BUG#12574820: binlog.binlog_tmp_table timing out in daily and weekly trunk run

Problem: MYSQL_BIN_LOG::reset_logs acquires mutexes in wrong order.
The correct order is first LOCK_thread_count and then LOCK_log. This function
does it the other way around. This leads to deadlock when run in parallel
with a thread that takes the two locks in correct order. For example, a thread
that disconnects will take the locks in the correct order.
Fix: change order of the locks in MYSQL_BIN_LOG::reset_logs:
first LOCK_thread_count and then LOCK_log.


mysql-test/suite/binlog/r/binlog_reset_master.result:
  added result file
mysql-test/suite/binlog/t/binlog_reset_master.test:
  Added test case that demonstrates deadlock because of wrong mutex order.
  The deadlock is between two threads:
   - RESET MASTER acquires mutexes in wrong order.
   - client thread shutdown code acquires mutexes in right order.
  Actually, this test case does not produce deadlock in 5.1, probably
  the client thread shutdown code does not hold both mutexes at the same
  time. However, the bug existed in 5.1 (mutexes are taken in the wrong
  order) so we push the test case to 5.1 too, to prevent future
  regressions.
sql/log.cc:
  Change mutex acquisition to the correct order:
  first LOCK_thread_count, then LOCK_log.
sql/mysqld.cc:
  Add debug code to synchronize test case.

2 files changed, 14 insertions, 7 deletions

diff --git a/sql/log.cc b/sql/log.cc
index 17642696e7d..77d12641442 100644
--- a/sql/log.cc
+++ b/sql/log.cc
@@ -2989,12 +2989,6 @@ bool MYSQL_BIN_LOG::reset_logs(THD* thd)
   DBUG_ENTER("reset_logs");
 
   ha_reset_logs(thd);
-  /*
-    We need to get both locks to be sure that no one is trying to
-    write to the index log file.
-  */
-  pthread_mutex_lock(&LOCK_log);
-  pthread_mutex_lock(&LOCK_index);
 
   /*
     The following mutex is needed to ensure that no threads call
@@ -3002,7 +2996,14 @@ bool MYSQL_BIN_LOG::reset_logs(THD* thd)
     thread. If the transaction involved MyISAM tables, it should go
     into binlog even on rollback.
   */
-  VOID(pthread_mutex_lock(&LOCK_thread_count));
+  pthread_mutex_lock(&LOCK_thread_count);
+
+  /*
+    We need to get both locks to be sure that no one is trying to
+    write to the index log file.
+  */
+  pthread_mutex_lock(&LOCK_log);
+  pthread_mutex_lock(&LOCK_index);
 
   /* Save variables so that we can reopen the log */
   save_name=name;
diff --git a/sql/mysqld.cc b/sql/mysqld.cc
index 54850f36d10..2d1290bf88a 100644
--- a/sql/mysqld.cc
+++ b/sql/mysqld.cc
@@ -1903,6 +1903,12 @@ void unlink_thd(THD *thd)
   pthread_mutex_unlock(&LOCK_connection_count);
 
   (void) pthread_mutex_lock(&LOCK_thread_count);
+  /*
+    Used by binlog_reset_master.  It would be cleaner to use
+    DEBUG_SYNC here, but that's not possible because the THD's debug
+    sync feature has been shut down at this point.
+  */
+  DBUG_EXECUTE_IF("sleep_after_lock_thread_count_before_delete_thd", sleep(5););
   thread_count--;
   delete thd;
   DBUG_VOID_RETURN;