[PATCH] sync_filesystems commentary and latency fix

- Add some commentary to this function

- Add a mutex to prevent new callers of sync_filesytems() from DoSing
  currently-running caller.

fs/super.c

@@ -303,11 +303,25 @@ void sync_supers(void)
 /*
  * Call the ->sync_fs super_op against all filesytems which are r/w and
  * which implement it.
+ *
+ * This operation is careful to avoid the livelock which could easily happen
+ * if two or more filesystems are being continuously dirtied.  s_need_sync_fs
+ * is used only here.  We set it against all filesystems and then clear it as
+ * we sync them.  So redirtied filesystems are skipped.
+ *
+ * But if process A is currently running sync_filesytems and then process B
+ * calls sync_filesystems as well, process B will set all the s_need_sync_fs
+ * flags again, which will cause process A to resync everything.  Fix that with
+ * a local mutex.
+ *
+ * FIXME: If wait==0, we only really need to call ->sync_fs if s_dirt is true.
  */
 void sync_filesystems(int wait)
 {
-	struct super_block * sb;
+	struct super_block *sb;
+	static DECLARE_MUTEX(mutex);
 
+	down(&mutex);		/* Could be down_interruptible */
 	spin_lock(&sb_lock);
 	for (sb = sb_entry(super_blocks.next); sb != sb_entry(&super_blocks);
 			sb = sb_entry(sb->s_list.next)) {
@@ -337,6 +351,7 @@ void sync_filesystems(int wait)
 		goto restart;
 	}
 	spin_unlock(&sb_lock);
+	up(&mutex);
 }
 
 /**