(The table already existed in the storage engine)

Wrong algorithm of closing partitions on error doesn't close last
partition.

when checking for free port, use the same logic (IPv6 socket address
/ dual socket), like the server would.

Previous solution for testing whether port is free was trying to bind
IPv4 socket on INADDR_ANY.

This not work now on some reason, that attempt succeeds, even if there is
an existing IPv6-dual socket listening on 0.0.0.0:3306

It works, but irritates people who look into the log and see
traces of 32bit custom action server.

In the merge 9e6e4355 we replaced
direct use of std::atomic with a wrapper class, so that
dict_index_t::lock will support the default assignment operator.

As part of that change, one occurrence of std::memory_order_release
was accidentally replaced with std::memory_order_relaxed.

Thanks to Sergey Vojtovich for noticing this.

Respect system fields in NO_ZERO_DATE mode.

This is the subject for refactoring in MDEV-19597

Conflict resolution from 7d5223310789f967106d86ce193ef31b315ecff0

Respect system fields in NO_ZERO_DATE mode.

This is the subject for refactoring in MDEV-19597

UPDATE gets access to history records because versioning conditions
are not set for VIEW. This leads to endless loop of inserting history
records when clustered index is rebuilt and ha_rnd_next() returns
newly inserted history record.

Return back original behavior of failing on write-locked table in
historical query.

35b679b9 assumed that SELECT_LEX::lock_type influences anything, but
actually at this point table is already locked. Original bug report
was tempesta-tech/mariadb#102

Bogus if() logic inside the func.

System versioning assertion fix. Since DROP SYSTEM VERSIONING does not
change list of dropped keys we should handle a special case.

Caused by MDEV-19751. This fix deprecates MDEV-17091.

otherwise it reaches "0 tests left" state and then waits for
a few minutes  for all workers to complete their tests.

show failures. account for retries.

don't fetch the complete history of all submodules, it's rarely needed

Revert "MDEV-12445 : Rocksdb does not shutdown worker threads and aborts in memleak check on server shutdown"

This reverts commit 6f1f9114.

because it doesn't do anything now (the server doesn't check
my_disable_leak_check) and it never did anything before
(because without `extern` it simply created a local instance of
my_disable_leak_check, did not affect server's my_disable_leak_check).

WITH_ROCKSDB_ZSTD must use the same capitalization as
in the foreach loop

MDEV-21958 Query having many NOT-IN clauses running forever and causing available free memory to use completely

let thd->killed to abort range optimizer

cannot use the current THD::mem_root, because it can be temporarily
reassigned to something with a very different life time
(e.g. to TABLE::mem_root or range optimizer mem_root).

the existing ".git/" rule only filters out .git directories,
but as we have submodules now, we have to filter out .git files too.

data_file_length == 0 in mi_repair() is normal for REPAIR ... USE_FRM.
But in-file links (for blocks and deleted chain) must be compared with
the real file length to avoid spurious "link points outside datafile"
warnings and arbitrary block skipping.

The build was broken on CentOS 5 and CentOS 6 as a result of
commit 18d8f06f introducing
some constructs that old GCC versions could not cope with.

The build was broken in the parent
commit 18d8f06f.

- Forget to record .result for instant_alter_limit 16k test case

namespace intrusive: removed

split class into two: ilist<T> and sized_ilist<T> which has a size field.

ilist<T> no more NULLify pointers to bring a slignly better performance.
As a consequence, fil_space_t::is_in_unflushed_spaces and
fil_space_t::is_in_rotation_list boolean members are needed now.