bk commit into 5.0 tree (aelkin:1.2595) BUG#26622

[Andrei Elkin]

Below is the list of changes that have just been committed into a local
5.0 repository of aelkin.  When aelkin does a push these changes
will be propagated to the main repository and, within 24 hours after the
push, to the public repository.
For information on how to access the public repository
see http://dev.mysql.com/doc/mysql/en/installing-source-tree.html

ChangeSet@stripped, 2008-03-06 14:49:21+02:00, aelkin@mysql1000.(none) +3 -0
  Bug#26622  MASTER_POS_WAIT does not work as documented
  
  MASTER_POS_WAIT return values are different than expected when the server is not a
slave. 
  It returns -1 instead of NULL.
  
  Fixed with correcting  st_relay_log_info::wait_for_pos() to return the proper
  value in the case of rli info is not inited.

  mysql-test/r/rpl_master_pos_wait.result@stripped, 2008-03-06 14:49:18+02:00,
aelkin@mysql1000.(none) +6 -0
    results changed

  mysql-test/t/rpl_master_pos_wait.test@stripped, 2008-03-06 14:49:19+02:00,
aelkin@mysql1000.(none) +11 -0
    the new test to check that select master_pos_wait() to a server does 
    not have master info returns NULL as specified.

  sql/slave.cc@stripped, 2008-03-06 14:49:19+02:00, aelkin@mysql1000.(none) +1 -1
    changing the return value to correspond the specification.

diff -Nrup a/mysql-test/r/rpl_master_pos_wait.result
b/mysql-test/r/rpl_master_pos_wait.result
--- a/mysql-test/r/rpl_master_pos_wait.result	2006-10-03 17:02:39 +03:00
+++ b/mysql-test/r/rpl_master_pos_wait.result	2008-03-06 14:49:18 +02:00
@@ -16,3 +16,9 @@ select master_pos_wait('master-bin.99999
 stop slave sql_thread;
 master_pos_wait('master-bin.999999',0)
 NULL
+"*** must be empty ***"
+show slave status;
+"*** must be NULL ***"
+select master_pos_wait('foo', 98);
+master_pos_wait('foo', 98)
+NULL
diff -Nrup a/mysql-test/t/rpl_master_pos_wait.test b/mysql-test/t/rpl_master_pos_wait.test
--- a/mysql-test/t/rpl_master_pos_wait.test	2005-09-15 17:17:19 +03:00
+++ b/mysql-test/t/rpl_master_pos_wait.test	2008-03-06 14:49:19 +02:00
@@ -15,4 +15,15 @@ stop slave sql_thread;
 connection slave;
 reap;
 
+#
+# bug#26622 MASTER_POS_WAIT does not work as documented
+#
+
+connection master;
+echo "*** must be empty ***";
+query_vertical show slave status;
+
+echo "*** must be NULL ***";
+select master_pos_wait('foo', 98);
+
 # End of 4.1 tests
diff -Nrup a/sql/slave.cc b/sql/slave.cc
--- a/sql/slave.cc	2008-02-14 13:55:09 +02:00
+++ b/sql/slave.cc	2008-03-06 14:49:19 +02:00
@@ -2685,7 +2685,7 @@ int st_relay_log_info::wait_for_pos(THD*
                                     longlong timeout)
 {
   if (!inited)
-    return -1;
+    return -2;
   int event_count = 0;
   ulong init_abort_pos_wait;
   int error=0;