Last part of fix for BUG#7998 "Replication should be more clever about when to replicate RELEASE_LOCK()" + fixes after merge

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

First part of fix for BUG#7998 "Replication should be more clever about when to replicate RELEASE_LOCK()"
(next part is updating test results)

into mysql.com:/M50/mysql-5.0

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

After merge fix.
The bugfix results in a different error message in this situation.

use function-wide declaration also for loop body.

into mysql.com:/home/mysql_src/mysql-4.1-clean

in slave SQL thread: if a transaction fails because of InnoDB deadlock or innodb_lock_wait_timeout exceeded,
optionally retry the transaction a certain number of times (new variable --slave_transaction_retries).

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

deadlock in MYSQL_LOG::new_file()
style fixes

Added a check, if the table, which we are going to create, is open.
This can happen if a MERGE mapped table is TRUNCATEd.

into hundin.mysql.fi:/home/jan/mysql-5.0

  Correct indentation + some small style corrections

  updated ndb test makefile to make src dist complete

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

  delete_all_rows(): Revert previous change.
  The transaction has been marked started already
  in external_lock().

  row_sel_store_mysql_rec(): Do not try to space-pad BLOB fields. (Bug #8771)

into hundin.mysql.fi:/home/heikki/mysql-5.0

  In TRUNCATE TABLE, MySQL may call innobase_xa_prepare() even though there is no active transaction because it was committed in ha_innobase::delete_all_rows(); in innobase_xa_prepare() InnoDB starts then a new transaction; let us adjust the trx->active_trans flag accordingly

into hundin.mysql.fi:/home/marko/mysql-5.0

CREATE INDEX and DROP INDEX every 10,000 rows,
similar to ALTER TABLE.

into hundin.mysql.fi:/home/jan/mysql-5.0

into hundin.mysql.fi:/home/marko/mysql-4.1

  Changed #warning pragma at the bottom to only occur when being used on non-Windows OS

remove old junk and to improve compression ratio.
InnoDB: Make implicit type conversions explicit. (Bug #8826)

with view-merge algorithm and prepared statements: in case when some
Item_ref pointing to a view column was substituted with a reference 
pointing to the view expression for that column
Item_ref::ref member of the original Item_ref was left pointing to 
not_found_item (0x1).
As we currently perform expression substition part of the view-merge 
algorithm per each execution of a prepared statement or stored procedure, 
we need to preserve original Item_ref objects usable.

into hundin.mysql.fi:/home/jan/mysql-5.0

into mysql.com:/usr/home/ram/work/4.1.b6089

into mysql.com:/usr/home/ram/work/4.1.b6089

into mysql.com:/usr/home/ram/work/4.1.b5615