This is an addition to original patch. Some platforms are strict about atomic
op argument signedness.

Spider support for new xid cache implementation.

XID cache is now based on lock-free hash.
Also fixed lf_hash_destroy() to call alloc destructor.

Note that previous implementation had race condition when thread was accessing
XA owned by different thread. This new implementation doesn't fix it either.

Replaced all references to LINT_INIT with UNINIT_VAR and LINT_INIT_STRUCT.
Removed LINT_INIT macro.

to my_convert_fix() - a new function in /strings.

Checked for missing synonims, added those found.

Problem was that we considered the point itself as the 'border' object. Instead
of that the 'border' of a POINT is an empty set, and the point is the 'interior'.
Another error fixed by the way - not all operations of the resulting function were properly
allocated.

Need to take into account cases of a polygon shaped as a very thin line.

The variables now return real values. They can't be made usual variables because they do not have DEFAULT value.

Problem was in reinitialization of first_successful_insert_id_in_prev_stmt_for_binlog after setting variables.

The problem is in the IsSimple function. If the first and the last points
of a curve coincide it's and exception and the line is still 'simple'.

Index names were not formatted correctly.

The 'srid' field's copying was missing in the copying Create_field::Create_field() constructor.

Merge from Percona Server enforced use of a specific storage engine
authored by Stewart Smith.

Modified to be session variable and modifiable only by SUPER. Use
similar implementation as default_storage_engine.

- Renaming Item::is_bool_func() to is_bool_type(), to avoid assumption
  that the item is an Item_func derivant.
- Deriving Item_func_spatial_rel from Item_bool_func rather than Item_int_func

fixed embedded server tests
MDEV-7009: SET STATEMENT min_examined_row_limit has no effect
MDEV-6948:SET STATEMENT gtid_domain_id = ... FOR has no effect (same for gtid_seq_no and server_id)

old values of SET STATENENT variables now saved in its own Query_arena and restored later

the problem was in assigning default value during parsing.

(changes of backported patch are very small: strlen removed, error processing fixed, view open statistics added)

Double error reporting removed.

* watch for "redefined macro" warning
* don't enable -D_FORTIFY_SOURCE=2 in debug builds (it requires -O)

common parent class for Item_func_isnull and Item_func_isnotnull

Add holder thread_id, file_name and line to output.

MDEV-7025 ANALYZE SELECT/INSERT/UPDATE/DELETE from a view does not check access permissions on the view

Added access checking for the ANALYZE statement command.

Fix Item_ident::print() to work when there is no current database

Change r_rows to be double