MDEV-31256 fil_node_open_file() releases fil_system.mutex allowing other thread to open its file node

There is room between mutex_exit(&fil_system.mutex) and
mutex_enter(&fil_system.mutex) calls in fil_node_open_file(). During this
room another thread can open the node, and ut_ad(!node->is_open())
assertion in fil_node_open_file_low() can fail.

The fix is not to open node if it was already opened by another thread.

trx_purge_truncate_history(): Only call trx_purge_truncate_rseg_history()
if the rollback segment is safe to process. This will avoid leaking undo
log pages that are not yet ready to be processed. This fixes a regression
that was introduced in
commit 0de3be8c (MDEV-30671).

trx_sys_t::any_active_transactions(): Separately count XA PREPARE
transactions.

srv_purge_should_exit(): Terminate slow shutdown if the history size
does not change and XA PREPARE transactions exist in the system.
This will avoid a hang of the test innodb.recovery_shutdown.

Tested by: Matthias Leich

The CODEOWNERS was added almost 3 years ago but never saw any adoption.
Only one person used it (me) to mark what files I maintain and for which
I wish to review commits. No other maintainers or code paths were added,
so clean it away for clarity.

buf_read_page_low(): Remove an error message and a debug assertion
that can be triggered when using innodb_page_size=4k and
innodb_file_per_table=0. In that case, buf_read_ahead_linear()
may be invoked on page 255, which is one less than the first
page of the doublewrite buffer (256).

fil_space_t::flush_freed(): Renamed from buf_flush_freed_pages();
this is a backport of aa458506 from 10.6.
Invoke log_write_up_to() on last_freed_lsn, instead of avoiding
the operation when the log has not yet been written.
A more costly alternative would be that log_checkpoint() would invoke
this function on every affected tablespace.

buf_read_ahead_linear(): Correct some calculations that were broken
in commit b1ab211d (MDEV-15053).

Thanks to Daniel Black for providing a test case and initial debugging.

Tested by: Matthias Leich

The problem, introduced in patch for MDEV-26301:

When check_join_cache_usage() decides not to use join buffer, it must
adjust the access method accordingly. For BNL-H joins this means switching
from pseudo-"ref access"(with index=MAX_KEY) to some other access method.

Failing to do this will cause assertions down the line when code that is
not aware of BNL-H will try to initialize index use for ref access with
index=MAX_KEY.

The fix is to follow the regular code path to disable the join buffer for
the join_tab ("goto no_join_cache") instead of just returning from
check_join_cache_usage().

Do not try to decide merge/materialize for derived if it was already decided
(even if it is a view).

RocksDB (in a submodule) has to include <cstdint> to use uint64_t
but it doesn't. Until the submodule is upgraded, let's replace
problematic types with something that's available

select_insert::store_values() must reset
has_value_set bitmap before every row, just like mysql_insert() does.
because ON DUPLICATE KEY UPDATE and triggers modify it

Temporary fix to avoid the server crash.

Add Optimizer Trace printouts.

This patch optimizes the number of refills for the lateral derived table
to which a materialized derived table subject to split optimization is
is converted. This optimized number of refills is now considered as the
expected number of refills of the materialized derived table when searching
for the best possible splitting of the table.

(part 1 is in the previous commit)
to [ pass ] galera.MDEV-18832, galera.MDEV-27862

1. log_event.cc stuff should go into log_event_server.cc
2. the test's wait condition is textually different in 10.5, fixed.
3. pre-exec 'optimistic' global var value is correct for 10.5 indeed.

- Update wsrep-lib which contains fix for the assertion
- Fix error handling for appending fragment to streaming log,
  make sure tables are closed after rollback.
Signed-off-by: Julius Goryavsky <julius.goryavsky@mariadb.com>

changed tast case 2 to be deterministic
Signed-off-by: Julius Goryavsky <julius.goryavsky@mariadb.com>

Adding virtual methods to class Schema:

  make_item_func_replace()
  make_item_func_substr()
  make_item_func_trim()

This is a non-functional preparatory change for MDEV-27744.

The fix was introduced, along with re-ordering to do other macros that check test environment capabilities before master/slave is set up.

Variant #2.

When Histogram::point_selectivity() sees that the point value of interest
falls into one bucket, it tries to guess whether the bucket has many
different (unpopular) values or a few popular values. (The number of
rows is fixed, as it's a Height-balanced histogram).
The basis for this guess is the "width" of the value range the bucket
covers. Buckets covering wider value ranges are assumed to contain
values with proportionally lower frequencies.

This is just a [brave] guesswork. For a very narrow bucket, it may
produce an estimate that's larger than total #rows in the bucket
or even in the whole table.

Remove the guesswork and replace it with basic logic: return
either the per-table average selectivity of col=const, or selectivity
of one bucket, whichever is lower.