functions.

There is no reason to export IO this functionality from server, if it can
be easily implemented with CRT.

Note: for the sake of compatibility, we do not use _open() on Windows,
but open file with CreateFile, with FILE_SHARE_DELETE flag,
so it can be renamed or deleted when open. Also atomic append flag is used.

Fixes some warnings.

fix --ps-protocol

move binlog-dependent part (~6%) into a separate file, to be able
to run the rest of multi_update with embedded server

(include/have_log_bin.inc disables embedded)

fil_space_extend_must_retry(): If the table is being truncated,
do not call fil_flush_low(). The operation is covered by the
truncate log. File extension during TRUNCATE only occurs
if there are many indexes on the table. With smaller innodb_page_size,
the file extension occurs already with fewer indexes on the table.

trx_t::in_rollback: Make the field available in non-debug builds.

row_undo_ins_remove_clust_rec(): On the rollback of an insert into
SYS_COLUMNS, revert instant ADD COLUMN in the cache by removing the
last column from the table and the clustered index.

page_set_instant(): Remove a too strict assertion. A rollback would
not reset the root page from FIL_PAGE_TYPE_INSTANT to FIL_PAGE_INDEX,
so the PAGE_INSTANT field could already be set.

rec_init_offsets_comp_ordinary(), rec_copy_prefix_to_buf(),
row_undo_search_clust_to_pcur(): Adjust too strict assertions.

row_search_on_row_ref(), row_undo_mod_parse_undo_rec(), row_undo_mod(),
trx_undo_update_rec_get_update(): Handle the 'default row'
as a special case.

btr_cur_trim(): In an update, trim the tuple as needed. For the
'default row', handle rollback specially. For user records, omit
fields that match the 'default row'.

btr_cur_optimistic_delete_func(), btr_cur_pessimistic_delete():
Skip locking and adaptive hash index for the 'default row'.

A dictionary transaction must be flagged as such before it has generated
any undo log records. This is because trx_undo_assign_undo() will mark
the transaction as a dictionary transaction in the undo log header
right before the very first undo log record is being written.

This is error-prone and causes warnings on Windows

wrep_sst_common and wsrep_sst_xtrabackup-v2 bugfixes to read sst parameters from configs

Fixing the data type for the "fuzzydate" parameter to
Item_func_hybrid_field_type::date_op() from uint to ulonglong,
for consistency with Item::get_date().

The function dict_disable_redo_if_temporary() was supposed to
disable redo logging for temporary tables. It was invoked
unnecessarily for two read-only operations:
row_undo_search_clust_to_pcur() and
dict_stats_update_transient_for_index().

When a table is not temporary and not in the system tablespace,
the tablespace should be flagged for MLOG_FILE_NAME logging.
We do not need this overhead for temporary tables. Therefore,
either mtr_t::set_log_mode() or mtr_t::set_named_space() should
be invoked.

dict_table_t::is_temporary(): Determine if a table is temporary.

dict_table_is_temporary(): Redefined as a macro wrapper for
dict_table_t::is_temporary().

dict_disable_redo_if_temporary(): Remove.

Fix truncation warning on Windows, in a populart header, with a cast.

Changing datatypes for:
- Item_spvar_args::m_table
- sp_rcontext::m_var_table
- return value of create_virtual_tmp_table()
from TABLE* to Virtual_tmp_table*

Advantages:
- Stricter data type control
- Removing the duplicate code (a loop with free_blobs)
  from destructors ~sp_rcontext() and ~Item_spvar_args(),
  using "delete m_(var_)table" in both instead.
- Using Virtual_tmp_table::delete makes the code call Field::delete,
  which calls TRASH() for the freed fields,
  which is good for valgrind test runs.

dict_col_t::name(): Also work with virtual columns

dict_v_idx_t: Add a constructor

dict_index_t::commit_instant_copy(): Work with virtual columns

ha_innobase_inplace_ctx::commit_instant(): Properly adjust base_col[]
and clear v_indexes, to be populated by dict_index_t::commit_instant_copy().

We must split the creation of table or index metadata from the
creation of the dictionary table records and the creation of
the data. In this way, we can transform a table-rebuilding operation
into an instant ADD COLUMN operation. Dictionary objects will only
be added to cache when table rebuilding or index creation is needed.
The ctx->instant_table will never be added to cache.

innodb_update_n_cols(): Renamed from innobase_update_n_virtual().
Call my_error() if an error occurs.

innobase_add_instant_try(): Call my_error() if an error occurs.

create_index_dict(): Replaces row_merge_create_index_graph().

prepare_inplace_alter_table_dict(): First create the ctx->new_table,
and only then decide if the table really needs to be rebuilt.

dict_create_sys_tables_tuple(), dict_create_table_step(): Omit the
system columns (which will now exist in the dict_table_t object
already at this point).

dict_create_table_step(): Expect the callers to invoke
dict_table_add_system_columns().

pars_create_table(): Before creating the table creation execution graph,
invoke dict_table_add_system_columns().

row_create_table_for_mysql(): Expect all callers to invoke
dict_table_add_system_columns().

The method handlerton::replace_native_transaction_in_thd was added
into MySQL 5.7 in WL#6860, Binlogging XA-prepared transaction.
In MariaDB we do not have that change yet.

dict_table_t::add_to_cache(): Modified from dict_table_add_to_cache().
Let the callers invoke dict_table_add_system_columns() and if needed,
set can_be_evicted.