forgotten when I reverted the kernel mutex split patch.

rb://335, Approved by Sunny Bains

  ------------------------------------------------------------
  revno: 3462
  revision-id: marko.makela@oracle.com-20100512104212-e2h9n1obxjb8tfg4
  parent: marko.makela@oracle.com-20100512060922-oh716ugpukfakwvh
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb
  timestamp: Wed 2010-05-12 13:42:12 +0300
  message:
    ha_innobase::add_index(): Reset trx->error_state in error handling.
    (Bug #53591)

Support returning 512 and 511 pages for the buffer pool size, this
is undeterministic and probably depends on alignment issues.

The default buffer pool size is 8M (512) pages, which is set in
include/default_mysqld.cnf. So the previous "replace_result 8192 8191"
had no effect.

The --send command does not seem to work as expected in that mode.

  ------------------------------------------------------------
  revno: 3460
  revision-id: marko.makela@oracle.com-20100512053925-ovwmpo0838fs2roo
  parent: marko.makela@oracle.com-20100511165845-lsw7seixftgzpfqt
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb
  timestamp: Wed 2010-05-12 08:39:25 +0300
  message:
    row_merge_drop_temp_indexes(): Do not reference freed memory.
    (Bug #53471)

the revert of kernel mutex split patch.

in the code but they have nothing to do with the kernel mutex split code.
Some subsequent commits use the new functions. This patch has been tested
with: ./mtr --suite=innodb with UNIV_DEBUG and UNIV_SYNC_DEBUG enabled.
All tests were successful.

  ------------------------------------------------------------
  revno: 3459
  revision-id: marko.makela@oracle.com-20100511165845-lsw7seixftgzpfqt
  parent: vasil.dimov@oracle.com-20100511110029-18xvuu6r7har8toh
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb2
  timestamp: Tue 2010-05-11 19:58:45 +0300
  message:
    Fix sys_vars.tx_isolation_func.test, which was broken in revno 3432
    when making READ UNCOMMITTED lock as little as READ COMMITTED.

  ------------------------------------------------------------
  revno: 3459
  revision-id: marko.makela@oracle.com-20100511105308-grp2t3prh3tqivw0
  parent: marko.makela@oracle.com-20100511105012-b2t7wvz6mu6bll74
  parent: marko.makela@oracle.com-20100505123901-xjxu93h1xnbkfkq0
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb
  timestamp: Tue 2010-05-11 13:53:08 +0300
  message:
    Merge a patch from Facebook to fix Bug #53290

    commit e759bc64eb5c5eed4f75677ad67246797d486460
    Author: Ryan Mack
    Date:   3 days ago

        Bugfix for 53290, fast unique index creation fails on duplicate null values

        Summary:
        Bug in the fast index creation code incorrectly considers null
        values to be duplicates during block merging.  Innodb policy is that
        multiple null values are allowed in a unique index.  Null duplicates
        were correctly ignored while sorting individual blocks and with slow
        index creation.

        Test Plan:
        mtr, including new test, load dbs using deferred index creation

        License:
        Copyright (C) 2009-2010 Facebook, Inc.  All Rights Reserved.

        Dual licensed under BSD license and GPLv2.

        Redistribution and use in source and binary forms, with or without
        modification, are permitted provided that the following conditions are met:
        1. Redistributions of source code must retain the above copyright notice,
           this list of conditions and the following disclaimer.
        2. Redistributions in binary form must reproduce the above copyright notice,
           this list of conditions and the following disclaimer in the documentation
           and/or other materials provided with the distribution.

        THIS SOFTWARE IS PROVIDED BY FACEBOOK, INC. ``AS IS'' AND ANY EXPRESS OR
        IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
        MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO
        EVENT SHALL FACEBOOK, INC. BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
        SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
        PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
        OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
        WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
        OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
        ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

        This program is free software; you can redistribute it and/or modify it
        under the terms of the GNU General Public License as published by the Free
        Software Foundation; version 2 of the License.

        This program is distributed in the hope that it will be useful, but WITHOUT
        ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
        FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
        more details.

        You should have received a copy of the GNU General Public License along with
        this program; if not, write to the Free Software Foundation, Inc., 59 Temple
        Place, Suite 330, Boston, MA  02111-1307  USA
      ------------------------------------------------------------
      revno: 3453.2.1
      revision-id: marko.makela@oracle.com-20100505123901-xjxu93h1xnbkfkq0
      parent: marko.makela@oracle.com-20100505120555-ukoq1gklpheslrxs
      committer: Marko Mäkelä <marko.makela@oracle.com>
      branch nick: 5.1-innodb
      timestamp: Wed 2010-05-05 15:39:01 +0300
      message:
        Merge a contribution from Ryan Mack at Facebook:

        Bugfix for 53290, fast unique index creation fails on duplicate null values

            Summary:
            Bug in the fast index creation code incorrectly considers null
            values to be duplicates during block merging.  Innodb policy is that
            multiple null values are allowed in a unique index.  Null duplicates
            were correctly ignored while sorting individual blocks and with slow
            index creation.

            Test Plan:
            mtr, including new test, load dbs using deferred index creation

            DiffCamp Revision: 110840
            Reviewed By: mcallaghan
            CC: mcallaghan, mysql-devel@lists
            Revert Plan:
            OK

  ------------------------------------------------------------
  revno: 3458
  revision-id: marko.makela@oracle.com-20100511105012-b2t7wvz6mu6bll74
  parent: marko.makela@oracle.com-20100511104910-nim8kgguawpis7zo
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb
  timestamp: Tue 2010-05-11 13:50:12 +0300
  message:
    Do not demand that buf_page_t be fully initialized on 64-bit systems.
    There may be padding before buf_page_t::zip. (Bug #53307)

  ------------------------------------------------------------
  revno: 3457
  revision-id: marko.makela@oracle.com-20100511104910-nim8kgguawpis7zo
  parent: marko.makela@oracle.com-20100511104500-c6kzd0bg5s42p8e9
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb
  timestamp: Tue 2010-05-11 13:49:10 +0300
  message:
    btr_page_split_and_insert(): Add an assertion
    suggested by Sunny Bains when reviewing Bug #52964.

  ------------------------------------------------------------
  revno: 3456
  revision-id: marko.makela@oracle.com-20100511104500-c6kzd0bg5s42p8e9
  parent: vasil.dimov@oracle.com-20100510132852-cz457uqvj8iiy9mm
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: mysql-5.1-innodb2
  timestamp: Tue 2010-05-11 13:45:00 +0300
  message:
    Remove a stray expression. Spotted by Sunny Bains.

before trying to access the transaction instance.

rb://336

when the purge thread completes by calling srv_suspend_thread().

for the new parameters introduced in the plugin 1.1. Now that we have
basic tests for all these new parameters it is no longer needed to
adjust the all_vars.result file.

  ------------------------------------------------------------
  revno: 3454
  committer: Marko Makela <mmakela@bk-internal.mysql.com>
  branch nick: mysql-5.1-innodb
  timestamp: Mon 2010-05-10 13:37:52 +0200
  message:
    Add an innodb test case for Bug #49164.

  ------------------------------------------------------------
  revno: 3450
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: 5.1-innodb
  timestamp: Wed 2010-05-05 14:24:11 +0300
  message:
    row_merge_drop_temp_indexes(): Load the table via the dictionary cache.
    Allow multiple indexes to be dropped. (Bug #53256)

Check that all background threads are suspended or shutdown instead of just
checking for the master thread.
rb://333

Change types of srv_purge_batch_size and srv_n_purge_threads
from ulint to ulong.

Suggested by Vlad.

  ------------------------------------------------------------
  revno: 3449
  revision-id: marko.makela@oracle.com-20100505104425-39y6qbffgotrhck2
  parent: marko.makela@oracle.com-20100505104001-883pqiepo384qr5h
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: 5.1-innodb
  timestamp: Wed 2010-05-05 13:44:25 +0300
  message:
    Factor out innodb_multi_update.test from innodb.test

      
The bug happened under the following condition:
  - there was a user variable of type REAL, containing NULL value
  - there was a table with a NOT_NULL column of any type but REAL, having
    default value (or auto increment);
  - a row was inserted into the table with the user variable as value.
    A warning was emitted here.

The problem was that handling of NULL values of REAL type was not properly
implemented: it didn't expect that REAL NULL value can be assigned to other
data type.

Basically, the problem was that set_field_to_null() was used instead of
set_field_to_null_with_conversions().

The fix is to use the right function, or more generally, to allow conversion of
REAL NULL values to other data types.

  
  Problem:
  item->name was NULL for Item_user_var_as_out_param
  which made strcmp(something, item->name) crash in the LOAD XML code.
  
  Fix:
  - item_func.h: Adding set_name() in constuctor for Item_user_var_as_out_param
  - sql_load.cc: Changing the condition in write_execute_load_query_log_event() which
  distiguished between Item_user_var_as_out_param and Item_field
  from
    if (item->name == NULL)
  to
    if (item->type() == Item::FIELD_ITEM)
  - loadxml.result, loadxml.test: adding tests

  ------------------------------------------------------------
  revno: 3446
  revision-id: marko.makela@oracle.com-20100505100507-6kcd2hf32hruxbv7
  parent: marko.makela@oracle.com-20100505095328-vetnl0flhmhao7p5
  committer: Marko Mäkelä <marko.makela@oracle.com>
  branch nick: 5.1-innodb
  timestamp: Wed 2010-05-05 13:05:07 +0300
  message:
    Add Valgrind diagnostics to track down Bug #38999.
  ------------------------------------------------------------

Problem: after introduction of "WL#2649 Number-to-string conversions"

This query:
  SET NAMES cp850; -- Or any other non-latin1 ASCII-based character set
  SELECT * FROM t1
  WHERE datetime_column='2010-01-01 00:00:00'
started to add extra character set conversion: 
  SELECT * FROM t1
  WHERE CONVERT(datetime_column USING cp850)='2010-01-01 00:00:00';

so index on DATETIME column was not used anymore.
Fix:
  avoid convertion of NUMERIC/DATETIME items
  (i.e. those with derivation DERIVATION_NUMERIC).