into  mysql.com:/home/hf/work/thfix/my50-thfix

the problem is that client tools are compiled with UNDEF_THREADS_HACK
flag, and my thread-related additions to the mysqltest.c can't be compiled.
Easy solution is to disable these in not-embedded case completely.

disable filling of I_S tables for EXPLAIN

into  mysql.com:/home/hf/work/mysql-5.0-mrg

into  mysql.com:/home/hf/work/mysql-5.0-mrg

into  mysql.com:/home/hf/work/mysql-5.0-0mrg

into  mysql.com:/home/hf/work/mysql-5.0-mrg

into  mysql.com:/home/hf/work/mysql-4.1-mrg

into  mysql.com:/home/bk/MERGE/mysql-5.0-merge

into  mysql.com:/home/bk/MERGE/mysql-5.0-merge

into  mysql.com:/home/bk/MERGE/mysql-5.0-merge

into  mysql.com:/home/bk/MERGE/mysql-4.1-merge

into  dsl-hkibras-fe30f900-107.dhcp.inet.fi:/home/elkin/MySQL/TEAM/BARE/5.0

mending windows test.

The regression is caused by the fix for bug 14767. When INSERT ... SELECT
used a view in the SELECT list that was not inlined, and there was an 
active transaction, the server could crash in Query_cache::invalidate.

On INSERT ... SELECT only the table being inserted into is invalidated.
Thus views that can't be inlined are skipped from invalidation.

The bug manifests itself in two ways so there is 2 test cases.
One checks that the only the table being inserted into is invalidated.
And the second one checks that there is no crash on INSERT ... SELECT.

into  kahlann.erinye.com:/home/df/mysql/build/mtr-5.0

into  kahlann.erinye.com:/home/df/mysql/build/mtr-5.0

into  kahlann.erinye.com:/home/df/mysql/build/mtr-5.0

into  kahlann.erinye.com:/home/df/mysql/build/mtr-5.0

into  trift2.:/MySQL/M50/mysql-5.0

ver 5.0 and 5.1 refinement.
adding to the MAX_SIZE_LOG_EVENT_STATUS estimation status vars and 
EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN

into  mysql.com:/home/bk/MERGE/mysql-5.0-merge

into  rakia.gmz:/home/kgeorge/mysql/autopush/B19216-4.1-opt

into  mysql.com:/home/bk/MERGE/mysql-5.0-merge

into  mysql.com:/home/bk/MERGE/mysql-4.1-merge

into  macbook.gmz:/Users/kgeorge/mysql/work/B19216-5.0-opt

 The server sends a number of columns to the client.
 It uses a limited "fast" function for that instead of the
 general one. This fast function cannot send numbers larger 
 than 2 bytes. 
 This causes the client to expect smaller number of columns. 
 The client writes outside of the allocated memory buffer 
 as a result.
 Fixed the server to use the general function to send column
 count.
 Fixed the client to check the column count before writing
 column data. 

comments are fixed as was suggested in reviews.

bugs/wls in cset's comments. The targetted BUG's or WL's identifier must be specified
the first in the comments. The referred bugs wls can be typed in same as the targeted
clickable format.
If the the target of the fix is a WL and there are some references to bugs
then the first found reference is regarded as "co-target" so that the bug's identifier
comes up on the subject line along with the WL's and commit mail will update
the bug page. The latter feature can be disarmed (if there is no need to update
the referred bug's page) with typing the first a pseudo-bug pattern (bug#0).

This paticular cset will generate subject line containing bug#0 (as it was the first
referred) whereas the old version would put in the subject line the last referred
pattern (e.g bug#2147483648).