Problem: show slave status may return different Slave_IO_Running values running some tests.
Fix: wait for a certain slave state if needed to get tests more predictable.


mysql-test/r/rpl_log_pos.result:
  Fix for bug #28144: "Slave_IO_Running" differs in replication tests
    - test result adjusted.
mysql-test/r/rpl_ssl.result:
  Fix for bug #28144: "Slave_IO_Running" differs in replication tests
    - test result adjusted.
mysql-test/t/rpl_log_pos.test:
  Fix for bug #28144: "Slave_IO_Running" differs in replication tests
    - test has been rewritten to get certain slave state, now we use wait_for_slave_param.inc
      to ensure we get proper IO/SQL slave's threads states.
mysql-test/t/rpl_ssl.test:
  Fix for bug #28144: "Slave_IO_Running" differs in replication tests
    - wait_for_slave_to_start.inc used to get certain slave states in order 
      to be more predictable.
mysql-test/include/wait_for_slave_param.inc:
  New BitKeeper file ``mysql-test/include/wait_for_slave_param.inc''
  
  allow to wait until SHOW SLAVE STATUS has returned a spicified value.