Issue #4892: multiprocessing Connections can now be transferred over multiprocessing Connections.

Patch by Richard Oudkerk (sbt).

Lib/multiprocessing/__init__.py

@@ -161,7 +161,9 @@ def allow_connection_pickling():
     '''
     Install support for sending connections and sockets between processes
     '''
-    from multiprocessing import reduction
+    # This is undocumented.  In previous versions of multiprocessing
+    # its only effect was to make socket objects inheritable on Windows.
+    import multiprocessing.connection
 
 #
 # Definitions depending on native semaphores

Lib/multiprocessing/connection.py

@@ -50,6 +50,7 @@ import _multiprocessing
 from multiprocessing import current_process, AuthenticationError, BufferTooShort
 from multiprocessing.util import (
     get_temp_dir, Finalize, sub_debug, debug, _eintr_retry)
+from multiprocessing.forking import ForkingPickler
 try:
     import _winapi
     from _winapi import WAIT_OBJECT_0, WAIT_TIMEOUT, INFINITE
@@ -227,8 +228,9 @@ class _ConnectionBase:
         """Send a (picklable) object"""
         self._check_closed()
         self._check_writable()
-        buf = pickle.dumps(obj, protocol=pickle.HIGHEST_PROTOCOL)
-        self._send_bytes(memoryview(buf))
+        buf = io.BytesIO()
+        ForkingPickler(buf, pickle.HIGHEST_PROTOCOL).dump(obj)
+        self._send_bytes(buf.getbuffer())
 
     def recv_bytes(self, maxlength=None):
         """
@@ -880,3 +882,21 @@ else:
                     raise
             if timeout is not None:
                 timeout = deadline - time.time()
+
+#
+# Make connection and socket objects sharable if possible
+#
+
+if sys.platform == 'win32':
+    from . import reduction
+    ForkingPickler.register(socket.socket, reduction.reduce_socket)
+    ForkingPickler.register(Connection, reduction.reduce_connection)
+    ForkingPickler.register(PipeConnection, reduction.reduce_pipe_connection)
+else:
+    try:
+        from . import reduction
+    except ImportError:
+        pass
+    else:
+        ForkingPickler.register(socket.socket, reduction.reduce_socket)
+        ForkingPickler.register(Connection, reduction.reduce_connection)

Lib/multiprocessing/forking.py

@@ -407,25 +407,6 @@ else:
 
         return d
 
-    #
-    # Make (Pipe)Connection picklable
-    #
-
-    # Late import because of circular import
-    from .connection import Connection, PipeConnection
-
-    def reduce_connection(conn):
-        if not Popen.thread_is_spawning():
-            raise RuntimeError(
-                'By default %s objects can only be shared between processes\n'
-                'using inheritance' % type(conn).__name__
-                )
-        return type(conn), (Popen.duplicate_for_child(conn.fileno()),
-                            conn.readable, conn.writable)
-
-    ForkingPickler.register(Connection, reduce_connection)
-    ForkingPickler.register(PipeConnection, reduce_connection)
-
 #
 # Prepare current process
 #

Lib/multiprocessing/reduction.py

@@ -33,7 +33,7 @@
 # SUCH DAMAGE.
 #
 
-__all__ = []
+__all__ = ['reduce_socket', 'reduce_connection', 'send_handle', 'recv_handle']
 
 import os
 import sys
@@ -42,9 +42,8 @@ import threading
 import struct
 
 from multiprocessing import current_process
-from multiprocessing.forking import Popen, duplicate, close, ForkingPickler
 from multiprocessing.util import register_after_fork, debug, sub_debug
-from multiprocessing.connection import Client, Listener, Connection
+from multiprocessing.util import is_exiting, sub_warning
 
 
 #
@@ -60,22 +59,91 @@ if not(sys.platform == 'win32' or (hasattr(socket, 'CMSG_LEN') and
 #
 
 if sys.platform == 'win32':
+    # Windows
+    __all__ += ['reduce_pipe_connection']
     import _winapi
 
     def send_handle(conn, handle, destination_pid):
-        process_handle = _winapi.OpenProcess(
-            _winapi.PROCESS_ALL_ACCESS, False, destination_pid
-            )
-        try:
-            new_handle = duplicate(handle, process_handle)
-            conn.send(new_handle)
-        finally:
-            close(process_handle)
+        dh = DupHandle(handle, _winapi.DUPLICATE_SAME_ACCESS, destination_pid)
+        conn.send(dh)
 
     def recv_handle(conn):
-        return conn.recv()
+        return conn.recv().detach()
+
+    class DupHandle(object):
+        def __init__(self, handle, access, pid=None):
+            # duplicate handle for process with given pid
+            if pid is None:
+                pid = os.getpid()
+            proc = _winapi.OpenProcess(_winapi.PROCESS_DUP_HANDLE, False, pid)
+            try:
+                self._handle = _winapi.DuplicateHandle(
+                    _winapi.GetCurrentProcess(),
+                    handle, proc, access, False, 0)
+            finally:
+                _winapi.CloseHandle(proc)
+            self._access = access
+            self._pid = pid
+
+        def detach(self):
+            # retrieve handle from process which currently owns it
+            if self._pid == os.getpid():
+                return self._handle
+            proc = _winapi.OpenProcess(_winapi.PROCESS_DUP_HANDLE, False,
+                                       self._pid)
+            try:
+                return _winapi.DuplicateHandle(
+                    proc, self._handle, _winapi.GetCurrentProcess(),
+                    self._access, False, _winapi.DUPLICATE_CLOSE_SOURCE)
+            finally:
+                _winapi.CloseHandle(proc)
+
+    class DupSocket(object):
+        def __init__(self, sock):
+            new_sock = sock.dup()
+            def send(conn, pid):
+                share = new_sock.share(pid)
+                conn.send_bytes(share)
+            self._id = resource_sharer.register(send, new_sock.close)
+
+        def detach(self):
+            conn = resource_sharer.get_connection(self._id)
+            try:
+                share = conn.recv_bytes()
+                return socket.fromshare(share)
+            finally:
+                conn.close()
+
+    def reduce_socket(s):
+        return rebuild_socket, (DupSocket(s),)
+
+    def rebuild_socket(ds):
+        return ds.detach()
+
+    def reduce_connection(conn):
+        handle = conn.fileno()
+        with socket.fromfd(handle, socket.AF_INET, socket.SOCK_STREAM) as s:
+            ds = DupSocket(s)
+            return rebuild_connection, (ds, conn.readable, conn.writable)
+
+    def rebuild_connection(ds, readable, writable):
+        from .connection import Connection
+        sock = ds.detach()
+        return Connection(sock.detach(), readable, writable)
+
+    def reduce_pipe_connection(conn):
+        access = ((_winapi.FILE_GENERIC_READ if conn.readable else 0) |
+                  (_winapi.FILE_GENERIC_WRITE if conn.writable else 0))
+        dh = DupHandle(conn.fileno(), access)
+        return rebuild_pipe_connection, (dh, conn.readable, conn.writable)
+
+    def rebuild_pipe_connection(dh, readable, writable):
+        from .connection import PipeConnection
+        handle = dh.detach()
+        return PipeConnection(handle, readable, writable)
 
 else:
+    # Unix
     def send_handle(conn, handle, destination_pid):
         with socket.fromfd(conn.fileno(), socket.AF_UNIX, socket.SOCK_STREAM) as s:
             s.sendmsg([b'x'], [(socket.SOL_SOCKET, socket.SCM_RIGHTS,
@@ -94,136 +162,109 @@ else:
                 pass
             raise RuntimeError('Invalid data received')
 
+    class DupFd(object):
+        def __init__(self, fd):
+            new_fd = os.dup(fd)
+            def send(conn, pid):
+                send_handle(conn, new_fd, pid)
+            def close():
+                os.close(new_fd)
+            self._id = resource_sharer.register(send, close)
+
+        def detach(self):
+            conn = resource_sharer.get_connection(self._id)
+            try:
+                return recv_handle(conn)
+            finally:
+                conn.close()
+
+    def reduce_socket(s):
+        df = DupFd(s.fileno())
+        return rebuild_socket, (df, s.family, s.type, s.proto)
+
+    def rebuild_socket(df, family, type, proto):
+        fd = df.detach()
+        s = socket.fromfd(fd, family, type, proto)
+        os.close(fd)
+        return s
+
+    def reduce_connection(conn):
+        df = DupFd(conn.fileno())
+        return rebuild_connection, (df, conn.readable, conn.writable)
+
+    def rebuild_connection(df, readable, writable):
+        from .connection import Connection
+        fd = df.detach()
+        return Connection(fd, readable, writable)
 
 #
-# Support for a per-process server thread which caches pickled handles
+# Server which shares registered resources with clients
 #
 
-_cache = set()
+class ResourceSharer(object):
+    def __init__(self):
+        self._key = 0
+        self._cache = {}
+        self._old_locks = []
+        self._lock = threading.Lock()
+        self._listener = None
+        self._address = None
+        register_after_fork(self, ResourceSharer._afterfork)
 
-def _reset(obj):
-    global _lock, _listener, _cache
-    for h in _cache:
-        close(h)
-    _cache.clear()
-    _lock = threading.Lock()
-    _listener = None
+    def register(self, send, close):
+        with self._lock:
+            if self._address is None:
+                self._start()
+            self._key += 1
+            self._cache[self._key] = (send, close)
+            return (self._address, self._key)
 
-_reset(None)
-register_after_fork(_reset, _reset)
+    @staticmethod
+    def get_connection(ident):
+        from .connection import Client
+        address, key = ident
+        c = Client(address, authkey=current_process().authkey)
+        c.send((key, os.getpid()))
+        return c
 
-def _get_listener():
-    global _listener
+    def _afterfork(self):
+        for key, (send, close) in self._cache.items():
+            close()
+        self._cache.clear()
+        # If self._lock was locked at the time of the fork, it may be broken
+        # -- see issue 6721.  Replace it without letting it be gc'ed.
+        self._old_locks.append(self._lock)
+        self._lock = threading.Lock()
+        if self._listener is not None:
+            self._listener.close()
+        self._listener = None
+        self._address = None
 
-    if _listener is None:
-        _lock.acquire()
-        try:
-            if _listener is None:
-                debug('starting listener and thread for sending handles')
-                _listener = Listener(authkey=current_process().authkey)
-                t = threading.Thread(target=_serve)
-                t.daemon = True
-                t.start()
-        finally:
-            _lock.release()
+    def _start(self):
+        from .connection import Listener
+        assert self._listener is None
+        debug('starting listener and thread for sending handles')
+        self._listener = Listener(authkey=current_process().authkey)
+        self._address = self._listener.address
+        t = threading.Thread(target=self._serve)
+        t.daemon = True
+        t.start()
 
-    return _listener
+    def _serve(self):
+        while 1:
+            try:
+                conn = self._listener.accept()
+                key, destination_pid = conn.recv()
+                send, close = self._cache.pop(key)
+                send(conn, destination_pid)
+                close()
+                conn.close()
+            except:
+                if not is_exiting():
+                    import traceback
+                    sub_warning(
+                        'thread for sharing handles raised exception :\n' +
+                        '-'*79 + '\n' + traceback.format_exc() + '-'*79
+                        )
 
-def _serve():
-    from .util import is_exiting, sub_warning
-
-    while 1:
-        try:
-            conn = _listener.accept()
-            handle_wanted, destination_pid = conn.recv()
-            _cache.remove(handle_wanted)
-            send_handle(conn, handle_wanted, destination_pid)
-            close(handle_wanted)
-            conn.close()
-        except:
-            if not is_exiting():
-                import traceback
-                sub_warning(
-                    'thread for sharing handles raised exception :\n' +
-                    '-'*79 + '\n' + traceback.format_exc() + '-'*79
-                    )
-
-#
-# Functions to be used for pickling/unpickling objects with handles
-#
-
-def reduce_handle(handle):
-    if Popen.thread_is_spawning():
-        return (None, Popen.duplicate_for_child(handle), True)
-    dup_handle = duplicate(handle)
-    _cache.add(dup_handle)
-    sub_debug('reducing handle %d', handle)
-    return (_get_listener().address, dup_handle, False)
-
-def rebuild_handle(pickled_data):
-    address, handle, inherited = pickled_data
-    if inherited:
-        return handle
-    sub_debug('rebuilding handle %d', handle)
-    conn = Client(address, authkey=current_process().authkey)
-    conn.send((handle, os.getpid()))
-    new_handle = recv_handle(conn)
-    conn.close()
-    return new_handle
-
-#
-# Register `Connection` with `ForkingPickler`
-#
-
-def reduce_connection(conn):
-    rh = reduce_handle(conn.fileno())
-    return rebuild_connection, (rh, conn.readable, conn.writable)
-
-def rebuild_connection(reduced_handle, readable, writable):
-    handle = rebuild_handle(reduced_handle)
-    return Connection(
-        handle, readable=readable, writable=writable
-        )
-
-ForkingPickler.register(Connection, reduce_connection)
-
-#
-# Register `socket.socket` with `ForkingPickler`
-#
-
-def fromfd(fd, family, type_, proto=0):
-    s = socket.fromfd(fd, family, type_, proto)
-    if s.__class__ is not socket.socket:
-        s = socket.socket(_sock=s)
-    return s
-
-def reduce_socket(s):
-    reduced_handle = reduce_handle(s.fileno())
-    return rebuild_socket, (reduced_handle, s.family, s.type, s.proto)
-
-def rebuild_socket(reduced_handle, family, type_, proto):
-    fd = rebuild_handle(reduced_handle)
-    _sock = fromfd(fd, family, type_, proto)
-    close(fd)
-    return _sock
-
-ForkingPickler.register(socket.socket, reduce_socket)
-
-#
-# Register `_multiprocessing.PipeConnection` with `ForkingPickler`
-#
-
-if sys.platform == 'win32':
-    from multiprocessing.connection import PipeConnection
-
-    def reduce_pipe_connection(conn):
-        rh = reduce_handle(conn.fileno())
-        return rebuild_pipe_connection, (rh, conn.readable, conn.writable)
-
-    def rebuild_pipe_connection(reduced_handle, readable, writable):
-        handle = rebuild_handle(reduced_handle)
-        return PipeConnection(
-            handle, readable=readable, writable=writable
-            )
-
-    ForkingPickler.register(PipeConnection, reduce_pipe_connection)
+resource_sharer = ResourceSharer()