instead of making it experimental.

The value was changed from 1024 to 2048,
because 1024 is already used.

Original revision:
------------------------------------------------------------
revno: 2476.657.259
committer: tsmith@bk-internal.mysql.com
timestamp: Tue 2008-01-29 19:32:44 +0100
message:
  WL#4098: Add a status flag for slow queries
  
  Add new result set status flag SERVER_QUERY_WAS_SLOW in mysql_com.h.
  
  There is no server-side implementation to actually set this flag at this
  time.  This patch simply defines the interface so that applications can
  make use of it.  See the worklog for background.
------------------------------------------------------------

```---------------------------------------------------------
revno: 2476.714.9
committer: markleith@medusa.mysql.com
timestamp: Mon 2007-09-03 23:18:50 +0100
message:
  Bug#25822 / WL#3953 - Log Access Denied errors for bad passwords in the error log
  
  Added logging of Access Denied messages on failed new connections to the 
  error log when log-warnings = 2.
  
  This allows a user to log these messages without having to incur the overhead
  of using the general query log on production systems.
```

---------------------------------------------------------

Remove the extra line which try to use the name INNOBASE again. The IF ELSE
loop above declares the library name based on the version of mysql

push_warning(MYSQL_ERROR::WARN_LEVEL_ERROR)).

The Signal/Resignal patch changes the push_warning() API: now
it silently downgrades WARN_LEVEL_ERROR to WARN_LEVEL_WARN.

This patch should be rolled back when Bug#47233 is fixed.

During insert, we are not reading the rows in a referring table but
instead using the last read row that happens to be in table->record[0].

Now INSERT into such view is denied. 

BUG#41759: "Valgrind warning: Invalid read of size 1 open_tables() with ps-protocol"
(from revid:sergefp@mysql.com-20081227013233-f9n2by8romt37hyi)

BUG#38049 "incorrect rows estimations with references from preceding table"
(from revid:sergefp@mysql.com-20090126194259-ue20il3qro529l4d).
Compared to 6.0 where EXPLAIN indicates "Using index condition", here in join_optimizer.result
we see "Using where"; it's normal; 6.0 shows the same if disabling Index Condition Pushdown.

backport to next-mr

s/return/DBUG_RETURN/

EXPLAIN EXTENDED warning.

Query optimizer searches for the constant tables and optimizes them away. This
means that fields of such tables are substituted for their values and on later
phases they are treated as constants. After this constant tables are removed
from the query execution plan. Nevertheless constant tables were shown in 
the EXPLAIN EXTENDED warning thus producing query that might be not an
equivalent of the original query.
        
Now the print_join function skips all tables that were optimized away from
printing to the EXPLAIN EXTENDED warning. If all tables were optimized away it
produces the 'FROM dual' clause.

use function prototypes from header in conjunction with K&R-style
function definitions.  The symptom is a type mismatch between
int and wchar_t.