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.

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.

The field dict_table_t::big_rows was only used for determining if
the adaptive hash index should be used when the internal InnoDB SQL
parser is used. That parser is only used for modifying the InnoDB
data dictionary, updating persistent tables, and for fulltext indexes.

btr_cur_pessimistic_delete(): Discard a possible record lock also in
the case when the record was the only one in the page. Failure to
do this would corrupt the record lock data structures in a partial
rollback (ROLLBACK TO SAVEPOINT or rolling back a row operation due
to some error, such as a duplicate key in a unique secondary index).

MySQL 5.7 allows the creation of InnoDB tables in a user-created
tablespace, but MariaDB does not. Remove the remaining references.

This is a non-functional change.

- Implementing stricter data type control for Item_long_func descendants
- Cleanup: renaming Type_handler::can_return_str_ascii() to can_return_text()
  (a better name).

warning

Only microsoft compiler, which can be correctly
tested with #ifdef _MSC_VER) has __declspec(noreturn)

Prior to this patch, creating or even opening any innodb file in 10.2
would set a sparse flag on file. The file extension was done by setting
end of file, without writing zeros. This technique is fine, however
due to sparsedness, it created a hole at the end of the file, which
lead to much higher fragmentation subsequently.

The fix is only to use sparse flag for compressed tables, where holes
are actually wanted, but not for normal tables.

'version' variables.

The warnings occur on Windows build, yet they are also are valid
on 32bit Unix.

Fix is to consistently use 64bit integer on all platforms.

some ulonglong variables were printed as %lu

upd_field_set_field_no(): Remove the unused parameter trx, and
remove a debug message. In debug builds, the out-of-bounds access
would still be caught by dict_index_get_nth_col().

trx_undo_update_rec_get_update(): Remove the unnecessary parameter trx.

The InnoDB native partitioning (ha_innopart) is a MySQL 5.7
code refactoring exercise that was not merged to MariaDB.

Patch contibuted by Gordon Fisher. Thank you for your contribution!

Fixes some warnings in popular headers.

- Fix win64 pointer truncation warnings
(usually coming from misusing 0x%lx and long cast in DBUG)

- Also fix printf-format warnings

Make the above mentioned warnings fatal.

- fix pthread_join on Windows to set return value.

The intent of this patch is to avoid copying arguments from
a pair "Item **args, uint arg_count" to List<Item> in
sp_head::execute_function(). If the number of a stored function parameters
is huge, such copying can affect performance.

Change:
1. Adding a new method Row_definition_list::adjust_formal_params_to_actual_params,
   which accepts a pair of "Item **, uint".
2. Modifying the code to use the new method:
   - the calls for sp_rcontext::retrieve_field_definitions() and
     Row_definition_list::adjust_formal_params_to_actual_params() have
     been moved from sp_rcontext::create() to sp_head::rcontext_create(),
     to handle different argument notations easier (Item** vs List<Item>).
   - sp_rcontext::create() now assumes that the passed Row_definition_list
     is already adjusted to the actual SP parameters, and all "TYPE OF"
     and "ROWTYPE OF" references are resolved.
3. Removing creation of List<Item> in sp_head::execute_procedure(),
   using the code with "Item**, uint" notation instead.
4. Improvement of the code for MDEV-10577:
   As a good side effect, this patch gets rid of double security context
   switch inside sp_head::execute_trigger():
   sp_rcontext is created when the context is already switched,
   so the second context switch inside sp_head::rcontext_create() was
   redundant. This is solved by adding a "bool switch_secutiry_ctx" parameter
   to rcontext_create(), so now execute_function() and execute_procedure()
   pass "true", while execute_trigger() passes "false".

Moving a few methods from sp_rcontext to different classes:

- Table_ident::resolve_table_rowtype_ref
- Qualified_column_ident::resolve_type_ref
- Row_definition_list::resolve_table_rowtype_ref
- Row_definition_list::adjust_formal_params_to_actual_params

It easier to reuse these methods this way in the future.

After review cleanup.

- The line was accidently removed by dd8474b1
- The effect of the missing test was just a few extra malloc when creating
  internal temporary tables. Nothing critical, but better got get fixed.

Update.rdiff fle

Also, introduce my_popen()/my_fgets() wrapper function for popen()/fgets()
in mysqltest to workaround a popen() bug in Windows C runtime,
mentioned in MDEV-9409

This workaround was used previously for "exec". From now on, it is also
used are used also for "perl" snippets.