Current table is placed into read_tables set of the current join_tab->select
so get_mm_parts function thinks that current table's record is read and
tries to calculate WHERE condition with the fields of the record.
Result of these calculations is unpredictable.
Looks funny - outcome of the SELECT depends on the queries executed before.

Anyway i think we should have testcase on this part of the code.

  Remove outdated comment and the corresponding assertion in debug version code

so I commit again in a fresh tree.
Fix for bug#763 (Relay_log_space too big by 4 bytes),
plus comments and DBUG_PRINT, and we don't start replication
if --bootstrap.

(Needed for platforms with the traditional bzero prototype.)
Fixes BUG#460.

  Remove all nondeterministic tests from innodb.test; but I have to study why on 64 bits we get different estimates

into serg.mylan:/usr/home/serg/Abk/mysql-4.0

into serg.mylan:/usr/home/serg/Abk/mysql-4.0

into eagle.mysql.r18.ru:/home/vva/work/BUG_672/mysql-4.0

into sinisa.nasamreza.org:/mnt/work/mysql-4.0

CREATE / INSERT ... (SELECT ...) UNION (SELECT ...) LIMIT;

  Remove nondeterministic EXPLAIN SELECT which gives on 64-bit platforms a slightly different estimate (sometimes we should study why we get different results on 64-bit and 32-bit)

  Make execution with UNIV_SYNC_DEBUG faster

  Fix a latching order violation in the previous (2003-06-15) push: could lead to a hang on the btr0sea.c semaphore

  Cleanup

  Remove still some potential hangs over the btr0sea.c semaphore

into sinisa.nasamreza.org:/mnt/work/mysql-4.0

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

  Fix a possible hang over the btr0sea.c latch

into serg.mylan:/usr/home/serg/Abk/mysql-4.0

into mashka.mysql.fi:/home/my/mysql-4.0

into eagle.mysql.r18.ru:/home/vva/work/BUG_671/mysql-4.0

Don't set field to DEFAULT value when set to NULL

into mashka.mysql.fi:/home/my/mysql-4.0

into mashka.mysql.fi:/home/my/mysql-3.23

  CREATE TABLE ...(SELECT ...) UNION ...
  INSERT ...      (SELECT ...) UNION ...

into serg.mylan:/usr/home/serg/Abk/mysql-4.0

   paths for the binary distributions (this will hopefully resolve the
   "file /usr/local/mysql/libexec/mysqld doesn't exist" error that many
   people are facing)