Stop event and bugs the test suite could not catch
Did some big restructuring of binlog event classes - most important
change is that now each event class has exec_event method and one does
not need to modify slave core code to add a new event. Slave code is
now much smaller and easier to read



include/my_sys.h:
  pre_code and arg in IO_CACHE
mysql-test/r/rpl_log.result:
  updated result for LOAD DATA INFILE fix
mysys/mf_iocache.c:
  pre_close routine and arg pointer for callback magic
sql/log.cc:
  changed MYSQL_LOG so that write() method is for generic
  Log_event - removed redundant code
sql/log_event.cc:
  added classes for file events
  added exec_event() method to all classes
  restructured/cleaned up event classes
sql/log_event.h:
  added classes for file events
  added exec_event() method to all classes
  restructured/cleaned up event classes
sql/mf_iocache.cc:
  pre_close/arg
sql/mysqld.cc:
  added slave-load-tmpdir and old-rpl-compat options
sql/slave.cc:
  changed exec_event() to use Log_event::exec_event()
  some routines are now needed in log_event.cc and cannot be static/inline
  general cleanup
sql/slave.h:
  some routines are now extern because they are called from log_event.cc
sql/sql_class.cc:
  added slave_net
sql/sql_class.h:
  added slave_net to THD
  MYSQL_LOG::write now handles generic Log_event
sql/sql_load.cc:
  changes for new handling of LOAD DATA INFILE replication
sql/sql_repl.cc:
  added log_loaded_block() callback for IO_CACHE
sql/sql_repl.h:
  added structure to pass args to IO_CACHE callback from mysql_load