main.derived_cond_pushdown: Move all 10.3 tests to the end,
trim trailing white space, and add an "End of 10.3 tests" marker.
Add --sorted_result to tests where the ordering is not deterministic.

main.win_percentile: Add --sorted_result to tests where the
ordering is no longer deterministic.

Remove trailing white space, and use lower case for "eval" and "let".
Add an "End of 10.2 tests" marker to ease future merges.

Part#2: take into account that join nest that we are marking as constant
might already have constant tables in it. Don't count these tables twice.

Update .result files after the previous patch

Continuation of the fix: Make condition selectivity estimate use the
right estimate, too.

Reuse the fix for MDEV-17518 here, too.

The relevant InnoDB/XtraDB fixes up to 5.6.42 had already
been applied to MariaDB in commit 30c3d6db.

Revert some changes that appeared in
the merge commit 87d852f1.

Problem affects INPLACE ALTER rename columns.

innobase_rename_column_try(): some strcmp() was replaced with my_strcasecmp(),
queries to update data dictionary was updated to not match column name case.

MDEV-17349 Assertion `!table || (!table->read_set || bitmap_is_set(table->read_set, field_index))' failed on concurrent SELECT and DELETE after RENAME from table with index on virtual column

Race condition. field->flags were copied from s->field->flags during
field->clone(), early in open_table_from_share(). But s->field->flags
were getting their PART_INDIRECT_KEY_FLAG bit much later in
TABLE::mark_columns_used_by_virtual_fields() and only once per share.

If two threads were executing the code between field->clone()
and mark_columns_used_by_virtual_fields() at the same time, only
one would get PART_INDIRECT_KEY_FLAG bits in field[].

C++ does not guarantee the order of parameter evaluation.
It was wrong to pass item->val_int() and item->null_value
at the same time to any function or constructor.
Adding a new helper class Longlong_null, and new methods
Item::to_longlong_null() and Item_func_hybrid_field_type::to_longlong_null_op(),
which make sure to properly call val_int()/int_op() and test null_value.
Reorganizing the rest of the code accordingly.

LOCK view WRITE shouldn't block FLUSH view.
So we set the view's mdl_request type to it's tables.

thd_rpl_stmt_based(): A new predicate to check if statement-based
replication is active. (This can also hold when replication is not
in use, but binlog is.)

que_thr_stop(), row_ins_duplicate_error_in_clust(),
row_ins_sec_index_entry_low(), row_ins(): On a duplicate key error,
only lock all index records when statement-based replication is in use.

Users expect window functions to produce a certain ordering of rows in
the final result set. Although the standard does not require this, we
already have the filesort result done for when we computed the window
function. If there is no ORDER BY attached to the query, just keep it
till the SELECT is completely evaluated and use that to print the
result.

Update test cases as many did not take care to guarantee a stable
result.

ha_innobase::check_if_supported_inplace_alter(): Remove a redundant
condition and defer some computations.

If INNOBASE_ALTER_VERSIONED_REBUILD is set, innobase_need_rebuild()
will necessarily hold. Therefore, it is not necessary to assign
need_rebuild at the start of the function.

set both `password` and `authentication_string` columns in `mysql`.`user`
table for now.

Suppress the "password was ignored" warning if the password is
the same as the authentication string

MDEV-17038 ALTER TABLE CHANGE COLUMN c1 c1 bigint NOT NULL - generates error if table uses SYSTEM VERSIONING

* Fine-grained inplace skipping by INNOBASE_ALTER_VERSIONED_REBUILD;
* Fixed column WITHOUT SYSTEM VERSIONING + ADD COLUMN;
* Fixed instant field change (MDEV-16330);
* Revisited test versioning.online;
* Merged the test versioning.trx_id_versioning_attribute_persistence
  to versioning.online;
* Renamed some versioning functions:
** change_fields_versioning_cache() -> vers_change_fields_cache()
** change_field_versioning_try() -> vers_change_field_try()
Skip condition moved out of func.

Closes tempesta-tech/mariadb#414
Closes tempesta-tech/mariadb#540
Related to tempesta-tech/mariadb#281

 which did not always correctly simulated io-cache::end_of_file.
The error was caused by implicit cast to unsigned of an intemediate term
in a formula.

It seems Item_cache_time::val_datetime_packed() is never caller in the
current code. Fixing it to use to_packed() rather than to_longlong() anyway.

The parameters bool sync=true, bool read_only=false of mtr_t::start()
were added in
https://github.com/mysql/mysql-server/commit/eca5b0fc17a5bd6d4833d35a0d08c8549dd3b5ec
(MySQL 5.7.3).

The parameter read_only was never used anywhere.
The parameter sync was only copied around, and would be returned
by the unused function mtr_t::is_async().

We do not need this dead code in MariaDB.

At the start of the function, we already dereferenced node,
so the checks for NULL are unnecessary and redundant.

For index merge union[or sort union], the estimates are not taken into account while calculating the selectivity of
a condition. So instead of showing the estimates of the index merge union[or sort union], it shows estimates equal to
all the records of the table.
The fix for the issue is to include the selectivity of index merge
union[or sort union] while calculating the selectivity of a condition.

branch.

Query INFORMATION_SCHEMA.INNODB_SYS_TABLES only once, and cache results.

As a small cleanup, remove  mdl_lock_con_mutex, the MDL handling
connection is never accessed by multiple threads at the same time.

  sql\sql_acl.cc(3114): error C2220: warning treated as error - no 'object' file generated
  sql\sql_acl.cc(3114): warning C4390: ';': empty controlled statement found; is this the intent?

Support SET PASSWORD for authentication plugins.

Authentication plugin API is extended with two optional methods:
* hash_password() is used to compute a password hash (or digest)
  from the plain-text password. This digest will be stored in mysql.user
  table
* preprocess_hash() is used to convert this digest into some memory
  representation that can be later used to authenticate a user.
  Build-in plugins convert the hash from hexadecimal or base64 to binary,
  to avoid doing it on every authentication attempt.

Note a change in behavior: when loading privileges (on startup or on
FLUSH PRIVILEGES) an account with an unknown plugin was loaded with a
warning (e.g. "Plugin 'foo' is not loaded"). But such an account could
not be used for authentication until the plugin is installed. Now an
account like that will not be loaded at all (with a warning, still).
Indeed, without plugin's preprocess_hash() method the server cannot know
how to load an account. Thus, if a new authentication plugin is
installed run-time, one might need FLUSH PRIVILEGES to activate all
existing accounts that were using this new plugin.

* remove dead code (from .yy)
* remove redundant commands from the test
* extract common code into a reusable function
  (get_auth_plugin, push_new_user)
* rename update_user_table->update_user_table_password
* simplify acl_update_user
* don't strdup a string that's already in a memroot
  (in ACL_ROLE::ACL_ROLE(ACL_USER*))
* create parent_grantee and role_grants dynamic arrays with size 0.
  to avoid any memory allocations when roles aren't used.

Don't distinguish between a "password hash" and "authentication string"
anymore. Now both are stored in mysql.user.authentication_string, both
are handled identically internally. A "password hash" is just how some
particular plugins interpret authentication string.

Set mysql.user.plugin even if there is no password. The server will use
mysql_native_password plugin in these cases, let's make it expicit.

Remove LEX_USER::pwhash.