This should help with the  PURGE STALE SESSIONS problem.


ndb/src/common/transporter/TransporterRegistry.cpp:
  No longer need to do funny connecting and port reporting in start_service as we
  convert the configuration mgm after we have bound to the ports we have to report.
  
  use a BaseString for constructing the connect string.
ndb/src/kernel/main.cpp:
  Reuse the mgm handle used to fetch configuration as a transporter connection.
ndb/src/kernel/vm/Configuration.cpp:
  m_mgmd_host is now a BaseString.
ndb/src/kernel/vm/Configuration.hpp:
  use BaseString for m_mgmd_host

into mysql.com:/home/stewart/Documents/MySQL/5.0/ndb-dynamic-ports

ndb/include/mgmapi/mgmapi.h:
  ndb_mgm_convert_to_transporter may destroy the handle, now takes a pointer.
ndb/include/mgmcommon/ConfigRetriever.hpp:
  If outside code is going to manipulate the NdbMgmHandle, allow it to do it
  via a get_mgmHandlePtr() call
ndb/include/transporter/TransporterRegistry.hpp:
  connect_client and connect_ndb_mgmd may destroy the handle, now they take a pointer.
ndb/src/common/transporter/TransporterRegistry.cpp:
  When start_service is binding to ports, report back the port numbers.
  
  We need this here now, as we re-use the initial mgm connection as a transporter, which
  is connected *before* start_service has allocated the dynamic port numbers. So the creation
  of this early transporter cannot be used to send the dynamic ports to the mgmd.
  
  We connect to the mgm server (using the handle that will be used in the client_Connect thread)
  if needed. This is thread safe as start_service is only ever called once, before
  the client connect thread starts.
  
  connect_client,connect_ndb_mgmd may destroy the NdbMgmHandle. It now accepts a pointer to it.
ndb/src/kernel/vm/Configuration.cpp:
  Copy the m_mgmd_host string from the config_retreiver as the NdbMgmHandle in the
  ConfigRetreiver will be destroyed later (along with the host string).
  
  globalTransporterRegistry.connect_client will destroy the mgm handle, use a pointer to the handle.
ndb/src/kernel/vm/Configuration.hpp:
  allow the dynamic allocation of m_mgmd_host.
ndb/src/mgmapi/mgmapi.cpp:
  accept a pointer for ndb_mgm_convert_to_transporter as we destroy the handle.

BitKeeper/etc/logging_ok:
  auto-union
include/mysql_com.h:
  Auto merged
ndb/src/mgmsrv/MgmtSrvr.cpp:
  Auto merged
sql/handler.h:
  Auto merged
sql/item.cc:
  Auto merged
sql/item_func.cc:
  Auto merged
sql/log.cc:
  Auto merged
sql/set_var.cc:
  Auto merged
sql/sql_class.h:
  Auto merged
sql/sql_parse.cc:
  Auto merged
sql/sql_select.cc:
  Auto merged

into mysql.com:/home/jonas/src/mysql-5.0


ndb/src/mgmsrv/MgmtSrvr.cpp:
  Auto merged

into mysql.com:/home/jonas/src/mysql-5.0

into mysql.com:/home/jonas/src/mysql-4.1

mysys.dsp:
  Added my_mmap


VC++Files/mysys/mysys.dsp:
  Added my_mmap

sql_map.cc:
  moved include of sys/stat outside of HAVE_SYS_MMAN_H define
my_decimal.h:
  Added define for EMBEDDED_SERVER for call to string2my_decimal
mi_packrec.c:
  removed cast to caddr_t


myisam/mi_packrec.c:
  removed cast to caddr_t
sql/my_decimal.h:
  Added define for EMBEDDED_SERVER for call to string2my_decimal
sql/sql_map.cc:
  moved include of sys/stat outside of HAVE_SYS_MMAN_H define

mysql-test/r/rpl_insert_id.result:
  fixed error format after mats merge
mysql-test/r/sp.result:
  fixed typo

into 51.0.168.192.in-addr.arpa:/Users/bell/mysql/bk/work-repl-5.0

into mysql.com:/home/bkroot/mysql-5.0

into 51.0.168.192.in-addr.arpa:/Users/bell/mysql/bk/work-repl-5.0


BitKeeper/etc/logging_ok:
  auto-union
sql/sql_parse.cc:
  Auto merged

logging_ok:
  Logging to logging@openlogging.org accepted
mi_packrec.c:
  Using HAVE_SYS_MMAN_H to protect include of sys/mman.h and replaced calls to mmap and munmap with my_mmap and my_munmap
mi_extra.c:
  Now using HAVE_SYS_MMAN_H to protect include of sys/mman.h
my_mmap.c:
  Changed how no mmap message is printed for Windows
sql_map.cc:
  Changed HAVE_MMAP to HAVE_SYS_MMAN_H and replaced calls to mmap and munmap with my_mmap and my_munmap


sql/sql_map.cc:
  Changed HAVE_MMAP to HAVE_SYS_MMAN_H and replaced calls to mmap and munmap with my_mmap and my_munmap
mysys/my_mmap.c:
  Changed how no mmap message is printed for Windows
myisam/mi_extra.c:
  Now using HAVE_SYS_MMAN_H to protect include of sys/mman.h
myisam/mi_packrec.c:
  Using HAVE_SYS_MMAN_H to protect include of sys/mman.h and replaced calls to mmap and munmap with my_mmap and my_munmap
BitKeeper/etc/logging_ok:
  Logging to logging@openlogging.org accepted

sql/log.cc:
  Merge from 4.1

into 51.0.168.192.in-addr.arpa:/Users/bell/mysql/bk/work-repl-5.0


BitKeeper/etc/logging_ok:
  auto-union

into 51.0.168.192.in-addr.arpa:/Users/bell/mysql/bk/work-qc_item_view-5.0


BitKeeper/etc/logging_ok:
  auto-union

into 51.0.168.192.in-addr.arpa:/Users/bell/mysql/bk/work-repl-5.0

sql/log.cc:
  Setting error code to 0 on SQL statements that cannot fail.

into serg.mylan:/usr/home/serg/Abk/mysql-5.0

mysql-test/r/binlog.result:
  binlog rotation after a big transation
mysql-test/r/rpl_rotate_logs.result:
  innodb-dependent test removed
mysql-test/t/binlog.test:
  binlog rotation after a big transation
mysql-test/t/rpl_rotate_logs.test:
  innodb-dependent test removed

into snake-hub.snake.net:/src/extern/MySQL/bk/mysql-5.0

into serg.mylan:/usr/home/serg/Abk/mysql-5.0


sql/log_event.cc:
  Auto merged

into mysql.com:/usr/local/home/marty/MySQL/test/mysql-5.0-ndb


sql/log.cc:
  Auto merged
sql/sql_class.h:
  Auto merged
sql/sql_select.cc:
  Auto merged

into mysql.com:/home/mysql_src/mysql-5.0-xa

pushed some minutes ago)



mysql-test/r/ctype_ucs.result:
  result fix
mysql-test/r/mysqlbinlog.result:
  result fix
mysql-test/r/mysqlbinlog2.result:
  result fix
mysql-test/r/rpl_charset.result:
  result fix
mysql-test/r/user_var.result:
  result fix

mysql-test/r/rpl_insert_id.result:
  Auto merged
mysql-test/t/rpl_insert_id.test:
  Auto merged
sql/log.cc:
  Merge from 4.0

into mysql.com:/home/timka/mysql/src/5.0-bug-8576

This fix addresses a bug of the HPUX C++ compiler, in the way it uses registers.
As result, the last comparison of the variable 'result' failed, and next_prefix()
looped forever.


sql/opt_range.cc:
  - fix for BUG#8576
  - updated function name in DBUG_ENTER

into mysql.com:/home/bk/b8412-mysql-4.0

into serg.mylan:/usr/home/serg/Abk/mysql-5.0


sql/log.cc:
  Auto merged
sql/sql_class.h:
  Auto merged

auto-end-rollback in case crashed binlog) which cannot hurt.


client/mysqlbinlog.cc:
  safety measures (completion_type, crashed binlog)

  cast sql_mode to ulonglong before int8store, otherwise the implementation of int8store
  based on >> complains that we do 32-bit var >> 32 which is undefined according to C standard.
  (compiler warning appeared on SGI Irix)


sql/log_event.cc:
  cast sql_mode to ulonglong before int8store, otherwise the implementation of int8store
  based on >> complains that we do 32-bit var >> 32 which is undefined according to C standard.
  (compiler warning appeared on SGI Irix)

into mysql.com:/home/mysql_src/mysql-5.0-xa


sql/lock.cc:
  Auto merged

correct end_log_pos for Xid_log_event


mysql-test/r/binlog.result:
  don't depend on the previous tests
mysql-test/r/rpl_rotate_logs.result:
  correct end_log_pos for Xid_log_event
mysql-test/t/binlog.test:
  don't depend on the previous tests
sql/handler.h:
  comment
sql/log_event.cc:
  advance position for Xid
sql/log_event.h:
  comment
sql/sql_class.h:
  correct end_log_pos for Xid_log_event
sql/sql_parse.cc:
  make sure commit handler knows whether it's autocommit or not

into 52.0.168.192.in-addr.arpa:/Users/bell/mysql/bk/work-repl-5.0


BitKeeper/etc/logging_ok:
  auto-union
sql/item.cc:
  Auto merged
sql/item_func.cc:
  Auto merged
sql/sql_acl.cc:
  Auto merged
sql/sql_parse.cc:
  Auto merged
sql/sql_select.cc:
  Auto merged
sql/protocol.cc:
  Auto merged

into ua141d10.elisa.omakaista.fi:/home/my/bk/mysql-5.0