Add thread_win32.h header

Workaround slow std::thread implementation in mingw
and gcc for Windows with our own old low level thread
functions.

No functional change.

src/misc.cpp

@@ -146,7 +146,7 @@ void dbg_print() {
 
 std::ostream& operator<<(std::ostream& os, SyncCout sc) {
 
-  static std::mutex m;
+  static Mutex m;
 
   if (sc == IO_LOCK)
       m.lock();