and fixing it another way (per Monty; a simpler solution which does not increase the number
of binlog events is to always execute DROP TEMPORARY TABLE IF EXISTS on slave). A new test rpl_drop_temp.
2) fixing BUG#8436 "Multiple "stacked" SQL statements cause replication to stop" by setting
thd->query_length to the length of the query being executed, not counting the next queries
if this is a multi-query. Should also improve display of SHOW PROCESSLIST. A new test rpl_multi_query.

  Alter handling for UNION syntax
  Tests for UNION and parentheses

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

  Backport the column prefix memory corruption fix from 5.0

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

into deer.(none):/home/hf/work/mysql-4.1.clean

new static archive_inited variable, so that archive_db_end() will do something only if archive_db_init() was run before
(protection against destroying uninited mutex in the case where mysqld fails early at startup (before archive_db_init() was called)
and then calls archive_db_end() to clean up).

into deer.(none):/home/hf/work/mysql-4.1.clean

into serg.mylan:/usr/home/serg/Abk/mysql-4.1

into serg.mylan:/usr/home/serg/Abk/mysql-4.1

into rurik.mysql.com:/home/igor/dev/mysql-4.1-0

  Adjustment of the result file after the revision of the fix for bug #7520.

into rurik.mysql.com:/home/igor/dev/mysql-4.0-0

  After revision of the fix for bug #7520.
table.cc:
  Revised the fix for bug #7520.
  Made it compliant with 5.0 code where the bug does not exist.

into mysql.com:/M41/mysql-4.1

into mysql.com:/M40/mysql-4.0

into mysql.com:/home/mydev/mysql-4.1-bug8306

into rurik.mysql.com:/home/igor/dev/mysql-4.1-0

  Adjusted results of the test case for bug #7520 for 4.1.

into rurik.mysql.com:/home/igor/dev/mysql-4.0-0

into rurik.mysql.com:/home/igor/dev/mysql-4.1-0

  Added a test case for bug #7520.
table.cc:
  Fixed bug #7520.
  The bug was caused by a wrong calculation of the field max_key_length for
  a TABLE structure when there was an index on a blob field.

Set of fixes requested by Kent in IRC. Tested (except the windows changes since I am trusting Kent...). No windows compiles here folks...

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

binlog-do-db or binlog-ignore-db are in effect.   (In the future 5.1? 5.0? I think 
each statement should be verified against the filtering criteria based on the database
it *uses* and not the *current* one.  But, right now the *current* database is what 
counts according to the semantics of the manual.)

It was a thread stack overrun.
IA64 had its own stack size section already.
Enlarged its default stack size from 192K to 256K.

reset mysql->net.last_error": the solution is to clear
MYSQL->net error before performing COM_CLOSE: if the call
succeeds, the connection is usable for other statements.
More comprehensive fix is to clear MYSQL->net for all
recoverable errors at the time they happen, it will be
implemented in 5.0 as it introduces incompatibility in behavior.