Remove use of prebuilt->trx, better play safe since we do not know MySQL call order of external_lock

  Small improvements
row0mysql.c:
  Small improvements + fix the ALTER TABLE problem by introducing a lazy drop table it can use
ha_innobase.cc:
  Some fine-tuning of optimization

  Fix the crashes reported by Peter; SHOW CREATE does not call external_lock before calling ::info, and consequently the trx struct in prebuilt is obsolete; setting trx->op_info caused erratic memory corruption in mysqldump

into donna.mysql.fi:/home/heikki/mysqln

  Fine-tune optimization parameters; in small tables the estimates are accurate; in big tables let us not be too optimistic about index selectivity

  Assume selectivity of any key prefix is better than 1 / 100, regardless of what the estimator returns; this balances the pessimistic assumption of the optimizer thatany secondary key access to the full row incurs a page read from disk

  Add more diagnostic code to cath possible memory corruption

  Add an assert
dyn0dyn.h:
  Reduce stack consumption

  Merge InnoDB-.48

  Tell table handlers where a binlog segment for a trx ends

  Inc the version

  Added missing 'Q' for -Q option

  Better error message if we hit the max process size in FreeBSD

  Add better error message if access rights are wrong

  Fix typo

  Improve error messages which may occur in installation

  Remove erroneous warning of a duplicate key when the key in a UNIQUE secondary index contains a NULL

into donna.mysql.fi:/home/heikki/mysqlm

  Make InnoDB startup option samples easier to understand, recommend not setting log files too big, because recovery slow

  Eliminate a gcc compiler bug in inlining

  Make os_file_flush to ignore the error we get from a raw device in fsync

  Changed default of file_io_threads to 4 and lock_wait_timeout to 50 seconds
  Added a missing break: if not specified, innodb_fast_shutdown got the same value as ..flush_log_at_trx commit

  Print progress info of the applying log records to the database phase in recovery
srv0srv.c:
  Do buffer pool flush and checkpoints more often to make recovery faster

  Increase table comment print size to 500 bytes to be able to print more foreign key constraint infos
ha_innobase.h:
  Change max key len back to 500 bytes because MySQL interpreter cannot handle more

  Allow foreign keys refer between fixed and var length strings

  Lower max key len to 3500 bytes because a secondary index record must also contain the primary key value

  Make sure no memory overrun of key buffer in range estimation