- When the optimizer chose LooseScan, make_join_readinfo() should
  use the index that was chosen for LooseScan, and should not try 
  to find a better (shortest) index.

The bug was in the mysqltest.c. It didn't check the send_query result
properly, so when that fails and the thread gets disconnected,
it goes unnoticed, so the consequitive read_result call crashes.
Fixed by checking the sernd_query return.

            ! pthread_equal(pth read_self(), (&(&LOCK_open)->m_mutex)->thread)'
            fails in intern_sys_var_ptr on server shutdown after uninstalling
            TokuDB plugin at runtime

This assertion was introduced by patch for MDEV-5089 to ensure proper lock order
among LOCK_open and LOCK_global_system_variables: LOCK_open must not be held
while acquiring LOCK_global_system_variables.

intern_sys_var_ptr() may be called while freeing storage engine variables with
PLUGIN_VAR_MEMALLOC flag (when destroying table share after storage engine was
uninstalled). In this case LOCK_open is held, which is harmless because we need
global value pointer and thus won't acquire LOCK_global_system_variables.

Relaxed assertion so it is valid only for session variables.

Analysis: Can't disable the error message because you may get database
started with incorrect log file size. 

Fix: Thus only improve the error message to give more information 
to users.

Analysis: By default 10.0 creates 48M log files and 5.5 assumes they
are 5M.

Fix: Remove the error and do size comparison later.

- Fixed bug that we where using wrong checksum algorithm when using VARCHAR with fixed lenth rows
- Ensure in myisampack that HA_OPTION_NULL_FIELDS is set for tables with null fields.

mysql-test/r/myisampack.result:
  Updated results
mysql-test/t/myisampack.test:
  Added more tests
storage/myisam/mi_open.c:
  Use correct checksum algorithm when we have VARCHAR fields with fixed length records
storage/myisam/myisampack.c:
  Ensure HA_OPTION_NULL_FIELDS is set for tables with null fields.
  (This was not set by default for not compressed tables without checksums to keep MyISAM tables compatible with MySQL)

added the test case

on select from I_S.INNODB_CHANGED_PAGES

Analysis: limit_lsn_range_from_condition() incorrectly parses
start_lsn and/or end_lsn conditions.

Fix from SergeyP. Added some test cases.

All underlying tables should share the same lock type.

"Table upgrade required..."

The row format is only different for the case where we have a very old MyISAM table with varchar fields, null fields and created with CHECKSUM=1
The table is usable, except that CHECKSUM TABLE will give a wrong result and CHECK TABLE will warn about this.

I added a test for this to warn when a table needs to be upgraded, but forgot to add a test that this was only relevant for tables with CHECKSUM=1
This is now fixed.



storage/myisam/ha_myisam.cc:
  Fixed wrong test.

git://github.com/Tokutek/ft-index.git
git://github.com/Tokutek/ft-engine.git
at the tag tokudb-7.1.6

Add charsets to the client API
(they are declared with MYSQL_PLUGIN_IMPORT, meaning the intention is to have them public)

It is triple bug with one test suite:
1. Incorrect outer table detection
2. Incorrect leaf table processing for multi-update (should be full like for usual updates and inserts)
3. ON condition fix_fields() fould be called for all tables of the query.

MDEV-6091 mysqldump goes in a loop and segfaults if --dump-slave is specified and it cannot connect to the server

do_start_slave_sql() is called from maybe_exit().
We should not recurse when maybe_exit() is called for an error during do_start_slave_sql().
Also remove a meaningless (but safe) "goto err".

fixes #230

because it's not a handler convenience wrapper