MDEV-4506: parallel replication.

Remove some unnecessary mutex locking.

sql/log.cc

@@ -6756,29 +6756,23 @@ MYSQL_BIN_LOG::queue_for_group_commit(group_commit_entry *orig_entry)
     DBUG_ASSERT(entry != NULL);
   }
 
-  /* Now we need to clear the wakeup_subsequent_commits_running flags. */
+  /*
+    Now we need to clear the wakeup_subsequent_commits_running flags.
+
+    We need a full memory barrier between walking the list above, and clearing
+    the flag wakeup_subsequent_commits_running below. This barrier is needed
+    to ensure that no other thread will start to modify the list pointers
+    before we are done traversing the list.
+
+    But wait_for_commit::wakeup(), which was called above for any other thread
+    that might modify the list in parallel, does a full memory barrier already
+    (it locks a mutex).
+  */
   if (list)
   {
     for (;;)
     {
-      if (list->wakeup_subsequent_commits_running)
-      {
-        /*
-          ToDo: We should not need a full lock/unlock of LOCK_wait_commit
-          here. All we need is a single (full) memory barrier, to ensure that
-          the reads of the list above are not reordered with the write of
-          wakeup_subsequent_commits_running, or with the writes to the list
-          from other threads that is allowed to happen after
-          wakeup_subsequent_commits_running has been set to false.
-
-          We do not currently have explicit memory barrier primitives in the
-          source tree, but if we get them the below mysql_mutex_lock() could
-          be replaced with a full memory barrier just before the loop.
-        */
-        mysql_mutex_lock(&list->LOCK_wait_commit);
-        list->wakeup_subsequent_commits_running= false;
-        mysql_mutex_unlock(&list->LOCK_wait_commit);
-      }
+      list->wakeup_subsequent_commits_running= false;
       if (list == last)
         break;
       list= list->next_subsequent_commit;

sql/sql_class.cc

@@ -5624,6 +5624,10 @@ wait_for_commit::wakeup()
     Otherwise we would need to somehow ensure that they were done
     waking up before we could allow this THD to be destroyed, which would
     be annoying and unnecessary.
+
+    Note that wakeup_subsequent_commits2() depends on this function being a
+    full memory barrier (it is, because it takes a mutex lock).
+
   */
   mysql_mutex_lock(&LOCK_wait_commit);
   waiting_for_commit= false;
@@ -5755,21 +5759,15 @@ wait_for_commit::wakeup_subsequent_commits2()
   }
 
   /*
-    ToDo: We should not need a full lock/unlock of LOCK_wait_commit here. All
-    we need is a (full) memory barrier, to ensure that the reads of the list
-    above are not reordered with the write of
-    wakeup_subsequent_commits_running, or with the writes to the list from
-    other threads that is allowed to happen after
-    wakeup_subsequent_commits_running has been set to false.
-
-    We do not currently have explicit memory barrier primitives in the source
-    tree, but if we get them the below mysql_mutex_lock() could be replaced
-    with a full memory barrier. It is probably not important, the lock is not
-    contented and will likely be in the CPU cache since we took it just before.
+    We need a full memory barrier between walking the list above, and clearing
+    the flag wakeup_subsequent_commits_running below. This barrier is needed
+    to ensure that no other thread will start to modify the list pointers
+    before we are done traversing the list.
+
+    But wait_for_commit::wakeup() does a full memory barrier already (it locks
+    a mutex), so no extra explicit barrier is needed here.
   */
-  mysql_mutex_lock(&LOCK_wait_commit);
   wakeup_subsequent_commits_running= false;
-  mysql_mutex_unlock(&LOCK_wait_commit);
 }