Do not accessing trx->mysql_log_file_name
after transaction end.

Do not use CloseHandle() on thread id. It is not a kernel handle.

Also MDEV-10385 Threadpool refactoring

* update net_field_length() to match the prototype
* various bug fixes in C/C

fix a bad merge

and INSERT ... SELECT.
Added test cases.

Problem was NULL-pointer reference. Fixed by removing unnecessary and not
really working recursive reserver thread lookup.

because it counts number of bytes sent and expects it to be independent
from the database name length

and a couple of collaterals:
* debugging assert in my_valgrind.h
* trivial cleanup in sql_union.cc

fix a race condition in the test

In some cases the function compare_order_elements() erroneously
returned CMP_EQ for not equal elements.

finally removed, after being deprecated since 2005

do not put there *.so symlinks

fixes: federated.federatedx rpl.rpl_sp rpl.rpl_colSize
rpl.rpl_extra_col_slave_innodb rpl.rpl_extra_col_master_innodb
rpl.rpl_udf rpl.rpl_extra_col_slave_myisam rpl.rpl_extra_col_master_myisam

Updated mysqld--help,win.rdiff.

Some compilers are unhappy about mixed struct/class declarations.

Fixed type cast.

Fixed sysvars_server_[not]embedded failure: changed type of
table_open_cache_instances from ulong to uint.

Added casts foratomic operations around tc_active_instances and
tc_contention_warning_reported: needed on some platforms.

Table cache instances autosizing.

Improve scalability by implementing multi-instance table cache.

because InnoDB 5.7 tdecided to store a full path there

test. We should not take X-lock for index in the begining, setting
tablespace as named space is enough.

add_identifier change comes from 5.7, everything else is a follow-up

* wait_until_connected_again.inc should expect ER_CONNECTION_KILLED now
* also fix MySQL BUG#36228