Parallel applier recovery, which happens after the sequence:
STOP SLAVE, START SLAVE; is based on master log name and position,
on Group Replication we have several masters what makes impossible
to recover parallel applier from that information.
Since we always have GTID_MODE=ON on Group Replication, we can
ignore the positions completely, seek the current relay log to the
beginning and start from there. Already applied transactions will be
skipped due to GTIDs auto skip feature and applier will resume from
the last applied transaction.

This patch fix the above issue and also introduce some more changes:
 1) Parallel applier recovery based on GTIDs instead of
    file+position.
    This change will only affect Group Replication applier.
    Files changed: rpl_slave.cc

 2) Sequential execution of transactions on logical clock parallel
    applier, by setting transaction logical timestamps to (0,0).
    This was already implemented on server, but since it was not a
    valid code path, there was a warning. Now we have a valid code
    path, so warning was removed.
    Files changed: log_event.cc, rpl_mts_submode.cc

 3) Adapt UNTIL_SQL_VIEW_ID to parallel applier.
    Parallel applier cannot be stopped on a ongoing transaction, so
    UNTIL_SQL_VIEW_ID condition, the stop condition for Group
    Replication distributed recovery, was adjusted to take that into
    consideration.
    Files changed: rpl_rli.h, rpl_rli.cc, rpl_slave.cc

 4) Disallow FLUSH RELAY LOGS FOR CHANNEL "group_replication_applier".
    To avoid that DBA splits transactions among relay logs, the
    command FLUSH RELAY LOGS was disabled for channel
    "group_replication_applier".
    Files changed: rpl_slave.cc, share/errmsg-utf8.txt

 5) Extend Trans_context_info structure with parallel applier
    options.
    Group Replication doesn't support database parallel applier, in
    order to prevent its use, it will check if it is in use and
    error out.
    Files changed: replication.h, rpl_group_replication.cc,
                   rpl_handler.cc

 6) Add channel_flush() function to rpl_channel_service_interface
    Provide a API function to flush relay logs from within a plugin,
    so that START GROUP_REPLICATION can rotate applier relay log.
    Files changed: rpl_channel_service_interface.h,
                   rpl_channel_service_interface.cc