NOT NULL DEFAULT NULL and thus makes no sense.
           - changed the test to not request DEFAULT NULL in the primary key
           - also see bug13995622
        ------------------------------------------------------------
        revno: 2876.563.10
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2013-10-15 12:58:32 +0200
        message:
          Second follow up fix for "BUG11748566 - MYSQL SHOWS WRONG DEFAULT VALUE FOR PK IF  NOT NULL NOT SPECIFIED"
            
             - forgot one test in the first round
        ------------------------------------------------------------
        revno: 2876.563.9
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2013-10-15 12:55:00 +0200
        message:
          Follow up fix for "Bug #17258782 MASTER DIDN'T WRITE ERROR MESSAGE IN THE LOG FILE FOR INCIDENT"
          
           - ignore the warnings produced by ndbcluster's binlog thread when the mysqld is restarted
            or cluster is restarted. These warnings are expected.
        ------------------------------------------------------------
        revno: 2876.563.8
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Tue 2013-10-15 11:07:06 +0200
        message:
          Cherrypick fix  for problem compiling with VS9
           - Added MCP's since this is not yet in trunk.
        ------------------------------------------------------------
        revno: 2876.563.7
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2013-10-14 14:14:20 +0200
        message:
          Follow up fix for "BUG11748566 - MYSQL SHOWS WRONG DEFAULT VALUE FOR PK IF  NOT NULL NOT SPECIFIED"
          
           - Tables created with CREATE TABLE t (a INT PRIMARY KEY) and CREATE TABLE t (a INT, PRIMARY KEY(a)) now
            get the same default values.
           - One small change in sql_table.cc, a big change for .result files
        ------------------------------------------------------------
        revno: 2876.563.6 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2013-10-14 11:43:04 +0200
        message:
          Merge trunk -> trunk-cluster
        ------------------------------------------------------------
        revno: 2876.563.5 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Mon 2013-10-14 10:33:59 +0200
        message:
          Merge trunk -> trunk-cluster
          
           - -r 6478.1.1
           - revert reverted MCPs
           - remove MCP marks for parts which has been incorporated into trunk
           - closing the door to trunk
        ------------------------------------------------------------
        revno: 2876.563.4 [merge]
        committer: magnus.blaudd@oracle.com
        branch nick: trunk-cluster
        timestamp: Fri 2013-10-11 16:06:24 +0200
        message:
          Merge trunk -> trunk-cluster
------------------------------------------------------------
revno: 8337 [merge]
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-01 11:47:24 +0300
message:
  Merge mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.338.7 [merge]
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-01 11:44:09 +0300
    message:
      Merge mysql-5.6.20-release to mysql-5.6.
        ------------------------------------------------------------
        revno: 3902.339.2
        committer: Marko M?kel? <marko.makela@oracle.com>
        branch nick: mysql-5.6.20-release
        timestamp: Tue 2014-07-01 11:41:45 +0300
        message:
          Revert Bug#15923864 fix (revno 6008) from MySQL 5.6.20.
        ------------------------------------------------------------
        revno: 3902.339.1
        committer: Venkata Sidagam <venkata.sidagam@oracle.com>
        branch nick: 5.6
        timestamp: Tue 2014-07-01 14:03:12 +0530
        message:
          Bug#11758766:MYSQLD CONTINUES OPERATION WITHOUT LOGGING WHEN
                BINLOGS CANNOT BE WRITTEN
          
          Post push fix.
          Reverting back the file permissions to 0664
------------------------------------------------------------
revno: 8336 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug11758766_mysql-trunk
timestamp: Tue 2014-07-01 13:49:45 +0530
message:
  Bug#11758766:MYSQLD CONTINUES OPERATION WITHOUT LOGGING WHEN
  BINLOGS CANNOT BE WRITTEN
  
  Null merge
    ------------------------------------------------------------
    revno: 3902.338.6
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug11758766_mysql-5.6
    timestamp: Tue 2014-07-01 13:48:26 +0530
    message:
      Bug#11758766:MYSQLD CONTINUES OPERATION WITHOUT LOGGING WHEN
      BINLOGS CANNOT BE WRITTEN
      
      Post push fix.
      Reverting back the file permissions to 0664
------------------------------------------------------------
revno: 8335 [merge]
author: 
committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-07-01 08:50:30 +0200
message:
  Empty version change upmerge
    ------------------------------------------------------------
    revno: 3902.338.5 [merge]
    author: 
    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-07-01 08:42:51 +0200
    message:
      Raise version number after cloning  and merging back the changes from 5.5
        ------------------------------------------------------------
        revno: 2875.598.6
        author: murthy.narkedimilli@oracle.com
        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-07-01 08:31:52 +0200
        message:
          Raise version number after cloning 5.5.39
------------------------------------------------------------
revno: 8334 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Tue 2014-07-01 08:45:12 +0530
message:
  Bug #17357528 BACKPORT BUG#16513435 TO 5.5 AND 5.6
  
  Follow-up patch to fix testcase.
  Null merge from mysql-5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.338.4
    tags: clone-5.6.20-build
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Tue 2014-07-01 08:40:26 +0530
    message:
      Bug #17357528 BACKPORT BUG#16513435 TO 5.5 AND 5.6
      
      Follow up patch to fix testcase failure.
------------------------------------------------------------
revno: 8333
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-07-01 08:19:52 +0530
message:
  - Follow-up performance fix for WL#7682.
  
    Intrinsic Table: Optimization to use dict_table_t for caching row_id/trx_id
    instead of innodb_session_t
  
    With WL#7682 we introduced intrinsic tables. Given that these tables has
    session scope row_id and trx_id needed for these tables were cached in
    innodb_session_t (session structure cached in thd).
  
    Unfortunately, accessing these ids through session is performance costly as
    it involves lookup into map using table name and loading of thd that in turn
    causes invalidation the CPU cache.
  
    We observed a regression upto 35% with Wisconsin-DB queries.
  
    So we are moving back to original idea of caching these ids as part of
    dict_table_t.
  
    Approved by: Jimmy (rb#5881)
------------------------------------------------------------
revno: 8332 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-30 19:41:34 +0530
message:
  Bug #17657223	 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Reverting the patch due to pb2 failure.
    ------------------------------------------------------------
    revno: 3902.338.3
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-30 19:38:45 +0530
    message:
      Bug #17657223	 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
      	Reverting the patch due to pb2 failure.
------------------------------------------------------------
revno: 8331 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Mon 2014-06-30 19:30:13 +0530
message:
  Bug #17357528 BACKPORT BUG#16513435 TO 5.5 AND 5.6
  
  Null merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.2 [merge]
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-06-30 19:27:01 +0530
    message:
      Bug #17357528 BACKPORT BUG#16513435 TO 5.5 AND 5.6
      
      Merging from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.598.5
        tags: clone-5.5.39-build
        committer: Venkata Sidagam <venkata.sidagam@oracle.com>
        branch nick: 5.5
        timestamp: Mon 2014-06-30 19:24:25 +0530
        message:
          Bug #17357528 BACKPORT BUG#16513435 TO 5.5 AND 5.6
          
          Description: Backporting BUG#16513435 to 5.5 and 5.6
          This is a fix for REMOTE PREAUTH USER ENUMERATION FLAW bug
------------------------------------------------------------
revno: 8330 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug11758766_mysql-trunk
timestamp: Mon 2014-06-30 19:00:42 +0530
message:
  Bug#11758766:MYSQLD CONTINUES OPERATION WITHOUT LOGGING WHEN
  BINLOGS CANNOT BE WRITTEN
  
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.338.1
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug11758766_mysql-5.6
    timestamp: Mon 2014-06-30 18:46:22 +0530
    message:
      Bug#11758766:MYSQLD CONTINUES OPERATION WITHOUT LOGGING WHEN
      BINLOGS CANNOT BE WRITTEN
      
      Fixing a typo.
------------------------------------------------------------
revno: 8329
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-cleanup
timestamp: Mon 2014-06-30 14:13:43 +0200
message:
  fixed build warning when compiling without the stage instrumentation
------------------------------------------------------------
revno: 8328 [merge]
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-30 13:02:01 +0100
message:
  BUG#14135002
  
  Automerged mysql-5.6 into latest mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.165 [merge]
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-30 13:00:36 +0100
    message:
      BUG#14135002
      
      Automerged into latest mysql-5.6
        ------------------------------------------------------------
        revno: 3902.337.1
        committer: Luis Soares <luis.soares@oracle.com>
        branch nick: mysql-5.6
        timestamp: Fri 2014-06-27 00:02:09 +0100
        message:
          BUG#14135002: RPL.RPL_ROTATE_PURGE_DEADLOCK FAILS WITH TIMEOUT ON PB2.
          
          The test case makes use of the fine DEBUG_SYNC facility to synchronize
          the execution of different threads.
          
          The problem was that the test was using DEBUG_SYNC to signal the
          waiting threads and then right away resetting it (SET
          DEBUG_SYNC=RESET). This could in some cases clear the signal before
          the waiting thread received it. In those cases the waiting thread
          would remain waiting ad eternum, thus causing the test case to
          timeout.
          
          We fix this by waiting for the threads to receive the signal and only
          then issue the reset.
------------------------------------------------------------
revno: 8327 [merge]
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-30 12:57:38 +0200
message:
  BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
  
  Reverted change due to mtr test failure.
    ------------------------------------------------------------
    revno: 3902.324.164 [merge]
    committer: Marcin Babij <marcin.babij@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-30 12:34:23 +0200
    message:
      BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
      
      Reverted change due to mtr test failure.
        ------------------------------------------------------------
        revno: 2875.598.4
        committer: Marcin Babij <marcin.babij@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-06-30 12:31:44 +0200
        message:
          BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
          
          Reverted change due to mtr test failure.
------------------------------------------------------------
revno: 8326
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-review
timestamp: Mon 2014-06-30 10:11:33 +0200
message:
  Bug#19048609 EXPORT_SET SUFFERS FROM INTEGER OVERFLOW AND FAILURE TO DETECT LARGE RESULT
  
  Do all size calculations in Item_func_export_set::val_str() in ulonglong.
  Calculations using uint could lead to failure to reject too long results,
  and undefined behaviour during string concatenation.
------------------------------------------------------------
revno: 8325 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-30 11:16:09 +0530
message:
  WL#7894 - Improve dtrace support on Oracle Linux 6
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.163
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-30 11:11:59 +0530
    message:
      WL#7894 - Improve dtrace support on Oracle Linux 6
      
      Follow up patch to fix pb2 test failure.
      This test runs dtrace.d script in background to get trace output.
      But because of some permission issues on Mac, dtrace.d script
      execution was failing. Changed script to run test script
      with dtrace command instead of running it directly.
------------------------------------------------------------
revno: 8324 [merge]
committer: Mayank Prasad <mayank.prasad@oracle.com>
branch nick: trunk
timestamp: Mon 2014-06-30 10:47:03 +0530
message:
  Merge from 5.6 to trunk for bug#17729044
    ------------------------------------------------------------
    revno: 3902.324.162
    committer: Mayank Prasad <mayank.prasad@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-06-30 10:16:56 +0530
    message:
      Bug #17729044 : PERFORMANCE SCHEMA QUERY ON EVENTS_STATEMENTS_CURRENT WRONG WITH ORDER BY
            
      Details : 
        - Issue was seen when ORDER BY clause is used. 
            
      Issue:
        - For each table, size of position structure (m_pos) supposed to be the size of a record in that table. In this case, reason 
          for wrong result is the wrong size of position structure aka m_pos (ref length) used in statement_current table to 
          traverse through records.  
            
      Fix:
       - Corrected the size of position structure, m_pos (ref length) in statement_current table so that it can correctly 
          traverse thorough all records.
------------------------------------------------------------
revno: 8323
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Fri 2014-06-27 16:48:35 +0200
message:
  Bug#19076686 GET RID OF DYNAMIC_ARRAY IN SID_MAP
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
      
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Do not say 'using std::list' at global scope in a header file,
  it pollutes the namespace of every file that happens to include it.
------------------------------------------------------------
revno: 8322 [merge]
committer: Gopal Shankar <gopal.shankar@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 19:38:06 +0530
message:
  Bug#18776592 INNODB: FAILING ASSERTION: PRIMARY_KEY_NO == -1 ||
                                          PRIMARY_KEY_NO == 0 
  
  Fixing regression. Some SELECT's needs ORDER BY.
    ------------------------------------------------------------
    revno: 3902.324.161 [merge]
    committer: Gopal Shankar <gopal.shankar@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 19:32:58 +0530
    message:
      Bug#18776592 INNODB: FAILING ASSERTION: PRIMARY_KEY_NO == -1 ||
                                              PRIMARY_KEY_NO == 0 
      
      Fixing regression. Some SELECT's needs ORDER BY.
        ------------------------------------------------------------
        revno: 2875.598.3
        committer: Gopal Shankar <gopal.shankar@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-06-27 19:30:19 +0530
        message:
          Bug#18776592 INNODB: FAILING ASSERTION: PRIMARY_KEY_NO == -1 ||
                                                  PRIMARY_KEY_NO == 0 
          
          Fixing regression. Some SELECT's needs ORDER BY.
------------------------------------------------------------
revno: 8321 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 19:04:42 +0530
message:
  Bug#18684393 - METADATA CHANGES MIGHT CAUSE PROBLEMS WITH TRIGGER
                 EXECUTION.
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.160
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql_5_6_new
    timestamp: Fri 2014-06-27 18:54:20 +0530
    message:
      Bug#18684393 - METADATA CHANGES MIGHT CAUSE PROBLEMS WITH TRIGGER
                     EXECUTION.
------------------------------------------------------------
revno: 8320
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-2
timestamp: Fri 2014-06-27 11:51:42 +0200
message:
  Fix for Bug#19063289 CRASH IN PRINT_TABLE_ARRAY ON 2ND EXECUTION OF PS WITH SUBQUERY AND VIEW
  The fix for bug 18945693 missed something: reinit_before_use() does not cover join nests.
  In this testcase, the join condition is on a join nest:
      C AS SQ2_alias1 LEFT  JOIN
         ( B AS SQ2_alias2 INNER JOIN C AS SQ2_alias3
           ON 1)
         ON 1
  So again, we get a dangling optim_join_cond pointer at start of second execution.
  Fix: set join->optimized a few lines later, so it can tell if optimized conditions are readable.
  Testcase is added to opt trace suite because opt trace prints the query in many places
  so gives more coverage.
------------------------------------------------------------
revno: 8319
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 13:22:00 +0100
message:
  BUG#19010820 START SLAVE UNTIL WILL TIMEOUT IF WAITING FOR
               THE LAST TRANSACTION IN RELAYLOG
  
  Problem:
  =======
  
  "START SLAVE UNTIL" with master binlog file/position or slave relay
  log file/position clauses won't stop if the position in the until
  clause is of the last transaction present on the relay log.
  
  Analysis:
  ========
  
  The UNTIL verification for master binlog file/position or slave
  relay log file/position clauses is only checked inside next_event()
  function, after calling read_log_event(). If read_log_event() return
  no event because it reached the end of the relay log, it will not
  check the UNTIL clause, and will wait for the relay log update by
  calling relay_log.wait_for_update_relay_log().
  
  There is a UNTIL verification in the exec_relay_log_event() function,
  before calling next_event() function, but only for until clauses
  containing SQL_AFTER_GTIDS option.  
  
  Fix:
  ===
  
  @sql/rpl_slave.cc
  
  Added both UNTIL_MASTER_POS and UNTIL_RELAY_POS options to the
  verification at the exec_relay_log_event() function that is called
  before the next_event() function.
------------------------------------------------------------
revno: 8318 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 17:21:22 +0530
message:
  Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
  
  Null merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.159 [merge]
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 17:19:28 +0530
    message:
      Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
      
      Null merge from 5.5.
        ------------------------------------------------------------
        revno: 2875.598.2
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-06-27 17:17:04 +0530
        message:
          Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
          
          Post-push patch. Changing file permission of "scripts/mysqlaccess.conf".
------------------------------------------------------------
revno: 8317 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 17:07:13 +0530
message:
  Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
  
  Null merge from 5.6.
    ------------------------------------------------------------
    revno: 3902.324.158 [merge]
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 17:05:16 +0530
    message:
      Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
      
      Merging from 5.5 to 5.6.
        ------------------------------------------------------------
        revno: 2875.598.1
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-06-27 17:04:08 +0530
        message:
          Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
          
          Backporting patch committed for bug 18008907 to 5.5
          and 5.6.
------------------------------------------------------------
revno: 8316 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 16:49:01 +0530
message:
  Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
  
  Null merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.336.1 [merge]
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 16:45:24 +0530
    message:
      Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
      
      Merging from 5.5 to 5.6.
        ------------------------------------------------------------
        revno: 2875.597.6
        committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-06-27 16:43:18 +0530
        message:
          Bug#18903155: BACKPORT BUG-18008907 TO 5.5+ VERSIONS.
          
          Backporting patch committed for bug 18008907 to 5.5
          and 5.6.
------------------------------------------------------------
revno: 8315 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk5
timestamp: Fri 2014-06-27 19:09:48 +0800
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.157
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 19:07:21 +0800
    message:
      Back port fixes for innodb_fts_misc.test from trunk, issue introduced by
      bug #18711306 checkin
------------------------------------------------------------
revno: 8314
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 16:22:11 +0530
message:
  Removing the test mysql_secure_installation.test from experimental.
------------------------------------------------------------
revno: 8313 [merge]
committer: Terje Rosten <terje.rosten@oracle.com>
branch nick: mysql-trunk.postfix
timestamp: Fri 2014-06-27 12:47:28 +0200
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.156 [merge]
    committer: Terje Rosten <terje.rosten@oracle.com>
    branch nick: mysql-5.6.postfix
    timestamp: Fri 2014-06-27 12:44:24 +0200
    message:
      Merge from mysql-5.5 to mysql-5.6.
        ------------------------------------------------------------
        revno: 2875.597.5
        committer: Terje Rosten <terje.rosten@oracle.com>
        branch nick: mysql-5.5.postfix
        timestamp: Fri 2014-06-27 12:41:49 +0200
        message:
          Bug#16395459 TEST AND RESULT FILES WITH EXECUTE BIT
          
          Post push fix: add execute bit on perl script.
------------------------------------------------------------
revno: 8312 [merge]
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 12:07:09 +0200
message:
  BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
  
  Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
  
  Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
  Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
    ------------------------------------------------------------
    revno: 3902.324.155 [merge]
    committer: Marcin Babij <marcin.babij@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 11:57:21 +0200
    message:
      BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
      
      Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
      
      Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
      Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
------------------------------------------------------------
revno: 8311 [merge]
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 11:50:42 +0200
message:
  BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
  
  Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
  
  Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
  Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
    ------------------------------------------------------------
    revno: 3902.335.1 [merge]
    committer: Marcin Babij <marcin.babij@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 11:29:32 +0200
    message:
      BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
      
      Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
      
      Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
      Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
        ------------------------------------------------------------
        revno: 2875.597.4
        committer: Marcin Babij <marcin.babij@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-06-27 11:27:27 +0200
        message:
          BUG#18779944: MYSQLDUMP BUFFER OVERFLOW
          
          Mysqldump overflows stack buffer when copying table name from commandline arguments resulting in stack corruption and ability to execute arbitrary code.
          
          Fix: Check length of all positional arguments passed to mysqldump is smaller than NAME_LEN.
          Note: Mysqldump heavily depends on that database objects (databases, tablespaces, tables, etc) are limited to small size (now it is 64).
------------------------------------------------------------
revno: 8310 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 15:14:48 +0530
message:
  WL#7894 - Improve dtrace support on Oracle Linux 6
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.154
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 15:07:44 +0530
    message:
      WL#7894 - Improve dtrace support on Oracle Linux 6
      
      Follow up patch. Test 'dynamic_tracing.test" is failing on
      some pb2 platforms because of not having permisssion to execute 
      dtrace.d.
------------------------------------------------------------
revno: 8309
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Fri 2014-06-27 11:33:14 +0200
message:
  Bug#19068046 GET RID OF DYNAMIC_ARRAY IN SQL/AUTH/*.CC
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
      
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8308 [merge]
committer: Anil Toshniwal <anil.toshniwal@oracle.com>
branch nick: trunk
timestamp: Fri 2014-06-27 14:01:34 +0530
message:
  Bug#13939924 MYSQLBACKUP REFERENCES IBBACKUP IN MYSQL ERROR LOG
  
  Fixed:The reference of "ibbackup" has to be replace by "mysqlbackup".
  Approved by: Shaohua(rb#5821)
    ------------------------------------------------------------
    revno: 3902.324.153
    committer: Anil Toshniwal <anil.toshniwal@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-27 13:59:51 +0530
    message:
      Bug#13939924 MYSQLBACKUP REFERENCES IBBACKUP IN MYSQL ERROR LOG
      
      Fixed:The reference of "ibbackup" has to be replace by "mysqlbackup".
      Approved by: Shaohua(rb#5821)
------------------------------------------------------------
revno: 8307 [merge]
committer: Terje Rosten <terje.rosten@oracle.com>
branch nick: mysql-trunk.chkout
timestamp: Fri 2014-06-27 09:29:58 +0200
message:
  Null merge from mysql-5.6 to mysql-trunk. 
    ------------------------------------------------------------
    revno: 3902.324.152
    committer: Terje Rosten <terje.rosten@oracle.com>
    branch nick: mysql-5.6-skip.my.cnf.ung
    timestamp: Fri 2014-06-27 09:26:01 +0200
    message:
      Bug#18205019: RPM INSTALLATION GENERATES /USR/MY.CNF
      
      A feature added to mysql_install_db in MySQL 5.6 was to generate my.cnf config file
      from template. The reason was a wish to set a specific value of sql_mode option
      for new installations. This is useful for many systems, however for some install
      methods and layouts this was not needed, and for some, even not wanted.
      
      Solution is add new option --keep-my-cnf to mysql_install_db.
      When used mysql_install_db will not generate my.cnf from template.
      
      
      Patch will also resolve issue mentioned in bug #68117 and #68318.
------------------------------------------------------------
revno: 8306
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 10:30:14 +0530
message:
  Bug#18778043: ERROR_LOG_THROTTLE OBJECT CREATED IN MYSQLD.CC
                IS NOT USED. 
  
  Object "err_log_throttle" was created to avoid spamming error
  log file on failure to create thread for new connection. After
  moving logic to add new connection to "Per_thread_connection_handler
  ::add_connection", this object was not used. Removed this object
  now. Also changed code to log error message on failure to create 
  thread in "Per_thread_connection_handler::add_connection" function
  and used new static member "ERROR_LOG_THROTTLE 
  create_thd_err_log_throttle" in file "connection_handler_per_thread.cc"
  to avoid spamming error log file on thread creation failure for new
  connection. This restores the functionality.
------------------------------------------------------------
revno: 8305
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-27 08:12:27 +0800
message:
  Bug #18963555 	TAKING WRITE LOCK WHILE ADDING A GTID INTO GTID SET IS NOT GOOD TO PERFORMANCE
  
  In Gtid_state::update_on_commit(), it takes global_sid_lock.wrlock()
  while adding a GTID into gtid set, which is not good to performance.
  
  Optimize it to take global_sid_lock.rdlock() and lock a mutex for
  the given SIDNO while adding the GTID into gtid set for improving
  performance.
------------------------------------------------------------
revno: 8304
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-cleanup
timestamp: Fri 2014-06-27 01:43:12 +0200
message:
  Build cleanup, when compiling without the performance schema
------------------------------------------------------------
revno: 8303 [merge]
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 15:33:48 +0100
message:
  BUG#13874553
  
  Automerged mysql-5.6 into latest mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.151 [merge]
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 15:30:38 +0100
    message:
      BUG#13874553
      
      Manually merged  mysql-5.5 into mysql-5.6
      
      CONFLICTS
      =========
      
      Text conflict in mysql-test/suite/rpl/r/rpl_stop_slave.result
      Text conflict in mysql-test/suite/rpl/t/rpl_stop_slave.test
        ------------------------------------------------------------
        revno: 2875.597.3
        committer: Luis Soares <luis.soares@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-06-26 12:54:27 +0100
        message:
          BUG#13874553: rpl.rpl_stop_slave fails sporadically on pb2
          
          The test case makes use of the fine DEBUG_SYNC facility. Furthermore,
          since it needs synchronization on internal threads (dump and SQL
          threads) the server code has DEBUG_SYNC commands internally deployed
          and activated through the DBUG_EXECUTE_IF macro. The internal
          DBUG_SYNC commands are then controlled from the test case through the
          DEBUG variable.
          
          There were three problems around the DEBUG + DEBUG_SYNC facility
          usage:
          
          1. When signaling the SQL thread to continue, the test would reset
             immediately the DEBUG_SYNC variable. This could mean that the SQL
             thread might loose the signal and continue to wait forever;
          
          2. A similar scenario was happening with the dump thread on the
             master. This thread was instructed to wait, and later it would be
             signaled to continue, but immediately after the DEBUG_SYNC would be
             reset. This could lead to the dump thread missing the signal and
             wait forever;
          
          3. The test was not cleaning itself up with respect to the
             instrumentation of the dump thread. This would leave the
             conditional execution of an internal DEBUG_SYNC command active
             (through the usage of DBUG_EXECUTE_IF). 
          
          We fix #1 and #2 by waiting for the threads to receive the signal and
          only then issue the reset. We fix #3 by reseting the DEBUG variable,
          thus deactivating the dump thread internal DEBUG_SYNC command.
------------------------------------------------------------
revno: 8302 [merge]
committer: Akhil Mohan <akhil.mohan@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 18:40:15 +0530
message:
  19020385 Added packaging source for Debian7, Ubuntu12.04, Ubuntu14.04
    ------------------------------------------------------------
    revno: 3902.324.150
    committer: Akhil Mohan <akhil.mohan@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 18:36:58 +0530
    message:
      19020385 Added packaging source for Debian7, Ubuntu12.04, Ubuntu14.04
------------------------------------------------------------
revno: 8301
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 16:22:46 +0530
message:
  - Reverting back the code that was rolled back in revno 8284.
    Actually we failed to realize the reason why code was added
    and wrongly assumed it to be redundant.
    Also made the code mutex get/release conditional only in case
    ATOMICS is enabled.
------------------------------------------------------------
revno: 8300 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 11:10:27 +0200
message:
  Null merge from 5.6
    ------------------------------------------------------------
    revno: 3902.324.149 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 10:08:11 +0200
    message:
            Merge from 5.5 => 5.6
            Bug#19063012 fix embedded-devel conflict issue
        ------------------------------------------------------------
        revno: 2875.597.2
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-06-26 09:39:29 +0200
        message:
          Bug#19063012 fix embedded-devel conflict issue
------------------------------------------------------------
revno: 8299
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 10:15:56 +0200
message:
              Merge from 5.6 => trunk
              Bug#19063012 fix embedded-devel conflict issue
------------------------------------------------------------
revno: 8298
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Thu 2014-06-26 09:46:28 +0200
message:
  Bug#19060799 GET RID OF DYNAMIC_ARRAY IN MY_TMPDIR
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Move mysys/mf_tmpdir.c to mysys_ssl (which is a misnomer, it's the C++ part of mysys)
  Move mutex_lock.h to the include directory, so it can be used outside of server code.
------------------------------------------------------------
revno: 8297 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 12:49:23 +0530
message:
  WL#7894 - Improve dtrace support on Oracle Linux 6
  
  Null merge from 5.6.
    ------------------------------------------------------------
    revno: 3902.324.148
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 12:41:13 +0530
    message:
      WL#7894 - Improve dtrace support on Oracle Linux 6
      
      Follow patch to change file permission of file "scripts/mysqlaccess.conf".
------------------------------------------------------------
revno: 8296 [merge]
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 10:13:21 +0530
message:
  Null merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.147 [merge]
    committer: Arun Kuruvila <arun.kuruvila@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 10:09:51 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.597.1
        committer: Arun Kuruvila <arun.kuruvila@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-06-26 10:08:55 +0530
        message:
          Bug#18463911 : SERVER CRASHES ON CREATING A TEMP TABLE WITH
                         CERTAIN MAX_HEAP_TABLE_SIZE VALUES
          
          Followup patch to fix failure on Window machine.
------------------------------------------------------------
revno: 8295 [merge]
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 10:00:03 +0530
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.334.1 [merge]
    committer: Arun Kuruvila <arun.kuruvila@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 09:53:39 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.24
        committer: Arun Kuruvila <arun.kuruvila@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-06-26 09:51:58 +0530
        message:
          Bug#18463911 : SERVER CRASHES ON CREATING A TEMP TABLE WITH
                         CERTAIN MAX_HEAP_TABLE_SIZE VALUES
          
          Followup patch to fix failure on Window machine.
------------------------------------------------------------
revno: 8294 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-26 09:41:53 +0530
message:
  WL#7894 - Improve dtrace support on Oracle Linux 6
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.146
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-26 09:36:07 +0530
    message:
      WL#7894 - Improve dtrace support on Oracle Linux 6
      
      Dtrace support is enabled and built if the build system(cmake)
      detects the "dtrace" command on the Solaris, Mac OS X and FreeBSD.
      As part of this WL, dtrace support is extented to the Oracle 
      Linux 6 with UEK3 kernel. Now Dtrace support is enabled and built
      if Oracle Linux 6 with UEK3 has real dtrace.
------------------------------------------------------------
revno: 8293
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Thu 2014-06-26 10:58:58 +0800
message:
  BUG#19066657 - INNODB-READ-VIEW.TEST MISSING DEBUG SYNC POINT
  
  Fix two problems in the test case:
  1. Missing debug sync point;
  2. Wrong connection in the test from where we send the signal.
  
  Approved by Jimmy.Yang in IM.
------------------------------------------------------------
revno: 8292
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt
timestamp: Wed 2014-06-25 17:06:14 +0200
message:
  bring n test bug signature and error message lists up to date
------------------------------------------------------------
revno: 8291 [merge]
committer: Raghav Kapoor <raghav.kapoor@oracle.com>
branch nick: mysql-trunk-bug-17665767
timestamp: Wed 2014-06-25 18:15:36 +0530
message:
  BUG#17665767 - FAILING ASSERTION: PRIMARY_KEY_NO == -1 || PRIMARY_KEY_NO == 0 
  
  BACKGROUND:
  This bug is a followup on Bug#16368875.
  The assertion failure happens because in SQL layer the key
  does not get promoted to PRIMARY KEY but InnoDB takes it
  as PRIMARY KEY.
  
  ANALYSIS:
  Here we are trying to create an index on POINT (GEOMETRY)
  data type which is a type of BLOB (since GEOMETRY is a
  subclass of BLOB).
  In general, we can't create an index over GEOMETRY family
  type field unless we specify the length of the
  keypart (similar to BLOB fields).
  Only exception is the POINT field type. The POINT column
  max size is 25. The problem is that the field is not treated
  as PRIMARY KEY when we create a index on POINT column using
  its max column size as key part prefix. The fix would allow
  index on POINT column to be treated as PRIMARY KEY.
  
  FIX:
  Patch for Bug#16368875 is extended to take into account
  GEOMETRY datatype, POINT in particular to consider it
  as PRIMARY KEY in SQL layer.
    ------------------------------------------------------------
    revno: 3902.324.145 [merge]
    committer: Raghav Kapoor <raghav.kapoor@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-25 18:10:29 +0530
    message:
      BUG#17665767 - FAILING ASSERTION: PRIMARY_KEY_NO == -1 || PRIMARY_KEY_NO == 0 
      
      BACKGROUND:
      This bug is a followup on Bug#16368875.
      The assertion failure happens because in SQL layer the key
      does not get promoted to PRIMARY KEY but InnoDB takes it
      as PRIMARY KEY.
      
      ANALYSIS:
      Here we are trying to create an index on POINT (GEOMETRY)
      data type which is a type of BLOB (since GEOMETRY is a
      subclass of BLOB).
      In general, we can't create an index over GEOMETRY family
      type field unless we specify the length of the
      keypart (similar to BLOB fields).
      Only exception is the POINT field type. The POINT column
      max size is 25. The problem is that the field is not treated
      as PRIMARY KEY when we create a index on POINT column using
      its max column size as key part prefix. The fix would allow
      index on POINT column to be treated as PRIMARY KEY.
      
      FIX:
      Patch for Bug#16368875 is extended to take into account
      GEOMETRY datatype, POINT in particular to consider it
      as PRIMARY KEY in SQL layer.
        ------------------------------------------------------------
        revno: 2875.595.23
        committer: Raghav Kapoor <raghav.kapoor@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-06-25 18:06:28 +0530
        message:
          BUG#17665767 - FAILING ASSERTION: PRIMARY_KEY_NO == -1 || PRIMARY_KEY_NO == 0 
          
          BACKGROUND:
          This bug is a followup on Bug#16368875.
          The assertion failure happens because in SQL layer the key
          does not get promoted to PRIMARY KEY but InnoDB takes it
          as PRIMARY KEY.
          
          ANALYSIS:
          Here we are trying to create an index on POINT (GEOMETRY)
          data type which is a type of BLOB (since GEOMETRY is a
          subclass of BLOB).
          In general, we can't create an index over GEOMETRY family
          type field unless we specify the length of the
          keypart (similar to BLOB fields).
          Only exception is the POINT field type. The POINT column
          max size is 25. The problem is that the field is not treated
          as PRIMARY KEY when we create a index on POINT column using
          its max column size as key part prefix. The fix would allow
          index on POINT column to be treated as PRIMARY KEY.
          
          FIX:
          Patch for Bug#16368875 is extended to take into account
          GEOMETRY datatype, POINT in particular to consider it
          as PRIMARY KEY in SQL layer.
------------------------------------------------------------
revno: 8290
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Wed 2014-06-25 13:46:01 +0200
message:
  Bug#18913949: USE REASONABLE DEFAULTS WHEN INITIALIZING
                THD/TRANS MEM_ROOTS
  
  Post-push fix: add DTOR to Transaction_ctx to ensure we do not leak memory from
  transaction_cache_insert_recovery()
------------------------------------------------------------
revno: 8289 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18405221
timestamp: Wed 2014-06-25 16:43:57 +0530
message:
  Merge from mysql-5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.324.144 [merge]
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18405221
    timestamp: Wed 2014-06-25 16:37:27 +0530
    message:
      Merge from mysql-5.6 to trunk.
        ------------------------------------------------------------
        revno: 2875.595.22
        committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
        branch nick: mysql-5.5-18405221
        timestamp: Wed 2014-06-25 16:33:04 +0530
        message:
          BUG#18405221: SHOW CREATE VIEW OUTPUT INCORRECT
          
          Fix:
          ---
          The issue reported is same as the BUG#14117018.
          Hence backporting the patch from mysql-trunk
          to mysql-5.5 and mysql-5.6
------------------------------------------------------------
revno: 8288 [merge]
committer: Terje Rosten <terje.rosten@oracle.com>
branch nick: mysql-trunk-rpmlint
timestamp: Wed 2014-06-25 12:44:30 +0200
message:
  Null merge from mysql-5.6 to mysql trunk.
    ------------------------------------------------------------
    revno: 3902.324.143 [merge]
    committer: Terje Rosten <terje.rosten@oracle.com>
    branch nick: mysql-5.6-rpmlint
    timestamp: Wed 2014-06-25 12:41:06 +0200
    message:
      Null merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.21
        committer: Terje Rosten <terje.rosten@oracle.com>
        branch nick: mysql-5.5-rpmlint
        timestamp: Wed 2014-06-25 12:35:50 +0200
        message:
          Bug#16395459 TEST AND RESULT FILES WITH EXECUTE BIT
          Bug#16415173 CRLF INSTEAD OF LF IN SQL-BENCH SCRIPTS
                
          Correct perms and converts from Windows style to UNIX style line endings on some files.
          Fix perms on installed ini files.
          
          (MySQL 5.5 version)
    ------------------------------------------------------------
    revno: 3902.324.142
    committer: Terje Rosten <terje.rosten@oracle.com>
    branch nick: mysql-5.6-rpmlint
    timestamp: Wed 2014-06-25 12:39:11 +0200
    message:
      Bug#16395459 TEST AND RESULT FILES WITH EXECUTE BIT
      Bug#16415173 CRLF INSTEAD OF LF IN SQL-BENCH SCRIPTS
                  
      Correct perms and converts Windows style to UNIX style line endings.
      Fix perms on installed files by tiny cmake patch.
            
      (MySQL 5.6 version)
------------------------------------------------------------
revno: 8287
committer: Terje Rosten <terje.rosten@oracle.com>
branch nick: mysql-trunk-rpmlint
timestamp: Wed 2014-06-25 12:43:00 +0200
message:
  Bug#16395459 TEST AND RESULT FILES WITH EXECUTE BIT
  Bug#16415173 CRLF INSTEAD OF LF IN SQL-BENCH SCRIPTS
        
  Correct perms and converts Windows style to UNIX style line endings.
  Fix perms on installed files by tiny cmake patch.
  
  (MySQL 5.7 version)
------------------------------------------------------------
revno: 8286 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 12:23:20 +0200
message:
  Merge from 5.6 => trunk - Bug#18321083 -Added bench package, enabled dtrace
    ------------------------------------------------------------
    revno: 3902.324.141 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-25 12:01:23 +0200
    message:
      Merge from 5.5 => 5.6 - Bug#18321083 -Added bench package, enabled dtrace
        ------------------------------------------------------------
        revno: 2875.595.20
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-06-25 11:52:13 +0200
        message:
          Bug#18321083 - Added bench package, enabled dtrace
------------------------------------------------------------
revno: 8285
committer: Narendra Chauhan <narendra.chauhan@oracle.com>
branch nick: mysql-5.7
timestamp: Wed 2014-06-25 15:47:42 +0530
message:
  Bug#18128323 - RPL.RPL_OPTIMIZE FAILED WITH TIMEOUT 900 IN DAILY-TRUNK.
  
  Enabling testcase for valgrind.
------------------------------------------------------------
revno: 8284
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 13:12:17 +0300
message:
  Bug#19061440 CMAKE -DINNODB_PAGE_ATOMIC_REF_COUNT:BOOL=OFF IS BROKEN
  
  Remove some unnecessary block->mutex acquisition that was added in WL#7682.
  Approved by Krunal Bauskar.
------------------------------------------------------------
revno: 8283 [merge]
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 12:14:35 +0530
message:
  Merging from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.140 [merge]
    committer: Arun Kuruvila <arun.kuruvila@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-25 11:53:34 +0530
    message:
      Merging from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.19
        committer: Arun Kuruvila <arun.kuruvila@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-06-25 11:42:41 +0530
        message:
          Bug #18463911 : SERVER CRASHES ON CREATING A TEMP TABLE
                          WITH CERTAIN MAX_HEAP_TABLE_SIZE VALUES
          
          Description:
          When the  system variable 'max_heap_table_size'
          is set to 20GB, the server crashes on creation of a
          temporary tables or tables using MEMORY storage engine.
          
          Analysis:
          The variable 'max_record' determines the amount heap
          allocated for the records of the table. This value
          is determined using the 'max_heap_table_size' variable.
          'records_in_block' in turn uses the max_records to
          determine the number of records per block.
          
          When the 'max_heap_table_size' is set to 20GB, then
          the 'records_in_block' is calculated to a value of
          2^28.
          
          The size of the block determined by multiplying the
          'records_in_block' and 'recbuffer' results in overflow
          and hence the value becomes zero. As a result, zero bytes
          of the heap is allocated for the table. This will
          result in a server crash when the table is accessed.
          
          Fix:
          The variables 'records_in_block' and 'recbuffer' are
          typecasted to 'unsigned long' while calculating the
          size of the block.
------------------------------------------------------------
revno: 8282
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 11:18:14 +0530
message:
  WL#7894 - Improve dtrace support on Oracle Linux 6
  
  Dtrace support is enabled and built if the build system(cmake)
  detects the "dtrace" command on the Solaris, Mac OS X and FreeBSD.
  As part of this WL, dtrace support is extented to the Oracle 
  Linux 6 with UEK3 kernel. Now Dtrace support is enabled and built
  if Oracle Linux 6 with UEK3 has real dtrace.
------------------------------------------------------------
revno: 8281
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 10:36:21 +0530
message:
  - Bug #19053151: SERVER CRASH IF INNODB TEMP TABLE IS USED
    (bug was logged on WL branch but is generic)
  
    Following workflow was not considered while adding an assert.
  
    Server -> Plugin -> SE
  
    In this case Plugin (I_S) share latch vector of SE and assert enforces
    latching semantics. I_S latches as such has nothing to do with SE latching
    order so Plugin should use separate namespace for its latches or we should
    relax the assert to consider the case.
  
    We used latter solution for now.
  
    Approved by: Sunny (over IM)
------------------------------------------------------------
revno: 8280 [merge]
committer: Gopal Shankar <gopal.shankar@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 09:55:03 +0530
message:
  Bug#18776592 INNODB: FAILING ASSERTION: PRIMARY_KEY_NO == -1 ||
                                          PRIMARY_KEY_NO == 0 
  
  null merge
    ------------------------------------------------------------
    revno: 3902.324.139 [merge]
    committer: Gopal Shankar <gopal.shankar@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-25 09:52:25 +0530
    message:
      Bug#18776592 INNODB: FAILING ASSERTION: PRIMARY_KEY_NO == -1 ||
                                              PRIMARY_KEY_NO == 0 
      
      null merge
        ------------------------------------------------------------
        revno: 2875.595.18
        committer: Gopal Shankar <gopal.shankar@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-06-25 09:50:17 +0530
        message:
          Bug#18776592 INNODB: FAILING ASSERTION: PRIMARY_KEY_NO == -1 ||
                                                  PRIMARY_KEY_NO == 0 
          
          This bug is a backport of the following revision of 5.6 source tree:
          # committer: Gopal Shankar <gopal.shankar@oracle.com>
          # branch nick: priKey56
          # timestamp: Wed 2013-05-29 11:11:46 +0530
          # message:
          #   Bug#16368875 INNODB: FAILING ASSERTION:
------------------------------------------------------------
revno: 8279 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-25 05:40:00 +0200
message:
  Disable memory barrier only for Intel CPU, because performance regression was observed at some conditions for Intel CPU.
  follow up for Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
    ------------------------------------------------------------
    revno: 3902.324.138
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-25 05:36:54 +0200
    message:
      Disable memory barrier only for Intel CPU, because performance regression was observed at some conditions for Intel CPU.
      follow up for Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
------------------------------------------------------------
revno: 8278 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-24 23:59:24 +0200
message:
  Merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.137
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-24 23:57:39 +0200
    message:
      fix build fail for win32.
      follow up for Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
------------------------------------------------------------
revno: 8277 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Tue 2014-06-24 21:16:48 +0200
message:
  WL#7415 PERFORMANCE SCHEMA, STAGE PROGRESS
  
  Merge to mysql-trunk
    ------------------------------------------------------------
    revno: 7409.1.15
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415-b
    timestamp: Tue 2014-06-24 00:11:00 +0200
    message:
      Improved test coverage
    ------------------------------------------------------------
    revno: 7409.1.14 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415-b
    timestamp: Mon 2014-06-23 20:58:16 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 7409.1.13 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415-b
    timestamp: Mon 2014-06-23 15:42:31 +0200
    message:
      merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.12
    committer: Tarique Saleem <tarique.saleem@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Mon 2014-06-16 19:51:55 +0530
    message:
      Added testcase to check if the new added columns WORK_ESTIMATED and WORK_COMPLETED exists. 
    ------------------------------------------------------------
    revno: 7409.1.11 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Fri 2014-06-13 02:39:33 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.10 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Mon 2014-06-02 14:00:45 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.9 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Fri 2014-04-18 14:39:34 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.8
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Fri 2014-04-18 13:38:04 +0200
    message:
      WL#7415 PERFORMANCE_SCHEMA STAGE PROGRESS
      
      Implemented code review comments
    ------------------------------------------------------------
    revno: 7409.1.7 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Mon 2014-04-07 15:20:19 +0200
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.6 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Tue 2014-03-04 11:13:53 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.5 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Tue 2014-03-04 09:51:01 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.4 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Wed 2014-02-26 13:55:47 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.3 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Tue 2014-02-11 16:12:28 +0100
    message:
      Merge mysql-trunk --> mysql-trunk-wl7415
    ------------------------------------------------------------
    revno: 7409.1.2
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Thu 2014-02-06 04:03:15 +0100
    message:
      WL#7415 PERFORMANCE SCHEMA, STAGE PROGRESS
    ------------------------------------------------------------
    revno: 7409.1.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-wl7415
    timestamp: Wed 2014-02-05 10:05:04 +0100
    message:
      New branch
------------------------------------------------------------
revno: 8276
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-sunstudio
timestamp: Tue 2014-06-24 17:54:17 +0200
message:
  Bug#19010286 BUILD WITH SUN C++ 5.13 SUNOS_SPARC BETA
  
  Look for libslport.so in a different location for Sun C++ 5.13
  Fix unit test item-t to work with Sun Studio.
  Fix a compiler warning in item_func.h
------------------------------------------------------------
revno: 8275
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Tue 2014-06-24 17:05:24 +0200
message:
  Bug#19054551 GET RID OF DYNAMIC_ARRAY IN HA_FEDERATED
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
  
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8274
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-winwarn
timestamp: Tue 2014-06-24 15:10:24 +0200
message:
  Bug#11752665: WINDOWS SERVER X64: SO MANY COMPILER WARNINGS
  
  Patch #9:
  
  This patch reduces the compiler warning count on Win64 by an additional
  350, as reported by VS2013.
------------------------------------------------------------
revno: 8273 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-24 14:25:42 +0200
message:
  Bug #19048563 : PERFORMANCE REGRESSION IN UPDATE OPS CAUSED BY TRUNK REVNO 8242 (BUG#11755438)
  
  Some of the added memory barrier (internal of spin loops of mutex/rw_lock) was too expensive. Removed.
  This is partial reverting of the fix for Bug#11755438.
  
  ------------------------------------------------------------
  revno: 8242 [merge]
  committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
  branch nick: mysql-trunk
  timestamp: Thu 2014-06-19 07:36:42 +0200
  message:
    Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
  
    Because of difference about memory ordering, some critical flags of mutex/rw_lock might be missed to read on non-Intel CPUs.
    Even for Intel-CPUs, the explicit memory barrier instruction might cause positive effects for performance.
  
    Approved by Kevin Lewis in rb#5466
  ------------------------------------------------------------
    ------------------------------------------------------------
    revno: 3902.324.136
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-24 14:22:27 +0200
    message:
      Bug #19048563 : PERFORMANCE REGRESSION IN UPDATE OPS CAUSED BY TRUNK REVNO 8242 (BUG#11755438)
      
      Some of the added memory barrier (internal of spin loops of mutex/rw_lock) was too expensive. Removed.
      This is partial reverting of the fix for Bug#11755438.
      
      ------------------------------------------------------------
      revno: 6004
      committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
      branch nick: mysql-5.6
      timestamp: Thu 2014-06-19 07:33:57 +0200
      message:
        Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
      
        Because of difference about memory ordering, some critical flags of mutex/rw_lock might be missed to read on non-Intel CPUs.
        Even for Intel-CPUs, the explicit memory barrier instruction might cause positive effects for performance.
      
        Approved by Kevin Lewis in rb#5561
      ------------------------------------------------------------
------------------------------------------------------------
revno: 8272 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-24 13:55:56 +0530
message:
  Merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.135
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-24 13:44:39 +0530
    message:
      Bug #19027905 ASSERT RET.SECOND DICT_CREATE_FOREIGN_CONSTRAINTS_LOW DICT_CREATE_FOREIGN_CONSTR
      
      Problem:
      
      This is a regression introduced by the fix of bug#18806829.  The assert
      fails when duplicate foreign key constraint names are there in a
      CREATE TABLE statement.
      
      Solution:
      
      The assert is replaced by an error return.
      
      approved by Vasil and Marko over IM.
------------------------------------------------------------
revno: 8271 [merge]
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Tue 2014-06-24 09:19:03 +0200
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.134 [merge]
    committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
    branch nick: mysql-5.6-test
    timestamp: Tue 2014-06-24 09:16:08 +0200
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.17
        committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
        branch nick: mysql-5.5-test
        timestamp: Tue 2014-06-24 09:13:01 +0200
        message:
          Bug#19001781: ADD SUPPORT FOR CMAKE 3
          
          Set CMP0026 and CMP0045 policies when using CMake 
          version 3 or higher to restore old CMake behavior.
------------------------------------------------------------
revno: 8270
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-06-24 08:33:25 +0200
message:
  Disabled until solution will be found,
------------------------------------------------------------
revno: 8269 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18618561_1
timestamp: Tue 2014-06-24 11:19:15 +0530
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.133
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18618561_1
    timestamp: Tue 2014-06-24 11:16:11 +0530
    message:
      BUG#18618561 - FAILED ALTER TABLE ENGINE CHANGE WITH 
                     PARTITIONS CORRUPTS FRM.
      
      Follow up patch to fix the test case.
------------------------------------------------------------
revno: 8268 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18618561
timestamp: Tue 2014-06-24 10:21:18 +0530
message:
  Merge from mysql-5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.132 [merge]
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18618561
    timestamp: Tue 2014-06-24 10:19:06 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.16
        committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
        branch nick: mysql-5.5-18618561
        timestamp: Tue 2014-06-24 10:15:53 +0530
        message:
          BUG#18618561: FAILED ALTER TABLE ENGINE CHANGE WITH PARTITIONS
                        CORRUPTS FRM
          
          Analysis:
          ---------
          ALTER TABLE on a partitioned table resulted in the wrong
          engine being written into the table's FRM file and displayed
          in SHOW CREATE TABLE.
          
          The prep_alter_part_table() modifies the partition_info object
          for TABLE instance representing the old version of table.
          If the ALTER TABLE ENGINE statement fails, the partition_info
          object for the TABLE contains the altered storage engine name.
          The SHOW CREATE TABLE uses the TABLE object to display the table
          information, hence displays incorrect storage engine for the table.
          Also a subsequent successful ALTER TABLE operation will write the
          incorrect engine information into the FRM file.
          
          Fix:
          ---
          A copy of the partition_info object is created before modification so
          that any changes would not cause the the original partition_info object
          to be modified if the ALTER TABLE fails.(Backported part of the code
          provided as fix for bug#14156617 in mysql-5.6.6).
------------------------------------------------------------
revno: 8267 [merge]
committer: Gleb Shchepa <gleb.shchepa@oracle.com>
branch nick: 18978946-trunk
timestamp: Mon 2014-06-23 20:07:06 +0400
message:
  null-merge: mysql-5.6 -> mysql-trunk (18978946)
    ------------------------------------------------------------
    revno: 3902.324.131 [merge]
    committer: Gleb Shchepa <gleb.shchepa@oracle.com>
    branch nick: 18978946-5.6
    timestamp: Mon 2014-06-23 20:03:06 +0400
    message:
      manual up-merge: mysql-5.5 -> mysql-5.6 (18978946)
        ------------------------------------------------------------
        revno: 2875.595.15
        committer: Gleb Shchepa <gleb.shchepa@oracle.com>
        branch nick: 18978946-5.5
        timestamp: Mon 2014-06-23 19:59:15 +0400
        message:
          Bug #18978946: BACKPORT TO 5.6: BUGFIX FOR 18017820 "BISON 3 BREAKS MYSQL BUILD"
          
          Backport of the fix:
          
          : Bug 18017820: BISON 3 BREAKS MYSQL BUILD
          : ========================================    
          : 
          : The source of the reported problem is a removal of a few deprecated
          : things from Bison 3.x: 
          : * YYPARSE_PARAM macro (use the %parse-param bison directive instead),
          : * YYLEX_PARAM macro (use %lex-param instead),
          : 
          : The fix removes obsolete macro calls and introduces use of
          : %parse-param and %lex-param directives.
------------------------------------------------------------
revno: 8266
committer: Bharathy Satish <bharathy.x.satish@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-23 16:06:38 +0530
message:
  Bug #18636291 VALIDATE_PASSWORD_STRENGTH() FLOORS CPU ON CASTED INPUTS
  Problem: In mysql_string_iterator_next() the call to function
  cs->cset->ctype() returns MY_CS_TOOSMALL2 (whose value is -102) which added
  iterator->iterator_ptr thus makes ptr go beyond the existing input string
  length and causes the function to iterate in an infinite loop.
  Fix would be to have proper check to see if the iterator_ptr is < end before
  adding the next value.
------------------------------------------------------------
revno: 8265 [merge]
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-23 12:16:12 +0200
message:
  Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.324.130 [merge]
    committer: Erlend Dahl <erlend.dahl@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-23 12:13:12 +0200
    message:
      Bug#18850241 WRONG COPYRIGHT HEADER IN SOME STRINGS/CTYPE-* FILES
      
      Merging to 5.6
        ------------------------------------------------------------
        revno: 2875.595.14
        committer: Erlend Dahl <erlend.dahl@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-06-23 12:11:13 +0200
        message:
          Bug#18850241 WRONG COPYRIGHT HEADER IN SOME STRINGS/CTYPE-* FILES
------------------------------------------------------------
revno: 8264
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-23 16:51:29 +0800
message:
  Fix bug#18684241 ASSERT SLOT PAGE_ZIP_REC_SET_DELETED()
  RTR_PAGE_COPY_REC_LIST_END_NO_LOCKS().
  
  Issue 1 in this bug has been fixed in previous patch.
  For issue 2, there're 2 identical recs in 2 different pages cause it.
  We need to compare the page_no field of the parent node of these 2
  identical recs, to find the right parent.
  
  rb://5782 approved by Jimmy.
------------------------------------------------------------
revno: 8263
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk2
timestamp: Sat 2014-06-21 14:11:39 +0800
message:
  Bug#18911119 assertion failed: no_data(nbytes)
  Check length for point GEOMETRY buffer to survive brutal-force truncates
  of geometry blobs; Correct length checking bugs in init_from_wkb functions.
  
  Before this change, if a GEOMETRY blob containing a point is truncated by cast
  or other function and the result is sent to GIS functions,it might cause
  buffer overrun asserts, because we didn't expect such truncated and incomplete
  blobs for points, we assumed the GEOMETRY buffer for point is always intact
  and valid.
  Now when scanning the point GEOMETRY blob buffer, we always do buffer length
  check so that we can detect truncated/incomplete buffer and avoid buffer
  overrun asserts. Also, length checking for WKB of geometry collection,
  multipolygon and polygon is corrected.
------------------------------------------------------------
revno: 8262
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Fri 2014-06-20 17:31:14 +0200
message:
  Bug#19029426 GET RID OF DYNAMIC_ARRAY IN MY_LOAD_DEFAULTS()
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
  
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8261
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-decimal
timestamp: Fri 2014-06-20 17:13:01 +0200
message:
  Bug#18563112 SELECT INTO VARIABLE CRASH IN USER_VAR_ENTRY::STORE
  
  In Item_func_set_user_var::update_hash:
  call entry->set_null_value() rather than entry->store() for NULL input.
------------------------------------------------------------
revno: 8260
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Fri 2014-06-20 21:19:54 +1000
message:
  Revert Bug #18805275 : PAGE CLEANER SHOULD BE MULTI-THREADED ALSO FOR RECOVERY AND SHUTDOWN
------------------------------------------------------------
revno: 8259 [merge]
committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 16:07:26 +0530
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.129
    committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-20 15:59:53 +0530
    message:
      Bug#18853696 : WRONG RESULT FOR COUNT DISTINCT OF RESULT OF
                           JOINED DERIVED TABLE
            
      Problem:
      Post the bugfix for Bug#11760197 count distinct on a const value was
      always made to return 1 irrespective of whether the join results were
      empty.
            
      Solution:
      count is set to '1' only after finding atleast one tuple in the result.
------------------------------------------------------------
revno: 8258
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 11:02:36 +0200
message:
  Bug #18805275 : PAGE CLEANER SHOULD BE MULTI-THREADED ALSO FOR RECOVERY AND SHUTDOWN
  
  From 5.7, page cleaner is multi-threaded for performance scalability.
  But it is not used during shutdown and recovery phases.
  
  It should be multi-threaded during shutdown and recovery phases for their
  optimal performance.
  
  Approved by Sunny in rb#5465
------------------------------------------------------------
revno: 8257
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Fri 2014-06-20 10:30:24 +0200
message:
  Bug#18994745 GET RID OF DYNAMIC_ARRAY IN MYSQL.CC MYSQLBINLOG.CC
    
  Our array abstractions should be removed, and substituted with modern data
  structures.
          
  Templatized vectors/maps are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8256
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 09:51:25 +0200
message:
  Bug#18977128 : ASSERT RW_LATCH == RW_NO_LATCH || BUF_BLOCK_GET_STATE(BLOCK) == BUF_BLOCK_FILE_P
  
  The assertion code is invalid for current implementaion. removed it.
  
  Approved by Sunny on IM
------------------------------------------------------------
revno: 8255 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 08:51:00 +0200
message:
  Merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.128
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-20 08:48:48 +0200
    message:
      Bug#15923864 : Bug#67718 : INNODB DRASTICALLY UNDER-FILLS PAGES IN CERTAIN CONDITIONS
      
      InnoDB should try to insert to the next page before split the page, if the insert record for split_and_insert is last of the page.
      
      Otherwise, the (reverse) sequential inserts to the middle of the index might cause 1 record per 1 page.
      
      Approved by Marko in rb#5563
------------------------------------------------------------
revno: 8254
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 07:48:37 +0200
message:
  Fixes result mismatch.
------------------------------------------------------------
revno: 8253
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 11:10:50 +0530
message:
  Adding the suite interactive_utilities to run on pb2
------------------------------------------------------------
revno: 8252
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-20 06:07:21 +0200
message:
  Removed rpl_binlog_n_mix_MTS-debug and rpl_binlog_ps_row-debug runs from Per push runs. This is expected to save around 30 min. These tests are covered in the daily by ps-debug-big and rpl_MTS-debug_db_part respectively.
------------------------------------------------------------
revno: 8251
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-06-19 17:42:23 +0200
message:
  Review 1 of wl6647. 
------------------------------------------------------------
revno: 8250 [merge]
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-wl7436
timestamp: Thu 2014-06-19 17:30:09 +0200
message:
  WL#7436: Deprecate and remove timed_mutexes system variable 
  
  This is the 5.7 version of the patch.
  
  Remove the timed_mutexes system variable.
    ------------------------------------------------------------
    revno: 3902.324.127 [merge]
    committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-19 17:21:29 +0200
    message:
      WL#7436: Deprecate and remove timed_mutexes system variable
      
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.13
        committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-06-19 16:47:41 +0200
        message:
          WL#7436: Deprecate and remove timed_mutexes system variable 
          
          This is the 5.5/5.6 version of the patch.
          
          Add deprecation warning for timed_mutexes.
------------------------------------------------------------
revno: 8249 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-19 20:46:50 +0530
message:
  Bug#18734396    INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
          Reverting the patch due to failure in trunk.
    ------------------------------------------------------------
    revno: 3902.324.126
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: t-5.6
    timestamp: Thu 2014-06-19 20:41:09 +0530
    message:
      Bug#18734396	INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
      	Reverting the patch due to failure in trunk.
------------------------------------------------------------
revno: 8248 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Thu 2014-06-19 15:06:04 +0200
message:
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 8233.1.2 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18991366
    timestamp: Thu 2014-06-19 15:02:55 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 8233.1.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18991366
    timestamp: Wed 2014-06-18 12:11:47 +0200
    message:
      Bug#18991366 SAFE_HASH::MUTEX (MF_KEYCACHES.C) IS NOT INSTRUMENTED
      
      Before this fix, the rwlock used in the SAFE_HASH implementation
      was not instrumented for the performance schema.
      
      With this fix, the rwlock is now instrumented.
      
      Also fixed the misleading variable name.
------------------------------------------------------------
revno: 8247 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Thu 2014-06-19 13:49:07 +0200
message:
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 8197.1.3 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18900309
    timestamp: Thu 2014-06-19 13:40:58 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 8197.1.2
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18900309
    timestamp: Wed 2014-06-18 07:23:21 +0200
    message:
      Bug#18900309 PERFORMANCE SCHEMA MEMORY INSTRUMENTATION IGNORES SETUP TABLES
      
      This is a performance bug, in pfs_memory_alloc_v1().
      
      The recently added performance schema instrumentation for memory
      does not honor:
      - the per instrument enabled flag in table setup_instruments
      - the consumers in table setup_consumers
      
      As a result, memory allocation is always instrumented,
      even when the instrumentation is supposed to be turned off.
      This leads to un necessary statistics collection,
      and to performance overhead.
    ------------------------------------------------------------
    revno: 8197.1.1 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18900309
    timestamp: Wed 2014-06-18 05:17:14 +0200
    message:
      Local merge
------------------------------------------------------------
revno: 8246
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-19 13:16:50 +0300
message:
  Bug#19020434 REMOVE ORPHAN METHOD HANDLER::UPDATE_TABLE_COMMENT
  
  Before MySQL 5.5, the method handler::update_table_comment used to return
  additional metadata in certain SHOW statements.
  The method is not used any more.
------------------------------------------------------------
revno: 8245
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-winwarn
timestamp: Thu 2014-06-19 09:46:48 +0200
message:
  Bug#11752665: WINDOWS SERVER X64: SO MANY COMPILER WARNINGS
  
  Patch #8:
  
  This patch reduces the compiler warning count on Win64 by an additional
  230, as reported by VS2013.
------------------------------------------------------------
revno: 8244
committer: Narendra Chauhan <narendra.chauhan@oracle.com>
branch nick: mysql-5.7
timestamp: Thu 2014-06-19 12:38:37 +0530
message:
  Bug#18128323 - RPL.RPL_OPTIMIZE FAILED WITH TIMEOUT 900 IN DAILY-TRUNK.
  
  With InnoDB engine on SLOW Solaris/WINT machines and debug builds 
  test is failing due to timeout. So, we will be using MyISAM engine 
  as default until that issue is sorted out.
------------------------------------------------------------
revno: 8243 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-19 11:18:13 +0530
message:
  Bug #17657223   EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.125
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: t-5.6
    timestamp: Thu 2014-06-19 11:16:15 +0530
    message:
      Bug #17657223	EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
      
      Analysis:
      Temporary file usage during alter table operation can be avoided
      during clustered index rebuild when new primary key follow existing
      pk order. So it will reduce the file usage during clustered index rebuild.
      Delaying the temporary file creation and log file creation
      for alter table operation will lead to avoid the file creation
      for smaller tables.
      
      	Approved by marko, shaohua (rb-4788)
------------------------------------------------------------
revno: 8242 [merge]
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-19 07:36:42 +0200
message:
  Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
  
  Because of difference about memory ordering, some critical flags of mutex/rw_lock might be missed to read on non-Intel CPUs.
  Even for Intel-CPUs, the explicit memory barrier instruction might cause positive effects for performance.
  
  Approved by Kevin Lewis in rb#5466
    ------------------------------------------------------------
    revno: 3902.324.124
    committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-19 07:33:57 +0200
    message:
      Bug#11755438: Bug#47213: INNODB MUTEX/RW_LOCK SHOULD BE CONSCIOUS ABOUT MEMORY ORDERING OTHER THAN INTEL
      
      Because of difference about memory ordering, some critical flags of mutex/rw_lock might be missed to read on non-Intel CPUs.
      Even for Intel-CPUs, the explicit memory barrier instruction might cause positive effects for performance.
      
      Approved by Kevin Lewis in rb#5561
------------------------------------------------------------
revno: 8241
committer: Gleb Shchepa <gleb.shchepa@oracle.com>
branch nick: 18759387-trunk
timestamp: Wed 2014-06-18 17:59:56 +0400
message:
  Bug #18759387: PROBLEM IN ITEM_FUNC_XOR::NEG_TRANSFORMER
  
  Negation of expressions that include native function calls with wrong parameter counts caused incorrect server execution.
  
  This bug is a result of the typo in the WL7199 code: "false" value instead of "true".
------------------------------------------------------------
revno: 8240
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug11751331
timestamp: Wed 2014-06-18 12:52:13 +0200
message:
  Bug#11751331: CONCURRENT DML AND LOCK TABLE ... READ FOR INNODB
                TABLE CAUSE WARNINGS IN ERRLOG
  
  The original problem was that concurrent execution of
  LOCK TABLES ... READ statement and DML statements affecting
  the same InnoDB table on debug builds of MySQL server might lead
  to "Found lock of type 6 that is write and read locked" warnings
  appearing in error log.
  
  This could happen since one connection might  acquire TL_READ_NO_INSERT
  on a table while another connection held TL_WRITE_ALLOW_WRITE lock on
  the same table. At the same time, the locking code assumes that that
  such locks are incompatible.
  
  After WL#6671 this issue is no longer repeatable since InnoDB no
  longer uses thr_lock.c locks. It is also not repeatable using other
  in-house engines since they only use TL_READ_NO_INSERT for 
  LOCK TABLES READ and this statement is now handled by the MDL
  subsystem.
  
  This patch removes the MTR suppressions that were added to
  suppress the warnings generated and adjust a code comment.
------------------------------------------------------------
revno: 8239
committer: viswanatham gudipati <viswanatham.gudipati@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-18 15:22:55 +0530
message:
  These are very old version ,causes confusion,hence removing 
------------------------------------------------------------
revno: 8238
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-18 17:39:22 +0800
message:
  Fix intermittent failure on innodb_fts_misc.test. Remove the ranking value
  from result. The ranking is depending on statistics row count, and can vary.
------------------------------------------------------------
revno: 8237 [merge]
committer: viswanatham gudipati <viswanatham.gudipati@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-18 14:55:01 +0530
message:
  Disabling the test to run on Solaris,Osx
    ------------------------------------------------------------
    revno: 3902.324.123
    committer: viswanatham gudipati <viswanatham.gudipati@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-18 14:48:25 +0530
    message:
      Disabling the test to run on Solrais,Osx 
------------------------------------------------------------
revno: 8236 [merge]
committer: Namit Sharma<namit.sharma@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-18 12:28:03 +0530
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.122 [merge]
    committer: Namit Sharma<namit.sharma@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-18 12:24:17 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.12
        committer: Namit Sharma<namit.sharma@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-06-18 12:22:09 +0530
        message:
          Bug#18949527 SUITE/BINLOG/T/BINLOG_KILLED.TEST FORGETS TO 
                       DISCONNECT CON1 AND CON2
            
          Problem:
          The test suite/binlog/t/binlog_killed.test makes the connections
          con1 and con2 but forgets to disconnect them + wait till that
          operation is finished at test end.
          This mistake has the potential to harm subsequent tests in
          case these tests depend on the content of the processlist.
           
          Solution:
          Added disconnect + wait_until_disconnected.inc 
          within the test cleanup.
------------------------------------------------------------
revno: 8235
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-18 12:14:40 +0530
message:
  Modifying the secure_installation test to use the correct  inc fileto skip on windows
------------------------------------------------------------
revno: 8234
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Wed 2014-06-18 13:31:43 +0800
message:
  Bug #18934104 WL7220: MEMORY ISSUE FOUND BY VALGRIND/ASAN IN UNIT TEST RUNS
  Fix a warning that's reported a lot on MSVC, by using char constant to assign to
  an element of a char array.
------------------------------------------------------------
revno: 8233
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-18 12:33:09 +0800
message:
  Fix valgrind error introduced by Bug #18674219 checkin
------------------------------------------------------------
revno: 8232
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-wl6671
timestamp: Wed 2014-06-18 08:28:30 +0400
message:
  WL#6671 "Improve scalability by not using thr_lock.c locks for InnoDB tables".
  
  Avoid scalability bottleneck associated with THR_LOCK::mutex locks for InnoDB
  tables by not using thr_lock.c locks for them. The patch tries to make minimal
  changes to SE API and locking code on SQL-layer. Further improvements in these
  areas enabled by this change will be done as separate WLs.
  
  Before this patch InnoDB downgraded strong TL_READ_NO_INSERT/TL_WRITE
  thr_lock.c locks to weaker ones compatible with each other in most cases.
  So it has really relied on thr_lock.c locking only in a few scenarios:
  
  1) To isolate HANDLER READ statements from LOCK TABLES WRITE statements.
  2) To isolate LOCK TABLES statements that lock tables for write implicitly
     or for read, both explicitly or implicitly, from concurrent DML statements.
  3) Due to coding mistake thr_lock.c lock was necessary to isolate ALTER TABLE
     IMPORT/DISCARD TABLESPACE under LOCK TABLES from concurrent I_S queries/
     open HANDLERs.
  4) To indicate that InnoDB tables don't support LOCK TABLES READ LOCAL by
     upgrading TL_READ lock requested by statement to TL_READ_NO_INSERT
  
  After addressing these scenarios it became possible to completely abandon
  thr_lock.c locking for InnoDB tables. To do this this patch:
  
  1)   Changes code for HANDLER READ statements to upgrade S metadata lock to
       SR metadata lock for the duration of read. This allows us properly
       isolate HANDLER READ from LOCK TABLES WRITE and makes metadata locking
       for these statements consistent with locking for other DML.
  2.a) Introduces new type of metadata lock - MDL_SHARED_READ_ONLY. This lock
       is similar to SR lock with exception that it is not compatible with SW
       locks. This type of lock is used as replacement for TL_READ_NO_INSERT
       thr_lock.c locks for tables locked by LOCK TABLES for read (both
       explicitly or implicitly).
       To preserve backward compatibility SRO lock was assigned lower priority
       than SW locks (acquired by DML that modify data). This means that stream
       of DML can lead to starvation of LOCK TABLES READ statement. To provide
       a way out from such a situation, MDL subsystem was changed to respect
       max_write_lock_count limit for SW locks as well. Also new
       MDL_SHARED_WRITE_LOW_PRIO type of lock was introduced. It has lower
       priority than SRO locks and is used by DML with LOW_PRIORITY clause
       (i.e. for the same DML which had lower priority than LOCK TABLES READ
       before the patch).
  2.b) Changes code for LOCK TABLES to acquire SNRW lock on tables implicitly
       locked for write to compensate for removal of TL_WRITE lock.
  2.c) After 2.a) and 2.b) were implemented it became impossible to predict
       in which order SNRW and SRO locks will be acquired, so we no longer
       can rely on that "strong" locks are always acquired in the same order
       to avoid deadlocks for DDL.
       To solve this issue and keep behavior compatible we had to change
       function which chooses deadlock resolution victim to prefer
       waits for "strong" locks from LOCK TABLES over locks from other DDL
       (MDL subsystem was extended for this) and to ensure that attempt to
       acquire locks for LOCK TABLES is restarted when we get ER_LOCK_DEADLOCK
       error.
  3)   Changes ALTER TABLE IMPORT/DISCARD TABLESPACE code to acquire X
       lock on table being imported/discarded even under lock tables.
  4)   New storage engine flag HA_NO_READ_LOCAL_LOCK was introduced to mark
       storage engines which don't support LOCK TABLES READ LOCAL but don't
       want to use thr_lock.c locks to indicate this. LOCK TABLES READ LOCAL
       automatically acquires SRO locks for them.
  
  After the above steps InnoDB code was changed to indicate that InnoDB
  no longer needs thr_lock.c locks acquired. This was done by changing
  ha_innobase::lock_count() to return 0 and by ensuring that
  ha_innobase::store_lock() doesn't try to store type of thr_lock.c lock
  in MYSQL_LOCK::locks[] array it gets as a parameter.
  
  
  It is worth to emphasize the following non-obvious behavior changes
  caused by this patch:
  
  *) LOCK TABLES READ blocks and is blocked by concurrent transactions
     changing the table, for all storage engines, similarly to how
     LOCK TABLES WRITE work now.
  *) Tables which are implicitly used by LOCK TABLES (e.g. through view
     or trigger) are locked using metadata locks in addition to (all SEs
     except InnoDB) or instead of (InnoDB) THR_LOCK locks. This means that
     previous item also applies to such tables.
  *) Multi-update is blocking/starts to be blocked by concurrent LOCK TABLES
     READ on any table from its join, even though such table will be only used
     for reading and won't be updated.
  
  Many test cases which were relying on old behavior had to be adjusted.
  Particularly:
  
  *) HANDLER-related tests had to be adjusted to take into account that
     HANDLER READ will wait for and acquire SR lock.
  *) Some tests using thr_lock.c locks had to be adjusted to use statements
     other than LOCK TABLES READ.
  *) Some tests which waited until LOCK TABLES READ will block DML on thr_lock.c
     locks now wait for blocking on MDL.
  *) P_S test coverage for aggregates now either uses MyISAM tables or were
     updated to take into account that InnoDB doesn't acquire thr_lock.c locks.
  *) Coverage for new behavior and new types of MDL locks were added, as well
     as unit tests for the latter.
------------------------------------------------------------
revno: 8231
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 17:45:15 +0530
message:
  Disabling these tests since they are not stable.
------------------------------------------------------------
revno: 8230
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 17:43:39 +0530
message:
  Adding the interactive_utilities suite to default.daily 
------------------------------------------------------------
revno: 8229
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 17:27:54 +0530
message:
  Removing interactive_utilities from the default list of suites to be run
------------------------------------------------------------
revno: 8228 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 16:21:47 +0530
message:
  Bug#18742916 MYSQLBINLOG --RAW DOES NOT CHECK FOR ERRORS
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.121
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-17 16:20:31 +0530
    message:
      Bug#18742916 MYSQLBINLOG --RAW DOES NOT CHECK FOR ERRORS
      
      Problem: mysqlbinlog --raw does not check for fail write errors
      
      Fixing after push pb2 failure
------------------------------------------------------------
revno: 8227
committer: Gleb Shchepa <gleb.shchepa@oracle.com>
branch nick: 18484088-trunk
timestamp: Tue 2014-06-17 13:50:14 +0400
message:
  Bug #18484088: PROBLEMS IN CREATE_FUNC_CAST ON QUERY ERRORS...
        
  Second minor after-push updates to tests results.
------------------------------------------------------------
revno: 8226
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 14:30:44 +0530
message:
  Adding a new test perl expect to automate mysql_secure_install. Include an inc file to skip
  the test if the expect module is not found.
------------------------------------------------------------
revno: 8225 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 10:05:41 +0200
message:
  Merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.120 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-17 10:03:17 +0200
    message:
      Merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.595.11
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-06-17 09:59:46 +0200
        message:
          Bug#18972488 Remove packaging/rpm-uln directory from source - Updated CMakeLists.txt
------------------------------------------------------------
revno: 8224 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 09:40:04 +0200
message:
  Merge from 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.119 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-17 09:31:44 +0200
    message:
      Merge from 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.595.10
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-06-17 09:22:26 +0200
        message:
          Bug#18972488 Remove packaging/rpm-uln directory from source
------------------------------------------------------------
revno: 8223
committer: Gleb Shchepa <gleb.shchepa@oracle.com>
branch nick: 18484088-trunk
timestamp: Tue 2014-06-17 11:11:57 +0400
message:
  Bug #18484088: PROBLEMS IN CREATE_FUNC_CAST ON QUERY ERRORS...
  
  Minor after-push updates to tests results and to unit tests.
------------------------------------------------------------
revno: 8222
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 08:53:16 +0200
message:
  Fixed a formatting problem in the file
------------------------------------------------------------
revno: 8221 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug18432495_mysql-trunk
timestamp: Tue 2014-06-17 10:43:30 +0530
message:
  Bug#18432495:RBR REPLICATION SLAVE CRASHES WHEN DELETE
  NON-EXISTS RECORDS
  
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.118 [merge]
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug18432495_mysql-5.6
    timestamp: Tue 2014-06-17 10:40:00 +0530
    message:
      Bug#18432495:RBR REPLICATION SLAVE CRASHES WHEN DELETE
      NON-EXISTS RECORDS
      
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.9
        committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
        branch nick: Bug18432495_mysql-5.5
        timestamp: Tue 2014-06-17 10:38:27 +0530
        message:
          Bug#18432495:RBR REPLICATION SLAVE CRASHES WHEN DELETE
          NON-EXISTS RECORDS
          
          Renamed test script.
------------------------------------------------------------
revno: 8220
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-17 09:49:47 +0800
message:
  Simplify rtree_create_inplace.test for bug #18674219 to avoid intermittent
  failure. Will extend it again once the issue is resolved.
------------------------------------------------------------
revno: 8219
committer: Gleb Shchepa <gleb.shchepa@oracle.com>
branch nick: 18484088-trunk
timestamp: Tue 2014-06-17 03:48:52 +0400
message:
  Bug #18484088: PROBLEMS IN CREATE_FUNC_CAST ON QUERY ERRORS...
  
  Nested call of CONVERT() function with oversized DECIMAL type caused
  a NULL pointer access on error message sending.
  
  This patch fixes two problems:
  
  1. After the WL#7200 some syntax/type error processing is deferred: for
  example, the create_func_cast can return NULL pointer in case of
  type error, then the parser continues its processing until the error
  check at the top level: the NULL result of the create_func_cast is just an
  part of the parse tree that we don't normally access.
  However, the create_func_cast function tries to access internals of it
  1st parameter at the parse stage (before the contextualization of the parse
  tree): in case of type error it refers the Item::item_name field.
  If that 1st parameter if a NULL pointer (as a result of some previous
  syntax error), then the server fails with a NULL pointer access error.
  
  That issue has been fixed by using the part of parser input data 
  instead of Item::item_name field value for error messaging.
  
  2. Before the fix the CONVERT() function sent malformed error messages when the
  1st parameter was an expression: for trivial constant/identifier parameter the
  Item::item_name field contains something reasonable (literal constant
  value, identifier name etc). For expression parameter that value is an empty
  string, so the CONVERT() function send malformed error messages in case of
  type errors, for example:
  
    Too big precision 66 specified for column ''. Maximum is 65.
  
  The fix for the first problem have fixed this one as well automatically.
  
  The
     "Too big precision %d specified for column '%-.192s'. Maximum is %lu."
  error message has been replaced with:
    "Too-big precision %d specified for '%-.192s'. Maximum is %lu.".
  
  Also some minor Bison warnings have been fixed in the sql_yacc.yy file.
------------------------------------------------------------
revno: 8218
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Mon 2014-06-16 17:16:52 +0200
message:
  Bug#18957679 GET RID OF DYNAMIC_ARRAY IN MYSQLTEST.CC
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
        
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8217
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Mon 2014-06-16 16:06:00 +0200
message:
  Bug#18913949: USE REASONABLE DEFAULTS WHEN INITIALIZING
                THD/TRANS MEM_ROOTS
  
  Post-push fix: Use old defaults for TC unit test.
------------------------------------------------------------
revno: 8216
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Mon 2014-06-16 15:09:21 +0200
message:
  Bug#18920203 GET RID OF DYNAMIC_ARRAY IN SERVER_IDS::DYNAMIC_I
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Also: remove some dead code.
  The class Database_ids is unused, which means we can remove it,
  and the base class Dynamic_ids.
  The member function Server_ids::do_search_id is also unused,
  and can be removed.
  
  The macro sort_dynamic() is no longer in use, and can be removed.
  
  There was a bug in change_receive_options():
  If we are inserting multiple ids into mi->ignore_server_ids
  then we need to re-sort before doing a binary_search.
  The old code worked "by accident", bsearch() found an element,
  even though the array was no longer sorted after an insert_dynamic()
  
  The fix for the sort/bsearch() bug: implement insertion-sort in Prealloced_array,
  and use insert_unique() to keep the array sorted at all times.
  This is a short array of ints, with pre-allocation, so insert-sort should be fairly efficient.
------------------------------------------------------------
revno: 8215
committer: Shivji Jha <shivji.jha@oracle.com>
branch nick: reset-slave
timestamp: Mon 2014-06-16 18:03:24 +0530
message:
  Bug #18816897- IGNORE_SERVER_IDS NOT RESETTING POST RESET SLAVE ALL
  
  RESET SLAVE ALL is supposed to clear everything
  that is set by CHANGE MASTER TO.
  Following RESET SLAVE ALL, the server should
  forget that it was used as a slave. Because of
  the buggy behaviour RESET SLAVE ALL did not
  clear ignore_server_ids. This followed by
  by a new replication set up still contained the
  old list of server-ids.
  
  Fixed by emptying the list contaning server-ids
  to ignore on RESET SLAVE ALL.
------------------------------------------------------------
revno: 8214
committer: Narendra Chauhan <narendra.chauhan@oracle.com>
branch nick: mysql-5.7
timestamp: Mon 2014-06-16 17:34:56 +0530
message:
  Bug#18977833 - RPL_AUTO_INCREMENT.TEST IS NOT REPEATABLE
  
  The testcase rpl_auto_increment violates the rule that a test must be repeatable.
  There was no line in the cleanup which removes master-bin.saved.
  As a side note - A file with a name like 'master-bin.saved' is some serious 
  risky for tests being developed in future.
  
  FIX:
  1) Renaming 'master-bin.saved' to 'master-bin-rpl-auto-increment.saved'.
  2) Removing 'master-bin-rpl-auto-increment.saved', once no longer required.
------------------------------------------------------------
revno: 8213
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Mon 2014-06-16 13:33:34 +0200
message:
  WL#7914: Remove custom rwlock Windows implementation
  
  Additional patch: Remove native_condattr_t since it is never used.
  (Similar to pthread_rwlockattr_t)
------------------------------------------------------------
revno: 8212
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-16 16:57:47 +0530
message:
  WL#6614 : Define and reimplement IGNORE
  WL#6891 : Define and reimplement STRICT MODE
  
  Post push patch
  Remove unused parameter from mysql_prepare_insert().
------------------------------------------------------------
revno: 8211
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-decimal
timestamp: Mon 2014-06-16 12:28:00 +0200
message:
  Bug#18839617 GLOBAL-BUFFER-OVERFLOW IN DECIMAL_BIN_SIZE
  
  For decimal modulo operations, we need an extra byte to hold the result for the case:
  <signed number> % <unsigned decimal where precision == scale>
------------------------------------------------------------
revno: 8210
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B18487724-trunk
timestamp: Mon 2014-06-16 11:51:34 +0300
message:
  Bug #18487724 SETTING DEFAULT_STORAGE_ENGINE CAN CRASH IN INTERN_PLUGIN_LOCK
  
  When for some reason the evaluation of the SET default_storage_engine returned a NULL value 
  the sys_var::check() code decided to just set the new storage engine to C NULL and continue.
  This caused problems down the road.
  Fixed by properly checking for a null value (val_str() returning a NULL) and propagating 
  the error up.
  Fixed spacing
------------------------------------------------------------
revno: 8209
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-16 09:35:09 +0100
message:
  WL#7742
  
  Follow up patch to make test run only on debug builds (since
  it relies on DBGU_SYNC facility).
------------------------------------------------------------
revno: 8208
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-16 14:58:28 +0800
message:
  Fix Bug #18674219 - IMPLEMENT ALTER TABLE...ALGORITHM=INPLACE FOR INNODB
  GIS INDEX CREATION 
  
  rb://5389 approved by marko
------------------------------------------------------------
revno: 8207 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug18432495_mysql-trunk
timestamp: Mon 2014-06-16 10:24:15 +0530
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.117 [merge]
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug18432495_mysql-5.6
    timestamp: Mon 2014-06-16 10:17:42 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.8
        committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
        branch nick: Bug18432495_mysql-5.5
        timestamp: Mon 2014-06-16 10:06:44 +0530
        message:
          Bug#18432495:RBR REPLICATION SLAVE CRASHES WHEN DELETE
          NON-EXISTS RECORDS
          
          Problem:
          ========
          In RBR replication, master deletes a record but the record
          don't exist on slave. when slave tries to apply the
          Delete_row_log_event from master, it will result in an
          assert on slave.
          
          Analysis:
          ========
          This problem exists not only with Delete_rows event but also
          with Update_rows event as well. Trying to update a non
          existing row on the slave from the master will cause the
          same assert.  This assert occurs only for the tables that
          doesn't have primary keys and which basically require
          sequential scan to be done to locate a record. This bug
          occurs only with innodb engine not with myisam.
          
          When update or delete rows is executed on a slave on a table
          which doesn't have primary key the updated record is stored
          in a buffer named table->record[0] and the same is copied to
          table->record[1] so that during sequential scan
          table->record[0] can reloaded with fetched data from the
          table and compared against table->record[1].  In a special
          case where there is no record on the slave side scan will
          result in EOF in that case we reinit the scan and we try to
          compare record[0]  with record[1] which are basically the
          same. This comparison is incorrect. Since they both are the
          same record_compare() will report that record is found and
          we try to go ahead and try to update/delete non existing
          row. Ideally if the scan results in EOF means no data found
          hence no need to do a record_compare() at all.
          
          Fix:
          ===
          Avoid comparision of records on EOF.
------------------------------------------------------------
revno: 8206 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-06-14 12:25:59 +0530
message:
  Merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.116
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Sat 2014-06-14 12:12:49 +0530
    message:
      Bug #18806829 OPENING INNODB TABLES WITH MANY FOREIGN KEY REFERENCES IS
      SLOW/CRASHES SEMAPHORE
      
      Third post push fix.  Fixing a memory leak reported by valgrind.
      
      approved by Sunny over IM.
------------------------------------------------------------
revno: 8205
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-wl6074
timestamp: Fri 2014-06-13 14:07:24 +0200
message:
  WL#6074: Memroot allocator for C++ STL
  
  This patch implements two custom memory allocators for C++ STL:
  1) Memroot_allocator
     A memory allocator which uses MEM_ROOT.
  
  2) Malloc_allocator
     A memory allocator based on my_malloc/my_free.
     This allows for P_S memory instrumentation of heap memory allocation
     done internally by STL container classes.
------------------------------------------------------------
revno: 8204 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-13 15:04:25 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.115
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-13 15:00:06 +0530
    message:
      Bug #18806829 OPENING INNODB TABLES WITH MANY FOREIGN KEY REFERENCES IS
      SLOW/CRASHES SEMAPHORE
      
      Second follow on push. This is to fix a use of uninitialized variable
      (identified by compilation failure of optimized build).
      
      approved by Sunny over IM.
------------------------------------------------------------
revno: 8203
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-plain
timestamp: Fri 2014-06-13 10:24:11 +0200
message:
  Unbreak Solaris 32 bit builds:
  error: "_FILE_OFFSET_BITS" redefined
  
  Solution: Include my_global.h first for correct defines.
------------------------------------------------------------
revno: 8202 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-13 13:10:39 +0530
message:
  Merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.114
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-13 12:54:57 +0530
    message:
      Bug #18806829 OPENING INNODB TABLES WITH MANY FOREIGN KEY REFERENCES IS
      SLOW/CRASHES SEMAPHORE
      
      Follow on push to fix a use of uninitialized variable (identified by
      compilation failure of optimized build).
      
      approved by Marko over IM.
------------------------------------------------------------
revno: 8201 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: trunk
timestamp: Fri 2014-06-13 12:34:47 +0530
message:
  Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
  	Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.113
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-13 12:25:56 +0530
    message:
      Bug #18734396	INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
      
      Analysis:
      	When an InnoDB in-place ALTER fails, it leaves behind both a
      temporary filename like "#sql-ibtid" where tid represents the table ID of
      the table being altered) in its data dictionary. This makes future in-place
      ALTERs of the same table impossible because the temporary table is named
      only with the table ID.
      
      Solution:
      	This patch is adding more uniqueness to the temporary file name. It
      creates a temporary tablename like
      "#sql-ibtid-inc" where 
               tid = the table ID
               inc = static global number that is initialized to a random
      distributed 32-bit number using ut_time() and ut_crc32().It is then 
      incremented atomically for each temporary file name assigned.
      
      	rb5580 Approved by Kevin
------------------------------------------------------------
revno: 8200
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-13 12:56:25 +0800
message:
  Fixed bug#18971483 INNOBASE_TRX_INIT() IS CALLED TWICE IN
  CHECK_TRX_EXISTS WHEN TRX == NULL.
  
  Approved by Sunny on IM.
------------------------------------------------------------
revno: 8199 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-13 10:17:31 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.112
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-13 10:16:40 +0530
    message:
      Bug #18806829 OPENING INNODB TABLES WITH MANY FOREIGN KEY REFERENCES IS
      SLOW/CRASHES SEMAPHORE
      
      Problem:
      
      There are 2 lakh tables - fk_000001, fk_000002 ... fk_200000.  All of them
      are related to the same parent_table through a foreign key constraint.
      When the parent_table is loaded into the dictionary cache, all the child table
      will also be loaded.  This is taking lot of time.  Since this operation happens
      when the dictionary latch is taken, the scenario leads to "long semaphore wait"
      situation and the server gets killed.
      
      Analysis:
      
      A simple performance analysis showed that the slowness is because of the
      dict_foreign_find() function.  It does a linear search on two linked list
      table->foreign_list and table->referenced_list, looking for a particular
      foreign key object based on foreign->id as the key.  This is called two
      times for each foreign key object.
      
      Solution:
      
      Change the linked lists table->foreign_list and table_referenced_list to
      std::set structures table->foreign_set and table->referenced_set.
      
      rb#5673 approved by Vasil.
------------------------------------------------------------
revno: 8198 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-13 10:00:02 +0530
message:
  Merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.333.1
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-12 18:28:31 +0530
    message:
      Bug #18806829 OPENING INNODB TABLES WITH MANY FOREIGN KEY REFERENCES IS
      SLOW/CRASHES SEMAPHORE
      
      Problem:
      
      There are 2 lakh tables - fk_000001, fk_000002 ... fk_200000.  All of them
      are related to the same parent_table through a foreign key constraint.
      When the parent_table is loaded into the dictionary cache, all the child table
      will also be loaded.  This is taking lot of time.  Since this operation happens
      when the dictionary latch is taken, the scenario leads to "long semaphore wait"
      situation and the server gets killed.
      
      Analysis:
      
      A simple performance analysis showed that the slowness is because of the
      dict_foreign_find() function.  It does a linear search on two linked list
      table->foreign_list and table->referenced_list, looking for a particular
      foreign key object based on foreign->id as the key.  This is called two
      times for each foreign key object.
      
      Solution:
      
      Change the linked lists table->foreign_list and table_referenced_list to
      std::set structures table->foreign_set and table->referenced_set.
      
      rb#5673 approved by Vasil.
------------------------------------------------------------
revno: 8197 [merge]
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-06-12 13:22:34 -0500
message:
  null merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.111
    committer: kevin.lewis@oracle.com
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-12 13:17:17 -0500
    message:
      Bug#18767811-INNODB REPORT "SPACE ID IN FSP HEADER" ERRORS WITH OWN
                   MULTI-FILE TABLESPACE
      
      In the current mysql-5.5 code, fil_read_first_page() contained an error
      in that even though a boolean was sent in to indicate whether this was
      the first file node or not, it always read the FSP_FLAGS.  The FSP header
      is only valid for page zero, in the first data file of the tablespace.
      So fil_read_first_page() in 5.5 is returning junk in the FSP_FLAGS
      pointer. This does not matter because the caller that is reading the
      system tablespace, the only tablespace with more than one datafile, does
      not use the flags returned for datafiles after the first one.
      
      For mysql-5.6, fil_read_first_page() was changed so that in addition to
      reading the flags field from the FSP header, it would also read the
      space_id using fsp_header_get_space_id(). But this function has a
      validity check in it that logs an error message if the space_id is not
      correct;
      
       [ERROR] InnoDB: Space id in fsp header 19742095, but in the page header 0
      
      So this error message is always put into the error log when there is
      data in more than one datafile of the system tablespace.  
      
      This error has no effect because the caller ignores the returned space_id
      for datafiles after the first one, just like it ignores the FSP_FLAGS.
      
      The fix is very simple.  Do not read these two FSP_HEADER flags if the
      datafile is not the first one for the tablespace.
      
      Approved by Marko and Vasil in RB#5630
------------------------------------------------------------
revno: 8196 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-12 12:30:26 -0300
message:
  Null merge from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.110
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-12 12:07:03 -0300
    message:
      Post fix patch for BUG#17620053 because of a variable that could be used uninitialized.
------------------------------------------------------------
revno: 8195
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-wl7914
timestamp: Thu 2014-06-12 17:18:57 +0200
message:
  WL#7914: Remove custom rwlock Windows implementation
  
  This patch:
  1) Removes custom rwlock implementation for Windows since all supported
     Windows versions now have TryAcquireSRWLockExclusive().
     This also means increasing _WIN32_WINNT to 0x0601.
  2) Enables SAFE_MUTEX for Windows.
  3) Refactors mysys mutex, rwlock and condition variable code.
------------------------------------------------------------
revno: 8194
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-12 16:55:17 +0300
message:
  Fix a compilation failure on non-debug builds.
------------------------------------------------------------
revno: 8193 [merge]
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-12 13:41:17 +0300
message:
  Bug#18965518 ADD DEBUG CHECK FOR TABLESPACE MATCH IN THE ADAPTIVE HASH INDEX
  
  The InnoDB adaptive hash index is lacking debug assertions for checking that
  the tablespace ID in the buffer blocks matches index->space.
  
  Various functions in btr0sea.cc: Add debug assertions
  
  	ut_ad(block->page.id.space() == index->space);
  
  buf_LRU_block_remove_hashed(): Do not reset the page_id.
  
  buf_LRU_block_free_hashed_page(): Reset the page_id here, after
  btr_search_drop_page_hash_index() has been invoked.
  
  rb#5698 approved by Jimmy Yang
    ------------------------------------------------------------
    revno: 7868.1.25
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Thu 2014-06-05 11:59:48 +0300
    message:
      Try to get a test to work on Windows.
    ------------------------------------------------------------
    revno: 7868.1.24
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-06-04 14:45:00 +0300
    message:
      WL#7806: Test InnoDB startup with a corrupted MLOG_FILE_NAME record.
    ------------------------------------------------------------
    revno: 7868.1.23
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-06-04 14:17:05 +0300
    message:
      WL#7806: Fix some diagnostics messages.
      
      ut_print_buf() could dereference memory outside of recv_sys->buf.
      
      Approved by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7868.1.22
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-06-04 14:14:03 +0300
    message:
      Correct a typo in a comment.
    ------------------------------------------------------------
    revno: 7868.1.21
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-06-04 11:46:20 +0300
    message:
      Add a test for InnoDB startup and recovery.
    ------------------------------------------------------------
    revno: 7868.1.20
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-06-04 11:34:42 +0300
    message:
      Enable a test on non-debug binaries, and on Valgrind.
    ------------------------------------------------------------
    revno: 7868.1.19
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Tue 2014-06-03 11:28:22 +0300
    message:
      WL#7806: Improve test coverage.
      
      Start an empty instance when an undo tablespace file cannot be created.
    ------------------------------------------------------------
    revno: 7868.1.18
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Tue 2014-06-03 11:05:34 +0300
    message:
      WL#7806: Improve test coverage.
      
      Start up with a system tablespace mismatch, using innodb-force-recovery
      to ignore the error (unknown file mentioned in a MLOG_FILE_NAME record).
    ------------------------------------------------------------
    revno: 7868.1.17
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Tue 2014-06-03 11:01:51 +0300
    message:
      WL#7806: Fix a test to work with innodb-page-size=4k.
      
      innodb.innodb-undo: New test, forked from innodb.innodb-wl6445-1.
    ------------------------------------------------------------
    revno: 7868.1.16
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Mon 2014-06-02 17:47:21 +0300
    message:
      WL#7806 test coverage improvement.
      
      Test undo log page overflow, and too big undo log records.
      
      Previously, there was no test that would exceed the maximum undo log
      record size when using temporary tables.
    ------------------------------------------------------------
    revno: 7868.1.15
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Mon 2014-06-02 17:00:29 +0300
    message:
      Clean up a test (non-functional change).
    ------------------------------------------------------------
    revno: 7868.1.14
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Fri 2014-05-30 23:02:29 +0300
    message:
      WL#7806: Fix a typo that was found during gcov analysis.
      
      fil_spaces_lookup(): Swap spaces->sys and spaces->undo with each other.
      
      This typo has no correctness implications, because the two fields are being
      treated symmetrically. The typo was found because the gcov run was only using
      the default parameter --innodb-undo-tablespaces=0, and the
      fil_names_write_if_was_clean() call was reported to never flagging
      spaces->sys dirty. It should have been spaces->undo instead.
      
      Approved by Kevin Lewis on IM.
    ------------------------------------------------------------
    revno: 7868.1.13
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Fri 2014-05-30 16:50:10 +0300
    message:
      WL#7806: Correct a comment.
    ------------------------------------------------------------
    revno: 7868.1.12
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Fri 2014-05-30 12:40:16 +0300
    message:
      WL#7806: Adjust a comment to address a review request.
    ------------------------------------------------------------
    revno: 7868.1.11
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Fri 2014-05-30 12:32:27 +0300
    message:
      WL#7806: Remove a branch prediction hint.
      
      While this hint should be correct, it did not have the desired impact
      in my single-threaded microbenchmark, on a non-debug server
      compiled with clang for GNU/Linux on AMD64.
    ------------------------------------------------------------
    revno: 7868.1.10
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Tue 2014-05-27 16:10:02 +0300
    message:
      WL#7806: Address some review comments from Jimmy Yang.
      
      fil_names_write_if_was_clean(), fil_names_dirty_and_write(): New
      functions, combining fil_names_dirty() and fil_names_write().
      
      fil_names_write(): Renamed from fil_names_write_low(). Always invoked
      while holding fil_system->mutex.
      
      mtr_t::Command::prepare_write(): Replace the calls to
      fil_names_dirty() and fil_names_write() with a single call to
      fil_names_write_if_was_clean(). This removes the speculative writing
      of MLOG_FILE_NAME records for user tablespaces (*.ibd files).
      Adjust some comments and debug assertions as well.
      
      dyn_buf_t::set_size(): Remove; this method is no longer needed.
    ------------------------------------------------------------
    revno: 7868.1.9 [merge]
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-05-21 16:48:54 +0300
    message:
      Merge mysql-trunk to mysql-trunk-wl7806.
    ------------------------------------------------------------
    revno: 7868.1.8
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-05-21 16:28:48 +0300
    message:
      WL#7806: Address review comments from Jimmy Yang.
      
      buf_LRU_remove_pages(), fil_check_pending_operations():
      Revert some unintended changes (leftover from discarded approach).
      
      fil_space_undo_check(): Rename to fil_space_undo_check_if_opened().
      
      fil_space_validate_for_mtr_commit(): New predicate for checking that
      a tablespace cannot be dropped or evicted during mtr_commit().
      
      fil_space_lookup(): Invoke fil_space_validate_for_mtr_commit().
      
      fil_spaces_lookup(): Replace the array of fil_space_t* with a struct.
      
      fil_names_write_low(): Renamed from fil_space_write(). This function
      will only be called from fil_names_clear() during log checkpoint.
      
      fil_names_write(): Invoke fil_space_validate_for_mtr_commit() and
      fil_names_write_low(). This will be invoked during mtr_commit() only.
    ------------------------------------------------------------
    revno: 7868.1.7 [merge]
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-05-14 23:53:20 +0300
    message:
      Merge mysql-trunk to mysql-trunk-wl7806.
    ------------------------------------------------------------
    revno: 7868.1.6
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-05-14 23:22:43 +0300
    message:
      WL#7806: Non-functional changes.
      
      Address some review comments by Kevin Lewis.
    ------------------------------------------------------------
    revno: 7868.1.5 [merge]
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Fri 2014-05-09 13:06:32 +0300
    message:
      Merge mysql-trunk to mysql-trunk-wl7806.
    ------------------------------------------------------------
    revno: 7868.1.4
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Tue 2014-04-22 10:25:52 +0300
    message:
      WL#7806: Try to address a test failure on Windows.
      
      fil_space_undo_check(): If the file node is not open, do not expect
      to find it in the fil_system->LRU.
    ------------------------------------------------------------
    revno: 7868.1.3
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Tue 2014-04-22 08:53:09 +0300
    message:
      WL#7806: Detect the current system tablespace size.
      
      We still skip the test if the tablespace name differs. The reason for this
      is that the test would hang if --mysqld=--innodb-data-file-path is passed.
      This is because the parameter would override whatever is specified in the
      configuration file. Also, we do not escape any ; that might have been
      specified in --innodb-data-file-path.
    ------------------------------------------------------------
    revno: 7868.1.2
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Sat 2014-04-19 10:11:19 +0300
    message:
      WL#7806: Call mtr_t::set_sys_modified() before trx_commit_low(),
      because a commit can update the TRX_SYS page.
    ------------------------------------------------------------
    revno: 7868.1.1
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7806
    timestamp: Wed 2014-04-16 14:50:23 +0300
    message:
      WL#7806 InnoDB: Log-based discovery of built-in tablespaces
      
      This is follow-up to
      WL#7142 InnoDB: Simplify tablespace discovery during crash recovery
      
      We will write MLOG_FILE_NAME for all persistent tablespaces, not only
      for *.ibd files. Currently, this includes the following:
      
      * The InnoDB system tablespace (ibdata*)
      * The InnoDB undo log tablespaces (undo*)
      
      On startup, the InnoDB system tablespace and the InnoDB redo log will be opened.
      If there are redo log records to be applied since the latest checkpoint,
      any tablespaces requiring cleanup will be opened and recovered based on
      MLOG_FILE_NAME records in the redo log.
      
      If the MLOG_FILE_NAME records for the system tablespace disagree with
      the server configuration affecting the data file names for the system
      tablespace, recovery will be aborted with an error message, before
      applying any redo log.
      
      After recovery, any undo log tablespaces for which no redo log records were
      applied will be opened based on existing mechanism. The system tablespace will
      remain open at all times.
      
      is_predefined_tablespace(): Remove. All redo-logged tablespaces will
      be treated in the same way.
      
      mtr_t::m_undo_space, mtr_t::set_undo_space(): New field and method, to
      associate an undo tablespace associate with the mini-transaction.
      
      mtr_t::m_modifies_sys_space, mtr_t::set_sys_modified(): New field and
      method, to note that the mini-transaction is modifying the system
      tablespace.
      
      mtr_t::set_spaces(): A kind of copy constructor that copies the
      information on modified tablespaces from another mini-transaction.
      
      mtr_t::is_undo_space(): A debug method to ensure that set_undo_space()
      has been called.
      
      mtr_t::Command::prepare_write(): Invoke fil_spaces_lookup() before
      log_mutex_enter(), to look up all tablespaces that were flagged as
      modified by the mini-transaction, and optimistically invoke
      fil_names_write() for the system tablespace. (This is equivalent to
      old behaviour introduced in WL#7142, with the exception that we may
      look up an undo tablespace and the system tablespace.) After
      log_mutex_enter(), truncate the log if the fil_names_write() was not
      needed. If an undo tablespace or the system tablespace were flagged
      and fil_names_dirty() holds for them, invoke fil_names_write() for
      them.
      
      dyn_buf_t::set_size(): New method, used for truncating unneeded
      MLOG_FILE_NAME records from the tail of mtr_t::log.
      
      mtr_write_log_t::m_len: Remove. We will write the entire log.
      The log can be truncated by dyn_buf_t::set_size().
      
      srv_undo_tablespaces, srv_undo_tablespaces_open: Set the initial value
      to 0 on server startup, so that fil_space_belongs_in_lru() will behave
      in a predictable way during redo log apply.
      
      trx_rseg_t: Note that space,page_no are constant and need not be
      protected by mutex.
      
      fil_load_single_file_tablespace(): Renamed from
      fil_load_single_table_tablespace().
      
      fil_space_system_check(): New function, to check that MLOG_FILE_NAME
      records match the system tablespace data files.
      
      fil_space_undo_check(): New function, to reopen possibly existing undo
      log files after redo log apply has completed.
      
      fil_spaces_lookup(): New function, to replace previous usage of
      fil_names_write(). At mini-transaction commit, this looks up all
      modified redo logged tablespaces.
      
      fil_names_write(): Renamed from fil_names_write_low(). We will get the
      fil_space_t* looked up by fil_spaces_lookup().
      
      fil_name_parse(): Support undo tablespaces and multi-file system
      tablespace.
------------------------------------------------------------
revno: 8192 [merge]
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-12 12:37:26 +0200
message:
  Null merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.109
    committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
    branch nick: mysql-5.6-bug18913935
    timestamp: Thu 2014-06-12 12:35:55 +0200
    message:
      Bug#18913935: REMOVE SUPPORT FOR LINUXTHREADS
      
      This patch removes support for LinuxThreads.
      It was superseded by NPTL in Linux 2.6 (2003).
      
      Partial backport of Bug#17007529 from 5.7 to 5.6.
------------------------------------------------------------
revno: 8191 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk3
timestamp: Thu 2014-06-12 12:39:06 +0800
message:
  Merge mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.108
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-06-12 12:05:19 +0800
    message:
      Fix testcase for Bug #18711306. The table row count is statistics value,
      so it could vary, result in indeterministic results.
------------------------------------------------------------
revno: 8190
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Thu 2014-06-12 09:53:27 +0800
message:
  Bug #18934104 WL7220: MEMORY ISSUE FOUND BY VALGRIND/ASAN IN UNIT TEST RUNS               
  In Gis_wkb_vector<>'s copy constructor, allocate extra memory and set the extra bytes with defined pattern.                                                
  After a Gis_polygon object is converted to unparsed WKB structure, it owns and frees such memory buffer.
  
  Coding style fix: define variables where they are to be used immediately in C++ code.
------------------------------------------------------------
revno: 8189 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-11 17:59:52 +0530
message:
  Bug#18742916 MYSQLBINLOG --RAW DOES NOT CHECK FOR ERRORS
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.107
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-11 16:18:57 +0530
    message:
      Bug#18742916 MYSQLBINLOG --RAW DOES NOT CHECK FOR ERRORS
      
      Problem: mysqlbinlog --raw does not check for fail write errors
      
      Analysis: We have two my_fwrite calls in mysqlbinlog.cc file
      which does not check for failures. If my_fwrite returns error and
      if we are not catching the error we could end up in corrupting
      written out binary logs without reporting any errors/warnings.
      
      Fix: catch my_fwrite return values in those two places
      and throw error if fails.
------------------------------------------------------------
revno: 8188
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-05-31 12:05:03 +0300
message:
  Remove unused code (index_update_t type definition).
------------------------------------------------------------
revno: 8187
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-11 11:38:20 +0200
message:
  Bug 16368498 - MYSQL_CONFIG_EDITOR FORGETS ABOUT FREEING SOME MEMORY 
  
  Memory tools detects memory leaks in mysql_config_editor.
  
  Fix:
  Closing opened files when they are not needed anymore.
  Reviewed by Kristofer.
------------------------------------------------------------
revno: 8186
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Wed 2014-06-11 09:48:30 +0200
message:
  Bug#18947996 GET RID OF DYNAMIC_ARRAY IN IGNORE_DB_DIRS_ARRAY
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
      
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8185
committer: Martin Hansson <martin.hansson@oracle.com>
branch nick: mysql-trunk-commit
timestamp: Thu 2014-06-05 14:39:48 +0200
message:
  Bug#18498344: SELECT WITH ALIAS NOT WORKING IN 5.7
  
  If you push a Name_resolution_context on the stack,
  you had better also pop it, or things will go awry.
------------------------------------------------------------
revno: 8184
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug18913949
timestamp: Wed 2014-06-11 08:59:07 +0200
message:
  Bug#18913949: USE REASONABLE DEFAULTS WHEN INITIALIZING
                THD/TRANS MEM_ROOTS
  
  When first initializing THD::main_mem_root and
  Transaction_ctx::m_mem_root, use the global values of the relevant
  system variables rather than dummy values. This prevents these
  memroots from causing excessive calls to malloc() when they are used
  before THD::init_for_queries() is called. For example, this reduces
  the number of malloc() calls during main.1st by almost 300.
  
  The patch also removes the unused set_prealloc_root() function.
------------------------------------------------------------
revno: 8183 [merge]
committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-11 10:47:15 +0530
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.106
    committer: Chaithra Reddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-06-11 10:42:29 +0530
    message:
      Fixing a test case failure on weekly-5.6 build
------------------------------------------------------------
revno: 8182 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk3
timestamp: Wed 2014-06-11 10:27:56 +0800
message:
  Merge with mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.105
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-10 18:50:10 +0800
    message:
      Fix Bug #18711306 - DELETED DOCUMENTS NOT GETTING SKIPPED WHEN DOING IDF
      CALCULATIONS 
      
      rb://5426 approved by Sunny Bains
------------------------------------------------------------
revno: 8181
committer: susan.koerner@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-06-10 10:57:30 -0600
message:
  20140609 - skoerner - changed innodb-thread-concurrency from 1000 to 50 to support the server release testing for 5.5, 5.6, and 5.7.
------------------------------------------------------------
revno: 8180
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-2
timestamp: Tue 2014-06-10 14:46:49 +0200
message:
  Bug #18945693 CRASH IN PRINT_TABLE_ARRAY AT SQL/SQL_LEX.CC ON 2ND EXEC OF PREPARED STATEMENT
  Introduced by the fix for Bug 18791851 .
  SQ4_alias1 has a join condition.
  In first execution, some optim_join_cond is created for this table,
  by get_optimizable_join_conditions().
  At end of this execution, the optim_join_cond pointer remains dangling
  (points to freed memory).
  In second execution, we prepare/optimize the top SELECT; this wants to
  print HAVING, and thus print the subquery containing SQ4_alias1. Note
  that this subquery has _not_ been optimized yet, in this second
  execution; so get_optimizable_join_conditions() has _not_ been called,
  so SQ4_alias1->optim_join_cond() is still the dangling pointer.
  We try printing the join condition of that table; because the pointer
  is dangling, it is not (Item*)1, so we assume it is usable, and crash:
      // Print join condition
      Item *const cond=
        (curr->select_lex->join && curr->optim_join_cond() != (Item*)1) ?
        curr->optim_join_cond() : curr->join_cond();
  To sum up: pointer is not reset to (Item*)1 at end of first execution,
  so is not trustable at second execution.
  Fix: reset dangling pointer at start of second execution.
  Moreover, I remove a redundant setting of having_for_explain
  which is already done at the goto label.
------------------------------------------------------------
revno: 8179 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk3
timestamp: Tue 2014-06-10 16:47:49 +0800
message:
  merge mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.104
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-10 15:59:35 +0800
    message:
      Fix Bug #18655281 - INNODB FTS INFORMATION SCHEMA PLUGINS CANNOT BE LOADED
      WHEN BUILT AS SHARED LIB 
      
      rb://5255 approved by Sunny Bains
------------------------------------------------------------
revno: 8178
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-10 13:36:09 +0530
message:
  Renaming the folder for which changes are being committed.
------------------------------------------------------------
revno: 8177
committer: Sneha Modi <sneha.modi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-10 13:26:51 +0530
message:
  Adding a new test to automate mysql_secure_installation client. Keeping it experimental now to check for stability.
------------------------------------------------------------
revno: 8176
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Tue 2014-06-10 17:58:35 +1000
message:
  Bug#18940008 COMMENT OF FUNCTION OS_EVENT_IS_SET IS INCORRECT
  
  Copy comment from os/os0event.h. Non-functional change.
------------------------------------------------------------
revno: 8175
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-calloc
timestamp: Tue 2014-06-10 09:19:05 +0200
message:
  Bug#18846682: REPLACE MALLOC+MEMSET WITH CALLOC
  
  Use calloc() rather than malloc() + memset(..., 0, ...).
  Microbenchmark shows this is faster for both glibc malloc and jemalloc.
------------------------------------------------------------
revno: 8174
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-10 10:33:06 +0400
message:
  Bug#18923685: PROPERLY INITIALIZE DB OPTION HASH TABLE
  
  Afterpush fix - made the test engine independent.
------------------------------------------------------------
revno: 8173 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-10 09:39:19 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.103 [merge]
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-10 09:37:00 +0530
    message:
      Null merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.595.7
        committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-06-10 09:35:50 +0530
        message:
          Bug #18806829 OPENING INNODB TABLES WITH MANY FOREIGN KEY REFERENCES IS
          SLOW/CRASHES SEMAPHORE
          
          Problem:
          
          There are 2 lakh tables - fk_000001, fk_000002 ... fk_200000.  All of them
          are related to the same parent_table through a foreign key constraint.
          When the parent_table is loaded into the dictionary cache, all the child table
          will also be loaded.  This is taking lot of time.  Since this operation happens
          when the dictionary latch is taken, the scenario leads to "long semaphore wait"
          situation and the server gets killed.
          
          Analysis:
          
          A simple performance analysis showed that the slowness is because of the
          dict_foreign_find() function.  It does a linear search on two linked list
          table->foreign_list and table->referenced_list, looking for a particular
          foreign key object based on foreign->id as the key.  This is called two
          times for each foreign key object.
          
          Solution:
          
          Introduce a rb tree in table->foreign_rbt and table->referenced_rbt, which
          are some sort of index on table->foreign_list and table->referenced_list
          respectively, using foreign->id as the key.  These rbt structures will be
          solely used by dict_foreign_find().  
          
          rb#5599 approved by Vasil
------------------------------------------------------------
revno: 8172 [merge]
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-06-09 09:46:22 -0500
message:
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.102
    committer: kevin.lewis@oracle.com
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-09 09:38:23 -0500
    message:
      Revert revision 5978 for Bug#18767811
      It contains a last minute unreveiwed change that caused a regression in some tests.
------------------------------------------------------------
revno: 8171 [merge]
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-09 12:02:26 +0400
message:
  Automerge 5.6 => trunk.
    ------------------------------------------------------------
    revno: 3902.332.1
    committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
    branch nick: b18923691-5.6
    timestamp: Sat 2014-06-07 11:59:37 +0400
    message:
      Fix for bug #18923691: KEEP CLIENT_REMEMBER_OPTIONS FOR COMPRESSED CONNECTIONS
      
      Don't reset CLIENT_REMEMBER_OPTIONS flag in case of compressed slave connection.
      So, in case of connection failure we don't call mysql_close_free_options() 
      and keep the options intact.
------------------------------------------------------------
revno: 8170 [merge]
committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-09 11:58:57 +0400
message:
  Automerge 5.6 => trunk.
    ------------------------------------------------------------
    revno: 3902.331.1
    committer: Ramil Kalimullin <ramil.kalimullin@oracle.com>
    branch nick: b18923685-5.6
    timestamp: Sat 2014-06-07 11:50:17 +0400
    message:
      Fix for bug #18923685: PROPERLY INITIALIZE DB OPTION HASH TABLE
      
      We should use case sensitive db names search if lower_case_table_names=0
      and case insensitive if 1 or 2.
      
      Changed initialization of the db option hash table.
      Test case added.
------------------------------------------------------------
revno: 8169 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-09 12:55:02 +0530
message:
  Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
          - NULL merge from mysql-5.6.
    ------------------------------------------------------------
    revno: 3902.324.101
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-09 12:52:50 +0530
    message:
      Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
      	- Reverting the patch due to test case faiure.Forget to add the
      	  test case.
------------------------------------------------------------
revno: 8168 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-09 12:43:06 +0530
message:
  Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
          - Reverting the patch due to test case faiure.
    ------------------------------------------------------------
    revno: 3902.324.100
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-09 12:41:22 +0530
    message:
      Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
      	- Reverting the patch due to test case faiure.
------------------------------------------------------------
revno: 8167
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk3
timestamp: Mon 2014-06-09 14:26:14 +0800
message:
  Fix Bug #18816381 - INNODB MEMCACHED PLUGIN CONFIGURATION - TEST DATABASE  
  
  Approved by Marko
------------------------------------------------------------
revno: 8166 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-09 10:07:29 +0530
message:
  Bug #18734396   INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
  	merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.99
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-06-09 10:04:33 +0530
    message:
      Bug #18734396	INNODB IN-PLACE ALTER FAILURES BLOCK FUTURE ALTERS
      
      This patch is adding more uniqueness to the temporary file name. It
      creates a temporary tablename like
      "#sql-ibtid-inc" where 
               tid = the table ID
               inc = static global number that is initialized to a random
      distributed 32-bit number using ut_time() and ut_crc32().It is then 
      incremented atomically for each temporary file name assigned.
------------------------------------------------------------
revno: 8165 [merge]
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Sat 2014-06-07 01:17:21 -0500
message:
  Merge Bug#18767811 from mysql-5.6.
  The fix is the same, but a little simpler in mysl-trunk due to the Datafile class.
    ------------------------------------------------------------
    revno: 3902.324.98
    committer: kevin.lewis@oracle.com
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-06 20:34:03 -0500
    message:
      Bug#18767811-INNODB REPORT "SPACE ID IN FSP HEADER" ERRORS WITH OWN
                   MULTI-FILE TABLESPACE
      
      In the current mysql-5.5 code, fil_read_first_page() contained an error
      in that even though a boolean was sent in to indicate whether this was
      the first file node or not, it always read the FSP_FLAGS.
      The FSP header is only valid for page zero, in the first data file of
      a tablespace.  So fil_read_first_page() in 5.5 is returning junk in the
      FSP_FLAGS pointer. This does not matter because the caller that is
      reading the system tablespace, the only tablespace with more than one
      datafile, does not use the flags returned for datafiles after the first
      one.
      
      For mysql-5.6, fil_read_first_page() was changed so that in addition to
      reading the flags field from the FSP header, it would also read the
      space_id using fsp_header_get_space_id(). But this function has a
      validity check in it that logs an error message if the space_id is not
      correct;
      
       [ERROR] InnoDB: Space id in fsp header 19742095,but in the page header 0
      
      The bug is that this error message is always put into the error log when
       there is data in more than one datafile of the system tablespace.  
      
      This error has no effect because the caller ignores the returned space_id
      for datafiles after the first one, just like it ignore the FSP_FLAGS.
      
      The fix is very simple.  Do not read these two FSP_HEADER flags if the
      datafile is not the first one for the tablespace.
      
      In addition, the text of the error message grammar is improved to be like this;
      
       [ERROR] InnoDB: Space id in fsp header is 19742095, but in the page header it is 0.
      
      Approved by Vasil at RB#5630
------------------------------------------------------------
revno: 8164 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Fri 2014-06-06 17:58:46 +0200
message:
  merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.97 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Fri 2014-06-06 17:23:29 +0200
    message:
      merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.595.6
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5-merge
        timestamp: Fri 2014-06-06 16:49:25 +0200
        message:
          Bug#18786138 SHA/MD5 HASHING FUNCTIONS DIE WITH "FILENAME" CHARACTER SET
          
          For charsets with no binary collation: use my_charset_bin.
------------------------------------------------------------
revno: 8163
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bugbug18701868decimal
timestamp: Fri 2014-06-06 14:36:32 +0200
message:
  Bug#18701868 CRASH IN MY_DECIMAL::SANITY_CHECK, ORDER BY CAST (INVALID GIS FUNCTION RESULT)
  
  Do not set null_value=true to signal error, in decimal contexts it may return
  a null_pointer, which may later crash in contexts where maybe_null == false.
------------------------------------------------------------
revno: 8162
committer: Andrei Elkin <andrei.elkin@oracle.com>
branch nick: trunk-fixes
timestamp: Fri 2014-06-06 14:43:55 +0300
message:
  Bug #11752299 REPLICATION SLAVE TRUNCATES MASTER_PASSWORD > 32 CHARACTERS
  
  The master_user password in Change-master can't exceed 32 chars. Attempt to submit a larger password
  is silently accepted only to cause a later failure at START SLAVE.
  
  Fixed with erroring out at once when Change-master password string size is found to be over 32 chars
  in the parser.
  
  As a side effect, the patch makes rpl_change_master to run by mtr only
  with MIXED format.
------------------------------------------------------------
revno: 8161
committer: Shivji Jha <shivji.jha@oracle.com>
branch nick: reset-slave
timestamp: Fri 2014-06-06 16:57:48 +0530
message:
  BUG#18791604 SERVER HITS ER 1200 ON START SLAVE CMD, STILL HEARTBEAT_PERIOD SETS TO DEFAULT,
  BUG#18777899 HEARTBEAT_PERIOD AUTOMATICALLY SETS TO DEFAULT POST RESET SLAVE COMMAND,
  BUG#18778485 SSL_VERIFY_SERVER_CERT AUTOMATICALLY SETS TO NO POST RESET SLAVE COMMAND)
  
  Problem:
  =======
  
  There are three problems we address in this patch.
  The cause of these is more or less the same problem.
  But this exhibits different symptoms as described in
  the three bugs listed above, namely:
  1) reset slave automatically sets heartbeat_period to
     default.
  2) reset slave automatically sets ssl_verify_server_cert
     to default.
  3) When a server is not configured as a slave
     (no CHANGE MASTER done yet), START SLAVE will fail
     and that is fine. But this resets the heartbeat_period
     to default which is strange and un-intuitive.
  
  Analysis:
  ========
  
  The function init_master_log_pos() reset heartbeat_period
  to default and ssl_verify_server_cert= 0. This was called
  from reset slave as:
   reset_slave() => clear_in_memory_info() => init_master_log_info
  
  Fix:
  ===
  
  1) relocated code to reset heartbeat period to default to
     change_receive_options() and
     removed ssl_verify_server_cert= 0 in init_master_log_pos().
     init_master_log_pos() is a different thing altogether,
     it shouldn't care about heartbeat or ssl certificate at
     all.
  2) As a side-effect of (1), on a CHANGE MASTER if host and
     port are given but heartbeat_period is not specified,
     it remains 0 which means heartbeats are disabled.
     While adding/switching master, we want to keep heartbeat
     enabled by default. So in such a situation we force
     heartbeat_period to default heartbeat_count= 0 and
     last_heartbeat_timestamp= 0 unless the user specifically
     chooses to disable heartbeats.
------------------------------------------------------------
revno: 8160
committer: Thayumanavar <thayumanavar.x.sachithanantha@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 15:16:31 +0530
message:
  BUG#16664462 - MYSQL_TZINFO_TO_SQL CREATES BAD DATA
  Problem Description And Fix:
  mysql_tzinfo_to_sql utility creates bad data. In tz_load
  function, the string length for abbreviation list (read
  from timezone information file) doesn't account for the 
  the null character that is appended to the end of the 
  char array. This possibly can result in a buffer overflow
  as indicated by the valgrind warning (shown on bug
  http://bugs.mysql.com/bug.php?id=20545) based on type of
  data.
------------------------------------------------------------
revno: 8159
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 17:11:38 +0800
message:
  Bug#18706554 ASSERT BLOCK->PAGE.BUF_FIX_COUNT > 0 IN ROLLBACK,
               MISSING INDEX RECORD MESSAGE
  
  Follow-up patch for fixing test case i_innodb.innodb_bug18706554
  failure.
------------------------------------------------------------
revno: 8158 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 14:44:10 +0800
message:
  Null merge from mysql-5.6.
    ------------------------------------------------------------
    revno: 3902.324.96
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-06 14:41:46 +0800
    message:
      Fixed the failure on PB2 Windows, it should be due to the different
      error number returned from file operation functions.
            
      The same patch for trunk is already approved by Jimmy over IM.
------------------------------------------------------------
revno: 8157
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 13:46:53 +0800
message:
  Bug#18835117 RPL.RPL_SEMI_SYNC_INSTALL_AT_START_SERVER IS FAILING ON WINDOWS
  
  It is fixed by BUG#18466390.
------------------------------------------------------------
revno: 8156 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 11:01:17 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.95
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-06 11:00:16 +0530
    message:
      Fixing a PB2 failure.  The previous push invalidates the intention of the
      test case.  To avoid changing the intention of the test case, instead of
      allowing the update statement to fail, we increase the size of the redo
      log file size.  Increased the redo log file size to 160M.
      
      approved by Yasufumi over IM.
------------------------------------------------------------
revno: 8155 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 10:11:55 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.94
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-06-06 10:10:06 +0530
    message:
      Fixing a pb2 issue.  This is a test only patch. 
------------------------------------------------------------
revno: 8154
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 11:28:36 +0800
message:
  Fixed bug#18706554 ASSERT BLOCK->PAGE.BUF_FIX_COUNT > 0 IN ROLLBACK,
  MISSING INDEX RECORD MESSAGE
  
  When table format is Barrcuda, we will log prefix data for externally
  stored field.
  But for spatial index, these prefix data will generate a wrong mbr in
  rollback. And this cause it can't find the matched spatial index rec to
  undelete-marked.
  The solution is skip these prefix data, just use external field
  reference to build the right mbr.
  
  rb://5527
  Approved by Jimmy.
------------------------------------------------------------
revno: 8153
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-06-06 09:30:54 +0800
message:
  WL6559    Optimize GTIDs for passive slave - store GTIDs in table
  
  Post fix failed test.
------------------------------------------------------------
revno: 8152
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-06-05 13:21:58 -0500
message:
  Bug #18684389-SPAM ERROR 'FILE NOT FOUND' MSG WHEN 'DATA DIRECTORY'
  CLAUSE IN 'CREATE TABLE'
  
  The patch for WL7142 caused this regression in which an error message
  is written to the log at startup for every remote tablespace that was
  created with DATA DIRECTORY.  A 'strict' parameter was added to
  Datafile::open_read_only(strict) in order to determine if the message
  would be logged.  The function fil_open_single_table_tablespace() tries
  to find the file in three possible locations and it always looks at the
  default location.  If the file is not there, this strict parameter causes
  the spam message.
  
  But by the time the default-location filepath is attempted,
  fil_open_single_table_tablespace() has already tried the other two
  locations.  So the fix is to print the message only if the file is not
  found elsewhere.
  
  Approved by marko in RB#5536
------------------------------------------------------------
revno: 8151
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Thu 2014-06-05 17:10:11 +0200
message:
  Bug#18904069 GET RID OF DYNAMIC_ARRAY IN LEAST_OCCUPIED_WORKERS
  
  Post-push fix: link error on windows.
------------------------------------------------------------
revno: 8150
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Thu 2014-06-05 15:21:07 +0200
message:
  Bug#18904069 GET RID OF DYNAMIC_ARRAY IN LEAST_OCCUPIED_WORKERS
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 8149
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Thu 2014-06-05 13:04:26 +0200
message:
  Bug#18855155 GET RID OF RAW ARRAY IN MDL_CONTEXT::ACQUIRE_LOCKS
  
    Our array abstractions should be removed, and substituted with modern data
    structures.
      
    Templatized vectors are:
     - type safe
     - easier to read/maintain
     - faster
  
  Using Prealloced_array, we also avoid a malloc/free of the buffer in most cases.
------------------------------------------------------------
revno: 8148 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Thu 2014-06-05 18:07:47 +0530
message:
  Bug#18000079 - BINLOG_DUMP_NON_BLOCK DISAPPEARED IN MYSQL 5.6
  		(null merge)
    ------------------------------------------------------------
    revno: 3902.324.93
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-06-05 18:06:08 +0530
    message:
      Bug#18000079 - BINLOG_DUMP_NON_BLOCK DISAPPEARED IN MYSQL 5.6
      
      	--Mysqlbinglog can't able to check whether it is a 
      	  debug build or not with have_debug.inc. So used 
      	  mysqlbinlog_have_debug.inc.			
------------------------------------------------------------
revno: 8147 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Thu 2014-06-05 18:01:55 +0530
message:
  Bug#18000079 - BINLOG_DUMP_NON_BLOCK DISAPPEARED IN MYSQL 5.6	
  		(null merge)
    ------------------------------------------------------------
    revno: 3902.330.1
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-06-05 18:00:30 +0530
    message:
      Bug#18000079 - BINLOG_DUMP_NON_BLOCK DISAPPEARED IN MYSQL 5.6
      
      	--Mysqlbinglog can't able to check whether it is a 
      	  debug build or not with have_debug.inc. So used 
      	  mysqlbinlog_have_debug.inc.			
------------------------------------------------------------
revno: 8146
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-05 13:07:07 +0300
message:
  Bug#18911060 FIL_VALIDATE() IS NOT VALIDATING THE SIZE
  Bug#18419443 RPL.RPL_DROP_DB RUN INTO ASSERTION: COUNT == LIST.COUNT
  
  The validation function fil_validate() is not computing the sizes of the
  files belonging to the tablespace (space->chain), and comparing that to the
  aggregate value space->size.
  
  UT_LIST_VALIDATE: Remove.
  
  ut_list_validate(), ut_list_map(): Take Functor& instead of Functor.
  
  ut_list_validate(): Do not use the functor for visiting the nodes when
  traversing the reverse links. If the functor is computing some sums,
  we only want to visit each element once.
  
  CheckInLRUList::validate(), CheckInFreeList::validate(),
  CheckUnzipLRUAndLRUList::validate(): New static methods.
  
  buf_buddy_list_validate(): A function to replace BUF_BUDDY_LIST_VALIDATE().
  
  fil_node_create(): Protect the fil_space_t change with fil_system->mutex.
  This is what seemed to trigger the assertion failure in fil_validate().
  
  rb#5612 approved by Jimmy Yang
------------------------------------------------------------
revno: 8145
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-05 12:18:11 +0300
message:
  Bug#18742393 DUPLICATE CALL TO MTR_T::SET_NAMED_SPACE() IN
  BTR_CREATE() ERROR HANDLING
  
  The error handling path of btr_create() is calling
  btr_free_root(). btr_free_root() is invoking mtr_t::set_named_space(),
  even though it has already been invoked earlier during the
  mini-transaction.
  
  btr_free_root(): Replace the call to mtr_t::set_named_space() with a
  debug assertion on mtr_t::is_named_space().
  
  Make the callers of btr_free_root() responsible for calling
  mtr_t::set_named_space().
  
  rb#5605 approved by Kevin Lewis, Bin Su
------------------------------------------------------------
revno: 8144
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: recperkey
timestamp: Thu 2014-06-05 10:17:30 +0200
message:
  Bug#18497308 WRONG COST ESTIMATE FOR LOOSE INDEX SCAN WHEN
               INDEX STATISTICS IS MISSING
        
  Follow-up patch to remove one compiler warning that
  the Windows compiler complains about when using a "double" 
  constant in an expression that is assigned to 
  a rec_per_key_t variable (which is float).
------------------------------------------------------------
revno: 8143
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Thu 2014-06-05 08:43:07 +0200
message:
  Bug#18767104 MALLOC OF UNINITIALIZED MEMORY SIZE, CRASH WITH BLOB + TMPTABLE
  
  Post-push fix: skip in embedded mode.
------------------------------------------------------------
revno: 8142 [merge]
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-06-05 10:00:25 +0800
message:
  WL6559    Optimize GTIDs for passive slave - store GTIDs in table
  
  If gtid_mode is enabled, store GTIDs in table always as below:
    - If binlog is disabled, store transaction owned GTID in table
      within each transaction.
    - If binlog is enabled and log_slave_updates is enabled, store
      set of GTIDs of the last binlog in table on each binlog
      rotation and on server shutdown.
    - If binlog is enabled and log_slave_updates is disabled, do below:
        - slave SQL thread or slave worker stores transaction owned GTID
          in table within each transaction.
        - store set of GTIDs of the last binlog in table on each binlog
          rotation and on server shutdown.
    ------------------------------------------------------------
    revno: 7095.1.85 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-06-05 09:20:20 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
            
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.84 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-06-03 19:00:15 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
          
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.83
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-05-27 17:40:22 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
          
      Commit current transaction to save GTIDs into 'gtid_executed' table
      on binlog rotation and server shutdown.
    ------------------------------------------------------------
    revno: 7095.1.82 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-05-20 17:59:50 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Manually merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.81
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-05-11 12:22:27 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix failed rpl_change_repository_stress.test after merge from mysql-trunk.
      Fix compile error: 'new_name_ptr' may be used uninitialized in this function.
    ------------------------------------------------------------
    revno: 7095.1.80 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-05-09 12:50:34 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.79
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-05-08 23:22:26 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Update affected test result on embedded platform
    ------------------------------------------------------------
    revno: 7095.1.78
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-05-08 11:52:42 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Using exact table nmae 'gtid_executed' in comments.
      Update colume name and comments for gtid_executed table.
      Optimize code and remove unnecessary changes.
      Fix affected tests.
      Resolve compile error on some platforms.
      Fix random failure of binlog_simulate_crash_and_error_on_compress_gtid_table.test on windows.
    ------------------------------------------------------------
    revno: 7095.1.77
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-05-05 12:06:42 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix failed test 'rpl_innodb_disabled.test' on windows platform.
    ------------------------------------------------------------
    revno: 7095.1.76
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-04-30 17:11:28 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      1. If binlog is enabled and log_slave_updates is disabled, slave SQL thread or slave worker thread saves transaction owned gtid into table before transaction prepare.
      2. Slave SQL thread or slave worker thread saves gtid into table for a DDL statement if binlog is enabled and log_slave_updates is disabled.
      3. Slave SQL thread or slave worker thread adds transaction owned gtid into global executed_gtids if binlog is enabled and log_slave_updates is disabled.
      4. Slave SQL thread or slave worker thread adds transaction owned gtid into global gtids_only_in_table if binlog is enabled and log_slave_updates is disabled.
      5. If binlog is enabled and log_slave_updates is disabled, slave SQL thread or slave worker thread does not log a empty gtid group.
      6. Write new 'rpl_gtids_table_disable_log_slave_updates.test' to cover above changes.
    ------------------------------------------------------------
    revno: 7095.1.75
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-04-24 10:57:20 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Update test result.
    ------------------------------------------------------------
    revno: 7095.1.74
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-04-23 10:16:13 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Update copyright.
    ------------------------------------------------------------
    revno: 7095.1.73
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-04-23 09:59:14 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      when binlog is enabled, notify compression thread to compress gtid
      table on binlog rotation.
      When binlog is disabled, compression thread reads each row by the
      PK(sid, gno_start) in increasing order, and compresses the first
      consecutive range of gtids in a single transaction, end the
      compression if the last transaction reached the end of the table,
      otherwise continue to compress the table by employing one more
      transactions until the gtid table is compressed completely.
      Optimize code and update test cases.
    ------------------------------------------------------------
    revno: 7095.1.72
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-04-21 16:18:16 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Save the set of GTIDs of the last binlog into table on server shutdown.
      Optimize code and rollback changes.
      Added test cases to cover error branches and update test result.
    ------------------------------------------------------------
    revno: 7095.1.71 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-04-18 17:20:15 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Manually merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.70
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-04-18 15:30:09 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Add code to save GTIDS into gtid table on binlog rotation, and remove
      related code of saving GTID into gtid table within each transaction
      when binlog is enabled. Write test cases for current approach, rollback
      the unnecessary changes of tests with previous approach.
    ------------------------------------------------------------
    revno: 7095.1.69
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-02-27 10:37:58 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix the test by restarting and resetting the server after simulate error on delete gtid from table.
      Fix the test by restarting slave server to get ride of 'simulate_create_compress_thread_failure'.
      Update comment and reset 'SERVER_STATUS_IN_TRANS' flag conditionally.
    ------------------------------------------------------------
    revno: 7095.1.68 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-25 09:23:56 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7095.1.67
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-25 08:29:50 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Write a XA transaction rollback test case to cover error branch.
      Write a test case to cover error branch of generating automatic gtid.
      Write a test case to cover error branch of creating compression thread
    ------------------------------------------------------------
    revno: 7095.1.66 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-21 12:27:50 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.65
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-21 11:12:46 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Write test case to cover error branches of resetting gtid table.
      Write a transaction rollback test case when binlog is disabled and gtid_mode is on.
      Remove redundant sync_slave_sql_with_master.inc before rpl_end.inc.
    ------------------------------------------------------------
    revno: 7095.1.64
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-02-19 13:56:55 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.63
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-02-19 12:49:58 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Updated test cases.
    ------------------------------------------------------------
    revno: 7095.1.62
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-02-19 09:59:01 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Consolidate code to handle the spurious wakeups for
      mysql_cond_wait(&COND_compress_gtid_table, ...).
    ------------------------------------------------------------
    revno: 7095.1.61
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-18 19:15:25 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize to terminate compression thread.
    ------------------------------------------------------------
    revno: 7095.1.60
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-18 16:27:37 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize to compress the first consecutive range of gtids.
    ------------------------------------------------------------
    revno: 7095.1.59
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-18 14:18:48 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.58
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-18 11:47:15 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize conditions of saving gtid into table when binlog is off.
    ------------------------------------------------------------
    revno: 7095.1.57
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-18 10:11:42 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      When binlog is off, report @@GLOBAL.GTID_PURGED from executed_gtids,
      since @@GLOBAL.GTID_PURGED and @@GLOBAL.GTID_EXECUTED are always
      same, so we did not save gtid into lost_gtids for every transaction
      for improving performance.
      Optimize conditions of saving gtid into table when binlog is off.
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.56
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-02-17 11:57:48 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Revert code change for purging binlog files.
    ------------------------------------------------------------
    revno: 7095.1.55
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-02-17 10:49:14 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Don't save gtid into table for the following case, if the transaction
      is being rolled back and contains changes that cannot be rolled back,
      the trx-cache's content is flushed.
    ------------------------------------------------------------
    revno: 7095.1.54
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-02-17 09:20:22 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Write test cases to cover error branches when saving gtid into table
      and compressing gtid table.
    ------------------------------------------------------------
    revno: 7095.1.53
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-14 14:12:07 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Add comments for compression thread functions.
    ------------------------------------------------------------
    revno: 7095.1.52
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-02-13 09:47:37 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix failed 'rpl.rpl_blackhole' after merge.
    ------------------------------------------------------------
    revno: 7095.1.51
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-02-12 09:46:36 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Saving thd->variables.sql_mode is not necessary, remove it.
    ------------------------------------------------------------
    revno: 7095.1.50
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-11 21:25:44 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Fix failed 'rpl.rpl_slow_query_log' after merge
    ------------------------------------------------------------
    revno: 7095.1.49
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-11 11:05:11 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Fix failed 'mysqld--help-win' on windows.
    ------------------------------------------------------------
    revno: 7095.1.48
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-02-11 10:37:51 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix failed 'i_rpl.rpl_commit_errchk' after merge
    ------------------------------------------------------------
    revno: 7095.1.47
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-02-10 18:20:37 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Remove options '--enforce-gtid-consistency --log-slave-updates --gtid-mode=ON'
      and replace '--sync_slave_with_master' with 'sync_slave_sql_with_master.inc'
    ------------------------------------------------------------
    revno: 7095.1.46
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-02-10 12:05:28 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Revert the change to write_one_empty_group_to_cache(...) and remove an invokation
      to write_empty_groups_to_cache(...) before transaction commit phase.
    ------------------------------------------------------------
    revno: 7095.1.45
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-02-09 18:54:52 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Remove copyright for public test case.
    ------------------------------------------------------------
    revno: 7095.1.44
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-02-09 18:51:21 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Remove copyright for public test cases.
    ------------------------------------------------------------
    revno: 7095.1.43
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-02-09 17:37:02 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Add a new table category for GTID table.
    ------------------------------------------------------------
    revno: 7095.1.42
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-02-09 15:59:56 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.41
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-02-09 10:11:51 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Add transaction's gtid into executed_gtids at binlog flush time when
      binlog is enabled and at right after commit time when binlog is
      disabled.
    ------------------------------------------------------------
    revno: 7095.1.40 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sat 2014-02-08 21:23:52 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
          
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.39
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sat 2014-02-08 20:52:49 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.38 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-07 18:04:59 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.37 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-07 17:30:33 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.36
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-07 16:47:35 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.35
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-02-07 13:47:32 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Add Gtid_table_access_context class to simplify code.
    ------------------------------------------------------------
    revno: 7095.1.34
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-02-06 21:46:49 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
        
      Fix compression problem that the field gno_end contains the
      value 0 after compression of a range of gtids.
    ------------------------------------------------------------
    revno: 7095.1.33
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-02-06 21:35:12 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix failed test case 'rpl_group_commit_deadlock'.
    ------------------------------------------------------------
    revno: 7095.1.32
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-01-28 16:32:55 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.31 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-01-28 14:22:08 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.30
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-01-28 13:21:20 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Hide the public interfaces of Gtid_table_persistor into Gtid_state.
    ------------------------------------------------------------
    revno: 7095.1.29
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-27 17:02:38 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Define a separate function for the debug code for compress().
      Rename function name 'fetch_gtids' with fetch_gtids.
      Do not disable binlog when fetch gtids from gtid table.
    ------------------------------------------------------------
    revno: 7095.1.28
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-27 13:27:55 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Rename function name and optimize comment.
    ------------------------------------------------------------
    revno: 7095.1.27
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-27 11:36:43 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Save the logged_gtids_binlog set into gtid table in upgrading case.
      Optimize code to calculate gtids_only_in_table first and then
      calculate lost_gtids.
    ------------------------------------------------------------
    revno: 7095.1.26
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-27 11:06:42 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      If gtid_mode is enabled and binlog is disabled, initialize executed_gtids
      from gtid table and lost_gtids has a same gtid set with executed_gtids
      during server startup.
    ------------------------------------------------------------
    revno: 7095.1.25 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-01-26 23:06:30 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.24
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-01-26 18:54:11 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Define compression thread related functions in rpl_gtid_persist.cc and optimize code.
    ------------------------------------------------------------
    revno: 7095.1.23
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Sun 2014-01-26 11:39:58 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Resolve InnoDB possible native deadlock by compressing first consecutive rows
      range (delete consecutive rows from the second row, then update the first row)
      in one transaction. Sure, we must lock the cluster index from left to right,
      and do not use ha_index_prev(...) when deleting and updating rows, and
      compression thread does not skip a gap to compress the next consecutive
      rows in one transaction.
    ------------------------------------------------------------
    revno: 7095.1.22 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-20 13:13:41 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.21
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-20 10:57:40 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Resolve InnoDB possible native deadlock caused by complex transaction
      while compressing the gtid table by executing delete operations and
      insert operations in separate statement in one transaction.
    ------------------------------------------------------------
    revno: 7095.1.20
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-01-17 19:06:22 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Resolve possible deadlock between 'RESET MASTER' thread and transactions' threads.
      The deadlock is dug out by sysbench performance test.
      Add lock to prevent from compressing gtid table while reseting the gtid table.
    ------------------------------------------------------------
    revno: 7095.1.19
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-01-09 16:46:23 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Fix failed test cases on embedded platform.
    ------------------------------------------------------------
    revno: 7095.1.18
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-01-09 12:01:14 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Clean code.
    ------------------------------------------------------------
    revno: 7095.1.17 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-01-09 11:22:55 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk.
    ------------------------------------------------------------
    revno: 7095.1.16
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-01-08 18:43:57 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Update Copyright.
    ------------------------------------------------------------
    revno: 7095.1.15
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-01-08 17:37:00 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code.
    ------------------------------------------------------------
    revno: 7095.1.14
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2014-01-08 13:09:11 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      1. Optimize update_on_rollback(...), update_on_commit(...) and gtid_rollback(...) functions and their invokers.
      2. Fix build error on windows.
      3. Remove unused COND_terminate_compress_thread.
    ------------------------------------------------------------
    revno: 7095.1.13
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-01-07 14:58:22 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Terminate compression thread by invoking pthread_join(...).
      MySQL main thread will wait until timeout on condition variable
      'COND_terminate_compress_thread' if it missed signal from
      compression thread, which will cause timeout of some test cases
      after hundreds repeat.
    ------------------------------------------------------------
    revno: 7095.1.12 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2014-01-07 12:14:36 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7095.1.11
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-06 12:54:29 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Optimize code to set m_is_resetting after acquire LOCK_reset_binlog.
    ------------------------------------------------------------
    revno: 7095.1.10
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2014-01-06 09:48:39 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      1. Remove the command 'COM_COMPRESS_GTID_TABLE' and command name 'Compress' for compression thread.
         Since it is good enough to give a good thread name 'compress_gtid_table'.
      2. Move generate_and_save_gtid(...) into Gtid_state class.
    ------------------------------------------------------------
    revno: 7095.1.9
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2014-01-03 15:51:46 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      1. Remove 'thd->is_operating_gtid_table' check by setting MYSQL_OPEN_IGNORE_KILLED when opening gtid table.
      2. Remove unnecessary 'gtid_table_persistor != NULL' check.
      3. Optimize code.
       
    ------------------------------------------------------------
    revno: 7095.1.8
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Thu 2014-01-02 17:11:37 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Resolve deadlock and crash caused by introduing gtid counter and
      LOCK_reset_binlog lock.
    ------------------------------------------------------------
    revno: 7095.1.7
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2013-12-31 18:42:13 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Resolve the problem between 'RESET MASTER' thread and 'TRANSACTION' theads,
      which will cause inconsistency between GLOBAL.GTID_EXECUTED and gtid table
      after resetting the master. And optimize code.
    ------------------------------------------------------------
    revno: 7095.1.6
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2013-12-31 11:22:26 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Revert the change by makeing thd->in_active_multi_stmt_transaction()
      be false after commit the transaction with saving gtid into table
      when binlog is disabled.
    ------------------------------------------------------------
    revno: 7095.1.5
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Fri 2013-12-27 10:17:59 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      1. Correct the behavior of the transaction commit when saving gtid into the gtid table. We just need to commit the transaction when compressing the gtid table, reseting the gtid table, and fetching gtids from the gtid table.
      2. Fix bugs caused by XA transaction.
      3. Remove an assertion as we generate transaction's gtid and save it into table on the top of XA_END. So the thd->owned_gtid.sidno is assigned in XA_PREPARE and XA_COMMIT phase.
      4. Set default value of system variable 'executed_gtids_compression_period' to 1000, as we got a good performance with the value in performance test.
      5. Add test cases to verify that XA transactions' gtids are stored into gtid table and can be reported from global.gtid_executed correctly.
    ------------------------------------------------------------
    revno: 7095.1.4
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Tue 2013-12-24 18:49:09 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Update sizing_default.result
    ------------------------------------------------------------
    revno: 7095.1.3 [merge]
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2013-12-23 13:32:09 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Merge mysql-trunk to mysql-trunk-wl6559
    ------------------------------------------------------------
    revno: 7095.1.2
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Mon 2013-12-23 11:47:50 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      Replace Gtid_Table_Persistor with Gtid_table_persistor in DBUG.
    ------------------------------------------------------------
    revno: 7095.1.1
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-trunk-wl6559
    timestamp: Wed 2013-12-18 14:36:19 +0800
    message:
      WL6559    Optimize GTIDs for passive slave - store GTIDs in table
      
      
      1. Create a mysql system table named gtid_executed, a row in the table can
      contain an interval of several GTIDs, all having the same SID and consecutive
      GNOs as following:
      
        CREATE TABLE gtid_executed(
          sid CHAR(36) NOT NULL,       -- Source ID
          gno_start BIGINT NOT NULL,   -- First GNO of interval
          gno_end BIGINT NOT NULL,     -- Last GNO of interval
          PRIMARY KEY(sid, gno_start)  -- PK on interval start
        ) ENGINE = InnoDB;
      
      
      2. Store gtids into the gtid_executed table always if gtid_mode is enabled as
      following:
        INSERT INTO gtid_executed VALUES (SID, GNO, GNO)
        - If binlog is disabled and gtid_mode is enabled, store gtids from master's
      transactions, relayed BINLOG transactions and 'SET @@SESSION.GTID_NEXT'
      statement into gtid_executed table right before transaction prepare.
        - If binlog is enabled and gtid_mode is enabled, store GTIDs into both binlog
      and gtid_executed table.
      
      
      3. The gtid_executed table would be filled with singleton interval as we cannot
      use UPDATE instead of INSERT, because any concurrently executing transactions
      would conflict on the update. So we propose to create a FOREGROUND thread (we
      can not create a BACKGROUND thread as we have to create THD object to open
      gtid_executed table) and introduce a user variable
      'executed_gtids_compression_period' to control how often the thread should be
      woken up to compress the gtid_executed table regularly.
        - To the user variable 'executed_gtids_compression_period', if value is 0,
      the thread never wakes up, if value is 1, wakes up every transaction. If value
      is 1000, wakes up every 1000 transactions.
        - We also added the thread's state info (suspending, compressing gtid_executed
      table) into performance_schema.threads table, so that user can monitor it and
      tune the user variable 'executed_gtids_compression_period' (The thread is in
      suspended mode most of the time, wakes up only when it needs).
        - THe thread runs always, user cannot kill it as we didn't add it into
      PROCESSLIST.
        - The thread opens a transactional context to execute the following within a
      single transaction on an non-autocommit mode:
          - Read each row by the PK in increasing order, delete consecutive rows from
      the gtid_executed table and fetch these deleted gtids at the same time.
          - Store compressed intervals from these deleted gtids into the gtid_executed
      table.
      
      
      4. Report @@GLOBAL.GTID_EXECUTED and @@GLOBAL.GTID_PURGED from both binlog and
      gtid_executed table as following:
        - GLOBAL.GTID_EXECUTED = all gtids of gtid_executed table;
        - GLOBAL.GTID_PURGED = GLOBAL.GTID_EXECUTED - (GTID_EXECUTED_BINLOG -
      GTID_PURGED_BINLOG);
        (GTID_EXECUTED_BINLOG: gtid_set of gtids is ever logged in binary logs.
         GTID_PURGED_BINLOG  : gtid_set of gtids is purged from binary logs.)
      
        - GLOBAL.GTID_EXECUTED is initialized with all gtids of gtid_executed table
      during server restarting. Every transaction's gtid is added into the
      GLOBAL.GTID_EXECUTED and gtid_executed table if gtid_mode is enabled after
      server restarts.
        - GLOBAL.GTID_PURGED is initialized with gtids from gtid_executed table and
      binlogs and precomputed as above during server restarting. After server
      restarts, added purged gtids into GLOBAL.GTID_PURGED when purging logs
      (regardless binlog is enabled or not), and added every transaction's gtid into
      GLOBAL.GTID_PURGED if binlog is disabled and gtid_mode is enabled.
        - A previous log event with GTID_EXECUTED_BINLOG (GTID_EXECUTED -
      GTID_ONLY_IN_TABLE) set is created when rotating binlog, so GTID_ONLY_IN_TABLE
      variable is maintained when binlog is enabled. GTID_ONLY_IN_TABLE is initialized
      with difference gtids from gtid_executed table and the latest binlog during
      server restarting. GTID_ONLY_IN_TABLE will never change since starting server.
      (we don't need maintain GTID_ONLY_IN_TABLE when binlog is disabled)
        - If gtid_executed_table is empty, add all gtids in GTID_EXECUTED_BINLOG into
      gtid_executed table during server restarting (Handle the upgrade case, and the
      case that a slave is provisioned from a backup of the master and the slave is
      cleaned by RESET MASTER and RESET SLAVE before this.).
        - The gtid_executed table is also initialized with gtid_purged when user is
      initializing it through SET GLOBAL gtid_purged. (It is possible to update the
      value of gtid_purged, but only by adding GTIDs to those already listed, and only
      when gtid_executed is unset?that is, on a new server.)
      
      
      5. The gtid_executed table is reset when resetting master.
      
      
      6. Gtid_table_persistor class
         It manages all operations on the gtid table.
      
         - m_count
           Count the append rows of the table.
      
         - save(Gtid *gtid)
           Insert the gtid into table.
      
         - save(Gtid_set *gtid_set)
           Store gtid set into the table.
      
         - compress()
           Compress intervals into consecutive GNOs in the table.
      
         - reset()
           Delete all rows from the table.
      
         - fetch_gtids_from_table(Gtid_set *gtid_set)
           Fetch gtids from the table and store them into gtid set.
------------------------------------------------------------
revno: 8141
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-06-04 15:14:17 -0500
message:
  Bug#17713871-ASSERT ERROR == DB_SUCCESS, COMMIT_CACHE_REBUILD(),
  ALTER TABLE, PRBLEM RENAMING
  
  In order to add even more uniqueness to the temporary filename, this patch replaces the LSN in the filename with a more unique another number.  It is just a static global number that is initialized to a random distributed 32-bit number using ut_time() and ut_crc32().  It is then incremented atomically for each temporary file name assigned.  This should not cause performance regression since an atomic_increment is nothing compared to a OS file rename. 
  
  Approved by Marco in RB#5541
------------------------------------------------------------
revno: 8140
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18767104blob
timestamp: Wed 2014-06-04 16:32:41 +0200
message:
  Bug#18767104 MALLOC OF UNINITIALIZED MEMORY SIZE, CRASH WITH BLOB + TMPTABLE
  
  In Item_func_sha2::fix_length_and_dec() :
  Do fix_length_and_charset even if input argument to function is wrong.
------------------------------------------------------------
revno: 8139
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: recperkey
timestamp: Wed 2014-06-04 15:51:58 +0200
message:
  Bug#18497308 WRONG COST ESTIMATE FOR LOOSE INDEX SCAN WHEN
               INDEX STATISTICS IS MISSING
  
  Follow-up patch to remove two compiler warnings that
  gcc 4.1.2 complained about when converting to uint from
  rec_per_key_t (which is a float type).
------------------------------------------------------------
revno: 8138
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18840123decimal
timestamp: Wed 2014-06-04 14:07:15 +0200
message:
  Bug#18840123 CRASH IN DECIMAL_MUL WITH STRICT MODE AND EXISTING INVALID DATE
  
  In Item::val_decimal_from_date()
  Do not return a NULL pointer in case of conversion failure,
  the result may be needed in subsequent arithmetic operations.
------------------------------------------------------------
revno: 8137
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Wed 2014-06-04 13:27:08 +0200
message:
  Bug#18759597 MISSING ROWS ON WHERE .. IN QUERY WITH VARIABLES AND CONCAT
  
  cmp_item_sort_string_in_static::store_value() assumed that its input
  string could not change. It turns out it can.
  
  In this case we stored 'm' 'n' 'a' 'l' 'xm' in column one of each row.
  We sort the data to get 'a' 'l' 'm' 'n' 'xm'
  
  Then the expression is analyzed again, and Item_func_concat::tmp_value
  changes value. So when we do lookup, we search in 'a' 'l' 'm' 'n' 'm'
  
  The binary-search failed, since the array of rows is no longer
  properly sorted.
  
  Fix: Save string results locally in 
  cmp_item_sort_string_in_static::store_value()
  This is done only once during the prepare phase, so it should not have
  any negative performance impact.
  
  After doing that change, we see that the classes
  cmp_item_sort_string and cmp_item_sort_string_in_static are nearly identical.
  We can remove them both, and move all member functions into cmp_item_string instead.
  
  Also: remove some C-style casts, introduce new template function down_cast instead.
  
  Where/why is it analyzed again?
  The range optimizer does some analysis and ends up calling Item::save_in_field_no_warnings
  which ends up calling Item_func_concat::val_str() again.
  Looking at the source code of that function, we see there are many
  ways that tmp_value can change.
------------------------------------------------------------
revno: 8136
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-06-04 14:54:09 +0800
message:
  Fixed the failure on PB2 Windows, it should be due to the different
  error number returned from file operation functions.
  
  Approved by Jimmy over IM.
------------------------------------------------------------
revno: 8135
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-06-04 08:16:09 +0200
message:
  1. Added big-test flag to tests that take more than a minute on the fastest platform OEL6 on PB2. This will help reduce time taken for per push runs
  2. Disabled rpl.rpl_stm_mixed_mts_rec_crash_safe_small. This is supposed to be shorter version of .rpl_stm_mixed_mts_rec_crash_safe suitable for PerPush runs. This is taking about 2 minutes and hence needs to be tuned down
  3. Moved rpl.rpl_optimize from experimental to disabled list. It is timing out on windows leading to very long running time for windows in daily.
------------------------------------------------------------
revno: 8134 [merge]
committer: Luis Soares <luis.soares@oracle.com>
branch nick: mysql-trunk-latest
timestamp: Tue 2014-06-03 14:58:41 +0100
message:
  WL#7742
  
  Automerged local mysql-trunk into latest mysql-trunk.
    ------------------------------------------------------------
    revno: 8127.1.1
    committer: Luis Soares <luis.soares@oracle.com>
    branch nick: mysql-trunk
    timestamp: Mon 2014-06-02 16:13:36 +0100
    message:
      WL#7742: BGC: waiting more transactions entering bgc queues
                      
      Adds two new options to introduce an artificial delay to make the
      binary log group commit procedure wait. This gives a chance that
      more transactions are flushed and synced together to disk, thus
      reducing the overall time spent to commit a group of transactions
      (the bigger the groups the less number of sync operations).
                      
      These options are named: binlog-group-commit-sync-delay and 
      binlog-group-commit-sync-no-delay-count. The former takes as 
      input the number of microseconds to wait. The latter, takes as 
      input the number of  transactions that the server waits for, 
      before deciding to abort the waiting.
------------------------------------------------------------
revno: 8133
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-2
timestamp: Tue 2014-06-03 12:53:05 +0200
message:
  Merge of the fix for Bug#18791851 CRASH IN ST_SELECT_LEX::PRINT WITH OPTIMIZER_TRACE ON SUBQUERY.
  Because of wl7082 the same bug cannot happen (we still add the testcase), but this variant can crash:
        - query is 
        SET @a:=(SELECT ROW(1, 2)=
        ROW((SELECT 1 FROM t1 LEFT JOIN t1 t2 ON 1
                       HAVING 3 IN (SELECT 2 FROM DUAL)),
            1));
        (i.e. the rare conditions of fixed bug18345786 CRASH AROUND ST_JOIN_TABLE::AND_WITH_CONDITION).
        - simplify_joins() runs in JOIN::optimize(), when JOIN::where_cond/having_cond and TABLE_LIST::optim_join_cond()
        are still unset (==0x1).
        - it prints the transformed query; select_lex::print() sees JOIN::optimized==true,
        so uses unset members above. Crash.
        
  Fix: break the link between JOIN::optimized and conditions: in select_lex::print(), do
        cur_where = (join && join->where_cond != 0x1) ?
        // it is ready for reading, and the most up-to-date condition
                                 join->where_cond : 
                                select_where->where;
------------------------------------------------------------
revno: 8132
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk-2
timestamp: Tue 2014-05-20 15:31:45 +0200
message:
  Fix for Bug #18770217 	DIFFERENT RESULT WITH SP ON 2ND EXECUTION OF QUERY WITH STRAIGHT_JOIN IN NOT IN
  We have this subquery predicate:
  NOT IN (
  SELECT sq1_alias1.col_varchar_nokey
  FROM t3 AS sq1_alias1
  JOIN ( t1 AS sq1_alias2
  STRAIGHT_JOIN view_d AS sq1_alias3 )
  WHERE (
  SELECT col_int_key
  FROM t1 AS c_sq1_alias1
  WHERE c_sq1_alias1.col_varchar_key = sq1_alias3.col_varchar_nokey
  )
  )
  which is handled with subquery materialization.
  It contains a WHERE clause involving a scalar subquery which returns
  two rows. Thus, it is expected that when we fill the temporary table,
  we hit and report an error.
  However, in the second execution of a stored procedure it is not the
  case. A necessary ingredient is the view column of view_d, used as
  outer reference in the scalar subquery.
  First execution:
  - view_d is merged
  - the view's column is resolved by fix_outer_field():
            prev_subselect_item->used_tables_cache|=
              (*reference)->used_tables();     // 1
            prev_subselect_item->const_item_cache&=
              (*reference)->const_item();
            mark_as_dependent(thd, last_checked_context->select_lex, // 2
  
  *reference is Item_direct_view_ref. Its depended_from is NULL at
  (1). So its used_tables() returns the map of t2.col_varchar_nokey
  (==4). This goes into used_tables_cache of the scalar subquery item,
  which is correct - from the POV of the materialized subquery which
  owns the scalar subquery, this scalar subquery depends on view_d, the
  table of map 4 in the materialized subquery.
  At (2), depended_from becomes "select#2" i.e. the materialized
  subquery. It is correct: the item_direct_view_ref has the scalar
  subquery as context, so it is an outer reference.
  
  Second execution:
  - in find_field_in_tables() we go through code specific of any not-first
  execution:
    if (item->cached_table) <<<< this is true on non-first execution
    {
    ...
          /*
            If the field was an outer referencee, mark all selects using this
            sub query as dependent on the outer query
          */
          if (current_sel != last_select)
            mark_select_range_as_dependent(thd, last_select, current_sel,
                                           found, *ref, item);
  This sets depended_from of the item_direct_view_ref to "select#2"
  (correct).
  - then we go into the same fix_outer_field() as in the first
  execution, but note that this time we enter it with
  depended_from!=NULL, so:
            prev_subselect_item->used_tables_cache|=
              (*reference)->used_tables();     // 1
  adds OUTER_REF_TABLE_BIT to used_tables_cache of the scalar
  subquery. That is wrong.
  - then, this scalar subquery gets lost by make_join_select() because
  this function (as it is operating for a materialized subquery)
  excludes conditions using OUTER_REF_TABLE_BIT (see
  join->allow_outer_refs).
  - the condition being lost, it won't throw the expected "more than one
  row" error.
  
  Fix: use resolved_used_tables(). Indeed, prev_subselect_item is owned
  by the qualifying query of the item. A second similar line is changed too.
------------------------------------------------------------
revno: 8131
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: loosescan-bug2
timestamp: Tue 2014-06-03 14:12:37 +0200
message:
  Fix for Bug#18497308 WRONG COST ESTIMATE FOR LOOSE INDEX SCAN WHEN
                       INDEX STATISTICS IS MISSING 
  
  The cost for doing loose index scan is calculated in
  cost_group_min_max(). To estimate the size of the groups to read,
  index statistics from rec_per_key is used. The code works correctly
  when we have index statistics but if index statistics is not
  available, there is a bug in the code that in some cases will cause
  wrong cost estimates to be produced.
  
  For a query like this:
  
    SELECT DISTINCT a FROM t1 WHERE b=4;
  
  with an index on (a,b,c), cost_group_min_max() will estimate the "keys
  per group" both for the index prefix (a) and the index
  prefix(a,b). The "keys per group" the first of these, is estimated by
  looking up in rec_per_key[0] and using the following code to adjust
  the estimate if there is no rec_per_key information:
  
    /* Compute the number of keys in a group. */
    keys_per_group= index_info->rec_per_key[group_key_parts - 1];
    if (keys_per_group == 0) /* If there is no statistics try to guess */
      /* each group contains 10% of all records */
      keys_per_group= (uint)(table_records / 10) + 1;
  
  For estimating the second "keys per group", rec_per_key[1] is used and
  the code looks like this:
  
    keys_per_subgroup= index_info->rec_per_key[used_key_parts - 1];
  
  There is no code that handles the case when index statistics is
  missing. In this case keys_per_subgroup get the value 0. Using 0 as an
  estimate in the code that follows causes wrong cost estimates to be
  produced.
  
  To ensure that keys_per_subgroup always has a sensible value, this
  patch adds code that handles the case where index statistics is
  missing. A new function guess_rec_per_key() is added. This function
  is based on existing code in find_best_ref() that assumes for an
  index with multiple parts, the first key part will match 1% of the
  records and the whole key will match 10 records in the case of
  a non-unique key and 1 record in the case of a unique key. For
  other numbers of used key parts, a formula is used for computing
  an estimated rec_per_key that is between the rec_per_key value for
  the first key part and the rec_per_key value for the whole key.
  
  This new function is used in both cases in cost_group_min_max().
  Replacing the current handling of missing index statistics causes
  changes to explain output in multiple test cases. The main reason
  for these changes are that the current code assumes that the
  index will have ten unique values, while the new
  guess_rec_per_key() function assumes that for each value in a
  non-unique index there are ten entries. For small tables, this
  results in different rec_per_key estimates.
------------------------------------------------------------
revno: 8130
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-06-03 16:26:48 +0530
message:
  - Bug#18808086: TEMP TABLE ASSERT LATCHES->EMPTY()
    || LATCH->M_LEVEL == SYNC_LEVEL_VARYING || LA
   
    Reverting the change as the workaround fix is not able
    to take care of all the cases.
    We need to wait for actual latching order fix being
    tracked by the bug#18883885
------------------------------------------------------------
revno: 8129 [merge]
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: trunk3
timestamp: Tue 2014-06-03 12:20:38 +0200
message:
  Null-Merge of the fix for Bug#18791851 CRASH IN ST_SELECT_LEX::PRINT WITH OPTIMIZER_TRACE ON SUBQUERY.
  A different fix will be pushed for trunk.
    ------------------------------------------------------------
    revno: 3902.324.92
    committer: Guilhem Bichot <guilhem.bichot@oracle.com>
    branch nick: 5.6
    timestamp: Fri 2014-05-30 10:37:00 +0200
    message:
      Bug#18791851 CRASH IN ST_SELECT_LEX::PRINT WITH OPTIMIZER_TRACE ON SUBQUERY
      
      The top query ends with:
      HAVING  alias1 . col_int_nokey  IN ( SELECT 2 FROM DUAL ) ;
      
      JOIN::prepare on the query owning HAVING saves select_lex->having to
      having_for_explain.
      Item_in_subselect::fix_fields (select_lex->having->fix_fields) replaces
      select_lex->having with "alias1 . col_int_nokey=2".
      SELECT 2 FROM DUAL becomes "detached", "orphan" (unit::exclude_level())),
      it won't be optimized, which is normal.
      having_for_explain is not changed, still points to item_in_subselect (bad).
      Later JOIN::optimize on the query owning HAVING wants to print this query,
      (in simplify_joins),
      select_lex::print gets HAVING from having_for_explain (stale item)
      and crashes when it reads master_unit() of 'SELECT 2 FROM DUAL' (NULL pointer,
      because 'detached' query).
      In trunk this does not happen because of WL7082: simplify_joins moved to the
      prepare phase, so join->optimized is false then, so select_lex::print uses
      select_lex::having which is correct
      (however, this bug has put me on the track of variant queries which do crash
      trunk, so a trunk-specific patch will be submitted; the present patch will not
      be put in trunk).
      
      Fix (conservative): update having_for_explain after fix_fields.
      Additionnally, a related robustness fix is added in select_lex::print
      (back-ported from trunk).
------------------------------------------------------------
revno: 8128 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-06-03 14:08:00 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.91
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-06-03 14:06:35 +0530
    message:
      Fixes a compilation problem on 32-bit platform.
      Use UINT64PF instead of ULINTPF for ib_uint64_t.
      approved by Marko over IM.
------------------------------------------------------------
revno: 8127
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-02 14:52:14 +0200
message:
  Bug#18869394 FIX COPYRIGHT HEADERS ON 5.7/TRUNK
  
  Approved by Murthy over e-mail.
------------------------------------------------------------
revno: 8126
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-06-02 11:35:56 +0530
message:
  - Bug#18808086: TEMP TABLE ASSERT LATCHES->EMPTY()
    || LATCH->M_LEVEL == SYNC_LEVEL_VARYING || LA
  
    Latching order during blob pessmistic insert is as follows:
    SYNC_FSP -> SYNC_NO_ORDER_CHECK -> SYNC_INDEX_TREE
  
    SYNC_NO_ORDER_CHECK is introduced by debug latch.
    For temporary tables we suppressed taking debug latch
    and so the latching order violating occured.
  
    Fixing the regression by re-enabling debug latch for
    temporary tables.
  
    Original issue of latching still needs to be reviewed
    from broader and general perspective through a separate bug.
  
    Approved by: Sunny (rb#5576) 
------------------------------------------------------------
revno: 8125
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-06-02 10:31:27 +0900
message:
  Bug#17869571 : CRASH/ASSERT AT BUF_PAGE_RELEASE_LATCH AT BUF0BUF.IC
  
  The assertion code was just too much.
  
  approved by Sunny in rb#5549
------------------------------------------------------------
revno: 8124
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-05-31 21:21:40 +0800
message:
  BUG#18466390 INCORRECT BEHAIVIOR WHEN SETTING UP SEMISYNC VARIABLES AS
               STARTUP OPTIONS
  
  Semisync didn't work as exepected if below two variables were
  set at starting server.
  - rpl_semi_sync_master_wait_no_slave
  - rpl_semi_sync_master_wait_for_slave_count
  The reason was that the variables were changed but the related
  internal status were not changed when initializing semisync.
  
  To fix it, internal actions are executed to change the internal status
  when initializing semisync at starting server.
------------------------------------------------------------
revno: 8123
committer: Pavan Naik<pavan.naik@oracle.com>
branch nick: mysql-trunk-gis
timestamp: Sat 2014-05-31 16:18:02 +0530
message:
  Follow up patch for WL#7220, WL#7221 and WL#7280. 
  Made some minor fixes and added new test cases to the tests under gis test suite.
------------------------------------------------------------
revno: 8122 [merge]
committer: Laasya Moduludu <laasya.moduludu@oracle.com>
branch nick: mysql-trunk
timestamp: Sat 2014-05-31 06:52:28 +0200
message:
  Upmerge of the 5.6.19 build
    ------------------------------------------------------------
    revno: 3902.324.90 [merge]
    author: 
    committer: Laasya Moduludu <laasya.moduludu@oracle.com>
    branch nick: mysql-5.6
    timestamp: Sat 2014-05-31 06:38:16 +0200
    message:
      Merge from mysql-5.6.19-release
        ------------------------------------------------------------
        revno: 3902.329.1
        tags: mysql-5.6.19
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.6.19-release
        timestamp: Tue 2014-05-06 12:13:29 +0200
        message:
          Disable dtrace for el7
------------------------------------------------------------
revno: 8121 [merge]
committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
branch nick: trunk
timestamp: Fri 2014-05-30 18:22:31 +0530
message:
  Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
  
  Problem:
  As the title mentions, on windows mysqldump tool is creating the dump file 
  without the content (tables, views etc) of a database that is dumped when the 
  error log file is deleted (by the user). Similar behaviour is not happening 
  on Linux flavours. The difference is due to the difference in handling of the 
  deleted files. In Windows, the line "if (flush_logs || opt_delete_master_logs)"
  in main.c of mysqldump is returning error whereas on linux this line is passing
  
  Solution:
  The core problem is that, in Windows, the new error log file is being created but
  the old file descriptor which is associated with both stdout and stderr stream
  is not being closed by the program. So a retry logic has been implemented where
  we close the error log's file descriptor (associated open fds of stderr and
  stdout) in the first attempt and open a new file and associating it with
  stdout/stderr streams.
    ------------------------------------------------------------
    revno: 3902.324.89
    committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
    branch nick: 5.6
    timestamp: Fri 2014-05-30 18:04:44 +0530
    message:
      Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
      
      Problem:
      As the title mentions, on windows mysqldump tool is creating the dump file 
      without the content (tables, views etc) of a database that is dumped when the 
      error log file is deleted (by the user). Similar behaviour is not happening 
      on Linux flavours. The difference is due to the difference in handling of the 
      deleted files. In Windows, the line "if (flush_logs || opt_delete_master_logs)"
      in main.c of mysqldump is returning error whereas on linux this line is passing
      
      Solution:
      The core problem is that, in Windows, the new error log file is being created but
      the old file descriptor which is associated with both stdout and stderr stream
      is not being closed by the program. So a retry logic has been implemented where
      we close the error log's file descriptor (associated open fds of stderr and
      stdout) in the first attempt and open a new file and associating it with
      stdout/stderr streams.
------------------------------------------------------------
revno: 8120
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-30 11:34:05 +0100
message:
  Post fix patch for BUG#17620053 because of a variable that could be used uninitialized.
------------------------------------------------------------
revno: 8119
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Fri 2014-05-30 11:41:19 +0200
message:
  Pre-requisite patch for WL6671 "Improve scalability by
  not using thr_lock.c locks for InnoDB tables".
  
  Change return type from bool to void for functions
  where we never use the returned value.
------------------------------------------------------------
revno: 8118
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-30 11:16:53 +0300
message:
  Make two tests compatible with multiple undo tablespaces:
  ./mysql-test-run --{mysqld,bootstrap}=--innodb-undo-tablespaces=2
------------------------------------------------------------
revno: 8117
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-30 10:04:02 +0200
message:
  Bug#18870850: CRASH WHEN WRITING TO WINDOWS EVENT LOG DURING SHUTDOWN
  
  Don't write to windows event log during shutdown.
  
  This is a temporary fix to make Windows pushbuild issues
  easier to identify and handle.
------------------------------------------------------------
revno: 8116
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Fri 2014-05-30 14:46:28 +0800
message:
  WL7220 Spatial Relation Check Functions
  WL7221 Geometry Set Operations 
  WL7280 WKB geometry container
  Add more files, this time all files added, sorry for the mess.
------------------------------------------------------------
revno: 8115
committer: David Zhao <david.zhao@oracle.com>
branch nick: trunk
timestamp: Fri 2014-05-30 14:35:29 +0800
message:
  WL7220 Spatial Relation Check Functions
  WL7221 Geometry Set Operations 
  WL7280 WKB geometry container
  
  Entire implementation of new GIS which uses Boost.Geometry(BG)'s GIS algorithm to do geometric computations. Currently we use Boost1.55.0 and some old GIS algorithms are still used if BG doesn't yet support the feature; OTOH, if BG supports a feature, we use BG's implementation. From now on server internally uses boost library, there are several implications: 
  1. We should follow the coding guideline when taking care of exceptions; 
  2. Boost is more picky to C++ compilers than existing MySQL code, so one might has to upgrade his C++ compiler. 
  3. You have to specify -DWITH_BOOST=/your/local/boost/root/dir when doing cmake configuration, and for only once you have to either manually download boost1.55.0 or specify -DDOWNLOAD_BOOST=1.
  
  We implemented a WKB container (WL7280) as an efficient adapter layer between MySQL GIS and BG, WKB data exchange between them has little cost(data copy/conversion is completely avoided), and this adapter is fit into existing Geometry classes to maximize code reuse and logical OO model.
  
  Based on that we reimplemented spatial relation check functions(WL7220: within, contains, intersects, disjoint, equal, touches, crosses, overlaps) and geometry set operations(WL7221: union, intersection, difference, symdifference) using BG, by dispatching to specific BG function calls according to the triplet: { operation-type, type-of-argument1, type-of-argument2}. And we implemented support for many of the triplet combinations that are not directly supported by BG but can be computed using BG's existing features.
  
  We also implemented support for geometry collection for above two groups of GIS operations indirectly using BG, because BG doesn't directly support geometry collections. We had to implement each GIS operation according to its OGC specification using BG's support for the six basic geometry types, i.e point, linestring, polygon, multipoint, multilinestring, multipolygon. We also implemented support for empty geometry collections and nested geometry collections, enabling both of them to be operand of above GIS operations, which makes GIS computation more practically usable. This is especially true for the support for empty geometry collections because in a tree of set operations any node can produce an empty geometry collection, which shouldn't stop the computation as old GIS did before, but allow operating on this empty collection and proceed GIS computation.
  
  This commit also contains some general purpose data structures and algorithms, including an Inplace_vector and its unit tests; an implementation of GIS algorithm to detect and reverse polygon ring order and its unit tests; as well as other unit tests for various functionality of new GIS.
  
  It also contains Tor's cmake build system support for using boost, which is able to download and make use of boost library, and also make use of our customized patches for multiple versions of BG as first priority over standard boost released code, by automatically choose the right version of customized patches according to the version of boost library used.
  
  It also contains Adam and Menelaos's customized BG patch for boost1.55.0.
  
  Finally, it contains QA team's entire test work which are prepared for the three GIS WLs.
  
  Some bugs that are found on old GIS are now naturally gone because we are using BG, and all bugs that are found on new GIS(WL7220, WL7221 and WL7280) are fixed.
  
  Bugs fixed:
  
  Bug #18423730 - WL7220: ASSERTION FAILED: LIST ITERATOR NOT INCREMENTABLE
  Bug #18701114 - WL7220: ASSERTION FAILED: (PTR != 0 && NBYTES > 0) || (PTR == 0 && NBYTES == 0),
  Bug #18851092 - GEOMFROMTEXT DOES NOT UNDERSTAND NESTED GEOMETRYCOLLECTIONS
  Bug 18408988 - WL7220 : OVERLAPPING MEMCPY CRASH IN BG_GEOMETRY_COLLECTION::CUT
  Bug 18408919 - WL7220 : OVERLAPPING MEMCPY CRASH IN GEOMETRY::AS_WKB
  Bug 18408875 - WL7220 : ASSERTION FAILED: *P == 0 || *P == 1
  Bug #17168699 OPEN POLYGON LISTED IN OGC GUIDELINES BUT NOT SUPPORTED IN MYSQL
  Bug #17168643 GEOMETRY COLLECTION IGNORES THE PROPER GEOMETRIES FOR ONE OPEN POLYGON
  Bug #17167765 ST_EQUALS FUNCTION RETURNS TRUE FOR THE SPATIALLY DIFFERENT GEOMETRIES.
  
  Bug#18492266 BOOST VERSION CHECK IN 5.7.5 LAB RELEASE
  Bug#72172    Boost version check in 5.7.4 lab release
  Bug#18503366 CHANGE OF WITH_BOOST SETTING IGNORED?
  Bug#72186    Change of WITH_BOOST setting ignored
  Bug#18497231 RELATIVE PATH FOR WITH_BOOST DOESN'T WORK
  Bug#72185    Relative path for WITH_BOOST doesn't work
  Bug#18499359
  Bug#72188    WITH_BOOST not shown in cmake-gui
------------------------------------------------------------
revno: 8114
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-29 20:00:49 +0300
message:
  Non-functional formatting fixup in ib_logf() calls.
------------------------------------------------------------
revno: 8113
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-29 09:15:38 +0100
message:
  Post fix for BUG#17620053, updating a mtr testcase result.
------------------------------------------------------------
revno: 8112 [merge]
committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-29 09:47:30 +0200
message:
  Upmerge of the 5.5.38 build
    ------------------------------------------------------------
    revno: 3902.324.88 [merge]
    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-29 09:39:00 +0200
    message:
      Upmerge of the 5.5.38 build
        ------------------------------------------------------------
        revno: 2875.595.5 [merge]
        author: murthy.narkedimilli@oracle.com
        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-05-29 09:34:50 +0200
        message:
          Merge from mysql-5.5.38-release
            ------------------------------------------------------------
            revno: 2875.596.1
            tags: mysql-5.5.38
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.38-release
            timestamp: Sun 2014-05-11 18:24:12 +0200
            message:
              Increment release version to resolve upgrade conflict issue
------------------------------------------------------------
revno: 8111 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-28 23:18:40 +0100
message:
  Merge of BUG#17620053 patch from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.87
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-28 22:09:56 +0100
    message:
      Bug#17620053 SET GTID_NEXT AND DROP TABLE WITH BOTH REGULAR AND
                   TEMPORARY TABLES
      
      Problem:
      =======
      DROP TABLE statement may be split before be sent to binlog if it
      contains regular tables and temporary tables or if it contains
      temporary tables with transactional and non-transactional storage
      engines.
      
      Issuing a DROP TABLE like the described above with GTIDs enables
      and having only one GTID associated to the statement (like the SQL
      thread does after issuing a SET GTID_NEXT='UUID:NUMBER') would lead to
      a problem as there is no GTIDs for all statements after splitting the
      original DROP TABLE.
      
      Analysis:
      ========
      DROP TABLE statements might be split because the behavior of the
      command in respect to the current transaction vary depending on table
      characteristics as follows:
      a) DROP TABLE <regular table> will be committed immediately;
      b) DROP TABLE <temp trans table> will be committed with the current
         transaction (after COMMIT);
      c) DROP TABLE <temp non-trans table> will be committed immediately.
      
      So, mixing these types of tables in a single DROP statement will make
      MySQL server to split the statement in up to three DROP statements in
      the binlog.
      
      Without using GTID, there is no problem with this approach.
      
      If GTIDs are enabled, and it is not in AUTOMATIC mode, issuing a DROP
      TABLE statement that mixes table types above described would lead the
      server to have no GTIDs enough to log all split statements into
      binlog.
      
      Also, DROP TABLE using IF EXISTS will always binlog the drop for all
      tables specified in the statement, even if the tables doesn't exist.
      
      With regular tables it doesn't represent a problem, but with
      temporary tables this was leading to the following problem: as
      temporary tables are split into transactional and non-transactional
      ones, the non-existent tables of a DROP TEMPORARY TABLE statement
      were assumed as transactional ones. So, if a DROP TEMPORARY TABLE
      with two non-transactional temporary tables were issued at the master
      server, it would binlog only one DROP statement containing the two
      tables. But if a filter made one of the temporary tables inexistent
      on the slave, when the SQL thread executes the statement it would
      try to split the statement as it would have a non-transactional
      existent temporary table and a transactional non-existent temporary
      table. And this split will be a problem because the SQL thread will
      have only one GTID for the original DROP statement but having to
      binlog two DROP statements.
      
      Found also a problem with the slave dropping temporary tables when
      it detects that the master has restarted, by calling
      close_temporary_tables()@sql_base.cc, as it was "binlogging" one DROP
      statement per pseudo-thread and per database, but was mixing
      transactional and non-transactional temporary tables in a single
      DROP statement. 
      
      
      Fix:
      ===
      
      The first part of the patch was to throw an error in the client
      session if GTID_NEXT is set to an 'UUID:NUMBER' and a DROP TABLE
      statement is issued mixing the table types above described.
      
      The second part of the patch was to group the inexistent temporary
      tables and only assume them as transactional if at least one
      transactional temporary table is dropped. If no transactional
      temporary tables are dropped, the inexistent temporary tables are
      assumed as non-transactional temporary tables.
      
      The third part of the patch fixed the problem with
      close_temporary_tables().
      
      @ mysql-test/include/save_master_pos.inc
      
        Moved the code that saved the master position to a place that it
        will always be executed.
      
      @ mysql-test/include/sync_slave_sql.inc
      
        Added a new parameter to permit forcing the sync using master
        position instead of GTID_EXECUTED, even if the use_gtids option
        is enabled.
      
        Added a code before the sync to verify is the SQL slave is in the
        current master position (it is already synced).
      
        If the SQL slave is assumed as already synced, we reduce the sync
        timeout to only 1 second. So, if the test is using GTIDs and not
        specified to force sync using master position, the sync command
        will use the master GTID_EXECUTED to sync and will timeout only
        if there is a discrepancy between master and slave GTID_EXECUTED.
      
      @ mysql-test/*/rpl/?/rpl_gtid_drop_table*
      
        mtr files for testing the bug/patch.
      
      @ mysql-test/suite/rpl/r/*.result
      
        Some tests had their result file changed because they listed the
        binlog events for DROP TABLE statements that were split both at
        master and slave and now some of these statements do not split
        anymore.
      
      @ sql/share/errmsg-utf8.txt
      
        Added a new error message for statements that would be split in the
        binary log when @@SESSION.GTID_NEXT is set to 'UUID:NUMBER'.
      
      @ sql/sql_table.cc
      
        Added code to mysql_rm_table() to verify if the statement is safe
        to be executed (if it is not mixing table types or if
        @@SESSION.GTID_NEXT is not set to 'UUID:NUMBER').
      
        Added code to mysql_rm_table_no_locks() to group the inexistent
        temporary tables and assume them as non-transactional if there is no
        transactional temporary tables dropped in the statement.
      
      @sql/sql_base.cc
      
        Changed close_temporary_tables() function to binlog transactional
        and non-transactional temporary tables in distinct DROP statements.
------------------------------------------------------------
revno: 8110
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-wl6671-pre
timestamp: Wed 2014-05-28 20:16:41 +0400
message:
  Pre-requisite patch for WL6671 "Improve scalability by
  not using thr_lock.c locks for InnoDB tables".
  
  Changed MDL_thread class to require explicit request to
  enable release-on-notification logic. Such approach should
  cause less "gotcha" situations during test development.
  
  Fixed HogLockTest5 to not rely on old behavior on clean-up.
------------------------------------------------------------
revno: 8109
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Wed 2014-05-28 17:56:34 +0200
message:
  Bug#18799511 GET RID OF DYNAMIC_ARRAY IN LEX::PLUGINS
  
  Post-push fix: avoid size_t <=> uint warnings on windows.
------------------------------------------------------------
revno: 8108
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Wed 2014-05-28 17:39:31 +0200
message:
  Bug#18851597 GET RID OF DYNAMIC_ARRAY IN REPL_IGNORE_SERVER_IDS
  
    Our array abstractions should be removed, and substituted with modern data
    structures.
      
    Templatized vectors are:
     - type safe
     - easier to read/maintain
     - faster
------------------------------------------------------------
revno: 8107
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-28 17:09:05 +0200
message:
  Follow-up to WL#7854 Deprecate and remove mysqlhotcopy.
        
  Remove removal of man pages from packaging scripts.
------------------------------------------------------------
revno: 8106
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-joh-signal
timestamp: Wed 2014-05-28 10:37:40 +0200
message:
  Bug#18730437: REFACTOR SIGNAL HANDLING CODE
  
  This is a follow-up to Bug#18411456: REMOVE THR_ALARM.C
  
  This patch updates signal handling code to match coding standard,
  removes duplicated or dead code, and improves code documentation.
  
  No functional changes.
------------------------------------------------------------
revno: 8105 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-28 09:48:39 +0200
message:
  Null merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.86
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-28 09:46:01 +0200
    message:
      Bug#18281535 - Updated usergroup to mysql on datadir
------------------------------------------------------------
revno: 8104 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-28 09:38:30 +0200
message:
  Merge from 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.328.1
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-28 09:35:01 +0200
    message:
      Bug#18281535 - Updated usergroup to mysql on datadir
------------------------------------------------------------
revno: 8103 [merge]
committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
branch nick: trunk
timestamp: Wed 2014-05-28 12:54:15 +0530
message:
  Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
  
  Reverting patch due to failures.
    ------------------------------------------------------------
    revno: 3902.324.85
    committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
    branch nick: 5.6
    timestamp: Wed 2014-05-28 12:50:37 +0530
    message:
      Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
      
      Reverting the patch due to failures.
------------------------------------------------------------
revno: 8102
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Wed 2014-05-28 12:33:20 +0800
message:
  BUG#17447086 - FTS: DUPLICATE FTS_DOC_ID VALUE ON TABLE
  
  It's a race condiction in fts_get_next_doc_id and was not completely
  fixed in rb://2171(bug#16469399).
  
  rb://5523 approved by Jimmy.Yang
------------------------------------------------------------
revno: 8101
committer: HaixLi < haixiang.li@oracle.com >
branch nick: mysql-trunk-bug-18487060-20140528
timestamp: Tue 2014-05-27 22:10:55 -0400
message:
  Bug #18487060: ASSERTION FAILED: !TABLE || (!TABLE->READ_SET || 
                 BITMAP_IS_SET(TABLE->READ_SET,
  
  Description:
  ------------
  This bug trigger a assert when SQL statement call "make_set()" function object 
  of GROUP BY clause.
  
  
  Analysis:
  ---------
  This bug can not handle "make_set()" function of GROUP BY clause.
  
  Fix:
  ----
  If a node is "func item" node, we should walk this node and its args which
  maybe covers other item. So we extend with support for walking the "item" 
  field.
  
  Test case added.
------------------------------------------------------------
revno: 8100
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-wl6671-pre
timestamp: Tue 2014-05-27 21:01:57 +0400
message:
  Pre-requisite patch for WL6671 "Improve scalability by
  not using thr_lock.c locks for InnoDB tables".
  
  Changed name of MDL_context::is_lock_owner() method to
  owns_equal_or_stronger_lock(). New name better reflects
  its function.
------------------------------------------------------------
revno: 8099
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-27 15:04:53 +0200
message:
  Disabling a failing test on windows
------------------------------------------------------------
revno: 8098 [merge]
committer: Marcin Babij <marcin.babij@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-27 14:52:18 +0200
message:
  BUG#18830493 - REQUEST TO BACKPORT BUG#18311024 TO 5.6
  
  mysql_config_editor: invalid write in remove_login_path when given empty path.
  
  Fix was to calculate correct indexes to copy.
  By the way memory check tool found memory leaks, which were easy to get fixed, so I attach them in this commit.
    ------------------------------------------------------------
    revno: 3902.324.84
    committer: Marcin Babij <marcin.babij@oracle.com>
    branch nick: 18830493-5.6
    timestamp: Tue 2014-05-27 13:52:47 +0200
    message:
      BUG#18830493 - REQUEST TO BACKPORT BUG#18311024 TO 5.6
      
      mysql_config_editor: invalid write in remove_login_path when given empty path.
      
      Fix was to calculate correct indexes to copy.
      By the way memory check tool found memory leaks, which were easy to get fixed, so I attach them in this commit.
------------------------------------------------------------
revno: 8097 [merge]
committer: Aditya A <aditya.a@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-27 17:09:50 +0530
message:
  Bug#18652854 INNODB MEMCACHED ASSERTION 
  	     `!CONN_DATA->CRSR_TRX' FAILED
  
  [ Merged from mysql-5.6 ]
    ------------------------------------------------------------
    revno: 3902.324.83
    committer: Aditya A <aditya.a@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-05-27 16:45:49 +0530
    message:
      Bug#18652854 INNODB MEMCACHED ASSERTION 
      	     `!CONN_DATA->CRSR_TRX' FAILED
      
      PROBLEM
      -------
      
      While calling the memcached command 
      "flush_all" , we first try to initialize
      the connection and transaction .If the 
      transaction is in TRX_STATE_NOT_STARTED 
      state we do not set CONN_DATA->CRSR_TRX
      as NULL.This was causing the assert.
      
      FIX
      --- 
      If the transaction is in TRX_STATE_NOT_STARTED
      state ,we will reuse the transaction. 
      
      [Approved by Jimmy #rb5469]
------------------------------------------------------------
revno: 8096
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: multi-update-bug
timestamp: Tue 2014-05-27 13:24:16 +0200
message:
  WL#6635 Make use of condition filtering in the optimizer
  
  Stabilizing test results: Added FORCE INDEX and STRAIGHT_JOIN
  to one multi-table update statement in innodb_multi_update.test
  to increase likelihood that the same query plan is produced on
  every run.
------------------------------------------------------------
revno: 8095
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-27 10:56:34 +0200
message:
  Fix two Clang-specific unit test compiler warnings with MERGE_UNITTTESTS=0
------------------------------------------------------------
revno: 8094
committer: Atanu Ghosh <atanu.ghosh@oracle.com>
branch nick: wl3836-complete
timestamp: Tue 2014-05-27 11:57:38 +0530
message:
  Bug#18826853: SERVER HANG WHEN ENABLE/DISABLE OFFLINE_MODE
  
  Problem: When offline_mode was set to on, other global variables
           were allowed to be updated. This caused a deadlock on the
           mutex LOCK_global_system_variables.
  
  Fix: When offline_mode is set to on the the LOCK_global_system_variables
       mutex is not unlocked until the update completes, thereby
       preventing a deadlock.
------------------------------------------------------------
revno: 8093
committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
branch nick: trunk
timestamp: Tue 2014-05-27 09:42:00 +0530
message:
  Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
  
  Problem:
  As the title mentions, on windows mysqldump tool is creating the dump file 
  without the content (tables, views etc) of a database that is dumped when the 
  error log file is deleted (by the user). Similar behaviour is not happening 
  on Linux falvours. The difference is due to the difference in handling of the 
  deleted files. In Windows, the line "if (flush_logs || opt_delete_master_logs)"
  in main.c of mysqldump is returning error whereas on linux this line is passing
  
  Solution:
  The core problem is that, in Windows, the new error log file is being created but
  the old file descriptor which is associated with both stdout and stderr stream
  is not being closed by the program. So a retry logic has been implemented where
  we close the error log's file descriptor (associated open fds of stderr and
  stdout) in the first attempt and open a new file and associating it with
  stdout/stderr streams.
------------------------------------------------------------
revno: 8092
committer: Sunny Bains <Sunny.Bains@Oracle.Com>
branch nick: trunk
timestamp: Tue 2014-05-27 11:10:18 +1000
message:
  Revert anirudh.mangipudi@oracle.com-20140526151526-cqaqbh3gxcdt9jba due to build
  issues on PB2.
------------------------------------------------------------
revno: 8091
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-05-27 08:44:37 +0800
message:
  BUG#18729428 - GB18030_CHINESE_CI ERRONEOUSLY CONSIDERS 'Y' IS EQUAL TO '~'
  
  A followup patch for this bug, re-record 3 mtr test cases.
------------------------------------------------------------
revno: 8090 [merge]
committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
branch nick: trunk
timestamp: Mon 2014-05-26 20:45:26 +0530
message:
  Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
  
  Problem:
  As the title mentions, on windows mysqldump tool is creating the dump file 
  without the content (tables, views etc) of a database that is dumped when the 
  error log file is deleted (by the user). Similar behaviour is not happening 
  on Linux falvours. The difference is due to the difference in handling of the 
  deleted files. In Windows, the line "if (flush_logs || opt_delete_master_logs)"
  in main.c of mysqldump is returning error whereas on linux this line is passing
  
  Solution:
  The core problem is that, in Windows, the new error log file is being created but
  the old file descriptor which is associated with both stdout and stderr stream
  is not being closed by the program. So a retry logic has been implemented where
  we close the error log's file descriptor (associated open fds of stderr and
  stdout) in the first attempt and open a new file and associating it with
  stdout/stderr streams.
    ------------------------------------------------------------
    revno: 3902.324.82
    committer: Anirudh Mangipudi <anirudh.mangipudi@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-05-26 20:39:56 +0530
    message:
      Bug #17076131 MYSQLDUMP FAILS WHEN ERR LOG ON RUNNING SERVER IS DELETED ON WINDOWS
      
      Problem:
      As the title mentions, on windows mysqldump tool is creating the dump file 
      without the content (tables, views etc) of a database that is dumped when the 
      error log file is deleted (by the user). Similar behaviour is not happening 
      on Linux falvours. The difference is due to the difference in handling of the 
      deleted files. In Windows, the line "if (flush_logs || opt_delete_master_logs)"
      in main.c of mysqldump is returning error whereas on linux this line is passing
      
      Solution:
      The core problem is that, in Windows, the new error log file is being created but
      the old file descriptor which is associated with both stdout and stderr stream
      is not being closed by the program. So a retry logic has been implemented where
      we close the error log's file descriptor (associated open fds of stderr and
      stdout) in the first attempt and open a new file and associating it with
      stdout/stderr streams.
------------------------------------------------------------
revno: 8089
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Mon 2014-05-26 15:32:39 +0200
message:
  Bug#18799511 GET RID OF DYNAMIC_ARRAY IN LEX::PLUGINS
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Also: fix memory leak in unit tests: call cleanup_variables()
------------------------------------------------------------
revno: 8088
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-05-26 14:36:03 +0800
message:
  BUG 18726196 - BROKEN TABLE IN MYSQL COMMAND LINE CLIENT WHEN USING GB18030
  
  The output of SELECT with gb18030 charset looks incorrect, the width of some
  field could be too narrow. Instead of using my_numcells_8bit, we should use
  my_numcells_mb for gb18030 charset to fix this issue.
  
  rb#5397, approved by Ramil.
------------------------------------------------------------
revno: 8087
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-05-26 14:29:12 +0800
message:
  BUG 18700668 - GB18030 + FTS: ASSERTION FAILED: ((UCHAR) (*SRC) <= 0X7F)
  
  GB18010 has a strict assertion when we pass single-byte invalid GB18030
  char to do caseup/casedn, which should be removed away, since other
  charsets also don't do this check but do caseup/casedn for the char
  directly.
  
  Also, the caseup_multiply/casedn_multiply of
  my_charset_gb18030_unicode_520 are not correct. They should be both 2
  instead of 1, like other GB18030 charsets. This is a follow-up fix
  of this bug, which makes the original test passed for gb18030_unicode_520.
  
  rb#5350, approved by Ramil.
------------------------------------------------------------
revno: 8086
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-05-26 14:21:00 +0800
message:
  BUG#18729428 - GB18030_CHINESE_CI ERRONEOUSLY CONSIDERS 'Y' IS EQUAL TO '~'
  
  According to Bernt's comment, this issue is a result from a typo and
  copy-and-paste coding. So we should change the sort weight of '~' to '~'
  instead of 'Y'.
  
  rb#5478, approved by Ramil.
------------------------------------------------------------
revno: 8085
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-26 06:18:53 +0200
message:
  Disabled a failing test on windows
------------------------------------------------------------
revno: 8084
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Mon 2014-05-26 11:24:22 +0800
message:
  Fix Bug #18772498 - ASSERT PAGE_ZIP_VALIDATE(PAGE_ZIP, PAGE_ALIGN(REC), NULL),
  GIS, PURGE
  
  
  rb://5463 approved by Sunny Bains
------------------------------------------------------------
revno: 8083 [merge]
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Fri 2014-05-23 20:34:17 +0800
message:
  Null merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.81
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-5.6-bugfix1
    timestamp: Fri 2014-05-23 20:29:33 +0800
    message:
      Bug#18683832 ASSERT RESULT != FTS_INVALID, FTS_TRX_ROW_GET_NEW_STATE()
      
      Backport rb://1673, approved by Sunny & Sveta.
------------------------------------------------------------
revno: 8082 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Fri 2014-05-23 12:52:05 +0200
message:
  NULL merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.80
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-cmake
    timestamp: Fri 2014-05-23 12:50:24 +0200
    message:
      Bug#18605389 PROVIDE OPTION TO LINK AGAINST LIBCSTD INSTEAD OF STLPORT4 ON SOLARIS 10+
      Bug#72352    Provide option to link against libCstd instead of stlport4 on Solaris 10+
      
      Patch for 5.6
      This patch enables the use of Cstd when building with SunStudio.
      It only works for client code, as the server depends on C++98
      Usage:
      cmake -DWITHOUT_SERVER=1 -DSUNPRO_CXX_LIBRARY=Cstd
      
      Also backport:
      Bug#14367046: PROBLEM BUILDING CLIENT ONLY
------------------------------------------------------------
revno: 8081 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-23 10:45:53 +0530
message:
  Bug#18000079 - binlog_dump_non_block disappeared in mysql 5.6
  
  Null merge from mysql-5.6 ( Fixing pb2 Werror build failure in 5.6)
    ------------------------------------------------------------
    revno: 3902.324.79
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-05-23 10:42:55 +0530
    message:
      Bug#18000079 - binlog_dump_non_block disappeared in mysql 5.6
      
      Fixing post push pb2 Werror build failure
------------------------------------------------------------
revno: 8080 [merge]
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-22 21:24:44 +0530
message:
  Null merge from mysql-5.6 to mysql-5.7
    ------------------------------------------------------------
    revno: 3902.324.78
    committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-22 21:23:03 +0530
    message:
      Bug #16963396 INNODB: USE OF LARGE EXTERNALLY-STORED FIELDS MAKES CRASH
      RECOVERY LOSE DATA
      
      Problem:
      
      When too-large blob fields are used, InnoDB overwrites its most recent
      checkpoint in its redo logs.
      
      Solution:
      
      Ensure that the total blob length does not exceed 10% of the redo log file
      size.
      
      rb#5399 approved by Marko, Nuno, Manish. 
      Venkat also contributed to patch (in replication related test case).
------------------------------------------------------------
revno: 8079 [merge]
committer: Atanu Ghosh <atanu.ghosh@oracle.com>
branch nick: wl3836-complete
timestamp: Thu 2014-05-22 19:06:39 +0530
message:
   WL#3836:  Method to bring servers off line
    
    In order to be able to upgrade a server, a method is
    developed to bring a server off line. It does the following:
    
    - Gracefully force connected clients to disconnect 
    except the clients with SUPER privilege.
    
    - Allow super users to connect to manage the system.
    
    - Deny any other user access to the system.
    
    - Allow the replication slave thread to keep applying data to the server.
    ------------------------------------------------------------
    revno: 8021.1.1
    committer: Atanu Ghosh <atanu.ghosh@oracle.com>
    branch nick: mysql-trunk-wl3836
    timestamp: Fri 2014-05-16 19:44:48 +0530
    message:
       WL#3836:  Method to bring servers off line
        
        In order to be able to upgrade a server, a method is
        developed to bring a server off line. It does the following:
        
        - Gracefully force connected clients to disconnect 
        except the clients with SUPER privilege.
        
        - Allow super users to connect to manage the system.
        
        - Deny any other user access to the system.
        
        - Allow the replication slave thread to keep applying data to the server.
------------------------------------------------------------
revno: 8078 [merge]
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-22 14:58:36 +0300
message:
  Null merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.77 [merge]
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql_5.6
    timestamp: Thu 2014-05-22 14:55:39 +0300
    message:
      Merge local copy to mysql-5.6.
------------------------------------------------------------
revno: 8077
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-22 13:12:01 +0200
message:
  Re-enable tests which now pass with updated PB2 configurations.
------------------------------------------------------------
revno: 8076
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-22 13:59:26 +0300
message:
  Bug#18815992 REMOVE UNUSED CODE FROM THE INNODB/MEMCACHED API
  
  Remove some unused function definitions and declarations.
  
  Correct the return types of cb_trx_release_t, ib_cfg_trx_level and
  cb_cfg_trx_level_t.
  
  rb#5480 approved by Jimmy Yang
------------------------------------------------------------
revno: 8075 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: trunk-revert
timestamp: Thu 2014-05-22 16:24:01 +0530
message:
  Bug #17657223	EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Reverting the patch for the above bug due to regression.
    ------------------------------------------------------------
    revno: 3902.324.76
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: revert-5.6
    timestamp: Thu 2014-05-22 16:15:37 +0530
    message:
      Bug #17657223   EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
              Reverting the patch for the above bug due to regression.
------------------------------------------------------------
revno: 8074
committer: Abhishek Ranjan <abhishek.ar.ranjan@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-22 15:56:54 +0530
message:
  WL#6614 : Define and reimplement IGNORE
  WL#6891 : Define and reimplement STRICT MODE
  
  IGNORE Reimplementation :
  This worklog reimplements IGNORE. In the previous implementation,
  errors were downgraded to warnings inside THD::raise_condition()
  based on SELECT_LEX::no_error. This patch changes it to use an
  Internal_error_handler activated just for statements which support
  IGNORE keyword to downgrade specific errors to warnings.
  
  Bugs Fixed by this worklog :
  
  - Bug#11744960 : INSERT IGNORE SHOULD RETURN WARNINGS
  - Bug#11752648 : MULTI-DELETE IGNORE DOES NOT REPORT WARNINGS
  - Bug#16522924 : UPDATE TRIGGER INVOKED WHEN UPDATE IGNORE
                   MEANS THAT NO UPDATE IS PERFORMED
  - Bug#16860715 : ASSERT IN PROTOCOL::END_STATEMENT DURING DELETE
                   IGNORE
  - Bug#16860829 : ASSERT IN DIAGNOSTICS_AREA::SET_ERROR_STATUS
  - Bug#17550423 : DELETE IGNORE GIVES INCORRECT RESULT WITH FOREIGN
                   KEY FOR PARENT TABLE
  
  
  STRICT Mode Reimplementation :
  This worklog reimplements STRICT mode. In the previous implementation
  When STRICT MODE was enabled, warning were upgraded to errors inside
  THD::raise_condition() based on THD::abort_on_warning flag.
  This patch uses an Internal_error_handler activated just for
  STRICT MODE to do the upgrade of specific warnings to errors.
  
  Bugs Fixed by this worklog :
  
  - BUG#11751889: TRIGGERS OVERRIDE STRICT SQL_MODE
  - Bug#16976939: FIX ERROR MESSAGE ON DUPLICATE INDEX CREATION AND
                  STRICT MODE
  - BUG#18526888: STRICT MODE DOES NOT APPLY TO MULTI DELETE STATEMENT
------------------------------------------------------------
revno: 8073 [merge]
committer: Mauritz Sundell <mauritz.sundell@oracle.com>
branch nick: mysql-5.7
timestamp: Thu 2014-05-22 11:53:07 +0200
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.75
    committer: Mauritz Sundell <mauritz.sundell@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-22 11:42:51 +0200
    message:
      WL#7774 Make MTR sync_slave_sql_with_master.inc handle NDB
      
        Make source include/sync_slave_sql_with_master.inc work as a
        replacement to sync_slave_with_master for NDB in MTR tests.
      
        The usage of mysqltest command sync_slave_with_master in MTR
        tests should in time be replaced.
      
        The replacement source include/sync_slave_sql_with_master.inc
        did not work in exactly the same way for NDB.
      
        In WL#7205 some uses of sync_slave_with_master was replaced in
        some rpl tests.  These changes make the corresponding NDB test
        to fail.
      
        This patch make changes needed for sync_slave_sql_with_master.inc
        to work with NDB.
      
        Committed transactions using the NDB engine are not directly binlogged
        but first packed in epochs.  This introduce an extra delay between
        commit and binlogging that must be taken into account when one wants
        to make sure a slave have replayed the committed transactions.
      
        In MySQL Cluster sync_slave_with_master is patched to make sure that
        current epoch is binlogged before taking master binlog position.
        See Bug#18375920.
------------------------------------------------------------
revno: 8072 [merge]
committer: Harin Vadodaria <harin.vadodaria@oracle.com>
branch nick: trunk
timestamp: Thu 2014-05-22 14:33:04 +0530
message:
  Bug#17201924 and Bug#18178997 : YASSL:MISSING CLOSEDIR()
                                  IN
                                  SSL_CTX_LOAD_VERIFY_
                                  LOCATIONS()
                                  and
                                  OFF-BY-ONE PROBLEM IN
                                  VOID CERTDECODER::
                                  GETDATE(DATETYPE DT)
                                  IN ASN.CPP
  
  Description : merge from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.74 [merge]
    committer: Harin Vadodaria <harin.vadodaria@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-05-22 14:29:52 +0530
    message:
      Bug#17201924 and Bug#18178997 : YASSL:MISSING CLOSEDIR()
                                      IN
                                      SSL_CTX_LOAD_VERIFY_
                                      LOCATIONS()
                                      and
                                      OFF-BY-ONE PROBLEM IN
                                      VOID CERTDECODER::
                                      GETDATE(DATETYPE DT)
                                      IN ASN.CPP
      
      Description : merge from 5.5 to 5.6.
        ------------------------------------------------------------
        revno: 2875.595.4
        committer: Harin Vadodaria <harin.vadodaria@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-05-22 14:26:09 +0530
        message:
          Bug#17201924 and Bug#18178997 : YASSL:MISSING CLOSEDIR()
                                          IN
                                          SSL_CTX_LOAD_VERIFY_
                                          LOCATIONS()
                                          and
                                          OFF-BY-ONE PROBLEM IN
                                          VOID CERTDECODER::
                                          GETDATE(DATETYPE DT)
                                          IN ASN.CPP
          
          Description : Fixes corner cases in yassl code.
                        Refer to bug page for details.
------------------------------------------------------------
revno: 8071
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-22 10:40:54 +0300
message:
  Skip a kill/restart test on embedded server.
------------------------------------------------------------
revno: 8070 [merge]
committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-21 21:54:35 +0530
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.73
    committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-21 21:51:16 +0530
    message:
      Post push fix for Bug#11760197
------------------------------------------------------------
revno: 8069
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Wed 2014-05-21 21:31:36 +0530
message:
  Bug #18057562: PROXY USERS LOCKED OUT WHEN UNDERLYING PROXIED USER
                 PASSWORD IS EXPIRED
  Followup patch for trunk, to fix the main.plugin_auth.test testcase. 
------------------------------------------------------------
revno: 8068 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Wed 2014-05-21 18:58:30 +0530
message:
  Bug #18057562: PROXY USERS LOCKED OUT WHEN UNDERLYING PROXIED USER PASSWORD IS EXPIRED
  
  Merging from mysql-5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.324.72
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Wed 2014-05-21 18:53:12 +0530
    message:
      Bug #18057562: PROXY USERS LOCKED OUT WHEN UNDERLYING PROXIED USER PASSWORD IS EXPIRED
      
      Description: The proxy user can't execute any commands when
      the proxied user's password is expired.
      
      Analysis: During the authentication of a proxy user, the
      "password_expired" variable is set to "password_expired"
      value of the proxied user. Because of this, the proxy user
      also can't execute any operations when the proxied user's
      password is expired.
      
      Fix: Allowing proxy users to execute operations even if
      proxied user's password is expired. At the same time,
      expired passwords should prevent the proxied user from
      performing operations until SET PASSWORD is issued (this
      behavior is already present).
      As a fix, We are copying the acl_user password expired flag
      to the security context. Before the acl_user gets updated.
      So, that the current user password_expired flag will be
      maintained.
      We are checking for mpvio.acl_use->password_expired  
      status in stead of acl_user, by this the older proxy 
      clients(5.5 client) can connect and execute the quires 
      successfully even the proxied user password expires.
------------------------------------------------------------
revno: 8067
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Wed 2014-05-21 15:23:08 +0200
message:
  Bug#18808513: MAIN.GREEDY_SEARCH TRIGGERS DBUG_ASSERT(JOIN->BEST_READ < DBL_MAX) ON 32BIT WIN
  Bug#18808559: SUBQUERY TESTS FAIL WITH WRONG RESULT ON 32BIT WIN
  
  Disable affected tests on weekly-trunk since they only fail on 32 bit Windows
  and 32 bit Windows is only tested on weekly-trunk.
------------------------------------------------------------
revno: 8066
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Wed 2014-05-21 15:05:18 +0200
message:
  Bug#18605389 PROVIDE OPTION TO LINK AGAINST LIBCSTD INSTEAD OF STLPORT4 ON SOLARIS 10+
  Bug#72352    Provide option to link against libCstd instead of stlport4 on Solaris 10+
  
  Patch for 5.7
  This patch enables the use of Cstd when building with SunStudio.
  It only works for client code, as the server depends on C++98
------------------------------------------------------------
revno: 8065
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug18718491
timestamp: Wed 2014-05-21 13:39:05 +0200
message:
  Bug#18718491: MSI PACKING BROKEN WITH VS2013
  
  This patch fixes problems building MSI packages with Visual Studio 2013.
  VS2013 requires WiX Toolkit 3.8 and before this patch the code only worked
  with 3.5 and 3.6.
  
  The patch raises the minimum required version of CMake to 2.8.12 for Windows
  only, as this version of CMake is required for VS2013. This also allows removing
  a workaround for an old CMake bug (CPACK_MONOLITHIC_INSTALL).
  
  The recommended way of building ZIP and MSI packages is:
  1: Run CMake as normal, specifying VS2013 as generator.
  2: cmake --build . --config (e.g debug) --target package
  3: cmake --build . --config (e.g. debug) --target msi
------------------------------------------------------------
revno: 8064 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Wed 2014-05-21 17:01:19 +0530
message:
  Bug#18000079 - binlog_dump_non_block disappeared in mysql 5.6
  
  
  Problem:-
  Before 5.6.5 or so, it was possible for a client(like Mysql-connector) to 
  set the BINLOG_DUMP_NON_BLOCK flag in the initial handshake 
  packet (COM_BINLOG_DUMP). Then this was (wrongly) removed, and the current 
  patch adds it back. As such, clients connecting to a server issuing a 
  COM_BINLOG_DUMP with the flag not set will not get an EOF when the server 
  has sent the last event in the binary log. The connection will block.
  
  In all versions, it is possible to get the same behaviour by setting 
  server_id=0(while is the default value of server_id).
  A slave server never sets the flag and never sets server_id=0, because it never 
  needs non-blocking behaviour.
  Mysqlbinlog never sets the flag. If it needs the non-blocking behaviour, 
  it sets server_id=0.
  
  In order to test the patch, we need mysqlbinlog to set the flag when it needs 
  non-blocking behaviour. However, since mysqlbinlog also sets server_id=0, 
  *just changing it to allow* setting the flag in mysqlbinlog will not verify the bug 
  in an unpatched server.
  
  Currently it is possible to set the server_id used by mysqlbinlog, but only 
  if *blocking* behaviour is requested (using mysqlbinlog --stop-never-slave-server-id=X), 
  and then the value 0 is not allowed.
  
  Solution:-
  
  So in order to test the bug, We are introducing a new option:
    --connection-server-id=NUMBER
  When mysqlbinlog is run in blocking mode:
  - default is 1
  - range is 1..maxint
  When mysqlbinlog is run in non-blocking mode:
  - default is 0
  - range is 0..maxint
  The difference in default value is needed for backward compatibility. 
  The difference in range is needed since 0 forces non-blocking mode regardless of the flag.
  
  With this in place, we can verify the bug in a non-patched server by running mysqlbinlog 
  in non-blocking mode and set connection-server-id=1.
  
  NOTES:- A few remarks about this flag and also about the clients shipped with the server 
  that make use of COM_BINLOG_DUMP.
  
    - An IO thread never issues (and has never issued)
      COM_BINLOG_DUMP with the flag set;
  
    -  Mysqlbinlog will set the flag always, unless --stop-never is specified.
  
    - The dump thread falls back to non-blocking behavior if the
      connecting client states that it has server_id=0;
    ------------------------------------------------------------
    revno: 3902.324.71
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Wed 2014-05-21 16:27:56 +0530
    message:
      Bug#18000079 - binlog_dump_non_block disappeared in mysql 5.6
      
      
      Problem:-
      Before 5.6.5 or so, it was possible for a client(like Mysql-connector) to 
      set the BINLOG_DUMP_NON_BLOCK flag in the initial handshake 
      packet (COM_BINLOG_DUMP). Then this was (wrongly) removed, and the current 
      patch adds it back. As such, clients connecting to a server issuing a 
      COM_BINLOG_DUMP with the flag not set will not get an EOF when the server 
      has sent the last event in the binary log. The connection will block.
      
      In all versions, it is possible to get the same behaviour by setting 
      server_id=0(while is the default value of server_id).
      A slave server never sets the flag and never sets server_id=0, because it never 
      needs non-blocking behaviour.
      Mysqlbinlog never sets the flag. If it needs the non-blocking behaviour, 
      it sets server_id=0.
      
      In order to test the patch, we need mysqlbinlog to set the flag when it needs 
      non-blocking behaviour. However, since mysqlbinlog also sets server_id=0, 
      *just changing it to allow* setting the flag in mysqlbinlog will not verify the bug 
      in an unpatched server.
      
      Currently it is possible to set the server_id used by mysqlbinlog, but only 
      if *blocking* behaviour is requested (using mysqlbinlog --stop-never-slave-server-id=X), 
      and then the value 0 is not allowed.
      
      Solution:-
      
      So in order to test the bug, We are introducing a new option
      (available only for debug builds):
        --connection-server-id=NUMBER
      When mysqlbinlog is run in blocking mode:
      - default is 1
      - range is 1..maxint
      When mysqlbinlog is run in non-blocking mode:
      - default is 0
      - range is 0..maxint
      The difference in default value is needed for backward compatibility. 
      The difference in range is needed since 0 forces non-blocking mode regardless of the flag.
      
      With this in place, we can verify the bug in a non-patched server by running mysqlbinlog 
      in non-blocking mode and set connection-server-id=1.
      
      NOTES:- A few remarks about this flag and also about the clients shipped with the server 
      that make use of COM_BINLOG_DUMP.
      
        - An IO thread never issues (and has never issued)
          COM_BINLOG_DUMP with the flag set;
      
        -  Mysqlbinlog will set the flag always, unless --stop-never is specified.
      
        - The dump thread falls back to non-blocking behavior if the
          connecting client states that it has server_id=0;
------------------------------------------------------------
revno: 8063
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-21 13:10:11 +0200
message:
  This change 
  
  1) removes the special run of innodb_gis from default.push (innodb_gis is a default suite, so it will run anyways)
  
  2) fixes default.daily so that the 4k and 8k runs of the innodb suite don't share the same vardir
  
  3) fixes default.weekly so that 4k and 8k runs of the innodb suite don't share the same vardir
  
  4) adds innodb_gis to the 4k run in default.weekly per Jimmy's request
  
  Approved by Anitha over e-mail.
------------------------------------------------------------
revno: 8062
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-21 13:31:29 +0300
message:
  Bug#18755095 REDO LOG SIZE CHANGE AFTER CRASH RESULTS IN CHECKPOINT AGE
  ERROR MESSAGE
  
  This is a regression from fixing
  Bug#18730524 REPEATED KILL+RESTART FAILS DUE TO MISSING MLOG_FILE_NAME RECORD
  
  innobase_start_or_create_for_mysql(): Invoke fil_names_clear() before
  creating the "checkpoint" when changing redo log files.
  
  Approved by Jimmy Yang on IM.
------------------------------------------------------------
revno: 8061
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: multi-update-bug
timestamp: Wed 2014-05-21 12:05:28 +0200
message:
  WL#6635 Make use of condition filtering in the optimizer
              
  Stabilizing test results: Added ANALYZE TABLE to two tables
  in innodb_multi_update.test to increase likelihood that the
  same query plan is produced on every run.
------------------------------------------------------------
revno: 8060 [merge]
committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-21 14:00:42 +0530
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.70
    committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-21 12:46:14 +0530
    message:
      Bug#11760197:INCORRECT RESULTS WITH COUNT DISTINCT AND BIG_TABLES
      Problem:
      When big_tables is enabled, COUNT DISTINCT on simple join with
      constant equality condition gives wrong result.
      
      Analysis:-
      In case of count distinct with sql_big_tables enable, optimizer
      saves data in myisam file instead of heap unique tree.
      When a constant equality condition is present, it does not
      detect duplicate field values when inserting the records into the
      temporary table.
      
      Original solution:
      While creating temporary table, allow creations of fields for constant
      items. When we have distinct query, this will make sure the duplicate
      value filtered out when inserting rows in the temporary table.
      
      Side Effects:
      Bug#17555462 - SELECT DISTINCT OF VIEW WITH CONSTANT STRING
                     RETURNS WRONG RESULT
      Problem:
      In a temporary table if a field is created for a const ref_item,
      and if this ref_item is a reference over a constant string and not
      on a field results are not as expected.
      
      Analysis:
      This is a regression from the patch for Bug#11760197. Post
      this bugfix, a field gets created in temporary table
      even for const_item's. If the const_item is of type Item_ref,
      while creating the temporary field, item->result_field
      is made to point to the newly created field.
      While the original item would hold the constant's value,
      the new field created while changing ref's to use temporary
      fields, will not be having the value.
      As a result we loose the value when fetching the results.
      
      In similar lines:
      Bug #17607155: ASSERTION FAILED: ((ITEM_RESULT_FIELD*)ITEM)->RESULT_FIELD
      Bug #17957913: ASSERTION `((ITEM_RESULT_FIELD*)ITEM)->RESULT_FIELD' FAILS
                     IN CREATE_TMP_FIELD
      Problem:
      Query having a GROUP BY CLAUSE with ROLLUP modifier and a
      GROUP_CONCAT/COUNT(DISTINCT) function with constant expression,
      causes an assert.
      
      Analysis:
      The GROUP_CONCAT/COUNT(DISTINCT) uses its own temporary table.
      When ROLLUP is present it creates the second copy of temporary
      table.
      This copy receives the same list of arguments that original
      group_concat/count(distinct) does which will help to copy
      the content of the result_field for the function under
      GROUP_CONCAT/COUNT from  the first temporary table to the second
      temporary table.
      
      In the case, when constant item is present, result_field will carry
      null value. After the fix for Bug#11760197, while creating field
      for second temporary table as result_field for the constant
      expression is not set, it asserts.
      
      Bug#17634335: SELECT DISTINCT...GROUP BY RETURNS WRONG RESULTS
      IN SOME CASES
      
      Query creating temporary table to find the distinct value and has
      constant value in projected list doesn't give correct result.
      
      Analysis:
      After the fix for Bug#11760197 const_items also were created as
      part of temporary tables. In the call to remove_duplicates() an
      assumption against the same was made which resulted in the above bug.
      
      All the above bugs were side effects of the fix made for Bug#11760197.
      
      Current solution:
      Distinct of a constant value will always be the constant value and
      count distinct of the same will always be one. Based on this,
      a new variable const_distinct is introduced. If enabled, temporary
      table is not created and aggregation is also avoided as the result
      will always be one.
      
      Works in similar lines to always_null.
------------------------------------------------------------
revno: 8059
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-sb4
timestamp: Wed 2014-05-21 09:54:23 +0200
message:
  Bug#14743180: Refactor pushdown_on_conditions()
  
  The current implementation of pushdown_on_conditions() is rather badly
  structured and it is possible to make it simpler and slightly more efficient.
  
   - Comments are added all over the place and use of local variables cleaned up.
  
   - Function pushdown_on_conditions() is renamed to
     JOIN::attach_join_conditions().
     The name is based on the fact that conditions are isolated and "attached"
     to the table where they can be evaluated in an optimal way.
  
   - pushdown_on_conditions() has a separate section that analyzes const
     conditions for outer join. This code is called for every table that
     is added to the plan and when called, it visits every table present
     in the plan. However, it is more efficient to call it only when a
     complete outer join operation has been formed and then attach the
     condition to the first inner table of the outer join.
  
   - pushdown_on_conditions() loops over all tables added to the plan,
     but it is only needed to loop over all inner tables of the plan.
     This is handled by setting the "prefix_tables" field for the first
     inner table to cover all tables up to and including this table.
  
   - The effect of these changes is also a simplification of some redundant
     table conditions.
  
   - make_join_select() has a first stage that looks at const join conditions.
     However, this is easily handled exclusively within
     JOIN::attach_join_conditions().
  
   - Function add_found_match_trig_cond() can be slightly rewritten and
     comments can be improved. Recursion is replaced with iteration (which
     causes slight change to optimizer trace).
     add_found_match_trig_cond() is called by pushdown_on_conditions().
  
   - The test file join_nested.test contains a few test cases that check
     nested outer join operations. However, these test cases use MySQL-
     specific syntax making it difficult to compare test results with
     other SQL databases. The patch adds duplicate test cases with
     standard compliant syntax.
  
  mysql-test/r/join_nested.result
  mysql-test/r/join_nested_bka.result
  mysql-test/r/join_nested_bka_nixbnl.result
    Test results for duplicate tests with standard compliant SQL syntax.
  
  mysql-test/r/subquery_sj_dupsweed.result
  mysql-test/r/subquery_sj_dupsweed_bka.result
  mysql-test/r/subquery_sj_dupsweed_bka_nixbnl.result
  mysql-test/r/subquery_sj_dupsweed_bkaunique.result
  mysql-test/r/subquery_sj_firstmatch.result
  mysql-test/r/subquery_sj_firstmatch_bka.result
  mysql-test/r/subquery_sj_firstmatch_bka_nixbnl.result
  mysql-test/r/subquery_sj_firstmatch_bkaunique.result
  mysql-test/r/subquery_sj_loosescan.result
  mysql-test/r/subquery_sj_loosescan_bka.result
  mysql-test/r/subquery_sj_loosescan_bka_nixbnl.result
  mysql-test/r/subquery_sj_loosescan_bkaunique.result
    Some simplified conditions.
  
  mysql-test/suite/opt_trace/r/general2_no_prot.result
  mysql-test/suite/opt_trace/r/general2_ps_prot.result
    Optimizer trace: Reversed wrapping order of triggered conditions.
  
  mysql-test/t/join_nested.test
    Added some duplicate nested join tests with standard SQL syntax.
  
  sql/sql_optimizer.cc
    Altered make_join_select(), attach_join_conditions() and
    add_found_match_trig_cond() according to above specification.
    quick_fix_fields() replaced with fix_fields() because hitting a problem
    where data is not correctly propagated using update_used_tables().
  
  sql/sql_optimizer.h
    Added attach_join_conditions() as member function in class JOIN.
  
  sql/sql_select.h
    Added comments.
  
  sql/table.h
    Added comments.
------------------------------------------------------------
revno: 8058
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: trunk
timestamp: Tue 2014-05-20 20:53:00 +0200
message:
  Don't run rpl_parallel_seconds_behind_master in valgrind, it times out.
------------------------------------------------------------
revno: 8057
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: multi-update-bug
timestamp: Tue 2014-05-20 15:25:22 +0200
message:
  WL#6635 Make use of condition filtering in the optimizer
        
  Stabilizing test results: remove an EXPLAIN that temporarily
  was added to one query in innodb_multi_update.test in order
  to understand what triggered the test failure.
------------------------------------------------------------
revno: 8056
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18163964ltrim
timestamp: Tue 2014-05-20 09:26:28 +0200
message:
  Bug#18315770 BUG#12368495 FIX IS INCOMPLETE
  
  Item_func_ltrim::val_str did not handle multibyte charsets.
  Fix: remove code duplication, i.e. remove Item_func_ltrim and Item_func_rtrim.
------------------------------------------------------------
revno: 8055 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: trunk
timestamp: Tue 2014-05-20 11:56:51 +0530
message:
  BUG#18432770 - SYS_VARS.LOG_SLOW_ADMIN_STATEMENTS_FUNC TEST
                 IS FAILING ON WINDOWS ON 5.6 
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.69
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-05-20 11:54:50 +0530
    message:
      BUG#18432770 - SYS_VARS.LOG_SLOW_ADMIN_STATEMENTS_FUNC TEST
                     IS FAILING ON WINDOWS ON 5.6 
      
      Analysis:
      ---------
      In test file "SYS_VARS.LOG_SLOW_ADMIN_STATEMENTS_FUNC.test",
      logging of slow admin statement such as ALTER, OPTIMIZE and ANALYZE
      operations is verified . When "long_query_time" is zero then
      admin statement which runs for  at least 1 microsecond is
      considered for logging. On some platforms (special on Windows)
      time taken for OPTIMIZE and ANALYZE table is less than 1 
      microsecond sometimes. So test is failed sporadically.
      
      Fix: 
      ----
      To make sure OPTIMIZE table operation take enough time
      so that it is logged, increased the number of rows in table.
      With the current changes also ANALYZE table returned within 1
      microsecond sometimes.Removed ANALYZE table operation from the
      test to avoid sporadic failure of this test case.
      CHECK table is another admin table which is logged to slow
      log table. Added CHECK table operation to test case.
      
      To test the changes, ran SYS_VARS.LOG_SLOW_ADMIN_STATEMENTS_FUNC
      for many times on windows machine.
------------------------------------------------------------
revno: 8054
committer: Krunal Bauskar <krunal.bauskar@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-20 07:29:21 +0200
message:
  - Bug #18795594 - TRX_CLEANUP_AT_DB_STARTUP SHOULD RESET
    TRX->RSEGS->M_REDO CONTENTS IN DEBUG CODE
  
    Ensure that half-cooked transaction content are reset back to default
    before freeing them and adding them to transaction pool.
  
    Approved by: Sunny (over IM)
------------------------------------------------------------
revno: 8053 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Mon 2014-05-19 22:07:49 +0530
message:
  Bug #17309863 AUTO RECONNECT DOES NOT WORK WITH 5.6 LIBMYSQLCLIENT
  
  Merging from mysql-5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.324.68
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-05-19 22:01:55 +0530
    message:
      Bug #17309863 AUTO RECONNECT DOES NOT WORK WITH 5.6 LIBMYSQLCLIENT
      
      Problem Statement: Automatic reconnection does not work for MySQL client
      programs linked with 5.6 libmysqlclient, even if MYSQL_OPT_RECONNECT is enabled.
      
      Analysis:
      When we have two connections (say con_1 and con_2) in which 'con_1' has
      auto-reconnect enabled. In such case if 'con_2' sends 'KILL <con_1 id>'
      (killing 'con_1'), then the server closes the socket for 'con_1'.
      After that when we send any query to 'con_1' it is failing with "Lost
      connection to MySQL server during query" error, even though auto-reconnect
      is enabled for 'con_1'.
      This is because send() which sends query might still succeed on client
      even though connection has been already closed on server. Since send()
      returns success at client side, client tries to recv() the data. Now
      client receives '0' means that the peer has closed the connection.
      Hence the query fails with the error mentioned above.
      
      Problem didn't exist in 5.5 and earlier versions because in them we tried
      to read-up all data remaining from previous command before sending new one
      to the server. As result we detected that connection was closed before
      query was sent and re-established connection.
      
      Fix:
      Check if socket is alive using vio_is_connected() call in case if
      auto-reconnect is enabled before sending query to server. If socket
      was disconnected by server set net->error to 2 so the socket on the
      client will be closed as well and reconnect will be initiated before
      sending the query. Reconnect doesn't make sense in case of COM_QUIT
      so skip the connection check for this command.
      
      Note: This fix doesn't solve the problem fully as bug still can
      occur if connection is killed exactly after this check and before
      query is sent. But this is acceptable since similar problem
      exists in 5.5.
      
      Also note that this patch might cause slight performance degradation,
      but it affects only auto-reconnect mode and therefore acceptable.
------------------------------------------------------------
revno: 8052
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-05-19 17:54:25 +0530
message:
  - WL#7682: Optimizing InnoDB temporary table.
    - Fixed wrong/invalid passing of argument while dropping intrinsic table.
    - Value of intrinsic table handler was being assigned to non_atomic variable.
  
   Approved by: Jimmy (over IM)
------------------------------------------------------------
revno: 8051
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk-bug18107853
timestamp: Mon 2014-05-19 18:14:23 +0700
message:
  This is a patch for bug#18107853 - XA LIST GETS CORRUPT, CRASH AND/OR HANG
  AND/OR ASSERTION
  
  From user point of view an XA command that got XID value with too long
  format ID part could cause crash or hang in the presence of concurrent
  running XA commands.
  
  The original test case attached to the bug report demonstrates problem
  on 64-bit Windows. It assigns too long hex value to a format ID part of
  XA xid, which leads to its overflow. This overflow condition wasn't
  handled correctly within a parser that resulted in assignment of the
  value -1 to the data member XID::formatID.
  
  The reason for getting -1 from too long hex string value is that
  my_strtoll() function, which is used to convert value from hex format,
  returns the special value LLONG_MAX (0x7fffffffffffffff) in this case.
  This 64-bit value later is casted to 32-bit unsigned long that leads to
  discarding highest significant bits and in result the formatID gets value
  0xffffffff (i.e. -1). Similar effect can be achieved on all platforms by
  trying to assign decimal 18446744073709551615 to a format ID part of XA xid.
  
  The xid value with format ID equal to -1 is interpreted by the implementation
  of XA as a xid whose value is NULL. Further usage of such XID value leads to
  error condition during operations on transaction_cache. This error condition
  is expressed in the fact that the NULL xid value stored in the
  transaction_cache can be modified later during executing of DML statement
  within the same transaction. That is, if an user executes
                XA START 'xid1', 1, 0xfffffffffffffffffff;
                INSERT INTO some_table VALUES (some_value);
  then during execution of INSERT statement the method XID_STATE::set_query_id
  is called when an InnoDB transaction is registered  with the MySQL 2PC
  coordinator and since the value of XID::formatID has value -1 as result of
  overflow then existing xid value is considered as NULL xid and new xid value
  is assigned. As result xid value which is already stored in the
  transaction_cache is modified without recalculation of its hash key.
  Modification of a value stored in the cache without a key recalculation could
  lead to a crash or hang in case when several threads running XA-related
  transaction are executed in parallel and all of them uses too long value that
  overflow the values of formatID. Since it is difficult to create a reproducible
  test that can lead to the crash this patch only provides test case that
  prohibits execution of XA-transaction with value of formatID that resulted in
  formatID equal to -1 in the past.
  
  Full-blown checking of xid's formatID value requires significant changes
  to source code and therefore has to be done in the scope of separate
  bug/worklog. This patch takes simplistic approach to the problem and ensures
  that formatID value passed from parser never gets negative value.
------------------------------------------------------------
revno: 8050
committer: Sergey Glukhov <sergey.glukhov@oracle.com>
branch nick: mysql-trunk-new
timestamp: Mon 2014-05-19 12:50:55 +0400
message:
  WL#7123 Additional query optimization for Fulltext Search
------------------------------------------------------------
revno: 8049
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Mon 2014-05-19 10:04:07 +0200
message:
  Bug#18694751 72512: Non-aggregated query with set function in ORDER BY should be rejected
  
  A non-aggregated query with an ORDER BY applied to it cannot contain
  aggregate functions. This patch prevents such queries.
  
  A new error message is added: ER_AGGREGATE_ORDER_NON_AGG_QUERY.
  
  Notice that one test case with HAVING and an aggregated argument fails but
  should not. This is because of a wrong resolve order in
  SELECT_LEX::prepare().
------------------------------------------------------------
revno: 8048
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-sb5
timestamp: Mon 2014-05-19 09:14:03 +0200
message:
  Bug#17748273: Failure in get_sort_by_table
  
  The problem occurs when Item_field::fix_fields() resolves a column
  reference from a list (find_item_in_list), and the field is an outer
  reference. In this case, the depended_from field was not set, and the
  used_tables property of the field was later calculated wrongly.
  
  The solution is to set the depended_from field to the same value as
  the item found in the list, by using the function mark_as_dependent().
------------------------------------------------------------
revno: 8047
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-19 08:48:22 +0200
message:
  Move 42 tests to "no_valgrind_without_big" to reduce PB2 per-push running time.
  
  Approved by Anitha over IM.
------------------------------------------------------------
revno: 8046
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-05-19 08:53:00 +0530
message:
  - WL#7682: Optimizing InnoDB temporary table.
    - Fixed a complier warning to remove uninitialized variable.
  
   Approved by: Sunny (over IM)
------------------------------------------------------------
revno: 8045
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Sat 2014-05-17 00:16:08 -0500
message:
  Fix recent Windows compiler warnings with static_cast<>()
------------------------------------------------------------
revno: 8044 [merge]
committer: Raghav Kapoor <raghav.kapoor@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-16 16:34:11 +0530
message:
  BUG#11756690 - ERROR 22012 DIVISION BY 0 HANDLED BY HANDLER FOR SQLSTATE 23000 (ER_DUP_ENTRY)
  BUG#11744940 - TRADITIONAL MODE: CAST ALLOWS INVALID VALUE
  BUG#11744941 - TRADITIONAL MODE: BIGINT RANGE NOT CORRECTLY DELIMITED
  
  BACKGROUND:
  The above bugs are not reproducible in current trunk.
  
  ANALYSIS:
  BUG#11756690:
  This bug is reproducible in 5.1, but not reproducible in 5.5,
  5.6 and trunk versions.
  It is verified that Due to refactoring that was done by 
  Worklog WL#5986 Make condition handlers closer to
  the standard, this bug has gone.
  Test Coverage of this bug is included in the Patch submitted.
  
  BUG#11744940 and BUG#11744941:
  The above two bugs were due to faulty arithmetic that was 
  done in 5.0.
  These bugs have been fixed and are not reproducible in 
  5.1+ versions.
  It has been verified by running all the scenarios mentioned 
  in the two bug reports.
  Test coverage of these two bugs is also included in the
  patch submitted.
  
  FIX:
  Since the above bugs are not reproducible, test coverage 
  is submitted as patch.
    ------------------------------------------------------------
    revno: 6658.1.1
    committer: Raghav Kapoor <raghav.kapoor@oracle.com>
    branch nick: mysql-trunk-strict
    timestamp: Wed 2013-10-09 16:21:15 +0530
    message:
      BUG#11756690 - ERROR 22012 DIVISION BY 0 HANDLED BY HANDLER FOR SQLSTATE 23000 (ER_DUP_ENTRY)
      BUG#11744940 - TRADITIONAL MODE: CAST ALLOWS INVALID VALUE
      BUG#11744941 - TRADITIONAL MODE: BIGINT RANGE NOT CORRECTLY DELIMITED
      
      BACKGROUND:
      The above bugs are not reproducible in current trunk.
      
      ANALYSIS:
      BUG#11756690:
      This bug is reproducible in 5.1, but not reproducible in 5.5,
      5.6 and trunk versions.
      It is verified that Due to refactoring that was done by 
      Worklog WL#5986 Make condition handlers closer to the 
      standard, this bug has gone.
      Test Coverage of this bug is included in the Patch submitted.
      
      BUG#11744940 and BUG#11744941:
      The above two bugs were due to faulty arithmetic that was 
      done in 5.0.
      These bugs have been fixed and are not reproducible in 
      5.1+ versions.
      It has been verified by running all the scenarios mentioned 
      in the two bug reports.
      Test coverage of these two bugs is also included in the
      patch submitted.
      
      FIX:
      Since the above bugs are not reproducible, test coverage 
      is submitted as patch.
------------------------------------------------------------
revno: 8043 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Fri 2014-05-16 11:04:37 +0200
message:
  NULL merge 5.6 => trunk (separate patch to be written)
    ------------------------------------------------------------
    revno: 3902.324.67 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Fri 2014-05-16 11:00:39 +0200
    message:
      merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.595.3
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5-review
        timestamp: Fri 2014-05-16 10:18:43 +0200
        message:
          Bug#18315770 BUG#12368495 FIX IS INCOMPLETE
          
          Item_func_ltrim::val_str did not handle multibyte charsets.
          Fix: factor out common code for Item_func_trim and Item_func_ltrim.
------------------------------------------------------------
revno: 8042
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: multi-update-bug
timestamp: Fri 2014-05-16 11:02:00 +0200
message:
  WL#6635 Make use of condition filtering in the optimizer
  
  Stabilizing test results: adding EXPLAIN to one query in 
  innodb_multi_update.test in order to understand what triggers
  the test failure.
------------------------------------------------------------
revno: 8041
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-16 13:58:24 +0530
message:
  Follow up patch to fix a test case.
------------------------------------------------------------
revno: 8040 [merge]
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: trunk
timestamp: Fri 2014-05-16 10:18:46 +0200
message:
  WL#7854 Deprecate and remove mysqlhotcopy
    ------------------------------------------------------------
    revno: 3902.324.66
    committer: Erlend Dahl <erlend.dahl@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-05-16 10:13:12 +0200
    message:
      WL#7854 Deprecate and remove mysqlhotcopy
------------------------------------------------------------
revno: 8039 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Fri 2014-05-16 14:22:23 +0800
message:
  Merge mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.65
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-05-16 14:19:52 +0800
    message:
      Fix a regex expression for innodb-wl5980-alter.test. Needed for bug #18635485
      testing change
------------------------------------------------------------
revno: 8038 [merge]
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-16 09:52:07 +0530
message:
  Merging from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.64 [merge]
    committer: Arun Kuruvila <arun.kuruvila@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-05-16 09:24:08 +0530
    message:
      Merging from mysql-5.5 to Mysql-5.6
      
      Fix: The new password also masked in the similar manner
      that of the --password argument. Warning message is also
      shown while giving new password via command line.
        ------------------------------------------------------------
        revno: 2875.595.2
        committer: Arun Kuruvila <arun.kuruvila@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-05-16 09:16:39 +0530
        message:
          Bug #18163964 PASSWORD IS VISIBLE WHILE CHANGING IT FROM
                        MYSQLADMIN IN PROCESSES LIST
          
          Description: Checking the process status (with ps -ef)  
          while executing "mysqladmin" with old password and new 
          password via command-line will show the new password in the
          process list sporadically.
          
          Analysis: The old password is being masked by "mysqladmin".
          So masking the new password in the similar manner would 
          reduce hitting the bug. But this would not completely fix
          the bug, because if "ps -ef " command hits the mysqladmin
          before it masks the passwords it will show both the old and
          new passwords in the process list. But the chances of
          hitting this is very less.
          
          Fix: The new password also masked in the similar manner
          that of the --password argument.
------------------------------------------------------------
revno: 8037
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Fri 2014-05-16 11:44:27 +0800
message:
  Fix Bug #18743725 - SPORADIC VALGRIND FAILURE IN 
  INNODB_GIS_ALTER_SPATIAL_INDEX 
  
  
  rb://5413 approved by Sunny Bains
------------------------------------------------------------
revno: 8036 [merge]
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: trunk2
timestamp: Fri 2014-05-16 10:59:52 +0800
message:
  Merge mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.63
    committer: Jimmy Yang <jimmy.yang@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-05-16 10:21:05 +0800
    message:
      Fix Bug #18635485 - INNODB_FTS: FTS INDEX TABLE SHOULD HAVE THEIR OWN SPACE ID 
            
      rb://5242 approved by Kevin Lewis
------------------------------------------------------------
revno: 8035
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-15 15:07:12 +0300
message:
  Fix Bug#18253089 BUF_POOL->FLUSH_RBT IS CREATED WHEN RECOVERY IS NOT NEEDED,
  THEN NEVER FREED
  
  * recv_recovery_from_checkpoint_start() is only called from
    innobase_start_or_create_for_mysql() and
    recv_recovery_from_checkpoint_start() calls recv_sys_create()&recv_sys_init()
    but those two functions have already been called earlier in
    innobase_start_or_create_for_mysql(). Those two functions return immediately
    if recv_sys is created/initialized. From this follows that their invokations
    from recv_recovery_from_checkpoint_start() are noops and thus I am removing
    them.
  
  * The code used to initialize buf_pool->flush_rbt (buf_flush_init_flush_rbt())
    from recv_sys_init(). Change this so that it is initialized directly from
    recv_recovery_from_checkpoint_start(). This is the only functional change
    in this patch - buf_pool->flush_rbt will now be initialized later in the
    startup code path.
  
  * The code used to free buf_pool->flush_rbt (buf_flush_free_flush_rbt()) from
    recv_sys_debug_free(). Change this so that the free is done from
    recv_recovery_from_checkpoint_finish(), which is the only caller of
    recv_sys_debug_free(). This is a noop, but made for consistency wrt where
    alloc/free is done - call
    buf_flush_init_flush_rbt() from recv_recovery_from_checkpoint_start() and
    buf_flush_free_flush_rbt() from recv_recovery_from_checkpoint_finish().
  
  This way the code will be restored as of before
  annamalai.gurusami@oracle.com-20140210082850-vqlzadis1asdws0e (the fix of
  Bug#18144349 INNODB CANNOT USE THE DOUBLEWRITE BUFFER FOR THE FIRST PAGE OF
  SYSTEM TABLESPACE) wrt buf_pool->flush_rbt initialization.
  
  Approved by:	Yasufumi (rb:5409)
------------------------------------------------------------
revno: 8034
committer: Aditya A <aditya.a@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-15 17:05:07 +0530
message:
  Bug #16874873 "ROW SIZE TOO LARGE" ERROR THROWN 
  	      EVEN WHEN IT'S NOT
  PROBLEM
  -------
  
  User is creating a table with many TEXT/BLOB fields 
  and getting error which says that the row size 
  is too large.
  
  FIX
  ---
  This behavior is expected because inoodb design
  requires the record to fit in the B-tree leaf 
  page and if unsuccessful we return error.During
  the creation of BLOB or TEXT field we allocate
  41 Bytes (dynamic or compressed row format) for
  the field which is considered for row size 
  calculation .During insertion if the BLOB or
  TEXT is less than 40 Bytes  we store it internally 
  with in these 41 bytes and if it is more than 40 
  bytes,we store it off page and store 20 Bytes 
  reference to the page.We have to update the 
  documentation with this information . To give 
  more information to the user we are adding extra 
  information in the error log about the maximum
  record size allowed,the present record size and
  till what field we can accommodate so that the
  record fits in the page. 
  
  Approved by Kevin [#rb5384]
------------------------------------------------------------
revno: 8033 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-05-15 19:05:17 +0800
message:
  BUG 18277305 - FTS: FAILING ASSERTION: PTR[1] == '\"' IN
  FTS_AST_CREATE_NODE_TEXT
  
  For collation ujis_japanese_ci, utf8mb4_turkish_ci, eucjpms_bin, if we have
  a 0x00 in the fts query string, it might hit an assertion and server crash.
  
  We shouldn't truncate the trailing characters after the first 0x00 in query
  strings. Instead of using strdup to copy the query string between quotes, we
  save the whole string and its length during parsing. After that, the whole
  query string would be parsed again when we do query. And the current parsing
  during query would treat 0x00 as a separator, like myisam.
  
  Also, a memory leak issue is fixed. After we find a syntax error and try to
  abort, we have to free the last token first.
  
  rb#5152, approved by Jimmy
    ------------------------------------------------------------
    revno: 3902.324.62
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-15 19:04:21 +0800
    message:
      BUG 18277305 - FTS: FAILING ASSERTION: PTR[1] == '\"' IN
      FTS_AST_CREATE_NODE_TEXT
      
      For collation ujis_japanese_ci, utf8mb4_turkish_ci, eucjpms_bin, if we have
      a 0x00 in the fts query string, it might hit an assertion and server crash.
      
      We shouldn't truncate the trailing characters after the first 0x00 in query
      strings. Instead of using strdup to copy the query string between quotes, we
      save the whole string and its length during parsing. After that, the whole
      query string would be parsed again when we do query. And the current parsing
      during query would treat 0x00 as a separator, like myisam.
      
      Also, a memory leak issue is fixed. After we find a syntax error and try to
      abort, we have to free the last token first.
      
      rb#5152, approved by Jimmy
------------------------------------------------------------
revno: 8032 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Thu 2014-05-15 15:54:22 +0530
message:
  Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
  		(null merge)
    ------------------------------------------------------------
    revno: 3902.324.61 [merge]
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-05-15 15:52:58 +0530
    message:
      Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
      		(updated patch for windows failure.)
        ------------------------------------------------------------
        revno: 2875.595.1
        committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-05-15 15:50:52 +0530
        message:
          Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
          
          Problem:
          Load_log_event::print_query() function does not put escape character in file name 
          for "LOAD DATA INFILE" statement.
          
          Analysis:
          When we have "'" in our file name for "LOAD DATA INFILE" statement,
          Load_log_event::print_query() function does not put escape character 
          in our file name.
          
          This one result that when we show binary-log, we get file name without 
          escape character.
          
          Solution:
          To put escape character when we have "'" in file name, for this instead of using 
          simple memcpy() to put file-name, we will use pretty_print_str().
------------------------------------------------------------
revno: 8031 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Thu 2014-05-15 15:37:22 +0530
message:
  Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
  		null merge
    ------------------------------------------------------------
    revno: 3902.327.1 [merge]
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-05-15 15:35:45 +0530
    message:
      Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
      	(updated patch- change taking care of windows failure.)
        ------------------------------------------------------------
        revno: 2875.594.1
        committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-05-15 15:34:24 +0530
        message:
          Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
          
          Problem:
          Load_log_event::print_query() function does not put escape character in file name 
          for "LOAD DATA INFILE" statement.
          
          Analysis:
          When we have "'" in our file name for "LOAD DATA INFILE" statement,
          Load_log_event::print_query() function does not put escape character 
          in our file name.
          
          This one result that when we show binary-log, we get file name without 
          escape character.
          
          Solution:
          To put escape character when we have "'" in file name, for this instead of using 
          simple memcpy() to put file-name, we will use pretty_print_str().
------------------------------------------------------------
revno: 8030 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Thu 2014-05-15 15:30:51 +0530
message:
  Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
  		(updated patch- taking care of failure in trunk)
    ------------------------------------------------------------
    revno: 3902.326.1 [merge]
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-05-15 15:19:50 +0530
    message:
      Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
      	(updated patch- change taking care of windows failure.)
        ------------------------------------------------------------
        revno: 2875.592.41
        committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-05-15 15:13:06 +0530
        message:
          Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
          
          Problem:
          Load_log_event::print_query() function does not put escape character in file name 
          for "LOAD DATA INFILE" statement.
          
          Analysis:
          When we have "'" in our file name for "LOAD DATA INFILE" statement,
          Load_log_event::print_query() function does not put escape character 
          in our file name.
          
          This one result that when we show binary-log, we get file name without 
          escape character.
          
          Solution:
          To put escape character when we have "'" in file name, for this instead of using 
          simple memcpy() to put file-name, we will use pretty_print_str().
------------------------------------------------------------
revno: 8029 [merge]
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Thu 2014-05-15 17:56:07 +0800
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.60
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-5.6-bugfix1
    timestamp: Thu 2014-05-15 17:53:33 +0800
    message:
      Followup fix of BUG#18079671 - INNODB FTS : ASSERT IN FTS_CACHE_APPEND_DELETED_DOC_IDS
        
      Approved by Jimmy.Yang in IM
------------------------------------------------------------
revno: 8028
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Thu 2014-05-15 17:36:38 +0800
message:
  Followup fix of BUG#18079671 - INNODB FTS : ASSERT IN FTS_CACHE_APPEND_DELETED_DOC_IDS
  
  Approved by Jimmy.Yang in IM.
------------------------------------------------------------
revno: 8027
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Thu 2014-05-15 16:04:31 +0800
message:
  BUG#18593561: INNODB TRIES TO MERGE B-TREE CHANGE BUFFER FOR UNDO TABLESPACE
  
  Analysis:
  The root cause is that we call buf_mark_space_corrupt when we simulate a page
  corruption. It returns false, but reduces n_pend_reads when the page is in undo
  tablespace, so we get assert failure.
  
  Solution:
  1. Skip undo tablespace when simulating page corruption;
  2. Avoid ibuf_merge_or_delete_for_page for undo tablespace in buf_page_io_complete.
  
  rb://5390 approved by Marko.
------------------------------------------------------------
revno: 8026 [merge]
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-15 11:53:11 +0530
message:
  Bug#17217128 : BAD INTERACTION BETWEEN MIN/MAX AND
                 "HAVING SUM(DISTINCT)": WRONG RESULTS.
  Merge from Mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.59 [merge]
    committer: mithun <mithun.c.y@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-15 11:51:13 +0530
    message:
      Bug#17217128 : BAD INTERACTION BETWEEN MIN/MAX AND
                     "HAVING SUM(DISTINCT)": WRONG RESULTS.
      
      Merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.40
        committer: mithun <mithun.c.y@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-05-15 11:46:57 +0530
        message:
          Bug#17217128 : BAD INTERACTION BETWEEN MIN/MAX AND
                         "HAVING SUM(DISTINCT)": WRONG RESULTS.
          ISSUE:
          ------
          If a query uses loose index scan and it has both
          AGG(DISTINCT) and MIN()/MAX()functions. Then, result values
          of MIN/MAX() is set improperly.
          When query has AGG(DISTINCT) then end_select is set to
          end_send_group. "end_send_group" keeps doing aggregation
          until it sees a record from next group. And, then it will
          send out the result row of that group.
          Since query also has MIN()/MAX() and loose index scan is
          used, values of MIN/MAX() are set as part of loose index
          scan itself. Setting MIN()/MAX() values as part of loose
          index scan overwrites values computed in end_send_group.
          This caused invalid result.
          For such queries to work loose index scan should stop
          performing MIN/MAX() aggregation. And, let end_send_group to
          do the same. But according to current design loose index
          scan can produce only one row per group key. If we have both
          MIN() and MAX() then it has to give two records out. This is
          not possible as interface has to use common buffer
          record[0]! for both records at a time.
          
          SOLUTIONS:
          ----------
          For such queries to work we need a new interface for loose
          index scan. Hence, do not choose loose_index_scan for such
          cases. So a new rule SA7 is introduced to take care of the
          same.
          
          SA7: "If Q has both AGG_FUNC(DISTINCT ...) and
                MIN/MAX() functions then loose index scan access
                method is not used."
------------------------------------------------------------
revno: 8025
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 18:05:24 +0530
message:
  - WL#7682: Optimizing InnoDB temporary table.
    - Follow-up patch to fix valgrind issue.
------------------------------------------------------------
revno: 8024
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B12368203-trunk
timestamp: Tue 2014-05-13 18:42:21 +0300
message:
  Bug#12368203 - 60878: mysqladmin doesn't support flushing
    specific logs
  
  mysqladmin currently supports a command 'flush-logs'
  to flush all log types. However, mysqladmin offered
  no support to flush specific logs.
  
  Added support for additional arguments to 
  mysqladmin fllush-logs. The syntax is : 
  
  flush-logs [specific specific ...]
  specific = binary
                  | engine
                  | error
                  | general
                  | relay
                  | slow
  
  Any other "specific" is interpretted as the next mysqladmin
  command.
  
  This results in COM_QUERY (vs. COM_REFRESH that
  was used before, but found lacking in functionality).
  
  Test case added.
------------------------------------------------------------
revno: 8023
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B17457755-trunk
timestamp: Thu 2014-04-17 18:04:54 +0300
message:
  Bug #17457755: MYISAM FTS: CRASH IN FT_SIMPLE_GET_WORD
  
  The UPDATE command was trying to prepare (by calling 
  setup_ftfuncs() ) the full text index search functions before 
  calling fix_fields() on the update values.
  This obviously is not a good idea, specially if the fulltext 
  search is part of the update value.
  
  Moved the call to setup_ftfuncs() after the preparation
  of the UPDATE values. 
  
  Test case added.
------------------------------------------------------------
revno: 8022 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Wed 2014-05-14 16:04:40 +0530
message:
  Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
  		(merge from 5.6 to trunk)
    ------------------------------------------------------------
    revno: 3902.324.58
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Wed 2014-05-14 16:00:44 +0530
    message:
      Bug#18207212 : FILE NAME IS NOT ESCAPED IN BINLOG FOR LOAD DATA INFILE STATEMENT
      
      Problem:
      Load_log_event::print_query() function does not put escape character in file name 
      for "LOAD DATA INFILE" statement.
      
      Analysis:
      When we have "'" in our file name for "LOAD DATA INFILE" statement,
      Load_log_event::print_query() function does not put escape character 
      in our file name.
      
      This one result that when we show binary-log, we get file name without 
      escape character.
      
      Solution:
      To put escape character when we have "'" in file name, for this instead of using 
      simple memcpy() to put file-name, we will use pretty_print_str().
------------------------------------------------------------
revno: 8021
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 13:23:19 +0300
message:
  Addendum to BUG#16272328
  
  Fixed various tests that depended on the presence of --debug-info and its help text
------------------------------------------------------------
revno: 8020
committer: viswanatham gudipati <viswanatham.gudipati@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 14:29:30 +0530
message:
  incorporated comments from rb#4822
------------------------------------------------------------
revno: 8019 [merge]
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 15:18:51 +0800
message:
  Fix Bug#18409840 MEMCACHED_SHUTDOWN IS CALLING "PLUGIN_DEL"
  WITHOUT ACQUIRING LOCK_PLUGIN MUTEX.
  
  NULL merge from 5.6.
    ------------------------------------------------------------
    revno: 3902.324.57
    committer: Allen lai <zheng.lai@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-14 15:17:39 +0800
    message:
      Fix Bug#18409840 MEMCACHED_SHUTDOWN IS CALLING "PLUGIN_DEL"
      WITHOUT ACQUIRING LOCK_PLUGIN MUTEX.
      
      We also fixed a race condition bug in ib_cursor_delete_row in this
      patch.
      
      rb://5359 approved by Jimmy.
------------------------------------------------------------
revno: 8018 [merge]
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 15:07:35 +0800
message:
  Fix Bug#18409840 MEMCACHED_SHUTDOWN IS CALLING "PLUGIN_DEL"
  WITHOUT ACQUIRING LOCK_PLUGIN MUTEX.
  
  Merge from mysql-5.6.
    ------------------------------------------------------------
    revno: 3902.325.1
    committer: Allen lai <zheng.lai@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-14 14:52:13 +0800
    message:
      Fix Bug#18409840 MEMCACHED_SHUTDOWN IS CALLING "PLUGIN_DEL"
      WITHOUT ACQUIRING LOCK_PLUGIN MUTEX.
      
      We also fixed a race condition bug in ib_cursor_delete_row in this
      patch.
      
      rb://5359 approved by Jimmy.
------------------------------------------------------------
revno: 8017 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug11758766_mysql-trunk
timestamp: Wed 2014-05-14 12:14:24 +0530
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.56
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug11758766_mysql-5.6
    timestamp: Wed 2014-05-14 12:01:43 +0530
    message:
      Bug#11758766:MYSQLD CONTINUES OPERATION WITHOUT LOGGING WHEN
      BINLOGS CANNOT BE WRITTEN
      
      Problem:
      ========
      If an error occurs that prevents mysqld writing to the
      binary logs (disk full, readonly file system, etc) then the
      logs are disabled and operations continue. This can lead to
      out of sync slaves and improper backups.
      
      Analysis:
      =========
      When binlogging becomes impossible due to readonly file
      system at present the binlog gets closed along with an error
      message in error log which says binlog is turned off for the
      whole duration of the server to fix the cause shutdown the
      server and fix the problem. The master continues without
      binlogging which causes the slave to go out of sync. At
      present there is no way user can control the behaviour so it
      will be better to let the user provide an option to stop the
      server immediately on the error.
      
      Fix:
      ====
      A new option "binlogging_impossible_mode" has been
      introduced whose values are "IGNORE" or "ABORT". When
      binlogging becomes impossible if user sets the variable to
      "ABORT" server will stop if user sets it to "IGNORE" binlog
      will be turned off and server will continue.
------------------------------------------------------------
revno: 8016
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 11:01:50 +0530
message:
  - WL#7682: Optimizing InnoDB temporary table.
    - Follow-up patch to fix valgrind issue.
------------------------------------------------------------
revno: 8015 [merge]
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk-for-mts
timestamp: Wed 2014-05-14 10:44:57 +0530
message:
  - WL#7682: Optimizing InnoDB temporary tables.
  
    There are specific use-cases where-in modules needs a temporary table for
    stagging data for a while and doesn't really care about transaction semantics.
  
    Given the use-case we have designed a special type of temporary table that
    are light weight (no-undo, no-transactional semantics, no-locking,
    no-doublewrite-buffer, operational even in read-only, etc....) and are
    optimized for performance.
  
    These tables will come into existence for a while
    (for example: during query exeuction) for stagging a data of subquery
    and so shortlived. Being temporary table they would continue to reside
    in shared temporary tablespace.
  
    Approved by: rb#4870 (Sunny + BinSu)
    ------------------------------------------------------------
    revno: 7583.1.143 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-05-14 10:28:55 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.142 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-05-14 09:34:38 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.141 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-13 16:09:48 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.140 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-13 14:37:36 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.139 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-13 12:01:19 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.138
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-13 11:07:19 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Removed redundant query from test-file.
    ------------------------------------------------------------
    revno: 7583.1.137 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-13 07:34:13 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.136 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-05-12 08:42:34 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.135 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-05-09 14:34:14 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.134 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-05-09 10:08:57 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.133 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-05-08 15:38:02 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.132
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-05-08 15:36:45 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Turned safety condition to debug assert.
    ------------------------------------------------------------
    revno: 7583.1.131
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-05-08 12:27:07 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Further optimized to avoid passing extra arguments.
    ------------------------------------------------------------
    revno: 7583.1.130
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-05-08 11:06:47 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Turned off redundant calculation.
    ------------------------------------------------------------
    revno: 7583.1.129
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-05-08 11:01:48 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Directly invoke appropriate version of row_search_for_xxxx
          This helps in saving some computation that otherwise seems to cost
          on performance.
    ------------------------------------------------------------
    revno: 7583.1.128
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-05-08 10:18:44 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Inline row_search_for_mysql that just takes decision of which version
          of sub-function to call (no_mvcc or mvcc)
    ------------------------------------------------------------
    revno: 7583.1.127
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-05-07 14:09:15 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Removed redundant initialization to fix performance drop.
    ------------------------------------------------------------
    revno: 7583.1.126 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-06 08:12:46 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.125
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-05-06 08:11:33 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed review comments.
    ------------------------------------------------------------
    revno: 7583.1.124
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-05-05 16:20:09 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed test-case that use to show different result when table
          hit full-case for page-size = 4K
    ------------------------------------------------------------
    revno: 7583.1.123 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-05-05 07:58:21 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.122 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-05-02 11:30:52 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.121
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-05-02 11:30:01 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Addressed review comments from Bin.
    ------------------------------------------------------------
    revno: 7583.1.120 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-05-02 08:58:34 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.119 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-30 13:40:06 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.118
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-30 13:31:48 +0530
    message:
      - WL#7682: removed dead variables and param
    ------------------------------------------------------------
    revno: 7583.1.117
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-30 13:27:21 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Realigned dict_index_t to use optimized packing.
    ------------------------------------------------------------
    revno: 7583.1.116
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-30 09:48:40 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Further review comments follow-up.
        - Moved the intrinsic table specific variable at table level to session level.
    ------------------------------------------------------------
    revno: 7583.1.115 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-29 15:56:11 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.114
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-29 15:27:49 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Another set of review comments (Part-2).
    ------------------------------------------------------------
    revno: 7583.1.113
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-29 13:48:04 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Another set of review comments (Part-1).
    ------------------------------------------------------------
    revno: 7583.1.112 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-29 07:51:49 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.111
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-04-28 17:05:21 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If table has auto-increment column and there are multiple insert
          operating on it parallely then predicting the value of auto-increment
          column is bit difficult. This is solved by taking an AUTOINC lock
          on table that is valid for statement lifetime.
        - As intrinsic table are not shared there is no need to take such lock.
    ------------------------------------------------------------
    revno: 7583.1.110
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-04-28 13:54:56 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Added test-scenario to cover case where-in server is started
          with auto-extension of temporary tablespace turned-off.
    ------------------------------------------------------------
    revno: 7583.1.109
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-04-28 12:15:57 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Added condition to handle tablespace full error during update.
    ------------------------------------------------------------
    revno: 7583.1.108
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-04-28 09:57:44 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If pessmistic insert fails then release the cached last insert state
          and commit mtr.
    ------------------------------------------------------------
    revno: 7583.1.107 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-04-28 09:31:37 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.106
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-25 12:56:04 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - In read-only mode now that we create temporary tablespace ensure
          that is is removed on clean shutdown.
    ------------------------------------------------------------
    revno: 7583.1.105
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-25 11:47:04 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If temporary table is created under LOCK table statement
          then external_lock is not called instead start_stmt is called.
          Hook to block ALTER table was only present in external_lock.
          Replicated it in start_stmt too.
    ------------------------------------------------------------
    revno: 7583.1.104 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-25 08:22:31 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.103
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 16:34:04 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected param qualifier
    ------------------------------------------------------------
    revno: 7583.1.102
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 16:30:28 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected param qualifier
    ------------------------------------------------------------
    revno: 7583.1.101 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 16:25:48 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.100
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 16:24:57 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Address few more review comments and probably last for current
          round.
    ------------------------------------------------------------
    revno: 7583.1.99
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 15:21:27 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Address another round of review comments. Remove goto from SELECT
          path of intrinsic table.
    ------------------------------------------------------------
    revno: 7583.1.98
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 14:29:38 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Address another round of review comments.
    ------------------------------------------------------------
    revno: 7583.1.97 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 14:17:12 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.96
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 14:16:27 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Address another round of review comments.
    ------------------------------------------------------------
    revno: 7583.1.95
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 13:11:59 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Address another round of review comments.
    ------------------------------------------------------------
    revno: 7583.1.94
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 12:05:41 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Address review comments. Mainly cosmetic. 
    ------------------------------------------------------------
    revno: 7583.1.93 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 10:40:32 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.92 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-24 10:38:57 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.91 [merge]
    committer: Krunal Bauskar <krunal.bauskar@oracle.com>
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-23 06:14:20 +0200
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.90
    committer: Krunal Bauskar <krunal.bauskar@oracle.com>
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-23 05:53:53 +0200
    message:
      - WL#7682: Optimizing InnoDB Temporary Tables.
        - Addressed review comments.
    ------------------------------------------------------------
    revno: 7583.1.89 [merge]
    committer: Krunal Bauskar <krunal.bauskar@oracle.com>
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-22 06:24:44 +0200
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.88 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-04-21 12:13:08 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.87 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-16 10:37:57 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.86
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-11 18:18:22 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.85
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-11 18:16:58 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - Corrected shutdown state to check while forcing buffer to flush.
    ------------------------------------------------------------
    revno: 7583.1.84 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-11 15:27:21 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.83
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-11 12:19:46 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - During shutdown don't demand even if buf-fix-count > 0 continue to flush it.
          This is safe given that it is for temporary tablespace only (where-in we
          can avoid flush completely on shutdown) + user threads are already done
          performing the action.
    ------------------------------------------------------------
    revno: 7583.1.82 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-10 12:54:09 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.81 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-10 10:38:49 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.80 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-10 10:31:54 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.79 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-09 15:19:04 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.78
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-09 14:21:34 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - Make sure you acquire block_mutex before changing fix_count
          as it is being used to synchronized flush thread and user thread.
    ------------------------------------------------------------
    revno: 7583.1.77 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-09 08:25:25 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.76 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-04-04 13:59:22 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.75
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-03 17:26:45 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - Formatting changes.
    ------------------------------------------------------------
    revno: 7583.1.74 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-03 14:57:49 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.73 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-03 14:33:48 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.72
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-04-03 09:37:01 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - Blocked creation of foreign key for temporary table and in turn relax
          check for dict_sys mutex and read only check.
    ------------------------------------------------------------
    revno: 7583.1.71
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-02 16:13:56 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - Added check to ensure there is error if there is attempt to create
          table with duplicate column name.
    ------------------------------------------------------------
    revno: 7583.1.70 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-04-02 08:28:06 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.69 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-01 19:58:16 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.68 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-01 19:55:31 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.67 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-04-01 08:10:21 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.66
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-28 18:56:28 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - During log-checkpoint it needs all the pages to be flushed
          before progressing checkpoint.
          User thread might still hold reference to pages residing in
          shared/system temporary tablespace because we cache the last
          inserted position.
        - These pages are not flushed if fix_count > 0 and so log-checkpoint
          can't proceed.
        - Avoid considering these pages and let log-checkpoint to proceed
          as anyway there is no redo logging associated with these pages.
    ------------------------------------------------------------
    revno: 7583.1.65 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-28 09:52:55 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.64
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-27 21:53:50 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - While traversing to previous record for intrinsic table
          we don't do optimsitic restore and so there is no left
          block that is pinned as done in normal cases always.
          Skip the section that after restore frees the pinned
          left block.
    ------------------------------------------------------------
    revno: 7583.1.63
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-27 21:01:13 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - If undo log is turned off then free extent are not reserved
          during DML. Enable this if table is intrinsic as undo log
          is turned off and this table can expect some DML that would
          need free extent.
    ------------------------------------------------------------
    revno: 7583.1.62 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-27 18:26:34 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.61
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-27 14:13:00 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - Fixed issues where-in if block with fix_count > 0 should never
          be selected for flush.
    ------------------------------------------------------------
    revno: 7583.1.60 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-27 13:25:00 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.59
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-27 11:09:34 +0530
    message:
      - WL#7682: Optimize InnoDB temporary tables.
        - For all the tables residing in system temporary tablespace
          sync between flush list and user thread is done using IO_FIX
          state.
    ------------------------------------------------------------
    revno: 7583.1.58 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-26 13:51:48 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.57
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-26 13:27:25 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If insert fails with TOO_BIG_RECORD then ensure that rollback
          action is trigger. No other error is expected while inserting
          as part of UPDATE as other errors are checked before starting
          insert.
    ------------------------------------------------------------
    revno: 7583.1.56
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-26 11:27:33 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Before allocatig pages for blob ensure that extent is reserved
          to avoid facing failure during allocation.
    ------------------------------------------------------------
    revno: 7583.1.55 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-26 08:43:40 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.54
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-25 16:31:41 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If column is too big for a page then InnoDB generate exepcted error.
        - Newly added path for intrinsic table was asserting if such error was
          seen.
    ------------------------------------------------------------
    revno: 7583.1.53
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-25 14:35:07 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Invalid select cursor if structure of index is going
          to chance even if the change is limited within the page.
    ------------------------------------------------------------
    revno: 7583.1.52
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-25 11:42:11 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Blob field_ref correctness check updated.
    ------------------------------------------------------------
    revno: 7583.1.51
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-03-24 15:25:50 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Missed marking buffer as dirty even though it was being used
          for DML action.
    ------------------------------------------------------------
    revno: 7583.1.50 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-03-24 10:58:25 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.49
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-03-24 10:57:16 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed race condition between User/Connection thread
          and background flush thread in accessing the block.
        - For normal block this is handled using SX-latch.
    ------------------------------------------------------------
    revno: 7583.1.48
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-20 18:29:48 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Set trx->error_state to avoid surprises.
    ------------------------------------------------------------
    revno: 7583.1.47
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-20 10:45:23 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected bug that avoided flushing of modified page.
          During insert if we hit duplicate then we mark the record deleted.
          For intrinsic table we don't latch the page and so even though
          record on the page was delete-marked it was not being scheduled for
          flushing.
    ------------------------------------------------------------
    revno: 7583.1.46
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-19 21:14:32 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - reverted back specialized condition for intrinsic table that was
          added when intrinsic table didn't supported inplace update.
    ------------------------------------------------------------
    revno: 7583.1.45
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-19 20:15:31 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed valgrind warning of un-initialized value.
    ------------------------------------------------------------
    revno: 7583.1.44
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-19 20:06:58 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Added few more test-scenarios to test-case related
          to repeated insert of same entry.
    ------------------------------------------------------------
    revno: 7583.1.43 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-19 19:43:46 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.42
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-19 19:31:19 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Re-enabled inplace update as code is not meant to handle duplicate
          node-ptr.
    ------------------------------------------------------------
    revno: 7583.1.41
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-18 16:12:52 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Improved condition to generate row-id if index is auto-generated.
    ------------------------------------------------------------
    revno: 7583.1.40
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-18 12:07:09 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - While comparing rec to rec it was assumed that rec are never equal
          in any b-tree page. They would differ in key part.
          For intrinsic table there is no inplace update and so delete-marked
          record = non-delete-marked record in all key parts.
          This use to cause the traversal loop to even compare node-ptr
          which is never expected when records are compared as node-ptr
          is not part of key.
          Corrected comparision condition to exit the loop once key part
          is compared.
    ------------------------------------------------------------
    revno: 7583.1.39 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-18 10:58:54 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.38
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-18 10:53:16 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed issue where-in left and right node can have same record
          with one delete-marked. Traversal algorithm needed correction
          to handle this case.
    ------------------------------------------------------------
    revno: 7583.1.37
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-14 16:49:26 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If during optimisitic insert page doesn't have enough space
          option-1 is to try re-organizing the page.
        - For intrinsic table we select a consistent option to use
          pessmisitic insert.
    ------------------------------------------------------------
    revno: 7583.1.36 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-14 14:31:52 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.35
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-14 14:30:47 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - While traversing through the deleted-records to reach the record of interest
          we might hit page boundary where-in infimum and supermum records.
          Skip them during traversal.
    ------------------------------------------------------------
    revno: 7583.1.34
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 16:43:21 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Expanded test-scenario to cover indexes on multiple column case.
    ------------------------------------------------------------
    revno: 7583.1.33
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 16:22:32 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - If an auto-gen clustered index is being updated then all the indexes
          needs to be updated as DB_ROW_ID is getting updated.
    ------------------------------------------------------------
    revno: 7583.1.32
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 12:07:32 +0530
    message:
      - W#7682: corrected assert check condition after the cursor has been re-position to the previous record
    ------------------------------------------------------------
    revno: 7583.1.31
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 11:36:17 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - fixed compilation error
    ------------------------------------------------------------
    revno: 7583.1.30
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 11:04:32 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected issue to return proper record to check-table api.
          (Note: check table api needs record in InnoDB format and not
           in MySQL format.)
    ------------------------------------------------------------
    revno: 7583.1.29
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 10:23:22 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - For intrinsic table, update = delete + insert
        - while searching for entry to delete cursor is located
          on last known entry even if there are multiple entry.
          If explicit_rollback has taken place before update
          (on duplicate update case) then this can lead to setting
          of cursor on deleted entry. Search backward for active
          entry to mark as deleted and insert new updated entry.
    ------------------------------------------------------------
    revno: 7583.1.28 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 08:37:14 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.27
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-13 08:34:42 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - With intrinsic table split can occur such that last record on left page
          = first record on right page with one of the record marked as deleted.
        - If this happenes parent verificiation can be confusing if started
          from last record of left page so detect such condition and avoid
          parent in these cases only.
    ------------------------------------------------------------
    revno: 7583.1.26
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-12 21:23:23 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected search-mode while rolling back delete.
    ------------------------------------------------------------
    revno: 7583.1.25
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-12 21:00:08 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - For check-table case InnoDB request the record in InnoDB format only
          (instead of converting it to MySQL) with some special prefix header.
        - This condition was missed in cursor interface that we developed for
          intrinsic table causing check table to fail.
    ------------------------------------------------------------
    revno: 7583.1.24 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-12 09:08:28 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.23
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-11 09:57:55 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected function/api signature to work with only UNIV_DEBUG flag defined.
    ------------------------------------------------------------
    revno: 7583.1.22 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-11 09:18:32 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.21 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Mon 2014-03-10 09:08:33 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.20
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-07 17:00:00 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Disabled test-case in embedded mode as it involves server restart.
    ------------------------------------------------------------
    revno: 7583.1.19
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-07 10:41:00 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Trailing spaces fixed.
    ------------------------------------------------------------
    revno: 7583.1.18
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-07 10:04:24 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected WL number.
    ------------------------------------------------------------
    revno: 7583.1.17 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Fri 2014-03-07 09:15:59 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.16
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 17:13:44 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Added test-scenarios for read-only cases
    ------------------------------------------------------------
    revno: 7583.1.15
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 16:02:44 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Added some comments for better understanding
    ------------------------------------------------------------
    revno: 7583.1.14
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 14:47:23 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Second round of cleanup
    ------------------------------------------------------------
    revno: 7583.1.13
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 12:09:40 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - First level code-cleanup
    ------------------------------------------------------------
    revno: 7583.1.12
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 11:26:22 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Improved code comments.
    ------------------------------------------------------------
    revno: 7583.1.11
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 10:22:49 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Removed some dead-code
    ------------------------------------------------------------
    revno: 7583.1.10
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 09:39:09 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed failing pb2 test-case.
        - Initially InnoDB just cached trx object in thread private cache.
          With introduction of intrinsic table we now cache other variable too.
          With that change if trx object can be null even though private
          cache object continue to remain valid.
        (Check for more detailed comments in innobase_close_connection)
    ------------------------------------------------------------
    revno: 7583.1.9 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Thu 2014-03-06 09:00:13 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.8
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 16:05:19 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Re-recording test-case to cover the newly added session/global
          variable use for creating intrinsic table directly.
    ------------------------------------------------------------
    revno: 7583.1.7
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 15:25:36 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Corrected optimized condition that was used to update secondary
          index only if needed.
        - Added test-sceanrio to cover big-data-types (blob).
    ------------------------------------------------------------
    revno: 7583.1.6
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 14:40:56 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Fixed complier warning.
    ------------------------------------------------------------
    revno: 7583.1.5
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 13:23:25 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Further improved test sceanrios.
    ------------------------------------------------------------
    revno: 7583.1.4
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 12:50:52 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Further improved test sceanrios.
    ------------------------------------------------------------
    revno: 7583.1.3
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 11:29:56 +0530
    message:
      - WL#7682: Optimizing InnoDB temporary tables.
        - Introduced special type of temporary tables named as
           intrinsic tables. These tables will inherit properties
            of temporary tables with added semantics:
        
            - intrinsic tables can reside only in shared temporary
              tablespace.
            - DML on intrinsic tables is not redo logged so no
              rollback action is possible.
            - intrinsic table uses optimized row-structure.
              (roll-ptr hidden column is not appened.)
            - most of the intrinsic table operation are optimized
              to take advantage of non-shared architecture
              and so locking is optimized + table and index ids
              are locally generated, no entry to dictionary (instead
              a reference is maintained in user thread), etc.
    ------------------------------------------------------------
    revno: 7583.1.2 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Wed 2014-03-05 10:54:06 +0530
    message:
      - WL#7682: merge from mysql-trunk
    ------------------------------------------------------------
    revno: 7583.1.1 [merge]
    committer: Krunal Bauskar krunal.bauskar@oracle.com
    branch nick: mysql-trunk-wl7682
    timestamp: Tue 2014-03-04 09:23:29 +0530
    message:
      - WL#7682: merge from mysql-trunk
------------------------------------------------------------
revno: 8014
committer: Atanu Ghosh <atanu.ghosh@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-14 10:21:11 +0530
message:
  Bug#18443096:  --DEFAULTS-EXTRA-FILE WITH NONEXISTENT FILE CAUSES SEGFAULT
  
  Problem: --defaults-file with a nonexistent file argument results 
  in a segfault.
  
  Fix: For a non-existent file, the free_used_memory need not be called.
  A check is made, as to whether the arguments are valid before
  freeing that memory.
------------------------------------------------------------
revno: 8013
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt5
timestamp: Wed 2014-05-14 00:08:13 +0200
message:
  n: bring up to date with current development version
------------------------------------------------------------
revno: 8012
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: topush-trunk
timestamp: Tue 2014-05-13 17:37:28 +0200
message:
  Bug#18694052: SERVER CRASH IN HA_PARTITION::INIT_RECORD_PRIORITY_QUEUE
  
  Missed to remove record priority queue if init_index failed for
  a partition.
  
  Approved by Vasil in rb#5406.
------------------------------------------------------------
revno: 8011
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Tue 2014-05-13 17:34:18 +0200
message:
  Bug#18702758 GET RID OF DYNAMIC_ARRAY IN DISPLAY_TABLE_LOCKS
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
    
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
  
  Extend Prealloced_array to be closer to std::vector
------------------------------------------------------------
revno: 8010
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 22:07:14 +0700
message:
  This patch is a follow-up of patch for bug#14670465 - PLEASE PRINT
  HUMAN READABLE, ESCAPED XID DATA IN XA RECOVER OUTPUT.
  
  This patch updates P_S digests after new symbol was added to the parser.
------------------------------------------------------------
revno: 8009
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 17:14:58 +0300
message:
  Non-functional followup to
  annamalai.gurusami@oracle.com-20140210082850-vqlzadis1asdws0e:
  whitespace and stylistic fixup and loop simplification.
------------------------------------------------------------
revno: 8008
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 16:30:17 +0300
message:
  Non-functional change around buf_pool->flush_rbt:
  use NULL when comparing and remove UNIV_LIKELY_NULL.
------------------------------------------------------------
revno: 8007
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-29 16:20:35 +0300
message:
  Whitespace indentation fixup.
------------------------------------------------------------
revno: 8006
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: trunk
timestamp: Tue 2014-05-13 16:16:02 +0200
message:
  Add missing '-' in Anitha's previous push.
------------------------------------------------------------
revno: 8005
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B16272328-trunk
timestamp: Thu 2014-04-10 16:56:58 +0300
message:
  Bug #16272328 --DEBUG IN MYSQLIMPORT DOES NOT WORK
  
  Unified the use of --debug/--debug-check/--debug-info across 
  the client tools.
------------------------------------------------------------
revno: 8004
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-sb1
timestamp: Tue 2014-05-13 14:17:57 +0200
message:
  bug#18558561: Replace Cost_estimate with a double when estimating join cost
  
  The member variable prefix_cost of class POSITION is made a double.
  Since partial plans and complete plans are compared with a cost represented
  as a double, it makes sense to store this as a double too.
  
  The member variable prefix_record_count of class POSITION is renamed to
  prefix_rowcount for more consistent naming.
  
  The function set_prefix_costs() is renamed to set_prefix_cost().
  
  A new function set_prefix_join_cost() that calculates the cost of a prefix
  plan has been added.
  
  The above change of type has made it possible to eliminate the arguments
  record_count and read_time that were currently shipped to some functions
  of class Optimize_table_order: consider_plan(),
  best_extension_by_limited_search(), eq_ref_extension_by_limited_search()
  and advance_sj_state(). These functions use member variables from
  POSITION instead.
  
  A new const property has_sj has been added to class Optimize_table_order,
  so it does not have to be calculated more than once.
  
  In JOIN::setup_materialized_table(), the cost of a semi-join
  materialization is correctly calculated, to keep a correct cost for a
  final plan.
------------------------------------------------------------
revno: 8003 [merge]
committer: Andrei Elkin <andrei.elkin@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 14:53:48 +0300
message:
  merging post-push Bug18563480 from 5.6
    ------------------------------------------------------------
    revno: 8001.1.1 [merge]
    committer: Andrei Elkin <andrei.elkin@oracle.com>
    branch nick: mysql-trunk
    timestamp: Tue 2014-05-13 14:43:03 +0300
    message:
      merging post-push Bug18563480 from 5.6
        ------------------------------------------------------------
        revno: 3902.324.55
        committer: Andrei Elkin <andrei.elkin@oracle.com>
        branch nick: 5.6-fixes
        timestamp: Tue 2014-05-13 14:35:58 +0300
        message:
          bug18563480 post-push: compile on win is fixed.
------------------------------------------------------------
revno: 8002
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-valgrind
timestamp: Tue 2014-05-13 13:38:47 +0200
message:
  Bug#18693654 VALGRIND WARNINGS IN INFLATE ON UNCOMPRESS
  
  The value of the sourceLen argument to uncompress() was wrong,
  and we got valgrind warnings when trying to verify the zip header of the compressed data.
------------------------------------------------------------
revno: 8001
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 13:31:10 +0200
message:
  Fixed duplicate entry for vardir
------------------------------------------------------------
revno: 8000 [merge]
committer: Andrei Elkin <andrei.elkin@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 12:42:38 +0300
message:
  merging Bug#18563480 from 5.6
    ------------------------------------------------------------
    revno: 3902.324.54
    committer: Andrei Elkin <andrei.elkin@oracle.com>
    branch nick: 5.6-fixes
    timestamp: Tue 2014-05-13 12:09:33 +0300
    message:
      Bug#18563480 CRASH WHEN SLAVE WORKER TRY TO EXECUTE A BIG STATEMENT THAT SHOULD FAIL
      
      The failure was caused by exceeding of capacity of the internal buffer
      by ultimate error reporting message text.
      
      It is fixed by submitting the error text to the size of the
      buffer. Any excess is simply cut off.
------------------------------------------------------------
revno: 7999
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk-bug14670465_2push
timestamp: Tue 2014-05-13 15:44:44 +0700
message:
  This patch is a follow-up of patch for bug#14670465 - PLEASE PRINT
  HUMAN READABLE, ESCAPED XID DATA IN XA RECOVER OUTPUT.
  
  This patch adds doxygen-style comments for private methods
  Sql_cmd_xa_commit::trans_xa_commit, Sql_cmd_xa_rollback::trans_xa_rollback
  Sql_cmd_xa_start::trans_xa_start, Sql_cmd_xa_prepare::trans_xa_prepare,
  Sql_cmd_xa_end::trans_xa_end, Sql_cmd_xa_recover::trans_xa_recover.
------------------------------------------------------------
revno: 7998
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk-bug14670465_2push
timestamp: Tue 2014-05-13 15:01:27 +0700
message:
  This patch is a follow-up of patch for bug#14670465 - PLEASE PRINT
  HUMAN READABLE, ESCAPED XID DATA IN XA RECOVER OUTPUT.
  
  The original patch changed the behaviour of XA RECOVER to use hex
  encoding for output of XID values which contained non-printable
  characters. These changes affected the Connector/J codebase in the
  way that led to additional work for detecting the format of XA
  RECOVER output.
  
  This patch reverts the original patch. To fix the original bug and don't
  introduce incompatible changes the XA RECOVER statement has been extended
  to add support for output of XID as hex string on user request.
  
  To do so the additional clause CONVERT XID has been added for the statement
  XA RECOVER. Also as part of this bug fix the refactoring has been done that
  allows to represent XA-related SQL-commands by classes derived from
  the base class Sql_cmd. The following derived classes have been introduced:
    Sql_cmd_xa_start,Sql_cmd_xa_end, Sql_cmd_xa_prepare,
    Sql_cmd_xa_recover, Sql_cmd_xa_commit, Sql_cmd_xa_rollback
  Moreover, since dedicated classes allows to store XA-specific information
  this patch also gets rid of using LEX::xid and LEX::xa_opt.
------------------------------------------------------------
revno: 7997
committer: Namit Sharma<namit.sharma@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 12:17:30 +0530
message:
  Bug#18620520 CORRECT TYPO IN ER 1755(ER_MTS_CANT_PARALLEL)
  
  Problem
  ------ 
  TYPO in ER 1755.
  Show Slave Status:
  ................
  ................
                     Last_Errno: 1755
                     Last_Error: Cannot execute the current event group in the
  parallel mode. Encountered event Gtid, relay-log name
  ./slave-relay-bin.000002, position 363 which prevents execution of this event
  group in parallel mode. Reason: The master does not support the selected
  parallelization mode. It may be too old, or replication was started from an
  event internal to a transaaction..
  
  Fix
  ------
  Change in:
  File: log_event.cc
  Line number: 2894
  Line: "to a transaaction.");
------------------------------------------------------------
revno: 7996
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 10:00:20 +0530
message:
  Bug #18723872  FK: ASSERTION: *CURSOR->INDEX->NAME == TEMP_INDEX_PREFIX
  
  This is a regression caused by the fix for Bug #11758237 INSERT ON DUPLICATE
  KEY UPDATE SOMETIMES WRITES BINLOG POSITION INCORRECT.  The trx->error_state
  will be modified when a thread is suspended.  But logic of placing just the gap
  locks on unique secondary indexes without actually performing the insert
  depends on the value of trx->error_state.  So the value of trx->error_state is
  restored for insert into every index.  Refer to rb#3196 for the original fix.
  
  approved by Marko over IM.
------------------------------------------------------------
revno: 7995
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 11:16:26 +0800
message:
  Bug#18266989 ACK_RECEIVER THREAD DOESN'T START WHEN STARTING
               SEMISYNC THROUGH SERVER OPTIONS
  
  Ack_receiver thread on the master didn't start successfully
  when setting up semisync replication through server options.
  
  To fix it, ack receiver thread is created if
  rpl_semi_sync_master_enabled is set 1 through server option.
  
  It also corrected the defination of REPLY_MESSAGE_MAX_LENGTH.
------------------------------------------------------------
revno: 7994 [merge]
committer: Shaohua Wang <shaohua.wang@oracle.com>
branch nick: mysql-trunk-bugfix1
timestamp: Tue 2014-05-13 10:34:38 +0800
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.53
    committer: Shaohua Wang <shaohua.wang@oracle.com>
    branch nick: mysql-5.6-bugfix1
    timestamp: Tue 2014-05-13 10:10:13 +0800
    message:
      Commit Message:
      BUG#18079671 - INNODB FTS : ASSERT IN FTS_CACHE_APPEND_DELETED_DOC_IDS
      
      Analysis:
      When query thread accesses deleted_doc_ids in fts_cache_append_deleted_doc_ids,
      fts sync thread sets deleted_doc_ids to NULL in fts_cache_clear, so it causes
      the first thread to access violation.
      
      Solution:
      1. Protect deleted_doc_ids in fts_cache_clear and fts_cache_init;
      2. Check whether deleted_doc_ids is NULL in fts_cache_append_deleted_doc_ids.
      
      rb://4740 approved by Jimmy.Yang
------------------------------------------------------------
revno: 7993
committer: Anitha Gopi <anitha.gopi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-13 04:23:55 +0200
message:
  Test fails only on freeBSD and OSX. Restruct disabling to affected platforms
------------------------------------------------------------
revno: 7992 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Mon 2014-05-12 15:25:07 +0200
message:
  empty merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.52 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Mon 2014-05-12 15:23:29 +0200
    message:
      merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.592.39
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5
        timestamp: Mon 2014-05-12 15:21:23 +0200
        message:
          Backport from trunk:
          Bug #18382225 MYSQL_CONFIG CAN'T HANDLE RELOCABLE PACKAGES THAT USES "LIB64" OR "-64" SUFFIX
            
          'lib' is hardcoded into mysql_config, so 'cmake -DINSTALL_LIBDIR=lib64' will not work.
          Use INSTALL_LIBDIR when generating mysql_config.
            
          mysql_config may be renamed to e.g. mysql_config-32, fix the basedir pattern matching.
------------------------------------------------------------
revno: 7991
committer: Vinay Fisrekar <vinay.fisrekar@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-12 17:31:33 +0530
message:
  Disable federated test dur to regression failure.
------------------------------------------------------------
revno: 7990 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: trunk
timestamp: Mon 2014-05-12 17:17:48 +0530
message:
  Bug #17657223	 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.324.51
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: t-5.6
    timestamp: Mon 2014-05-12 17:16:50 +0530
    message:
      Bug #17657223    EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLEi
      		Post push fix.
------------------------------------------------------------
revno: 7989 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: trunk
timestamp: Mon 2014-05-12 16:02:48 +0530
message:
  Bug #17657223	 EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
  	Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.324.50
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: t-5.6
    timestamp: Mon 2014-05-12 15:58:24 +0530
    message:
      Bug #17657223	EXCESSIVE TEMPORARY FILE USAGE IN ALTER TABLE
      
      Analysis:
      Temporary file usage during alter table operation can be avoided
      during clustered index rebuild when new primary key follow existing
      pk order. 
      Scenarios for primary key in same order are
           1) when non primary key is added or removed.
           2) Dropping columns at the tail of the primary key.
           3) Newly added column with default value can be anywhere between
      the old pk fields.
      
      So it will reduce the IO usage during clustered index rebuild.
      Delay the temporary file creation and log file creation
      for alter table operation will lead to avoid the file creation
      for smaller tables.
      
      	Approved by marko, shaohua, matthias leich (rb-4788)
------------------------------------------------------------
revno: 7988
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-12 14:45:16 +0530
message:
  Bug #18646430 ASSERT F_LEN >= BTR_EXTERN_FIELD_REF_SIZE, ROW_EXT_CACHE_FILL()
  
  Post push fix.  There is a address sanitizer error reported.  This
  will fix it.  The dfield_t object must be cloned and then put into
  the update vector.
  
  approved by Marko over IM.
------------------------------------------------------------
revno: 7987 [merge]
committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
branch nick: trunk
timestamp: Mon 2014-05-12 12:26:09 +0530
message:
  Bug#17532932 - TIMESTAMP AND DATETIMES SELF-INCOMPATIBLE DURING REPLICATION
  	(merge from 5.6 to trunk)
    ------------------------------------------------------------
    revno: 3902.324.49
    committer: Neeraj Bisht <neeraj.x.bisht@oracle.com>
    branch nick: 5.6
    timestamp: Mon 2014-05-12 12:18:27 +0530
    message:
      Bug#17532932 - TIMESTAMP AND DATETIMES SELF-INCOMPATIBLE DURING REPLICATION
      
      Problem:
      In RBR, replication between master and slave tables are failing with an 
      error (1677), when the tables contain temporal type fields(TIMESTAMP,
      DATETIME,TIME). 
      
      Analysis:
      In the following scenarion:
      Master(mysql-5.6.12)
      -->created a table with TIMESTAMP field.
      Slave(mysql-5.6.14)
      -->created a table with TIMESTAMP field in mysql-5.5.
      -->upgraded the 5.5 data directory for 5.6.14 and used that 
      as slave for replication.
      
      Now when we are trying to insert a record with row based replication. 
      We will get an error saying 
      "Column * of table '****' cannot be converted from type 'timestamp' to 
      type 'timestamp'"
      
      The Problem is as in mysql-5.6 we introduce a new type 
      MYSQL_TYPE_TIMESTAMP2(this will carry fraction part for timestamp field) 
      and in mysql-5.5 we dont have MYSQL_TYPE_TIMESTAMP2 type.
      So when we create a field of TIMESTAMP(sql type)  in 5.6 it will be 
      of MYSQL_TYPE_TIMESTAMP2(internal type) and in 5.5 it will create a 
      MYSQL_TYPE_TIMESTAMP(internal type).
      
      According to documentation, when we upgrade from 5.5 to 5.6, 
      there is some Incomatible changes document in
      http://dev.mysql.com/doc/refman/5.6/en/upgrading-from-previous-series.html
       
      "Due to the temporal type changes described in the previous 
      incompatible change item above, importing pre-MySQL 5.6.4 tables that 
      contain DATETIME and TIMESTAMP types into MySQL 5.6.4 (or later) fails."
      
      Which result that, in upgraded slave we have old TIMESTAMP type field.
      So when we are trying to insert a value in RBR mode, we get an error as the 
      type mismatch happen.
      
      
      In replication, we are supporting MYSQL_TYPE_TIMESTAMP->MYSQL_TYPE_TIMESTAMP2
      but the other way is not supported(i.e., Master with MYSQL_TYPE_TIMESTAMP2, slave 
      with MYSQL_TYPE_TIMESTAMP), which happens in upgraded server because of the 
      limitation of mysql_upgrade.
      
      Solution:
      Added code to do conversion between MYSQL_TYPE_TIMESTAMP2->MYSQL_TYPE_TIMESTAMP 
------------------------------------------------------------
revno: 7986
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-05-11 06:52:08 +0530
message:
  Bug #17059925: UNIONS COMPUTES ROWS_EXAMINED INCORRECTLY
  
  Post-push-fix: Added some sleep to selects. If not on
                 some machines queries are not logged to
                 slog_log table because they are executed
                 too fast.
------------------------------------------------------------
revno: 7985 [merge]
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-05-11 07:58:05 +0800
message:
  Manual Merge
    ------------------------------------------------------------
    revno: 7717.1.1
    committer: Libing Song <libing.song@oracle.com>
    branch nick: mysql-trunk-wl6813-new
    timestamp: Wed 2014-03-19 13:13:28 +0800
    message:
      WL#6813: Intra-schema MTS: ordered commits (sequential consistency)
      
      DESCRIPTION
      ===========
      This worklog ensures that the commits by slave worker threads running in
      parallel are done in the same order as on the master. This also means that
      the slave will never externalize a database state which was never externalized
      by the master.
      
      DESIGN
      ======
      - slave_preserve_commit_order
        The behavior is optional. Users can control it through a new system variable
        named slave_preserve_commit_order.
      
      - Commit_order_manager
        We implement a commit order manager which maintains a queue of transactions
        which are applying by workers.
      
        Coordinator registers them into the queue when dispatching them to a worker.
        So the transactions in the queue have the order as they are in the relay log.
      
        All the transactions should go into the flush stage(see Binlog Group commit)
        in the queue sequence. When a worker commits its transaction(before the
        transaction goes into flush stage), it must wait until it becomes the queue
        head. It does means all transactions before it have entered into flush stage.
      
        Once the transaction enters flush stage, It is popped from the queue
        and signals the next transaction to go ahead.
      
      The feature is for only intra-schema multi-threaded slave, but not database
      partitioned multi-threaded slave. And it is not supported when log_bin or
      log_slave_updates is OFF.
------------------------------------------------------------
revno: 7984
committer: Libing Song <libing.song@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-05-11 07:51:21 +0800
message:
  Bug#18363515 MULTIPLE THREADED SLAVE MAY BE CRASHED BY
               STOP SLAVE SQL
  
  Worker error will crash MTS if coordinator is waiting for a
  free worker to dispatch next event. The crash happens because
  the caller of Mts_submode_logical_clock::get_least_occupied_worker()
  supposes it always return a valid worker. But that is not true.
  It may turn NULL when coordinator is aborted at the same time.
  
  This patch makes the caller of get_least_occupied_worker() to
  check if the function returns a valid value.
------------------------------------------------------------
revno: 7983
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: topush-trunk
timestamp: Sat 2014-05-10 20:33:40 +0200
message:
  post-push fix for test failures from bug#18330694.
  
  Due to added sleep + retry on failed remove/move/copy_file
  operations
  
  Fixed by not trying to remove_files possible non-existing
  files.
------------------------------------------------------------
revno: 7982
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-09 20:40:36 +0530
message:
  Bug #18646430 ASSERT F_LEN >= BTR_EXTERN_FIELD_REF_SIZE, ROW_EXT_CACHE_FILL()
  
  This patch solves the following problems:
  
  * In the function btr_store_big_rec_extern_fields(), update the offsets
    so that the "external storage" flags are correct when it ends. 
  * After the call to btr_store_big_rec_extern_fields(), the any cached
    record pointer must be updated in the caller.
  * In fts code, there was one uninitialized dfield->ext.
    Initialize it correctly. 
  
  rb#5316 approved by Marko
------------------------------------------------------------
revno: 7981
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-09 11:22:52 +0300
message:
  Bug#18730524 REPEATED KILL+RESTART FAILS DUE TO MISSING MLOG_FILE_NAME RECORD
  
  There were several spurious test failures on PB2 where InnoDB failed to start
  due to missing MLOG_FILE_NAME record:
  
  ... [ERROR] [FATAL] InnoDB: Missing MLOG_FILE_NAME or MLOG_FILE_DELETE for
  redo log record ...
  
  I analyzed one of them (innodb.innodb-double-write on daily-trunk 2014-05-07
  carefully. I was only able to repeat the bug by tweaking code in debugger:
  
  create table ...; insert lots of data;
  [kill the server]
  Set a breakpoint on log_checkpoint()
  restart with --debug=d,ib_log
  
  Observe the LSN of the last MLOG_FILE_NAME record (rec 55) for this
  tablespace.
  
  When the breakpoint on log_checkpoint() is hit
  (the master thread does periodic checkpoints,
  even if there is no user activity),
  tweak buf_flush_ready_for_flush() to always return false.
  
  Check the value of
  
  	oldest_lsn = log_buf_pool_get_oldest_modification();
  
  With our tweak to buf_flush_ready_for_flush(), this should be not later than
  the MLOG_FILE_NAME record, which is fine.
  
  When buf_flush_ready_for_flush() is returning suitable values, the
  oldest_lsn could end up being somewhere between the LSN of the
  MLOG_FILE_NAME record and the last logged modification for the
  tablespace.
  
  In my manual repeat, oldest_lsn happened to be exactly the LSN of the
  MLOG_FILE_NAME record. I faked oldest_lsn to be immediately after the
  MLOG_FILE_NAME record, and let the execution continue.
  
  Immediately after log_checkpoint() returns, kill and restart the
  server. You should get the message:
  
  ... [ERROR] [FATAL] InnoDB: Missing MLOG_FILE_NAME or MLOG_FILE_DELETE for
  redo log record ...
  
  The fix:
  
  recv_init_crash_recovery_spaces(): Invoke fil_names_dirty() on the
  tablespaces for which MLOG_FILE_NAME records were scanned during crash
  recovery.
  
  recv_recovery_from_checkpoint_start(): Initialize log_sys->lsn a
  little earlier, so that fil_names_dirty() will get the correct value.
  
  Patch approved by Jimmy Yang on IM.
------------------------------------------------------------
revno: 7980
committer: Satya Bodapati <satya.bodapati@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-09 11:49:44 +0530
message:
  BUG#18642372 - I_INNODB.INNODB-LOG-FILE THROW TIMEOUT ERROR ONLY ON WINDOWS
  
  Problem:
  
  The reason for hang is because DeleteFile() hangs on ibtmp1. 
  The file is opened by the page cleaner thread. The handle is not in
  Datafile Object but it is stored in fil_node_t.
  
  The explicit close() on Datafile will not solve the issue because the
  handle is not opened by Datafile Object. It is stored in fil_node_t by page
  cleaner thread.
  
  Fix:
  Exit flush thread early incase of startup failure. This will make sure
  that we don't have handle open on 'ibtmp1' by the page cleaner thread.
  Without any open handles, deletion of ibtmp1 will be successful
  
  Note:
  Although it->close() in Tablespace::delete_files() will not solve the
  current hang problem, it will make sure that we close the file before deleting.
  
  Approved by Kevin,Yasufumi,Krunal rb#5258
------------------------------------------------------------
revno: 7979 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-09 10:00:22 +0530
message:
  Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
  SHOW PROCESSLIST, SHOW BINLOGS
        
  Null merge from mysql-5.6 
    ------------------------------------------------------------
    revno: 3902.324.48 [merge]
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-05-09 09:53:50 +0530
    message:
      Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
      SHOW PROCESSLIST, SHOW BINLOGS
      
      Fixing post push test failure (MTR does not like giving
      127.0.0.1 for localhost incase of --embedded run, it thinks
      it is an external ip address)
        ------------------------------------------------------------
        revno: 2875.592.38
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-05-09 09:52:15 +0530
        message:
          Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
          SHOW PROCESSLIST, SHOW BINLOGS
          
          Fixing post push test failure (MTR does not like giving
          127.0.0.1 for localhost incase of --embedded run, it thinks
          it is an external ip address)
------------------------------------------------------------
revno: 7978
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Thu 2014-05-08 16:38:56 +0200
message:
  Bug#18715670 CRASH IN DECIMAL_ACTUAL_FRACTION COMPARING DECIMAL TO NULLS
  
  Another bug in the range optimizer, uncovered by the patch for
  Bug#18556403 USE STD LIBRARY FOR SORTING AND MATCHING IN SUBCLASSES OF
  
  The range optimizer would build predicates for empty in-lists
  (since NULL values are removed from NOT IN (in-list))
------------------------------------------------------------
revno: 7977 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-08 18:23:44 +0530
message:
  Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
  SHOW PROCESSLIST, SHOW BINLOGS
        
  Null merge from mysql-5.6 ( Issue is not present in mysql-trunk)
  except the test script (and required debug sync points)
    ------------------------------------------------------------
    revno: 3902.324.47 [merge]
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-08 18:17:46 +0530
    message:
            Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
            SHOW PROCESSLIST, SHOW BINLOGS
            
            Merge from mysql-5.5 ( Fix is changed from
            mysql-5.5's patch)
            
            Problem:  A deadlock was occurring when 4 threads were
            involved in acquiring locks in the following way
            Thread 1: Dump thread ( Slave is reconnecting, so on
                          Master, a new dump thread is trying kill
                          zombie dump threads. It acquired thread's
                          LOCK_thd_data and it is about to acquire
                          mysys_var->current_mutex ( which LOCK_log)
            Thread 2: Application thread is executing show binlogs and
                           acquired LOCK_log and it is about to acquire
                           LOCK_index.
            Thread 3: Application thread is executing Purge binary logs
                           and acquired LOCK_index and it is about to
                           acquire LOCK_thread_count.
            Thread 4: Application thread is executing show processlist
                           and acquired LOCK_thread_count and it is
                           about to acquire zombie dump thread's
                           LOCK_thd_data.
            Deadlock Cycle:
                 Thread 1 -> Thread 2 -> Thread 3-> Thread 4 ->Thread 1
            
            The same above deadlock was observed even when thread 4 is
            executing 'SELECT * FROM information_schema.processlist' command and
            acquired LOCK_thread_count and it is about to acquire zombie
            dump thread's LOCK_thd_data.
            
            Analysis:
            There are four locks involved in the deadlock.  LOCK_log,
            LOCK_thread_count, LOCK_index and LOCK_thd_data.
            LOCK_log, LOCK_thread_count, LOCK_index are global mutexes
            where as LOCK_thd_data is local to a thread.
            We can divide these four locks in two groups.
            Group 1 consists of LOCK_log and LOCK_index and the order
            should be LOCK_log followed by LOCK_index.
            Group 2 consists of other two mutexes
            LOCK_thread_count, LOCK_thd_data and the order should
            be LOCK_thread_count followed by LOCK_thd_data.
            Unfortunately, there is no specific predefined lock order defined
            to follow in the MySQL system when it comes to locks across these
            two groups. In the above problematic example,
            there is no problem in the way we are acquiring the locks
            if you see each thread individually.
            But If you combine all 4 threads, they end up in a deadlock.
            
            Fix:
            Since everything seems to be fine in the way threads are taking locks,
            In this patch We are changing the duration of the locks in Thread 4
            to break the deadlock. i.e., before the patch, Thread 4
            ('show processlist' command) mysqld_list_processes()
            function acquires LOCK_thread_count for the complete duration
            of the function and it also acquires/releases
            each thread's LOCK_thd_data. Instead of it, Now it will take
            a copy of THDs from global_thread_list and perform traversal
            (on copied THDs) only after releasing  LOCK on LOCK_thread_count.
            During traversal(on copied THDs), removal from global_thread_list is
            blocked using another mutex LOCK_thd_remove such that
            THD copied are valid during traversal(otherwise remove destroys THD).
            
            Now the new locking order after this patch is:
            LOCK_thd_remove -> LOCK_thd_data -> LOCK_log ->
            LOCK_index -> LOCK_thread_count
        ------------------------------------------------------------
        revno: 2875.592.37
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-05-08 18:13:01 +0530
        message:
          Bug#17283409  4-WAY DEADLOCK: ZOMBIES, PURGING BINLOGS,
          SHOW PROCESSLIST, SHOW BINLOGS
          
          Problem:  A deadlock was occurring when 4 threads were
          involved in acquiring locks in the following way
          Thread 1: Dump thread ( Slave is reconnecting, so on
                        Master, a new dump thread is trying kill
                        zombie dump threads. It acquired thread's
                        LOCK_thd_data and it is about to acquire
                        mysys_var->current_mutex ( which LOCK_log)
          Thread 2: Application thread is executing show binlogs and
                         acquired LOCK_log and it is about to acquire
                         LOCK_index.
          Thread 3: Application thread is executing Purge binary logs
                         and acquired LOCK_index and it is about to
                         acquire LOCK_thread_count.
          Thread 4: Application thread is executing show processlist
                         and acquired LOCK_thread_count and it is
                         about to acquire zombie dump thread's
                         LOCK_thd_data.
          Deadlock Cycle:
               Thread 1 -> Thread 2 -> Thread 3-> Thread 4 ->Thread 1
          
          The same above deadlock was observed even when thread 4 is
          executing 'SELECT * FROM information_schema.processlist' command and
          acquired LOCK_thread_count and it is about to acquire zombie
          dump thread's LOCK_thd_data.
          
          Analysis:
          There are four locks involved in the deadlock.  LOCK_log,
          LOCK_thread_count, LOCK_index and LOCK_thd_data.
          LOCK_log, LOCK_thread_count, LOCK_index are global mutexes
          where as LOCK_thd_data is local to a thread.
          We can divide these four locks in two groups.
          Group 1 consists of LOCK_log and LOCK_index and the order
          should be LOCK_log followed by LOCK_index.
          Group 2 consists of other two mutexes
          LOCK_thread_count, LOCK_thd_data and the order should
          be LOCK_thread_count followed by LOCK_thd_data.
          Unfortunately, there is no specific predefined lock order defined
          to follow in the MySQL system when it comes to locks across these
          two groups. In the above problematic example,
          there is no problem in the way we are acquiring the locks
          if you see each thread individually.
          But If you combine all 4 threads, they end up in a deadlock.
          
          Fix: 
          Since everything seems to be fine in the way threads are taking locks,
          In this patch We are changing the duration of the locks in Thread 4
          to break the deadlock. i.e., before the patch, Thread 4
          ('show processlist' command) mysqld_list_processes()
          function acquires LOCK_thread_count for the complete duration
          of the function and it also acquires/releases
          each thread's LOCK_thd_data.
          
          LOCK_thread_count is used to protect addition and
          deletion of threads in global threads list. While show
          process list is looping through all the existing threads,
          it will be a problem if a thread is exited but there is no problem
          if a new thread is added to the system. Hence a new mutex is
          introduced "LOCK_thd_remove" which will protect deletion
          of a thread from global threads list. All threads which are
          getting exited should acquire LOCK_thd_remove
          followed by LOCK_thread_count. (It should take LOCK_thread_count
          also because other places of the code still thinks that exit thread
          is protected with LOCK_thread_count. In this fix, we are changing
          only 'show process list' query logic )
          (Eg: unlink_thd logic will be protected with
          LOCK_thd_remove).
          
          Logic of mysqld_list_processes(or file_schema_processlist)
          will now be protected with 'LOCK_thd_remove' instead of
          'LOCK_thread_count'.
          
          Now the new locking order after this patch is:
          LOCK_thd_remove -> LOCK_thd_data -> LOCK_log ->
          LOCK_index -> LOCK_thread_count
------------------------------------------------------------
revno: 7976
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-review
timestamp: Thu 2014-05-08 12:53:04 +0200
message:
  Bug#18675099: WINDOWSCACHE.CMAKE MUST BE UPDATED FOR VS2013
  
  Post-push fix: build was broken on 32bit windows.
  Fix: use a separate cmake variable HAVE_EPOLL_WAIT rather than HAVE_EPOLL
------------------------------------------------------------
revno: 7975 [merge]
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-08 14:56:03 +0530
message:
  Bug #17059925: UNIONS COMPUTES ROWS_EXAMINED INCORRECTLY
  
  ISSUE:
  ------
  For UNION of selects, rows examined by the query will be sum
  of rows examined by individual select operations and rows
  examined for union operation. The value of session level
  global counter that is used to count the rows examined by a
  select statement should be accumulated and reset before it
  is used for next select statement. But we have missed to
  reset the same. Because of this examined row count of a
  select query is accounted more than once.
  
  SOLUTION:
  ---------
  Since we use a single session level global counter which we
  only increment we do not need to accumulate it in a local
  counter while handling union operations.
  To achieve this we do following steps
  1. For every statement reset the global counter.
  2. Make sure we do not reset this global counter during the
     query execution.
  With these things we can remove the local counter used in
  Union operation.
    ------------------------------------------------------------
    revno: 3902.324.46 [merge]
    committer: mithun <mithun.c.y@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-08 14:52:17 +0530
    message:
      Bug #17059925 : UNIONS COMPUTES ROWS_EXAMINED INCORRECTLY
      Merge from 5.5
        ------------------------------------------------------------
        revno: 2875.592.36
        committer: mithun <mithun.c.y@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-05-08 14:49:53 +0530
        message:
          Bug #17059925: UNIONS COMPUTES ROWS_EXAMINED INCORRECTLY
          
          ISSUE:
          ------
          For UNION of selects, rows examined by the query will be sum
          of rows examined by individual select operations and rows
          examined for union operation. The value of session level
          global counter that is used to count the rows examined by a
          select statement should be accumulated and reset before it
          is used for next select statement. But we have missed to
          reset the same. Because of this examined row count of a
          select query is accounted more than once.
          
          SOLUTION:
          ---------
          In union reset the session level global counter used to
          accumulate count of examined rows after its value is saved.
------------------------------------------------------------
revno: 7974 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Thu 2014-05-08 14:45:14 +0530
message:
  Bug #18045646 LOCAL USER CAN RUN ARBITRARY CODE IN THE CONTEXT OF THE MYSQL SERVER
  
  Merging from 5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.324.45 [merge]
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Thu 2014-05-08 14:42:57 +0530
    message:
      Bug #18045646 LOCAL USER CAN RUN ARBITRARY CODE IN THE CONTEXT OF THE MYSQL SERVER
      
      Merging from 5.5 to 5.6
        ------------------------------------------------------------
        revno: 2875.592.35
        committer: Venkata Sidagam <venkata.sidagam@oracle.com>
        branch nick: 5.5
        timestamp: Thu 2014-05-08 14:41:01 +0530
        message:
          Bug #18045646 LOCAL USER CAN RUN ARBITRARY CODE IN THE CONTEXT OF THE MYSQL SERVER
          
          Description: Using the temporary file vulnerability an
          attacker can create a file with arbitrary content at a
          location of his choice. This can be used to create the
          file /var/lib/mysql/my.cnf, which will be read as a
          configuration file by MySQL, because it is located in the
          home directory of the mysql user. With this configuration
          file, the attacker can specify his own plugin_dir variable,
          which then allows him to load arbitrary code via
          "INSTALL PLUGIN...".
          
          Analysis: While creating the ".TMD" file we are not checking
          if the file is already exits or not in mi_repair() function.
          And we are truncating if the ".TMD" file exits and going ahead
          This is creating the security breach.
          
          Fix: We need to use O_EXCL flag along with O_RDWR and O_TRUNC
          which will make sure if any user creates ".TMD" file, will
          fails the repair table with "cannot create ".TMD" file error".
          Actually we are initialing "param.tmpfile_createflag" member
          with O_RDWR | O_TRUNC | O_EXCL in myisamchk_init(). And we
          are modifying it in ha_myisam::repair() to O_RDWR | O_TRUNC.
          So, we need to remove the line which is modifying the
          "param.tmpfile_createflag".
------------------------------------------------------------
revno: 7973
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-08 09:55:26 +0200
message:
  Bug#17007529: REMOVE MYSQL SERVER SUPPORT FOR LINUXTHREADS
  
  Follow-up patch: Remove unneeded valgrind suppressions for
  nptl_pthread_exit_hack_handler() as the original patch removed
  this function.
------------------------------------------------------------
revno: 7972
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug18675099
timestamp: Thu 2014-05-08 08:02:59 +0200
message:
  Bug#18675099: WINDOWSCACHE.CMAKE MUST BE UPDATED FOR VS2013
  
  WindowsCache.cmake hardcodes results for CMake configure checks so they
  don't have to be performed on Windows (as this is very slow).
  
  The problem was that the cache file had several values which were no
  longer correct. The consequence that we were using Windows specific
  workarounds in cases where it was no longer needed.
  
  This patch removes the outdated values from the cache file so that the
  checks are actually made. It also reorganizes the cache file so that it
  matches config.h.cmake, removes duplicate entries, removes dead
  entries and adds missing entries. The patch also removes some
  unneeded Visual Studio version checks.
------------------------------------------------------------
revno: 7971
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-05-08 08:00:13 +0200
message:
  var-dir and mem cannot be set at the same time. So, removed mem.
------------------------------------------------------------
revno: 7970
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-05-08 10:41:12 +0800
message:
  Fixed bug#18671824 ASSERT INFO_BITS ==
  MIN_MARK,PAGE_ZIP_DIR_ENCODE(),RTR_UPDATE_MBR_FIELD(), PURGE
  
  This bug is caused by there're 2 min-rec marked recs in a page
  of R-tree index.
  
  Approved by Jimmy on IM.
------------------------------------------------------------
revno: 7969 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-review
timestamp: Wed 2014-05-07 17:11:18 +0200
message:
  empty merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.44 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Wed 2014-05-07 17:10:14 +0200
    message:
      NULL merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.592.34
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5-merge
        timestamp: Wed 2014-05-07 17:09:14 +0200
        message:
          Backport from trunk:
          Bug#18187290 ISSUE WITH BUILDING MYSQL USING CMAKE 2.8.12
          
          We want to upgrade to VS2013 on Windows.
          In order to do this, we need to upgrade to cmake 2.8.12
          This has introduced some incompatibilities for .pdb files,
          and "make install" no longer works.
          
          To reproduce:
            cmake --build . --target package --config debug
          
          The fix:
          Rather than installing .pdb files for static libraries, we use the /Z7 flag
          to store symbolic debugging information in the .obj files.
------------------------------------------------------------
revno: 7968 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-review
timestamp: Wed 2014-05-07 17:06:55 +0200
message:
  empty merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.43
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Wed 2014-05-07 17:04:20 +0200
    message:
      Backport from trunk:
      Bug#18187290 ISSUE WITH BUILDING MYSQL USING CMAKE 2.8.12
      
      We want to upgrade to VS2013 on Windows.
      In order to do this, we need to upgrade to cmake 2.8.12
      This has introduced some incompatibilities for .pdb files,
      and "make install" no longer works.
      
      To reproduce:
        cmake --build . --target package --config debug
      
      The fix:
      Rather than installing .pdb files for static libraries, we use the /Z7 flag
      to store symbolic debugging information in the .obj files.
------------------------------------------------------------
revno: 7967
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Wed 2014-05-07 09:38:11 -0500
message:
  Code cleanup on 2 SQL syntax files.
  Make lex.h consistent by replacing tabs with spaces and lining up the code.
  Remove trailing spaces in noth files.
  
  Patch approved by Gleb Shchepa over email
------------------------------------------------------------
revno: 7966 [merge]
committer: Satya Bodapati <satya.bodapati@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-07 19:37:54 +0530
message:
  BUG#17798076 - BUG#16418661 CLEANUP: REMOVE UNNECESSARY BUF_FLUSH_LIST()
                 DURING RECOVERY
  
  Why and when the extra buf_flush_list() was added?
  ---------------------------------------------------
  The extra buf_flush_list() was added as part of
  Bug#16418661 CHANGING NAME IN FOR INNODB_DATA_FILE_PATH
  SHOULD NOT SUCCEED WITH LOG FILES.
  The testcase for rb#2293 failed randomly reporting inconsistent number
  of undo tablespaces.
  MTR didn't provide a way to use custom innodb_data_file_path,
  innodb_undo_tablespaces. See Bug#17059436.
  So the workaround is to use "--exec $MYSQLD --args my.cnf" on a empty
  data directory. This will create necessary files and aborts after that.
  
  The extra flush made sure that correct the number of undo tablespaces
  reported. This is not correct because the changes are already covered
  by redo.
  
  So the the ideal solution would be have been to run "--exec $MYSQLD"
  on the datadir again to apply redo or allow the bootstrapping to
  finish properly and exit.
  
  Fix:
  ----
  Remove uncessary flush. Fix the inconsistent number of undo
  tablespaces issue by passing '--skip-grant-tables
  --innodb-unkown-parameter' to $MYSQLD. This would allow the
  bootstrapping to finish and properly exit.
  
  Another side effect of the unnecessary flush is, it masked a problem
  in WL#7142:
  
  Side effect Issue:
  ------------------
  An MLOG_CHECKPOINT record for an earlier checkpoint can be mistaken as
  the MLOG_CHECKPOINT record for the latest checkpoint.
  
  Fix:
  ----
  The fix is to grow the MLOG_CHECKPOINT record size from 1 to 9 bytes,
  so that it will contains the checkpoint LSN. On the first scan of the
  redo log, we will ignore MLOG_CHECKPOINT records, unless the LSN
  matches the latest checkpoint.
  
  SIZE_OF_MLOG_CHECKPOINT: Grow from 1 to 9.
  
  mtr_t::commit_checkpoint(), recv_parse_log_recs(), recv_scan_log_recs():
  Add the parameter checkpoint_lsn.
  
  Approved by Marko, Kevin. rb#5234 rb#5320
    ------------------------------------------------------------
    revno: 3902.324.42
    committer: Satya Bodapati <satya.bodapati@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-07 19:36:11 +0530
    message:
      BUG#17798076 - BUG#16418661 CLEANUP: REMOVE UNNECESSARY BUF_FLUSH_LIST()
                     DURING RECOVERY
      
      Why and when the extra buf_flush_list() was added?
      --------------------------------------------------
      The extra buf_flush_list() was added as part of rb#2293.
      The testcase for rb#2293 failed randomly reporting inconsistent number
      of undo tablespaces. MTR didn't provide a way to use custom
      innodb_data_file_path, innodb_undo_tablespaces. See Bug#17059436 So
      the workaround is to use "--exec $MYSQLD --args my.cnf" on a empty data
      directory. This will create  necessary files and aborts after that.
      
      The extra flush made sure that correct the number of undo tablespaces
      reported. This is not correct because the changes are already covered by redo.
      
      So the the ideal solution would be have been to run "--exec $MYSQLD on
      the datadir" again to apply redo or allow the bootstrapping to finish
      properly and exit.
      
      Fix:
      ----
      Remove uncessary flush. Fix the inconsistent number of undo tablespaces
      issue by passing '--skip-grant-tables --innodb-unkown-parameter' to $MYSQLD.
      This would allow the bootstrapping to finish and properly exit.
      
      Approved by Marko, Kevin. rb#5234, rb#5320
------------------------------------------------------------
revno: 7965 [merge]
committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-07 16:58:50 +0530
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.41 [merge]
    committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-07 16:56:08 +0530
    message:
      Merge from 5.5 to 5.6
        ------------------------------------------------------------
        revno: 2875.592.33
        committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-05-07 16:55:03 +0530
        message:
          Fixing compilation error. Post push fix for Bug#17909656
------------------------------------------------------------
revno: 7964 [merge]
committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-07 15:58:29 +0530
message:
  Merge from 5.6 to 5.7
    ------------------------------------------------------------
    revno: 3902.324.40 [merge]
    committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-07 15:47:10 +0530
    message:
      Merge from 5.5 to 5.6
        ------------------------------------------------------------
        revno: 2875.592.32
        committer: Chaithra Gopalareddy <chaithra.gopalareddy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-05-07 14:59:23 +0530
        message:
          Bug#17909656  - WRONG RESULTS FOR A SIMPLE QUERY WITH GROUP BY
          
          Problem:
          If there is a predicate on a column referenced by MIN/MAX and
          that predicate is not present in all the disjunctions on
          keyparts earlier in the compound index, Loose Index Scan will
          not return correct result.
          
          Analysis:
          When loose index scan is chosen, range optimizer currently
          groups all the predicates that contain group parts separately
          and minmax parts separately. It therefore applies all the
          conditions on the group parts first to the fetched row.
          Then in the call to next_max, it processes the conditions
          which have min/max keypart.
          
          For ex in the following query:
          Select f1, max(f2) from t1 where (f1 = 10 and f2 = 13) or
          (f1 = 3) group by f1;
          Condition (f2 = 13) would be applied even for rows that
          satisfy (f1 = 3) thereby giving wrong results.
          
          Solution:
          Do not choose loose_index_scan for such cases. So a new rule
          WA2 is introduced to take care of the same.
          
          WA2: "If there are predicates on C, these predicates must
          be in conjuction to all predicates on all earlier keyparts
          in I."
          
          Todo the same, fix reuses the function get_constant_key_infix().
          Since this funciton will fail for all multi-range conditions, it
          is re-written to recognize that if the sub-conditions are
          equivalent across the disjuncts: it will now succeed.
          And to achieve this a new helper function is introduced called
          all_same().
          
          The fix also moves the test of NGA3 up to the former only
          caller, get_constant_key_infix().
------------------------------------------------------------
revno: 7963 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-05-07 14:37:38 +0530
message:
  Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
  
  Null merge from mysql-5.5
    ------------------------------------------------------------
    revno: 3902.324.39 [merge]
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-05-07 14:35:08 +0530
    message:
      Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
      
      Null merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.31
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-05-07 14:33:58 +0530
        message:
          Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
          
          Fixing post push failure
------------------------------------------------------------
revno: 7962
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: topush-trunk
timestamp: Wed 2014-05-07 10:35:42 +0200
message:
  post-push fix for test timeout related failures from bug#18330694.
------------------------------------------------------------
revno: 7961
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-06 17:50:13 +0200
message:
  Bug#18710679 INCREASE SERVER RESTART TIMEOUT FOR VALGRIND RUNS IN MTR
------------------------------------------------------------
revno: 7960
committer: horst.hunger@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-05-06 17:19:17 +0200
message:
  A core get lost. So, max_parts-innodb gets an own var dir.
------------------------------------------------------------
revno: 7959
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: b18330694-t
timestamp: Mon 2014-05-05 19:03:38 +0200
message:
  Bug#18330694: INNODB TESTCASES RUN TOO SLOW ON WINDOWS
  
  This fix contains 3 different parts
  
  1) Reason for mysqltest command shutdown server always waited the full
  timeout:
  During shutdown of the server there are still handles open to the process
  (by mtr?), 
  resulting in that it is possible to retrieve a new handle to the process,
  which leads to the current my_kill(pid, 0) windows implementation thinks
  the process is still alive, when it infact already exited.
  
  So it needs to either check with WaitForSingleObject or GetExitCodeProcess
  to verify if the process handle really has exited or not.
  
  2) Reason for spurious test failures when doing copy_file/remove_file/
  move_file:
  Even if the process signaled that it exited, Windows may not yet closed
  all the process files, resulting in failures (ERROR_SHARING_VIOLATION)
  during move/delete/copy. So to be more tolerant I added on retry in such
  cases.
  
  3) Killing the server if the shutdown takes too long time.
  Changing shutdown_server to:
  If timeout is non-zero, send shutdown to the server and wait until the
  server terminates or timeout is exceeded. If timeout is exceeded return
  error.
  If timeout is zero, then kill/terminate the server (kill -9 on Unix,
  TerminateProcess on Windows).
  This since killing the server if timeout is exceeded can have side-effects
  that is hard to predict. Perhaps we should never kill in shutdown_server
  (not allowing zero as timeout) and introduce kill_server
  (current shutdown_server 0 behavior)? I have not done this since it
  would need to update several test files.
  
  I also included a test change that I used during testing.
------------------------------------------------------------
revno: 7958
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-06 14:04:19 +0200
message:
  Bug#18709121: UNRESOLVED EXTERNAL SYMBOL _XMM@ WHEN LINKING 32 BIT BUILDS USING VS2013
  
  Add "__xmm@" to the list of compiler defined symbols not to be exported.
  These are SSE instruction set eqvivalents of "__real@" which we already
  don't export.
------------------------------------------------------------
revno: 7957 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug17588419_mysql-trunk
timestamp: Tue 2014-05-06 16:44:53 +0530
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.38
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug17588419_mysql-5.6
    timestamp: Tue 2014-05-06 16:41:02 +0530
    message:
      Bug#17588419: INSERT ON DUPLICATE KEY UPDATE FAILING AFTER
      MYSQL 5.6 UPGRADE.
      
      Problem:
      ========
      After upgrading a MySQL 5.5.23 slave to 5.6.14 (the master
      still runs 5.5.23) on executing INSERT ON DUPLICATE KEY
      following error is reported in SBR.
      
      [ERROR] Slave SQL: Error 'Auto-increment value in UPDATE
      conflicts with internally generated values' on query.
      
      Analysis:
      ========
      On master when user specifies an autoincrement value in
      multi insert statement the user given value is compared with
      current auto increment value and current + number of rows
      affected value. Whenever the user specified value falls
      within the given range ER_AUTO_INCREMENT_CONFLICT error is
      generated.
      
      
      On the slave the range is set to the value ULONGULONG_MAX
      by default. Whenever user specifies any value it will
      always fall within the above range and error gets generated.
      
      On slave each DML operation will result in n number of rows
      being manipulated. When number of manipulated rows is known
      the value range can be reset to currect value to current+n.
      
      Fix:
      ===
      On slave identify the number of rows being manipulated.
      Reset the value ranage from currnet value to current
      value + number of maninupated rows. If the number is not
      known go ahead with default values.
------------------------------------------------------------
revno: 7956 [merge]
committer: Mattias Jonsson <mattias.jonsson@oracle.com>
branch nick: topush-trunk
timestamp: Tue 2014-05-06 12:55:09 +0200
message:
  Merge of bug#17909699 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.37 [merge]
    committer: Mattias Jonsson <mattias.jonsson@oracle.com>
    branch nick: topush-5.6
    timestamp: Tue 2014-05-06 11:25:08 +0200
    message:
      Merge of bug#17909699 into mysql-5.6
        ------------------------------------------------------------
        revno: 2875.592.30
        committer: Mattias Jonsson <mattias.jonsson@oracle.com>
        branch nick: b17909699-55_2
        timestamp: Tue 2014-05-06 11:05:37 +0200
        message:
          Bug#17909699: WRONG RESULTS WITH PARTITION BY LIST COLUMNS()
          
          Typo leading to not including the last list values (partition).
          
          Also improved pruning to skip last partition if not used.
          
          rb#4762 approved by Aditya and Marko.
------------------------------------------------------------
revno: 7955
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-06 13:51:25 +0300
message:
  Fix Bug#18704384 regression (memory leak).
  
  dict_ind_init(): Do not create any objects for the removed dict_ind_compact.
------------------------------------------------------------
revno: 7954
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18486460inner_refs
timestamp: Tue 2014-05-06 12:10:10 +0200
message:
  Bug#18486460 ASSERTION FAILED: N < M_SIZE AFTER FIX_INNER_REFS
  
  The function st_select_lex::setup_ref_array allocates a Ref_ptr_array
  which is too small. This is a regression from
  WL#7200: True bottom-up server parser: refactoring of the SELECT statement
  
  Prior to that worklog, we would increment select->select_n_where_fields in 
  Item_field::Item_field(Name_resolution_context *context_arg, ...)
  
  The parser now uses Item_field::Item_field(const POS &pos, ...)
  immediately followed by Item_field::itemize(...)
  and this function failed to increment select->select_n_where_fields
  when the parsing context is CTX_SELECT_LIST.
------------------------------------------------------------
revno: 7953
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-06 11:03:37 +0300
message:
  Bug#18704384 REMOVE DICT_IND_COMPACT() AND SIMPLIFY PAGE_CREATE()
  
  page_create_low(): Use low-level byte access to initialize the page,
  instead of invoking the high-level methods.
  
  page_dir_get_nth_slot(): Add parenthesis to help the compiler perform
  constant folding, when calling the macro with a constant n.
  
  dict_ind_compact: Remove.
  
  rb#5322 approved by Kevin Lewis and Sunny Bains
------------------------------------------------------------
revno: 7952
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-review
timestamp: Tue 2014-05-06 09:41:41 +0200
message:
  Bug#18503515: UNION with set function in ORDER BY should be rejected
  
  A UNION query is not an aggregated query, and an ORDER BY applied to it
  cannot contain aggregate functions. This patch prevents such queries.
  
  A new error message is added: ER_AGGREGATE_ORDER_FOR_UNION.
  
  It was also found that nest_level is not correctly updated for
  fake_select_lex. This has been remedied, but to our knowledge, there
  were no error from this mistake.
------------------------------------------------------------
revno: 7951 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-05-06 11:32:34 +0530
message:
  Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
  
  Null merge from mysql-5.5
    ------------------------------------------------------------
    revno: 3902.324.36 [merge]
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-05-06 11:29:16 +0530
    message:
      Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
      
      Null merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.29
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-05-06 11:23:42 +0530
        message:
          Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
          
          Fixing post push failure
------------------------------------------------------------
revno: 7950
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 21:03:11 +0200
message:
    Follow-up patch for
    
      WL#7689 Deprecate and remove mysqlbug
      WL#7826 Deprecate and remove mysql_zap and mysql_waitpid
    
    Remove removal of man pages from packaging scripts.
------------------------------------------------------------
revno: 7949 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 22:29:32 +0530
message:
  Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.35 [merge]
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-05-05 22:26:08 +0530
    message:
      Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
      
      Merging fix from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.28
        committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-05-05 22:22:15 +0530
        message:
          Bug#17638477 UNINSTALL AND INSTALL SEMI-SYNC PLUGIN CAUSES SLAVES TO BREAK
          
          Problem: Uninstallation of semi sync plugin causes replication to
          break.
          
          Analysis: A semisync enabled replication is mutual agreement between
          Master and Slave when the connection (I/O thread) is established.
          Once I/O thread is started and if semisync is enabled on both
          master and slave, master appends special magic header to events
          using semisync plugin functions and sends it to slave. And slave
          expects that each event will have that special magic header format
          and reads those bytes using semisync plugin functions.
          
          When semi sync replication is in use if users execute
          uninstallation of the plugin on master, slave gets confused while
          interpreting that event's content because it expects special 
          magic header at the beginning of the event. Slave SQL thread will
          be stopped with "Missing magic number in the header" error.
          
          Similar problem will happen if uninstallation of the plugin happens
          on slave when semi sync replication is in in use. Master sends
          the events with magic header and slave does not know about the
          added magic header and thinks that it received a corrupted event.
          Hence slave SQL thread stops with "Found  corrupted event" error.
          
          Fix: Uninstallation of semisync plugin will be blocked when semisync
          replication is in use and will throw 'ER_UNKNOWN_ERROR' error.
          To detect that semisync replication is in use, this patch uses
          semisync status variable values.
           > On Master, it checks for 'Rpl_semi_sync_master_status' to be OFF
              before allowing the uninstallation of rpl_semi_sync_master plugin.
              >> Rpl_semi_sync_master_status is OFF when
                  >>> there is no dump thread running
                  >>> there are no semisync slaves
           > On Slave, it checks for 'Rpl_semi_sync_slave_status' to be OFF
              before allowing the uninstallation of rpl_semi_sync_slave plugin.
              >> Rpl_semi_sync_slave_status is OFF when
                 >>> there is no I/O thread running
                 >>> replication is asynchronous replication.
------------------------------------------------------------
revno: 7948 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Mon 2014-05-05 16:46:23 +0200
message:
  Empty merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.34 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Mon 2014-05-05 16:41:44 +0200
    message:
      Backport from trunk:
      Bug #18593044 COMPILE FLAGS NOT PASSED TO DTRACE, BREAKS CROSS BUILD
        ------------------------------------------------------------
        revno: 2875.592.27
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5-merge
        timestamp: Mon 2014-05-05 16:39:14 +0200
        message:
          Backport from trunk:
          Bug #18593044 COMPILE FLAGS NOT PASSED TO DTRACE, BREAKS CROSS BUILD
------------------------------------------------------------
revno: 7947
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: group-by
timestamp: Mon 2014-05-05 14:52:28 +0200
message:
  Bug#18035906 TEST_IF_SKIP_SORT_ORDER INCORRECTLY CHOSES
               NON-COVERING INDEX FOR ORDERING
  
  Follow-up patch to stabilize test output: Adding
  ANALYZE TABLE to increase likelihood of updated
  index statistics.
------------------------------------------------------------
revno: 7946
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 16:47:13 +0530
message:
  Bug #17449901	 TABLE DISAPPEARS WHEN ALTERING WITH
  			FOREIGN KEY CHECKS OFF
  
  Analysis:
  By removing the FK index with out removing the foreign key constraint,
  we are moving the table into inconsistent state. FK index lookup happens
  during load table process. So it fails to open the table when FK index
  is missing.
  
  Fix:
  Don't allow to drop an index if it is needed for an FK constraint,
  even if foreign_key_checks is disabled.
  
  		Approved by Jimmy (rb-5274)
------------------------------------------------------------
revno: 7945 [merge]
committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 12:42:58 +0200
message:
  Empty version change upmerge
    ------------------------------------------------------------
    revno: 3902.324.33 [merge]
    author: murthy.narkedimilli@oracle.com
    committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-05-05 12:33:33 +0200
    message:
      Raise version number after cloning 5.5.38 and upmerge the 5.5 build
        ------------------------------------------------------------
        revno: 2875.592.26
        author: murthy.narkedimilli@oracle.com
        committer: Murthy Narkedimilli <murthy.narkedimilli@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-05-05 12:15:12 +0200
        message:
          Raise version number after cloning 5.5.38
------------------------------------------------------------
revno: 7944
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 16:28:59 +0800
message:
  Fixed bug#18644435 ASSERT 0 IN RTR_PAGE_SPLIT_AND_INSERT(), GIS
  
  This assertion should be removed, since we have re-split now.
  This patch also diabled valgrind test on some of gis test cases,
  since it will take a long time.
  
  Approved by Jimmy on IM.
------------------------------------------------------------
revno: 7943
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 10:05:58 +0200
message:
  Bug#18694238 REMOVE UNDOCUMENTED SCRIPT SUPPORT-FILES/BINARY-CONFIGURE.SH FROM 5.7
------------------------------------------------------------
revno: 7942
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-winwarn
timestamp: Mon 2014-05-05 09:42:38 +0200
message:
  Bug#11752665: WINDOWS SERVER X64: SO MANY COMPILER WARNINGS
  
  Patch #7:
  
  This patch reduces the compiler warning count on Win64 by an additional
  200, as reported by VS2013.
------------------------------------------------------------
revno: 7941
committer: Dmitry Shulga <Dmitry.Shulga@oracle.com>
branch nick: mysql-trunk-bug18068253
timestamp: Mon 2014-05-05 11:06:33 +0700
message:
  This is a patch for Bug#18068253 - XA START WITH THE EXIST XID, AND THEN
  XA COMMIT WITH THE XID WILL FAILED.
  
  When a xa transaction is prepared and the server crashed, then after restart the 
  MySQL server,  first, do XA START with the same xid, says:" XAER_DUPID: The 
  XID already exists", that's OK, but do XA COMMIT with this xid, says that:
  XAER_RMFAIL: The command cannot be executed when global transaction is 
  in the  NON-EXISTING state. So, the prepared transaction can't be committed. 
  
  The reason for the bug was that the method eq() of class xid didn't take into
  account the value of xid::format during comparison. First time when XA  START
  is executed the server found prepared transaction and return error but for all
  of that the value of thd->transaction.xid_state.xid is reset assigning the
  value -1 to the the attribute xid.format. Later when the statement XA COMMIT
  is executed comparison of  thd->transaction.xid_state.xid with xid value
  specified in the statement is made without taking into account
  the value of format.
------------------------------------------------------------
revno: 7940
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Mon 2014-05-05 10:49:23 +0800
message:
  Bug 18368241 - ASSERT FILE != -1 FIL_FUSIONIO_ENABLE_ATOMIC_WRITE()
  DICT_CHECK_TABLESPACES_AND_
  
  We need to check if the Datafile is already opened before calling
  fil_fusionio_enable_atomic_write, so that the assert wouldn't fail.
  
  Patch provided by Sunny, Approved by Jimmy via IM.
------------------------------------------------------------
revno: 7939
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-05-02 12:34:20 -0500
message:
  Win32 nd Win64 compiler warning fixes.  
  This is replacement for r7886 which was undone in r7909.
  Approved by Mattias Jonsson in RB#5249
------------------------------------------------------------
revno: 7938
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Wed 2014-04-30 15:36:58 +0200
message:
  Bug#18590516 TESTCASE MULTI_PLUGIN_LOAD BROKEN
  
  Testcase multi_plugin_load is always skipped due to the way it's written:
  Revert previous patch, since that include file was used by other tests.
  Write the new check in multi_plugin_load.test instead.
------------------------------------------------------------
revno: 7937
committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-02 10:41:32 +0530
message:
  Added "--source include/not_valgrind.inc" so that the tests skip on valgrind. These tests take too much time on valgrind. 
------------------------------------------------------------
revno: 7936
committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-05-02 10:07:28 +0530
message:
  Removed the file ./mysql-test/include/have_daemon_example_plugin.inc.moved
------------------------------------------------------------
revno: 7935 [merge]
committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
branch nick: trunk-build
timestamp: Wed 2014-04-30 21:08:08 +0400
message:
  Null merge from 5.6. The bug will be fixed in scope of 18497612.
    ------------------------------------------------------------
    revno: 3902.324.32 [merge]
    tags: clone-5.6.19-build
    committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
    branch nick: 5.6-build
    timestamp: Wed 2014-04-30 20:53:43 +0400
    message:
      Manual merge to 5.6.
        ------------------------------------------------------------
        revno: 2875.592.25
        tags: clone-5.5.38-build
        committer: Alexander Nozdrin <alexander.nozdrin@oracle.com>
        branch nick: 5.5-build
        timestamp: Wed 2014-04-30 20:48:29 +0400
        message:
          Patch for Bug#18511348 (DDL_I18N_UTF8 AND DDL_I18N_KOI8R
          ARE PERMANENTLY SKIPPED IN 5.5/5.6).
          
          The problem was that some result files were not updated,
          so the tests were skipped.
          
          The fix is to record updated result files.
------------------------------------------------------------
revno: 7934 [merge]
committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
branch nick: trunk
timestamp: Thu 2014-05-01 09:12:24 +0530
message:
  Bug#18596756 - FAILED PREPARING OF TRIGGER ON TRUNCATED TABLES
                 CAUSE ERROR 1054.
  
  Merging from 5.6 to trunk.
    ------------------------------------------------------------
    revno: 3902.324.31
    committer: Praveenkumar Hulakund <praveenkumar.hulakund@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-05-01 09:10:19 +0530
    message:
      Bug#18596756 - FAILED PREPARING OF TRIGGER ON TRUNCATED TABLES
                     CAUSE ERROR 1054.
      
      Analysis:
      --------
      Issue here is, the re-parse of the 'SELECT' query in the 'SET'
      statement for the NEW field of a row (triggered due to a DDL
      operation), removes current object and creates a new object for
      the "NEW.<field>" information. But the binding between the new
      object and the actual field in the table is not set. Hence while
      setting the value, an "unknown column" error is reported.
      
      For the SET operation with NEW.<field>, object of type
      "sp_instr_set_trigger_field" is created while parsing the trigger.
      This object has the member "m_trigger_field" of type
      "Item_trigger_field"(Represents NEW/OLD field of row). This
      trigger is bound with actual field of the row by calling method
      "Item_trigger_field::fix_field".
      
      During the cleanup process, performed before the re-parse of the 
      "SELECT" query of "SET" operation, "Item_trigger_field" 
      field held by "m_trigger_field" is unlinked. After the "SELECT"
      query is re-parsed, the "Item_trigger_field" for "NEW.id" is created
      and held with "m_trigger_field" member of "sp_instr_set_trigger_field"
      object. But it is not bound with the actual field object of table. 
      Hence an "unknown column" is reported. ?
      
      Fix:
      ---
      Modified code to bind the "Item_trigger_field" created for "NEW.id"
      to the Field object in table by calling "Item_trigger_field::fix_field"
      method.
------------------------------------------------------------
revno: 7933
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Wed 2014-04-30 12:58:04 +0200
message:
  Bug#18663095 GET RID OF DYNAMIC_ARRAY IN THD
  
  Our array abstractions should be removed, and substituted with modern data
  structures.
  
  Templatized vectors are:
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 7932 [merge]
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-30 16:22:49 +0530
message:
  Bug #17156940 : THE UPDATE AND SELECT BEHAVE DIFFERENTLY
                  UNDER THE SAME CONDITIONS.
  Merge from Mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.30
    committer: mithun <mithun.c.y@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-30 16:20:10 +0530
    message:
      Bug #17156940 : THE UPDATE AND SELECT BEHAVE DIFFERENTLY
                      UNDER THE SAME CONDITIONS.
      ISSUE   :
      In myisam, Suppose in btree index of varchar type we have
      keys with and without trailing spaces example keys 'abc',
      'abc ', 'abc '. During index search based on same key,
      length of the lastkey get changed if lastkey read is the
      one with trailing spaces. If x is the length of key 'abc'.
      After reading 'abc ' length will be x + 1. And,
      last_rkey_length should be recalculated as accordingly
      whenever lastkey changes.
      But in function mi_rnext_same during BTREE search when we
      tried to copy lastkey to lastkey2 we have used non updated
      last_rkey_length even though lastkey and its length might
      have been changed as explained above. And, last_rkey_length
      is computed only once during mi_rkey. Because of this
      invalid length usage compare_key failed and scan got
      terminated. And, hence Update command ended before updating
      further tuples which will satisfy the condition.
      
      SOLUTION :
      In function mi_rnext_same the input key lastkey2 can remain
      constant if we use a separate buffer other than lastkey2.
      And, we can fill this key just for one time for the complete
      scan in mi_rnext_same. Since key and length is computed only
      once for the entire scan, issues related to invalid length
      during compare keys will not arise.
------------------------------------------------------------
revno: 7931 [merge]
committer: Bill Qu <bill.qu@Oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-30 18:43:56 +0800
message:
  Bug #17305385 	I_RPL.RPL_GROUP_COMMIT_WITH_SESSION_ATTACH_ERROR FAILS ON SOLARIS
  
  The slave IO thread will try to connect master server and report
  an error with Error_code: 2003, after the master server is
  shut down when simulating session attach error. PB2 tree will start
  up mutiple processes to run all tests in parallel in general. So
  the problem will happen when one process is checking if their is
  an error on slave, when executing 'rpl_end.inc' and the slave
  thread in another process is tring to connect master server and
  report an error with Error_code: 2003.
  
  To solve the problem, we stop slave IO thread before master server
  is shut down, and start the slave IO thread after the master server
  is started, then the slave IO thread does not report the error
  any more when connecting to master server.
    ------------------------------------------------------------
    revno: 3902.324.29
    committer: Bill Qu <bill.qu@Oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-30 18:39:19 +0800
    message:
      Bug#17305385 	I_RPL.RPL_GROUP_COMMIT_WITH_SESSION_ATTACH_ERROR FAILS ON SOLARIS
        
      The slave IO thread will try to connect master server and report
      an error with Error_code: 2003, after the master server is
      shut down when simulating session attach error. PB2 tree will start
      up mutiple processes to run all tests in parallel in general. So
      the problem will happen when one process is checking if their is
      an error on slave, when executing 'rpl_end.inc' and the slave
      thread in another process is tring to connect master server and
      report an error with Error_code: 2003.
        
      To solve the problem, we stop slave IO thread before master server
      is shut down, and start the slave IO thread after the master server
      is started, then the slave IO thread does not report the error
      any more when connecting to master server.
------------------------------------------------------------
revno: 7930
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-30 12:21:19 +0200
message:
  Fix Bug #18619945 - AFTER OPTIMIZE TABLE A SELECT ON INNODB SPATIAL
  INDEX CRASHES THE SERVER 
  
  rb://5219 approved by Sunny Bains
------------------------------------------------------------
revno: 7929 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Wed 2014-04-30 10:26:48 +0200
message:
  Push to mysql-trunk
    ------------------------------------------------------------
    revno: 7867.2.3 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18259356
    timestamp: Wed 2014-04-30 10:19:42 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 7867.2.2 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18259356
    timestamp: Tue 2014-04-29 18:14:00 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 7867.2.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18259356
    timestamp: Thu 2014-04-17 14:52:26 +0200
    message:
      Bug#18259356 - PROCESSLIST_ID FOR SLAVE SQL THREAD IS MISSING
      
      For historical reasons, background replication threads are
      assigned a connection_id(), and displayed in:
      - SHOW PROCESSLIST
      - table INFORMATION_SCHEMA.PROCESSLIST.
      
      In the performance schema, these threads are also displayed,
      but with a NULL PROCESSLIST_ID.
      
      Having the same thread displayed:
      - with a PROCESSLIST_ID (in the information_schema)
      - without a PROCESSLIST_ID (in the performance_schema)
      is inconsistent, and is a bug.
      
      With this fix, the ID assigned in the server code
      is also given to the instrumentation,
      so that column PROCESSLIST_ID for replication threads
      is correctly populated in table performance_schema.threads.
------------------------------------------------------------
revno: 7928
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Tue 2014-04-29 14:25:15 +0200
message:
  Bug#18590516 TESTCASE MULTI_PLUGIN_LOAD BROKEN
  
  Testcase multi_plugin_load is always skipped due to the way it's written:
  Fix:
  In have_plugin_auth.inc check for the presence of qa_auth_server instead.
------------------------------------------------------------
revno: 7927
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Tue 2014-04-29 14:21:02 +0200
message:
  Bug #18593044 COMPILE FLAGS NOT PASSED TO DTRACE, BREAKS CROSS BUILD
  
  RHEL7/Oracle Linux 7 comes in 64 bits variants only, however at
  least libraries are required in 32 bits versions too.  Going
  forward there might be need for 32 bits clients.
   
  RHEL/OL7 supports cross build of 32 bits binaries on 64 bits platform.
   
  This works just fine for MySQL server also, however when dtrace
  is enabled build fails.
   
  Due to the fact that dtrace on Linux is running gcc and reading
  CFLAGS from environment only.
  (Just reading last part of /usr/bin/dtrace to verify this).
   
  When cross building a 32 bits server package, flags includes
  -m32, however this is not passed to dtrace, which builds a 64
  bits probes_mysql.o, which cause link error by mixing 32 of 64
  bits object files.
------------------------------------------------------------
revno: 7926 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-29 17:22:53 +0530
message:
  Bug #17026898 PREVIOUS GTID EVENT IS NOT WRITTEN WHEN BINLOG IS ROTATED VIA
  SIGHUP
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.28
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-29 17:03:20 +0530
    message:
      Bug #17026898 PREVIOUS GTID EVENT IS NOT WRITTEN WHEN BINLOG IS ROTATED VIA
      SIGHUP
      
      Problem: When Binlog is rotated via SIGHUP signal, the newly generated binlog
      does not contain previous gtid event which is very important for processing
      that binlog's gtid events later.
      
      If there are any transactions written to this binlog then on next restart,
      while server is processing available binary logs, it was failing with
      following error:
      "The first global transaction identifier was read, but no other information
      regarding identifiers existing on the previous log files was found."
      and the server refuses to start.
      
      Or If the new GTID transactions which were written to this new binlog
      are replicated, Slave gets confused after seeing a GTID event without 
      a previous_gtid_event and enters into "Fatal 1236" error.
      
      Analysis: SIGHUP siganl causes the server to reload the grant tables
      and to flush tables, logs, the thread cache, and the host cache.
      As part of flush logs, server rotates binary log as well.
      
      When server receives SIGHUP signal, it calls reload_acl_and_cache and
      which eventually executes the following code to write PREVIOUS_GTID_EVENT.
      if (current_thd && gtid_mode > 0)
      {
      /* write previous gtid event */
      }
      
      And current_thd is NULL when server reaches this code through signal handler.
      Hence the newly generated binary log is not containing previous gtid event
      which resulted in reported issue at the time of restart.
      
      Fix: If reload_acl_and_cache() is called from SIGHUP handler, then
      allocate temporary THD before execution of binary log rotation function.
      The same above problem can be seen with relay log as well. Hence this
      temporary THD will be allocated even before relay log rotation function.
      And delete the THD object after finishing the task.
------------------------------------------------------------
revno: 7925
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-joh-string
timestamp: Tue 2014-04-29 13:47:26 +0200
message:
  Bug#18525978: COMBINE CLIENT/SQL_STRING AND SQL/SQL_STRING
  
  This patch removes client/sql_string.cc and moves sql/sql_string.cc to
  sql-common so that it can be used by both client and server.
  
  This removes code duplication. It turns out that the differences between
  client/sql_string.cc and sql/sql_string.cc were mostly due to bugfixes only
  applied to the latter (Bug#56492, Bug#44793, WL#1213, Bug#24293).
------------------------------------------------------------
revno: 7924 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-04-29 18:15:44 +0800
message:
  Bug #18634201 - Upgrade from 5.6.10 to 5.6.16 crashes and leaves unausable DB
  
  When user tries to upgrade from 5.6.10 to latest 5.6, the rename of aux
  tables could make server crash, since we don't get rid of those obsolete
  tables, which were removed away since 5.6.11.
  
  The patch fixed 2 issues:
  1) We shouldn't try to rename those obsolete tables.
  2) We should try to drop those obsolete tables when upgrade, so that those
     obsolete tables would not be left in the data directory.
  
  rb#5202, approved by Jimmy.
    ------------------------------------------------------------
    revno: 3902.324.27
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-29 18:06:44 +0800
    message:
      Bug #18634201 - Upgrade from 5.6.10 to 5.6.16 crashes and leaves unausable DB
      
      When user tries to upgrade from 5.6.10 to latest 5.6, the rename of aux
      tables could make server crash, since we don't get rid of those obsolete
      tables, which were removed away since 5.6.11.
      
      The patch fixed 2 issues:
      1) We shouldn't try to rename those obsolete tables.
      2) We should try to drop those obsolete tables when upgrade, so that those
         obsolete tables would not be left in the data directory.
      
      rb#5202, approved by Jimmy.
------------------------------------------------------------
revno: 7923 [merge]
committer: Marc Alff <marc.alff@oracle.com>
branch nick: mysql-trunk-push
timestamp: Tue 2014-04-29 11:22:09 +0200
message:
  Merge to mysql-trunk
    ------------------------------------------------------------
    revno: 7867.1.3 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18304086
    timestamp: Tue 2014-04-29 11:18:37 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 7867.1.2 [merge]
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18304086
    timestamp: Tue 2014-04-29 11:04:45 +0200
    message:
      Local merge
    ------------------------------------------------------------
    revno: 7867.1.1
    committer: Marc Alff <marc.alff@oracle.com>
    branch nick: mysql-trunk-bug18304086
    timestamp: Thu 2014-04-17 11:04:28 +0200
    message:
      Bug#18304086 - P_S DIGEST LOOKS WRONG FOR SYSTEM VARIABLES, SHOWN AS @ @ VARIABLE
      
      Before this fix, a query containing "@@variable",
      when printing the query digest text,
      was printed as "@ @ variable".
      
      This is a bug, because the syntax "@ @ variable" is actually not accepted
      by the server.
      
      "@@variable" is parsed as 3 tokens, '@', '@' and 'variable'.
      When printing the digest text, a space is added to separate each token,
      causing the resulting string.
      
      The root cause of the bug is in the lexer:
      if '@' was a real token, "@ @ variable" would also be legal,
      and the server would also accept the "@ @ variable" syntax,
      so returning '@' as a token alone is incorrect.
      
      Now, independently of the lexer, query digest text printed could use
      a better formatting in general, so the token description has been expanded
      to have an append_space flag, so that a space is not systematically added
      for each token.
      
      This fix does not change the lexer in any way.
      This fix only prints the '@' without a trailing space,
      in the query digest text.
------------------------------------------------------------
revno: 7922
committer: Sivert Sorumgard <sivert.sorumgaard@oracle.com>
branch nick: trunk-7593
timestamp: Tue 2014-04-29 10:37:07 +0200
message:
  WL#7593: New data dictionary: don't hold LOCK_open while
  reading table definition
  
  While retrieving a table definition when getting a table
  share, the LOCK_open mutex may be temporarily released.
  This allows a higher degree of concurrency when opening
  tables, and is needed for the new data dictionary where
  dictionary tables may need to be opened recursively.
  
  The following changes have been implemented:
  
  1. A new TABLE_SHARE member m_open_in_progress is introduced
  to flag that a share is being initialized. Concurrent threads
  opening the same share must check this flag to see whether
  they must wait for the share to be initialized.
  
  2. A condition variable COND_open is introduced for threads
  that need to wait for a share which is being opened.
  
  3. We must wait for the share to be completely opened
  before returning from get_cached_table_share().
  
  4. There are loops iterating over all elements in the
  table definition cache. One loop is in close_cached_tables(),
  this is safe to keep as is because shares being opened are
  handled. The other loop is in list_open_tables(), here we found
  it better to ignore shares being opened to avoid situations
  where we would incorrectly present a table as being open
  (e.g. if the opening actually failed).
  
  5. We must assert that the newly inserted share is actually
  present in the cache before returning from get_table_share().
  
  6. Shares being opened (m_open_in_progress == true) in
  close_cached_connection_tables() are skipped. Otherwise, due
  to the way TABLE_SHARE::connect_string is initialized in
  open_binary_frm(), there is a risk that connect_string.length
  is initialized while connect_string.str is not, hence
  making us read into uninitialized memory further down in
  close_cached_connection_tables().
  
  Thus, in theory, there is a risk that shares are left in the
  cache that should really be closed (matching the submitted
  connection string), and this risk is already present since
  LOCK_open is unlocked before calling
  close_cached_connection_tables().
  
  Now, close_cached_connection_tables() is called as the final
  step of DROP/ALTER SERVER, so its goal is to flush all tables
  which were open before DROP/ALTER SERVER started. Thus, if a
  share gets opened after close_cached_connection_tables() has
  started, the information about the server has already been
  updated, so the new table will use the new definition of the
  server.
  
  It might have been an issue, however if one thread started
  opening a federated table, read the old server definition into a
  share, and then a switch to another thread doing ALTER SERVER
  happened right before setting m_open_in_progress to false for
  the share. Because in this case ALTER SERVER would not flush
  the share opened by the first thread as it should have been. But
  luckily, server definitions affected by * SERVER statements are
  not read into TABLE_SHARE structures, but are read when we
  create the TABLE object in ha_federated::open().
  
  This means that in close_cached_connection_tables(), ignoring
  shares that are in the process of being opened is safe, because
  such shares don't have TABLE objects associated with them yet.
------------------------------------------------------------
revno: 7921
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt5
timestamp: Mon 2014-04-28 19:53:41 +0200
message:
  n tests: routine update of bug signatures and error messages,
    minor changes to server parameters in default.tcnf
------------------------------------------------------------
revno: 7920
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt5
timestamp: Mon 2014-04-28 19:40:07 +0200
message:
  Implement enhancement 18631086:
    by default, enable all workload features in n workload ibtest_ctd_w.pl
  ibtest_ctd_w.pl:
  * enable the following options by default: --ufk, --utm, --umd, --ugis,
    to test, respectively,
    foreign keys, temporary tables, multiple data directories, spatial data.
  * remove option --uall, previously used to enable most of the above.
  * increase the default of --slaves, from 4 to 7.
  * make --rd the default
  * add option --ugtpm N
    to specify the probability of using a GIS table in the workload loop
  rb://5223 approved by Vin
------------------------------------------------------------
revno: 7919 [merge]
committer: mithun <mithun.c.y@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-28 21:13:42 +0530
message:
  Bug #18167356: EXPLAIN W/ EXISTS(SELECT* UNION SELECT*)
                 WHERE ONE OF SELECT* IS DISTINCT FAILS.
  NULL Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.324.26 [merge]
    committer: mithun <mithun.c.y@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-04-28 21:09:44 +0530
    message:
      Bug #18167356: EXPLAIN W/ EXISTS(SELECT* UNION SELECT*)
                     WHERE ONE OF SELECT* IS DISTINCT FAILS.
      
      NULL MERGE from 5.5
        ------------------------------------------------------------
        revno: 2875.592.24
        committer: mithun <mithun.c.y@oracle.com>
        branch nick: mysql-5.5
        timestamp: Mon 2014-04-28 21:07:27 +0530
        message:
          Bug #18167356: EXPLAIN W/ EXISTS(SELECT* UNION SELECT*)
                         WHERE ONE OF SELECT* IS DISTINCT FAILS.
          ISSUE:
          ------
          There are 2 issues related to explain union.
          1. If we have subquery with union of selects. And, one of
             the select need temp table to materialize its results
             then it will replace its query structure with a simple
             select from temporary table. Trying to display new
             internal temporary table scan resulted in crash. But to
             display the query plan, we should save the original
             query structure.
          2. Multiple execution of prepared explain statement which
             have union of subqueries resulted in crash. If we have
             constant subqueries, fake select used in union operation
             will be evaluated once before using it for explain.
             During first execution we have set fake select options to
             SELECT_DESCRIBE, but did not reset after the explain.
             Hence during next execution of prepared statement during
             first time evaluation of fake select we had our select
             options as SELECT_DESCRIBE this resulted in improperly
             initialized data structures and crash.
          
          SOLUTION:
          ---------
          1. If called by explain now we save the original query
             structure. And this will be used for displaying.
          2. Reset the fake select options after it is called for
             explain of union.
------------------------------------------------------------
revno: 7918
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Mon 2014-04-28 16:43:25 +0200
message:
  Followup to vasil.dimov@oracle.com-20140428124411-5q3o4pfy85husey7
  
  In the merged unit test: include univ.i before other files,
  because other files may pull in stdint.h without defining __STDC_LIMIT_MACROS
------------------------------------------------------------
revno: 7917
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Mon 2014-04-28 16:27:27 +0200
message:
  Bug#18521389 CMAKE -DFEATURE_SET=XSMALL FAILS WITH MTR BOOTSTRAPPING
  
  When building with FEATURE_SET=xsmall (or small or classic) mysql-test-run.pl
  will not succeed in bootstrapping the test due to default-storage-engine is
  InnoDB which is not included.
  
  Fix: remove feature sets, enable "xlarge"/"community" by default.
------------------------------------------------------------
revno: 7916 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-17994219
timestamp: Mon 2014-04-28 19:48:23 +0530
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.25 [merge]
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-17994219
    timestamp: Mon 2014-04-28 19:37:33 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.592.23
        committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
        branch nick: mysql-5.5-17994219
        timestamp: Mon 2014-04-28 16:28:09 +0530
        message:
          BUG#17994219: CREATE TABLE .. SELECT PRODUCES INVALID STRUCTURE,
                        BREAKS RBR
          
          Analysis:
          --------
          A table created using a query of the format:
          CREATE TABLE t1 AS SELECT REPEAT('A',1000) DIV 1 AS a;
          breaks the Row Based Replication.
          
          The query above creates a table having a field of datatype
          'bigint' with a display width of 3000 which is beyond the
          maximum acceptable value of 255.
          
          In the RBR mode, CREATE TABLE SELECT statement is
          replicated as a combination of CREATE TABLE statement
          equivalent to one the returned by SHOW CREATE TABLE and
          row events for rows inserted. When this CREATE TABLE event
          is executed on the slave, an error is reported:
          Display width out of range for column 'a' (max = 255)
          
          The following is the output of 'SHOW CREATE TABLE t1':
          CREATE TABLE t1(`a` bigint(3000) DEFAULT NULL)
                            ENGINE=InnoDB DEFAULT CHARSET=latin1;
          
          The problem is due to the combination of two facts:
          
          1) The above CREATE TABLE SELECT statement uses the display
             width of the result of DIV operation as the display width
             of the column created without validating the width for out
             of bound condition.
          2) The DIV operation incorrectly returns the length of its first
             argument as the display width of its result; thus allowing
             creation of a table with an incorrect display width of 3000
             for the field.
          
          Fix:
          ----
          This fix changes the DIV operation implementation to correctly
          evaluate the display width of its result. We check if DIV's
          results estimated width crosses maximum width for integer
          value (21) and if yes set it to this maximum value.
          
          This patch also fixes fixes maximum display width evaluation
          for DIV function when its first argument is in UCS2.
------------------------------------------------------------
revno: 7915
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-28 15:44:11 +0300
message:
  Followup to vasil.dimov@oracle.com-20140428083718-ysum5udw25eosie6:
  
  in mem0mem-t.cc - include univ.i before handler.h and gtest/gtest.h
  because they may pull stdint.h without defining __STDC_LIMIT_MACROS
  before that.
------------------------------------------------------------
revno: 7914
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Mon 2014-04-28 11:05:53 +0200
message:
  Bug#18187290 ISSUE WITH BUILDING MYSQL USING CMAKE 2.8.12
  
  We want to upgrade to VS2013 on Windows.
  In order to do this, we need to upgrade to cmake 2.8.12
  This has introduced some incompatibilities for .pdb files,
  and "make install" no longer works.
  
  To reproduce:
    cmake --build . --target package --config debug
  
  The fix:
  Rather than installing .pdb files for static libraries, we use the /Z7 flag
  to store symbolic debugging information in the .obj files.
------------------------------------------------------------
revno: 7913
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-28 11:37:18 +0300
message:
  Remove ib_int64_t, partial fix for Bug#16923874 REPLACE CUSTOM INNODB
  INTEGER TYPES WITH STANDARD ONES DEFINED IN C99
  
  * Replace ib_int64_t with int64_t
  * Use the standard PRId64 when printing via printf(3) family of functions
------------------------------------------------------------
revno: 7912
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-28 12:39:24 +0530
message:
  Bug #16963396 INNODB: USE OF LARGE EXTERNALLY-STORED FIELDS MAKES CRASH
  RECOVERY LOSE DATA
  
  Post push fix.  The newly added assert is failing for the tests
  innodb_wl6501_crash_2 and innodb_wl6501_crash_7.  Removing the assert.
  This will be tracked using a separate bug.
  
  Approved by sunny over IM.
------------------------------------------------------------
revno: 7911
committer: Prabeen Pradhan <prabeen.pradhan@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-28 11:01:43 +0530
message:
  It is failing on daily. Will be disabled till bug 16963396 is fixed.
------------------------------------------------------------
revno: 7910 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18158639
timestamp: Sat 2014-04-26 10:15:56 +0530
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.324.24
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18158639
    timestamp: Sat 2014-04-26 09:03:58 +0530
    message:
      BUG#18158639: MATERIALIZED CURSOR + FLUSH TABLES CRASH WHEN
                    FETCHING VARIABLE
      
      Analysis:
      --------
      Concurrent execution of a stored program having cursor and
      a FLUSH TABLE operation may cause the 'mysqld' to crash.
      The crash can be observed, when the events occurs in the
      following sequence:
      
      a) Cursor OPEN executes the SELECT statement for which the
         table is opened.
      b) Flush table operation is triggered and finds that a table
         share object is present. Its version is marked as zero to
         ensure that the share is removed when it is no longer
         referenced.
      c) Since the share version is old and is referenced, the
         flush table operation waits until the flush request is
         granted.
      d) The SELECT statement execution for cursor OPEN closes all
         tables except the internal temporary table used by cursor
         for saving the materialized records.
      e) While closing the table, since the table share is an old
         version and there is a pending flush request, the flush
         request is granted. Thus the table share is deleted by
         awakening the FLUSH TABLE operation.
      f) During the cursor FETCH operation, the column type is checked
         for field conversion. To perform the check, the table share of
         orig_table in the field definition of the cursor temporary
         table is accessed. Since the share was deleted by FLUSH
         operation, accessing the invalid memory may cause the
         server to crash.
      
      Fix:
      ---
      In case of cursors, since all tables other than the temporary
      table are closed, the orig_table in the field definition for
      the internal temporary table is set to NULL. This is done 
      once the metadata of the temporary table for the CURSOR is 
      sent.
------------------------------------------------------------
revno: 7909
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-04-25 13:43:07 -0500
message:
  Bug#18646983 - DAILY-TRUNK PERFORMANCE TESTS "CRASHING" MYSQLD FROM 2014-04-23 BUILD
  This was isolated to r7886 which i a few static_casts to fix Windows copiler warnings.
  A review of these changes did not reveal any issues and the it was not easy to identify
  the source of the problem. So this patch reverts the entire r7886 patch.
------------------------------------------------------------
revno: 7908
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-cmake
timestamp: Fri 2014-04-25 16:36:49 +0200
message:
  Bug#18599982: HAVE_DECL_FDATASYNC & HAVE_FEDISABLEEXCEPT
                CMAKE CONFIGURE CHECKS ARE BROKEN
  
  The problem was that the results of the HAVE_DECL_FDATASYNC and
  HAVE_FEDISABLEEXCEPT CMake configure checks were not stored
  in my_config.h due to them being absent from the config.h.cmake
  template.
  
  This meant that code relying on these checks was in fact broken.
  The affected code was used to prefer fdatasync() over fsync()
  and to disable floating point exceptions.
  
  The patch fixes the problem by updating config.h.cmake so
  that the results of these two checks are saved.
  
  The patch also reorganizes config.h.cmake to make it easier to
  match the file with configure.cmake and other CMake files where
  checks are defined. In addition, the patch removes some duplicate
  CMake checks and some unused items from config.h.cmake.
------------------------------------------------------------
revno: 7907
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-review
timestamp: Fri 2014-04-25 16:17:14 +0200
message:
  Bug#18561062: MYSQL-TRUNK REVNO 7736 CAUSED A -2% ~ -3% 
                PERFORMANCE REGRESSION (OLTP_RO)
  
  The regression is caused by calculating the condition filtering
  effect for simple single table queries where condition filtering
  has no effect.
  
  Whether or not condition filtering shall be calculated is based
  on a number of conditions. One of these conditions is whether
  join buffering may be used for the table.
  
  Optimize_table_order::calculate_scan_cost() called 
  calculate_condition_filter() with a wrong parameter that indicated
  that join buffering was used even when that was not the case.
  
  Fixed by passing "join_buffering_is_used" parameter instead of
  unconditional 'true'.
------------------------------------------------------------
revno: 7906
committer: Guilhem Bichot <guilhem.bichot@oracle.com>
branch nick: bef
timestamp: Fri 2014-04-25 14:54:21 +0200
message:
  Fix for Bug #18486607 ASSERTION FAILED:
  IN_SUBQ_PRED->LEFT_EXPR->FIXED IN CONVERT_SUBQUERY_TO_SEMIJOIN .
  Query:
   select # S1
       (
        select 1 # S2
        from t2
        where a in (
                              select 1 # S3
                              from t2
                              )
       ) from a group by a ;
  
  we prepare S1 (the "first SELECT" from the top). Do setup_fields() on it.
  This does fix_fields() on subquery containing S2. Which does prepare() on S2,
  thus setup_conds(), thus fix_fields() which does two things:
  1) fixes left arg "a" of "a in (select S3)"; this arg being an outer reference (to grouping column of S1)
  it's not quite fixed but rather added to S1::inner_refs list.
  2) fix_fields() on subquery containing S3, which calls resolve_subquery() on S3
     => S3 is candidate for sj-merging in S2.
  Back to preparation of S2: flatten_subqueries() runs, to merge S3, asserts that
  left arg is fixed and it's not.
  It would have been fixed only after fix_inner_refs() for S1.
  This didn't crash before WL 7082, because back then, sj was merged in JOIN::optimize() i.e. after all JOIN::prepare() are done, which included fix_inner_refs().
  Now sj is merged at end of preparation of the sj's owner (=parent of the subuery).
  The fix: relax the assertion, as in the fixes for:
  Bug 14601664 ASSERTION FAILED: *REF && (*REF)->FIXED, FILE ITEM.CC
  and
  Bug 13735050: CRASH IN END_WRITE WHEN SUBQUERY
  GROUPS BY OUTER FIELDS.
------------------------------------------------------------
revno: 7905
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-25 16:48:09 +0300
message:
  Bug#18651616 UT_AD(0) IN FIL_NAMES_WRITE() IN
  IBUF_MERGE_OR_DELETE_FOR_PAGE(), DROP TABLE
  
  A debug assertion can fail in fil_space_lookup() when a page I/O
  completion thread is performing a change buffer merge on a
  single-table tablespace while DROP TABLE is executing.
  
  The fix is to remove the debug assertion and to never skip writing a
  MLOG_FILE_NAME record due to space->stop_new_ops. There is no possible
  race condition thanks to the interlocking of space->stop_new_ops and
  space->n_pending_ops.
  
  fil_space_free_low(): Update a comment.
  
  fil_names_write(): Remove the debug assertion and expand the comment.
  
  rb#5244 approved by Yasufumi Kinoshita
------------------------------------------------------------
revno: 7904
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-dynarray
timestamp: Fri 2014-04-25 14:43:41 +0200
message:
  Bug#18636972 GET RID OF SORT_DYNAMIC IN SQL_SHOW
  
  Use standard algorithms, rather than dynamic_array.
  standard library functions are
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 7903
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-25 14:26:16 +0200
message:
  Bug#18651087 REMOVE UNNEEDED CATALOGUE SUPPORT-FILES/RHEL4-SELINUX
------------------------------------------------------------
revno: 7902
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-18035906
timestamp: Fri 2014-04-25 10:34:42 +0200
message:
  Bug#18035906: TEST_IF_SKIP_SORT_ORDER INCORRECTLY CHOSES 
                NON-COVERING INDEX FOR ORDERING
  
  test_if_skip_sort_order() didn't check if an index scan had 
  earlier been chosen and would therefore possibly replace it
  with a different index scan. If the first index scan was 
  covering and the one chosen by test_if_skip_sort_order() was not,
  a worse plan would be the result.
  
  Fixed by making test_if_skip_sort_order() index scan aware.
------------------------------------------------------------
revno: 7901
committer: Vamsikrishna Bhagi <vamsikrishna.bhagi@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-25 12:24:06 +0530
message:
  Bug #18659533 MYSQL_SECURE_INSTALLATION SKIPS REST OF
                OPTIONS ON FINDING UNKNOWN OPTIONS
                                                              
  Problem: While reading the start up options,
           mysql_secure_installation ignores any option after
           it encounters one it can't recognize. It is
           resulting in few options getting missed out.
  
  Solution: A flag is added to the option handling function
            according to which the function decides if it
            should continue reading options after encountering
            an invalid one, or abort.
------------------------------------------------------------
revno: 7900
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-invector
timestamp: Thu 2014-04-24 15:12:24 +0200
message:
  Bug#18556403 USE STD LIBRARY FOR SORTING AND MATCHING IN SUBCLASSES OF IN_VECTOR, PART 2
  
  As demonstrated by the patch for Bug#18486249
  in_vector and its derived classes can benefit from some code modernization.
  
  standard library functions are
   - type safe
   - easier to read/maintain
   - faster
  
  Introduce typed vectors of objects, rather than the "blob vector" in the base class in_vector.
  The re-write uncovered a bug in the range optimizer:
  it was building range predicates for unused parts of the IN-vector.
------------------------------------------------------------
revno: 7899
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-24 16:18:37 +0530
message:
  Bug #16963396 INNODB: USE OF LARGE EXTERNALLY-STORED FIELDS MAKES CRASH
  RECOVERY LOSE DATA
  
  Post push fix.  Fixing a valgrind leak issue reported in pb2.  Free the
  allocated memory in the error path.
  
  Approved by sunny over IM.
------------------------------------------------------------
revno: 7898 [merge]
committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-24 12:18:59 +0200
message:
  - Merge from 5.6 => trunk
  - Support for enterprise packages
  - Upgrade from MySQL-* packages
  - Fix Cflags for el7
    ------------------------------------------------------------
    revno: 3902.324.23 [merge]
    committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-04-24 11:56:18 +0200
    message:
      Merge from 5.5 - Updated for 5.6.18
        ------------------------------------------------------------
        revno: 2875.592.22 [merge]
        committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-04-24 11:06:02 +0200
        message:
          - Support for enterprise packages
          - Upgrade from MySQL-* packages
          - Fix Cflags for el7
            ------------------------------------------------------------
            revno: 2875.593.8
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.37-repo
            timestamp: Mon 2014-04-07 16:36:09 +0200
            message:
              updated optflags variable and cmake option for debug build
            ------------------------------------------------------------
            revno: 2875.593.7
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.37-repo
            timestamp: Mon 2014-04-07 14:51:44 +0200
            message:
              Fix Cflags for el7
            ------------------------------------------------------------
            revno: 2875.593.6
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.37-repo
            timestamp: Fri 2014-04-04 05:58:49 +0200
            message:
              Changed permisison for filter-requires.sh and filter-provides.sh
            ------------------------------------------------------------
            revno: 2875.593.5
            committer: Balasubramanian Kandasamy <balasubramanian.kandasamy@oracle.com>
            branch nick: mysql-5.5.37-repo
            timestamp: Thu 2014-04-03 12:56:26 +0200
            message:
              Support for enterprise packages
------------------------------------------------------------
revno: 7897
committer: Raghav Kapoor <raghav.kapoor@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-24 14:09:12 +0530
message:
  BUG#18484649 - STRICT MODE + CURSOR FETCH INTO WRONG TYPE,
                 ASSERTION IN PROTOCOL::END_STATEMENT
  
  BACKGROUND:
  An assertion failure in 5.7 in Protocol::end_statement()
  occurred when we try to fetch a large value (int) which
  cannot fit into a variable (tiny int) declared inside
  stored routine through a cursor. However, it works fine
  in 5.6.
  Also, This assertion failure happens only in debug builds,
  We do not get any assertion failure/crash on release builds,
  Because asserts are not active on release builds.
  
  ANALYSIS:
  On analysis, it is found that error comes in
  Materialized_cursor::fetch() function. But its return type
  is void. Therefore, the calls to upper level functions assumed
  everything went OK with it (however, it resulted in error),
  so it resulted in incorrect value of err_status in
  sp_head::execute() function.
  The documentation for
  virtual bool execute(THD *thd, uint *nextp) = 0;
  function in sp_instr.h says
  
      @return Error status.
  
  In this case execute() did not perform as documented - it has
  reported an error but still returned false.
  
  FIX:
  The fix is to set err_status variable correctly (to true if
  it is known that some error has occurred or the query was
  killed i.e thd->killed is set).
  The return type of function Materialized_cursor::fetch() is
  modified from void to bool, so that it reports the error to
  upper level functions correctly. This fix results in err_status
  variable inside sp_head::execute() to be set correctly
  after the execution of statement and perform according to
  the documentation.
------------------------------------------------------------
revno: 7896
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-24 16:15:46 +0800
message:
  Fixed test cases gis, innodb_gis_rtree_rollback1 failure on pb2.
  
  Approved by Jimmy on IM.
------------------------------------------------------------
revno: 7895
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-17903490
timestamp: Thu 2014-04-24 10:15:16 +0400
message:
  Fix for bug #17903490 "ASSERT FAILURE IN MDL.H WHEN DROP FUNCTION CALLED
  WITH INVALID FUNCTION NAME".
  
  DROP FUNCTION statement using too long function name caused assertion
  failure when executed against debug server. There was no problem for
  production builds.
  
  MDL subsystem asserts that names of databases and objects which passed
  to it has at most NAME_LEN bytes. It is responsibility of caller to check
  that database/object name satisfies this limit. Unlike other code,
  DROP FUNCTION implementation was not checking if function name satisfies
  this limit, so assert in MDL subsystem fired.
  
  Since in addition to asserting that database/object names have certain
  maximum length MDL subsystem also truncates too lengthy database/object
  names there was no problem in production builds.
------------------------------------------------------------
revno: 7894 [merge]
committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
branch nick: mysql-trunk-18080920
timestamp: Thu 2014-04-24 09:41:19 +0530
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.22 [merge]
    committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
    branch nick: mysql-5.6-18080920
    timestamp: Thu 2014-04-24 09:32:56 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6
        ------------------------------------------------------------
        revno: 2875.592.21
        committer: Nisha Gopalakrishnan <nisha.gopalakrishnan@oracle.com>
        branch nick: mysql-5.5-18080920
        timestamp: Thu 2014-04-24 09:30:21 +0530
        message:
          BUG#18080920: CRASH; MY_REALLOC_STR DEREFERENCES NEGATIVE VALUE 
                        INTO CLIENT_ERRORS ARRAY
                        
          Analysis:
          --------
          The client may crash while executing a statement due to
          the missing mapping of the server error to it's equivalent
          client error.
          
          When trying to reallocate memory for the packet buffer, if
          the system is out of memory or the packet buffer is large,
          the server errors 'ER_OUT_OF_RESOURCES' or 'ER_PACKET_TOO_LARGE'
          is returned respectively. The client error number calculated is
          negative and when trying to dereference the array of client 
          error messages with the calculated error number, the client
          crashes.
          
          Fix:
          ----
          Map the server error returned to it's equivalent client error
          prior to dereferencing the array of client error messages.
          
          Note: Test case is not added since it is difficult to simulate
          the error condition.
------------------------------------------------------------
revno: 7893 [merge]
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-merge
timestamp: Wed 2014-04-23 17:06:41 +0200
message:
  NULL merge 5.6 => trunk
    ------------------------------------------------------------
    revno: 3902.324.21 [merge]
    committer: Tor Didriksen <tor.didriksen@oracle.com>
    branch nick: 5.6-merge
    timestamp: Wed 2014-04-23 17:04:55 +0200
    message:
      merge 5.5 => 5.6
        ------------------------------------------------------------
        revno: 2875.592.20
        committer: Tor Didriksen <tor.didriksen@oracle.com>
        branch nick: 5.5-merge
        timestamp: Wed 2014-04-23 17:01:35 +0200
        message:
          Backport from trunk:
            Bug#18396916 MAIN.OUTFILE_LOADDATA TEST FAILS ON ARM, AARCH64, PPC/PPC64
            
            The recorded results for the failing tests were wrong.
            They were introduced by the patch for
            Bug#30946 mysqldump silently ignores --default-character-set when used with --tab
            
            Correct results were returned for platforms where 'char' is implemented as unsigned.
            This was reported as 
            Bug#46895 Test "outfile_loaddata" fails (reproducible)
            Bug#11755168 46895: TEST "OUTFILE_LOADDATA" FAILS (REPRODUCIBLE)
            The patch for that bug fixed only parts of the problem,
            leaving the incorrect results in the .result file.
            
            Solution: use 'uchar' for field_terminator and line_terminator on all platforms.
            Also: remove some un-necessary casts, leaving the ones we actually need.
------------------------------------------------------------
revno: 7892
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-23 18:45:35 +0530
message:
  Bug #16963396 INNODB: USE OF LARGE EXTERNALLY-STORED FIELDS MAKES CRASH
  RECOVERY LOSE DATA
  
  Problem:
  
  When too-large blob fields are used, InnoDB overwrites its most recent
  checkpoint in its redo logs.  
  
  Solution:
  
  While writing large blobs, call log_free_check() periodically to ensure that
  enough free space is there in the redo logs. 
  
  During an update operation, when a non-updated blob column is moved off page
  then the update vector is correctly updated before writing the undo log.
  This way special handling for update operation is avoided.
  
  rb#4862 approved by Marko.  Yasufumi also contributed (rb#5175).
------------------------------------------------------------
revno: 7891 [merge]
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk-clean
timestamp: Wed 2014-04-23 14:04:25 +0200
message:
  Remove utilities that have been deprecated in 5.6:
  
  WL#7689 Deprecate and remove mysqlbug
  WL#7826 Deprecate and remove mysql_zap and mysql_waitpid
    ------------------------------------------------------------
    revno: 3902.324.20
    committer: Erlend Dahl <erlend.dahl@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-23 13:43:28 +0200
    message:
      5.6 version of patch for
      
      WL#7689 Deprecate and remove mysqlbug
      WL#7826 Deprecate and remove mysql_zap and mysql_waitpid 
------------------------------------------------------------
revno: 7890 [merge]
committer: Igor Solodovnikov <igor.solodovnikov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-23 12:53:47 +0300
message:
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.19 [merge]
    committer: Igor Solodovnikov <igor.solodovnikov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-23 12:47:09 +0300
    message:
      Merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.19
        committer: Igor Solodovnikov <igor.solodovnikov@oracle.com>
        branch nick: mysql-5.5
        timestamp: Wed 2014-04-23 12:46:00 +0300
        message:
          Bug #17514920	MYSQL_THREAD_INIT() CALL WITHOUT MYSQL_INIT() IS CRASHING IN WINDOWS
          
          It is error to call mysql_thread_init() before libmysql is initialized with mysql_library_init(). Thus to fix this bug we need to detect if library was initialized and return error result if mysql_thread_init() is called with uninitialized library.
          
          Fixed by checking my_thread_global_init_done and returning nonzero if the library is not initialized.
------------------------------------------------------------
revno: 7889
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-23 15:22:44 +0800
message:
  Fixed the test case innodb_gis_alter_spatial_index sporadically failure
  on pb2.
  
  Approved by Jimmy on IM.
------------------------------------------------------------
revno: 7888
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-23 09:53:52 +0300
message:
  Bug#18544317 ASSERT IN IBUF_MERGE_OR_DELETE_FOR_PAGE BUF_PAGE_GET_GEN
  
  Fix a regression introduced in WL#7142.
  
  ibuf_merge_or_delete_for_page(): Even if block==NULL, do update
  the bitmap, in case update_ibuf_bitmap==true.
  
  Original patch provided by Michael Izioumtchenko, adjusted by me.
  Approved by Jimmy Yang.
------------------------------------------------------------
revno: 7887
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-23 09:51:39 +0300
message:
  Remove an invalid sentence from a comment that was introduced in WL#7142.
  
  During a log checkpoint, there may exist dirty blocks in the buffer pool.
  What matters is that before the checkpoint completes, we will have
  flushed all blocks up to the checkpoint LSN, and no blocks with a newer LSN.
  
  In the event of a crash, we will recover the system as it was at the
  checkpoint LSN, after applying any redo log that has been
  successfully written since the latest checkpoint.
------------------------------------------------------------
revno: 7886
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 16:52:54 -0500
message:
  Fix more Windows 32 and 64 bit compiler warnings introduced by revision 7874
  WL6968  InnoDB GIS: R-tree index support.
  WL6609: InnoDB GIS: Support Predicate Locking for GIS index
  WL6745: InnoDB GIS: Support DML for InnoDB GIS index
------------------------------------------------------------
revno: 7885
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 14:47:43 -0500
message:
  Make declaration of page_get_ssn_id() exactly like definition.
------------------------------------------------------------
revno: 7884
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 13:48:59 -0500
message:
  Fix recent Windows compiler warnings in InnoDB
------------------------------------------------------------
revno: 7883
committer: Dmitry Lenev <Dmitry.Lenev@oracle.com>
branch nick: mysql-trunk-6987-dead-code
timestamp: Tue 2014-04-22 19:53:18 +0400
message:
  Follow-up patch for WL#6987 "Refactoring: Remove TABLE object from updatable
  views".
  
  After this worklog has been implemented we no longer set
  TABLE_LIST::schema_table member for table elements representing
  mergeable views. As result code handling table elements with
  both TABLE_LIST::schema_table and TABLE_LIST::view members set
  in open_and_process_table() became dead. This patch replaces this
  code with assert.
------------------------------------------------------------
revno: 7882 [merge]
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 16:31:55 +0300
message:
  Null-merge mysql-5.6 -> mysql-trunk
  (fixed compilation error only reported in mysql-5.6)
    ------------------------------------------------------------
    revno: 3902.324.18
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-22 16:28:49 +0300
    message:
      Followup to vasil.dimov@oracle.com-20140409172641-jf52d93p1f1yo1x1:
      
      Add an explicit typecast, needed to fix a Windows 32 only compilation error:
      
      ...\os0once.h(89): error C2664: '_InterlockedCompareExchange' : cannot convert parameter 1 from 'volatile os_once::state_t *' to 'volatile long *'
      Types pointed to are unrelated; conversion requires reinterpret_cast, C-style cast or function-style cast
------------------------------------------------------------
revno: 7881 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 18:36:53 +0530
message:
  Bug#18069107 SLAVE CRASHES WITH GTIDS,TEMP TABLE, STOP IO_THREAD, START SLAVE
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.17
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-22 18:32:55 +0530
    message:
      Bug#18069107 SLAVE CRASHES WITH GTIDS,TEMP TABLE, STOP IO_THREAD, START SLAVE
      
      Fixing post push pb2 failure
------------------------------------------------------------
revno: 7880
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt
timestamp: Tue 2014-04-22 14:52:00 +0200
message:
  Bring up to date with n test development branch.
  Reviewed by: Vin.
------------------------------------------------------------
revno: 7879
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-review
timestamp: Tue 2014-04-22 14:41:33 +0200
message:
  WL#6635 - stabilize tests
------------------------------------------------------------
revno: 7878
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 12:15:35 +0200
message:
  Add order by to innodb_gis.innodb_gis_alter_spatial_index.test to avoid
  random result ordering mismatch
------------------------------------------------------------
revno: 7877
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-uintkorr
timestamp: Tue 2014-04-22 09:47:30 +0200
message:
  Bug#18557876 MOVE COMMON CODE OUT OF BYTE_ORDER_GENERIC[_X86].H
  
  Several of the functions in
    include/byte_order_generic.h
    include/byte_order_generic_x86.h
  are identical.
  Move them to include/my_byteorder.h
  
  Also: remove some un-necessary casting.
------------------------------------------------------------
revno: 7876
committer: Allen lai <zheng.lai@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-22 09:59:15 +0800
message:
  Fixed a UNIV_DEBUG issue found by Michael in GIS merge.
------------------------------------------------------------
revno: 7875
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-21 09:43:29 +0200
message:
  Update some files' copyright year
------------------------------------------------------------
revno: 7874
committer: Jimmy Yang <jimmy.yang@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-21 04:28:39 +0200
message:
  Check in WL #6968  InnoDB GIS: R-tree index support. This is the
  implementation to support Spatial Index in InnoDB. It is also
  umbralla worklog for following 2 worklogs:
  
  #WL 6609: InnoDB GIS: Support Predicate Locking for GIS index
  #WL 6745: InnoDB GIS: Support DML for InnoDB GIS index
  
  rb://4289 approved by Yasufumi and Bin
------------------------------------------------------------
revno: 7873 [merge]
committer: Nuno Carvalho <nuno.carvalho@oracle.com>
branch nick: mysql-trunk
timestamp: Sun 2014-04-20 20:20:13 +0100
message:
  BUG#18432744: RPL_CORRUPTION TEST FAILING ON 5.6
  
  Merge from mysql-5.6 into mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.16
    committer: Nuno Carvalho <nuno.carvalho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Sun 2014-04-20 20:11:56 +0100
    message:
      BUG#18432744: RPL_CORRUPTION TEST FAILING ON 5.6
      
      rpl_corruption test was sporadically failing with a different slave
      error code then expected on PB2 when it was run with gtid-mode=ON.
      
      When gtid-mode is ON, and in particular when MASTER_AUTO_POSITION=1,
      slave informs master which transactions it already retrieved and
      applied so that master (dump thread) can move forward on binary logs
      until it reaches the missing transactions and send them to slave.
      
      The injected event corruption through corrupt_read_log_event
      debug flag was also corrupting Previous_gtids_log_event and
      Gtid_log_event events disallowing dump thread to walk through the
      binary logs what was causing the unexpected
      ER_MASTER_FATAL_ERROR_READING_BINLOG error.
      
      Also the injected corruption was only corrupting the first read
      event from binary log, since dump thread was skipping the
      transactions already sent to slave the corrupted read event was not
      sent to slave.
      
      Fixed the failure by:
       1) Excluding Previous_gtids_log_event and Gtid_log_event events
      from the injected corruption.
       2) Corrupting all read events when corrupt_read_log_event debug
      flag is set.
------------------------------------------------------------
revno: 7872 [merge]
committer: Venkata Sidagam <venkata.sidagam@oracle.com>
branch nick: trunk
timestamp: Fri 2014-04-18 16:25:49 +0530
message:
  Bug #17235179 OPTIMIZE AFTER A DELETE RETURNS ERROR 0 CAN'T
                GET STAT OF MYD FILE
  
  Merging from mysql-5.6 to trunk
    ------------------------------------------------------------
    revno: 3902.324.15
    committer: Venkata Sidagam <venkata.sidagam@oracle.com>
    branch nick: 5.6
    timestamp: Fri 2014-04-18 16:22:06 +0530
    message:
      Bug #17235179 OPTIMIZE AFTER A DELETE RETURNS ERROR 0 CAN'T
                    GET STAT OF MYD FILE
      
      Description: Impossible to neither REPAIR nor OPTIMIZE .MYD files bigger 
      than 4GB. Only the Windows versions are affected.
      
      Analysis: In my_copystat() function while calling stat() system call for
      Windows it is unable to get the file information for files greater than
      4GB. Hence stat() fails with error and OPTIMIZE table on 4GB ".MYD" files  
      fails. And the corresponding .TMD file(temporary .MYD file) is not getting
      deleted.
      
      Fix: Now we are calling windows specific stat() system call i.e _stati64()
      instead of stat() for windows and stat() call for other OS's.
------------------------------------------------------------
revno: 7871 [merge]
committer: Igor Solodovnikov <igor.solodovnikov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-17 16:46:45 +0300
message:
  Merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.14 [merge]
    committer: Igor Solodovnikov <igor.solodovnikov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-04-17 16:35:23 +0300
    message:
      Merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.18
        committer: Igor Solodovnikov <igor.solodovnikov@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-04-17 16:33:55 +0300
        message:
          Bug #18053212	MYSQL_GET_SERVER_VERSION() CALL WITHOUT A VALID CONNECTION RESULTS IN SEG FAULT
          
          When there is no connection mysql_get_server_version() will return 0
          and report CR_COMMANDS_OUT_OF_SYNC error.
------------------------------------------------------------
revno: 7870
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B16102788-trunk
timestamp: Tue 2014-04-15 12:34:29 +0300
message:
  Bug #16102788: INDETERMINATE BEHAVIOR DUE TO EMPTY OPTION VALUES
  
  Currently mysql programs silently accept empty strings as option values,
  e.g. "--port=" will work the same as "--port=0" and -P "" will work the same as
  -P 0.
  This is unintuitive and can cause a lot of confusion.
  Fixed the options collection code to throw an error if the explicit argument to 
  an numeric or enum option is empty string.
  
  Test case added.
------------------------------------------------------------
revno: 7869 [merge]
committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-16 22:39:52 +0530
message:
  Bug#18069107 SLAVE CRASHES WITH GTIDS,TEMP TABLE, STOP IO_THREAD, START SLAVE
  
  Merging fix from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.324.13
    committer: Venkatesh Duggirala<venkatesh.duggirala@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-16 22:36:55 +0530
    message:
      Bug#18069107 SLAVE CRASHES WITH GTIDS,TEMP TABLE, STOP IO_THREAD, START SLAVE
      
      Problem:
      When Slave SQL thread detects that the master was restarted
      with the help of information sent by master's 'Format Description'
      event, slave SQL thread drops all the opened temporary tables inorder
      to have proper cleanup.
      
      When GTID mode is on and while slave SQL thread is generating 
      DROP TEMPORARY statement for all these temporary tables, server is
      hitting an assert DEBUG_ASSERT(gtid.spec_type != UNDEF_GROUP).
      
      Analysis: 
      When the server was cleaning up a threadd (client disconnected
      for example) it was setting up thd's GTID_NEXT variable to
      AUTOMATIC at THD::cleanup just before calling
      close_temporary_tables(thd) function in order to generate proper
      GTID events for any 'DROP TEMPORARY' query created. No problem here.
      
      When slave SQL thread applies a FD/Start_log_event_v3 supposed be from
      a master that had just (re)started, it assumes that the master
      doesn't has the temporary tables anymore. So, slave SQL thread needs to
      do a cleanup of the current temporary tables by calling 
      close_temporary_tables(thd).
      
      Slave SQL thread always starts with GTID_NEXT set to AUTOMATIC. So,
      when the first FD event sent by the master after slave SQL thread
      started was calling close_temporary_tables(thd), it had GTID_NEXT set
      to AUTOMATIC, not hitting the assert.
      
      When slave SQL thread applies transactions with GTID, at the time of
      commit/rollback of that transaction, GTID_NEXT is set to UNDEF_GROUP.
      Any thread with it's GTID_NEXT set to UNDEF_GROUP cannot generate
      Gtid events. Hence, after applying some transactions, slave SQL thread
      was hitting this assert when having to close temporary tables if it
      detected that the master had restarted.
      
      Fix:
      Moved the code that set GTID_NEXT to AUTOMATIC from THD::cleanup() to
      close_temporary_tables(THD *thd).
------------------------------------------------------------
revno: 7868
committer: Yasufumi Kinoshita <yasufumi.kinoshita@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-16 17:54:40 +0900
message:
  Bug#18598813 : MASTER THREAD CAUSES FAILING ASSERTION: SRV_GET_ACTIVE_THREAD_TYPE() == SRV_NONE
  
  Approved by Marko over IM.
------------------------------------------------------------
revno: 7867 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug17942050_mysql-trunk
timestamp: Tue 2014-04-15 15:35:25 +0530
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.12 [merge]
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug17942050_mysql-5.6
    timestamp: Tue 2014-04-15 15:26:56 +0530
    message:
      Merge from mysql-5.5 to mysql-5.6.
        ------------------------------------------------------------
        revno: 2875.592.17
        committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
        branch nick: Bug17942050_mysql-5.5
        timestamp: Tue 2014-04-15 15:17:25 +0530
        message:
          Bug#17942050:KILL OF TRUNCATE TABLE WILL LEAD TO BINARY LOG
          WRITTEN WHILE ROWS REMAINS
          
          Problem:
          ========
          When truncate table fails while using transactional based
          engines even though the operation errors out we still
          continue and log it to binlog. Because of this master has
          data but the truncate will be written to binary log which
          will cause inconsistency.
          
          Analysis:
          ========
          Truncate table can happen either through drop and create of
          table or by deleting rows. In the second case the existing
          code is written in such a way that even if an error occurs
          the truncate statement will always be binlogged. Which is not
          correct.
          
          Binlogging of TRUNCATE TABLE statement should check whether
          truncate is executed "transactionally or not". If the table
          is transaction based we log the TRUNCATE TABLE only on
          successful completion.
          
          If table is non transactional there are possibilities that on
          error we could have partial changes done hence in such cases
          we do log in spite of errors as some of the lines might have
          been removed, so the statement has to be sent to slave.
          
          Fix:
          ===
          Using table handler whether truncate table is being executed
          in transaction based mode or not is identified and statement
          is binlogged accordingly.
------------------------------------------------------------
revno: 7866
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-15 11:59:15 +0200
message:
  Stop running rpl.rpl_optimize in valgrind as it creates havoc in the weekly testing.
  
  The test is in any case experimental for the time being.
  
  Cf. Bug#18128323
  
  Approved by Anitha over IM.
------------------------------------------------------------
revno: 7865 [merge]
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-15 12:23:58 +0300
message:
  WL#6068 Remove InnoDB rec_per_keys hack
  
  Merge mysql-trunk-wl6068 -> mysql-trunk
    ------------------------------------------------------------
    revno: 5107.38.16 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Tue 2014-04-15 12:23:25 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.15 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Mon 2014-04-14 19:14:25 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.14 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Mon 2014-04-14 17:19:51 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.13
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Mon 2014-04-14 17:15:17 +0300
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Make sure that InnoDB does not return to the Optimizer rec_per_key
      values < 1.0 and also that it returns rec_per_key == 1.0 if the
      table is empty.
    ------------------------------------------------------------
    revno: 5107.38.12
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Thu 2014-04-10 16:06:12 +0300
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Add a comment explaining that index->stat_n_diff_key_vals[] and
      index->table->stat_n_rows could have been calculated at a different time.
      
      Remove a redundant comment - the same info is in a comment further below
      that starts with "/* The code below is legacy and should be"
    ------------------------------------------------------------
    revno: 5107.38.11 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Thu 2014-04-10 15:17:19 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.10 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Wed 2014-04-09 13:40:49 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.9
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Wed 2014-04-09 13:39:36 +0300
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Reword a comment, so that it is obvious and easy to delete the legacy code.
    ------------------------------------------------------------
    revno: 5107.38.8
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Wed 2014-04-09 13:32:08 +0300
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Use index->table->stat_n_rows for "number of rows in the table" insead of
      stats.records because the former is more up to date and thus possibly more
      in-sync with the stats that are read up from the dict_index_t object.
      
      index->table->stat_n_rows is calculated at the same time as
      index->stat_n_diff_key_vals[]. index->table->stat_n_rows is also adjusted
      on each DML.
    ------------------------------------------------------------
    revno: 5107.38.7 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Fri 2014-04-04 12:30:20 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.6
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Tue 2014-04-01 21:28:24 +0300
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Use the newly introduced type rec_per_key_t in InnoDB internal calculations
      involving rec_per_key.
    ------------------------------------------------------------
    revno: 5107.38.5 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Tue 2014-04-01 21:17:14 +0300
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.4
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Fri 2014-03-07 14:27:56 +0200
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Set the floating point rec_per_key from ha_innobase::info() as per the new
      interface added in "WL#7338 Interface for improved records per key estimates"
    ------------------------------------------------------------
    revno: 5107.38.3 [merge]
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Fri 2014-03-07 13:25:00 +0200
    message:
      Merge mysql-trunk -> mysql-trunk-wl6068
    ------------------------------------------------------------
    revno: 5107.38.2
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Fri 2014-03-07 13:15:41 +0200
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Revert the patch that ships n_distinct[] to the MySQL/Optimizer layer.
      Instead we will ship a floating point rec_per_key[] array.
    ------------------------------------------------------------
    revno: 5107.38.1
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-trunk-wl6068
    timestamp: Thu 2012-12-13 14:51:21 +0200
    message:
      WL#6068 Remove InnoDB rec_per_keys hack
      
      Introduce a new array KEY::n_distinct[], similar to KEY::rec_per_key[].
      
      Introduce new methods (in class handler) for access to the new
      KEY::n_distinct[]:
      
      void handler::set_n_distinct(key number, prefix, n_distinct value);
      void handler::set_n_distinct(key number, whole n_distinct array);
      n_distinct value handler::get_n_distinct(key number, prefix);
      
      In InnoDB - set the KEY::n_distinct[] array from ha_innobase::info() and
      override handler::get_n_distinct() in ha_innobase::get_n_distinct() to
      actually use the new array KEY::n_distinct[] instead of calculating the
      result via KEY::rec_per_key[] like done by default.
------------------------------------------------------------
revno: 7864
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-15 12:21:32 +0300
message:
  Non-functional change: remove duplicate prototype of
  dict_index_remove_from_cache() from dict0dict.h and
  convert the comments to the new @param style.
------------------------------------------------------------
revno: 7863
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-14 18:46:43 +0300
message:
  Reversal path of bug BUG#16102788. Needs more work.
------------------------------------------------------------
revno: 7862
committer: Vamsikrishna Bhagi <vamsikrishna.bhagi@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-14 21:03:52 +0530
message:
  Bug #18405677 CRASH WHEN SETTING SESSION_TRACK_SYSTEM_VARIABLES
  
  Problem: Server crashes when @@session_track_system_variables
           is set with a blank value under some special
           conditions.
  
  Solution: The key value of the hash member is not updated when
            session variable is changed in the current code.
            Used appropriate hash functions in the patch to update
            the key.
------------------------------------------------------------
revno: 7861
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B16102788-trunk
timestamp: Mon 2014-04-14 16:52:07 +0300
message:
  Bug #16102788: INDETERMINATE BEHAVIOR DUE TO EMPTY OPTION VALUES
  
  Currently mysql programs silently accept empty strings as option values,
  e.g. "--port=" will work the same as "--port=0" and -P "" will work the same as
  -P 0.
  This is unintuitive and can cause a lot of confusion.
  Fixed the options collection code to throw an error if the explicit argument to 
  an option is empty string and the option is not handled programatically and 
  does have an option value pointer to be set.
  
  Test case added.
------------------------------------------------------------
revno: 7860
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-14 16:07:26 +0300
message:
  Revert a test change due to
  Bug#18592693 INNOCHECKSUM.EXE DOES NOT EXPAND FILENAME WILDCARDS
------------------------------------------------------------
revno: 7859
committer: Shivji Jha <shivji.jha@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-04-14 12:29:00 +0530
message:
  Bug#18338203: P_S REPLICATION_CONNECTION_STATUS.SOURCE_UUID                      
                UNINITIALIZED VALUE                                                
                                     
  Problem:
  The master_uuid field was introduced in mysql-5.6.0.                            
  When replicating from a pre-5.6.0 master to a 5.7 slave,                        
  the slave will ask for the master_uuid field which the                          
  master doesnt have. Prior to fix, the P_S code made an                          
  assuption that if there is a master there will be a                             
  master_uuid for the master which is not true in the case                        
  of say 5.5->5.7 replication. Since the P_S code                                 
  blindly copied MASTER_UUID bytes from mi object,                                
  we ended up with a garbage value in the select                                  
  SOURCE_UUID field from P_S.replication_connect_status 
  
  Fix:
  - Added a check for empty master_uuid. In this case
    we print an empty value.
------------------------------------------------------------
revno: 7858 [merge]
committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
branch nick: Bug18542111_mysql-trunk
timestamp: Sat 2014-04-12 16:10:47 +0530
message:
  Merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.11
    committer: Sujatha Sivakumar <sujatha.sivakumar@oracle.com>
    branch nick: Bug18542111_mysql-5.6
    timestamp: Sat 2014-04-12 15:45:55 +0530
    message:
      Bug#18542111:ADD A TEST CASE TO TEST THE BINLOG TRANSACTION 
      CACHE SIZE TO 32768
------------------------------------------------------------
revno: 7857
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-11 21:41:38 +0300
message:
  Make a test work with --innodb-undo-tablespaces.
------------------------------------------------------------
revno: 7856
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-04-11 13:38:40 -0500
message:
  Bug #18558232 - SERVER CRASH IN FIL_INC_PENDING_OPS AT FIL0FIL.CC:1983
------------------------------------------------------------
revno: 7855
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18486249decimal
timestamp: Fri 2014-04-11 17:33:56 +0200
message:
  Bug#18497473 GET RID OF FIX_BUFFER_POINTER, COPY DECIMAL OBJECTS PROPERLY
  
  my_decimal::fix_buffer_pointer has been an abundant source of bugs.
  Remove it, and copy objects of type my_decimal properly.
------------------------------------------------------------
revno: 7854
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-11 17:06:09 +0300
message:
  Make the test work with --innodb-undo-tablespaces.
------------------------------------------------------------
revno: 7853
committer: Roy Lyseng <roy.lyseng@oracle.com>
branch nick: mysql-trunk-sb3
timestamp: Fri 2014-04-11 12:03:20 +0200
message:
  Bug#18064775 Assertion failed: fields.elements == values.elements
  Bug#11745889 Traditional: INSERT accepts invalid date from default value
  
  A recent bugfix (16820562) gave column count zero a special meaning,
  but failed to give an error for the case of zero column names but more
  than zero column values.
  
  The fix is to check explicitly for this case. The interface for
  check_insert_fields() has been extended with a bool value_count_known,
  to make it easy to remove it when the "known" property is no longer
  needed (all values are actually "known", but the code path is clumsy).
  
  The bugfix also re-introduced bug 11745889, which was fixed by bug
  16078943, alas it did not contain a test for this particular case.
  
  The fix for this bug is to treat INSERT statements with no column names
  and no column values specified as if a partial (but empty) column list
  has been specified. This way, the write_set for the table is not
  populated and the columns will be picked for validation of default
  values.
------------------------------------------------------------
revno: 7852
committer: Krunal Bauskar krunal.bauskar@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-04-11 15:08:50 +0530
message:
  - Bug #18511398: TRUNCATE TABLE IS WRITING PAGE-LEVEL REDO LOG RECORDS FOR THE
    "DROP TABLE" PART
    
    Truncate is done by dropping existing indexes and then re-creating the new
    ones. If tablespace is user-tablespace (innodb_file_per_table=1) then we can
    skip redo logging given that if server crashes then we don't apply redo logs
    instead we fix the truncate using ddl log. On successful completion
    log-checkpoint is done and so that, any crash post this point will not scan
    into previous redo-logs.
  
    Approved by: Sunny (rb#5145)
------------------------------------------------------------
revno: 7851 [merge]
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-11 11:11:25 +0300
message:
  merge
    ------------------------------------------------------------
    revno: 3902.324.10 [merge]
    committer: Georgi Kodinov <georgi.kodinov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-04-11 11:06:53 +0300
    message:
      merge
        ------------------------------------------------------------
        revno: 2875.592.16
        committer: Georgi Kodinov <georgi.kodinov@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-04-11 10:42:30 +0300
        message:
          Addendum #1 to the fix for bug #18359924
          
          Removed unused variable.
          Fixed long (>80 lines)
------------------------------------------------------------
revno: 7850 [merge]
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: B18359924-trunk
timestamp: Thu 2014-04-10 17:43:57 +0300
message:
  auto-merge
    ------------------------------------------------------------
    revno: 3902.324.9 [merge]
    committer: Georgi Kodinov <georgi.kodinov@oracle.com>
    branch nick: B18359924-5.6
    timestamp: Thu 2014-04-10 17:42:45 +0300
    message:
      auto-merge
        ------------------------------------------------------------
        revno: 2875.592.15
        committer: Georgi Kodinov <georgi.kodinov@oracle.com>
        branch nick: B18359924-5.5
        timestamp: Thu 2014-04-10 13:18:32 +0300
        message:
          Bug #18359924: INNODB AND MYISAM CORRUPTION ON PREFIX INDEXES
          
          The problem was in the validation of the input data for blob types.
          When assigned binary data, the character blob types were only checking if 
          the length of these data is a multiple of the minimum char length for the 
          destination charset. 
          And since e.g. UTF-8's minimum character length is 1 (becuase it's 
          variable length) even byte sequences that are invalid utf-8 strings (e.g. 
          wrong leading byte etc) were copied verbatim into utf-8 columns when
          coming from binary strings or fields.
          Storing invalid data into string columns was having all kinds of ill effects 
          on code that assumed that the encoding data are valid to begin with.
          
          Fixed by additionally checking the incoming binary string for validity when 
          assigning it to a non-binary string column.
          Made sure the conversions to charsets with no known "invalid" ranges 
          are not covered by the extra check.
          Removed trailing spaces.
          
          Test case added.
------------------------------------------------------------
revno: 7849
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-10 13:42:02 +0300
message:
  Work around a warning message on PB2 Valgrind runs.
------------------------------------------------------------
revno: 7848 [merge]
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-10 11:47:17 +0300
message:
  Null merge mysql-5.6 -> mysql-trunk
  (the changes are already present in mysql-trunk)
    ------------------------------------------------------------
    revno: 3902.324.8
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-09 20:27:43 +0300
    message:
      Non-functional change: rename dict_table_stats_latch_key to
      dict_table_stats_key, so that they have the same name in mysql-5.6 and
      mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.7
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-09 20:27:09 +0300
    message:
      Backport the following changeset from mysql-trunk to mysql-5.6:
      
        ** revision-id: vasil.dimov@oracle.com-20140401162003-zfhuxtu710846jd9
        ** committer: Vasil Dimov <vasil.dimov@oracle.com>
        ** branch nick: mysql-trunk
        ** timestamp: Tue 2014-04-01 19:20:03 +0300
        ** message:
        **   Fix Bug#71708 70768 fix perf regression: high rate of RW lock creation
        **   and destruction
        **
        **   Lazily create dict_table_t::stats_latch the first time it is used.
        **   It may not be used at all in the lifetime of some dict_table_t objects.
        **
        **   Approved by:	Bin (rb:4739)
    ------------------------------------------------------------
    revno: 3902.324.6
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-09 20:26:41 +0300
    message:
      Backport the following changeset from mysql-trunk to mysql-5.6:
      
        ** revision-id: vasil.dimov@oracle.com-20140401123646-qvr25kcca2bntx9q
        ** committer: Vasil Dimov <vasil.dimov@oracle.com>
        ** branch nick: mysql-trunk
        ** timestamp: Tue 2014-04-01 15:36:46 +0300
        ** message:
        **   Non-functional change: use InterlockedCompareExchange() instead of
        **   win_cmp_and_xchg_dword() for the macro os_compare_and_swap_uint32()
        **   on Windows. win_cmp_and_xchg_dword() only calls InterlockedCompareExchange()
        **   so it can be skipped altogether.
        **
        **   The macro os_compare_and_swap_uint32() is not used anywhere in the code.
------------------------------------------------------------
revno: 7847 [merge]
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: main-tr
timestamp: Thu 2014-04-10 10:30:17 +0200
message:
  Null upmerge
    ------------------------------------------------------------
    revno: 3902.324.5
    committer: Bjorn Munch <bjorn.munch@oracle.com>
    branch nick: main-56
    timestamp: Thu 2014-04-10 10:28:42 +0200
    message:
      Increase version number, this will not be 5.6.18
------------------------------------------------------------
revno: 7846
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-uintkorr
timestamp: Thu 2014-04-10 07:58:14 +0200
message:
  BUG#18368345 - ASSERT FIL_PAGE_GET_TYPE(PAGE) == 17855 AND OTHERS
  Fix broken build on 32bit platforms:
  error: format '%lu' expects argument of type 'long unsigned int', but argument 3 has type 'os_offset_t {aka long long unsigned int}' [-Werror=format=]
------------------------------------------------------------
revno: 7845 [merge]
committer: Arun Kuruvila <arun.kuruvila@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-10 11:15:49 +0530
message:
  Null merge from mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.324.4 [merge]
    committer: Arun Kuruvila <arun.kuruvila@oracle.com>
    branch nick: mysql-5.6
    timestamp: Thu 2014-04-10 11:12:28 +0530
    message:
      Null merge from mysql-5.5 to mysql-5.6.
        ------------------------------------------------------------
        revno: 2875.592.14
        committer: Arun Kuruvila <arun.kuruvila@oracle.com>
        branch nick: mysql-5.5
        timestamp: Thu 2014-04-10 11:10:31 +0530
        message:
          Description: When we execute a correlated subquery on an
          archive table which is using an auto increment column, the
          server hangs. In order to recover the mysqld process, it
          has to be terminated abnormally using SIGKILL. The problem
          is observed in mysql-5.5.
          Bug #18065452 "PREPARING" STATE HOGS CPU WITH ARCHIVE
                         + SUBQUERY
          
          Analysis: This happens because the server is trapped inside
          an infinite loop in the function,
          "subselect_indexsubquery_engine::exec()". This function
          resolves the correlated suquery by doing an index lookup
          for the appropriate engine. In  case of archive engine,
          after reaching the end of records, "table->status" is not
          set to STATUS_NOT_FOUND. As a result the loop is not 
          terminated.
          
          Fix: The "table->status" is set to STATUS_NOT_FOUND when
          the end of records is reached.
------------------------------------------------------------
revno: 7844
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-04-10 12:14:31 +0800
message:
  BUG 18522549 - HAVE_IB_LINUX_FUTEX TEST BROKEN IN DEBUG BUILDS,
  FUTEXES DISABLED
  
  Add #include <unistd.h> in innodb.cmake, so that the
  'Performing Test HAVE_IB_LINUX_FUTEX - Failed' warning disappears.
  
  Approved by Sunny over IM.
------------------------------------------------------------
revno: 7843
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-04-10 11:35:59 +0800
message:
  BUG 18368345 - ASSERT FIL_PAGE_GET_TYPE(PAGE) == 17855 AND OTHERS,
  TABLE CORRUPTED
  
  Table would get corrupted and mysqld eventually asserts
  fil_page_get_type(page) == 17855 through various code paths.
  
  There are mainly 3 issues here:
  1) We could only expect posix_fallocate returns 0 when success,
     but not -1 for all failures.
  2) Also, we used to pass the allocation end offset instead of
     allocation length to posix_fallocate, which was wrong.
  3) We should fill the newly allocated pages with zeros.
  
  In the future we could make it an optional parameter or make it a
  build conditional.
  
  Approved by Jimmy, rb#4978.
------------------------------------------------------------
revno: 7842
committer: Tarique Saleem <tarique.saleem@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-09 19:44:15 +0530
message:
  disabled i_main.partition_upgrade_big
------------------------------------------------------------
revno: 7841
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-invector
timestamp: Wed 2014-04-09 12:43:01 +0200
message:
  Bug#18503621 USE STANDARD LIBRARY FOR SORTING AND MATCHING IN SUBCLASSES OF IN_VECTOR
  
  As demonstrated by the patch for Bug#18486249
  in_vector and its derived classes can benefit from some code modernization.
  
  standard library functions are
   - type safe
   - easier to read/maintain
   - faster
------------------------------------------------------------
revno: 7840
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-18446462
timestamp: Wed 2014-04-09 10:05:06 +0200
message:
  WL#6635 - stabilize tests
------------------------------------------------------------
revno: 7839
committer: Bharathy Satish <bharathy.x.satish@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-09 11:40:26 +0530
message:
  Bug #18329348 RESETCONNECTION DOESN'T CLEAR TIMESTAMP
  Bug #18329560 RESETCONNECTION DOESN'T CLEAR RAND SEED
  Bug #18328396 RESETCONNECTION DOESN'T CLEAR WARNINGS
  Bug #18329452 RESETCONNECTION DOESN'T CLEAR PROFILING
  
  
  Problem: resetconnection does not reset some of the system variables.
           Most of the system variables are declared in THD::variables 
           struct. The variables which are declared outside this struct 
           are not handled in resetconnection ex: last_insert_id, insert_id,
           timestamp, rand_seed1, rand_seed2 etc.
  
  Fix: Fix is to reset these session variables to its default values in
       COM_RESET_CONNECTION. Even though @@profiling is reset, the profiling 
       information is not cleaned up before, with this fix this information 
       will be cleaned up.
------------------------------------------------------------
revno: 7838
committer: Sunanda Menon<sunanda.menon@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-09 06:56:59 +0200
message:
  Temporary changes to allow skip-random-passwords on Solaris
   - PKGVER is broken on weekly-trunk
   - mysql script is used for Solaris PKG format
------------------------------------------------------------
revno: 7837 [merge]
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-08 19:10:04 +0300
message:
  Null merge mysql-5.6 -> mysql-trunk
  (the changes are already present in mysql-trunk)
    ------------------------------------------------------------
    revno: 3902.324.3
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-08 19:08:35 +0300
    message:
      Backport the following changeset from mysql-trunk to mysql-5.6:
      
        ** revision-id: vasil.dimov@oracle.com-20140403163612-rjymqwuzkh6vs6dj
        ** committer: Vasil Dimov <vasil.dimov@oracle.com>
        ** branch nick: mysql-trunk
        ** timestamp: Thu 2014-04-03 19:36:12 +0300
        ** message:
        **   Followup to vasil.dimov@oracle.com-20140403070651-w1nefsafrqeid6ct:
        **   Increase the margin of allowed deviance for n_rows - with 4k page size
        **   it could sometimes be 745 (27% away from the actual value 1024).
    ------------------------------------------------------------
    revno: 3902.324.2
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-08 19:07:27 +0300
    message:
      Backport the following changeset from mysql-trunk to mysql-5.6:
      
        ** revision-id: vasil.dimov@oracle.com-20140403070651-w1nefsafrqeid6ct
        ** committer: Vasil Dimov <vasil.dimov@oracle.com>
        ** branch nick: mysql-trunk
        ** timestamp: Thu 2014-04-03 10:06:51 +0300
        ** message:
        **   Fix Bug#18384390 WRONG STATISTICS WITH BIG ROW LENGTH AND PERSISTENT STATS
        **
        **   Estimate the number of external pages when scanning any leaf page and subtract
        **   that estimate from index->stat_n_leaf_pages when calculating
        **   index->stat_n_diff_key_vals[].
        **
        **   Approved by:	Kevin, Satya (rb:4956)
    ------------------------------------------------------------
    revno: 3902.324.1
    committer: Vasil Dimov <vasil.dimov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-08 16:00:23 +0300
    message:
      Backport the following changeset from mysql-trunk to mysql-5.6:
      
        ** revision-id: vasil.dimov@oracle.com-20140403070224-eu2mw56ut6ydp354
        ** committer: Vasil Dimov <vasil.dimov@oracle.com>
        ** branch nick: mysql-trunk
        ** timestamp: Thu 2014-04-03 10:02:24 +0300
        ** message:
        **   Non-functional change:
        **
        **   Delay the calculation of each dict_index_t::stat_n_diff_key_vals[n_prefix]
        **   until after data is gathered for all n-column prefixes. This is a noop (non
        **   functional change) - still the same numbers will be used to calculate
        **   dict_index_t::stat_n_diff_key_vals[], but it will help fix
        **   Bug#18384390 WRONG STATISTICS WITH BIG ROW LENGTH AND PERSISTENT STATS
        **
        **   The problem in that bug is that index->stat_n_leaf_pages is bloated by
        **   the number of externally stored pages, which should not really be counted
        **   in the formula that derives dict_index_t::stat_n_diff_key_vals[].
        **
        **   To fix this we need to estimate the number of external pages in the
        **   index and subtract it from dict_index_t::stat_n_leaf_pages in that formula.
        **
        **   In a subsequent change we will estimate the number of external pages while
        **   sampling each page for all possible n-column prefixes and then this cumulative
        **   result will be used then calculating each member of
        **   dict_index_t::stat_n_diff_key_vals[].
        **
        **   The code before this change:
        **
        **   dict_stats_analyze_index()
        **     for each n prefix
        **       dict_stats_analyze_index_for_n_prefix()
        **         sample some pages and save the n_diff results in
        **         index->stat_n_diff_key_vals[] using index->stat_n_leaf_pages in the
        **         formula
        **
        **   The code after this change (equivalent):
        **
        **   dict_stats_analyze_index()
        **     for each n prefix
        **       dict_stats_analyze_index_for_n_prefix()
        **         sample some pages and save the n_diff results in a temporary place
        **     for each n prefix
        **       // new function, code moved from dict_stats_analyze_index_for_n_prefix()
        **       dict_stats_index_set_n_diff()
        **         set index->stat_n_diff_key_vals[] using
        **         index->stat_n_leaf_pages in the formula
        **
        **   Further planned change that will actually fix the bug:
        **
        **   dict_stats_analyze_index()
        **     for each n prefix
        **       dict_stats_analyze_index_for_n_prefix()
        **         sample some pages and save the n_diff results in a temporary place
        **         and also accumulate an estimate about the number of external pages
        **         when scanning each leaf page
        **     for each n prefix
        **       dict_stats_index_set_n_diff()
        **         set index->stat_n_diff_key_vals[] using
        **         "index->stat_n_leaf_pages - number_of_external_pages" in the formula
        **
        **   Approved by:	Satya (rb:4955)
------------------------------------------------------------
revno: 7836
committer: Annamalai Gurusami <annamalai.gurusami@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-08 13:37:59 +0530
message:
  Problem:
            
  In the clustered index, when an update operation is done the overall
  scenario is as follows:
            
  1.  Delete mark the old record that is to be updated.
  2.  The old record disowns the blobs. As part of this step, we assert that
      the blobs are owned by the old record.
  3.  Checks the foreign key constraints.
  4.  Insert the new record into clustered index.
            
  Scenario involving DB_LOCK_WAIT:
            
  If step 3 times out, then we will repeat steps 1 and 2.  Since the steps 1
  and 2 are already done, the assert that the blobs must be owned in step 2
  will fail.
  
  Solution:
            
  If step 3 times out, then skip steps 1 and 2 and proceed with 3.  This is
  achieved by checking if the record is already delete marked.
  
  rb#5081 approved by Marko.
------------------------------------------------------------
revno: 7835
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-uintkorr
timestamp: Mon 2014-04-07 16:53:55 +0200
message:
  Bug#18526237 BYTE_ORDER_GENERIC_X86[_64].H ARE ALMOST IDENTICAL
  
  After recent code cleanups, the two files
    include/byte_order_generic_x86.h
    include/byte_order_generic_x86_64.h
  are almost identical.
  Remove one of them.
------------------------------------------------------------
revno: 7834
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Mon 2014-04-07 10:02:12 +0200
message:
  Bug#18526007: STRINGS_SKIP_TRAILING UNIT TEST FAILS WITH LIBC++
  
  The unit test failed because it assumed that the first byte of
  a std::string would be aligned. This is not neccessarily true for 
  the libc++ implementation of std::string 
  
  This patch fixes the problem by removing the assert.
  
  Test change only, no changes to server code.
------------------------------------------------------------
revno: 7833
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-review
timestamp: Mon 2014-04-07 08:07:52 +0200
message:
  BUG#18446462
  
  Explain JSON tests too unstable on 32 bit platform (minor diffs
  in rows estimates etc). Only run on 64 bit platforms.
------------------------------------------------------------
revno: 7832
committer: Olav Sandstaa <olav.sandstaa@oracle.com>
branch nick: mrr-test
timestamp: Fri 2014-04-04 14:00:52 +0200
message:
  Follow-up patch for Bug#13106350: MRR initialization on a derived 
                                    table caused crash
  
  Updated test case by removing out-commented setting of
  optimizer_switch materialization to off.
  This was commented out due to at the time this test was
  back-ported to mysql-trunk, the materalization switch was
  not part of optimizer_switch on mysql-trunk.
------------------------------------------------------------
revno: 7831
committer: Tatjana Azundris Nuernberg <tatjana.nuernberg@oracle.com>
branch nick: 57-18127243
timestamp: Fri 2014-04-04 10:56:26 +0100
message:
  Bug#18127243: MYSQLD --GENERAL-LOG HAS INCORRECT HELP TEXT
  
  Add missing whitespace to and clarify --general-log help text.
------------------------------------------------------------
revno: 7830 [merge]
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-04 12:23:20 +0300
message:
  Merge mysql-5.6 to mysql-trunk.
    ------------------------------------------------------------
    revno: 3902.291.328
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-04-04 12:17:07 +0300
    message:
      Remove some Valgrind check requests that were broken by the following:
      
      Bug#16249481 INNODB DOES NOT SCALE WELL ON 12 CORE SYSTEM FOR
      SPECIFIC ALL IN MEMORY SELECT
      
           This fix increased sizeof(buf_page_t) from 64 bytes even on
           32-bit systems.
      
      WL#5522 Fix the IMPORT tablespace performance.
      
           This fix made sizeof(page_zip_des_t) something else than a power
           of two.
      
      The removed UNIV_MEM_ASSERT_RW() checks were only enabled on 32-bit
      systems. Since the above changes, they would emit bogus warnings about
      unused "pad" bytes being uninitialized.
------------------------------------------------------------
revno: 7829
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-04 12:17:58 +0300
message:
  WL#7142: Correct a type mismatch.
------------------------------------------------------------
revno: 7828 [merge]
committer: Serge Kozlov <serge.kozlov@oracle.com>
branch nick: mysql-trunk
timestamp: Fri 2014-04-04 12:11:45 +0400
message:
  Bug#18506556 Merge 5.6->trunk
    ------------------------------------------------------------
    revno: 3902.291.327 [merge]
    committer: Serge Kozlov <serge.kozlov@oracle.com>
    branch nick: mysql-5.6
    timestamp: Fri 2014-04-04 12:06:10 +0400
    message:
      Bug#18506556 Merge 5.5->5.6
        ------------------------------------------------------------
        revno: 2875.592.13
        committer: Serge Kozlov <serge.kozlov@oracle.com>
        branch nick: mysql-5.5
        timestamp: Fri 2014-04-04 10:42:25 +0400
        message:
          BUG#18506556. Added sync slave with master for clean-up
------------------------------------------------------------
revno: 7827 [merge]
committer: bin.x.su@oracle.com
branch nick: mysql-trunk
timestamp: Fri 2014-04-04 11:43:03 +0800
message:
  BUG 18277082 - FTS: STACK BUFFER OVERFLOW IN INNOBASE_STRNXFRM AND TIS620
  
  When we run windows debug build or use a linux Address Sanitizer build,
  there would be a failure or stack-buffer-overflow in innobase_strnxfrm.
  
  The root cause is that my_strnxfrm_tis620 would access one more byte of
  the dst parameter and set it to '\0', which is unnecessary in this case.
  So we just do not set this terminating '\0'.
  
  Approved by Tor, rb#5051.
    ------------------------------------------------------------
    revno: 3902.291.326
    committer: bin.x.su@oracle.com
    branch nick: mysql-5.6
    timestamp: Fri 2014-04-04 11:35:27 +0800
    message:
      BUG 18277082 - FTS: STACK BUFFER OVERFLOW IN INNOBASE_STRNXFRM AND TIS620
      
      When we run windows debug build or use a linux Address Sanitizer build,
      there would be a failure or stack-buffer-overflow in innobase_strnxfrm.
      
      The root cause is that my_strnxfrm_tis620 would access one more byte of
      the dst parameter and set it to '\0', which is unnecessary in this case.
      So we just do not set this terminating '\0'.
      
      Approved by Tor, rb#5051.
------------------------------------------------------------
revno: 7826
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-03 19:36:12 +0300
message:
  Followup to vasil.dimov@oracle.com-20140403070651-w1nefsafrqeid6ct:
  Increase the margin of allowed deviance for n_rows - with 4k page size
  it could sometimes be 745 (27% away from the actual value 1024).
------------------------------------------------------------
revno: 7825 [merge]
committer: Marko M?kel? <marko.makela@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-03 17:11:16 +0300
message:
  Merge mysql-trunk-wl7142 to mysql-trunk.
  
  NOTE: This will change the InnoDB redo log format.
  Crash recovery will fail if you upgrade or downgrade
  after a non-clean shutdown.
  
  rb#4700 approved by Kevin Lewis and Bin Su
    ------------------------------------------------------------
    revno: 7721.1.54
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-04-03 13:46:38 +0300
    message:
      WL#7142: Prevent a potential race condition between fil_space_free_low()
      and fil_names_dirty().
      
      fil_names_write(): If the tablespace is being dropped, trigger a
      debug assertion failure and return NULL, so that fil_names_dirty()
      cannot be dereferencing a freed object nor enter a race with
      fil_space_free_low().
      
      This addresses comments by Vasil Dimov, Bin Su and Michael Izioumtchenko.
    ------------------------------------------------------------
    revno: 7721.1.53
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-04-03 10:54:58 +0300
    message:
      WL#7142: Disable a failing test for which a work-in-progress fix exists.
      
      The fix for Bug#16963396 had to be reverted due to other test failures.
      The slightly added redo log volume due to MLOG_FILE_NAME from WL#7142
      is tipping the scales to make this BLOB redo logging test fail.
    ------------------------------------------------------------
    revno: 7721.1.52
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-04-03 10:43:08 +0300
    message:
      WL#7142: Address review comments from Bin Su.
      
      fil_mtr_rename_log(): Fix a memory leak.
    ------------------------------------------------------------
    revno: 7721.1.51
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 23:21:03 +0300
    message:
      WL#7142: Clean up a new method.
      
      Datafile::init(name, filepath, size, order): Assume filepath to be freeable
      at shutdown(), or to be NULL. Remove conditions from the function body.
      
      This addresses a review comment by Kevin Lewis.
    ------------------------------------------------------------
    revno: 7721.1.50
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 23:05:31 +0300
    message:
      WL#7142: Correct a comment.
      
      Requested by Kevin Lewis.
    ------------------------------------------------------------
    revno: 7721.1.49
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 23:04:17 +0300
    message:
      WL#7142: Do not replace ut_free() with free().
      
      While ut_malloc() and ut_realloc() currently are non-trivial wrappers for
      malloc() and free(), and ut_free() is a null wrapper for free(),
      we may want to change this later.
      
      Requested by Kevin Lewis on IM.
    ------------------------------------------------------------
    revno: 7721.1.48
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 22:19:37 +0300
    message:
      WL#7142: Clarify some comments.
      
      These changes were requested by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.47
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 22:09:18 +0300
    message:
      WL#7142: Address some review comments.
      
      Write x == 0 instead of 0 == x in comparisons.
      Add some const qualifiers.
      Reduce the scope of some variable declarations.
      
      These changes were requested by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.46
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 21:55:09 +0300
    message:
      WL#7142: Address a review comment.
      
      Explicitly compare the result of memcmp(), memchr(), strchr(), strstr()
      against 0 or NULL, instead of relying on the implicit
      integer-to-Boolean or pointer-to-Boolean conversion.
      
      This change was requested by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.45
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 21:48:09 +0300
    message:
      WL#7142: Address a review comment.
      
      fil_space_free_low(): Remove a type cast by replacing "%lu" with ULINTPF.
      
      This change was requested by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.44
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 16:37:22 +0300
    message:
      WL#7142: Test --innodb-force-recovery.
      
      Add a few test "stubs" to be fixed in
      Bug#18131883 IMPROVE INNODB ERROR MESSAGES REGARDING FILES
      and inject a few faults to ensure that --innodb-force-recovery
      will do the right thing.
      
      Datafile::validate_first_page(): Do not skip any validation even in
      innodb-force-recovery mode. Skipping validation of the header page
      would lead to all kinds of assertion failures in other code.
      Other code needs valid tablespace flags, for instance. To force
      loading a badly corrupted file after skipping the redo log for it,
      the DBA could always tweak the file manually.
      
      fil_op_log_parse_or_replay(): If the target file exists, abort
      the redo log replay so that the DBA can decide which file to keep.
      
      fil_load_single_table_tablespace(): Move the messages to
      the only caller fil_name_parse(), and correct the wording
      regarding the handling of missing tablespace files.
      
      This addresses the faults found during code coverage analysis.
    ------------------------------------------------------------
    revno: 7721.1.43
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 10:52:39 +0300
    message:
      WL#7142: Correct a comment.
    ------------------------------------------------------------
    revno: 7721.1.42
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 10:48:02 +0300
    message:
      WL#7142: Clarify comments regarding space->max_lsn.
    ------------------------------------------------------------
    revno: 7721.1.41
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 10:22:11 +0300
    message:
      WL#7142: Remove a goto from out-of-memory error handling.
      
      This addresses review comments by Kevin Lewis and Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.40
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 10:00:28 +0300
    message:
      WL#7142: Skip crash testing on embedded server.
    ------------------------------------------------------------
    revno: 7721.1.39
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 09:49:02 +0300
    message:
      WL#7142: Clarify some code with comments and assertions.
      
      mtr_t::Command::prepare_write(): Add comments and debug assertions to
      explain how we are manipulating the tail of the local mini-transaction log
      before it will be copied to the global redo log buffer.
      
      This addresses a review comment by Kevin Lewis.
    ------------------------------------------------------------
    revno: 7721.1.38
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 08:59:52 +0300
    message:
      WL#7142: Reduce some dependency between header files,
      to be able to use TRX_SYS_SPACE in an assertion in mtr0mtr.h.
      
      TRX_SYS_SPACE,TRX_SYS_PAGE_NO: Move the declaration
      from trx0sys.h to trx0types.h.
      
      This addresses a review comment by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.37
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-04-02 08:20:31 +0300
    message:
      WL#7142 cleanup: Remove redundant :: qualifiers for the global scope.
      
      Also, replace ::ut_free() with free().
      
      This addresses a review comment by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.36
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Tue 2014-04-01 08:45:04 +0300
    message:
      WL#7142: Simplify mtr_t::set_named_space().
      
      mtr_t::set_named_space(): Unconditionally set m_named_space.
      Assert that it was not set previously.
      
      ibuf_bitmap_page_set_bits(): Replace the mtr_t::set_named_space() call
      with mtr_t::is_named_space() assertions.
      
      ibuf_set_free_bits_func(), ibuf_insert_low(): Call mtr_t::set_named_space().
      
      There are a few other callers of ibuf_bitmap_page_set_bits() in
      ibuf0ibuf.cc, but these are accessing a predefined tablespace
      (tablespace 0, IBUF_SPACE_ID).
    ------------------------------------------------------------
    revno: 7721.1.35
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Tue 2014-04-01 07:42:44 +0300
    message:
      WL#7142: Improve the wording of an error message.
    ------------------------------------------------------------
    revno: 7721.1.34
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 23:07:15 +0300
    message:
      WL#7142: Fix Windows failures of innodb.innodb-multiple-tablespaces.
    ------------------------------------------------------------
    revno: 7721.1.33
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 23:02:09 +0300
    message:
      WL#7142: Fix a typo in a comment.
    ------------------------------------------------------------
    revno: 7721.1.32
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 23:01:02 +0300
    message:
      WL#7142: Clarify some comments.
    ------------------------------------------------------------
    revno: 7721.1.31
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 22:56:41 +0300
    message:
      WL#7142: Fix rb#4700 Problem 9.
    ------------------------------------------------------------
    revno: 7721.1.30
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 13:08:38 +0300
    message:
      WL#7142: Clean up innodb.innodb-multiple-tablespaces test.
      
      Use shutdown_server 0 for killing the server.
      Allow the test to run with non-debug servers, and with Valgrind.
      Clean up message suppressions. Remove some sleeps.
      
      Do not restart with force-recovery, but instead do a test restart
      that is expected to fail to start up. Then, clean up the injected
      error and restart normally.
      
      This test started failing after the change
      marko.makela@oracle.com-20140328123744-ewk9sgbtwxfmurio
      started to completely ignore *.isl files during redo log processing.
      
      The problem was that the redo log was being ignored for the affected
      tablespaces due to --innodb-force-recovery=1, and the tablespaces
      were corrupted. Enabling innodb_force_recovery should be a last resort,
      it can corrupt data also in other cases.
    ------------------------------------------------------------
    revno: 7721.1.29
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 11:55:59 +0300
    message:
      WL#7142: Properly adjust tablespace name in MLOG_FILE_RENAME2 apply.
      
      fil_op_log_parse_or_replay(): On Windows, replace OS_PATH_SEPARATOR with '/'
      in the new_table name.
      
      fil_rename_tablespace_in_mem(): Remove the buggy tweak.
      
      fil_rename_tablespace(): Assert that the names contain correct separators.
      
      This addresses a review comment from Kevin Lewis.
    ------------------------------------------------------------
    revno: 7721.1.28
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 10:43:16 +0300
    message:
      WL#7142: Use --console instead of --loose-console in tests.
    ------------------------------------------------------------
    revno: 7721.1.27
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-31 10:37:32 +0300
    message:
      WL#7142: Replace fsp_names_write().
      
      Call mtr_t::set_named_space() directly instead of invoking fsp_names_write().
      In most cases, we can call it immediately after mtr_t::start().
      When the tablespace is missing, we must avoid the call in order to
      avoid a debug assertion failure during mtr_commit().
      
      There were a few redundant calls in low-level code (btr0cur.cc).
      The high-level code (row0*.cc) is calling set_named_space() already.
      
      This addresses a review comment by Vasil Dimov.
    ------------------------------------------------------------
    revno: 7721.1.26
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 16:29:06 +0200
    message:
      WL#7142: Correct an error that was introduced in
      marko.makela@oracle.com-20140328100830-qyku3d65kn3thrln
      in the file name handling. This only affected debug builds and Windows.
    ------------------------------------------------------------
    revno: 7721.1.25
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 15:56:45 +0200
    message:
      WL#7142: Minor cleanup (non-functional change).
    ------------------------------------------------------------
    revno: 7721.1.24
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 14:37:44 +0200
    message:
      WL#7142: Completely ignore *.isl files during redo log apply.
      Instead, use the *.ibd file names from MLOG_FILE_NAME records.
      
      fil_load_single_table_tablespace(): Replace df_default and df_remote
      with Datafile file. Pass the ownership of the allocated name to
      Datafile, so that the ut_free() cleanup will be eliminated, done as
      part of Datafile::shutdown(), executed by the destructor. Likewise,
      remove calls to Datafile::close(), because that will be part of
      shutdown().
      
      Datafile::init(): Add a variant that takes ownership of the "name",
      and allows filepath to be initialized.
      
      Datafile::shutdown(): Remove a redundant check for m_name!=NULL.
      free(NULL) is documented as no-op in the C standard.
    ------------------------------------------------------------
    revno: 7721.1.23
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 12:08:30 +0200
    message:
      WL#7142: Add a helper function fil_path_to_space_name().
      
      HAVE_MEMRCHR: Feature macro for memrchr().
      
      fil_load_single_table_tablespace(): Make the filename const again.
      Make use of the new helper function fil_path_to_space_name().
    ------------------------------------------------------------
    revno: 7721.1.22
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 12:07:46 +0200
    message:
      WL#7142: Add quotes around file names in error messages.
    ------------------------------------------------------------
    revno: 7721.1.21
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 10:02:19 +0200
    message:
      WL#7142: Remove suppression expressions from a test result.
    ------------------------------------------------------------
    revno: 7721.1.20
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 09:57:58 +0200
    message:
      WL#7142: Introduce a local variable for a complex expression.
    ------------------------------------------------------------
    revno: 7721.1.19
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 09:53:23 +0200
    message:
      WL#7142: Introduce the constant SIZE_OF_MLOG_CHECKPOINT.
    ------------------------------------------------------------
    revno: 7721.1.18
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 09:47:06 +0200
    message:
      WL#7142: fil_op_write_log(): Rename a parameter.
    ------------------------------------------------------------
    revno: 7721.1.17
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-28 09:40:44 +0200
    message:
      WL#7142: Correct comments about fil_names_dirty() and fil_names_write().
    ------------------------------------------------------------
    revno: 7721.1.16
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-03-27 18:01:11 +0200
    message:
      WL#7142: Add a test case and adjust error handling.
      
      Datafile::validate_for_recovery(), Datafile::validate_first_page():
      Report a duplicate space_id by returning DB_TABLESPACE_EXISTS.
      
      fil_space_create(): Do not add the "dup" output parameter.
      
      fil_load_single_table_tablespace(): Detect duplicate space_id
      by checking for DB_TABLESPACE_EXISTS. Remove the dead code
      checking for dup!=NULL.
      
      The reason why dup!=NULL cannot ever hold is that
      Datafile::validate_first_page() invokes
      fil_space_read_name_and_filepath(), which invokes
      fil_space_get_by_id(). We could only get dup!=NULL when
      fil_space_get_by_id() in fil_space_create() already found the
      tablespace. Because we always invoke Datafile::validate_first_page()
      via Datafile::validate_for_recovery() before fil_space_create(),
      we will already have ruled out a duplicate space_id when invoking
      fil_space_create().
    ------------------------------------------------------------
    revno: 7721.1.15
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-03-27 17:22:20 +0200
    message:
      WL#7142: Fix some white space.
    ------------------------------------------------------------
    revno: 7721.1.14
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-03-26 13:33:46 +0200
    message:
      WL#7142: Check for space_id mismatch.
      
      fil_load_single_table_tablespace(): On df->space_id() mismatch, return
      FIL_LOAD_ID_CHANGED.
      
      This fixes random error messages from fil_space_free_low().
    ------------------------------------------------------------
    revno: 7721.1.13
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-03-26 11:58:40 +0200
    message:
      WL#7142: Fix some review comments.
      
      fil_space_free(): Move the function right after fil_space_free_low()
      that it is wrapping.
      
      fil_load_single_table_tablespace(): Correct an outdated comment.
    ------------------------------------------------------------
    revno: 7721.1.12
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-03-26 11:49:28 +0200
    message:
      WL#7142: Fix MLOG_FILE_RENAME2 handling on Windows.
      
      The space->name will use / even on Windows, while the file names will
      use OS_PATH_SEPARATOR.
      
      fil_op_log_parse_or_replay(): Look for OS_PATH_SEPARATOR instead of /.
      
      fil_rename_tablespace_in_mem(): Make space->name to use /
      instead of OS_PATH_SEPARATOR.
      
      This fix is originally from Mattias Jonsson.
    ------------------------------------------------------------
    revno: 7721.1.11
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-03-26 11:38:10 +0200
    message:
      WL#7142: Remove some redundant checks.
      
      fil_load_single_table_tablespace(): Remove the assertion about
      file name suffix. The only caller fil_name_parse() will already
      have checked the file name. Remove some dead code (impossible condition)
      and a check for an always valid condition.
    ------------------------------------------------------------
    revno: 7721.1.10
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-24 15:57:06 +0200
    message:
      WL#7142: Correct a function comment.
    ------------------------------------------------------------
    revno: 7721.1.9
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-24 15:41:35 +0200
    message:
      WL#7142: Try to fix a problem on Windows.
      
      fil_load_single_table_tablespace(): Use '/' instead of OS_PATH_SEPARATOR
      in the file name, for initializing space->name.
      
      Fix some tests to allow either form of OS_PATH_SEPARATOR.
    ------------------------------------------------------------
    revno: 7721.1.8
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-24 09:37:52 +0200
    message:
      WL#7142: Address rb#4700 Problem 6.
      
      Add fsp_names_write() calls and is_named_space() assertions to ensure
      that no tablespace is going to be changed without writing MLOG_FILE_NAME.
    ------------------------------------------------------------
    revno: 7721.1.7
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Mon 2014-03-24 09:34:54 +0200
    message:
      WL#7142: Fix rb#4700 Problem 4.
      
      fil_delete_tablespace(): Write the MLOG_FILE_DELETE record before deleting.
      Otherwise, crash recovery can be refused due to a missing file,
      in case we get killed after deleting but before writing the record.
      
      fil_load_single_table_tablespace(): Remove some type casts in error messages.
    ------------------------------------------------------------
    revno: 7721.1.6
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Fri 2014-03-21 10:14:44 +0200
    message:
      WL#7142: Do not write MLOG_LSN after MLOG_CHECKPOINT.
      
      This was reported as rb#4700 Problem 5, a warning about tablespace files
      being newer than the end of the redo log when the server was compiled
      with the extra debug option UNIV_LOG_LSN_DEBUG.
      
      This also caused a failure of innodb.innodb-log-file-size-1 due to
      the same warning.
    ------------------------------------------------------------
    revno: 7721.1.5
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-03-20 18:19:12 +0200
    message:
      WL#7142: Remove the function fil_space_get().
    ------------------------------------------------------------
    revno: 7721.1.4
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-03-20 18:12:43 +0200
    message:
      Revert an inadvertent change.
    ------------------------------------------------------------
    revno: 7721.1.3
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-03-20 17:56:22 +0200
    message:
      WL#7142 scalability improvement.
      
      During mtr_commit(), reduce the work done while holding log_sys->mutex.
      Also, do not acquire fil_system->mutex while holding log_sys->mutex.
    ------------------------------------------------------------
    revno: 7721.1.2
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Thu 2014-03-20 14:48:41 +0200
    message:
      WL#7142: Fix rb#4700 Problem 1 (failure with UNIV_LOG_LSN_DEBUG).
      
      recv_parse_log_rec(), recv_parse_log_recs(): Handle MLOG_LSN.
    ------------------------------------------------------------
    revno: 7721.1.1
    committer: Marko M?kel? <marko.makela@oracle.com>
    branch nick: mysql-trunk-wl7142
    timestamp: Wed 2014-03-19 15:07:55 +0200
    message:
      Test push of WL#7142 InnoDB: Simplify tablespace discovery during crash recovery
      
      When the setting innodb_file_per_table=ON was introduced in MySQL 4.1,
      InnoDB crash recovery was changed so that the directories will be
      searched for *.ibd files if any redo needs to be applied.
      
      The scanning and opening of all *.ibd files (including ones for which
      no redo log needs to be applied) can be very slow, especially on
      deployments that contain a large number of *.ibd files. Furthermore,
      if we allow a more liberal placement of tablespace files in the file
      system, we might have to extend the search to an even broader range of
      directories.
      
      This worklog eliminates the *.ibd file scan by guaranteeing the
      following:
      
      If there are redo log records for any non-predefined tablespace, there
      will also be an MLOG_FILE_NAME record.
      
      The InnoDB redo log format will be changed as follows:
      
      MLOG_FILE_NAME(space_id, filename): A new redo log record.
      Replaces MLOG_FILE_CREATE, MLOG_FILE_CREATE2.
      
      MLOG_FILE_RENAME2(space_id, old, new): The names will be file names
      (directory/databasename/tablename.ibd). Replaces MLOG_FILE_RENAME,
      which used table names (databasename/tablename).
      
      NOTE: We will write MLOG_FILE_NAME once since the latest redo log
      checkpoint. Immediately after a checkpoint, the log may contain some
      MLOG_FILE_NAME records that were "copied across the checkpoint" and a
      MLOG_CHECKPOINT marker to signal the end of a checkpoint.
      
      On redo log apply during crash recovery, we will scan the log up to
      three times:
      
      Recovery scan 1: Look for the first MLOG_FILE_CHECKPOINT marker since
      the latest checkpoint.
      
      If there is no MLOG_FILE_CHECKPOINT, we will skip the entire log. The
      data files will correspond to the system state as of the checkpoint.
      
      Recovery scan 2: Read the redo log since the latest checkpoint. Copy
      scanned records to recv_sys->addr_hash, and construct a map of
      recv_spaces, based on MLOG_FILE_NAME and MLOG_FILE_DELETE records.
      
      Before applying the records from recv_sys->addr_hash, we will check if
      any tablespace files are missing. If there are missing tablespaces, we
      will refuse to start up, so that the DBA can intervene, for example to
      manually rename files. This new safeguard of WL#7142 can be disabled
      by setting innodb_force_recovery.
      
      If not all redo log records in recv_sys->addr_hash, we will need a
      third log scan:
      
      Recovery scan 3: Read the redo log since the latest checkpoint. If
      recv_sys->addr_hash fills up, apply the batch of log records and read
      a new one.
      
      mlog_id_t: Remove MLOG_FILE_CREATE, MLOG_FILE_CREATE2, MLOG_FILE_RENAME.
      Add MLOG_FILE_NAME, MLOG_FILE_RENAME2, MLOG_CHECKPOINT.
      
      MLOG_FILE_FLAG_TEMP: Remove. This was a flag for MLOG_FILE_CREATE*.
      
      enum dict_check_t: Remove DICT_CHECK_ALL_LOADED. Crash recovery no
      longer loads all tablespaces.
      
      mtr_t::m_named_space: Associates a tablespace with a
      mini-transaction. A mini-transaction may be associated with up to one
      non-predefined tablespace. It may also modify predefined tablespaces
      for change buffering and undo logging.
      
      mtr_t::set_named_space(ulint space): Sets m_named_space.
      
      mtr_t::is_named_space(ulint space): Checks if the mini-transaction is
      associated with a given tablespace.
      
      mtr_t::Command::prepare_write(): Write an MLOG_FILE_NAME record if
      needed. This is executed as part of mtr_commit().
      
      mtr_t::commit_checkpoint(): A special method to emit redo log records
      to the redo log buffer when the caller already invoked
      log_mutex_enter(). This is only used by fil_names_clear().
      
      fil_space_t::max_lsn: LSN of the most recent fil_names_write() call,
      or 0 if the tablespace has not been dirtied since fil_names_clear().
      
      fil_space_t::named_spaces, fil_system_t::named_spaces: List of
      tablespaces for which MLOG_FILE_NAME has been written since the latest
      checkpoint.
      
      recv_sys_t: mlog_checkpoint_lsn: The LSN of the first scanned
      MLOG_CHECKPOINT record, or 0 if none was read yet.
      
      fil_space_get(): Look up a tablespace. This is invoked during
      mtr_t::Command::prepare_write() while not holding the log mutex, to
      prepare for a fil_names_write() call. The idea is to minimize the
      log_mutex hold time.
      
      fil_space_create(): Add an output parameter for returning a duplicate
      tablespace (same space_id).
      
      fil_space_free(): Make this an externally callable function, to free a
      tablespace from the cache when applying MLOG_FILE_DELETE.
      
      fil_space_free_low(): Renamed from fil_space_free(). The new wrapper
      fil_space_free() will acquire fil_system->mutex.
      
      fil_op_log_parse_or_replay(): Change the order of parameters. Remove
      log_flags, and rename parse_only to replay. We no longer attempt to
      replay log records of a multi-item mini-transaction, unless the
      MLOG_MULTI_REC_END was seen.
      
      fil_rename_tablespace(): Change the function signature. Take old_path,
      new_name, new_path_in. MLOG_FILE_RENAME2 is logging file names, not
      table names like MLOG_FILE_RENAME was. Also invoke fil_name_write().
      
      enum fil_load_status: Outcomes of fil_load_single_table_tablespace().
      
      fil_load_single_table_tablespace(): Do not exit on failure. Instead,
      return a status value to the caller.
      
      fil_load_single_table_tablespaces(): Remove. We no longer try to load
      all *.ibd files.
      
      fil_create_new_single_table_tablespace(): Do not write any
      MLOG_FILE_CREATE or MLOG_FILE_CREATE2. Instead, invoke
      fil_name_write() to write MLOG_FILE_NAME.
      
      fil_mtr_rename_log(): Change the signature. Take dict_table_t instead
      of names. Take a tmp_name.
      
      fil_names_write_low(): Write MLOG_FILE_NAME record(s) for a
      tablespace.
      
      fil_names_write(): Write MLOG_FILE_NAME record(s) for a tablespace if
      not already written since the latest checkpoint.
      
      fil_names_clear(): Write MLOG_FILE_NAME records and MLOG_CHECKPOINT on
      a log checkpoint or at system startup. If do_write=true, writes
      MLOG_CHECKPOINT even if no MLOG_FILE_NAME was written.
      Reset those fil_space_t::max_lsn for which fil_names_write() has not
      been invoked after the checkpoint LSN. Return true to the caller if
      any redo log was written.
      
      fil_op_write_log(): Replace log_flags with first_page_no, and replace
      table names with file paths. The parameter first_page_no is currently
      being passed as 0, because we do not have non-predefined multi-file
      tablespaces yet.
      
      fil_name_write(): Write an MLOG_FILE_NAME record for a file.
      
      Datafile::open_read_only(): Add the parameter bool strict.
      
      fsp_names_write(): Wrapper for mtr->set_named_space(). This must be
      called when a mini-transaction is going to modify a non-predefined
      tablespace.
      
      is_predefined_tablespace(): Check if a tablespace is a predefined one
      (system tablespace, undo tablespace or shared temporary tablespace).
      
      enum recv_addr_state: Add RECV_DISCARDED, so that buffered redo log
      records can be retroactively deleted if an MLOG_FILE_DELETE was
      later recovered for a tablespace.
      
      btr_free_but_not_root(), btr_free_root(): Call fsp_names_write().
      
      btr_cur_ins_lock_and_undo(), btr_cur_optimistic_insert(),
      btr_cur_pessimistic_insert(), btr_cur_update_in_place(),
      btr_cur_optimistic_update(), btr_cur_pessimistic_update(),
      btr_cur_del_mark_set_clust_rec_log(),
      btr_cur_del_mark_set_clust_rec(), btr_cur_optimistic_delete_func(),
      btr_cur_pessimistic_delete(): Call fsp_names_write() after successful
      locking and undo logging.
      
      btr_store_big_rec_extern_fields(), btr_free_externally_stored_field(),
      row_ins_index_entry_big_rec_func(): Call fsp_names_write().
      
      dict_build_tablespace(), dict_create_index_tree_step(),
      dict_recreate_index_tree(), fil_reinit_space_header(): Call
      fsp_names_write().
      
      page_cur_insert_rec_write_log(),
      page_copy_rec_list_to_created_page_write(),
      page_cur_delete_rec_write_log(), page_cur_delete_rec(), page_create():
      Assert that fsp_names_write() has been called.
      
      dict_table_rename_in_cache(): Pass old_path to
      fil_rename_tablespace().
      
      dict_check_tablespaces_and_store_max_id(): Remove the logic for
      DICT_CHECK_ALL_LOADED. We could probably remove this entire function,
      given that the maximum is also stored in the DICT_HDR page.
      
      mlog_write_initial_log_record_for_file_op(): Replaced by
      mlog_write_initial_log_record_low().
      
      log_checkpoint(): Before invoking log_write_up_to(), invoke
      fil_names_clear() to copy any MLOG_FILE_NAME records across the
      checkpoint. Flush the log up to the MLOG_CHECKPOINT marker, instead of
      only up to the checkpoint LSN. Without this step, the log between
      oldest_lsn and log_sys->lsn would be essentially corrupted (missing
      MLOG_FILE_NAME records on redo log apply). When the redo log scanner
      sees the first MLOG_CHECKPOINT since the latest checkpoint, it knows
      that there must be no missing MLOG_FILE_NAME record for any page
      operation on a non-predefined tablespace. If the MLOG_CHECKPOINT
      marker is missing, no redo log will be applied, and the system would
      be at the state of the checkpoint.
      
      fil_name_parse(): New function, to update the recv_spaces map based on
      MLOG_FILE_NAME and MLOG_FILE_DELETE records during recovery.
      
      recv_parse_or_apply_log_rec_body(), recv_parse_log_rec(): Add the
      parameter "apply". Do not apply file-level redo log records unless the
      entire mini-transaction has been recovered. Fail if an MLOG_FILE_NAME
      record is missing for a page-level operation.
      
      recv_recover_page_func(): Assert that no LSN is after the latest
      scanned redo log LSN.
      
      recv_parse_log_rec(): Check for some more log corruption.
      
      recv_parse_log_recs(): Add a parameter "store_to_hash" to control
      whether the records should be stored into recv_sys->addr_hash.
      Add a parameter "apply" to specify whether log records should be applied
      (apply=false during the first scan for MLOG_CHECKPOINT). Return true
      if an MLOG_CHECKPOINT record was seen for the first time.
      Improve DBUG_PRINT output, and detect some more log corruption.
      
      recv_scan_log_recs(): Add a parameter "store_to_hash" to control
      whether the records should be stored into recv_sys->addr_hash.
      
      recv_group_scan_log_recs(): Initialize the variables and data
      structures to begin reading redo log records. Add a parameter
      "last_phase" that is set when a multi-pass recovery is needed and we
      are scanning the redo log for a third time. In last_phase, we will
      invoke recv_apply_hashed_log_recs() to empty recv_sys->addr_hash
      between passes. If last_phase=false, we would stop filling
      recv_sys->addr_hash, only processing file-level redo log records.
      
      recv_init_crash_recovery(): Split some code into
      recv_init_crash_recovery_spaces(), to be invoked after the first call
      to recv_group_scan_log_recs().
      
      recv_recovery_from_checkpoint_start(): Invoke
      recv_group_scan_log_recs() up to 3 times if needed.
      After processing all redo log, write an MLOG_CHECKPOINT marker
      so that in case we will crash before making a checkpoint, the log
      will be replayed by subsequent crash recovery.
      
      checkpoint_now_set(): Avoid an infinite loop in case an MLOG_CHECKPOINT
      marker is the only thing that was written since the latest checkpoint.
      
      rb#4700r6
------------------------------------------------------------
revno: 7824
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-uintkorr
timestamp: Thu 2014-04-03 15:35:20 +0200
message:
  Bug#18481960 REPLACE LONGGET/LONGSTORE MACROS WITH INLINE FUNCTIONS
  
  For better type safety, replace longget/longstore and similar macros with
  inline functions
------------------------------------------------------------
revno: 7823
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-cmake
timestamp: Thu 2014-04-03 14:03:21 +0200
message:
  Bug#18510941: REMOVE CMAKE WORKAROUNDS FOR OLDER VERSIONS OF OS X/XCODE
  
  Remove workarounds for older OS X and XCode versions:
  - Always use Clang, even for 32 bit.
  - We no longer use different compilers for C and C++ (e.g. GCC+Clang)
  - CMake bug#9051 has been fixed for quite a while
  - Remove workaround for OS X < 10.5
------------------------------------------------------------
revno: 7822
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-cmake
timestamp: Thu 2014-04-03 12:48:59 +0200
message:
  Bug #18382225 MYSQL_CONFIG CAN'T HANDLE RELOCABLE PACKAGES THAT USES "LIB64" OR "-64" SUFFIX
  
  'lib' is hardcoded into mysql_config, so 'cmake -DINSTALL_LIBDIR=lib64' will not work.
  Use INSTALL_LIBDIR when generating mysql_config.
  
  mysql_config may be renamed to e.g. mysql_config-32, fix the basedir pattern matching.
------------------------------------------------------------
revno: 7821
committer: matthias.leich@oracle.com
branch nick: mysql-trunk
timestamp: Thu 2014-04-03 13:45:08 +0200
message:
  Remove some misleading comment.
------------------------------------------------------------
revno: 7820
committer: Georgi Kodinov <georgi.kodinov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-03 11:26:49 +0300
message:
  Bug #11746951 - 29965: USE DBNAME SHOULD INDICATE WARNING PRESENT
  
  Added code to notify number of warnings generated (if any) after
  the 'USE' command.
  
  No case for this bug is included as, currently, there are no ndb
  tests in trunk.
------------------------------------------------------------
revno: 7819
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-03 10:06:51 +0300
message:
  Fix Bug#18384390 WRONG STATISTICS WITH BIG ROW LENGTH AND PERSISTENT STATS
  
  Estimate the number of external pages when scanning any leaf page and subtract
  that estimate from index->stat_n_leaf_pages when calculating
  index->stat_n_diff_key_vals[].
  
  Approved by:	Kevin, Satya (rb:4956)
------------------------------------------------------------
revno: 7818
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Thu 2014-04-03 10:02:24 +0300
message:
  Non-functional change:
  
  Delay the calculation of each dict_index_t::stat_n_diff_key_vals[n_prefix]
  until after data is gathered for all n-column prefixes. This is a noop (non
  functional change) - still the same numbers will be used to calculate
  dict_index_t::stat_n_diff_key_vals[], but it will help fix
  Bug#18384390 WRONG STATISTICS WITH BIG ROW LENGTH AND PERSISTENT STATS
  
  The problem in that bug is that index->stat_n_leaf_pages is bloated by
  the number of externally stored pages, which should not really be counted
  in the formula that derives dict_index_t::stat_n_diff_key_vals[].
  
  To fix this we need to estimate the number of external pages in the
  index and subtract it from dict_index_t::stat_n_leaf_pages in that formula.
  
  In a subsequent change we will estimate the number of external pages while
  sampling each page for all possible n-column prefixes and then this cumulative
  result will be used then calculating each member of
  dict_index_t::stat_n_diff_key_vals[].
  
  The code before this change:
  
  dict_stats_analyze_index()
    for each n prefix
      dict_stats_analyze_index_for_n_prefix()
        sample some pages and save the n_diff results in
        index->stat_n_diff_key_vals[] using index->stat_n_leaf_pages in the
        formula
  
  The code after this change (equivalent):
  
  dict_stats_analyze_index()
    for each n prefix
      dict_stats_analyze_index_for_n_prefix()
        sample some pages and save the n_diff results in a temporary place
    for each n prefix
      // new function, code moved from dict_stats_analyze_index_for_n_prefix()
      dict_stats_index_set_n_diff()
        set index->stat_n_diff_key_vals[] using
        index->stat_n_leaf_pages in the formula
  
  Further planned change that will actually fix the bug:
  
  dict_stats_analyze_index()
    for each n prefix
      dict_stats_analyze_index_for_n_prefix()
        sample some pages and save the n_diff results in a temporary place
        and also accumulate an estimate about the number of external pages
        when scanning each leaf page
    for each n prefix
      dict_stats_index_set_n_diff()
        set index->stat_n_diff_key_vals[] using
        "index->stat_n_leaf_pages - number_of_external_pages" in the formula
  
  Approved by:	Satya (rb:4955)
------------------------------------------------------------
revno: 7817
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-clang-warning
timestamp: Wed 2014-04-02 16:30:28 +0200
message:
  Bug#18489724: FIX CLANG WARNINGS REPORTED BY -WUNUSED-PRIVATE-FIELD 
  
  Remove unused private fields as reported by Clang's -Wunused-private-field 
  compiler warning option.
------------------------------------------------------------
revno: 7816
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-18446462
timestamp: Wed 2014-04-02 14:43:55 +0200
message:
  Bug#18446462 ASSERT `ROWS >= 0.0' AT 
               COST_MODEL_SERVER::ROW_EVALUATE_COST IN 
               OPT_COSTMODEL.H
              
  <large_int_number> * 1.0 can be a slightly larger number which
  is problematic for calculate_scan_cost().
------------------------------------------------------------
revno: 7815
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-winwarn
timestamp: Wed 2014-04-02 12:58:38 +0200
message:
  Bug#11752665: WINDOWS SERVER X64: SO MANY COMPILER WARNINGS
  
  Patch #6:
  
  This patch reduces the compiler warning count on Win64 by an additional
  1400, as reported by VS2013.
------------------------------------------------------------
revno: 7814
committer: Atanu Ghosh <atanu.ghosh@oracle.com>
branch nick: mysql-trunk-new
timestamp: Wed 2014-04-02 13:58:45 +0530
message:
  Bug#11759453: UNINSTALL PLUGIN MISLEADING ERROR
  	      MESSAGE FOR NON-DYNAMIC PLUGINS
  
  Problem: For an existing non-dynamic plugin, the error message
           during uninstall - The plugin does exist - is misleading.
           It just cannot be uninstalled because it's not dynamically
           loaded, not that it does not exist.
  
  Fix: Previous to this fix, a warning (built-in plugin cannot be deleted)
       and an error message (plugin does not exist) was thrown.
       This fix renames the existing warning (WARN_PLUGIN_UNINSTALL_BUILTIN)
       to an error message ER_PLUGIN_UNINSTALL_BUILTIN, saying that
       built-in plugins cannot be uninstalled.
------------------------------------------------------------
revno: 7813
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-02 10:20:03 +0300
message:
  Non-functional change: elaborate the comment for dict_table_t::heap.
------------------------------------------------------------
revno: 7812
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 19:20:03 +0300
message:
  Fix Bug#71708 70768 fix perf regression: high rate of RW lock creation
  and destruction
  
  Lazily create dict_table_t::stats_latch the first time it is used.
  It may not be used at all in the lifetime of some dict_table_t objects.
  
  Approved by:	Bin (rb:4739)
------------------------------------------------------------
revno: 7811
committer: Vasil Dimov <vasil.dimov@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 15:36:46 +0300
message:
  Non-functional change: use InterlockedCompareExchange() instead of
  win_cmp_and_xchg_dword() for the macro os_compare_and_swap_uint32()
  on Windows. win_cmp_and_xchg_dword() only calls InterlockedCompareExchange()
  so it can be skipped altogether.
  
  The macro os_compare_and_swap_uint32() is not used anywhere in the code.
------------------------------------------------------------
revno: 7810 [merge]
committer: Aditya A <aditya.a@oracle.com>
branch nick: mysql-trunk
timestamp: Wed 2014-04-02 10:53:44 +0530
message:
  Bug #13029450	OFF BY ONE ERROR IN INNODB_MAX_DIRTY_PAGES_PCT
  		LOGIC
  
  [ Null Merge since issue already fixed in trunk ]
    ------------------------------------------------------------
    revno: 3902.291.325
    committer: Aditya A <aditya.a@oracle.com>
    branch nick: mysql-5.6
    timestamp: Wed 2014-04-02 10:50:50 +0530
    message:
      Bug #13029450	OFF BY ONE ERROR IN INNODB_MAX_DIRTY_PAGES_PCT
      		LOGIC
      
      If the percentage of dirty pages in the buffer pool
      exceeds innodb_max_dirty_pages_pct (set by the user)
      then we flush the pages.If user sets 
      innodb_max_dirty_pages_pct=0,then the flushing mechanism
      will not kick in unless the percentage of dirty pages 
      reaches at least 1%.For huge buffer pools even 1% of the
      buffer pool can be a huge number.
      
      FIX
      ---
      
      Flush the dirty pages in buffer pool if percentage of dirty 
      pages is greater than zero and innodb_max_dirty_pages_pct
      is set to zero.
      
      [Approved by vasil #rb4776] 
------------------------------------------------------------
revno: 7809
committer: kevin.lewis@oracle.com
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 22:45:55 -0500
message:
  Fix Visual Studio compiler warnings for InnoDB with appropriate static_casts.
------------------------------------------------------------
revno: 7808
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-bug18483776
timestamp: Tue 2014-04-01 18:39:12 +0200
message:
  Bug#18483776: ASSERTION FAILED: QUERY_SOURCE == 0,
                FILE SQL_PROFILE.CC, LINE 297
  
  This assert was triggered since PROFILING::set_query_source() by
  mistake was called twice for a prepared statement. On release
  builds this could lead to memory leaks.
  
  This bug was a regression introduced by the patch for Bug#17606098
  which moved the setting of the profiling query string from
  mysqld_stmt_execute() to Prepared_statement::execute_loop().
  This patch fixes the problem by moving the call back.
------------------------------------------------------------
revno: 7807
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-alarm2
timestamp: Tue 2014-04-01 16:52:22 +0200
message:
  Bug#18411456: REMOVE THR_ALARM.C 
  
  This patch removes thr_alarm.h and thr_alarm.c since they contain 
  almost only dead code. Remaining live code moved to mysqld.cc
  
  The patch also removes my_alarm.h and my_alarm.c and moves
  the code which is actually used to my_lock.c (only place where it is used).
------------------------------------------------------------
revno: 7806
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18396916loaddata
timestamp: Tue 2014-04-01 16:12:53 +0200
message:
  Bug#18396916 MAIN.OUTFILE_LOADDATA TEST FAILS ON ARM, AARCH64, PPC/PPC64
  
  The recorded results for the failing tests were wrong.
  They were introduced by the patch for
  Bug#30946 mysqldump silently ignores --default-character-set when used with --tab
  
  Correct results were returned for platforms where 'char' is implemented as unsigned.
  This was reported as 
  Bug#46895 Test "outfile_loaddata" fails (reproducible)
  Bug#11755168 46895: TEST "OUTFILE_LOADDATA" FAILS (REPRODUCIBLE)
  The patch for that bug fixed only parts of the problem,
  leaving the incorrect results in the .result file.
  
  Solution: use 'uchar' for field_terminator and line_terminator on all platforms.
  Also: remove some un-necessary casts, leaving the ones we actually need.
------------------------------------------------------------
revno: 7805
committer: Erlend Dahl <erlend.dahl@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 15:44:41 +0200
message:
  Reduce PB2 time, require --big-test to run some tests in Valgrind:
  
  innodb.innodb-import-partition
  innodb.innodb_update_time_wl6658
  main.merge_recover
  rpl.rpl_alter_repository
  rpl.rpl_extra_col_slave_innodb
  rpl.rpl_gtid_mode
  rpl.rpl_row_img_sanity
  
  Approved by Jon Olav over IM.
------------------------------------------------------------
revno: 7804
committer: Tor Didriksen <tor.didriksen@oracle.com>
branch nick: trunk-bug18486249decimal
timestamp: Tue 2014-04-01 14:06:41 +0200
message:
  Bug#18486249 ASSERTION FAILED: BUF == BUFFER IN MY_DECIMAL::SANITY_CHECK
  
  This bug was uncovered by the stricter sanity_check introduced in patch for:
  Bug#18335446 DECIMAL_ROUND: ASSERTION FAILED: FROM->LEN == TO->LEN
  
  The problem is that my_qsort2 sorts blobs, rather than objects.
  When sorting an array of decimals, it will sometimes leave
  objects with wrong buffer pointers.
  
  Solution: use std::sort, which is type-aware, and will copy/swap
  decimal objects properly.
------------------------------------------------------------
revno: 7803
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-wl6635
timestamp: Tue 2014-04-01 13:25:30 +0200
message:
  WL#6635: Check for dependent subqueries in IN list
------------------------------------------------------------
revno: 7802
committer: Vamsikrishna Bhagi <vamsikrishna.bhagi@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 14:27:59 +0530
message:
  Bug #18349102 RESULT CONTENTS MISMATCH IN THE
                NIST.NIST_ALL TEST
  
  info message in OK packet which is assigned to mysql->info
  does not have a null terminator which results in having
  garbage values in addition to expected string.
  
  Solution: mysql->info_buffer is used to hold the info and
            proper null termination is done. mysql->info
            points to this location.
------------------------------------------------------------
revno: 7801
committer: Jorgen Loland <jorgen.loland@oracle.com>
branch nick: mysql-trunk-review
timestamp: Tue 2014-04-01 10:37:00 +0200
message:
  Bug#18364815: OPTIMIZER PREFERS TABLE SCAN WHEN USING "IN" 
                WITH VALUE OF DIFFERENT TYPE
        
  Forgot to record result file
------------------------------------------------------------
revno: 7800
committer: viswanatham gudipati <viswanatham.gudipati@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 12:58:09 +0530
message:
  Disbaled testcases which are failing due to BUG#18449085
------------------------------------------------------------
revno: 7799 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 11:39:16 +0530
message:
  Bug #17858679   TOO MANY TIMES OF MEMSET DECREASE
                  THE PERFORMANCE UNDER HEAVY INSERT
          Null merge from mysql-5.6
    ------------------------------------------------------------
    revno: 3902.291.324 [merge]
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-01 11:38:02 +0530
    message:
      Bug #17858679   TOO MANY TIMES OF MEMSET DECREASE
                      THE PERFORMANCE UNDER HEAVY INSERT
              Null merge from mysql-5.5
        ------------------------------------------------------------
        revno: 2875.592.12
        committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-04-01 11:36:58 +0530
        message:
          Bug #17858679   TOO MANY TIMES OF MEMSET DECREASE
          		THE PERFORMANCE UNDER HEAVY INSERT
          	Fixing the build problem in 5.5.
------------------------------------------------------------
revno: 7798 [merge]
committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
branch nick: mysql-trunk
timestamp: Tue 2014-04-01 10:52:19 +0530
message:
  Bug #17858679   TOO MANY TIMES OF MEMSET DECREASE
                  THE PERFORMANCE UNDER HEAVY INSERT
          Merge from 5.6
    ------------------------------------------------------------
    revno: 3902.291.323 [merge]
    committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
    branch nick: mysql-5.6
    timestamp: Tue 2014-04-01 10:49:54 +0530
    message:
      Bug #17858679   TOO MANY TIMES OF MEMSET DECREASE
                      THE PERFORMANCE UNDER HEAVY INSERT
      	Merge from 5.5
        ------------------------------------------------------------
        revno: 2875.592.11
        committer: Thirunarayanan B<thirunarayanan.balathandayuth@oracle.com>
        branch nick: mysql-5.5
        timestamp: Tue 2014-04-01 10:46:13 +0530
        message:
          Bug #17858679	TOO MANY TIMES OF MEMSET DECREASE
          		THE PERFORMANCE UNDER HEAVY INSERT
          Problem:
          	There are three memset call to allocate memory for system fields
          in each insert.
          
          Solution:
          	Instead of calling it in 3 times, we can combine it into
          one memset call. It will reduce the CPU usage under heavy insert.
          
          	Approved by Marko rb-4916
------------------------------------------------------------
revno: 7797
committer: Benny.Wang<Benny.Wang@oracle.com>
branch nick: trunk
timestamp: Tue 2014-04-01 04:57:33 +0200
message:
  Add test cases for bug#17458054.
------------------------------------------------------------
revno: 7796
committer: michael.izioumtchenko@oracle.com
branch nick: mysql_src_cmt
timestamp: Mon 2014-03-31 18:38:42 +0200
message:
  WL#7535 
  
  Implements n testing framework as an internal part of MySQL source tree.
  Allows running n tests for mysql-trunk with no setup.
  
  See https://mediawiki.us.oracle.com/mysql/index.php/N_for_developers for a how-to.
  
  Approved by: Marko [rb#4661]
------------------------------------------------------------
revno: 7795 [merge]
committer: Joao Gramacho <joao.gramacho@oracle.com>
branch nick: mysql-trunk
timestamp: Mon 2014-03-31 17:00:33 +0100
message:
  Merge of BUG#18482854 patch from mysql-5.6 into mysql-trunk
    ------------------------------------------------------------
    revno: 3902.291.322
    committer: Joao Gramacho <joao.gramacho@oracle.com>
    branch nick: mysql-5.6
    timestamp: Mon 2014-03-31 16:53:58 +0100
    message:
      BUG#18482854 RPL : ROTATE_LOG_EVENT INCORRECTLY ADVANCES
                         GROUP_RELAY_LOG_POS IN A GROUP
      
      Problem:
      =======
      
      Rotate events can cause the group_relay_log_pos to be incorrectly moved
      forward within a group.
      
      This means that when the transaction is retried, or if you stop the SQL
      thread in the middle of a transaction after some Rotates (considering
      that the transaction/group was spanned into multiple relay log files),
      the group (or part of the group from the beginning) will be silently
      skipped.
      
      Analysis:
      ========
      
      It was found a problem in the logic for avoiding to touch SQL thread
      coordinates in Rotate_log_event::do_update_pos().
      
      The logic was allowing to update SQL thread coordinates if not using
      MTS regardless rli->is_in_group() signaling that it was in the middle
      of a group.
      
      Fix:
      ===
      
      Corrected the logic.
------------------------------------------------------------
revno: 7794 [merge]
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Mon 2014-03-31 15:47:54 +0200
message:
  Merge from mysql-5.6 to mysql-trunk
    ------------------------------------------------------------
    revno: 3902.291.321
    committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
    branch nick: mysql-5.6-test
    timestamp: Mon 2014-03-31 15:09:59 +0200
    message:
      Backport from mysql-trunk to mysql-5.6 of:
      ------------------------------------------------------------
      revno: 7033
      committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
      branch nick: mysql-trunk-c11
      timestamp: Wed 2013-11-27 13:54:59 +0100
      message:
        Bug#14631159: ALLOW COMPILATION USING CLANG IN C++11 MODE
        
        This patch fixes the new compilation errors that are reported by Clang and GCC
        when compiling in C++11 mode. 
        
        The patch is not based on the contribution in the bug report.
------------------------------------------------------------
revno: 7793 [merge]
committer: Bjorn Munch <bjorn.munch@oracle.com>
branch nick: main-tr
timestamp: Mon 2014-03-31 15:09:49 +0200
message:
  Merge from 5.7.4-m14-release
    ------------------------------------------------------------
    revno: 7618.1.8
    tags: mysql-5.7.4
    committer: Bjorn Munch <bjorn.munch@oracle.com>
    branch nick: rel-574
    timestamp: Wed 2014-03-19 11:32:19 +0100
    message:
      Fix for bugs 18402580 and 18402999, as suggested for 18402580
      
        Instead of relying on $HOME, use Perl's getpwuid() to get home dir.
