Galera MTR Tests: Use SET GLOBAL when setting wsrep_replicate_myisam, as it is a GLOBAL variable in MySQL Galera Cluster and SESSION in Percona XTraDB Cluster

Before this, FLUSH TABLES and FLUSH TABLES t1 (REFRESH_TABLES hereafter) were executed locally,
which also had implications for GTID. MySQL async replication does
replicate FLUSH TABLES.

Hence, this commit

a)  Adds REFRESH_TABLES to TOI.

b) Skips LOCK-based REFRESH_TABLES from TOI, ie. FLUSH TABLES WITH READ
LOCK and FLUSH TABLES FOR EXPORT.

c) Executes them after reload_acl_and_cache. This is done since
reload_acl_and_cache checks for thd->locked_tables_mode and does upgrade
of MDL locks accordingly. A failure here means REFRESH_TABLES cannot
run.

    i) This is also important since LOCK TABLES is not replicated in
    galera.

d) Test galera_flush has been modified for

    i) wsrep_replicate_myisam SESSION variable.

    ii) Testing REFRESH_TABLES.

    iii) Testing REFRESH_TABLES + LOCK TABLES.

e) Use different WSREP_TO_ISOLATION_BEGIN for global FLUSH TABLES and FLUSH TABLES <table,>.

  - errno handling in wsp::env::append() method, where error could be returned by mistake
  - return code of sst_prepare_other() when pthread_create() fails - it was returning positive error code which by convention is treated as success.

  1. Passes wsrep_sst_auth_value to SST scripts via WSREP_SST_OPT_AUTH envronmental variable, so it never appears on the command line
  2. In mysqldump and xtrabackup* SST scripts which rely on MySQL authentication, instead of passing password on the command line, SST script sets MYSQL_PWD environment variable, so that password also never appears on the mysqldump/innobackupex command line.

- fix for THD::m_digest initialization, according to Raghu

- changed BF thread's MDL wait to never timeout
- all explicit locks are now honored by BF threads

sessions to cover all exclusive MDL locks

Protecting non replicated FLUSH session from brute force aborts

Clear binlog table maps before writing SAVEPOINT query event into
binlog cache. This enforces recreation of table map events for the
following row event.

Galera MTR Tests: Attempt to remove rare sporadic failures in galera_transaction_replay.test by waiting for all transactions to get blocked before proceeding.

Galera MTR Tests: Enable the use of --parallel for port-intensive Galera tests by additionally specifying --port-group-size=50

  - errno handling in wsp::env::append() method, where error could be returned by mistake
  - return code of sst_prepare_other() when pthread_create() fails - it was returning positive error code which by convention is treated as success.

Synced xtrabackup SST fixes from Percona tree (as of PXC 5.6.24-25.11 release). This fixes/adresses the following LP bugs:
  - LP1380697: wsrep_sst_xtrabackup-v2 doesn't stop when mysql is SIGKILLed. (full fix for this (as engineeered by Percona) requires Linux-specific patch that we don't carry, but keep xtrabackup scripts as close as possible)
  - LP1399134: Log the innobackupex/SST logs in SST to syslog if possible. (fixed)
  - LP1405668: Race condition between donor and joiner in PXB SST. (fixed)
  - LP1405985: Fail early if xtrabackup_checkkpoints is missing. (fixed)
  - LP1407599: wsrep_sst_xtrabackup-v2 script causes innobackupex to print a false positive stack trace into the log. (fixed)
  - LP1441762: IST Fails with SST script error. (fixed)
  - LP1451670: Fail when move-back fails in xtrabackup SST. (fixed)

  1. Passes wsrep_sst_auth_value to SST scripts via WSREP_SST_OPT_AUTH envronmental variable, so it never appears on the command line
  2. In mysqldump and xtrabackup* SST scripts which rely on MySQL authentication, instead of passing password on the command line, SST script sets MYSQL_PWD environment variable, so that password also never appears on the mysqldump/innobackupex command line.

Do not replicate commands from check-warnings.test to other
nodes in the cluster.

When wsrep is enabled, for any update on innodb tables, the
corresponding keys are appended to galera's transaction writeset
(wsrep_append_keys()). However, for LOAD DATA, this got skipped
if binary logging was disabled or it was non-ROW based.
As a result, while the updates from LOAD DATA on non-partitioned
tables replicated fine as wsrep implicitly enables binary logging
(if not enabled, explicitly), the same did not work on partitioned
tables as for partitioned tables the binary logging gets disabled
temporarily (ha_partition::write_row()).

Fixed by removing the unwanted conditions from the check.
Also backported some changes from 10.0-galera to make sure
wsrep_load_data_splitting affects LOAD DATA commands only.