fix sporadic failures of main.alter_table_online

progress updates aren't deterministic, occasionally they can be skipped (if `LOCK_thd_data` is locked). This isn't a problem as such, but breaks tests that wait for a specific value of the progress (and threads waiting for a specific value of the progress, exactly, lock `LOCK_thd_data` which prevents the progress from being updated) In this particular case, though, there is no need to wait for progress = 100, because "Waiting for table metadata lock" happens only at the end of stage = 4 anyway.

Author: vuvova

mysql-test/main/alter_table_online.test

@@ -266,7 +266,7 @@ send alter table t1 add b int NULL, algorithm= copy, lock= none;
 --connection con2
 evalp set @con= $con;
 
-let $wait_condition= select stage = 4 and progress = 100
+let $wait_condition= select stage = 4
  and state= "Waiting for table metadata lock"
  from information_schema.processlist where id = @con;
 --source include/wait_condition.inc