(backports from fixes made in 4.1)


myisam/myisampack.c:
   - replaced "1ULL" with "((ulonglong)1)" to resolve a compile error on
     Windows
sql/sql_handler.cc:
   - removed some unused variables
   - added a (byte*) cast to fix a compile error on Windows (backport of a
     fix made in 4.1)

BitKeeper/deleted/.del-innodb-lock-master.opt~f76a4a1999728f87:
  Delete: mysql-test/t/innodb-lock-master.opt

innobase/include/srv0srv.h:
  Add configuration parameter srv_max_purge_lag.
  Add global variable srv_dml_needed_delay.
innobase/include/trx0sys.h:
  Add trx_sys->rseg_history_len
innobase/row/row0mysql.c:
  Add row_mysql_delay_if_needed() for delaying INSERTs, UPDATEs and
  DELETEs for srv_dml_needed_delay microseconds.
innobase/srv/srv0srv.c:
  Define global variable srv_dml_needed_delay.
  Define configuration parameter srv_max_purge_lag.
innobase/trx/trx0purge.c:
  Update trx_sys->rseg_history_len.
  trx_purge(): Compute srv_dml_needed_delay from srv_max_purge_lag
  and trx_sys->rseg_history_len.
innobase/trx/trx0rseg.c:
  Initialize trx_sys->rseg_history_len at InnoDB start-up.
sql/ha_innodb.h:
  Add configuration parameter srv_max_purge_lag.
sql/mysqld.cc:
  Add startup option innodb_max_purge_lag,
  with default value 0 (meaning infinite, disabling the feature).
sql/set_var.cc:
  Add global variable innodb_max_purge_lag.

into sanja.is.com.ua:/home/bell/mysql/bk/work-4.0


sql/sql_parse.cc:
  Auto merged

sql/mysql_priv.h:
  constant definition
sql/sql_base.cc:
  difine used instead of constant
sql/sql_cache.cc:
  difine used instead of constant
  typo fixed

into sanja.is.com.ua:/home/bell/mysql/bk/work-qc-4.0

into sanja.is.com.ua:/home/bell/mysql/bk/work-4.0

crashes server"

(in 4.0 we fix only connection stalling in case of error, crash itself is
fixed in 4.1, the test case for this code is also there).


sql/sql_parse.cc:
  Handling of COM_CREATE_DB, COM_DROP_DB:
  We should call send_error() if mysql_create_db or mysql_drop_db
  return error (like we do it for SQL versions of these commands).

  Add typecast from ulint to ssize_t in pread and pwrite, so that the type is according to the Linux man page; this will probably not help to fix the HP-UX 32-bit pwrite failure, since the compiler should do the appropriate typecasts anyway


innobase/os/os0file.c:
  Add typecast from ulint to ssize_t in pread and pwrite, so that the type is according to the Linux man page; this will probably not help to fix the HP-UX 32-bit pwrite failure, since the compiler should do the appropriate typecasts anyway

  Fix bug #5961: release the dictionary latch during a long cascaded FOREIGN KEY operation, so that we do not starve other users


innobase/row/row0ins.c:
  Fix bug #5961: release the dictionary latch during a long cascaded FOREIGN KEY operation, so that we do not starve other users
innobase/row/row0mysql.c:
  Fix bug #5961: release the dictionary latch during a long cascaded FOREIGN KEY operation, so that we do not starve other users

mysql-test/r/query_cache.result:
  hiding real table stored in query cache by temporary table
mysql-test/t/query_cache.test:
  hiding real table stored in query cache by temporary table
sql/sql_cache.cc:
  Check of temporary tables hiding for query fetched from QC
sql/sql_cache.h:
  Key length now stored in table record of query cache

mysql-test/r/innodb-lock.result:
  Fix test case (old one didn't test things correctly)
mysql-test/t/innodb-lock.test:
  Fix test case (old one didn't test things correctly)
mysys/thr_lock.c:
  More debugging information
sql/mysqld.cc:
  Enable innodb_table_locks as default, as otherwise there is a possibility for deadlocks
sql/sql_base.cc:
  More debug information

into mysql.com:/home/my/mysql-4.0


libmysql/libmysql.c:
  Auto merged

Rename innodb_table_locks_old_behavior -> innodb_table_locks
Set innodb_table_locks to off by default to get same behaviour as in MySQL 4.0.20
(This means that Innodb ignore table locks by default, which makes it easier to combine MyISAM and InnoDB to simulate a transaction)


libmysql/libmysql.c:
  Use ulong instead of unsigned long
  Reuse _dig_vec()
myisam/myisampack.c:
  Simplify code
mysql-test/r/innodb-lock.result:
  new test case
mysql-test/t/innodb-lock.test:
  new test case
sql/ha_innodb.cc:
  Rename innodb_table_locks_old_behavior -> innodb_table_locks
sql/mysqld.cc:
  Rename innodb_table_locks_old_behavior -> innodb_table_locks
  Set this off by default to get same behaviour as in MySQL 4.0.20
sql/set_var.cc:
  Rename innodb_table_locks_old_behavior -> innodb_table_locks
sql/sql_class.h:
  Rename innodb_table_locks_old_behavior -> innodb_table_locks

  test


innobase/trx/trx0rec.c:
  test

  Add note to mysql_hex_string() comment.


libmysql/libmysql.c:
  Add note to mysql_hex_string() comment.

  New function mysql_hex_string()


libmysql/libmysql.c:
  New function mysql_hex_string()

into goetia.(none):/home/dellis/mysqlab/bk/mysql-4.0


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

  BUG #5731 key_buffer_size not properly restricted to 4GB; use UINT_MAX32 for clarity.


sql/mysqld.cc:
  BUG #5731 key_buffer_size not properly restricted to 4GB; use UINT_MAX32 for clarity.

into mysql.com:/home/hartmut/projects/mysql/dev/4.0

  Change parsing of @image argument.


Docs/Support/texi2html:
  Change parsing of @image argument.

so lets get rid of it (Bug #6112)

into hundin.mysql.fi:/home/marko/k/mysql-4.0

into mysql.com:/home/mydev/mysql-4.0-bug5625


sql/ha_myisam.cc:
  Auto merged
sql/sql_table.cc:
  Auto merged

Added a try to a normal repair() if repair_by_sort() failed.
This was not done with ENABLE KEYS and OPTIMIZE TABLE.
Fixed error code handling in mysql_alter_table().


sql/ha_myisam.cc:
  BUG#5625 - MyISAM Index corruption on ALTER TABLE x ENABLE KEYS due to full tmpdir.
  Added a try to a normal repair() if repair_by_sort() failed.
  This was not done with ENABLE KEYS and OPTIMIZE TABLE.
sql/sql_table.cc:
  BUG#5625 - MyISAM Index corruption on ALTER TABLE x ENABLE KEYS due to full tmpdir.
  Added a translation from 'bool' return value to '< 0' error indication,
  which is used within mysql_execute_command() and must as such be
  handed over by mysql_alter_table(). A returncode >= 0 is interpreted
  as 'I have already called send_ok()'.

innodb_table_locks_old_behavior: do not acquire an
InnoDB table lock for LOCK TABLES, as in mysql-4.0.18
and earlier.


sql/ha_innodb.cc:
  Do not acquire an InnoDB table lock for LOCK TABLES if
  innodb_table_locks_old_behavior is enabled.
sql/mysqld.cc:
  Added innodb_table_locks_old_behavior
sql/set_var.cc:
  Added innodb_table_locks_old_behavior
sql/sql_class.h:
  Added innodb_table_locks_old_behavior

  Change error code to HA_ERR_ROW_IS_REFERENCED if we cannot DROP a parent table referenced by a FOREIGN KEY constraint; this error number is less misleading than the previous value HA_ERR_CANNOT_ADD_FOREIGN, but misleading still; we should introduce to 5.0 a proper MySQL error code


sql/ha_innodb.cc:
  Change error code to HA_ERR_ROW_IS_REFERENCED if we cannot DROP a parent table referenced by a FOREIGN KEY constraint; this error number is less misleading than the previous value HA_ERR_CANNOT_ADD_FOREIGN, but misleading still; we should introduce to 5.0 a proper MySQL error code

into hundin.mysql.fi:/home/heikki/mysql-4.0

into mysql.com:/mnt/tmp/mysql-4.0-bug-myisampack

into mysql.com:/home/mysql_src/mysql-4.0

when one connection had done FLUSH TABLES WITH READ LOCK, some updates, and then COMMIT,
it was accepted but my_error() was called and so, while client got no error, error was logged in binlog.
We now don't call my_error() in this case; we assume the connection know what it does.
This problem was specific to 4.0.21. The change is needed to make replication work with existing versions of innobackup.


sql/lock.cc:
  If a connection has done FLUSH TABLES WITH READ LOCK and now is doing COMMIT, don't give error
  (applies only if it's the same connection; others' COMMITs are still blocked).

  Fix bug #5960: if one updated a column so that its size changed, or updated it to an externally stored (TEXT or BLOB) value, then ANOTHER externally stored column would show up as 512 bytes of good data + 20 bytes of garbage in a consistent read that fetched the old version of the row


innobase/trx/trx0rec.c:
  Fix bug #5960: if one updated a column so that its size changed, or updated it to an externally stored (TEXT or BLOB) value, then ANOTHER externally stored column would show up as 512 bytes of good data + 20 bytes of garbage in a consistent read that fetched the old version of the row

(Fixed this by backporting some logic from 4.1)

BitKeeper/etc/logging_ok:
  Logging to logging@openlogging.org accepted

in the error monitor thread.  (Bug #5898)


innobase/include/sync0arr.h:
  sync_array_print_long_waits(): return error status
innobase/srv/srv0srv.c:
  srv_error_monitor_thread(): Keep track on successive fatal timeouts,
  and crash the server only if the timeouts have been exceeded for
  several times in succession.
innobase/sync/sync0arr.c:
  sync_array_print_long_waits(): return error status

into build.mysql.com:/users/rburnett/mysql-4.0

myisam/mi_locking.c:
  More comments
sql/mysql_priv.h:
  Change mode to uint (as it's a bitmap)
sql/sql_handler.cc:
  Change mode to uint (as it's a bitmap)
  Fixed DBUG_PRINT to use same format as other MySQL code

into gw.mysql.r18.ru:/usr/home/ram/work/4.0.b5733


sql/records.cc:
  Auto merged

into mysql.com:/home/my/mysql-4.0


sql/sql_table.cc:
  Auto merged