does not work in non-debug modes.

The bug manifested itself for recursive definitions that
used anchors over tables with blobs.

        mysql_prepare_create_table fixed so it doesn't let duplicating
        constraint names. Syntax for CONSTRAINT IF NOT EXISTS added
        and handled in mysql_alter_table.

Based on the patch from Daniil Medvedev (a Google Summer of Code task)

This bug in st_select_lex_node::move_node could result
in invalid select lists in recursive units that could
cause falling into infinite loops when iterating over
selects in such units.

The rdiff file needed to be adjusted after changes made in scope of MDEV-9864

Ability to print lock type added.
Restoring correct lock type for CREATE VIEW added.

Modifying innodb_ctype_ldml.test related files to take into account
the previous commit changes.

- Moving detection of the MY_CS_CSSORT, MY_CS_PUREASCII, MY_CS_NONASCII
  flags of loadable collations from add_collation() in mysys.c
  to my_cset_init_8bit() and my_coll_init_simple() in ctype-simple.c.

- Adding tests that these flags are set properly for loadable collations

- Moving LDML test related *.xml files from mysql-test/std_data/
  to mysql-test/std_data/ldml/, as there will be more *.xml test files

parallel replication

to know what goes wrong.
rpl_parallel_conflicts now contains the tests that
can cause row conflicts in replication.

Recent commits has added new variables and changed the default
for server-id, which caused the patch in rdiff to no longer apply.

Postreview fixes.
New MySQL tests fixes.

Transaction tracker

System variables tracking

initial commit to test

Markdown README for a prettier github representation

Added comments.
Added reaction for exceeding maximum number of elements in with clause.
Added a test case to check this reaction.
Added a test case where the specification of a recursive table
uses two non-recursive with tables.

- Sometimes privilege_table_io printed double rows of roles_mapping
  - Fixed by forcing restart of server when running test
- Wsrep_cluster_address_basic failed in some combinations because
  wsrep_cluster_address was set to NULL
  - Fixed by ensuring it's never set to NULL, only empty string

This makes it easier to setup master as on only have to set --log-bin.
Before this patch if one did set up the master with just --log-bin, slaves
could not connect until server_id was set on the master, which could be
both confusing and hard to do.

- When waiting for events, start time is now counted from start of wait
- Instead of having "Connect" as "Command" for all replication threads we
  now have:
  - Slave_IO for Slave thread reading relay log
  - Slave_SQL for slave executing SQL commands or distribution queries to
    Slave workers
  - Slave_worker for slave threads executin SQL commands in parallel replication

use the new approach with Field->default_value expressions.
But keep the old TIMESTAMP_UN_FIELD for ON UPDATE NOW().