Issue #14753: Make multiprocessing treat negative timeouts as it did in 3.2

In Python 3.2 and earlier, Process.join() and Connection.poll()
treated negative timeouts as zero timeouts.  Earlier versions from
the 3.3 line of development treat them as infinite timeouts.

The patch reverts to the old behaviour.

Lib/multiprocessing/forking.py

@@ -75,12 +75,9 @@ else:
 #
 
 if sys.platform != 'win32':
-    import select
-
     exit = os._exit
     duplicate = os.dup
     close = os.close
-    _select = util._eintr_retry(select.select)
 
     #
     # We define a Popen class similar to the one from subprocess, but
@@ -130,10 +127,10 @@ if sys.platform != 'win32':
         def wait(self, timeout=None):
             if self.returncode is None:
                 if timeout is not None:
-                    r = _select([self.sentinel], [], [], timeout)[0]
-                    if not r:
+                    from .connection import wait
+                    if not wait([self.sentinel], timeout):
                         return None
-                # This shouldn't block if select() returned successfully.
+                # This shouldn't block if wait() returned successfully.
                 return self.poll(os.WNOHANG if timeout == 0.0 else 0)
             return self.returncode