FIXME: rocksdb.drop_table hangs on shutdown, but is not disabled in 10.2

row_upd_build_difference_binary(): Correctly handle the
case where columns (or clustered index fields) have been added
since the 'entry' was originally created. In this case,
the update vector must replace any missing columns with the
default values of the instantly added columns.

Merge the changes to include/index_merge*inc from the upstream. The changes
add this command in many places:

 +if ($engine_type == RocksDB)
 +{
 +    set global rocksdb_force_flush_memtable_now=1;
 +}

also add it in one more place to make the test truly stable.

With --skip-debug-assert, DBUG_ASSERT(false) will allow execution to
continue. Hence, we will need /* fall through */ after them.

Some DBUG_ASSERT(0) were replaced by break; when the switch () statement
was followed by DBUG_ASSERT(0).

Adjust the testcase according to the review input

(Backported to 10.3, addressed review input)

Sj_materialization_picker::check_qep(): fix error in cost/fanout
calculations:
- for each join prefix, add #prefix_rows / TIME_FOR_COMPARE to the cost,
  like best_extension_by_limited_search does
- Remove the fanout produced by the subquery tables.
- Also take into account join condition selectivity

optimize_wo_join_buffering() (used by LooseScan and FirstMatch)
- also add #prefix_rows / TIME_FOR_COMPARE to the cost of each prefix.
- Also take into account join condition selectivity

The test occasionally fails with different table reference count
due to purge activity after INSERT operations (MDEV-12288).
Wait for purge before accessing dict_table_t::n_ref_count.

Merge a part of commit 25af2a18.
This was forgotten in the
merge commit e41eb044.

Commit 536215e3 in MariaDB Server 10.3.1
introduced the compiler flag (not cmake option) DBUG_ASSERT_AS_PRINTF
that converts DBUG_ASSERT in non-debug builds into printouts.

For debug builds, it could be useful to be able to convert DBUG_ASSERT
into a warning or error printout, to allow execution to continue.
This would allow debug builds to be used for reproducing hard failures
that occur with release builds.

my_assert: A Boolean flag (set by default), tied to the new option
debug_assert that is available on debug builds only.
When set, DBUG_ASSERT() will invoke assert(), like it did until now.
When unset, DBUG_ASSERT() will invoke fprintf(stderr, ...)
with the file name, line number and assertion expression.

Changes of PAGE_MAX_TRX_ID must be redo-logged for correctness.
That was fixed in the InnoDB Plugin for MySQL 5.1 already.

Improved handling of subdirectories in the xtrabackup-v2 SST scripts (similar to MDEV-18863) for more predictable test results (related to xtrabackup-v2 SST)

MDEV-16932: ASAN heap-use-after-free in my_charlen_utf8 / my_well_formed_char_length_utf8 on 2nd execution of SP with ALTER trying to add bad CHECK

Make automatic name generation during execution (not prepare).

Check result of memory allocation operation.

Some bugs are detected only after a table definition has been evicted
and then reloaded to the InnoDB data dictionary cache.

For debug builds, introduce the settable Boolean configuration parameter
innodb_evict_tables_on_commit_debug that can be set to request InnoDB
to attempt to evict table definitions from the data dictionary cache
whenever a transaction is committed.

This has been tested on 10.3 and 10.4 with the following:

./mysql-test-run.pl --mysqld=--loose-innodb-evict-tables-on-commit-debug

You can also use the following:

SET GLOBAL innodb_evict_tables_on_commit_debug=ON;
SET GLOBAL innodb_evict_tables_on_commit_debug=OFF;

The parameter affects the commit (or rollback or abort) of
transactions that have modified persistent InnoDB tables.

Revert part of fa2a74e0.

trx_reference(): Remove, and merge the relevant part to the only caller
trx_rw_is_active(). If the statements trx = NULL; were ever executed,
the function would have dereferenced a NULL pointer and crashed in
trx_mutex_exit(trx). Hence, those statements must have been unreachable,
and they can be replaced with debug assertions.

trx_rw_is_active(): Avoid unnecessary acquisition and release of trx->mutex
when do_ref_count=false.

lock_trx_release_locks(): Do not reset trx->id=0. Had the statement been
necessary, we would have experienced crashes in trx_reference().

Cherry picking:
Bug#25135304: RBR: WRONG FIELD LENGTH IN ERROR MESSAGE
commit 47bd3f7cf3c8518f62b1580ec65af2ba7ac13b95

Description:
============
In row based replication, when replicating from a table with a field with
character set set to UTF8mb3 to the same table with the same field set to
character set UTF8mb4 I get a confusing error message:

For VARCHAR: VARCHAR(1) 'utf8mb3' to VARCHAR(1) 'utf8mb4'
"Column 0 of table 'test.t1' cannot be converted from type 'varchar(3)' to
type 'varchar(1)'"

Similar issue with CHAR type as well.

Issue with respect to BLOB types:

For BLOB: LONGBLOB to TINYBLOB - Error message displays incorrect blob type.
"Column 0 of table 'test.t1' cannot be converted from type 'tinyblob' to type
'tinyblob'"

For BINARY to BINARY - Error message displays incorrect type for master side
field.
"Column 0 of table 'test.t' cannot be converted from type 'char(1)' to type
'binary(10)'"
Similar issue exists for VARBINARY type. It is displayed as 'VARCHAR'.

Analysis:
=========
In Row based replication charset information is not sent as part of metadata
from master to slave.

For VARCHAR field its character length is converted into equivalent
octets/bytes and stored internally. At the time of displaying the data to user
it is converted back to original character length.

For example:
VARCHAR(2)- utf8mb3 is stored as:2*3 = VARCHAR(6)
At the time of displaying it to user
VARCHAR(6)- charset utf8mb3:6/3= VARCHAR(2).

At present the internally converted octect length is sent from master to slave
with out providing the charset information. On slave side if the type
conversion fails 'show_sql_type' function is used to get the type specific
information from metadata. Since there is no charset information is available
the filed type is displayed as VARCHAR(6).

This results in confused error message.

For CHAR fields
CHAR(1)- utf8mb3 - CHAR(3)
CHAR(1)- utf8mb4 - CHAR(4)

'show_sql_type' function which retrieves type information from metadata uses
(bytes/local charset length) to get actual character length. If slave's chaset
is 'utf8mb4' then

CHAR(3/4)-->CHAR(0)
CHAR(4/4)-->CHAR(1).

This results in confused error message.

Analysis for BLOB type issue:

BLOB's length is represented in two forms.
1. Actual length
i.e
  (length < 256) type= MYSQL_TYPE_TINY_BLOB;
  (length < 65536) type= MYSQL_TYPE_BLOB; ...

2. packlength - The number of bytes used to represent the length of the blob
  1- tinyblob
  2- blob ...

In row based replication only the packlength is written in the binary log. On
the slave side this packlength is interpreted as actual length of the blob.
Hence the length is always < 256 and the type is displayed as tiny blob.

Analysis for BINARY to BINARY type issue:
The character set information is needed to identify a filed's type as char or
binary. Since master side character set information is not available on the
slave side both binary and char fields are displayed as char.

Fix:
===
For CHAR and VARCHAR fields display their length in octets for both source and
target fields. For target field display the charset information if it is
relevant.

For blob type changed the code to use the packlength and display appropriate
blob type in error message.

For binary and varbinary fields use the slave side character set as reference
to map them to binary or varbinary fields.

Also fixes:

MDEV-20431 GREATEST(int_col,date_col) returns wrong results in a view

After applying MDEV-18863, in some test configurations, SST
may fails due to duplication of some parameters (in particular
"--port") in the main part of the command line and after
"--mysqld-args", as well as due to incorrect interpretation
of the parameter "--port" passed after "--mysqld-args" when
the SST script is invoked without explicitly specifying a port
for SST. In addition, it is necessary to correctly handle spaces,
quotation marks and special characters when copying original
arguments from the argv[] array to a new command line (after
"--mysqld-args"). This patch resolves these shortcomings.

Depending on version, "handle.exe -?" can output either "Handle v4.0"
or "Nthandle v4.1"

Analysis:
========
As part of BUG#28642318 fix, two new test cases were added. The first test
case tests a scenario where two sessions are present, in which the first
session has a regular table named 't1' and another session has a temporary
table named 't1'. Test executes a DELETE statement on regular table. These
statements are captured from binary log and replayed back on new client
connection to prove that DELETE statement is applied successfully. Note that
the binlog contains only CREATE TEMPORARY TABLE part hence a temporary table
gets created in new connection. This replaying logic is implemented by using
'--exec $MYSQL' command. If the new connection gets disconnected within the
scope of first test case the test passes, i.e the temporary table gets dropped
as part thread cleanup. But on slow platforms the connection gets closed at
the time of execution of test case 2. When the temporary table is dropped as
part thread cleanup a "DROP TEMPORARY TABLE t1" is written into the binary
log. In test case two the same sessions continue to exist and and table names
are reused to test a new bug scenario. The additional "DROP TEMPORARY TABLE"
command drops second test specific tables which results in "Unknown table"
error.

Fix:
====
Rename the second case specific table to 't2'. Even if the close connection
from test case one happens later the drop command with has
'DROP /*!40005 TEMPORARY */ TABLE IF EXISTS `t1`' will not result in an error.

Some code was duplicated near the start of the function,
only for InnoDB, not XtraDB. This was noticed by
comparing the InnoDB between MariaDB and MySQL.

For the Sphinx storage engine, this is a functional change (bug fix):
we will ensure that the message buffer is always NUL-terminated.

Current easy fix is not possible, because SELECT clones ha_partition
and then closes the clone which leads to unclosed transaction in
partitions we forcely prune out. We cound solve this by closing these
partitions (and release from transaction) in
change_partitions_to_open() at versioning conditions stage, but this
is problematic because table lock is acquired for each partition at
open stage and therefore must be released when we close partition
handler in change_partitions_to_open(). More details in MDEV-20376.

This should change after MDEV-20250 where mechanism of opening
partitions will be improved.

This reverts commit cdbac54d.

Fix debug build failing with error:

extended initializer lists only available with -std=c++11 or -std=gnu++11

ha_innobase::open(): Always ignore problems with FOREIGN KEY constraints
(pass DICT_ERR_IGNORE_FK_NOKEY), no matter whether foreign_key_checks
is enabled. Instead, we must report errors when enforcing the FOREIGN KEY
constraints. As a result of ignoring these errors, the tables will be
loaded with dict_foreign_t objects whose foreign_index or referenced_index
will be NULL.

Also, pass DICT_ERR_IGNORE_FK_NOKEY instead of DICT_ERR_IGNORE_NONE
to dict_table_open_on_id_low() in many other cases. Notably, on
CREATE TABLE and ALTER TABLE, we will keep validating the FOREIGN KEY
constraints as before.

dict_table_open_on_name(): If no other flags than
DICT_ERR_IGNORE_FK_NOKEY are set, refuse access to unreadable tables.
Some encryption tests rely on this code path.

For the DML code path, we used to have the problem that when
one of the indexes was missing in dict_foreign_t, we would ignore
the FOREIGN KEY constraint altogether. The following changes
address that.

row_ins_check_foreign_constraints(): Add the parameter pk.
For the primary key, consider also foreign key constraints for which
foreign->foreign_index=NULL (no underlying index is available).

row_ins_check_foreign_constraint(): Report errors also for !check_ref.
Remove a redundant check for srv_read_only_mode.

row_ins_foreign_report_add_err(): Tolerate foreign->foreign_index=NULL.

fkerr_t: Errors for the foreign key checks. Replaces ulint,
which used #define that looked like dberr_t literals.

wsrep_dict_foreign_find_index(): Remove. Use
dict_foreign_find_index() instead, with default parameters.

dict_foreign_push_index_error(): Do not add redundant quotes
around quoted table names.

Add wait conditions and compare cardinality etc information
between nodes and print something only if they differ.