into  pilot.(none):/data/msvensson/mysql/mysql-5.0-maint

into  pilot.(none):/data/msvensson/mysql/mysql-5.0-maint


mysql-test/t/sp.test:
  Auto merged

 - Use SQL for diffing master and slave


mysql-test/r/rpl_misc_functions.result:
  Dump t1 on slave and load it back into temporary table on master
  to allow comapre with SQL
mysql-test/t/rpl_misc_functions.test:
  Dump t1 on slave and load it back into temporary table on master
  to allow comapre with SQL

into  pilot.(none):/data/msvensson/mysql/mysql-5.0-maint

 - Move the check of too long path to 'get_one_option'


client/mysqldump.c:
  Move the check of too long path to 'get_one_option'
mysql-test/r/mysqldump.result:
  Update result file after changing error message

sql/Makefile.am:
  Avoid "noinst_HEADERS" being sucked into mysql_LDADD 
  target if @MYSQLD_EXTRA_IBS@ are empty

mysql-test/t/sp.test:
  Change "exec rm" to "remove_file"
  Change two "remove file if exists" to "check that file not exist"

into  damien-katzs-computer.local:/Users/dkatz/mysql50

show that shm communication still works on windows,
and that we can't overflow the base-name.


mysql-test/t/windows_shm-master.opt:
  Bug#24924: shared-memory-base-name that is too long causes buffer overflow
  
  start a server with shm communication if we're on
  windows.
mysql-test/t/windows_shm.test:
  Bug#24924: shared-memory-base-name that is too long causes buffer overflow
  
  .opt has started a server with shm communication
  if we're on windows. now start a client with shm
  and connect to that server.

Windows fixies

into  sin.intern.azundris.com:/home/tnurnberg/22540/50-22540

windows fixies

into  pilot.(none):/data/msvensson/mysql/mysql-5.0-maint


mysql-test/lib/mtr_process.pl:
  Auto merged

into  pilot.(none):/data/msvensson/mysql/mysql-5.0-maint

into  sin.intern.azundris.com:/home/tnurnberg/22540/50-22540

into  damien-katzs-computer.local:/Users/dkatz/50

Added more sleep time befoe reap to allow query to be executed.


mysql-test/t/status.test:
  Added more sleep time befoe reap to allow query to be executed.

into  maint1.mysql.com:/data/localhome/tsmith/bk/maint/50

into  amd64.(none):/src/bug24732/my50-bug24732

- Corrected manifest attributes.


CMakeLists.txt:
  Bug#24732 Executables do not include Vista manifests
  - Use correct value for 64 built binaries.
win/create_manifest.js:
  Bug#24732 Executables do not include Vista manifests
  - Remove type attribute as the only valid value is win32.
win/mysql_manifest.cmake:
  Bug#24732 Executables do not include Vista manifests
  - Remove type attribute.

 - Various fixes for running mysql-test-run.pl on ActiveState perl
   for windows. Reading it's source in win32/win32.c helps...


mysql-test/lib/mtr_process.pl:
  Use "real_pid" when killing the process on ActiveState perl for windows
  Wait blocking for the pseudo pid to exit.
  Change "sleep_until_file_created" to return 1 when pidfile has been
  created - this should avoid early wakeup if $pid was 0
mysql-test/lib/mtr_timer.pl:
  Wake timer process with signal 15 to avoid to avoid resource leak
  on ActiveState perl for windows.
  Install signal handler in timer process to  exit gracefully
mysql-test/mysql-test-run.pl:
  Read "real_pid" of process from pid_file - to be used when killing the
  process on ActiveState perl
  Drop the --console option to mysqld
  Pass "real_pid" to mtr_check_stop_server so it can select to use it
  Correct two argument declarations

into  pilot.(none):/data/msvensson/mysql/bug25657/my50-bug25657-new2

end_log_pos data within a transaction are relative to
the start of the transaction rather than absolute.
we fix those groups in situ before writing the log out.

additional comments and handling for groups with very
large single events, as suggested by Guilhem.

fix binlog-writing so that end_log_pos is given correctly even
within transactions for both SHOW BINLOG and SHOW MASTER STATUS,
that is as absolute values (from log start) rather than relative
values (from transaction's start).
---
Merge tnurnberg@bk-internal.mysql.com:/home/bk/mysql-5.0-maint
into  sin.intern.azundris.com:/home/tnurnberg/22540/50-22540


mysql-test/r/binlog.result:
  Bug #22540: Incorrect value in column End_log_pos of SHOW BINLOG EVENTS using InnoDB
  
  show that end_log_pos in SHOW BINLOG EVENTS is correct even in transactions.
  show that SHOW MASTER STATUS returns correct values while in transactions
  (so that mysqldump --master-data will work correctly).
  also remove bdb dependency.
  ---
  manual merge
mysql-test/t/binlog.test:
  Bug #22540: Incorrect value in column End_log_pos of SHOW BINLOG EVENTS using InnoDB
  
  show that end_log_pos in SHOW BINLOG EVENTS is correct even in transactions.
  show that SHOW MASTER STATUS returns correct values while in transactions
  (so that mysqldump --master-data will work correctly).
  also remove bdb dependency.
sql/log.cc:
  Bug #22540: Incorrect value in column End_log_pos of SHOW BINLOG EVENTS using InnoDB
  
  fix output for SHOW BINLOG EVENTS so that end_log_pos is given correctly
  even within transactions. do this by rewriting the commit-buffer in place.

into  gleb.loc:/home/uchum/work/bk/5.0-opt

into  gleb.loc:/home/uchum/work/bk/5.0-opt

into  gleb.loc:/home/uchum/work/bk/4.1-opt

into  olga.mysql.com:/home/igor/dev-opt/mysql-5.0-opt-bug25602


sql/sql_select.cc:
  Auto merged

into  gleb.loc:/home/uchum/work/bk/5.0-opt


sql/log_event.cc:
  Auto merged

the loose scan optimization for grouping queries was applied returned 
a wrong result set when the query was used with the SQL_BIG_RESULT
option.

The SQL_BIG_RESULT option forces to use sorting algorithm for grouping
queries instead of employing a suitable index. The current loose scan
optimization is applied only for one table queries when the suitable
index is covering. It does not make sense to use sort algorithm in this
case. However the create_sort_index function does not take into account
the possible choice of the loose scan to implement the DISTINCT operator
which makes sorting unnecessary. Moreover the current implementation of
the loose scan for queries with distinct assumes that sorting will
never happen. Thus in this case create_sort_index should not call
the function filesort.


mysql-test/r/group_min_max.result:
  Added a test case for bug #25602.
mysql-test/t/group_min_max.test:
  Added a test case for bug #25602.

into  gleb.loc:/home/uchum/work/bk/5.0-opt

into  gleb.loc:/home/uchum/work/bk/5.0-opt

INSERT into table from SELECT from the same table
with ORDER BY and LIMIT was inserting other data
than sole SELECT ... ORDER BY ... LIMIT returns.

One part of the patch for bug #9676 improperly pushed
LIMIT to temporary table in the presence of the ORDER BY
clause.
That part has been removed.


sql/sql_select.cc:
  Fixed bug #29095.
  One part of the patch for bug #9676 improperly pushed
  LIMIT to temporary table in the presence of the ORDER BY
  clause.
  That part has been removed.
mysql-test/t/insert_select.test:
  Expanded the test case for bug #9676.
  Created a test case for bug #29095.
mysql-test/r/insert_select.result:
  Expanded the test case for bug #9676.
  Created a test case for bug #29095.

into  mysql.com:/home/ram/work/b29079/b29079.5.0

into  mysql.com:/home/ram/work/b29079/b29079.5.0


include/my_global.h:
  Auto merged

into  trift2.:/MySQL/M50/push-5.0

to be run only if it is available on the machine.

into  magare.gmz:/home/kgeorge/mysql/autopush/B28400-5.0-opt

The C optimizer may decide that data access operations
through pointer of different type are not related to 
the original data (strict aliasing).
This is what happens in fetch_long_with_conversion(),
when called as part of mysql_stmt_fetch() : it tries 
to check for truncation errors by first storing float
(and other types of data) into a char * buffer and then 
accesses them through a float pointer.
This is done to prevent the effects of excess precision
when using FPU registers.
However the doublestore() macro converts a double pointer
to an union pointer. This violates the strict aliasing rule.
Fixed by making the intermediary variables volatile (
to not re-introduce the excess precision bug) and using
the intermediary value instead of the char * buffer.
Note that there can be loss of precision for both signed
and unsigned 64 bit integers converted to double and back,
so the check must stay there (even for compatibility 
reasons).
Based on the excellent analysis in bug 28400.


libmysql/libmysql.c:
  Bug #27383: avoid pointer aliasing problems while 
  not re-violating the Intel FPU gcc bug.

Problem: long and long long types mess in a comparison may lead to wrong results on some platforms.
Fix: prefer [unsigned] long long as [u]longlong as it's used unconditionally in many places.


include/my_global.h:
  Fix for bug #29079: Semantics of "bigint" depend on platform specifics (size, signedness of char ?)
    - use [unsigned] long long as [u]longlong if sizeof(long long) == 8, to avoid type mess,
      as we use [unsigned] long long unconditionally in many places, for example in constants 
      with [U]LL suffix.