Merge joerg@mysql.com-20100322150231-vdq0afbqtmbs6phy from BZR,

Including univ.i before mysql/plugin.h is needed to avoid this
compiler error:

o  This is how gcc puts it:
o  > > ccache /usr/local/gcc-4.3.2/bin/gcc -static-libgcc -DHAVE_CONFIG_H -I. -I../../include -I../../include -I../../include -I../../regex -I./include -I../../sql -I. -I../../zlib    -g -O3 -march=i686   -DUNIV_LINUX -MT libinnobase_a-trx0i_s.o -MD -MP -MF .deps/libinnobase_a-trx0i_s.Tpo -c -o libinnobase_a-trx0i_s.o `test -f 'trx/trx0i_s.c' || echo './'`trx/trx0i_s.c
o  > > In file included from ./include/univ.i:114,
o  > >                  from trx/trx0i_s.c:36:
o  > > ../../include/my_pthread.h:628: error: expected ')' before '*' token
o  > > In file included from ../../include/my_pthread.h:732,
o  > >                  from ./include/univ.i:114,
o  > >                  from trx/trx0i_s.c:36:
o  > > ../../include/mysql/psi/mysql_thread.h:100: error: expected specifier-qualifier-list before 'pthread_rwlock_t'
o  > > ../../include/mysql/psi/mysql_thread.h:116: error: expected specifier-qualifier-list before 'pthread_rwlock_t'
o  > > ../../include/mysql/psi/mysql_thread.h: In function 'inline_mysql_rwlock_init':
o  > > ../../include/mysql/psi/mysql_thread.h:711: error: 'mysql_rwlock_t' has no member named 'm_psi'
o  > > ../../include/mysql/psi/mysql_thread.h:716: error: 'mysql_rwlock_t' has no member named 'm_rwlock'
o  > > .... ((continued))
o  
o  Intel's icc gives slightly clearer messages:
o  > > icc -static-intel -static-libgcc -DHAVE_CONFIG_H -I. -I../../include -I../../include -I../../include -I../../regex -I./include -I../../sql -I. -I../../zlib    -O3 -g -unroll2 -ip -mp -restrict -no-ftz -no-prefetch   -DUNIV_LINUX -MT libinnobase_a-trx0i_s.o -MD -MP -MF .deps/libinnobase_a-trx0i_s.Tpo -c -o libinnobase_a-trx0i_s.o `test -f 'trx/trx0i_s.c' || echo './'`trx/trx0i_s.c
o  > > ../../include/my_pthread.h(628): error: identifier "pthread_rwlock_t" is undefined
o  > >   extern int rw_pr_init(rw_pr_lock_t *);
o  > >                         ^
o  > > 
o  > > ../../include/mysql/psi/mysql_thread.h(100): error: identifier "pthread_rwlock_t" is undefined
o  > >     rw_lock_t m_rwlock;
o  > >     ^
o  > > 
o  > > ../../include/mysql/psi/mysql_thread.h(116): error: identifier "pthread_rwlock_t" is undefined
o  > >     rw_pr_lock_t m_prlock;
o  > >     ^

to speed up insertions into the flush_list and thus the recovery
process. The patch has been tested by Nokia.

comparing to builtin InnoDB (Windows only).

Disable Windows atomics by default.

Approved by: Inaam

when comparing the age of access_time to buf_LRU_old_threshold_ms.
This fixes a bug on 64-bit systems.

HAVE_IB_PAUSE_INSTRUCTION in CMakeLists.txt.

The rename was done as r5871, but CMakeLists.txt was
forgotten. Also, add INNODB_RW_LOCKS_USE_ATOMICS to
CMake.

Fixed compiler warning about uninitialized variable.

Non-functional change.

allowing mtr->state == MTR_COMMITTING.

zip only test innodb_bug44571 needs to be updated to reflect the
column name change would be successful be done in InnoDB as well.

buf_block_t::check_index_page_at_flush: Note that this field is not
protected by any mutex. Make it a separate field, not a bitfield that
could share the machine word with other fields.

buf_block_t::lock_hash_val: Note that this field is protected by
buf_block_t::lock (or during block creation, by buf_pool_mutex and
buf_block_t::mutex).

buf_block_get_lock_hash_val(): Assert that block->lock is held by the
current thread.

Issue #465, rb://267 approved by Inaam Rana

  ------------------------------------------------------------------------
  r6774 | calvin | 2010-03-03 23:56:10 +0200 (Wed, 03 Mar 2010) | 2 lines
  Changed paths:
     M /branches/5.1/trx/trx0sys.c

  branches/5.1: fix bug#51653: outdated reference to set-variable
  Non functional change.
  ------------------------------------------------------------------------
  r6780 | vasil | 2010-03-08 19:13:20 +0200 (Mon, 08 Mar 2010) | 4 lines
  Changed paths:
     M /branches/5.1/plug.in

  branches/5.1:

  Whitespace fixup.
  ------------------------------------------------------------------------
  r6783 | jyang | 2010-03-09 17:54:14 +0200 (Tue, 09 Mar 2010) | 9 lines
  Changed paths:
     M /branches/5.1/handler/ha_innodb.cc
     M /branches/5.1/mysql-test/innodb_bug21704.result
     A /branches/5.1/mysql-test/innodb_bug47621.result
     A /branches/5.1/mysql-test/innodb_bug47621.test

  branches/5.1: Fix bug #47621 "MySQL and InnoDB data dictionaries
  will become out of sync when renaming columns". MySQL does not
  provide new column name information to storage engine to
  update the system table. To avoid column name mismatch, we shall
  just request a table copy for now.

  rb://246 approved by Marko.
  ------------------------------------------------------------------------
  r6785 | vasil | 2010-03-10 09:04:38 +0200 (Wed, 10 Mar 2010) | 11 lines
  Changed paths:
     M /branches/5.1/mysql-test/innodb_bug38231.test

  branches/5.1:

  Add the missing --reap statements in innodb_bug38231.test. Probably MySQL
  enforced the presence of those recently and the test started failing like:

    main.innodb_bug38231                     [ fail ]
            Test ended at 2010-03-10 08:48:32

    CURRENT_TEST: main.innodb_bug38231
    mysqltest: At line 49: Cannot run query on connection between send and reap
  ------------------------------------------------------------------------
  r6788 | vasil | 2010-03-10 10:53:21 +0200 (Wed, 10 Mar 2010) | 8 lines
  Changed paths:
     M /branches/5.1/mysql-test/innodb_bug38231.test

  branches/5.1:

  In innodb_bug38231.test: replace the fragile sleep 0.2 that depends on timing
  with a more robust condition which waits for the TRUNCATE and LOCK commands
  to appear in information_schema.processlist. This could also break if there
  are other sessions executing the same SQL commands, but there are none during
  the execution of the mysql test.
  ------------------------------------------------------------------------

with concurrent ddl". Null terminate the name string returned
from innobase_convert_identifier() call when reporting DB_DUPLICATE_KEY
error in create_table_def().
rb://266 approved by Marko

unique index cannot be upgrade to primary index even if there
is no primary index already defined. Also fix possible corruption
when initialize "ref_length" value in case there is a mismatch
between MySQL and InnoDB primary key. Fix bug #51378: "Init
'ref_length'  to correct value, in case an out of bound MySQL
primary_key".
rb://262 approved by Marko.

TODO: We might also consider removing recv_max_parsed_page_no, because
it does not make much sense with *.ibd files.

recv_report_corrupt_log(), recv_scan_log_recs(): Abort when a
corrupted log record has been found, unless innodb_force_recovery has
been set.

This fixes Issue #464.
rb://265 approved by Heikki Tuuri

Fix typo in comment

for block mutexes and blocks.

Designed by Michael and Marko. rb://188, Issue #358

a wrong parameter was passed to buf_flush_init_for_writing().

fil_reset_too_high_lsns(): Set up page_zip and use it if needed.

rb://264, Issue #352

about a possibly uninitialized variable.

of possibly uninitialized variable format_in_hex.

This should fix Mantis Issue #461.

innodb.test, innodb.result, innodb-index.test, innodb-index.result:
Adjust the test result and mention that the introduced restriction
has been reported as MySQL Bug #51451.

innobase_check_index_keys(): Add a parameter for the InnoDB table and
check that no duplicate index name is added.  Report errors by
my_error() instead of sql_print_error().

rb://260 approved by Sunny Bains

Add a NOTE to the comment of btr_node_ptr_get_child_page_no()
to prevent mysterious bugs.

Non-functional change.

The merge sort can use up to 48KiB of buffers when merging blocks.
That can cause a stack overflow, especially on 64-bit systems when not
building with inlined functions.  This was reported as Issue #462.

row_merge_dup_report(): Allocate buf and offsets from a heap.

row_merge_heap_create(): Allocate space for buf[3] too. Fix bogus
sizeof arithmetics that happened to work, because
sizeof(ulint)==sizeof(void*).

row_merge_blocks(), row_merge_blocks_copy(): Allocate buf[3] from heap.

row_merge_insert_index_tuples(): Allocate buf from graph_heap.

rb://258 approved and tested by Sunny Bains

This was added to trunk in r435.

that have been modified after 2010-01-01 according to svn.

for f in $(svn log -v -r{2010-01-01}:HEAD |grep "^   M " |cut -b 16- |sort -u) ; do sed -i "" -E 's/(Copyright \(c\) [0-9]{4},) [0-9]{4}, (.*Innobase Oy.+All Rights Reserved)/\1 2010, \2/' $f ; done

found in the buffer pool. Only print the bits that are set and that
information is in the lock and not in the block.

See rb://256 approved by Marko.

This fixes the FILE_FLUSH_LSN printouts mentioned in Issue #341.
Suggested by Heikki.

for Valgrind.

for the first page of each ibdata* file, not *.ibd files.
Suggested by Heikki, in connection with the LSN warning noted in Issue #341.

  ------------------------------------------------------------------------
  r6669 | jyang | 2010-02-11 12:24:19 +0200 (Thu, 11 Feb 2010) | 7 lines

  branches/5.1: Fix bug #50691, AIX implementation of readdir_r
  causes InnoDB errors. readdir_r() returns an non-NULL value
  in the case of reaching the end of a directory. It should
  not be treated as an error return.

  rb://238 approved by Marko
  ------------------------------------------------------------------------

Return an incompatible lock or NULL instead of TRUE or FALSE.
Approved by Sunny over IM.

in UNIV_LOG_LSN_DEBUG.  This could have caused Issue #440.