Fix some more cases after merging
commit 31aef3ae.
Some warnings look possibly genuine, others are clearly bogus.

fil_page_decompress(): Remove a rather useless debug check.
We should have test coverage for reading page_compressed pages
from files, either due to buffer pool page eviction or due to
server restarts.

A similar check was removed from fil_space_encrypt() in
commit 0b36c27e (MDEV-20307).

The usage message for the innodb_compression_algorithm system variable
did not list snappy, which was added as an optional compression algorithm
in MariaDB 10.1.3 and might actually work since
commit 90c52e52 (MDEV-12615)
in MariaDB 10.1.24.

Unfortunately, we will include also unavailable compression algorithms
in the list, because ENUM parameters allow numeric values, and we do
not want innodb_compression_algorithm=3 to change meaning depending on
the way how the source code was compiled.

InnoDB only reserves 13 bits for the heap number in the record header,
limiting the heap number to be at most 8191. But, when using
innodb_page_size=64k and secondary index records of 7 bytes each,
it is possible to exceed the maximum heap number.

btr_cur_optimistic_insert(): Let the operation fail if the
maximum number of records would be exceeded.

page_mem_alloc_heap(): Move to the same compilation unit with the
only caller, and let the operation fail if the maximum heap number
has been allocated already.

The debug assertion is bogus, and we had removed it in
commit b1ab211d (MDEV-15053)
in the MariaDB Server 10.5 branch.

For a small data file, fil_space_extend_must_retry() would always
allocate a minimum size of 4*innodb_page_size.

It is possible that random read-ahead will be triggered for
a smaller file than this. In the observed case, the read-ahead
was triggered for a 6-page file that used ROW_FORMAT=COMPRESSED
with 8KiB page size. So, the desired file size was 49152 bytes,
but the actual size was 65536 bytes.

In 10.3, DBUG_ASSERT() may expand to something that includes
__builtin_expect(), which expects integer arguments, not pointers.
To avoid any compiler warnings, let us use an explicit rather than
implicit comparison to the null pointer.

innobase_pk_order_preserved(): Treat an added AUTO_INCREMENT
column in the same way as an added existing column.
In either case, the column values are not guaranteed to
be constant, and thus the ordering may change if such a column
is added before any existing PRIMARY KEY columns.

prepare_inplace_alter_table_dict(): Initialize
dict_table_t::persistent_autoinc before invoking
innobase_pk_order_preserved().

Patch 4aaa38d2 is replacing `my_config.h` with
`my_global.h` which is included in in-source build, but not for
out-of-source build tree.

Closes #1466

fil_system_t::keyrotate_next(): If space && space->is_in_rotation_list
does not hold, iterate from the start of the list.

In debug builds, we would typically have hit SIGSEGV because the
iterator would have wrapped a null pointer. It might also be that
we are dereferencing a stale pointer.

There is no test case, because the encryption is very nondeterministic
in nature, due to the use of background threads.

This scenario can be hit by setting the following:

SET GLOBAL innodb_encryption_threads=5;
SET GLOBAL innodb_encryption_rotate_key_age=0;

For some reason, GCC emits more -Wmaybe-uninitialized warnings
when using the flag -Og than when using -O2. Many of the warnings
look genuine.

The problem is caused by the fact that adding the
--defaults-group-suffix option to fix MDEV-18863 causes
mysqld to read all options from the appropriate sections
of the config file, including options specific to mysqld_multi.
Reading unknown options (which are not supported by mysqld)
causes mysqld to terminate with an error.

However, the MDEV-18863 problem has been completely fixed
by passing options on the command line, and now there is no
need to specify the --defaults-group-suffix option (we just
need to give priority to options passed through the command
line, so as not to break MDEV-18863).

During SST we need to let FTWRL to use normal timeout method
even when client is disconnected.

Some tests relied on the fact that DATETIME->DATE conversion
always produce a truncation (with a warning). This is not the case
when the SQL statement is executed at current time '00:00:00' sharp.

Adding a new SET TIMESTAMP statements to make sure time is not '00:00:00'.

Implement a workaround to shut the "memory not freed" message.

The test encryption.create_or_replace would occasionally fail,
because some fil_space_t::n_pending_ops would never be decremented.

fil_crypt_find_space_to_rotate(): If rotate_thread_t::should_shutdown()
holds due to innodb_encryption_threads having been reduced, do
release the reference.

fil_space_remove_from_keyrotation(), fil_space_next(): Declare the
functions static, simplify a little, and define in the same compilation
unit with the only caller, fil_crypt_find_space_to_rotate().

fil_crypt_key_mutex: Remove (unused).

Example of the failure:
http://buildbot.askmonty.org/buildbot/builders/bld-p9-rhel7/builds/4417/steps/mtr/logs/stdio
```
main.mysqld--help 'unix'                 w17 [ fail ]
        Test ended at 2020-06-20 18:51:45

CURRENT_TEST: main.mysqld--help
--- /opt/buildbot-slave/bld-p9-rhel7/build/mysql-test/main/mysqld--help.result	2020-06-20 16:06:49.903604179 +0300
+++ /opt/buildbot-slave/bld-p9-rhel7/build/mysql-test/main/mysqld--help.reject	2020-06-20 18:51:44.886766820 +0300
@@ -1797,10 +1797,10 @@
 sync-relay-log-info 10000
 sysdate-is-now FALSE
 system-versioning-alter-history ERROR
-table-cache 421
+table-cache 2000
 table-definition-cache 400
-table-open-cache 421
-table-open-cache-instances 1
+table-open-cache 2000
+table-open-cache-instances 8
 tc-heuristic-recover OFF
 tcp-keepalive-interval 0
 tcp-keepalive-probes 0

mysqltest: Result length mismatch
```
mtr: table_open_cache_basic autosized:

Lets assume that >400 are available and that
we can set the result back to the start value.

All of these system variables are autosized and can
generate MTR output differences.

Closes #1527

hang.

MDEV-22626: mysql_tzinfo_to_sql not replicates timezone to galeranodes if only 1 timezone will be loaded.

Move alter to InnoDB earlier to more correct place to handle
also if only a one timezone file is loaded.

fix_fields for the arguments of the NTH_VALUE function was updating the same reference,
so for the second argument (or after the first argument) the items were not resolved
to their corresponding field from the view as they were updating the reference to the
first argument.

depending on build config the error might be hidded,
in particular liblz4.so and libjemalloc.so make it to disappear,
but with -DWITH_INNODB_LZ4=NO -DWITH_JEMALLOC=NO it reappears.

Removing the ORDER BY clause from the UNION when UNION is inside an IN/ALL/ANY/EXISTS subquery.
The rewrites are done for subqueries but this rewrite is not done for the fake_select of
the UNION.

MDEV-17066: Bytes lost or Assertion `status_var.local_memory_used == 0 after DELETE with subquery with ROLLUP

The issue here is when records are read from the temporary file
(filesort result in this case) via a cache(rr_from_cache).
The cache is initialized with init_rr_cache.
For correlated subquery the cache allocation is happening at each execution
of the subquery but the deallocation happens only once and that was
when the query execution was done.

So generally for subqueries we do two types of cleanup

1) Full cleanup: we should free all resources of the query(like temp tables).
   This is done generally when the query execution is complete or the subquery
   re-execution is not needed (case with uncorrelated subquery)

2) Partial cleanup: Minor cleanup that is required if
   the subquery needs recalculation. This is done for all the structures that
   need to be allocated for each execution (example SORT_INFO for filesort
   is allocated for each execution of the correlated subquery).

The fix here would be free the cache used by rr_from_cache in the partial
cleanup phase.

Skip leading zeros when converting a string to decimal_t.